Preferential Recognition of Drug-Induced Altered Self-Presentation on Tumor Cells, but Not Host Cells, by CD8⁺ T Cells for Eliciting Anti-Tumor Immunity

Abstract

We previously reported that CD8⁺ T cell activation via drug-induced altered self-presentation increases the tumor immunogenicity and cancer immunotherapy efficacy. Although the anti-human immunodeficiency virus drug abacavir (ABC) increases the tumor immunogenicity and induces CD8⁺ T cell anti-tumor immune responses in mice inoculated with tumor cells ectopically expressing the human leukocyte antigen (HLA)-B*57:01, whether such anti-tumor immunity is also triggered in hosts with high HLA-B*57:01 expression levels remain unclear. To verify this, we investigated the anti-tumor effects of ABC on HLA-B*57:01-expressing tumor and normal host cells using HLA-B*57:01 transgenic (B*57:01-Tg) mice in this study. ABC suppressed the HLA-B*57:01-expressing B16F10 tumor growth and increased the CD8⁺ T cell tumor infiltration in B*57:01-Tg mice. ABC also activated the tumor-infiltrating CD8⁺ T cells to secrete interferon-γ but did not promote their proliferation in the tumors of B*57:01-Tg mice. Moreover, ABC did not increase the effector CD8⁺ T cell proportions in the tumor-draining lymph nodes of B*57:01-Tg mice. Overall, CD8⁺ T cells preferentially recognized the ABC-induced altered self-antigen-presenting HLA-B*57:01-expressing tumor cells, but not host cells, to elicit anti-tumor immunity.

INTRODUCTION

Immune checkpoint blockade (ICB) agents, specifically anti-cytotoxic T-lymphocyte-associated protein-4 and anti-programmed death-1 antibodies, target the dysfunctional immune system and stimulate CD8⁺ T cells to kill cancer cells.¹⁾ Although ICB agents targeting cytotoxic T-lymphocyte-associated protein-4 and programmed death-1 have revolutionized the management of various malignancies, including advanced melanoma,¹⁾ several patients remain resistant to ICB-based immunotherapy.²⁾ Over the past few years, overall response rate to ICB agents is 40–50% in endometrial cancer, with ICB-resistant tumors exhibiting drastically low CD8⁺ T cell infiltration, lack of terminally differentiated T cells, mature tertiary lymphoid structures, and dendritic cells, and loss of human leukocyte antigen (HLA) class I.³⁾ As ICB-resistant tumors are poorly immunogenic, novel cancer immunotherapies are necessary to improve their immunogenicity.

Abacavir (ABC), an anti-human immunodeficiency virus drug, activates the CD8⁺ T cells by interacting with HLA-B*57:01, leading to ABC hypersensitivity syndrome (AHS) development.^4–6) ABC alters the shape and chemistry of the antigen-binding cleft of HLA-B*57:01, resulting in the recruitment of approximately 1000 different abnormal peptide antigens, thereby causing abnormal CD8⁺ T cell activation.⁵⁾ We previously reported that the interaction between HLA-B*57:01 and ABC enhances tumor immunogenicity and induces CD8⁺ T cell-mediated anti-tumor immune responses in wild-type mice, proposing a novel strategy to target ICB-resistant tumors with poor immunogenicity.⁷⁾ However, these mice exhibited alloimmune responses to externally introduced HLA molecules, resulting in exaggerated immune responses. Furthermore, wild-type mice could not be used to evaluate the effects of non-cancer cell HLA expression on the anti-tumor effects of ABC in hosts. Therefore, whether ABC elicits immune-mediated anti-tumor responses in HLA-B*57:01-expressing hosts remains unclear.

In this study, we investigated the anti-tumor effects of ABC on HLA-B*57:01-expressing tumor and normal host cells using human–mouse chimeric HLA-B*57:01 transgenic (B*57:01-Tg) mice.⁸⁾ This mouse line was used to induce abnormal ABC-induced immune toxicity and reproduce the AHS symptoms.⁹⁾ We also evaluated the efficacy and safety of cancer immunotherapy based on the HLA-B*57:01–ABC interactions in B*57:01-Tg mice.

