High-Dose Ascorbate in Combination with Anti-PD1 Checkpoint Inhibition as Treatment Option for Malignant Melanoma

Ascorbate acts as a prooxidant when administered parenterally at high supraphysiological doses, which results in the generation of hydrogen peroxide in dependence on oxygen. Most cancer cells are susceptible to the emerging reactive oxygen species (ROS). Accordingly, we evaluated high-dose ascorbate for the treatment of the B16F10 melanoma model. To investigate the effects of ascorbate on the B16F10 cell line in vitro, viability, cellular impedance, and ROS production were analyzed. In vivo, C57BL/6NCrl mice were subcutaneously injected into the right flank with B16F10 cells and tumor-bearing mice were treated intraperitoneally with ascorbate (3 g/kg bodyweight), immunotherapy (anti-programmed cell death protein 1 (PD1) antibody J43; 2 mg/kg bodyweight), or both treatments combined. The efficacy and toxicity were analyzed by measuring the respective tumor sizes and mouse weights accompanied by histological analysis of the protein levels of proliferating cell nuclear antigen (Pcna), glucose transporter 1 (Glut-1), and CD3. Treatment of B16F10 melanoma-carrying mice with high-dose ascorbate yielded plasma levels in the pharmacologically effective range, and ascorbate showed efficacy as a monotherapy and when combined with PD1 inhibition. Our data suggest the applicability of ascorbate as an additional therapeutic agent that can be safely combined with immunotherapy and has the potential to potentiate anti-PD1-based immune checkpoint blockades.


Introduction
The lifetime risk to develop melanoma is about 2.4% for the Caucasian population in the United States, and malignant melanoma causes most skin cancer-associated deaths due to early metastasis [1,2]. Its incidence is steadily increasing worldwide [2]. Until 2010, chemotherapeutics had largely failed to prolong overall survival (OS) in metastatic melanoma patients, and significant survival benefits were first achieved with the introduction of selective inhibitors of oncogenic B-Raf kinase (B-rapidly accelerated fibrosarcoma; BRAFi), which is constitutively active in about 50% of melanoma patients (vemurafenib The vitality of the B16F10 cell line was investigated with the 4-methylumbelliferyl heptanoate (MUH) assay (Sigma-Aldrich, Taufkirchen, Germany). Therefore, 3 × 10 4 cells/well were seeded into 24-well plates, and after 24 h, B16F10 cells were treated in duplicates for 24 h with ascorbate concentrations of 0-5 mM. Quantitative cell decline as positive control was induced by killing cells with 1% (v/v) Triton X100 (Roth, Karlsruhe, Germany). Prior to analysis, B16F10 cells were washed with PBS and then incubated for 1 h with 100 µg/mL MUH in phosphate-buffered saline (PBS; stock solution 10 mg/mL diluted 1:100) at 37 • C, 5% CO 2 . Hydrolysis of MUH by intracellular esterases and lipases of living cells results in the formation of the strongly fluorescent 4-methylumbelliferone (λ ex 355 nm, λ em 460 nm) and was detected with the Synergy H1 fluorescence microplate reader (BioTek, Bad Friedrichshall, Germany). Fluorescence intensity was proportional to the count of living cells after treatment, yielding the relative number of vital cells (% control).

Sulforhodamine B Cytotoxicity Assay
B16F10 cells (3 × 10 4 cells/well) were placed into 24-well plates (500 µL medium/well) and incubated for 24 h; 37 • C, 5% CO 2 . After medium change, treatment was performed in duplicates with 0-5 mM ascorbate for 24 h and growth inhibition was evaluated. Quantitative cell death as positive control was induced by adding 1% (v/v) Triton X100 (Roth) prior to fixing the cells. Then, medium was removed and the wells were washed with cold PBS and fixed with 10% (v/v) trichloroacetic acid (TCA) for 30 min at 4 • C. After washing the cells with tap water and drying them at 40 • C, protein staining was performed for 10 min with sulforhodamine B reagent (SRB; 0.4% (w/v) in 1% (v/v) acetic acid; CAS 3520-42-1, Sigma-Aldrich). Following washing with tap water and 1% (v/v) acetic acid to remove unbound dye, wells were dried at 40 • C. The protein-bound dye was dissolved with 10 mM Tris base (pH 10.5) for 10 min and absorbance was measured after transferring triplicates of 80 µL/well into a 96-well plate with the SynergyH1 multiplate reader (BioTek; wavelength 550 nm, reference wavelength 620 nm). Data represent the mean of optical density values (% control) as a surrogate parameter for relative cell density.

