The NLRP3 inflammasome is associated with a variety of human diseases, including cryopyrin-associated periodic syndrome (CAPS). CAPS is a dominantly inherited disease with NLRP3 missense mutations. Currently, most studies on the NLRP3-inflammasome have been performed with mice, but the activation patterns and the signaling pathways of the mouse NLRP3 inflammasome are not always identical with those in humans. The NLRP3 inflammasome activation in pigs is similar to that in humans. Therefore, pigs with precise NLRP3-point mutations may model human CAPS more accurately. In this study, an NLRP3 gain-of-function pig model carrying a homozygous R259W mutation was generated by combining CRISPR/Cpf1-mediated somatic cell genome editing with nuclear transfer. The newborn NLRP3 R259W homozygous piglets showed early mortality, poor growth, and spontaneous systemic inflammation symptoms, including skin lesion, joint inflammation, severe contracture, and inflammation-mediated multiorgan failure. Severe myocardial fibrosis was also observed. The tissues of inflamed skins and several organs showed significantly increased expressions of NLRP3, Caspase-1, and inflammation-associated cytokines and factors (i.e., IL-1β, TNF-α, IL-6, and IL-17). Notably, approximately half of the homozygous piglets grew up to adulthood and even gave birth to offspring. Although the F1 heterozygous piglets showed improved survival rate and normal weight gain, 39.1% (nine out of 23) of the piglets died early and exhibited spontaneous systemic inflammation symptoms. In addition, similar to homozygotes, adult heterozygotes showed increased delayed hypersensitivity response. Thus, the NLRP3 R259W pigs are similar to human CAPS and can serve as an ideal animal model to bridge the gap between rodents and humans.
The NLRP3 belongs to the NLR family and is also known as NALP3 or cryopyrin. The NLRP3 inflammasome is an intracytoplasmic protein complex composed of NLRP3, ASC, and procaspase-1 (1). Typically, NLRP3 resides in the cytoplasm in its inactive form. Upon sensing microbial- and/or danger-associated molecular patterns, the pyrin domain of NLRP3 interacts with the pyrin domain of ASC, which further induces the interaction of the CARD domain of ASC with the inactive CARD domain of procaspase-1. This aggregation triggers the automatic cleavage of procaspase-1, thereby releasing the active caspase-1 fragment from its autoinhibitory domain. Thereafter, pro–IL-1β and pro–IL-18 are cleaved by caspase-1 into mature IL-1β and IL-18, respectively (2).
Studies show that the activity of the NLRP3 inflammasome is associated with many diseases, including cryopyrin-associated autoinflammatory syndrome (CAPS) (3, 4). CAPS is the disease spectra of three clinically defined autosomal dominant disorders, namely, familial cold autoinflammatory syndrome (FCAS), Muckle–Wells syndrome (MWS), and neonatal-onset multisystem inflammatory disease (5). Patients with CAPS have inflammatory symptoms involving the skin, muscles, joints, conjunctiva, and CNS (6–8). Cuisset et al. (9) have reported that the estimated prevalence of CAPS in France is 1/360,000. Of the 135 cases reported, 84% were family cases, and 16% were sporadic cases. Most (86%) of the patients have developed symptoms before the age of 10, whereas 96% of patients have developed symptoms before the age of 20. The three most common mutations in NLRP3 from patients with CAPS are Y348M (20%), V198M (16%), and R260W (15%). Notably, the R260W mutation is a frequent and unambiguous CAPS-linked mutation that causes two forms of CAPS (i.e., MWS and FCAS) (6). Given that CAPS is dominantly inherited, the vast majority of the reported patients with CAPS are heterozygous mutations (3, 9–11), and few homozygous NLRP3 missense mutations have been reported (5).
Mouse models play an important role in elucidating the molecular mechanism and the pathology of NLRP3-associated diseases. A mouse model shows that the heterozygous NLRP3 R258W (homologous to R260W in humans) mutation (12) reduces the activation threshold of the NLRP3 inflammasome. Thus, low amounts of ligand lead to strong activation. In mouse models with functionally enhanced NLRP3 missense mutations, including R258W, A350V, and L351P (corresponding to the R260W (12), A352V, and L353P (13) mutations in humans, respectively), only heterozygous (no homozygotes) mice are obtained. In addition, these heterozygous mice can be born alive, but a more severe phenotype than in patients with MWS and FCAS has been observed. Moreover, the L353P human mutation is milder than A352V, but the phenotype of L351P mice is more severe than A350V mice.
