Remember me
HLA-restricted T cell mediated drug hypersensitivity reactions (HLA-DHRs) are classified as type IV allergic reactions. These are idiosyncratic adverse reactions with delayed onset of T cell responses recognizing drug epitopes presented by HLA molecules that can result in unwanted clinical outcomes such as skin rashes, drug-induced liver injury (DILI) and even death. The low frequency and unpredictability of these events make the study of causal factors and immune mechanisms difficult in human. Genome-wide association studies have provided insights on genetic associations with DHRs, revealing a higher frequency of expression of certain HLAs (risk alleles) in individuals affected by specific DHRs (summarized in Table 1 of [1]). In the past 5 years, the generation of HLA transgenic (Tg) mice has emerged as an appealing strategy to study DHRs linked to particular HLA risk alleles. These animal models do not yet reflect the etiology of the disease in humans for a variety of reasons that include species-related differences (i.e., drug metabolism), lack of individual diversity of MHC class I molecules, and differences in drug dosing or treatment regimens. However, even with these limitations, HLA Tg mice are still viewed as an important step towards development of animal models with clinical characteristics similar to those in patients with high risk HLAs undergoing drug adverse reactions.
Box 1: no caption available
Table 1 - Studies using HLA class-I Tg mouse models of DHRs HLA-I Tg mouse strain Drug and route of administration Intended DHR Immune alterations (adopted strategy) Relevant findings Reference HLA-B∗57:01 Tg mice ABC (topical or oral) Cutaneous adverse reactions Immuno-competent • Topical application on ear (3 days) Increased skin inflammation and CD8+ TE/EM cells in draining LN.ABC, abacavir; CBZ, carbamazepine; DHRs, drug hypersensitivity reactions; DILI, drug-induced liver injury; FLX, flucloxacillin.
Here, we review the advances and lessons learned from HLA class-I Tg mouse models of DHRs. We highlight potential applications of these animal models towards the study of mechanisms of drug-induced immune activation and host tolerance, from in vitro responses to drugs to in vivo experiments that reveal immune pathogenic pathways that cannot be addressed in humans.
PROOF-OF-CONCEPTStudies of human disease and new drug treatment strategies rely heavily on animal studies. To date, in the field of idiosyncratic drug adverse reactions, efforts have focused primarily on generating Tg mice expressing HLA-B∗57:01 because of its linkage with skin hypersensitivity reactions caused by the reverse transcriptase inhibitor abacavir (ABC) prescribed for patients with AIDS [2], and with DILI caused by the antibiotic flucloxacillin (FLX) used to treat gram-positive bacterial infections [3]. Although similar, HLA-B∗57:01 Tg mouse strains have slight genetic differences in the HLA chimeric molecules [4,5▪] that may impact T cell receptor (TCR) engagement and levels of HLA expression and alter the quality of the subsequent T cell response. Proof-of-concept in vitro studies with ABC and drug-naïve cells from HLA-B∗57:01 Tg mice showed a rapid activation of CD8+ T cell upon ABC exposure with enhanced secretion of interferon gamma (IFN-γ), granzyme B, and interleukin 2 (IL-2), and cell proliferation. HLA expression on antigen presenting cells was required as well as CD28 signaling through engagement of co-stimulatory CD80/86 molecules on antigen presenting cells [5▪]. Similarly, CD8+ T cells from drug naïve HLA∗B:57:01 Tg mice responded to FLX in vitro after prolonged repeated activation of cells because of a lower frequency of drug reactive T cells [6▪] compared with the reported polyclonal ABC-induced response [7]. These results paralleled those shown in human peripheral blood mononuclear cells (PBMCs) expressing HLA-B∗57:01 [8–11], proving the value of expressing the HLA risk allele in Tg mice to study immune activation by drugs.