Real-Time Proliferation Assay
The xCELLigence ® SP system (Agilent Technologies Inc., Waldbronn, Germany) was used to evaluate the cellular impedance as a surrogate parameter for cell proliferation over time. B16F10 cells (4 × 10 3 cells/well) were seeded in 100 µL medium into a 96-well plate equipped with gold microelectrodes (E-Plate 96, OMNI Life Science GmbH & Co. KG, Bremen, Germany). Prior to this, 90 µL/well of medium had to be provided to measure the zero value per well. Ascorbate treatment was performed in quadruplicates after 24 h in 20 µL/well to reach final concentrations of 0-5 mM in a final volume of 210 µL/well. Quantitative cell decline as positive control was induced by adding 1% (v/v) Triton X100 (Roth). Measurements of electrical impedance were recorded in 30 min intervals over a period of 72 h. Each experiment was conducted three times. Finally, normalized cell

DCFH-DA-Assay for ROS Detection
B16F10 cells were seeded at a density of 3 × 10 4 cells per well in 24-well plates in duplicates and then incubated overnight at 37 • C. After 24 h, cells were treated with 0-5 mM ascorbate for different time periods. Three h before the end of treatment, 50 µL of medium were removed from the positive control wells and replaced with 50 µL 10 mM tert-butyl hydroperoxide (TBH; Sigma-Aldrich) solution to achieve a final concentration of 1 mM. Two, three, and four h after the start of treatments, medium was removed, washed with 500 µL of PBS, and cells were harvested with 150 µL trypsin at 37 • C for 5 min. For each well, 1 mL of medium was added to stop the trypsin reaction, cells were centrifuged (149× g, 5 min), and the supernatant was discarded. A working solution with 5 µM in phenol red-free, FCS-free medium was prepared from the 40 mM 2 ,7 -dichlorofluorescein diacetate (DCFH-DA; Sigma-Aldrich) stock solution, and the cell pellets were resuspended in 1 mL of this solution. After intracellular esterases increase the polarity of the molecule by deacetylation and prevent it from leaving the cell, the green fluorescent dichlorofluorescein (DCF) is formed by oxidation by various ROS [39]. After incubation at 37 • C for 30 min in the dark, the cells were centrifuged, the supernatant was discarded, and the cells were covered once again with PBS. Subsequently, pellets were resuspended in 400 µL of phenol red-free, FCS-free medium and were transferred to a flow cytometry tube. Flow cytometric analysis followed with absorbance at 488 nm and detection of emission at 530 nm (NovoCyte ® 2060R flow cytometer, Agilent Technologies Inc.). For this purpose, 10,000 events per sample were measured and the experiment was performed three times, independently.

In Vivo Mouse Experiment
Six-week-old female C57BL/6N Crl mice were purchased from Charles River Laboratory (Wiga, Germany) and accommodated in the animal care institution at the University of Tuebingen. After one week, C57BL/6N Crl females received a subcutaneous (s.c.) injection of 1 × 10 5 B16F10 cells suspended in 50 µL sterile PBS (Sigma-Aldrich) into the right flank. Once tumors were palpable, the mice were randomly divided into four groups (eight mice as controls and six mice for each treatment) and respective treatments were started. A total of 42 mice were injected, but only mice that had uniform tumor size at treatment initiation were used for randomization. All treatments were applied intraperitoneally (i.p.). The anti-PD1 antibody J43 was injected at 2 mg/kg bodyweight in 100 µL isotonic saline solution, twice weekly. Ascorbate (Pascorbin ® ) was injected daily at 3 g/kg bodyweight. Mixed syringes were used for combined treatments (up to 700 µL volume for i.p. injections; 23G needle) to minimize pain in mice by reducing the number of i.p. injections. The same volume of solvent (NaCl (Sigma-Aldrich)) was used for up to 20 i.p. injections for the control group. Animals were weighed every other day through the experiment and tumor size was evaluated daily (volume = length × width × width × 0.5). To directly compare tumor volumes and weights between the treatment groups, all animals in the four treatment groups were sacrificed by using CO 2 on the same day that sham-treated mice reached the abortion criterion. Tissue samples (organs, blood, and tumors) were collected and stored for further analysis.