MATERIALS AND METHODS

Materials

ABC sulfate was purchased from Carbosynth, Ltd. (Compton, Berkshire, U.K.). Minimum essential medium for B16F10 cell culture was purchased from Nissui Pharmaceutical (Tokyo, Japan). For fluorescence-activated cell sorting (FACS) analysis, phycoerythrin (PE)-Cy/7 anti-mCD8a (53–6.7), PE anti-mouse/human CD44 (IM7), fluorescein isothiocyanate anti-mCD62L (MEL-14), APC anti-mouse interferon-gamma (mIFNγ; XMG1.2), and fluorescein isothiocyanate anti-Ki-67 (JES6-5H4) antibodies were purchased from BioLegend (San Diego, CA, U.S.A.). PE anti-mCD8a antibodies (53–6.7) were purchased from TONBO Biosciences (San Diego, CA, U.S.A.). All other chemicals and solvents were of analytical grade, unless otherwise noted.

Animals

B*57:01-Tg mice carrying chimeric HLA-B*57:01 (in-house colony) were generated as previously described.⁸⁾ All animal experiments were conducted using 6–20-week-old male B*57:01-Tg mice or their littermates (LMs). The animal rooms were maintained under a temperature-controlled (23–25°C) state with 12 h day/light cycles. Polymerase chain reaction-based genotyping was performed using the genomic DNA extracted from mouse tails with the following chimeric HLA primer pair: 5′-GAGCTACTCTCAGGCTGCGTG-3′ (forward) and 5′-CATGTTAGCAGACTTCCTCTGCC-3′ (reverse). Notably, 3–5 mice per group were used in each independent experiment. All mouse-related procedures were approved by the Animal Care Committee of the University of Toyama (Toyama, Japan; Animal Experiment Protocols: A2020INM-5 and A2023INM-1). The mice were humanely treated according to the guidelines of the National Institutes of Health (Bethesda, MD, U.S.A.).¹⁰⁾

Tumor Inoculation and in Vivo Cancer Cell Growth Evaluation

Human–mouse chimeric HLA-B*57:01-expressing murine cancer cells (HLA-B*57:01/B16F10) were established as previously described.⁷⁾ The mice were subcutaneously inoculated with HLA-B*57:01-expressing or control tumor cells (5.0 ×10⁵ cells/mouse) in the right flank and intraperitoneally treated with ABC (5 mg) or vehicle (water) for 9 consecutive days after tumor inoculation. Tumor size was measured every 2 days using Vernier calipers. Tumor volume was calculated using the following equation: Volume (mm³) = (Length × width × width)/2.¹¹⁾

CD8 Immunohistochemistry

After treating the mice with ABC (5 mg) or vehicle for 9 consecutive days post-tumor inoculation, their ear or tumor biopsies were collected, embedded in the Tissue-Tek O.C.T. Compound (Sakura Finetek, Tokyo, Japan), and cryopreserved at −80°C. The tissues were sliced into 5-µm-thick sections using the Leica CM3050S cryotome (Leica Biosystems, Wetzlar, Germany).

CD8 immunohistochemistry was performed as previously reported.⁷⁾ Briefly, the sections were fixed with acetone (Nacalai Tesque, Kyoto, Japan) for 5 min, blocked with 5% fetal bovine serum at room temperature (15–25°C) for 30 min, and incubated with rat anti-CD8a antibodies (clone: YTS169.4; Abcam, Cambridge, U.K.) diluted at 1 : 250 in phosphate-buffered saline containing 0.1% bovine serum albumin at room temperature for 1 h. The sections were further incubated with the corresponding Alexa Fluor-conjugated secondary antibodies (Abcam) and Hoechst 33342 nuclear stain (Thermo Fisher Scientific, Waltham, MA, U.S.A.) diluted at 1 : 250 and 1 : 1000, respectively, in phosphate-buffered saline containing 0.1% bovine serum albumin at room temperature for 1 h. Vectashield (Vector Laboratories, Burlingame, CA, U.S.A.) was used to mount the stained sections.

All samples were imaged under the BZ-X800 microscope (Keyence, Osaka, Japan).