INITIAL IN VIVO MOUSE STUDIES WITH DRUGS ASSOCIATED TO DRUG HYPERSENSITIVITY REACTIONSIn in vivo studies with different immunocompetent HLA Tg murine strains, administration of drugs associated with DHRs has proven to be insufficient to trigger clinical disease, although T cell activation was demonstrated [4,5▪,6▪,12,13]. The lack of clinical symptoms in drug-treated mice reflects drug tolerance as experienced by most human subjects treated with the same medicines, indicating that in mice, like in humans, environmental and host factors other than the HLA risk allele are involved in the development of drug pathology. Of note, animal models with low predictive values as those in human are not feasible for experimental studies of DHRs. Thus, to improve the HLA Tg murine models, important factors need to be considered including: (i) inter-species differences at the level of immune-inhibitor pathways such as regulatory T cell (Treg) function and check-point molecules, (ii) inflammation from ongoing disease (i.e., infections) that activates antigen presenting cells for T cell priming, (iii) levels of preexisting drug reactive T cells perhaps from previous priming by drugs of a similar class, (iv) drug bioavailability, drug metabolism and adduct formation resulting in different concentrations of target drug neo-antigens required to stimulate T cells in each species.
EFFORTS TO BREAK IMMUNE-TOLERANCE AND CHARACTERIZATION OF IMMUNE REACTIONSDrug tolerance is the immune outcome desired for therapeutic drugs under development. However, HLA presentation of certain drugs to T cells [14] can tip the balance of tightly regulated host mechanisms and lead to activation of auto-immune-like adverse responses. Because of the infrequency and unpredictability of these drug-induced adverse reactions, the study of the underlying mechanisms has been challenging in human. In contrast, manipulation of immune effector versus regulatory responses to drugs is possible in experimental mice, thus HLA Tg mice provide a foundation to understand factors determining host tolerance and drug adverse reactions in vivo.
To initiate an immune response to foreign antigens, T cells require priming under immunogenic conditions. Topical application of drug is an effective method of T cell sensitization. Studies in HLA-B∗57:01 Tg mice have explored the topical route, among others, to induce immune responses to ABC. In contrast to the ABC hypersensitivity syndrome seen in 55% of HLA-B∗57:01+ treated patients [2], none of the ABC-administered HLA-B∗57:01 Tg mice experienced systemic symptoms or skin rashes even on sites where the drug was topically applied. In one report, a 3-day course of ABC application on the ear led to skin inflammation and accumulation of CD8+ effector/memory T cells in the draining lymph nodes. A similar CD8+ T cell response was also observed with a week of oral ABC [4]. In contrast, Cardone et al.[5▪] showed that ear painting with ABC was not sufficient to induce cutaneous immune infiltration on a similar strain of HLA-B∗57:01 Tg mice, even when drug was co-administered intraperitoneally. In these animals, systemic ABC-reactive CD8+ T cells displayed enhanced expression of PD-1 within days of drug treatment, peaking at 5 days but had decreased by 10 days even with continuous drug treatment [5▪]. The self-limited ABC-T cell response required early recognition of ABC-induced neo-antigens by the TCR along with CD28 engagement of co-stimulatory molecules on antigen presenting cells as evidenced by cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) Ig blocking [5▪,15]. It was concluded that ABC-induced CD8+ T cells either became anergized or deleted possibly due to neo-antigen presentation on resting “tolerogenic” dendritic cells (DCs), a cell type that is kept in check by immune regulatory mechanisms (Fig. 1A).
