Adverse drug reactions are a major cause of morbidity and mortality. Of the great diversity of drugs involved in hypersensitivity drug reactions, the most frequent are non-steroidal anti-inflammatory drugs followed by β-lactam antibiotics. The redox status regulates the level of reactive oxygen and nitrogen species (RONS). RONS interplay and modulate the action of diverse biomolecules, such as inflammatory mediators and drugs.
- redox
- hypersensitivity drug reaction
- non-steroidal anti-inflammatory drugs
- β-lactam antibiotics and SNPs
1. Introduction
According to the World Health Organization (WHO), adverse drug reactions (ADRs) are defined as noxious and unintended reactions to a drug that is administered in standard doses by the proper route for prophylaxis, diagnosis, or treatment, or for the modification of physiological function [1]. For the last 30 years, ADRs have been the leading cause of both morbidity and mortality in the emergency departments and hence pose a significant burden on health care resources. Furthermore, ADRs represent one of the four leading causes of death in the developed world [2,3]. Approximately, 6–10% of ADRs belong to the Type B reactions, which include drug hypersensitivity reactions (DHRs). These are unpredictable, not related to the pharmacological action of the drug, and can be very severe and life-threatening [4,5]. Concerning DHRs, the most frequent drugs involved are non-steroidal anti-inflammatory drugs (NSAIDs), which account for 20–25% of patients evaluated in allergy units. These causative drugs are followed in frequency by β-lactam antibiotics [4].
From a mechanistic point of view, the European Academy of Allergy, Asthma, and Clinical Immunology (EAACI) has classified the NSAIDs hypersensitivity reactions into two main groups: Single NSAID induced hypersensitivity reactions, also known as selective reactions (SR), are specific immunological reactions, IgE or T-cell mediated. These reactions occur with a single NSAID whereas there is tolerance to other NSAIDs. The second group of NSAIDs hypersensitivity reactions involve several NSAIDs not structurally related and are mediated through a pharmacological mechanism linked to cyclooxygenase inhibition, although, the underlying mechanism is not fully understood [6,7,8,9]. The clinical manifestation may vary from mild to severe reactions such as life-threatening anaphylaxis [10]. Hence, the heterogenous symptoms observed and the severity of certain reactions hamper the diagnosis of NSAIDs hypersensitivity reactions [7].
Regarding hypersensitivity reactions triggered by β-lactam antibiotics, they are induced by specific immunological mechanisms and are mediated through immediate and non-immediate reactions [11,12]. Immediate reactions usually appear within the first hour and are mediated by specific IgE antibodies. The clinical manifestations include urticaria, angioedema, rhinitis, bronchospasm, and/or anaphylaxis [13]. These reactions constitute the most frequent DHRs to β-lactam antibiotics and the most common cause of anaphylaxis [14]. On the contrary, the non-immediate reactions constitute a group of entities that occur within a period of 24–48 h after the intake of the drug, although the time may be as short as 1 to 2 h. They are mediated through T-cells. The most common clinical entities are benign, like urticaria; nevertheless, severe reactions such as DRESS and TEN syndromes can also occur [15]. Although the pattern of β-lactam antibiotics consumption has been altered over the past years, amoxicillin is still the drug that most frequently induces DHRs [11]. The details about the interaction between β-lactam antibiotics and immune cell receptors have not yet been fully determined.
