Introduction

Hepatocellular carcinoma (HCC) is the sixth most common cancer worldwide.1 With the progress of drug development, immune checkpoint inhibitor (ICI)-based combination therapy is the recommended first-line therapy for unresectable HCC.2–6 The best tumor response is around one-third in the patients with HCC under ICI-based therapy. Because the current drug therapy for HCC is still far from ideal, the critical demand for developing more potent therapies is needed.7 Enhanced immunogenic cell death (ICD) is one of the strategies to increase antitumor immune response.

ICD is a form of cell death that triggers adaptive immune responses and immunological memory through the simultaneous delivery of sufficient antigenicity and adjuvanticity for effective stimulation.8 9 The induction of ICD can be assessed by the appearance of danger-associated molecular patterns (DAMPs), such as the increased presence of the endoplasmic reticulum (ER)-resident chaperone calreticulin on the cell membrane and the enhanced extracellular release of high-mobility group box 1 (HMGB1) and ATP. Additionally, novel ICD-induced DAMPs such as Hsp70, Hsp90, interferons, and annexin A are continually being identified.9 10 Currently, the mouse vaccination assay is recognized as the gold standard for verifying ICD by demonstrating suppressed tumor growth in rechallenged tumors using a vaccine derived from ICD-induced cells.8

The molecular mechanisms of ICD remain largely unknown, although the induction of ER stress is recognized as a prerequisite.11 12 Typically, ER stress, marked by the accumulation of misfolded and unfolded proteins in the ER lumen, triggers the unfolded protein response (UPR) to restore ER function. This response activates at least three key signaling pathways—PERK, IRE1α, and ATF6. To date, only the phosphorylation of the translation elongation factor eIF2α is relevant as a pathognomonic and quintessential hallmark for ICD.13 14 Further, in addition to apoptosis, cell-death pathways such as necroptosis, pyroptosis, and ferroptosis also mediate ICD,15 16 while the inhibition of autophagy favors ICD progression.16

Doxorubicin was identified as the first ICD inducer in 2005.15 Subsequently, various cancer treatments, including other chemotherapeutic agents, targeted therapies, oncolytic viruses, radiotherapy, photodynamic therapy, natural products, synthetic metal-based drugs, and histotripsy-focused ultrasound tumor ablation have also been proven to induce ICD.17–23 Among the hundreds to thousands of compounds that have been screened for ICD induction, only a few have proved effective.24 Thus, searching for effective and available ICD inducers for anticancer therapy remains an important and necessary endeavor.

ICD-based cancer vaccines, including vaccination with ICD inducers or cells undergoing ICD, represent promising approaches in cancer immunotherapy treatment.25 Among the various ways that an ICD inducer can be administered, intratumoral injection can selectively target cancer cells and thus minimize systemic side effects and cytotoxicity.26 Researchers have demonstrated that ethanol can induce oxidation of cellular components, ER stress, and activation of various cell-death pathways.27 Since ER stress is a primary mechanism through which ICD is induced, we investigated the potential of ethanol as an ICD inducer in the current study. Our findings reveal that low-concentration ethanol acts as a potent ICD inducer, and intratumoral administration exhibited both direct and abscopal antitumor effects in mouse models of HCC.

Materials and methodsMice and tumor cell lines

C57BL/6 and BALB/c mice were obtained from the National Laboratory Animal Center (NARLabs, Taipei, Taiwan). The two mouse HCC cell lines (Hep-55.1C and BNL 1ME A.7R.1) and the two human HCC cell lines (HepG2 and PLC/PRF/5) were used in this study.

Measurement of cell viability and detection of markers of ICD

To determine the viability of each HCC cell line after treatment with low-concentration ethanol, cells were cultured in the absence or presence of graded concentrations of ethanol for 24 and 48 hours. Cell number per well was measured by the 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) assay. For ICD markers, calreticulin, HMGB1, and ATP were measured.