Flow Cytometry

Tumor-infiltrating lymphocytes were prepared as previously described.⁷⁾ Tumor biopsies were collected from mice bearing HLA-B*57:01-expressing B16F10 tumors treated with ABC (5 mg) or the vehicle for 9 consecutive days following tumor inoculation. This endpoint was based on the protocol established in our previous study,⁷⁾ in which tumor-infiltrating CD8⁺ T cells were evaluated, and a significant anti-tumor effect along with the induction of IFN-γ-secreting CD8⁺ T cells was observed. Briefly, tumor biopsies were minced and digested using the serum-free Roswell Park Memorial Institute-1640 medium supplemented with 2 mg/mL collagenase (Roche Diagnostics, Mannheim, Germany) and 0.1 mg/mL DNase I (Roche Diagnostics) at 37°C for 1 h. After filtration with an 80-µm mesh filter, the cell samples were subjected to FACS analysis.

To measure the cytokine production by tumor-infiltrating lymphocytes, the cells were incubated with 500 µL of the Roswell Park Memorial Institute-1640 medium containing 10% fetal bovine serum, 2 mM l-glutamine, 1 mM sodium pyruvate, 1× minimum essential medium with non-essential amino acids (Nacalai Tesque), 10 mM N-(2-hydroxyethyl)piperazine-N'-2-ethanesulfonic acid (HEPES), 0.05 mM 2-mercaptoethanol, penicillin (100 U/mL), and streptomycin (100 µg/mL) at 37°C in a humidified atmosphere with 95% air and 5% CO₂ for 4 h and stimulated with phorbol 12-myristate-13-acetate (50 ng/mL), ionomycin (1 µg/mL), and GolgiStop (BD Biosciences, San Jose, CA, U.S.A.).

For intracellular staining, the cell surfaces were initially stained with the anti-mCD4 and anti-mCD8a antibodies at 4°C for 30 min. Then, the cells were fixed and permeabilized with a fixation/permeabilization solution (BD Biosciences) at 4°C for 20 min and incubated with the anti-mIFNγ or anti-Ki antibodies at 4°C for 30 min.

CD8⁺ T cells isolated from cervical lymph nodes (LNs) were co-stained with fluorochrome-conjugated antibodies against surface markers at 4°C for 30 min.

FACS Canto II (BD Biosciences) was used for FACS analysis. Data were analyzed using the FlowJo software (BD Biosciences).

Statistical Analyses and Data Reproducibility

Statistical analyses were conducted using the GraphPad Prism 8 software (GraphPad Software, La Jolla, CA, U.S.A.). Significance was determined via one-way ANOVA or unpaired two-tailed t-test (Student’s t-test), followed by Bonferroni’s multiple-comparison test. Statistical significance was set at p < 0.05. All experiments were conducted at least twice to ensure reproducibility.

RESULTS

ABC Exerts Anti-Tumor Effects against HLA-B*57:01-Expressing Tumors in B*57:01-Tg Mice

To determine whether ABC exerts anti-tumor effects in B*57:01-Tg mice, we assessed its effects on the growth and host CD8⁺ T cell responses of HLA-B*57:01-expressing B16F10 melanoma cells.

Consistent with our previous study,⁷⁾ ABC significantly suppressed the tumor growth in wild-type LMs bearing HLA-B*57:01/B16F10 tumors (day 10: p < 0.01; Fig. 1a). Tumor growth rate was significantly higher in B*57:01-Tg mice than in wild-type LMs (day 10: p < 0.05). In contrast, growth rate of HLA-B*57:01/B16F10 tumors was lower and average tumor size was significantly lower in ABC-treated B*57:01-Tg mice than in vehicle-treated mice (day 10: p < 0.001; Fig. 1a). Notably, increased CD8⁺ T cell infiltration in ABC-treated HLA-B*57:01/B16F10 tumors was observed in both B*57:01-Tg mice and wild-type LMs (Fig. 1b). ABC did not affect the growth of control B16F10 tumors in B*57:01-Tg mice (Fig. 1c).

