Introduction
NLRP3 is an integral part of the NLRP3 inflammasome, a large cytosolic protein complex which acts as an intracellular innate immune sensor. NLRP3 inflammasome assembly occurs following the detection of a wide range of harmful stimuli and mediates caspase-1 autoproteolysis and activation, leading to caspase-1-mediated cleavage of the pro-inflammatory cytokines, IL-1β and IL-18, into their active forms, and cleavage of gasdermin-D, which subsequently forms membrane pores and induces pyroptotic cell death. Gain-of-function mutations in
NLRP3 are associated with a series of rare autoinflammatory diseases (AIDs), collectively known as
NLRP3-associated autoinflammatory diseases (
NLRP3-AIDs), previously referred to as cryopyrin-associated periodic syndromes (CAPS) [
1].
NLRP3-AIDs are typified by recurrent episodes of fever, urticaria, arthralgia, and chronic inflammation which can lead to long-term damage including sensorineural hearing loss and amyloidosis [
2].
The NLRP3 protein comprises three domains: the pyrin domain (PYD), the central NACHT domain, and the leucine-rich repeat (LRR) domain. Most of the reported
NLRP3-AID-related
NLRP3 mutations are located in or around the central NACHT domain [
3], whereas the C terminus LRR domain, the function of which is unclear due to evidence that it is dispensable for NLRP3 inflammasome activation [
4], is an uncommon site for pathogenic mutations [
3]. One of the few validated LRR domain mutations is the missense substitution c.2759G > A, coding for p.Arg920Gln (p.R920Q) in the LRR domain. This mutation was previously reported in two unrelated North American families, where it was previously known as p.Arg918Gln (p.R918Q) and primarily caused autosomal-dominant sensorineural hearing loss [
5] which is common in
NLRP3-AIDs [
6]. However, the symptoms observed in these families were varied; one family reported hearing loss with an age of onset in the second to fourth decade of life, with no additional physical signs or symptoms of
NLRP3-AID. In contrast, the second family experienced hearing loss in the first 10 years which was accompanied by autoinflammatory signs and symptoms, including oral ulcers, without serologic signs of inflammation, contributing to an atypical
NLRP3-AID phenotype [
5]. Hearing loss and autoinflammatory symptoms were generally reversed or improved following treatment with the interleukin-1 (IL-1) receptor antagonist anakinra, which has been historically used in the treatment of
NLRP3-AIDs [
7‐
9]. Here, we present another case of a patient with the
NLRP3-p.R920Q mutation. This patient’s symptoms spanned the spectrum of those seen in this previous study; she experienced mild hearing loss (only evident on audiogram) with some evidence of systemic inflammation, but her main complaint was persistent oropharyngeal ulcers. She showed only a partial response to anakinra, instead responding well to steroids and the phosphodiesterase type-4 (PDE4) inhibitor apremilast.
In this report, we investigated this atypical patient phenotype, the potential pathogenic mechanisms of this LRR domain mutation and signalling pathways which contribute to the disease. Altogether, our study provides a detailed exploration of this rare pathogenic mutation.
Methods
Human Subjects
Work using human samples was approved by the Health Research Authority (study number 18/YH/0070). Informed written consent was obtained from participants prior to sample collection. Age- and sex-matched HCs were recruited from St James’s University Hospital, Leeds, UK.
Samples
Patient and HC samples were collected in VACUETTE® tubes (Greiner Bio-One), coated with EDTA anticoagulant or serum clot activator gel for whole blood or serum, respectively. Serum samples were collected by allowing the sample to clot for 1 h, followed by centrifugation at 1000 × g for 15 min.
Sequencing
The coding sequence was captured from genomic DNA using the SureSelectXT target enrichment kit with All Exon v5 capture library (Agilent). Sequencing was performed on a HiSeq 3000 (Illumina) using a 2 × 150-bp paired-end sequencing protocol, and analysis restricted to the autoinflammatory/autoimmune gene panel (supplementary methods).
Cell Culture and Stimulation
HEK293T cells (ATCC) were cultured according to the manufacturer’s specifications. Monocytes were isolated from whole blood using Lymphoprep density gradient media (StemCell), as outlined in the supplementary methods. The NLRP3 inflammasome was stimulated with 10 ng/mL LPS (Ultrapure Escherichia coli K12, Invivogen) for 4 h, with 5 mM ATP (Invivogen) added for the last 30 min. After stimulation, cells were lysed with TRIzol (Invitrogen).
Cytokine Quantification by ELISA
Release of IL-1β, IL-18, TNF, IL-6, or IL-10 in patient sera or media from cultured cells was detected by enzyme-linked immunosorbent assay (ELISA). Nunc MaxiSorp 96-well plates (Thermo Fisher Scientific) and commercially available ELISA kits were used (Thermo Fisher Scientific, product codes listed in supplementary methods) according to the manufacturer’s specifications.