Mouse vaccination with ICD-induced tumor cells: rechallenge assay

To assess the effects of vaccination using a vaccine derived from ICD-induced tumor cells, 5×106 Hep-55.1C cells treated with ethanol or doxorubicin were injected subcutaneously into one flank of C57BL/6 mice. 10 days later, 2×106 live Hep-55.1C cells were injected into the contralateral flank of all mice. Tumor growth at the contralateral side was measured over time. For the repeated challenge assay, 2 months after the vaccination-rechallenge assay, live Hep-55.1C cells or live B16F1 melanoma cells were subcutaneously injected into the flank area of tumor-free mice, and tumor growth was measured over time.

Measurement of the ER stress response and the activation of cell-death pathways by Western blotting

Hep-55.1C cells were treated with ethanol for 6, 12, 24, or 48 hours. Cell lysates were probed with an antibody against eIF2α, phospho-eIF2α, ATF4, or ATF6. Signal intensities were normalized to that of vinculin or GAPDH.

Quantitative real-time reverse transcription-PCR to detect unspliced and spliced XBP1 mRNA

HCC cells were treated with 5% ethanol for 4 hours. Quantitative real-time PCR was performed using the Smart Quant Green Master Mix on a LightCycler 96. Primers were designed and synthesized as described.28 Total XBP1 messenger RNA (mRNA) was used as an endogenous internal reference.

Assessing cell-death pathways and HMGB1 release via specific pathway inhibitors

To determine which cell-death pathways became activated, inhibitors of apoptosis, necroptosis, ferroptosis, pyroptosis and autophagy were respectively added to the culture. The HMGB1 released into the medium was then measured by Western blotting as described above.

Mouse models of subcutaneous and orthotopic HCC

For subcutaneous implantation, 107 tumor cells were injected subcutaneously into one or two flanks of each mouse. For intrahepatic implantation, 2×106 tumor cells in 20 µL Matrigel were injected into the left lateral lobe. The volume of each subcutaneous tumor was measured twice per week, while the volume of each liver tumor was measured 1 day before treatment initiation, at regular intervals throughout the intratumoral injection treatment period during laparotomy, and at the end of each experiment. Tumor volume was calculated using the standard formula as previously described.29 Briefly, the maximum tumor diameters were measured in three dimensions (a, b, c), and the volume was calculated using the formula (a × b × c) × π/6.

Detection of ICD markers in vivo following intratumoral administration of low-concentration ethanol

After intratumoral administration of low-concentration ethanol or doxorubicin for 6 and 12 hours, tumors were harvested for immunohistochemical analysis using anti-HMGB1 and anti-calreticulin antibodies. Blood samples were collected at 6 and 12 hours postadministration to measure serum levels of HMGB1 and ATP.

Repeated intratumoral administration of low-concentration ethanol for established HCC tumors

Repeated intratumoral administration (50 µL for subcutaneous HCC tumors, 30 µL for orthotopic HCC tumors) of low-concentration ethanol was carried out 2 or 3 weeks after cancer-cell implantation. This treatment was repeated four times at 1 week intervals. For comparison, injections of equal volumes of saline or doxorubicin (2.5 mg/kg body weight) served as negative and positive controls, respectively.

Transcriptome analysis of primary and distant tumors

Total RNAs, extracted from primary and distant tumors, were hybridized to nCounter probes for 770 genes predefined by the PanCancer Immune Profiling Panel. The gene counts for each sample were normalized using the trimmed mean of the M-values method and quantile normalization and were then converted to log-transformed counts per million for further analysis. For each sample, the immune-related genes were categorized using nSolver V.2.6 and mouse PanCancer Immune advanced analysis software.

HCC metastasis using mouse models of intravenous and intrasplenic injection

For the Hep-55.1C intravenous HCC metastasis model, 1.5×106 cells were injected into the tail vein. For the intrasplenic HCC metastasis model, 1×106 cells were injected into the spleen parenchyma under anesthesia. Three weeks after injection, the lungs and liver were harvested for the measurement of organ weights and the determination of colony numbers through histopathological examination.

Statistical analysis

All statistical analyses were performed with Prism software (GraphPad, California, USA). An unpaired t-test was used for two-group comparisons. One-way analysis of variance with post hoc Tukey’s test was used for the experiments with more than two groups. P value<0.05 was considered statistically significant.