In addition, a species-specific response has been notified in terms of the activation patterns and the signaling pathways of the NLRP3 inflammasome. In a previous study, Gaidt et al. (14) have found that LPS triggers IL-1β secretion in human monocytes but not in murine cells, indicating that human monocytes activate an alternative inflammasome in response to LPS. Moreover, the NLRP3 inflammasome activation in pigs is similar to that in humans. The monocytes (14) and the bone marrow–derived macrophages (15) of pigs are reported to resemble those of humans in terms of their response to bacterial LPS. Furthermore, pig characteristics are close to human characteristics in terms of organ size, longevity, anatomy, and physiology. Therefore, pigs with precise NLRP3 point mutations may model human CAPS more accurately.
NLRP3-related studies have been carried out in pigs by using physical or chemical methods (16–19), but, to our knowledge, no NLRP3 gene–edited pigs have been generated because of the inefficiency of the precise genetic modification of somatic cells and somatic cell nuclear transfer (SCNT). In recent years, custom endonucleases, including Zinc-finger nuclease, transcription activator–like effector nuclease, and the clustered regularly interspaced short palindromic repeats (CRISPR) systems (i.e., CRISPR/Cas9, CRISPR/Cpf1, and base editors), have revolutionized genomic engineering in prokaryotes and eukaryotes (20, 21). To date, custom endonuclease-mediated genome engineering, including knockout (22, 23), knock-in (24, 25), point mutations (26–29), and in vivo genome editing (25), has been efficiently achieved in pigs. In thecurrent study, single-stranded oligonucleotides (ssODNs) combining CRISPR/Cpf1 technology with the SCNT approach were used to generate a pig model with NLRP3 R259W homozygous (NLRP3 R259W homo) mutation (homologous to R260W in humans). The newborn NLRP3 R259W homo piglets showed inflammasome hyperactivation, and about half of the piglets survived to adulthood. These adult homozygous pigs exhibited more severe hypersensitivity reactions induced by 1-chloro-2,4-dinitrobenzene (DNCB) than those in wild-type pigs. Furthermore, the NLRP3 R259W heterozygous (NLRP3 R259W hetero) pig population was established by mating the founder pigs with wild-type sows. The NLRP3 R259W hetero pigs showed similar spontaneous inflammatory responses and hypersensitivity reactions as NLRP3 R259W homo pigs, but the survival rate and the weight gain were significantly improved. Our results indicated that the pig model with NLRP3 gain-of-function point mutation may serve as an ideal animal model to investigate the pathogenesis of autoinflammatory diseases to identify new therapeutics.
Materials and Methods
Large white pigs with standard agricultural housing were used as subjects for precise genome editing. The animal experiments performed complied with the guidelines for animal experiments of the Department of Science and Technology of Guangdong Province (China) and were approved by the Animal Research Committee of the Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences (Animal Welfare Assurance #N2015046).
Design of CRISPR/Cpf1 system and ssODNs
In our previous study (27), the Cpf1-sgRNA transcription system was designed and optimized by inserting a tRNA precursor sequence downstream of the CRISPR RNA (crRNA), which could protect the crRNA from immediate digestion by 3′–5′ exonucleases. In the current study, Cpf1-sgRNA targeting the exon 3 of porcine NLRP3 gene was designed and cloned in the pCS2-crRNA-tRNA plasmid. In detail, two complementary oligo primers containing targeting sequences and 4 bp overhangs were synthesized, annealed, and ligated into the BpiI-digested pCS2-crRNA-tRNA cloning vector. The constructed sgRNA vectors were confirmed through Sanger sequencing by Guangzhou IGE Biotechnology (Guangzhou, China). ssODNs (98 nt) with two synonymous mutations (TTT > TTC and TGT > TGC) and a nonsynonymous mutation (CGG > TGG) were designed and synthesized by Guangzhou IGE Biotechnology. ssODNs (1 μg/μl) were diluted with RNase-free water and stored at −20°C for future use.