Furthermore, several factors may contribute to the occurrence of DHRs. In this regard, it has been pointed out that the reactive metabolites generated by the bioactivation of a particular drug and its binding to cellular macromolecules might be related to DHR development [16,17]. Thus, β-lactam antibiotics form adducts with high molecular weight proteins through a covalent bond generating a hapten-carrier formation for immune activation [18]. Also, it is demonstrated that sulfamethoxazole metabolites such as the nitroso sulfamethoxazole can form adducts with protein cysteine residues, and induce DHR reactions [19,20]. Furthermore, it has been documented that both a higher amount of reactive metabolite bound to cellular proteins and a higher total drug daily dose is correlated with an increased risk of DHR [21]. Supporting the concept that additional factors might contribute to DHR development, the danger hypothesis proposes that immune response to drugs is also determined by cell damage. Thus, injured cells might release alarm signals, such as reactive oxygen intermediates that might be capable of modifying the immune response. In this regard, the oxidative stress status may contribute as a key factor in the development of DHRs [16,22]. Genetic association studies have documented several findings supporting this hypothesis [23,24]. Thus, a previous report demonstrated an association between the risk of developing DILI in Spanish individuals carrying the double GSTT1-GSTM1 null genotypes [25]. Also, the haplotype conformed by the rs10735393, rs4964287, and rs4595619 single-nucleotide polymorphisms (SNPs) in the thioredoxin reductase 1 gene was found associated with a cohort of Asian patients with DILI [26]. Despite these previous findings, the role of redox status in the development of DHRs is scarcely studied.
The evidence raised from recent studies, particularly those from collaborative studies in the ARADyAL network suggests that redox status might play a key role in the development or the clinical presentation of DHRs [27].
2. Redox Status and Hypersensitivity Drug Reactions
Chemically, reactive oxygen species (ROS) and reactive nitrogen species (RNS) are a family of reactive species containing partially reduced oxygen and derived from oxygen or nitric oxide. These species include superoxide anion, hydroxyl radical, and peroxynitrite. These molecules are highly reactive and may modulate cellular signaling through the interaction with diverse biomolecules [41,42]. These biomolecules may suffer oxidative modifications which may undergo conformational changes or even severe denaturation. Oxidative stress is usually defined as increased production and/or decreased scavenging or metabolism of ROS [43]. The intracellular redox status determines the function of ROS and RNS. Thus, excessive production of these reactive species evokes cytotoxic effects that may lead to apoptotic cell death, whereas a lower level of these may mediate the regulation of certain cellular processes such as inflammation, protein expression, or posttranslational modifications. Moreover, ROS can play a role as second messengers in signaling pathways [41,44,45,46]. The mechanisms to protect biomolecules from oxidative stress are widely conserved among mammalians and constitute a complex composed of antioxidant enzymes, small antioxidant molecules, and molecules able to scavenge transition metal ions: Glutathione S-transferases (GSTM1, GSTP, GSTT1). These enzymes catalyze the nucleophilic attack by reduced glutathione (GSH) on nonpolar compounds that contain an electrophilic atom. Moreover, the glutathione peroxidase (GPX) family catalyzes the reduction of a variety of hydroperoxides using GSH. The thioredoxin reductase family (TXNRD1 and TXNRD2) degrades hydrogen peroxide and protects cells from cytotoxicity. The superoxide dismutase enzymes (SOD1, SOD2, and SOD3) decompose the free radical superoxide by converting it to hydrogen peroxide, which may, in turn, be destroyed by catalase (CAT). The aldo-keto reductase (AKR) family of enzymes which detoxifies reactive aldehydes to alcohols and peroxiredoxin system (PRDX1, PRDX2, PRDX3, PRDX4, PRDX5, PRDX6) have a crucial role as a scavenger of hydrogen peroxide [47,48,49,50].
3. Genetics of Redox System and Their Relationship with Hypersensitivity Drug Reactions
3.1. Glutathione Transferases
GSTs are phase II multifunctional enzymes that play an important role in cellular protection by detoxifying numerous electrophilic compounds as well as regulating cellular redox homeostasis [50]. GSTs are categorized into cytosolic, mitochondrial, and microsomal families. There are seven classes within cytosolic GSTs (GSTA, GSTM, GSTP, GSTT, GSTO, GSTZ, and GSTS) that differ in their chemical, physical, and structural properties [50]. The effect of polymorphisms in the genes encoding these cytosolic enzymes, as well as that of gene duplications and deletions, has been extensively studied and their functional relevance in the context of drug detoxication has been highlighted [25,76,77,78,79]. Here, we will focus on their impact on DHRs development.