A full description of the materials and methods used for this work is described in online supplemental material and methods.

ResultsLow-concentration ethanol induces ICD in mouse and human HCC cell lines

The viability of the four HCC cells decreased in a concentration-dependent manner after ethanol. Almost all cells were dead 24 hours under the concentration of 10% ethanol (online supplemental figure 1). Thus, the concentration of ethanol in the subsequent in vitro experiments was limited to 5% or lower. Treatment of Hep-55.1C cells with 2.5% or 5% ethanol, and BNL cells with 5% ethanol induced an increase in the presence of calreticulin on the plasma membrane (figure 1A and online supplemental figure 2). Similarly, treatment of the two human HCC cells, HepG2 and PLC/PRF/5, with 2.5% or 5% ethanol increased the translocation of calreticulin to the plasma membrane (figure 1B). To further confirm that ethanol induces ICD, we evaluated the release of HMGB1 and ATP under gradient ethanol treatment. Treatment with 2.5% or 5% ethanol enhanced the release of both ATP (figure 1C) and HMGB1 (figure 1D) into the culture medium in both mouse and human HCC cell lines. These results indicate that low-concentration ethanol effectively induces ICD, comparable to the positive control doxorubicin treatment (online supplemental figure 3), as evidenced by the increased translocation of calreticulin to the plasma membrane and the enhanced release of key DAMPs, including ATP and HMGB1, in both mouse and human HCC cell lines.

Figure 1

Low-concentration ethanol-induced ICD in HCC cell lines in vitro and in vivo vaccination-rechallenge assay. In vitro assay, (A) Ratio of membrane-exposed calreticulin (CRT+) cells, mean fluorescence intensity, and histogram of membrane-exposed calreticulin in mouse and (B) human HCC cell lines after 6 hours of titrated ethanol treatment. (C) ATP release in culture media at 12 hours and (D) HMGB1 levels in culture medium at 48 hours post-treatment, analyzed by Western blotting (Ponceau S staining on PVDF as control). In vivo vaccination-rechallenge assay: ethanol-treated Hep-55.1C cells were injected subcutaneously into one flank of C57BL/6 mice as a tumor vaccine, with untreated Hep-55.1C cells on the opposite flank. (E) Tumor-free mouse percentage and (F) tumor volume post-challenge. (G) Long-term memory in mice vaccinated with 5% ethanol-treated cells, rechallenged with Hep-55.1C or B16F1 cells. Sample sizes ranged from 4 to 6 per group. Data are presented as mean±SD. *p<0.05, **p<0.01, ***p<0.001 versus control group (EtOH 0%). DOXO, doxorubicin; EtOH, ethanol; FITC, fluorescein isothiocyanate; HCC, hepatocellular carcinoma; HMGB1, high-mobility group box 1; ICD, immunogenic cell death; MFI, mean fluorescence intensity; PBS, phosphate-buffered saline; PVDF, polyvinylidene difluoride membrane.

In vivo experiments using ethanol-treated cells as tumor vaccine

The mouse vaccination-rechallenge assay is the gold-standard method for demonstrating ICD. Ethanol-treated Hep-55.1C cells were used as tumor vaccines. At 30 days post-rechallenge, all mice in the negative-control group developed tumors, whereas tumor development was absent or delayed in mice that received Hep-55.1C cells pretreated with doxorubicin or various concentrations of ethanol (figure 1E,F). Notably, the prophylactic effect of treatment with 5% or 2.5% ethanol was superior to that of doxorubicin. For long-term specific immune memory, these mice did not develop tumors even 60 days after rechallenge and still showed a suppressive effect against Hep-55.1C cell tumor growth but not against B16F1 cell tumor growth, indicating the presence of long-term specific memory (figure 1G).