FIGURE 1: Drug-induced immune response in HLA-B∗57:01 Tg mouse models of skin hypersensitivity and liver injury – (A) Immunocompetent mice have an intact immune system. Upon drug administration to the animal (i.e., ABC), drug-naïve CD8+ T cells can sense drug-neoepitopes in the context of HLA through the TCR and get activated with increased PD-1. The presence of Treg, however, limits the effector response by (a) blocking CD80 on dendritic cells (DC) with CTLA-4, (b) sequestering IL-2 through the constitutively express IL-2R, (c) secreting inhibitory cytokines. Activated CD8+ T cells will be anergized/deleted by lack of co-stimulation and through PD-1/PDL-1 interactions. (B) Animals depleted of Treg experience drug-neoantigen presentation with enhanced co-stimulation by DC. The magnitude of the CD8+ T cell response is affected by the number of drug-epitopes presented (more in ABC than FLX, and more in HLA Tg/ H2KO than HLA Tg) (a), and whether PD-1 is being expressed (more in PD-1+/+ than in PD-1−/− (KO)) (b). Drug-induced CD8+ T cells upregulate activation molecules (i.e., PD-1, IL-2R) and secrete cytokines that will assist further in the maturation of DC and amplification of the T cell response. ABC-activated T cells can migrate to the skin by expressing skin-homing receptors (CLA) to cause inflammation and tissue damage (c). FLX presented by both HLA-B∗57:01 and non-HLA molecules will result in the expansion of FLX-CD8+ T cells with the potential to cause liver injury (d) (as seen in non-HLA mouse models). In Tg animals not expressing murine MHC I molecules (H-2), FLX-reactive cells are enriched in the effector CD8+ T cell pool (e), and even more in PD-1−/− KO mice (f). FLX-CD8+ T cells infiltrate the liver and show cytotoxic activity on primary hepatocytes in vitro. Hepatic inflammation and histopathology were observed without ALTs, due to liver tolerance. ABC, abacavir; FLX, flucloxacillin.
HIV patients lose CD4+ T cells including T regulatory cells (Treg) from both direct virus infection and a chronic state of “immune activation” [16,17]. To translate the host immune status of HIV patients prior to ABC treatment to the HLA Tg mice models, monoclonal anti-CD4 antibodies were administered prior to ABC drug dosing. Depletion of CD4+ T cells led to vigorous CD8+ T cell expansion and activation in HLA Tg animals receiving drug [5▪,18▪], associated with expression of activation/exhaustion markers (PD-1, CD25, LAG3, TIM3, CXCR3 and KLRG1) and skin homing receptor CLA, that was not seen in ABC only treated animals (Fig. 1B). ABC painted skin showed tissue scarring due to robust CD8+ T cell and macrophage infiltration along with inflammatory cytokines and chemokines by 2-3 weeks of treatment initiation, thus modeling ABC-induced skin hypersensitivity in humans [5▪].
Depletion of CD4+ T cells most likely allows CD8+ T cell activation to ABC secondary to depletion of Tregs. Removal of Tregs in HLA Tg mice was necessary for expansion and differentiation of ABC-reactive CD8+ T cells to induce pathogenesis by stimulating DCs to present drug-modified neo-epitopes on HLA-B∗57:01 (signal 1) together with optimal co-stimulation (signal 2). Tregs, through CTLA-4 can block or remove CD80/CD86 from DCs by trans-endocytosis [19] resulting in impaired T cell co-stimulation via CD28 (Fig. 1A). Depletion of Treg facilitates early CD80 expression on DCs in the absence of PD-L1 inhibitory effects on PD-1 on TCR-engaged CD8+ T cells [20]. Subsequently, CD80 molecules bind and signal through CD28 on the CD8+ T cell facilitating TCR signaling and T cell activation (Fig. 1B). In addition to effects on DCs, Tregs suppress T cells through competition for IL-2 [21,22] (Fig. 1B). Hence, depletion of Tregs reduces consumption of IL-2 through the Treg cell surface high affinity IL-2 receptor [23] allowing activated CD8+ T cells to expand and differentiate in response to autocrine or paracrine IL-2 (signal 3) leading to immune-mediated tissue injury. The findings in the HLA-B∗57:01 Tg mouse model parallel ABC hypersensitivity reactions in HIV infected patients. More recent experiments from our lab using strategies that specifically deplete Tregs have confirmed results observed with CD4 depletion in the HLA-B∗57:01 Tg mouse (Norcross M & Puig M, personal communication).
