Alternative splicing regulates stochastic NLRP3 activity

Abstract

Leucine-rich repeat (LRR) domains are evolutionarily conserved in proteins that function in development and immunity. Here we report strict exonic modularity of LRR domains of several human gene families, which is a precondition for alternative splicing (AS). We provide evidence for AS of LRR domain within several Nod-like receptors, most prominently the inflammasome sensor NLRP3. Human NLRP3, but not mouse NLRP3, is expressed as two major isoforms, the full-length variant and a variant lacking exon 5. Moreover, NLRP3 AS is stochastically regulated, with NLRP3 ∆ exon 5 lacking the interaction surface for NEK7 and hence loss of activity. Our data thus reveals unexpected regulatory roles of AS through differential utilization of LRRs modules in vertebrate innate immunity.

Fig. 1

LRR domains of the NLR family have a conserved multi-exon organization. a Scheme of domain and exon distribution in TLR4 and NLRP3. b Workflow for the selection of LRR exons used in c–f. c Length distribution of exons extracted in b. Box indicates 25th to 75th percentile, the middle line indicates the median, whiskers indicate min and max values. d All exons up to 200 bp in length plotted for their frequency distribution. Exons were divided in frame-shifting or frame-preserving. The typical length of LRRs (23 to 29 aa) is indicated as gray boxes. Colored boxes in the lower panel match assignment of groups in the following subfigures. e Genes in word clouds represent those genes whose exons contribute to the respective peaks in d. Word size is linearly dependent to the number of contributing exons. f Phylogenetic analysis of all genes contributing to the 4 major peaks in d. g Model of the NLRP3 LRR based on the human ribonuclease inhibitor LRR crystal structure. h LRR consensus sequences and structural alignments for the four groups identified in d–f. i Quantification of all LLR exons vs. non-LRR exons of all NLRs which would be frame-shifting if alternatively spliced. Independence of distribution was calculated using two-tailed Fischer’s exact test. See also Supplementary Fig. 1. Source data are provided as a Source Data file

Fig. 2

The splicing landscape of human NLRs. a Gene expression of all previously identified RI-type LRR encoding genes (Fig. 1d–f). b Sashimi plot of NLRP3 expressed in human monocyte-derived macrophages established from five healthy human blood donors created with MISO. Read frequency within exons is plotted as RPKM and exon spanning reads are labeled with the number of mapped reads. The NLRP3 gene structure is plotted above, with boxes indicating exonic regions and arrows within the intronic stretches indicating the reading directions. Short repetitive LRR exons are highlighted with a gray box. The genomic location is depicted below. c, d Sashimi plots as in b, focused on exons 4–5–6 and 6–7–8, respectively. MISO ψ (% spliced in) values (red bars in histogram) indicate the calculated frequency of exon inclusion. 95% confidence intervals are indicated as gray bars in the histogram. Ψ and CI values are listed as well numerically. See also Supplementary Fig. 2 and Supplementary Table 1. Source data are provided as a Source Data file

Fig. 3

The LRR domain of human NLRP3 is subject to alternative splicing a Scheme of the NLRP3 exons and domains. Arrows indicate primers used in b and Supplementary Fig. 3a, b. b PCR of the NLRP3 LRR on cDNA isolated from LPS primed mouse BMDMs, pig and human PBMCs, respectively. Representative of at least three (mouse, human) or two (pig) individuals. c Immunoblot of human NLRP3 from primary human monocyte-derived macrophages (hMDM) or THP-1 cells. Either whole cell lysates or NLRP3 immunoprecipitates, using mAb targeted against the NACHT domain of NLRP3, were immunostained with a mAb targeted against the PYD to ensure NLRP3 specificity. Representative of two experiments. d Scores for the probability to function as splice acceptor and donor sites were calculated for all human NLRP3 LRR exon boundaries using SplicePort. e Number of exonic splice enhancer (ESE) sites within the exons of the LRR as predicted by RESCUE-ESE. f Scheme of a splice-switching oligo (SSO), blocking the spliceosomal access to an intron-exon boundary inducing AS. g Changes in the NLPR3 alternative splicing pattern of M0 hMDMs were induced with an exon 5 SSO. NLRP3 isoform expression analysis by qPCR. Mean and SEM of three donors (untreated: n = 2). h Cytokine secretion of morpholino-treated cells after priming with LPS (TNF) and nigericin-induced activation of the NLRP3 inflammasome (IL-1β). Mean and SEM of 3 donors (LPS only: n = 2). Each donor is plotted using a unique symbol shape. See also Supplementary Fig. 3. Source data are provided as a Source Data file