ASC Speck Quantification by Flow Cytometry
A 2 μL of PE-conjugated ASC antibody (HASC-71 clone, BioLegend) was added to 100 μL of cell culture media, and the mixture incubated in FACS collection tubes on a shaker for 1 h. Size gating was carried out with Megamix-Plus beads (Biocytex) according to the manufacturer’s specifications and was used to threshold out readings below 0.9 μm (representative flow cytometry plots shown in supplementary methods). Samples were run and analysed on a CytoFLEX-S (Beckman Coulter).
Gene Expression Analysis
RNA isolation was carried out using TRIzol Reagent and Phasemaker tubes (Thermo Fisher Scientific) according to the manufacturer’s specifications. One hundred nanograms of RNA was converted to cDNA using the SuperScript IV one-step RT-PCR system (Invitrogen) according to the manufacturer’s specifications. Gene expression was analysed using TaqMan probes (Thermo Fisher Scientific); the supplementary methods detail the probes and cycle parameters used. Data were expressed as relative expression compared to the housekeeping genes, ACTB and HPRT1.
RNA Preparation and RNAseq
RNA was isolated using TRIzol (Invitrogen) and Phasemaker tubes (Thermo Fisher Scientific) according to the manufacturer’s instructions. RNA quality was assessed using an Agilent 4200 TapeStation (Agilent Technologies). Library preparation and RNA-seq was carried out on a high-throughput Illumina platform and paired-end reads generated (Novogene (UK) Company Limited); further information is provided in the supplementary methods.
In Silico Investigation
The cryo-EM structure of NLRP3 in complex with NEK7 (PDB 6NPY)[
10] was used for structural investigations. Interaction analysis was conducted using PISA [
11] and structure representations, including electrostatic surface representations, produced using Pymol version 2.3.3 [
12].
Immunoprecipitation Studies
Cells were transfected with 10 μg of plasmids for NLRP3-GFP and/or NEK7-His (details in supplementary methods) using Opti-MEM and Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer’s specifications. Cell lysates were harvested and used for immunoprecipitation, then analysed by SDS-PAGE and Western blot.
Statistical Analysis
Data were presented as the mean ± standard error of the mean (SEM). Analyses were performed using GraphPad Prism 9. A two-way ANOVA test with Tukey’s post-hoc analysis was performed when calculating variance between samples (p values *p < 0.05, **p < 0.01, ***p < 0.001). p < 0.05 was considered significant.
Discussion
The NLRP3 inflammasome is the most widely studied of the inflammasomes, with over 150 currently validated pathogenic mutations [
23]. Among these mutations, very few are in the LRR domain, and as a result, our knowledge of
NLRP3-AIDs caused by such mutations is incomplete. We show here that the
NLRP3-p.R920Q mutation causes enhanced interactions between NLRP3 and NEK7, leading to increased NLRP3 inflammasome activity and affecting macrophage cytokine signalling.
Mutations in the LRR domain are associated with atypical or milder
NLRP3-AIDs phenotypes [
37], as seen here. Our case report adds to the small number of patients known to carry the
NLRP3-p.R920Q mutation [
5]. In these patients, the primary symptom was sensorineural hearing loss, whereas our patient did not report hearing difficulties. Of the two families already described, one reported no further
NLRP3-AID symptoms, whereas members of the second family reported autoinflammatory signs and symptoms including oral ulcers but without serologic signs of inflammation. This disparity in symptoms shows that the p.R920Q mutation can be associated with different disease severity in different families, a phenomenon observed with other
NLRP3 mutations [
38]. However, arguably, the largest difference between our patient and previously reported families was response to treatment. Anakinra was effective in treating hearing loss and improving the symptoms of autoinflammation in both published families [
5]; in contrast, our patient only partially responded to anakinra, which improved her fevers and partly helped with headaches and arthralgia but did not improve mucosal ulceration. Oral ulcers were improved; however, following addition of the PDE4 inhibitor, apremilast, to her treatment regimen, suggesting that signalling pathways other than those directly mediated by the NLRP3 inflammasome were dysregulated. RNA-seq showed that the NF-кB signalling pathway was enriched in the patient’s M1 macrophages, and since apremilast acts via a NF-кB-dependent mechanism to reduce levels of inflammatory cytokines such as TNF [
39], this may provide an explanation for the efficacy of this treatment. The IL-17 signalling pathway was also enriched in M1 macrophages, which is notable because NLRP3 inflammasome activation and IL-1β release have been linked to enhanced inflammatory Th17 cell responses in diseases including ankylosing spondylitis [
40], inflammatory skin diseases including hidradenitis suppurativa [
41,
42], rheumatoid arthritis [
43] and obliterative bronchiolitis [
44]. A MWS-related
NLRP3 mutation also caused spontaneous skin inflammation in mice due to increased IL-1β production and consequent Th17 cell predominance [
45]. As such, enrichment of IL-17 signalling in our patient’s M1 macrophages may contribute to the inflammatory phenotype. It may also partly explain the efficacy of apremilast, as this agent reduces inflammatory cytokine release and increases anti-inflammatory cytokine production in response to IL-17 stimulation in psoriasis [
46‐
48], and IL-17A is a biomarker of apremilast efficacy in psoriasis [
49].