Low-concentration ethanol induces ICD through UPRs in response to ER stress and multiple death pathways

Phosphorylation of eIF2α has been recognized as an indicator of ER stress associated with ICD. In Hep-55.1C cells treated with varying concentrations of ethanol, 5% ethanol for 6 hours significantly enhanced eIF2α phosphorylation, as shown by Western blot analysis (figure 2A). This phosphorylation continued to increase over time, reaching a peak at 8 hours of 5% ethanol treatment (figure 2B). We then assayed the downstream signaling of UPR. Our results showed a significant increase in the expression of ATF4 and CHOP protein, consistent with the observations for eIF2α (figure 2C). We also analyzed the splicing of XBP1 mRNA as a marker of ER-stress IRE1α pathway activation. Spliced XBP1 mRNA dramatically increased after 4 hours of treatment with 5% ethanol, indicating the activation of the IRE1α pathway (figure 2D).

Figure 2

Induction of ICD by low-concentration ethanol via split unfolded protein response and ER stress-associated cell death pathways. (A) Immunoblot analysis of phosphorylated eIF2α (p-eIF2α) and total eIF2α in Hep-55.1C cells treated with ethanol at various concentrations for 6 hours. (B) Time-course analysis of p-eIF2α and total eIF2α expression in Hep-55.1C cells following treatment with 5% ethanol. (C) Expression levels of ATF4 and CHOP proteins after 24 hours of treatment with graded ethanol concentrations were detected by immunoblot. (D) Real-time PCR quantification of unspliced XBP1 (uXBP1) and spliced XBP1 (sXBP1) mRNA to assess IRE1 activation after 4 hours of ethanol treatment at different concentrations. (E) Western blot analysis of HMGB1 release in the medium after 12 hours of 5% ethanol treatment with inhibitors targeting apoptosis (Z-VAD-FMK), necroptosis (Necrostatin-1), ferroptosis (Ferrostatin-1), pyroptosis (Ac-FLTD-CMK), and autophagy (Chloroquine). Vinculin or GAPDH served as loading controls, and Ponceau S staining on PVDF membranes was used as an additional control specifically for medium HMGB1 detection. Each experimental group consisted of three biological replicates. Data are presented as mean±SD. *p<0.05, **p<0.01, ***p<0.001 versus control group (EtOH 0%). ER, endoplasmic reticulum; EtOH, ethanol; HMGB1, high-mobility group box 1; ICD, immunogenic cell death; mRNA, messenger RNA; PVDF, polyvinylidene difluoride membrane.

To further investigate the cell death mechanisms triggered by ER stress induced by low-concentration ethanol, key marker proteins for various cell death pathways were analyzed using Western blotting. The analysis confirmed the activation of apoptosis (cleaved-PARP, cleaved-caspase 3), necroptosis (RIP-1, RIP-3), autophagy (LC3B), and pyroptosis (cleaved-caspase 1), while no significant changes were detected in ferroptosis markers (GPx4, SLC7A11, TFRC), indicating the absence of ferroptosis involvement (online supplemental figure 4). Consistent with this, experiments using several different cell-death inhibitors demonstrated that ethanol-induced ICD involves apoptosis, necroptosis, pyroptosis, and autophagy. Inhibitors targeting these pathways significantly reduced HMGB1 release, whereas a ferroptosis inhibitor had no effect (figure 2E). Taken together, low-concentration ethanol may induce ICD through UPR activation in response to ER stress, mediated by multiple cell death pathways.

Intratumoral administration of low-concentration ethanol induces ICD in tumors

To determine whether intratumoral administration of low-concentration ethanol could also induce ICD in tumors, Hep-55.1C cells were inoculated subcutaneously into C57BL/6 mice. Two weeks later, mice received intratumoral administration of graded concentrations of ethanol. Tumor tissues were collected at 6 and 12 hours after intratumoral administration. Consistent with our in vitro studies, intratumoral administration of low-concentration ethanol or doxorubicin—but not saline—resulted in the release of HMGB1 as well as enhanced translocation of calreticulin to the plasma membrane of tumor cells in vivo (figure 3A,B). Moreover, the serum levels of HMGB1 and ATP in mice that received low-concentration ethanol or doxorubicin were also elevated 6 and 12 hours after intratumoral administration compared with mice that received saline (figure 3C,D). Notably, 1.25% ethanol was also sufficient to induce ICD in vivo. Collectively, these results demonstrated that intratumoral administration of low-concentration ethanol could induce ICD in tumors.