Immunohistochemistry of Murine Tumors and Organs
Murine tumors were initially fixed in formalin (4%), embedded in paraffin, serially sectioned, and finally stained with hematoxylin and eosin (H&E). Proteins of interest were stained using antibodies against CD3 (Dianova, Hamburg, Germany; DIA-303), PCNA (Abcam; #ab92552), hypoxia-inducible factor 1α (HIF-1α; Biozol, Eching, Germany; #IHC-IW-PA 1041), and GLUT-1 (Abcam, Berlin, Germany; #ab115730). Subsequently, bound antibodies were stained with the ultraView Universal Alkaline Phosphatase Red Detection Kit from Ventana (Roche, Rotkreuz; Switzerland). Quantification of immunohistochemical staining was performed as follows. Five representative samples were stained per group (only four tumors could be evaluated for Glut-1 staining of the combination group). For the proliferation marker Pcna, the percentage of tumor cells with a positively stained nucleus was calculated using the formula [% positive cells] = (100 * [number of positive tumor nuclei]/[total nuclei]), and the average value of three high magnifications per sample was calculated. For Glut-1, only the membranous signal was evaluated. Similar to the HercepTest TM , a score of 0-3 was defined for every sample, where 0 means that no staining or membrane staining is detected in <10% of B16F10 cells; 1 means that weak/weak perceptible membrane staining is detected in >10% of B16F10 cells; 2 means that weakto-moderate complete membrane staining is detected in >10% of B16F10 cells; 3 means that strong complete membrane staining is detected in >10% of B16F10 cells. For CD3, the absolute number of CD3-positive cells in a representative non-necrotic tumor area of 1 mm 2 per sample was determined.

Ethics
Investigations were carried out in accordance with the Declaration of Helsinki [40]. The mice experiments were conducted in agreement with the European Union and German law and were approved by local authorities (Regierungspraesidium Tuebingen, HT3/16).

Growth-Inhibiting Effects of Ascorbate on B16F10 Cells
To evaluate the direct effects of ascorbate, different concentrations of ascorbate were used to treat B16F10 cells (0-5 mM). Cell viability was determined using the fluorometric MUH assay after 24 h treatment with Triton-X100 as a positive control for complete cell death. Figure 1A shows that ascorbate concentrations of ≥2 mM reduced cell viability significantly, whereas ≥ 3 mM of ascorbate decreased the viability of B16F10 cells below a threshold of 50%, suggesting relevant antitumoral effects. These results were confirmed by the SRB assay, which determines the total protein content of the treated B16F10 cells and is proportionate to the number of living tumor cells ( Figure 1B). Interestingly, the ascorbate effect was even more pronounced, decreasing the cell number below 50% with ≥2 mM of ascorbate.
Based on the endpoint assays above, pharmacological ascorbate has a potent antitumor effect when appearing in the milieu of the melanoma cells. Although millimolar plasma levels of ascorbate are rapidly reached in the patient after infusion of ascorbate, they are also renally excreted within a short period of time. Accordingly, it is important to determine how long ascorbate must be in contact with B16F10 cells for to produce relevant antitumor effects. To determine the exact temporal onset and the extent of ascorbate effects in a real-time setting, B16F10 cells were treated with concentrations 0-5 mM and measured in real-time with the xCELLigence ® SP system (Agilent Technologies Inc.). Cell proliferation was tracked by measuring electrical impedance every 30 min over 72 h (Figure 2A). Detailed analysis revealed a fast antiproliferative effect of ascorbate treatment starting from 2 mM after 4 h of treatment ( Figure 2B). Based on the endpoint assays above, pharmacological ascorbate has a potent antitumor effect when appearing in the milieu of the melanoma cells. Although millimolar plasma levels of ascorbate are rapidly reached in the patient after infusion of ascorbate, they are also renally excreted within a short period of time. Accordingly, it is important to determine how long ascorbate must be in contact with B16F10 cells for to produce relevant antitumor effects. To determine the exact temporal onset and the extent of ascorbate effects in a real-time setting, B16F10 cells were treated with concentrations 0-5 mM and measured in real-time with the xCELLigence ® SP system (Agilent Technologies Inc.). Cell proliferation was tracked by measuring electrical impedance every 30 min over 72 h (Figure 2A). Detailed analysis revealed a fast antiproliferative effect of ascorbate treatment starting from 2 mM after 4 h of treatment ( Figure 2B).