Fig. 4

NLRP3 ∆ exon 5 is inactive. a Immunoblot expression control of NLRP3-tagRFP and ASC-mCerulean from HEK TREx Flp-In cells. b NLRP3-tagRFP was induced in HEK TREx with the respective concentrations of doxycycline. Expression levels were measured by FACS. c HEK TREx cells expressing inducible NLRP3-tagRFP and ASC-mCerulean were analyzed for ASC speck formation by fluorescence microscopy. Cell nuclei were counterstained with Draq5. Scale bar represents 50 μm. d Quantification of ASC speck formation after doxycycline induced NLRP3 overexpression (0–10 ng/mL). Mean and SD of nine frames per condition. Overlaid symbols represent single measurements. e Quantification of ASC speck formation after 2.5 h stimulation with nigericin (0–10 µM). Mean and SD of technical duplicates, nine frames per well, representative of three independent experiments. Overlaid symbols represent single measurements. f Co-immunprecipitation (IP) of ASC with NLRP3-tagRFP from HEK TREx Flp-In cells. NLRP3 was immunprecipitated using anti-tagRFP mAb. Asterisk indicates heavy band of IP mAb. Representative of two independent experiments. g and h NLRP3-deficient immortalized macrophages (iMos) were retrovirally reconstituted with either NLRP3 full-length or NLRP3 ∆ exon 5. g Cytokine secretion after priming with LPS (TNF) and activation of the NLRP3, NLRP1b, or AIM2 inflammasomes (IL-1β). Mean and SD of technical triplicates, representative of three independent experiments. Each individual data point from one experiment is plotted using a unique overlaid symbol shape. h Immunoblots of iMos after activation of the NLRP3 inflammasome (ATP, nigericin) or the NLRP1b inflammasome (lethal toxin). Blots are representative of two independent experiments. See also Supplementary Fig. 4. Source data are provided as a Source Data file

Fig. 5

NLRP3 splicing is regulated on a single cell level. a Single PI-negative GM-CSF derived hMDMs were FACS-sorted into individual wells and lysed. RNA was reverse transcribed and NLRP3 full-length, NLRP3 ∆ exon 5 and HPRT encoding mRNAs were pre-amplified. Transcripts were detected with nested TaqMan assays. 187 to 192 individual cells per donor. b Quantification of the single cell NLRP3 splice pattern. Shown as the mean of three donors from a. c Scheme of the burst-kinetic of gene-expressions on single-cell level, resulting in oscillations of produced mRNA levels per gene. d Human monocyte-derived macrophages were analyzed for ASC speck formation by fluorescence microscopy after NLRP3 activation with nigericin and NLRC4 activation with bacterial product PrgI (both 1.5 h). Cell nuclei were counterstained with Draq5. Scale bar represents 50 μm. Five images per well were captured, plotted are means and SD of two replicate wells, representative of four individual experiments. Overlaid symbols represent single measurements. e IL-1β ELISpot assay of hMDMs after NLRP3 or NLRC4 inflammasome activation. Shown are two independent donors. Mean and SD of technical duplicates, two independent donors. Overlaid symbols represent single measurements. See also Supplementary Fig. 5. Source data are provided as a Source Data file

Fig. 6

NLRP3 ∆ exon 5 lacks NEK7 interaction surface. a Co-immunprecipitation (IP) from iMos stably expressing the respective NLRP3-mCitrine variants. IP was performed in GFP-trap plates. b Models of the NLRP3 LRRs based on the crystal structure of human ribonuclease inhibitor. Shown are the LRR model structures of NLRP3 full-length and NLRP3 ∆ exon 5, as well as two artificially created hybrid LRRs: NLRP3 LRR lacking exon 5 but carrying a duplicate exon 6, and NLRP3 LRR carrying a duplicate exon 6 in which all surface amino acids of exon 5 were rescued. c NLRP3-deficient iMos were reconstituted with full-length or the hybrid NLRP3-mCitrine variants from b. IL-1β measured after priming with LPS and activation of the NLRP3, AIM2 or NLRC4 inflammasomes. Mean and SEM of three experiments. Each individual data point from one experiment is plotted using a unique overlaid symbol shape d Co-IP from iMos stably expressing the respective NLRP3-mCitrine variants. IP was performed in GFP-trap plates. Representative of two independent experiments. e Co-IP from HEK cells transiently transfected to express the respective NLRP3-mCitrine variants. IP was performed in GFP-trap plates. Representative of three independent experiments. See also Supplementary Fig. 6 and Supplementary Table 2. Source data are provided as a Source Data file