Figure 3

Inducing ICD in HCC in vivo with intratumoral low-concentration ethanol injections shows antitumor effects on primary and distant tumors. (A) Microscopy images showing HMGB1 translocation from the nucleus to the cytoplasm in tumor cells and (B) confocal images illustrating the translocation of calreticulin (CRT) to the plasma membrane, captured at 6 hours after intratumoral administration of varying concentrations of ethanol or doxorubicin (positive control). Serum levels of HMGB1 (C) and ATP (D) at 6 and 12 hours in mice treated with titrated concentration ethanol or doxorubicin compared with saline-treated mice. Sample sizes ranged from 4 to 6 per group. (E) HCC cells were implanted bilaterally, with the primary tumor receiving intratumoral injections of titrated concentrations of ethanol, normal saline (NS), or doxorubicin. The size of the distant tumor was also monitored to evaluate the abscopal effect and assess systemic anticancer immunity. (F) In a parallel setup, the primary tumor received ethanol combined with anti-PD1 (1.25% + α-PD1) to assess the combined systemic anticancer immunity. The number of mice analyzed is indicated in parentheses. Data are presented as mean±SD. *p<0.05, **p<0.01 versus control group (EtOH 0% or normal saline group for tumor size measurement). DOXO, doxorubicin; EtOH, ethanol; HCC, hepatocellular carcinoma; HMGB1, high-mobility group box 1; ICD, immunogenic cell death; IT, intratumoral; PD1, programmed cell death protein-1.

Repeated intratumoral administration of low-concentration ethanol exhibits both direct and abscopal antitumor effects in a mouse model of subcutaneous HCC

The potential antitumor effects of repeated intratumoral administration of low-concentration ethanol were evaluated using a subcutaneous mouse model. Hep-55.1C cells were inoculated into C57BL/6 mice, and 3 weeks later, the mice received weekly intratumoral injections of either saline (negative control) or graded concentrations of ethanol (1.25%, 2.5%, and 5%) for 4 weeks (online supplemental figure 5A). While low-concentration ethanol induced a dose-dependent increase in ICD markers both in vitro and within tumor tissues, the antitumor efficacy of 2.5% and 5% ethanol was notably less effective than that of 1.25% ethanol.

Immunohistochemistry results showed an increase in immunosuppressive cells following repeated intratumoral injections of 2.5% or 5% ethanol, specifically regulatory T cells (Tregs) (identified as FoxP3+ cells) and M2-like macrophages (quantified as the CD206+ stained area). These results suggest the possibility that 5% ethanol might contribute to tissue damage and immunosuppressive cell recruitment, potentially limiting its therapeutic efficacy (online supplemental figure 5 B,C).

Consistent results were also observed for both the direct and abscopal antitumor effects in a bilateral HCC tumor model. In this model, mice received repeated intratumoral administrations of saline, doxorubicin (as a positive control for ICD induction), or graded concentrations of ethanol into tumors on one flank. Mice treated with 1.25% ethanol or doxorubicin showed significantly delayed tumor growth on both flanks compared with the saline-treated group (figure 3E). Notably, the abscopal antitumor effect of repeated intratumoral administration of 1.25% ethanol was superior to that of doxorubicin. These results clearly demonstrated that repeated ethanol intratumoral administration of low-concentration ethanol had direct and abscopal antitumor effects in a mouse model of subcutaneous HCC.

Finally, to evaluate the potential of low-concentration ethanol as a sensitizer for ICI therapy, we conducted experiments using bilateral flank subcutaneous tumor models combined with repeated intratumoral administration of 1.25% ethanol (figure 3F). The results showed that repeated intratumoral injections of 1.25% ethanol combined with anti-programmed cell death protein-1 (PD-1) therapy significantly inhibited tumor growth in both primary and distant tumors compared with controls and the group treated with normal saline intratumoral injections combined with anti-PD-1 therapy. The reduction in tumor volume was particularly pronounced, demonstrating consistent suppression of both primary and distant tumors throughout the study period. These results suggest that low-concentration ethanol enhances the efficacy of ICI therapy, highlighting its potential as a sensitizer in combination with immunotherapy strategies.