CXCL8, encoding the neutrophil chemoattractant IL-8, was also upregulated in the patient’s M1 and M2 macrophages. Aberrant regulation of IL-8 has been implicated in numerous inflammatory diseases [
50‐
53], including mucosal tissues, which may partly explain the mucosal ulceration observed here.
The study which first reported pathogenic cases of this mutation also investigated the cochlea of mice and identified tissue-resident macrophage-like cells which secrete IL-1β in response to LPS and ATP stimulation, which may mediate local autoinflammation in the cochlea [
5]. In this report, we have expanded on this by elucidating a molecular mechanism of how the p.R920Q mutation leads to an atypical phenotype. Previously reported
NLRP3 mutations have been shown to affect interactions between NLRP3 and its endogenous regulator NEK7 [
10,
27]; the p.G755R and p.G755A mutations, which cause CINCA, increase the affinity between NLRP3 and NEK7 [
24,
25], while the hypomorphic
NLRP3 missense mutation p.D946G binds with less avidity to NEK7 than to the WT protein [
26]. The p.R920Q mutation increases NLRP3 interactions with NEK7, which was predicted by our in silico interrogation of the protein structure to be due to the reduction in positive charge density at the NLRP3/NEK7 interface when the positively charged arginine residue was substituted for an uncharged glutamine.
The function of the LRR domain remains unclear. The recently determined cryo-EM structure of NLRP3 in complex with NEK7 suggests that the LRR and NACHT domains of NLRP3 are important in protein–protein interactions with NEK7 [
10], and the second of the two major isoforms, identified in humans lacking exon 5 resulting in a truncated LRR domain, exhibits a loss of activity due to a loss of NLRP3/NEK7 interactions [
54]. However, other studies have shown that NLRP3 protein lacking the LRR domain can be activated by the canonical inflammasome pathway [
4,
55,
56]. Two recent preprint studies reporting cryo-EM structures of full-length NLRP3 in its native form indicate that interactions between LRR domains are important for the formation of its endogenous ‘ring cage’ structure [
57,
58]. In Hochheiser et al., residue R920 is proposed to lie within a concave site in the LRR domain which forms contacts between two opposing LRRs; this LRR-LRR interaction is mediated by an acidic loop extending from an LRR transition segment, suggesting that the p.R920Q mutation may affect electrostatic interactions between this loop and the LRR domain in the inactive form. Overall, our observations add to the LRR story, supporting the idea that missense mutations in the LRR domain can play a crucial role in NLRP3 inflammasome function.
Despite the increased understanding with regards to the structural basis of NLRP3 inflammasome function, efforts in understanding the structure–function relationship of NLRP3 are frequently centred on the central NACHT domain. This is due to a combination of factors, including the majority of the
NLRP3-AID-related mutations being found in this region [
59] and because many NLRP3 inhibitors bind to this domain [
60,
61]. The most well-studied pharmacological NLRP3 inflammasome inhibitor, MCC950, has been shown to target NLRP3 in the region of the Walker B motif of the NACHT domain [
60]. As such, previous studies utilising a structure-based approach to investigate small molecule inhibitors of the NLRP3 inflammasome, using either a homology model of NLRP3 produced using the NLRC4 crystal structure [
62] or the NLRP3/NEK7 cryo-EM structure [
10], have targeted the search of druggable binding sites close to the Walker B region [
63,
64]. However, the NLRP3 inflammasome inhibitor 3,4-Methylenedioxy-β-nitrostyrene (MNS) binds to the NACHT and LRR domains [
65], demonstrating that there is potential for small molecule inhibitors targeting domains other than NACHT, particularly in disease conditions involving LRR domain mutations.
Due to a scarcity of patient blood samples resulting from a combination of her pregnancy and COVID-19 restrictions in hospitals, the present study primarily investigated the effects of the p.R920Q variant in circulating myeloid cells, where the NLRP3 inflammasome is predominantly expressed (Guarda et al. 2011). However, a future avenue of investigation would be to explore the effect of this mutation in cells such as neutrophils, particularly as neutrophil-specific
Nlrp3 mutations have recently been shown to promote the development of
NLRP3-AIDs in mice [
66], and PDE4 inhibitors such as apremilast can inhibit neutrophil extracellular trap formation [
67,
68]. This may also offer further insights into the kinetics of the p.R920Q mutation, in addition to the immunoprecipitation studies carried out above.
In summary, we have shown that the rare NLRP3-p.R920Q mutation is associated with an atypical presentation. This is likely connected to enhanced NLRP3-NEK7 interactions, which lead to a lower activation threshold for the NLRP3 inflammasome, in addition to involvement of different convergent pathways. Our results highlight the importance of investigating alternative pathways involved in disease states and outline a potential course of treatment for atypical NLRP3-AIDs patients.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.