Porcine fetal fibroblasts culture, transfection, and single cell–derived colony selection
As previously reported (23, 27), porcine fetal fibroblasts (PFFs) were derived from the 35-d-old fetuses of large white pigs. For transfection, 106 PFFs were resuscitated from liquid nitrogen. On the next day, 5 μg of ssODNs, 10 μg of AsCpf1, and 3 μg of NLRP3-sgRNA expression vectors were coelectroporated into PFFs by using the Neon Transfection System (Life Technology) (1350 V, 30 ms, and one pulse). G418 (800 μg/ml) was used for selection, and survived single cell–derived colonies were retrieved and genotyped by using Sanger sequencing. The genotyping forward and reverse primers were 5ʹ-ACTGCCGCGACCTACTTC-3ʹ and 5ʹ-CTGACTGTCCCAGCCGTT-3ʹ, respectively. The PCR conditions were as follows: 95°C for 3 min; 98°C for 10 s, 60°C for 20 s, 72°C for 10 s, 35 cycles; and 72°C for 5 min. The cell colonies with precise NLRP3 R259W–point mutations were cultured, expanded, and frozen in liquid nitrogen for future use.
Production of NLRP3 R259W homo–cloned pigs by SCNT
The detailed methods for SCNT were reported in our previous study (22, 24). In brief, porcine oocytes were collected and matured in vitro for 44 h, and the screened NLRP3 R259W homo–positive cell colony was revived as donor nuclei on the day of preparation of oocytes. Oocytes were enucleated through blind aspiration, and a donor cell was drawn into the perivitelline space of the enucleated oocyte. Next, a complex of enucleated oocytes and somatic cells was fused and activated by electrical pulse stimulation. The reconstructed embryos were cultured in the cell incubator at 38.5°C, and 5% CO2. After 20 h of culture, the reconstructed embryos were transplanted into the oviduct of the recipient sows by surgery. Thirty days after embryo transfer, the pregnancy status was examined using an ultrasound scanner, and pregnant recipient sows were monitored every week to track the development of the cloned embryos. After ∼114 d of pregnancy, the recipient sows naturally gave birth, and the cloned piglets with expected precise point mutations were obtained.
Genomic DNA was extracted from the ear tissues of newborn cloned piglets, and the genotypes of the newborn piglets were analyzed the same as the above-mentioned method for genotyping selected cell colonies.
The method of prediction and search for potential off-target sites (OTSs) was performed as previously described (30). After retrieving a base-by-base scan of the whole pig genome, which allowed for ungapped alignment of, at most, five mismatches in the sgRNA target sequence, all OTSs homologous to the 24 bp sequence [PAM (TTTN) + sgRNA] were identified. All OTSs were amplified by PCR, and the PCR products were analyzed by using Sanger sequencing.
Total RNAs were extracted from the ear, skin, heart, liver, spleen, lung, kidney, and small intestine tissues of NLRP3 R259W homo and age-matched wild-type pigs by using TRIzol (Invitrogen). cDNAs were synthesized from 1 μg of total RNAs with the PrimeScript RT Reagent Kit with gDNA Eraser (Takara Bio) in a 20-μl reaction volume following the manufacturer’s instructions. Quantitative PCR (Q-PCR) was performed using the SYBR Green PCR Master Mix (Takara Bio) in the Bio-Rad CFX96 (Bio-Rad Laboratories) in a 10-μl reaction volume by using the following conditions: 95°C for 15 min; 40 cycles at 95°C for 10 s, 60°C for 30 s, plate read; and 95°C for 10 s, 65°C for 5 s, melt curve from 65.5 to 95°C, increment of 0.5°C for 5 s, plate read. Data were analyzed using the 2ΔΔCt method (31). Equal loading was confirmed by simultaneous β-actin amplification. The primers used to amplify cytokines were selected from the qPrimerDB (32), a thermodynamics-based gene-specific Q-PCR primer database.