Antitumor and abscopal effects via activation of multiple immune response pathways including T cells and the interferon pathway

To explore the underlying mechanisms of the antitumor and abscopal effects of repeated intratumoral administration of low-concentration ethanol, we employed nCounter gene expression analysis. The heatmaps and violin plots showed that in mice receiving repeated intratumoral administration of 1.25% ethanol, cell-type scores were increased relative to mice that received saline, including tumor-infiltrating lymphocytes, T cells, CD8 cells, cytotoxic cells, dendritic cells, natural killer cells, and natural killer CD56dim cells, while Treg cell scores were comparable between the two groups (figure 4A,B). Notably, the differences observed for distant tumors were larger than those for the primary tumor. The heatmap for NanoString Gene Set Enrichment Analysis significance scores showed marked differences in multiple immune-response pathways, including innate, adaptive, T-cell functions, and dendritic-cell functions, among others (figure 4C). The differentially expressed genes between the 1.25% ethanol and control groups, involved in the interferon pathway and T-cell functions, showed significant differences in expression in the primary tumor, as well as in T-cell functions and cytokines and receptors pathways in distant tumors (figure 4D). Finally, Gene Set Enrichment Analysis revealed significant activation of type II interferon signaling, immunoregulatory interactions, and the adaptive immune system of primary tumors as well as the immune system, interleukin 2 family signaling, and immunoregulatory interactions of distant tumors (figure 4E). Taken together, the analysis of primary and distant tumors revealed activation of multiple immune-response pathways including those involving CD8 T cells and the interferon pathway.

Figure 4

Immune cell infiltration and responses in primary and distant tumors after repeated intratumoral injections of low-concentration ethanol. (A, B) Heatmaps display the mean cell type scores of various immune cells in primary and distant tumors treated with 1.25% ethanol (EtOH) or saline (NS), while truncated violin plots display the distribution and median of these scores. These scores are calculated from the mean log2 expression levels of all probes and analyzed using nCounter’s Cell Type Profiling Capabilities. (C) A heatmap shows Gene Set Analysis Significance Scores for primary versus distant tumors in the EtOH and NS groups, derived from the mean log2 expression levels using nCounter’s gene set analysis. (D) Volcano plots illustrate differentially expressed genes between the EtOH and NS control groups. Red points indicate genes with adjusted p values<0.05 and absolute log2 fold changes>0.5. Dark gray points represent genes from the associated gene set, while light gray points encompass all genes analyzed with nCounter probes for the 770 genes predefined by the PanCancer Immune Profiling Panel. (E) The GSEA plot highlights the most enriched pathways from WikiPathways and Reactome, analyzing differences between the EtOH and NS control groups. Normalized Enrichment Score (NES) is noted. Each group consists of three samples. DC, dendritic cell; GSEA, Gene Set Enrichment Analysis; NK, natural killer; TIL, tumor-infiltrating lymphocyte; Treg, regulatory T cell.

Repeated intratumoral administration of low-concentration ethanol also exhibits direct and abscopal antitumor effects in a mouse model of orthotopic HCC

The tumor microenvironment may influence the antitumor immunity and responses. Thus, we conducted our experiment in a mouse model of orthotopic HCC. Hep-55.1C cells were inoculated into both the flank and liver of C57BL/6 mice 3 weeks before intratumoral administration. Thus, these mice had simultaneous liver tumors and subcutaneous tumors.