Fig. 1. Effects of ABC on the Growth and CD8⁺ T Cell Infiltration of HLA-Expressing B16F10 Tumors in HLA-B*57:01 Tg Mice

(a, b) Tumor growth (a) and representative images of nucleus- and immuno-stained (with CD8) tumor sections (b) of HLA-B*57:01 Tg (B*57:01-Tg) mice and their LMs bearing HLA-B*57:01/B16F10 tumors intraperitoneally treated with ABC (5 mg) or vehicle (water) for 9 consecutive days. Each plot represents the mean ± standard error of the mean (S.E.M.; n = 19–20). ^**p < 0.01 and ^***p < 0.001 vs. vehicle-treated group; ^#p < 0.05 vs. vehicle-treated LMs (one-way ANOVA, followed by Bonferroni’s multiple-comparison correction). Scale bar, 100 µm. Arrows indicate CD8⁺ T cell infiltration. Data are of individual mice in each group or representative of five independent experiments. (c) Growth of B*57:01-Tg mice bearing B16F10 tumors intraperitoneally treated with ABC (5 mg) or vehicle (water) for 9 consecutive days. Each plot represents the mean ± S.E.M. (n = 6–8; unpaired two-tailed t-test [Student’s t-test]). Data are of individual mice in each group or representative of two independent experiments. ABC: Abacavir; HLA: human leukocyte antigen; LMs: littermates; Tg: transgenic.

Overall, ABC exerted anti-tumor effects against B16F10 tumors expressing HLA-B*57:01 and promoted CD8⁺ T cell tumor infiltration in B*57:01-Tg mice. Notably, tumor HLA expression was essential for these effects.

ABC Induces IFN-γ Production, but Not Tumor-Infiltrating CD8⁺ T Cell Proliferation, in B*57:01-Tg Mice

To further assess the effect of ABC on the activation status of CD8⁺ T cells in B*57:01-Tg mice bearing HLA-B*57:01-expressing tumors, we examined IFN-γ production by the HLA-B*57:01-expressing tumor-infiltrating CD8⁺ T cells. ABC significantly increased the IFN-γ-producing CD8⁺ T cell proportions in HLA-B*57:01/B16F10 tumors (Fig. 2a) in B*57:01-Tg mice compared with those in the vehicle-treated group (p < 0.05; Fig. 2b). This effect was also observed in ABC-treated wild-type LMs bearing HLA-B*57:01/B16F10 tumors (p < 0.01; Figs. 2a, 2b).

Fig. 2. Analysis of IFN-γ-Secreting Tumor-Infiltrating CD8⁺ T Cells in ABC-Treated HLA-B*57:01-Expressing Tumors Inoculated into HLA-B*57:01 Tg Mice

(a, b) Representative dot plots (a) and percentages of (b) IFN-γ-producing cells among the tumor-infiltrating CD8⁺ T cells in HLA-B*57:01 Tg (B*57:01-Tg) mice and their LMs bearing HLA-B*57:01/B16F10 tumors. Each plot represents the mean ± S.E.M. (n = 16). ^*p < 0.05 and ^**p < 0.01 vs. vehicle-treated group (one-way ANOVA, followed by Bonferroni’s multiple-comparison correction). Data are of individual mice in each group across 4 independent experiments. IFN: interferon; LMs: littermates; Tg: transgenic.

Next, we analyzed the Ki-67 expression levels in tumor-infiltrating CD8⁺ T cells in ABC-treated HLA-B*57:01-expressing B16F10 tumors to examine cell proliferation. As previously reported,⁷⁾ ABC induced Ki-67 expression in HLA-B*57:01/B16F10 tumor-infiltrating CD8⁺ T cells (Fig. 3a) and significantly increased the Ki-67⁺ cell percentages in B*57:01-Tg mice compared to those in the vehicle-treated group (p < 0.01; Fig. 3b). Notably, comparable Ki-67 expression levels and Ki-67⁺ tumor-infiltrating CD8⁺ T cell percentages were observed between the ABC- and vehicle-treated B*57:01-Tg mice bearing HLA-B*57:01/B16F10 tumors (Figs. 3a, 3b).

Fig. 3. Analysis of Ki-67-Positive Tumor-Infiltrating CD8⁺ T Cells in ABC-Treated HLA-B*57:01-Expressing Tumors Inoculated into HLA-B*57:01 Tg Mice

(a, b) Representative histograms (a) and percentages of (b) Ki-67⁺ cells among the tumor-infiltrating CD8⁺ T cells in HLA-B*57:01 Tg (B*57:01-Tg) mice and their LMs bearing HLA-B*57:01/B16F10 tumors. Each plot represents the mean ± S.E.M. (n = 9–13). ^**p < 0.01 vs. vehicle-treated LMs (one-way ANOVA, followed by Bonferroni’s multiple-comparison correction). The mice were intraperitoneally treated with ABC (5 mg) or vehicle for 9 consecutive days after the subcutaneous inoculation of tumors into their flanks (5.0 ×10⁵ cells/mouse). Data are of individual mice in each group across three independent experiments. ABC: Abacavir; HLA: human leukocyte antigen; LMs: littermates; Tg: transgenic.