H&E and immunohistochemical staining
The tissues of the skin, heart, liver, spleen, lung, kidney, and small intestine from the sacrificed age-matched wild-type and short-lived NLRP3 R259W homo piglets were fixed in 4% paraformaldehyde. Similarly, the ear tissues of the DNCB-stimulated adult wild-type and NLRP3 R259W homo pigs were fixed in 4% paraformaldehyde. After 2 d of fixation, the tissues were subsequently dissected, embedded in paraffin wax, and cross-sectioned at 3 μM. The sections were deparaffinized with xylene and rehydrated using a graded series of alcohol (100, 90, 80, 70, and 50%), and water. For H&E staining, the rehydrated sections were stained with H&E, differentiated, and placed with a cover-slip. For immunohistochemical (IHC) staining, the sections were stained using the standard IHC staining protocols. The following Abs were used for IHC: anti-NLRP3 (GB11300, 1:300; Servicebio), anti–Caspase-1 (GB11383, 1:600; Servicebio), anti–IL-1β (GB11113, 1:600; Servicebio), anti–TNF-α (GB11188, 1:500; Servicebio), anti–IL-6 (GB11117, 1:300; Servicebio), anti–IL-17 (GB11110, 1:800; Servicebio), and anti-F4/80 (GB11027, 1:1000; Servicebio). HRP-labeled goat anti-rabbit IgG (GB23303, 1:200; Servicebio) was used as secondary Ab.
Blood samples derived from NLRP3 R259W point mutation and age-matched piglets were examined in the Guangdong Laboratory Animals Monitoring Institute.
The back skins of the adult wild-type and NLRP3 R259W homo pigs (or NLRP3 R259W hetero pigs) were shaved and exposed to contact irritants. DNCB (1 ml, 5%; Sigma-Aldrich) dissolved in acetone/olive oil solution (4:1, v/v) was used for sensitizing 10 sites (sensitizing phase), and back skin inflammation was observed daily. After 6 d, the ears were again exposed to irritants by using 1 ml of 0.5% DNCB at 10 sites (elicitation phase). The ears were observed, and the tissues of the inflammation sites were collected 24 h after irritation and subjected to H&E staining, IHC, and extraction of RNA for Q-PCR analysis.
The Graphpad Prism5 software was used for data analysis. The two-tailed Student t test was used to assess a significant difference. A p value <0.05 was considered statistically significant.
CRISPR/Cpf1–mediated gene targeting in PFFs
The arginine region of the 260th aa sequence in the human NLRP3 gene is conserved in pigs and mice as the 259 and 258th aa, respectively (Fig. 1A). To mimic R260W point mutation of NLRP3 gene in pigs, the NLRP3-Cpf1-sgRNA was designed and assembled following the protocol reported (27). ssODNs with lengths of 98 nt, which harbored two synonymous mutations (TTT > TTC and TGT > TGC) and a nonsynonymous mutation (CGG > TGG), were used as homology-directed repair donors (Fig. 1B). The introduced nonsynonymous mutation led to R-to-W amino acid substitution in the 259th region of the NLRP3 protein, whereas the synonymous mutations were designed to avoid repetitive targeting (Fig. 1B).
ssODNsR259W, Cpf1, and NLRP3-Cpf1-sgRNA–expressing vectors were cotransfected into PFFs derived from 35-d-old male large white pig fetuses through electroporation. After G418 selection for 10 d, 50 cell colonies were selected, and 46 (46 out of 50, 92%) were expanded. The Sanger sequencing results of the PCR products covering the target site showed that four cell colonies (#5, #7, #26, and #35; four out of 46, 8.7%) harbored the expected R259W-point mutation (Fig. 1C). The #26 and the #35 cell colonies contained the homozygous R259W–point mutation, and the #7 cell colony only harbored the R259W-point mutation in one allele. The #5 cell colony had the R259W-point mutation, an unwanted 2 bp deletion, and unexpected AG > CA substitutions in one allele, and the other allele was wild-type (Fig. 1E). In addition, 33 cell colonies (33 out of 46, 71.74%) were repaired by nonhomologous end joining.