In the first part of the orthotopic HCC model, saline, doxorubicin, or graded concentrations of ethanol were administered into liver tumors four times at 1 week intervals. The mice that received administration of 1.25% ethanol into liver tumors had significantly decreased tumor volume at the primary tumor in the liver and distant tumors in the subcutis (figure 5A). To further explore the immune response within the tumor microenvironment, immunohistochemistry was conducted to assess the number and functionality of CD8+ cytotoxic T cells (using CD8 and Granzyme B markers) and dendritic cells (using the CD11c marker) in both primary and distant tumors (online supplemental figure 6). The results demonstrated a significant increase in CD8+T cells, functional cytotoxic CD8+T cells, and dendritic cells in both primary and distant tumors following repeated intratumoral injections of 1.25% ethanol. This aligns with the transcriptome analysis results from the nCounter gene expression assay (figure 4), which indicated enhanced immune response pathways in both primary and distant tumors after intratumoral administration of low-concentration ethanol.

Figure 5

Impact of repeated intratumoral low-concentration ethanol administration on orthotopic HCC tumor growth and metastasis. (A) HCC cells were implanted into the liver and one side of the subcutaneous flank area of mice. Tumor sizes were monitored over time after orthotopic tumors received intratumoral injections (IT) of varying concentrations of ethanol, normal saline (NS), or doxorubicin (positive control). (B–C) Evaluation of lung weight, number of metastatic colonies in lung sections, and histological analysis of lung tissue using H&E staining at the conclusion of the experiment. Arrows indicated the presence of metastatic HCC tumor. The number of mice analyzed is indicated in parentheses. Statistical significance is indicated as *p<0.05, **p<0.01 versus the control group (NS). DOXO, doxorubicin; EtOH, ethanol; HCC, hepatocellular carcinoma; SC, subcutaneous.

Moreover, since the Hep-55.1C cell line has been reported in previous studies to develop spontaneous lung metastases from orthotopic liver tumors,29 our results further demonstrated that lung mass, representing both spontaneous metastatic tumor volume and the number of metastatic colonies, was significantly reduced in mice receiving repeated intratumoral administration of 1.25% ethanol compared with other groups (figure 5B). This reduction was additionally validated through histopathological examination of lung sections (figure 5C and online supplemental figure 7).

In the second part of the orthotopic HCC model, repeated intratumoral administration was done in subcutaneous tumors, not in the liver tumor. The intratumoral administration of 1.25% ethanol into subcutaneous tumors resulted in a decreased volume of primary tumors at the subcutis and distant tumors in the liver. Notably, the abscopal antitumor effect of repeated intratumoral administration of ethanol was superior to that of doxorubicin in the distant tumors in the liver (figure 6A). Furthermore, mice receiving repeated intratumoral administration of ethanol had significantly fewer lung metastases (figure 6B,C and online supplemental figure 8).

Figure 6

Abscopal effects of repeated intratumoral administration of low-concentration ethanol on HCC in diverse organs. (A) Mice were implanted with HCC cells in both the liver and subcutaneous flank areas. Both types of tumors were tracked following injections into the subcutaneous tumors with varying concentrations of ethanol, saline (NS), or doxorubicin (used as a control). (B–C) The experiment concluded with measurements of lung weight, counting of metastatic colonies in lung tissue sections, and histological evaluations of the lung. Arrows indicated the presence of metastatic HCC tumor. The number of mice analyzed is indicated in parentheses. Data are presented as mean±SD. *p<0.05, **p<0.01 versus the control group (NS). DOXO, doxorubicin; EtOH, ethanol; HCC, hepatocellular carcinoma; IT, intratumoral; SC, subcutaneous.

Taken together, these results demonstrate that repeated intratumoral administration of low-concentration ethanol has both direct and abscopal antitumor effects, as well as an antimetastatic effect, in a mouse model of orthotopic HCC.

Vaccination with cancer cells treated with low-concentration ethanol induces antimetastatic effects in both intravenous and intrasplenic models of metastasis

To further investigate the anti-metastatic effects of low-concentration ethanol-treated liver cancer cells as a potential “vaccine”, we injected doxorubicin or 5% ethanol-treated Hep-55.1C cells into subcutaneous sites, while phosphate-buffered saline was used as the negative control vaccine. Two weeks later, live, untreated Hep-55.1C cells were inoculated intravenously or intrasplenically to study experimentally induced metastasis models for the lung and liver, respectively.