CD8⁺ T Cells Are Not Systemically Activated in ABC-Treated Tumor-Bearing B*57:01-Tg Mice

ABC administration leads to AHS in HLA-B*57:01-positive patients with systemic symptoms mediated by CD8⁺ T cell activation.⁴⁾ We previously demonstrated that systemic activation of CD8⁺ T cells by ABC leads to skin toxicity in B*57:01-Tg mice.^8,9) To further evaluate the safety of cancer immunotherapy using ABC, we assessed systemic immunotoxicity in ABC-treated HLA-B*57:01-expressing tumor-bearing B*57:01-Tg mice by analyzing the effector CD8⁺ T cell (CD44^highCD62L^low) populations in pooled cervical LNs. ABC activated the CD8⁺ T cells in the LNs of B*57:01-Tg mice without tumors (Supplementary Fig. 1), consistent with our previous report.⁷⁾ Here, comparable effector CD8⁺ T cell populations were observed between the ABC-treated B*57:01-Tg mice bearing HLA-B*57:01/B16F10 tumors and vehicle-treated mice or LMs (Fig. 4a). The mean percentage of effector CD8⁺ T cells among all CD8⁺ T cells was not significantly different between the ABC-treated B*57:01-Tg and control mice (Fig. 4b). Moreover, CD8⁺ T cell infiltration was not observed in the ear of HLA-B*57:01-expressing B16F10-bearing B*57:01-Tg mice (Supplementary Fig. 2). Therefore, ABC was safe for the immunotherapy of tumor-bearing B*57:01-Tg mice.

Fig. 4. Effect of ABC on T Cell Activation in the Cervical Lymph Nodes of HLA-Expressing B16F10-Bearing HLA-B*57:01 Tg Mice

(a, b) Representative dot plots (a) and percentages (b) of effector T cells among the gated CD8⁺ T cells classified by CD44 and CD62L expression in the cervical lymph nodes of ABC- and vehicle-treated HLA-B*57:01 Tg (B*57:01-Tg) mice and their LMs bearing HLA-B*57:01/B16F10 tumors. Each plot represents the mean ± S.E.M. (n = 7–9; one-way ANOVA, followed by Bonferroni’s multiple-comparison correction). The mice were intraperitoneally treated with ABC (5 mg) or vehicle for 9 consecutive days after the subcutaneous inoculation of tumors into their flanks (5.0 ×10⁵ cells/mouse). Data are of individual mice in each group across two independent experiments. ABC: Abacavir; HLA: human leukocyte antigen; LMs: littermates; Tg: transgenic.

DISCUSSION

In this study, we demonstrated that ABC treatment induced CD8⁺ T cell-dependent anti-tumor immunity in B*57:01-Tg mice bearing poorly immunogenic tumors. Suppressed tumor growth in wild-type LMs compared to that in B*57:01-Tg mice suggests the development of an alloimmune response to externally introduced HLA molecules in LMs, considering that the specific immune response of the HLA-B*57:01–ABC interaction led to an anti-tumor effect in B*57:01-Tg mice. Furthermore, the failure to induce an anti-tumor effect in B*57:01-Tg mice bearing HLA-lacking murine tumors suggests that non-cancer cell host HLA expression does not affect the anti-tumor effect of ABC. These results highlight the preferential recognition of ABC-induced altered self-antigen-presenting HLA-B*57:01-expressing tumor cells, but not host cells, by CD8⁺ T cells to elicit anti-tumor immunity. Additionally, preferential CD8⁺ T cell priming via HLA-B*57:01–ABC interactions in tumors was important for eliciting anti-tumor immunity, as significant anti-tumor effects were elicited in HLA-B*57:01-expressing tumor-bearing mice lacking HLA-B*57:01-expressing non-cancerous cells (i.e., wild-type LMs).