Glutathione S-transferase alpha (GSTA) class genes, comprised of 5 genes, is placed in a cluster on chromosome 6 [80]. Certain polymorphisms have been identified in the coding region of GSTA1, GSTA2, and GSTA3, although there is no clear evidence of their influence on the GSTAs activity [77]. Moreover, further studies have shown that GSTA1 haplotypes comprised of SNPs in the 5’ non-coding region may affect GSTA1 expression levels [77]. Nevertheless, there are no reports of their impact on DHRs susceptibility.
The GST Mu class (GSTM) is located on chromosome 1. GSTM class are dimeric proteins with a tissue specificity expression [77]. The GSTM1 locus is highly polymorphic, being the GSTM1 null allele the most common allele in most populations studied so far. Thus, the homozygous deletion of the GSTM1 gene leads to the absence of enzymatic activity [81]. Evidence of an association between GSTM1 null genotypes and the development of skin eruptions caused by NSAIDs and β-lactam antibiotics has been reported [28], although this result was not confirmed in a recent study, possibly due to differences in the frequencies for the genetic variations between both cohorts [29]. Moreover, GSTM1 null genotype has been related to the most severe and rare form of cutaneous drug reactions, Steven-Johnson syndrome, and toxic epidermal necrolysis (SJS/TEN) in HIV-positive patients receiving nevirapine [30]. Individuals carrying the GSTM1 null genotype might show a low efficiency to detoxify the 12-sulfoxyl-nevirapine metabolite [30].
Drug-induced liver injury (DILI) is one of the most common DHRs and a recent updated meta-analysis has confirmed an association between GSTM1 null genotypes and susceptibility to antituberculosis-DILI in South Asian individuals [31]. GSTM enzymes exert an essential role in detoxifying ROS and xenobiotics and individuals holding GSTM1 null genotypes have shown an absence of GSTM activity [81]. Therefore, considering the interactions between the redox status and DHRs, it seems plausible that the absence of GSTM activity caused by GSTM1 null genotypes may trigger an imbalance in redox status that may have an impact on the susceptibility to develop a DHR. Apart from the aforementioned GSTM1 null genotype, the GSTM3*C haplotype has been functionally characterized, resulting in a gain of function, and several SNPs have been identified in the promoter region, but none have been shown to alter gene expression [77,82]. The effect of these SNPs in susceptibility to DHRs remains to be elucidated.
The GST Pi class (GSTP) is formed by a unique class gene, GSTP1 which is placed on chromosome 11. Two nonsynonymous variants, rs1695 and rs1138272 have been extensively characterized and have shown differences in the GSTP1 activity in the metabolism of certain drugs [77,83]. The effect of the variant GSTP1 rs1695 in DHRs is controversial. Thus, GSTP1 rs1695 mutant homozygous genotype confers risk for skin drug eruption induced by NSAIDs, β-lactam antibiotics, and other drugs in a Jewish population [29] whereas this association was not found in a cohort of Turkish patients [28]. The GSTP1 rs1695 causes the amino acid substitution Ile105Val, the mutant genotype has been associated with decreased activity in the GSH conjugation of the anticancer drug thiotepa [84]. Nevertheless, regarding NSAIDs and β lactams, this finding is still inconclusive, as there a lack of studies. In addition to these genetic associations, the characterization of GSTP1 oligomerization, which modulates its activity, by cyclopentanone prostaglandins and certain drugs exposes a direct crosstalk between inflammatory mediators and cellular redox homeostasis [85]. Therefore, these functional and genetic evidences support the need for additional studies in larger patient cohorts which might shed light on the role of this variant in DHRs.
