Genomic variants association with selective hypersensitivity reactions to amoxicillin and clavulanic acid

Betalactams (BLs) are the most frequent triggers of immediate drug hypersensitivity reactions (IDHRs), affecting up to 10% of general population, especially the amoxicillin (AX) and clavulanic acid (CLV) combination.1 Most patients are only sensitised to AX and some only to CLV, probably due to structural differences,1 but the reason why a person treated with AX-CLV reacts to one drug but not to the other remains unknown. This fact could be related to genetic differences, since associations between single nucleotide polymorphisms (SNPs) in genes related to the antigenic presentation, or effector mediators have been previously demonstrated.2 We aimed to analyze the exonic regions of the genome using RNA sequencing to find variants on genes related with the immune system possibly explaining mechanistic differences between AX and CLV allergic patients and healthy subjects. The identification of these variants may be a first step in the search for potential genetic biomarkers. Two types of variants, SNPs and insertion-deletions (INDELs), were detected applying the gatk bioinformatic workflow to 42 RNA-seq libraries generated from 17 subjects: five healthy individuals, seven patients with confirmed IDHR to AX (group A), and five with confirmed IDHR to CLV (group B) (Table S1). To increase the accuracy of the association analysis, we included exome sequencing samples from 107 additional healthy control individuals from the Iberian population stored at the 1000 genomes repository (Appendix S1). Association analysis found 159 exonic regions (Table 1) significantly (adjusted p-value < .01) associated with DHRs triggered by either AX or CLV (Figure 1A-left). These regions overlapped with 242 different genes and were affected by 822 non-synonymous rare and novel exonic variants. These genes are involved, as shown by reactome pathways enrichment (Figure 1A-right), in IL-1 signaling, neutrophil degranulation, antigen processing-cross presentation, regulation of TNFR1 signaling, TNFR2 noncanonical NF-kB pathway, downstream TCR and BCR signaling pathways, and programmed cell death. Some of them (Table S2), such as HLA-DQA2, ICOSLG, BIRC3, REL, and PSDM8, have been previously associated with allergy, mainly with Th2 diseases such as asthma, allergic airway and eosinophilic inflammation, B cells survival and proliferation, and antigen processing, although, to our knowledge, never with IDHRs to drugs.3, 4 Seventy-eight rare or novel variants in 48 regions overlapping 47 different genes were significantly associated only with patients with IDHRs to AX (Figure 1B-left and Table S2), including ENDOG and immunoproteasome components PSMB7, PSMC1, and PSDM2 (Table 1). Enrichment analysis (Figure 1B-right) found 151 enriched reactome pathways, highlighting: neutrophil degranulation, activation of NF-kappaB in B cells, FCεRI mediated NF-kB activation, downstream TCR signaling, signaling B Cell Receptor, and IL-1 signaling. Interestingly, no regions were found to be significantly associated only with CLV allergic patients, which could be explained by the low number of samples. However, rising up significance threshold to adjusted p-value < .05, 41 regions, comprising 85 novel or rare non-synonymous variants, could be significantly associated with CLV-induced hypersensitivity (Figure 1C-left). These regions overlapped with 47 different genes such as XIAP and PSMD8, which enriched REACT terms related to PTEN regulation (Figure 1C-right), linked to chronic asthma and other immuno-related diseases.5 In summary, there are rare variants found in our patients that could be related to the development of AX or CLV-induced IDHRs. Interestingly, some of these rare variants could be associated with selective response to AX, such as the proapoptotic endonuclease ENDOG and neutrophil degranulation pathway, previously associated with drug-induced anaphylaxis,6 or to CLV, like immunoproteasome component PSMD8. Differences could be related to nature and metabolic characteristics of each drug; however, more research is needed. These findings represent a first step to identify potential genetic biomarkers to predict the susceptibility of suffering AX and CLV IDHRs. CM, MJT and TDF designed the study and coordinated the work of the rest of the authors. MS and GB recruited the study individuals and managed the clinical procedures and obtained clinical data. RN, RFS and PDE performed the in vitro experiments and analysed the experimental results. TDF, CM, RN and RFS wrote the manuscript and figures, which was reviewed by the rest of the authors. We thank Claudia Corazza for her invaluable English language support; all sequencing experiments have been carried out thanks to the Supercomputing and Bioinnovation Center of the University of Malaga. This work has been supported by Institute of Health 'Carlos III' (ISCIII) of the Ministry of Economy and Competitiveness (MINECO), (grants co-funded by European Regional Development Fund: PI15/01206, PI17/01237, PI18/00095, PI20/01734, PI21/0329, RETICS ARADYAL RD16/0006/0001, RICORS REI, RD21/0002/0008); Andalusian Regional Ministry of Health (grants PI-0241-2016, PE-0172-2018, PI-0127-2020). GB is a Clinical Investigator (B-0007-2022) research contract by Andalusian Regional Ministry Health. PDE holds a Sara Borrell research contract by Institute of Health 'Carlos III' (CD22/00112). CM holds a 'Nicolas Monardes' research contract by Andalusian Regional Ministry Health (RC-0004-2021). Funding for open access charge: Univesidad de Malaga/CBUA. The authors declare that they have no conflict of interest. Research data are not shared. Appendix S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.