In contrast to those in wild-type LMs, activated tumor-infiltrating CD8⁺ T cells in ABC-treated B*57:01-Tg mice were Ki-67 negative, indicating that these CD8⁺ T cells did not proliferate in the tumors and that their proliferation was not essential for eliciting anti-tumor responses. Alternatively, ABC-stimulated CD8⁺ T cells in other lymphoid organs were possibly recruited to HLA-B*57:01-expressing tumors in B*57:01-Tg mice, as observed in a previous study.⁸⁾ Notably, this study found no significant increase in the effector CD8⁺ T cell proportions in the LNs of ABC-treated tumor-bearing B*57:01-Tg mice. IFN-γ-driven C–X–C chemokine receptor 3 and C–X–C chemokine ligand (CXCL)-9/10/11 axis is critical for effector anti-tumor CD8⁺ T cell recruitment to solid tumors, including melanomas.¹²⁾ We previously reported the infiltration of C–X–C chemokine receptor 3-dependent CD8⁺ T cells to ABC-treated HLA-B*57:01-expressing tumors.⁷⁾ In contrast, another study reported increased serum CXCL10 levels in non-tumor-bearing ABC-treated B*57:01-Tg mice with skin inflammation,¹³⁾ suggesting that CXCL10 is not responsible for CD8⁺ T cell recruitment to tumors in these mice. Although further investigations are essential, CXCL9 and CXCL11 possibly regulate CD8⁺ T cell recruitment to HLA-B*57:01-expressing tumors in B*57:01-Tg mice. Unlike that in non-tumor-bearing mice, tumors can locally modulate APC function in LNs to elicit anti-tumor CD8⁺ T cell immunity in tumor-bearing mice.¹⁴⁾ Therefore, localization of ABC-induced activated CD8⁺ T cells was possibly affected by tumor implantation into B*57:01-Tg mice, leading to the failure of effector CD8⁺ T cell accumulation in the LNs of ABC-treated tumor-bearing mice.

In conclusion, this study demonstrated that CD8⁺ T cell tumor recruitment via preferential HLA-B*57:01–ABC interactions in tumors induced tumor-selective CD8⁺ T cell activation in ABC-treated tumor-bearing B*57:01-Tg mice. This study also demonstrated the efficacy of cancer immunotherapy based on CD8⁺ T cell anti-tumor immunity mediated by HLA–drug interactions in B*57:01-Tg mice. Notably, ABC did not cause significant systemic immunotoxicity in tumor-bearing B*57:01-Tg mice. By contrast, ABC-induced skin toxicity was observed only when CD4⁺ T cells were depleted in B*57:01-Tg mice, reflecting the low CD4⁺ T cell counts observed in patients infected with human immunodeficiency virus (HIV).⁹⁾ Our findings highlight the potential of strategies targeting the HLA-B*57:01–ABC interactions for cancer immunotherapy of HLA-B*57:01-positive patients with ICB-resistant tumors with poor immunogenicity and use of ABC prodrugs to enhance tumor selectivity and treatment safety. Although the HLA-B protein has no direct effect on abacavir pharmacokinetics or pharmacodynamics, it is important to examine whether ABC pharmacokinetics are altered by HLA expression levels and affect the efficacy of the cancer immunotherapy.¹⁵⁾ Furthermore, strategies combining ABC with a chemokine ligand released from tumors can be used for targeted drug delivery to poorly immunogenic HLA-B*57:01-expressing tumors. ABC-induced anti-tumor immunity is important for AHS management in clinical settings. The above-mentioned methods can improve the tumor selectivity and reduce the risk of systemic AHS symptoms in HLA-B*57:01-positive patients. Overall, our findings can contribute to the development of novel strategies targeting drug–HLA interactions to treat poorly immunogenic ICB-resistant tumors in clinical settings.

Acknowledgments

This study was supported by the Japan Society for the Promotion of Science (JSPS KAKENHI; Grant Nos. 20K22801, 21H02640, 21H02783, and 21K15311) and Takeda Science Foundation.

Conflict of Interest

The authors declare no conflict of interest.