HLA-B∗57:01 Tg mice have also provided a model to study neoantigen presentation and liver damage using the antibiotic FLX. FLX causes DILI mostly in HLA-B∗57:01 and HLA-B∗57:03 positive patients [24▪]. Unlike ABC-DHR, which is exclusively linked to HLA-B∗57:01 [2], FLX-induced DILI has a lower negative predictive value [3,24▪]. FLX is thought to be presented through both covalent and noncovalent mechanisms [9–11,25]. To elucidate the factors leading to FLX-DILI associated with HLA-B∗57:01, Ananthula and collaborators [6▪] treated HLA-B∗57:01 Tg mice with FLX by skin sensitization with retinoic acid followed by oral gavage to enhance generation and mobilization of drug-reactive cells to the gut-liver axis and induce liver injury. In immune-competent mice, treatment with FLX was not sufficient to activate T cells responses to the drug. Similar to what was observed with ABC, depletion of CD4+ T cells was required to generate drug-specific T cells in FLX-treated animals. However, the FLX-generated immune response was of low magnitude, short-lived, and insufficient to cause liver injury in the animals, possibly due to a low number of T cell clones responding to drug (Table 1). Earlier studies in immunocompromised, non-HLA expressing mice by Nattrass and collaborators [26] showed that FLX administration led to transient and mildly elevated hepatic enzymes. The authors detected FLX-reactive CD8+ T cells that had hepatocyte cytolytic activity in vitro. These results indicate that FLX-epitopes are not exclusively presented by HLA-B∗57:01 to T cells, in agreement with GWAS [3,24▪] and in vitro studies with human PBMCs cultured with FLX [9–11,25], despite the higher association of HLA-B∗57:01 with FLX-DILI. Determining whether HLA-B∗57:01 has unique characteristics that make carriers of this allele more susceptible to DILI when treated with FLX is not possible in subjects or cells expressing multiple major histocompatibility complex class I (MHC-I) molecules, including the HLA Tg mice with intact murine MHC-I. Therefore, to gain knowledge on mechanisms leading to FLX-DILI when drug is presented by HLA-B:57:01, our group generated a strain that exclusively expresses the human allele [HLA Tg/ H2-KbDb KO mice (Tg/KO)] [27▪]. These animals had decreased numbers of CD8+ T cells compared to HLA Tg animals due to the H2-KbDb deletion [28] but increased expression of the HLA transgene per cell [6▪], thereby resulting in enrichment of drug-reactive T cells in the CD8+ T cell repertoire as seen in other models for nondrug antigens [29]. Tg/ KO mice treated with FLX and aCD4 Ab had increased levels of drug-induced PD1+CD8+ T cells in both lymphatic organs and liver (Fig. 1B). Nevertheless, liver pathology was not observed in the Tg/ KO strain despite of detection of infiltrating drug-reactive cytotoxic T cells, possibly due to factors controling liver tolerance such as the lack of co-stimulatory molecules on hepatic DCs. This second generation of HLA Tg mice expressing only the HLA-B∗57:01 allotype now provides an animal model with higher specificity, susceptibiltiy and enhanced drug-specific T cell responses.