Figure 1.
Ascorbate-induced growth inhibition on murine melanoma cells. (A) B16F10 mouse melanoma cells were exposed to varying amounts of ascorbate (0-5 mM). The viability of murine melanoma cells was evaluated using the MUH assay in comparison to the medium-treated control. Depicted are the mean values ± SD of five standalone experiments, measured in duplicate, respectively, and standardized to the medium-treated control. (B) B16F10 mouse melanoma cells were exposed to varying amounts of ascorbate (0-5 mM). The cell viability was assessed using the SRB assay in comparison to the medium-treated control. Depicted are the mean values ± SD of five standalone experiments, each measured in duplicate and standardized to the medium-treated control. ***: p ≤ 0.001. MUH, 4-methylumbelliferyl heptanoate; SD, standard deviation; SRB, sulforhodamine B. Figure 1. Ascorbate-induced growth inhibition on murine melanoma cells. (A) B16F10 mouse melanoma cells were exposed to varying amounts of ascorbate (0-5 mM). The viability of murine melanoma cells was evaluated using the MUH assay in comparison to the medium-treated control. Depicted are the mean values ± SD of five standalone experiments, measured in duplicate, respectively, and standardized to the medium-treated control. (B) B16F10 mouse melanoma cells were exposed to varying amounts of ascorbate (0-5 mM). The cell viability was assessed using the SRB assay in comparison to the medium-treated control. Depicted are the mean values ± SD of five standalone experiments, each measured in duplicate and standardized to the medium-treated control. ***: p ≤ 0.001. MUH, 4-methylumbelliferyl heptanoate; SD, standard deviation; SRB, sulforhodamine B.
Based on the endpoint assays above, pharmacological ascorbate has a potent antitumor effect when appearing in the milieu of the melanoma cells. Although millimolar plasma levels of ascorbate are rapidly reached in the patient after infusion of ascorbate, they are also renally excreted within a short period of time. Accordingly, it is important to determine how long ascorbate must be in contact with B16F10 cells for to produce relevant antitumor effects. To determine the exact temporal onset and the extent of ascorbate effects in a real-time setting, B16F10 cells were treated with concentrations 0-5 mM and measured in real-time with the xCELLigence ® SP system (Agilent Technologies Inc.). Cell proliferation was tracked by measuring electrical impedance every 30 min over 72 h (Figure 2A). Detailed analysis revealed a fast antiproliferative effect of ascorbate treatment starting from 2 mM after 4 h of treatment ( Figure 2B).

Effects of Ascorbate on ROS Production of Melanoma Cells
After having established that ascorbate in a concentration of ≥2 mM could quickly achieve antitumoral effects, we sought to determine the timing and extent of ascorbateinduced ROS formation and whether these actually correlate with cell death. Therefore, a DCFH-DA assay for ROS detection was performed (Figure 3). After treatment of B16F10 melanoma cells with ascorbate (0-5 mM), we observed that significant ROS production occurred with ascorbate levels ≥ 3 mM already after 2 h, with a peak at 3 h. These results were largely in line with the viability assays, which showed significant antitumoral effects with ≥ 2 mM ascorbate.
was monitored by impedance measurement at 30 min intervals and compared to the mediumtreated control. (A) The viability is shown in 4 h intervals over the complete period of 72 h. (B) The viability is shown in 1 h intervals in the first 6 h after treatment start. Depicted are the mean values ± SD of three stand-alone experiments, measured in quadruplicate, respectively (n = 11 for controls, n = 12 for treated cells). SD, standard deviation.