The GST Theta class (GSTT) is comprised of two functional genes (GSTT1 and GSTT2) which are placed on chromosome 22 [86]. Several SNPs in GSTT1 have been functionally characterized. Thus, the mutant variants rs11550605, rs2266633, and rs2234953 might cause instability and low activity of GSTT1 [87,88]. The majority of GSTT1 variants occur at a relatively rare frequency and are population-specific [89]. However, the gene deletion GSTT1*0 occurs frequently in most populations [77,89]. The GSTT1*0 variant has been associated with an increased risk of DHRs triggered by a variety of drugs, including NSAIDs and antibiotics in a cohort of Turkish patients [28]. Although this association was not confirmed in a Jewish population [29]. A certain number of studies have shown controversial results about the association between GSTT1 null variant and DILI [31,33,34]. Nonetheless, meta-analysis studies did not confirm that association [31,90]. Lucena et al. demonstrated an association of GSTM1-T1 double-null genotype and Spanish patients with co-amoxiclav hepatotoxicity [25]. Nevertheless, this finding was not observed in a larger cohort of European co-amoxiclav hepatotoxicity patients [32]. Therefore, the association found between GST variants and DILI patients seems to be more related to anti-tuberculosis drugs or NSAIDs than to co-amoxiclav.
The GSTT2 locus includes an inverted repeat of GSTT2 named GSTT2b which is commonly deleted and occurs in linkage disequilibrium with the GSTT1 null allele [91]. No association between GSTT2 variants and susceptibility to DHRs has been described. In addition, there is a lack of evidence of an association between GSTS, GSTZ, or GSTO variants and the risk of developing DHRs [77].
3.2. Glutathione Peroxidases
GPX family is composed of eight isoenzymes in mammals. Their expression varies across different cells and tissues. These enzymes catalyze the reduction of hydrogen peroxide and lipid hydroperoxides using GSH as a reducing agent [48,92]. GPX1 is ubiquitous and its action contributes to diminish DNA damage and inhibit the synthesis of inflammatory mediators whereas GPX2 and GPX3 are overexpressed, mainly, in diverse cancer types [92,93].
The GPX1 gene is located on chromosome 3p21.3 [93]. The non-synonymous GPX1 rs1050450 (Pro200Leu) variant has been extensively studied and the mutant allele has been identified as a risk factor for lung cancer, cardiovascular or aging brain [94,95]. In addition, the GPX1 rs1050450 mutant variant has been related to decreased GPX1 activity in cancer patients [96,97,98]. Concerning DHR, Lucena et al. described a significant association between the rs1050450 mutant homozygous genotype and the risk of developing cholestatic injury triggered by various drugs; including NSAIDs and β-lactam antibiotics [35]. This finding was not replicated in another cohort of European co-amoxiclav-related DILI patients [32]. Whether the impaired activity of GPX1 rs1050450 mutant might enhance oxidative stress and consequently contribute to activate the inflammation pathway or generate drug reactive metabolites needs to be extensively studied.
3.3. Thioredoxin Reductases
Thioredoxin reductases (TrxR) are a family of NADPH-dependent enzymes that make part of the thioredoxin system. This system reduces ROS and represents the major H2O2 scavenger in mitochondria [99]. Three isoforms have been characterized in mammalians. TrxR1 is mainly located in the cytosol, while TrxR2 is in the mitochondria and TrxR 3 is predominantly expressed in maturing spermatids [100,101]. TrxR1 and TrxR2 are encoded by TXNRD1 and TXNRD2 genes which are placed on chromosomes 12 and 22; respectively [102].
A few studies have explored the role of TXNRD1 genetic variants and the DHRs. These have been focused on DILI. Hence, Kwon et al. analyzed the SNPs TXNRD1 rs10735393, rs4964287, and rs4595619, rs10861201, rs11111997, rs4246270, and rs4246271. The analysis of isolated SNPs did not reveal any association. However, an association was identified between the TTA haplotype, composed of TXNRD1 rs10735393, rs4964287, and rs4595619, and DILI in a cohort of Korean patients [26]. In addition, the TCAGCC haplotype conformed by rs10735393, rs4964287, rs4595619, rs4246270, rs4246271, and rs11611385 was associated with hepatotoxicity induced by anti-tuberculosis drugs in Chinese female non-smokers [36]. However, these associations were not confirmed in a larger cohort of Chinese patients with DILI induced by anti-tuberculosis drugs [37]. The SNPs that conformed these targeted TXNRD1 haplotypes are mainly located in intronic regions and there are not functional studies related to these SNPs. Therefore, studies aimed to determine the impact of these variants in the TXNRD1 capacity for reducing thioredoxins and their role in DHRs underlying mechanism are warranted.