Generation of NLRP3 R259W homo piglets via SCNT
Given the continuous passage of cells, #5 and #35 cell colonies exhibited aging phenotypes during the passage to expand the cell number. Thus, these cell colonies were unsuitable for use as nuclear donor for SCNT. Therefore, the #26 cell colony was selected as nucleus donors to investigate whether live NLRP3 R259W homo piglets can be generated by SCNT. A total of 922 reconstructed embryos were transferred into four surrogate mothers. Three mothers (three out of four, 75.0%) developed to full term and gave birth to 28 cloned male piglets (Fig. 1D, 1F). The ear tissues were collected for genomic DNA extraction and genotyping. The Sanger sequencing results showed that all cloned piglets carried the NLRP3 R259W homo–point mutation, corresponded to the genotype of the donor nuclei (Fig. 1G). These results suggested that live NLRP3 R259W homo piglets were successfully generated via SCNT.
The off-target effect is the main concern for the CRISPR system–mediated genome editing. For NLRP3-Cpf1-sgRNA, five potential OTSs were predicted and analyzed by using Sanger sequencing. One piglet was randomly selected for off-target detection, as all cloned pigs were derived from the #26 cell colony. The sequencing results showed no sign of off-targeting at these five potential OTSs (Supplemental Fig. 1).
NLRP3 R259W homo piglets showed early mortality, poor growth, and systemic inflammation
Among the 28 NLRP3 R259W homo piglets, three were stillborn, and 25 live-born piglets were weak at birth and suffered from tremor and hind leg weakness (Fig. 1A, Supplemental Table I). Twelve of the live-born piglets died within a week (one, nine, one, and one piglets died at days 1, 2, 3, and 7, respectively) (Fig. 1B, Supplemental Table I). Of these 12 short-lived piglets, skin lesion (urticarial skin rash [n = 4], skin flushing and cyanosis [n = 4], and s.c. blisters [n = 1]), hind limb weakness (n = 12), severe contracture (n = 6), and shock (n = 12) were observed (Fig. 2A, Supplemental Table I). Notably, six piglets suffered with contracture on the first day after birth, five of which had a body temperature of 34°C on day 2, and one had relatively normal temperature (between 37 and 38°C) for 6 d and finally dropped to 34°C. All six piglets fell into shock and eventually died. The symptoms of these newborn homozygous piglets were very similar to those reported for newborn patients with CAPS (7, 8, 33).
The skin tissues of five short-lived homozygous (i.e., 800378-1, 800378-7, 800378-15, 800378-19, and 800378-23) and the age-matched wild-type piglets were collected for staining and Q-PCR. The H&E staining of cutaneous hives showed perivascular infiltration of neutrophils, indicating the automatic occurrence of inflammation (Fig. 2D). IHC staining showed that the expression of the NLRP3, Caspase-1, IL-1β, TNF-α, and IL-6 in these dead homozygous piglets were higher than those in wild-type piglets (Fig. 2E). IL-17 was observed in and around the s.c. blood vessel of homozygous piglets, whereas almost no IL-17–positive staining was found in the skin of wild-type piglets (Fig. 2E). The mRNA levels of proinflammatory cytokines (IL-1β, TNF-α, and IL-6), Th17 cell–related master regulator transcription factor and unique cytokines (RORγt and IL-17F) were dramatically increased (6.38, 13.61, 1.48, 2225, and 3.56 times, respectively), compared with age-matched wild-type piglets, implying that the Th17 cell differentiation in the skins of dead homozygous piglets was significantly promoted (Fig. 2F). The expression of Th1 cell–related master regulator transcription factor, T-bet, decreased significantly to 5% of wild-type piglets. The expression of IFN-γ increased to 3.77 times but without significant difference. In addition, the expression of GATA-3 increased 2.09-fold, whereas IL-5 decreased to 81%. These results suggested that the Th1 and the Th2 cell differentiation in the skins of dead homozygous piglets were inhibited to a certain extent (Fig. 2F).