Mice receiving ethanol-treated Hep-55.1C cells had significantly lower lung mass and fewer lung metastases than the negative control group 3 weeks after intravenous tumor inoculation (figure 7A–D). Furthermore, mice receiving ethanol-treated Hep-55.1C cells also had significantly lower liver mass and fewer liver metastases than the negative control group in the intrasplenic injection model (figure 7E–H). These results demonstrate that vaccination with cancer cells treated with low-concentration ethanol has antimetastatic effects in both the intravenous and intrasplenic models of metastasis.

Figure 7

Antimetastatic effects of vaccination with low-concentration ethanol-treated tumor cells in intravenous and intrasplenic metastasis models. (A–D) Lung weight, whole-section scans, relative metastatic tumor area percentages in the lung, and histological analysis of lung tissue using H&E staining were compared in the intravenous metastasis models. (E–H) Liver weight, whole-section scans of the left liver lobe, relative metastatic tumor area percentages in the liver, and histological analysis of liver tissue using H&E staining were compared in the intrasplenic metastasis models. Arrows indicated the presence of metastatic HCC tumor colonies. Results for mice vaccinated with 5% ethanol-treated Hep-55.1C cells (EtOH 5%), doxorubicin (DOXO, used as a positive control), and PBS (used as a negative control) are presented. Each dot represents an individual mouse. Statistical significance is indicated as *p<0.05, **p<0.01 versus the control group (NS). EtOH, ethanol; HCC, hepatocellular carcinoma; PBS, phosphate-buffered saline.

Discussion

Our results demonstrate that low-concentration ethanol is a potent inducer of ICD, mediated by the ER stress response and multiple cell death pathways. Repeated intratumoral administration of low-concentration ethanol exhibits direct and abscopal antitumoral effects in both subcutaneous and orthotopic HCC mouse models.

In this study, we employed intratumoral administration of low-concentration ethanol. The intratumoral route offered several benefits.26 First, it can achieve high concentrations in situ. Second, it delays the release of the agent into the systemic circulation and thus decreases systemic and off-target toxicities. Third, intratumoral administration could facilitate the entry of immunotherapeutic agents into the tumor. Additionally, our study shows that the intratumoral administration of an ICD inducer can generate tumor-specific adaptive immune responses. Nowadays, intratumoral administration has been used in numerous clinical trials to deliver small molecules or macromolecules, peptides or proteins, other gene products, viruses, bacteria or cells.26 We anticipate that further refinements of intratumoral immunotherapeutic approaches may enhance their clinical applicability and that they are worth further study.

Our results also reveal that low-concentration ethanol has a stronger abscopal antitumor effect compared with doxorubicin, while doxorubicin exhibits a more potent direct antitumor effect. We speculate that doxorubicin—as a chemotherapeutic agent—has greater cell-killing capability. However, not all cells that were killed by doxorubicin fulfilled the criteria of ICD. In other words, the portion of doxorubicin-treated cells that could induce ICD might be smaller than that of ethanol, thus resulting in a smaller abscopal antitumor effect. Our results from the vaccination-rechallenge assay also suggest that ethanol is superior to doxorubicin in ICD induction.

The specifics of how ethanol damages cells have been intensively investigated, including the induction of single-strand and double-strand DNA breaks, formation of DNA, protein or lipid adducts, acetaldehyde formation, reactive oxygen species (ROS) formation, and oxidative/ER stress.30 31 These types of damage may lead to the initiation of cell-death pathways, including apoptosis, necroptosis, pyroptosis, ferroptosis and autophagy in hepatocytes, cardiomyocytes and neurons depending on the concentration and timing of ethanol administration.27 These pathways are not independent but exhibit significant crosstalk, facilitated by shared molecules such as RIPK1, caspase-8, and DAMPs. This interconnectivity allows one pathway to influence or trigger another, amplifying cell death and promoting an ICD response, which enhances immune cell recruitment and inflammation.32–34 Our results are generally in agreement with previous reports with the exception of the ferroptosis pathway, which is characterized by a fatal accumulation of lipids and iron. While several clinical and non-