3.4. Superoxide Dismutase
SODs are a family of enzymes responsible for the detoxication of superoxide anion radicals acting as an important defense mechanism against cellular damage [103]. Three forms have been described (SOD1, SOD2, and SOD3) and they are found in the cytoplasm, mitochondria, and extracellular space; respectively [103].
SOD1 is localized on chromosome 21 and contains five exons. Several rare variants (rs202446, rs202447, rs4816405, and rs2070424) have been associated with an increased SOD1 expression [104] although their functional significance is still unclear [105]. Among these variants, rs2070424 has been associated with DILI induced by anti-tuberculosis drugs in Korean patients [38]. Conversely, SOD1 rs4816407 and rs1041740 were not associated in a Chinese cohort of patients with DILI induced by anti-tuberculosis drugs [39]. The essential SOD activity to protect against redox imbalance has been shown altered in patients with a non-immediate DHR [74]. This finding supports the link between the antioxidant scavenger enzymes and DHRs. Therefore, additional functional studies to determine whether this observed impaired activity is driven by SNPs in SOD1 are required.
Concerning SOD2, which is located on chromosome 6, the nonsynonymous SNP rs4880 (Val16Ala) is the most studied. The mutant variant is associated with an augment of the SOD2 import into the mitochondrial matrix and reduced catalytic activity [106,107]. The SOD2 homozygous genotype has been associated with the risk of developing DILI in Taiwanese and Spanish cohorts [35,40]. Both cohorts included a wide range of drugs inducing this pathology. Nevertheless, this association was not confirmed in Korean or Chinese patients with DILI induced by anti-tuberculosis drugs [37,38,39] nor European patients with co-amoxiclav DILI [32]. The discrepancies between these results might be explained due to ethnic differences in allele frequencies and the drugs involved in DILI development. These findings are keeping in line with the reduced catalytic activity associated with mutant homozygous SOD2 rs4880. The impaired activity associated with this genotype is related to SOD2 transport efficiency into the mitochondria [106]. The mitochondria are the main source of ROS and SOD2 exerts a primordial role by reducing superoxide anion radical and protecting cells from redox imbalance [48]. Thereby, an altered SOD2 activity may evoke a redox imbalance which contributes to an alteration in inflammatory signaling pathways and drug reactive metabolites generation, and consequently the occurrence of DHRs.
The SOD3 gene is placed on chromosome 4. The missense variants SOD3 rs2536512 and rs1799895 and the following variants in non-coding region SOD3 rs699473, rs2855262, and rs8192290, have been analyzed to investigate the possible association with DILI induced by anti-tuberculosis drugs. However, none were found to be associated with the risk [38,39]. Whether SOD3 polymorphisms have a relevant role in the development of DHRs needs to be investigated more extensively.
3.5. Catalase
CAT plays an important role by discomposing H2O2, mainly at high H2O2 concentrations [108]. CAT has been localized on chromosome 11 and its genetic variations have been the subject of a previous review [109]. Hence, it has been reported that the CAT rs1001179 SNP, which is placed in the promoter region, may modulate the specificity of different transcriptional factor binding. In addition, individuals carrying the mutant T allele showed higher CAT expression levels in blood than homozygous for the C allele [110]. However, the results about the influence of this variant on CAT activity are controversial [109]. In addition, carriers of the variant CAT rs769214 have demonstrated a higher CAT activity in basal conditions [109,111]. Also, a certain number of CAT SNPs in the exon regions have been described [109]. A plethora of studies have associated CAT polymorphisms with a wide range of pathologies [109]. Nevertheless, there is a lack of studies involving CAT variants and DHRs.