The sera of the six dead homozygous piglets (i.e., 800378-1, 800378-3, 800378-13, 800378-19, 800378-21, and 800228-1) were subjected to biochemical analysis. The results revealed that pronounced myocardial injury, hepatorenal dysfunction, and steady-state imbalance were observed in tested NLRP3 R259W homo piglets (Fig. 2G). Compared with age-matched wild-type piglets, the NLRP3 R259W homo piglets had significantly increased aspartate aminotransferase (AST), AST/alanine aminotransferase (ALT), and lactate dehydrogenase by 7.43, 4.63, and 3.31 times, respectively; significantly decreased glucose and high-density lipoprotein-cholesterols (HDL-C) to 31.32 and 16.06%; decreased albumin (ALB) and ALB/globulin (GLOB) to 32.44 and 23.01%, respectively; and increased urea by 3.14 times, and decreased triglyceride to 52.8%. The changes in creatine kinase, GLOB, ALT, CREA, and total cholesterol also showed the heart, liver, and kidney abnormalities of NLRP3 R259W homo piglets. These blood biochemical results preliminarily suggested that the death of newborn NLRP3 R259W homo piglets may be caused by excessive inflammatory reactions and severe multiorgan damage.
The remaining 13 homozygous-point mutation piglets showed significantly less weight gain than wild-type piglets within 4 wk after birth (Fig. 2C). However, no significant difference was found between the two groups after 6 mo of feeding.
Significant multiorgan pathological changes associated with inflammation in short-lived NLRP3 R259W homo piglets
The tissues of the heart, liver, spleen, lung, kidney, and small intestine of five spontaneously inflamed homozygous (i.e., 800378-1, 800378-2, 800378-7, 800378-15 and 800378-19) and wild-type piglets were collected and analyzed by using H&E staining (Fig. 3). The signs of fibrosis were observed in myocardial tissue from inflamed homozygous piglets, and this result was further supported by the evidence from Masson and Sirius red staining (Fig. 3A). The H&E staining of liver showed macrovesicular steatosis, hepatocellular ballooning, and lobular inflammation of homozygous piglets, and the hepatocytes were loosely arranged. In addition, the Sirius red staining revealed the disappearance of hepatic lobule structure. The number of F4/80+ granulocytes and macrophages in the liver of homozygous piglets increased (Fig. 3B). These pathological phenotypes of liver and myocardial injury were very similar to nonalcoholic fatty liver disease and serious nonalcoholic steatohepatitis reported in humans and mouse models (34, 35). Decreased white pulp cell density in the spleen of NLRP3 R259W homo piglets was found. The periarteriolar lymphoid sheath of white pulp appeared “moth-eaten,” which may be due to scattered lymphocyte apoptosis (Fig. 3C, left panel). Moreover, hyperemia and emphysema appeared in the lung of spontaneously inflamed homozygous piglets (Fig. 3C, right panel). The renal H&E staining results showed that the kidneys of spontaneously inflamed homozygous piglets exhibited inflammatory cell infiltration and decreased glomerular and tubular cells in the deep cortex (Fig. 3D, left panel). Furthermore, severe inflammation and structure disruptions occurred in the small intestines of short-lived homozygous piglets (Fig. 3D, right panel).
IHC assays were performed on three typical dead homozygous piglets (i.e., 800378-19, 800378-15, and 800378-7) to initially confirm the expression levels of NLRP3, NLRP3-associated genes, and inflammation-associated cytokines. Compared with wild-type piglets, short-lived NLRP3 R259W homo pigs had elevated the expression of NLRP3 in the heart, liver, spleen, lung, kidney, and small intestine, which confirmed that NLRP3 R259W homo mutation could enhance NLRP3 expression in autoinflammatory response (Fig. 4A). In mice and humans, the NLRP3 inflammasome activation triggers the autoactivation of caspase-1 and the maturation and secretion of proinflammatory cytokines, such as IL-1β (12, 14). In pigs, after activation, the NLRP3 overexpression increased the expression of Caspase-1 and IL-1β in multiple organs, and this finding was confirmed by Q-PCR or/and IHC and consistent with humans and mice (Figs. 4B, 4C, 5). Besides IL-1β, the expression levels of other proinflammatory cytokines, such as IL-6 and TNF-α, in the heart, liver, lung, kidney, and intestine in short-lived NLRP3 R259W homo piglets were higher than those in wild-type piglets (Figs. 4D, 4E, 5). The expression of Th2-related cytokines, like IL-5, were