CD8+ T cells that respond to ABC-neoepitopes presented on HLA-B∗57:01 express enhanced PD-1 and therefore should be inhibited by PD-L1 on resting DC that express low levels of CD80. These observations together with reports of adverse drug reactions in cancer patients undergoing checkpoint immune-therapy [30] prompted studies of interference with checkpoint pathways to enhance T cell responses to drug. Studies with human leukocyte antigens class I (HLA-I) Tg animals treated with ABC and anti-PD1 blocking antibody did not generate a robust CD8+ T cell response suggesting that Tregs can control T cell responses even when this immunoregulatory pathway was blocked [5▪]. This was also observed in non-HLA murine models of amodiaquine direct toxicity using PD-1 KO mice that required additional CTLA-4 blocking to sustain transient amodiaquine liver inflammation [31]. HLA-B∗57:01 Tg animals lacking PD-1 molecules were subsequently generated to enhance DHR to ABC and FLX (Fig. 1B). One report of a PD-1 KO strain expressing HLA-B∗57:01 observed some enhancement of CD8+ T cell responses to ABC, although deletion of CD4+ T cells was required to see more significant effects [18▪]. However, T cell activation by FLX was not detected when drug was administered to these same mice [32]. In another study with FLX, eliminating the expression of PD-1 in Tg/ KO mice (HLA Tg/double knock-out) and depleting CD4+ T cells resulted in exacerbation of FLX-induced liver inflammation with similar histological findings reported in humans experiencing FLX-DILI but without serum alanine transaminase (ALT) elevations. Hepatic infiltration of FLX-reactive cytotoxic CD8+ T cells was observed and thought to be controlled by liver-intrinsic immune suppressive mechanisms [6▪]. Additional studies are being pursued by these researchers to characterize the liver tolerogenic mechanisms.
Innate immune inflammation triggered by pathogen infections may also play a role in DHRs. Song and collaborators [12] administered a TLR ligand, CpG ODN, to ABC-treated HLA-B∗57:01 Tg mice to increase inflammation. Innate activation by the TLR9 agonist synergized with ABC-reactive CD8+ T cells to induce mild and transient liver injury. The authors hypothesized that liver infiltrating lymphocytes in HLA-B∗57:01 Tg mice, but not HLA-B∗57:03 Tg mice, were responsible for hepatic alterations.
A limitation of these preclinical models is that, unlike humans, animals do not have preexisting disease at the time of initiation of drug treatment, such as viral or bacterial infections. Underlying infections may alter the host immune system and contribute to DHRs. Thus, in the absence of disease, strategies such as the ones described in this article are required to lower the threshold of tolerance to drug-induced immune responses.
TRANSLATIONAL STUDIESIdentification of drug-modified epitopes has been possible by immunoproteomic analysis of human B cells expressing monoallelic HLA [27▪,33]. Several FLX-modified peptides on lysine residues, including one from the HLA-B∗57:01 itself, have been tested for immunogenicity in Tg/KO mice using chemically synthesized peptides [27▪]. These studies identified different immunogenicity patterns, depending on the peptide sequence and the position of the drug-covalent attachment. Peptides with FLX-modified lysine residues (FLX-Lys) at P4 or P6 generated specific CD8+ T cell responses to the immunogen and not to the unmodified peptide or versions of the peptide sequence with FLX-Lys in other positions in the sequence. In silico modeling suggested that the FLX-Lys at P4 may be recognized in a manner similar to longer peptides that form a bulge in the middle of the HLA presented peptide. In contrast, FLX-Lys at -1 PΩ near the C-terminus of the peptide were either not immunogenic or showed cross-reactivity with the parent peptide. Induction of an autoimmune response to the parent sequence by the FLX-modified sequence may be relevant to drug immune responses in vivo, especially in patients with DILI who continue to have liver inflammation long after drug therapy is stopped. In a report by Gao and collaborators [32], immunization of HLA-B57:01 Tg / PD-1 KO mice with human serum albumin previously incubated with FLX showed an increase in CD44+CD62L- CD8+ T cells in the LN of the animals, although it was not clear whether these were antigen specific, and no clinical symptoms were further explored in these animals.