Effects of Ascorbate on ROS Production of Melanoma Cells
After having established that ascorbate in a concentration of ≥2 mM could quickly achieve antitumoral effects, we sought to determine the timing and extent of ascorbateinduced ROS formation and whether these actually correlate with cell death. Therefore, a DCFH-DA assay for ROS detection was performed (Figure 3). After treatment of B16F10 melanoma cells with ascorbate (0-5 mM), we observed that significant ROS production occurred with ascorbate levels ≥ 3 mM already after 2 h, with a peak at 3 h. These results were largely in line with the viability assays, which showed significant antitumoral effects with ≥ 2 mM ascorbate.

Effects of Ascorbate and the Anti-PD1 Antibody J43 In Vivo
The evaluation of the effectiveness of treatment was carried out by s.c. injecting the syngeneic murine melanoma cell line B16F10 in the right flank of female C57BL/6N Crl mice that were treated with ascorbate, anti-PD1 immunotherapy, or the combination of both in the following. This represents a fully immunocompetent murine melanoma model. Once the tumors had developed, female C57BL/6N Crl mice were divided into four randomized groups (eight in the control group and six for each treatment), and treatment was started. Mice were treated via i.p. injections. Anti-PD1 antibody J43 was injected at 2 mg/kg body-weight, twice per week. Ascorbate was administered daily as injection at 3 g/kg bodyweight. In the combination treatment, both compounds were mixed immediately before application. The control group received the corresponding volume of isotonic saline as treatment. The treatment scheme is illustrated in Figure 4A. The weight of the mice was assessed every other day during the entire experiment. To determine the tumor volume, the length and width of each tumor were measured daily (volume = length × width × width × 0.5) ( Figure 4B). To enable intertreatment group comparison, the mice of

Effects of Ascorbate and the Anti-PD1 Antibody J43 In Vivo
The evaluation of the effectiveness of treatment was carried out by s.c. injecting the syngeneic murine melanoma cell line B16F10 in the right flank of female C57BL/6N Crl mice that were treated with ascorbate, anti-PD1 immunotherapy, or the combination of both in the following. This represents a fully immunocompetent murine melanoma model. Once the tumors had developed, female C57BL/6N Crl mice were divided into four randomized groups (eight in the control group and six for each treatment), and treatment was started. Mice were treated via i.p. injections. Anti-PD1 antibody J43 was injected at 2 mg/kg body-weight, twice per week. Ascorbate was administered daily as injection at 3 g/kg bodyweight. In the combination treatment, both compounds were mixed immediately before application. The control group received the corresponding volume of isotonic saline as treatment. The treatment scheme is illustrated in Figure 4A. The weight of the mice was assessed every other day during the entire experiment. To determine the tumor volume, the length and width of each tumor were measured daily (volume = length × width × width × 0.5) ( Figure 4B). To enable intertreatment group comparison, the mice of all four cohorts were sacrificed on day 11, 1 h after therapy initiation, with CO 2 , as soon as the animals in the control group reached the abortion criteria due to tumor size.
All tumors of treated mice (either with ascorbate, anti-PD1, or the combination of ascorbate and anti-PD1) showed decreased tumor volumes compared to control. Tumors in the groups receiving either ascorbate or the combination treatment displayed a significant decrease in tumor volumes compared to control tumors at the end of the experiment ( Figure 4C). To represent tumor growth over time, the calculated tumor volumes over time are relative to day 0 (therapy initiation) with 100% tumor volume ( Figure 4B). Anti-PD1 treatment with J43 (grey) decreased the tumor growth compared to the sham-treated group (black). Similarly, high-dose ascorbate (orange) significantly reduced tumor growth. The combined treatment with ascorbate and anti-PD1 (green) led to less tumor growth in comparison to the control group and the groups treated with ascorbate or anti-PD1 alone. all four cohorts were sacrificed on day 11, 1 h after therapy initiation, with CO2, as soon as the animals in the control group reached the abortion criteria due to tumor size. All tumors of treated mice (either with ascorbate, anti-PD1, or the combination of ascorbate and anti-PD1) showed decreased tumor volumes compared to control. Tumors in the groups receiving either ascorbate or the combination treatment displayed a significant decrease in tumor volumes compared to control tumors at the end of the experiment ( Figure 4C). To represent tumor growth over time, the calculated tumor volumes over time are relative to day 0 (therapy initiation) with 100% tumor volume ( Figure 4B). Anti-PD1 treatment with J43 (grey) decreased the tumor growth compared to the sham-treated group (black). Similarly, high-dose ascorbate (orange) significantly reduced tumor growth. The combined treatment with ascorbate and anti-PD1 (green) led to less tumor growth in comparison to the control group and the groups treated with ascorbate or anti-PD1 alone.
To evaluate the state of health of the mice at the end of the experiment, various organs (lung, liver, intestine, kidney, and spleen) were examined histologically ( Figure 5A). No To evaluate the state of health of the mice at the end of the experiment, various organs (lung, liver, intestine, kidney, and spleen) were examined histologically ( Figure 5A). No changes in tissue microstructure or other cellular alterations were detected in the organs of the respective treatment groups. Moreover, animal weight revealed no significant differences between the groups throughout the experiment ( Figure 5B). The change in mouse weight of the different treatment groups was normalized to the day of s.c. tumor cell injection and is shown in Figure 5C. All treatment groups gained weight, and none of the groups differed significantly from the others in this respect throughout the experiment. In addition, tumors of all treatment groups were stained with H&E ( Figure 5D), as well as for Pcna to assess proliferation status, glucose / dehydroascorbate transporter Glut-1 expression, and CD3 to determine the amount of T-cell infiltration ( Figure 5E). Pcna expression levels were higher in control tumors than in tumors of the three different treatment groups ( Figure 5E), with the quantification shown in Figure 5F. The results indicate the highest proliferation of the tumor cells in the control group and the lowest in the combination group. Glut-1 tended to be reduced in all treatment groups, including combination therapy, compared to the control group ( Figure 5E,G). CD3-positive cells were enhanced by trend in the areas of tumors treated with ascorbate or anti-PD1, whereas the combination resulted in a significant rise in these cells in line with increased immune cell infiltration into the tumor ( Figure 5E,H).