3.6. Aldo-Keto Reductase
AKR protein superfamily is comprised of 15 different families which also contain multiple subfamilies [112]. AKRs catalyze the reduction of carbonyl groups to yield primary and secondary alcohols utilizing NADPH as a cofactor on a wide range of substrates [112]. AKR1 is the largest family and is subdivided into different subgroups. AKR1B1 catalyzes the conversion of glucose to sorbitol leading to redox imbalances and resulting in tissue injury which is linked to diabetic complications [113]. Moreover, AKR1B1 exerts a role in inflammatory processes. Hence, AKR1B1 exerts a role in inflammation modulation by synthesizing prostaglandin PGF2α which can regulate PGE2 production [59]. In turn, several prostaglandins such as PGA1 and PGE2 are capable to bind and inhibit AKR1B1 [114]. Furthermore, structural analysis of sulindac, a precursor of the NSAID sulindac sulfide, has revealed its capacity to inhibit AKR1B1 and AKR1B10 isoforms [115,116]. These interactions open new possibilities to explore the role of the AKR1B subfamily and the cellular redox system in the imbalance in prostaglandins and lipid mediators detected in NSAIDs hypersensitivity patients. AKR1B1 is placed on chromosome 7 and it is highly polymorphic. Most of the association studies involving AKR1B1 SNPs are related to the development of diabetes [117]. In addition, several polymorphisms have been associated with diverse pathologies [118]. The AKR1B1 rs2229542 nonsynonymous variant has been associated with asthma and allergic rhinitis. It has been suggested that its occurrence might modify the AKR1B1 protein stability and consequently contribute to alterations in its function [119]. Moreover, the effect of AKR1B1 rs5054, rs5056, rs5057, rs2229542, rs5061, and rs5062 variants on the activity over anticancer drugs doxorubicin and daunorubicin has been evaluated. However, none of these variants caused significant activity changes [120]. Concerning to AKR1C subfamily, its different isoforms metabolize multiple drugs such as NSAIDs, antipsychotics, and anticancer agents [121]. Interestingly, certain NSAIDs, such as flufenamic acid and indomethacin, have the ability to inhibit AKR1C3 activity [122]. In addition, acetylsalicylic acid and its metabolite salicylic acid were found to inhibit the isoform AKR1C1 activity [123]. Besides, different enzymes in the AKR1C subfamily may be involved in the nabumetone metabolism [124]. These findings reveal crosstalk between NSAIDs and redox status, which might contribute to the underlying mechanism of DHRs triggered by NSAIDs. AKR1C1-4 genes are located on chromosome 10 [125]. Several polymorphisms have been described in these genes being associated with cancer or mental disorders [126,127,128] or with a decrease in enzyme activity [120]. However, there is a lack of data about the role of these variants on DHRs development.
3.7. Peroxiredoxin System
Peroxiredoxins (PRDXs) constitute a family of small non-seleno peroxidases. Six isoforms have been described in humans. They all catalyze peroxide reduction of H2O2, organic hydroperoxides, and peroxynitrite using the thioredoxin system as the electron donor [129,130]. There is scarce information on the involvement of PRDXs variants in the development of DHRs. Nevertheless, evidence for the contribution of several polymorphisms in these genes to drug metabolism variability and prognosis in breast cancer patients exist. Thus, the PRDX4 rs518329 mutant homozygous genotype has been associated with decreased clearance of the anti-cancer chemotherapy drug docetaxel [131]. As the rs518329 is an intronic variant, the functional basis for the observed effect remains to be elucidated. Besides, the PRDX6 rs4916362 which is located at a transcription factor binding site, and PRDX6 rs7314 placed at a miRNA binding site, may modify the mortality risk in breast cancer patients [132]. Whether PRDX variants might exert a role in the susceptibility of DHRs need to be further studied.
This entry is adapted from the peer-reviewed paper 10.3390/antiox10020294