Drug-induced immune-mediated tissue damage is dependent on TCRs that specifically recognize drug-modified epitopes generated by either covalent (haptenated) or noncovalent mechanisms. Several TCRs have been isolated from oligoclonal T cells from human blood, tissue or blisters from patients with drug reactions, and some TCRs show specific drug responses in vitro[13,34–36]. Mouse HLA Tg strains are ideal to test TCRs for function in causing disease in vivo once TCRs are transfected into normal T cells and transferred to mice. One example is the transfection of a “public” TCR to carbamazepine (CBZ) into T cells from an HLA-B∗15:02 Tg mouse lacking murine MHC-I [13] (Table 1). These mice do not respond to CBZ alone, similar to what was seen with ABC and FLX in HLA-B∗57:01 Tg mice. However, after transfer of normal T cells transfected with the public CBZ-specific TCR followed by treatment with CBZ, hair loss was noted along with mild skin inflammation by histology. Although these experiments were not definitive, strategies using transfer of TCRs into HLA Tg mice are promising to model drug specific immune recognition in vivo especially when the epitope specificity of the TCR is well characterized. This strategy has also been used to dissect in vivo responses to the heavy metal beryllium where TCRs to a chemokine epitope modified by beryllium was able to cause lung toxicity in a class II HLA-DP2 Tg mouse [37]. Transfer of HLA-B∗57:01 specific TCRs that recognize FLX-haptenated peptides into HLA Tg mice will be useful to document epitope targeting in liver injury.
Another intriguing finding generated from immunoproteome studies with FLX-treated B-cells was the detection of FLX conjugated to Lys146 of the HLA-B∗57:01 sequence. This residue is recognized by the inhibitory KIR3DL1 receptor and thus drug hapten binding to this site could prevent KIR binding resulting in NK or even CD8+ T cell activation [38]. The same region of HLA-B∗57:01 is presented by HLA as a haptenated peptide after processing. Modelling the peptide in HLA-B∗57:01 [27▪] would place the FLX in the -1 PΩ position which again is a key peptide amino acid site that would interfere with KIR3DL1 binding. The same drug adduct could interfere with KIR binding in two different ways again inducing NK/CD8 activation. These hypotheses for KIR interactions by drugs could also be studied in the HLA Tg mice.
CONCLUSIONHLA Tg mice have provided important insights into the immunobiology of DHRs. These in vivo models now allow not only the role of the HLA restriction element to be controlled and studied, but also the contribution of immuno-regulatory cell and molecular interactions for development of drug tolerance and adverse reactions. In particular, HLA-I Tg animal models show a central role for Treg in the outcome of drug exposure through interactions with presenting cells and effector T cells. Immune checkpoint pathways participate in drug tolerance along with factors that control HLA expression and drug concentrations. In the future, HLA Tg mice expressing HLAs that are linked to immune-mediated drug adverse events will be key to understanding drug-specific and host immune factors that cooperate in initiating and preventing severe skin reactions and DILI.
AcknowledgementsWe thank Kirthiram Krishnaveni Sivakumar and Tongzhong Ju for critically reading the manuscript.
Financial support and sponsorshipNone.
Conflicts of interestThere are no conflicts of interest.