Discussion
In this study, we demonstrate that highly dosed ascorbate exhibits cytotoxic effects on B16F10 murine melanoma cells both in vitro and in an immunocompetent mouse melanoma model (C57BL/6N Crl ). SRB assay results and impedance analysis (xCELLigence ® SP system) showed that ascorbate reduced the viability of cancerogenic B16F10 cells with high efficacy (≥2 mM). Using MUH assay, a peak effect was observed at a concentration ≥ 3 mM. These results were confirmed by our ROS analysis and are in line with the described ROS induction in the literature [23,41,42]. In recent studies, it was demonstrated that antitumoral ascorbate levels are only achieved by parenteral-and preferably intravenousadministration, and that ascorbate acts as a prodrug for the ascorbyl radical (Asc •− ) and H 2 O 2 in the blood and interstitium, respectively [43,44]. This is in contrast to the classification of ascorbate as a general antioxidative radical scavenger [23,24,42]. Nanoparticles have also been developed to optimize pharmacokinetics and tumor targeting to reduce the required high amount of ascorbate. Various formulations exist for these ascorbate nanodeliverers. It has been demonstrated that ascorbate administered in this manner can effectively induce cytotoxicity in cancerous cells in vitro and suppress tumor growth in mice, e.g., ascorbyl palmitate together with paclitaxel in a murine B16F10 model [45,46].
In recent years, various studies demonstrated an effect of ascorbate on varying melanoma cell lines, as well as cell lines from other tumor entities [41]. In A375.S2 melanoma cells, for example, ascorbate caused cell cycle arrest and apoptosis [47]. In B16F10 mouse melanoma cells, high-dose ascorbate led to the induction of caspase-8independent apoptosis via decrease in transferrin receptor-dependent cellular iron uptake [48,49]. Proliferation of B16F10 melanoma cells was inhibited by ascorbate-dependent regulation of the p53-p21 pathway [50].
Remarkably, plasma levels of up to 20 mM that are well tolerated by healthy cells in vitro [42], can be achieved safely in human tumor patients by the application of intravenous ascorbate [22,51,52]. Since a subset of patients with melanoma strongly benefits from the use of immunotherapies it is important to find suitable add-on therapy options that either improve their efficacy or mitigate the sometimes therapy-limiting side effects. In addition, there is an continuous need for new treatment strategies in case of primary or secondary resistance to checkpoint inhibition, which is observed in many melanoma patients [16].
However, the performed in vitro assays were not appropriate to investigate the effects of high-dose ascorbate therapy on the immune system and thus not suitable to investigate possible additive, synergistic, or even antagonistic effects caused by the combination of ascorbate and PD1 blockade, which was the main aim of this study. To overcome such shortfall, we used immunocompetent C57BL/6N Crl mice with a syngeneic background to the B16F10 cells studied in vitro. The B16F10 cells were administered to the mice by s.c. injection and the effect of both the respective monotherapy (PD1 antibody J43) and ascorbate (3 g/kg bodyweight) was compared with the combination of both. Repeated i.p. doses of 3-4 g/kg bodyweight were also used in other murine tumor models and additionally in a murine colitis model [37,[53][54][55]. The role of ascorbate as a prodrug for H 2 O 2 and the generation of the ascorbyl radical is well established in the literature, and future studies should quantify this whenever possible [42]. An additive effect of the combination therapy could at least be demonstrated, and at the same time, no significant additional toxicities appeared for the combination compared to the monotherapies, respectively. The