Data Availability

All data are available upon reasonable request from the corresponding authors.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1)  Carlino MS, Larkin J, Long GV. Immune checkpoint inhibitors in melanoma. Lancet, 398, 1002–1014 (2021).
• 2)  Haslam A, Prasad V. Estimation of the percentage of US patients with cancer who are eligible for and respond to checkpoint inhibitor immunotherapy drugs. JAMA Netw. Open, 2, e192535 (2019).
• 3)  Grau Bejar JF, Yaniz Galende E, Zeng Q, Genestie C, Rouleau E, de Bruyn M, Klein C, Le Formal A, Edmond E, Moreau M, Plat A, Gouy S, Maulard A, Pautier P, Michels J, Oaknin A, Colomba-Blameble E, Leary A. Immune predictors of response to immune checkpoint inhibitors in mismatch repair-deficient endometrial cancer. J. Immunother. Cancer, 12, e009143 (2024).
• 4)  Chessman D, Kostenko L, Lethborg T, Purcell AW, Williamson NA, Chen Z, Kjer-Nielsen L, Mifsud NA, Tait BD, Holdsworth R, Almeida CA, Nolan D, Macdonald WA, Archbold JK, Kellerher AD, Marriott D, Mallal S, Bharadwaj M, Rossjohn J, McCluskey J. Human leukocyte antigen class I-restricted activation of CD8+ T cells provides the immunogenetic basis of a systemic drug hypersensitivity. Immunity, 28, 822–832 (2008).
• 5)  Illing PT, Vivian JP, Dudek NL, Kostenko L, Chen Z, Bharadwaj M, Miles JJ, Kjer-Nielsen L, Gras S, Williamson NA, Burrows SR, Purcell AW, Rossjohn J, McCluskey J. Immune self-reactivity triggered by drug-modified HLA-peptide repertoire. Nature, 486, 554–558 (2012).
• 6)  Rodriguez-Nóvoa S, Soriano V. Current trends in screening across ethnicities for hypersensitivity to abacavir. Pharmacogenomics, 9, 1531–1541 (2008).
• 7)  Susukida T, Sasaki SI, Shirayanagi T, Aoki S, Ito K, Hayakawa Y. Drug-induced altered self-presentation increases tumor immunogenicity. Biomed. Pharmacother., 165, 115241 (2023).
• 8)  Susukida T, Aoki S, Kogo K, Fujimori S, Song B, Liu C, Sekine S, Ito K. Evaluation of immune-mediated idiosyncratic drug toxicity using chimeric HLA transgenic mice. Arch. Toxicol., 92, 1177–1188 (2018).
• 9)  Susukida T, Kuwahara S, Song B, Kazaoka A, Aoki S, Ito K. Regulation of the immune tolerance system determines the susceptibility to HLA-mediated abacavir-induced skin toxicity. Commun. Biol., 4, 1137 (2021).
• 10)  Guide for the Care and Use of Laboratory Animals. 8th ed., National Academic Press, Washington DC (2011).
• 11)  Faustino-Rocha A, Oliveira PA, Pinho-Oliveira J, Teixeira-Guedes C, Soares-Maia R, da Costa RG, Colaco B, Pires MJ, Colaco J, Ferreira R, Ginja M. Estimation of rat mammary tumor volume using caliper and ultrasonography measurements. Lab Anim. (NY), 42, 217–224 (2013).
• 12)  Kohli K, Pillarisetty VG, Kim TS. Key chemokines direct migration of immune cells in solid tumors. Cancer Gene Ther., 29, 10–21 (2022).
• 13)  Kazaoka A, Kumagai K, Matsushita J, Aida T, Kuwahara S, Aoki S, Ito K. Pathological changes in various organs in HLA-B*57: 01 transgenic mice with abacavir-induced skin eruption. Toxicol. Res., 40, 223–235 (2024).
• 14)  O’Melia MJ, Rohner NA, Manspeaker MP, Francis DM, Kissick HT, Thomas SN. Quality of CD8⁺ T cell immunity evoked in lymph nodes is compartmentalized by route of antigen transport and functional in tumor context. Sci. Adv., 6, eabd7134 (2020).
• 15)  Barbarino JM, Kroetz DL, Altman RB, Klein TE. PharmGKB summary: abacavir pathway. Pharmacogenet. Genomics, 24, 276–282 (2014).