REFERENCES AND RECOMMENDED READINGPapers of particular interest, published within the annual period of review, have been highlighted as:
▪ of special interest
▪▪ of outstanding interest
REFERENCES 1. Li Y, Deshpande P, Hertzman RJ, et al. Genomic risk factors driving immune-mediated delayed drug hypersensitivity reactions. Front Genet 2021; 12:1–14. 2. Mallal S, Phillips E, Carosi G, et al. HLA-B∗5701 screening for hypersensitivity to abacavir. N Engl J Med 2008; 358:568–579. 3. Daly AK, Donaldson PT, Bhatnagar P, et al. HLA-B∗5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet 2009; 41:816–819. 4. Susukida T, Aoki S, Kogo K, et al. Evaluation of immune-mediated idiosyncratic drug toxicity using chimeric HLA transgenic mice. Arch Toxicol 2018; 92:1177–1188. 5▪. Cardone M, Garcia K, Tilahun ME, et al. A transgenic mouse model for HLA-B∗57:01–linked abacavir drug tolerance and reactivity. J Clin Invest 2018; 128:2819–2832. 6▪. Ananthula S, Krishnaveni Sivakumar K, Cardone M, et al. Development of mouse models with restricted HLA-B∗57:01 presentation for the study of flucloxacillin-driven T-cell activation and tolerance in liver injury. J Allergy Clin Immunol 2023; https://doi.org/10.1016/j.jaci.2023.03.029 7. Redwood AJ, Rwandamuriye F, Chopra A, et al. Single-cell transcriptomics reveal polyclonal memory T-cell responses in skin with positive abacavir patch test results. J Allergy Clin Immunol 2019; 144:1413–1416. 8. Bell CC, Faulkner L, Martinsson K, et al. T-cells fromHLA-B-b∗57:01+ human subjects are activated with abacavir through two independent pathways and induce cell death by multiple mechanisms. Chem Res Toxicol 2013; 26:759–766. 9. Monshi MM, Faulkner L, Gibson A, et al. Human leukocyte antigen HLA-B∗57:01-restricted activation of drug-specific T cells provides the immunological basis for flucloxacillin-induced liver injury. Hepatology 2013; 57:727–739. 10. Yaseen FS, Saide K, Kim SH, et al. Promiscuous T-cell responses to drugs and drug-haptens. J Allergy Clin Immunol 2015; 136:474–476. 11. Faulkner L, Gibson A, Sullivan A, et al. Detection of primary T cell responses to drugs and chemicals in HLA-typed volunteers: Implications for the prediction of drug immunogenicity. Toxicol Sci 2016; 154:416–429. 12. Song B, Aoki S, Liu C, et al. An animal model of abacavir-induced HLA-mediated liver injury. Toxicol Sci 2018; 162:713–723. 13. Pan RY, Chu MT, Wang CW, et al. Identification of drug-specific public TCR driving severe cutaneous adverse reactions. Nat Commun 2019; 10:1–13. 14. Deshpande P, Hertzman RJ, Palubinsky AM, et al. Immunopharmacogenomics: mechanisms of HLA-associated drug reactions. Clin Pharmacol Therapeut 2021; 110:607–615. 15. Phillips EJ, Mallal SA. Active suppression rather than ignorance: Tolerance to abacavir-induced HLA-B∗57:01 peptide repertoire alteration. J Clin Invest 2018; 128:2746–2749. 16. Tenorio AR, Zheng Y, Bosch RJ, et al. Soluble markers of inflammation and coagulation but not T-cell activation predict nonaids-defining morbid events during suppressive antiretroviral treatment. J Infect Dis 2014; 210:1248–1259. 17. Younes S-A, Talla A, Pereira Ribeiro S, et al. Cycling CD4+ T cells in HIV-infected immune nonresponders have mitochondrial dysfunction. J Clin Invest 2018; 128:5083–5094. 18▪. Susukida T, Kuwahara S, Song B, et al. Regulation of the immune tolerance system determines the susceptibility to HLA-mediated abacavir-induced skin toxicity. Commun Biol 2021; 4:1–10. 19. Qureshi OS, Zheng Y, Nakamura K, et al. Trans-endocytosis of CD80 and CD86: a molecular basis for the cell-extrinsic function of CTLA-4. Science 2011; 332:600–603. 