suppression of human A375 melanoma cells by ascorbate treatment and caspase-3 and caspase-9 activation, as well as induction of the Bcl-2-and Bax-mediated mitochondrial pathway, was already described by Chen and coworkers [56]. Evidence that the combination of high-dose ascorbate with immune checkpoint inhibitors could be beneficial was provided by Luchtel and colleagues. They describe good effects of high-dose ascorbate in combination with anti-PD1 therapy in a lymphoma mouse model [35], whereas Magri et al. demonstrated that the tumor volume in colon, pancreatic, and breast cancers was significantly reduced by the combined administration of ascorbate with anti-PD1 or anti-CTLA4 in mice [36]. Recently, a study by Peng et al. described the combination of pharmacological ascorbate with PD-L1 blockade in a murine renal carcinoma model. The combination showed increased intratumoral CD4 + and CD8 + infiltration and elevated IFN-γ levels, as well as downstream signaling [57]. To the best of our knowledge, we have shown for the first time that the combination of ascorbate and PD1 inhibition with J43 is both highly effective and well tolerated in an immunocompetent melanoma mouse model resembling BRAF wild-type melanomas.
However, one limitation of the study was the short treatment time window until the discontinuation criterion was reached in the control animals, as subcutaneous B16F10 tumors grow rapidly once tumors are established [58]. Since all animals in this study were sacrificed at the same time point, some of the tumors in the therapy groups remained very small. Larger-sized preclinical studies with longer medication periods need to be performed to investigate these promising beneficial effects of high-dose ascorbate on melanoma growth and antitumoral immune response. It also remains open if the mechanistical aspects underlying the increased antitumor effects of the combination therapy are completely different and independent from each other or if high-dose ascorbate can enable an improved immune response via the immune checkpoint inhibitors, not only because of the prooxidative mechanisms, but also epigenetic modulation, ten eleven translocation 2 (TET2) enzyme activation (tet methylcytosine dioxygenase 2), and regulation of cytokine expression [57,59]. In a review by Bedhiafi et al., the role of high-dose ascorbate and its ability to modify the immune response against cancer is nicely summarized, and the authors describe how it may help to overcome the resistance mechanisms against immune checkpoint blockades [59].
An increased initial tumor cell death caused by high doses of ascorbate accompanied by increased antigen release could be also an additional mechanism. It seems plausible that dying tumor cells release more tumor-specific antigens, which in turn could further facilitate the immune response of the host under immunotherapy and fuel the cancer immunity cycle [60]. However, this hypothesis has to be proven more extensively in mice (preclinical immunocompetent melanoma models) and humans (clinical trials). Although no signs of toxicity in the C57BL/6N Crl mice were observed, long-term treatment must be carefully monitored, and clinical studies are required to evaluate this promising combination and optimize the results of anti-PD1-based immunotherapy against malignant melanoma.

Conclusions
Our data support the benefits of this dual treatment of immunotherapy with highly dosed ascorbate to target melanoma cells. Nevertheless, clinical trials could further evaluate the effects of ROS-inducing and redox-active agents such as pharmacological high-dose ascorbate in the context of malignant melanoma, as well as the role of a stimulated immune system through immune checkpoint inhibition.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.