20. Zhao Y, Lee CK, Lin C-H, et al. PD-L1:CD80 cis-heterodimer triggers the co-stimulatory receptor CD28 while repressing the inhibitory PD-1 and CTLA-4 pathways. Immunity 2019; 51:1059–1073. 21. Sakaguchi S, Yamaguchi T, Nomura T, et al. Regulatory T cells and immune tolerance. Cell 2008; 133:775–787. 22. Kalia V, Penny Laura A, Yuzefpolskiy Y, et al. Quiescence of memory CD8+ T cells is mediated by regulatory T cells through inhibitory receptor CTLA-4. Immunity 2015; 42:1116–1129. 23. Liu Z, Gerner MY, Van Panhuys N, et al. Immune homeostasis enforced by co-localized effector and regulatory T cells. Nature 2015; 528:225–230. 24▪. Nicoletti P, Aithal GP, Chamberlain TC, et al. Drug-induced liver injury due to flucloxacillin: Relevance of multiple human leukocyte antigen alleles. Clin Pharmacol Therapeut 2019; 106:245–253. 25. Wuillemin N, Adam J, Fontana S, et al. HLA haplotype determines hapten or p-i T cell reactivity to flucloxacillin. J Immunol 2013; 190:4956–4964. 26. Nattrass R, Faulkner L, Vocanson M, et al. Activation of flucloxacillin-specific CD8+ T-cells with the potential to promote hepatocyte cytotoxicity in a mouse model. Toxicol Sci 2015; 146:146–156. 27▪. Puig M, Ananthula S, Venna R, et al. Alterations in the HLA-B∗57:01 immunopeptidome by flucloxacillin and immunogenicity of drug-haptenated peptides. Front Immunol 2020; 11:1–13. 28. Sng XYX, Li J, Zareie P, et al. The impact of MHC Class I dose on development and maintenance of the polyclonal naive CD8+ T cell repertoire. J Immunol 2020; 204:3108–3116. 29. Boucherma R, Kridane-Miledi H, Bouziat R, et al. HLA-A∗01:03, HLA-A∗24:02, HLA-B∗08:01, HLA-B ∗27:05, HLA_B∗35:01, HLA-B∗44:02, and HLA-C∗07:01 monochain transgenic/H-2 Class I null mice: novel versatile preclinical models of human T cell responses. J Immunol 2013; 191:583–593. 30. Hammond S, Olsson-Brown A, Grice S, et al. Checkpoint inhibition reduces the threshold for drug-specific T-cell priming and increases the incidence of sulfasalazine hypersensitivity. Toxicol Sci 2022; 186:58–69. 31. Metushi IG, Hayes MA, Uetrecht J. Treatment of PD-1−/− mice with amodiaquine and anti-CTLA4 leads to liver injury similar to idiosyncratic liver injury in patients. Hepatology‘ 2015; 61:1332–1342. 32. Gao Y, Song B, Aoki S, et al. Conjugation of human serum albumin and flucloxacillin provokes specific immune response in HLA-B∗57:01 transgenic mice. Immunol Lett 2022; 249:5–11. 33. Waddington JC, Meng X, Illing PT, et al. Identification of flucloxacillin-haptenated HLA-B∗57:01 ligands: evidence of antigen processing and presentation. Toxicol Sci 2020; 177:454–465. 34. Villani AP, Rozieres A, Bensaid B, et al. Massive clonal expansion of polycytotoxic skin and blood CD8(+) T cells in patients with toxic epidermal necrolysis. Sci Adv 2021; 7:1–17. 35. Chung WH, Pan RY, Chu MT, et al. Oxypurinol-specific T cells possess preferential TCR clonotypes and express granulysin in allopurinol-induced severe cutaneous adverse reactions. J Invest Dermatol 2015; 135:2237–2248. 36. Mifsud NA, Illing PT, Lai JW, et al. Carbamazepine induces focused T cell responses in resolved Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis cases but does not perturb the immunopeptidome for T cell recognition. Front Immunol 2021; 12:1–20. 37. Falta MT, Crawford JC, Tinega AN, et al. Beryllium-specific CD4+ T cells induced by chemokine neoantigens perpetuate inflammation. J Clin Invest 2021; 131:1–17. 38. Vivian JP, Duncan RC, Berry R, et al. Killer cell immunoglobulin-like receptor 3DL1-mediated recognition of human leukocyte antigen B. Nature 2011; 479:401–405.
Comments (0)