significantly upregulated in the lung, kidney, and intestine but clearly downregulated in the spleen of inflamed NLRP3 R259W homo piglets (Fig. 5). The expression pattern of GATA-3 was consistent with that of IL-5. The high expression of the Th1-related factor (T-bet), and Th1-related cytokine (IFN-γ) were observed in the heart, lung, and kidney of gene-modified pigs. The expression levels of T-bet and IFN-γ were downregulated and upregulated, respectively, in the spleen and the small intestine (Fig. 5). Moreover, the expression levels of T-bet and IFN-γ in the liver were upregulated and downregulated, respectively. The Th17 cell differentiation in the skin, hearts, livers, spleens, lungs, kidneys, and intestines of all the dead homozygous piglets showed a significant increase (Figs. 2F, 4F, 5). The expression levels of Th17 cell–related cytokines, such as IL-17F, and Th17 cell–related factor (i.e., RORγt) in multiple organs of inflamed NLRP3 R259W homo piglets were significantly increased compared with those of wild-type piglets. These data suggested that the NLRP3 R259W mutation potentiated Th17 cell responses in our pig model, which was consistent with that in a mouse model with NLRP3 R258W mutation (12). Overall, these findings suggested that the gain-of-function of NLRP3 mutation resulted in changes in expression patterns of related genes and cytokines in multiple organs, which led to multiorgan failure and death.
Adult NLRP3 R259W homo pigs exhibited stronger inflammation than wild-type pigs
The adult homozygous pigs presented no morphological abnormality. The change in the blood indices of homozygotes compared with age-matched adult wild-type pigs was verified. Seven adult homozygous pigs (i.e., 903303, 903305, 903307, 903309, 903403, 903409, and 903413) were subjected to blood biochemical and routine examinations. The blood biochemical results showed that the ALT and the GLOB of adult homozygotes were significantly increased, whereas the AST/ALT and ALB/GLOB were significantly decreased, indicating impaired liver function in the adult homozygous pigs. The creatine kinase and the lactate dehydrogenase were significantly increased, indicating a myocardial injury in adult homozygous pigs. The HDL-C and total protein were significantly increased, suggesting a renal impairment in adult homozygous pigs (Supplemental Fig. 2A). Routine blood examination results showed the percentage of basophils, the absolute numbers of basophils, eosinophils, and lymphocytes of adult homozygous pigs were significantly decreased; the percentage of monocytes were significantly increased; the platelet-related indices, including average platelet volume, platelet distribution width and large platelet ratio were significantly decreased (Supplemental Fig. 2B). The abnormality of the platelet of NLRP3 R259W homo pigs was consistent with patients with immune thrombocytopenia, which was caused by aberrant NLRP3 inflammasome activity (36). These results suggested that adult NLRP3 R259W homo pigs had physiological abnormalities, but did not show morphological abnormalities.
Previous reports suggest that the NLRP3 inflammasome is activated during delayed-type hypersensitivity (DTH) (37, 38). DTH is a T cell–mediated skin-inflammatory reaction to repeated cutaneous exposure to small sensitizing chemicals or haptens. DNCB is a chemical that induces contact hypersensitivity in mice and rats (12, 39). Therefore, we verified whether DNCB could induce DTH in NLRP3 R259W homo pigs without spontaneous abnormalities. The homozygotes showed more intense inflammation at 24 h postapplication of 5% DNCB solution in the back with skin contraction and redness (Fig. 6A). On day 6, the ears of these pigs were reapplied with 0.5% DNCB. Edema and the redness of ears were observed in homozygous and wild-type groups on day 7, but the ear of the homozygotes showed a more severe inflammation (Fig. 6B). H&E staining showed that the infiltration of neutrophils and monocytes were more severe in the swollen part of NLRP3 R259W homo pig ears than those of wild-type pigs (Fig. 6C). The IHC staining of the swollen ears of NLRP3 R259W homo pigs showed that the expression levels of NLRP3, Caspase-1, IL-1β, TNF-α, IL-6, and IL-17 were more evident than those of wild-type pigs (Fig. 6D). In addition, the edema and the redness of ears of NLRP3 R259W homo pigs lasted longer than those of wild-type pigs.
The analysis of the swollen ears by Q-PCR and IHC revealed a significant increase in IL-1β, IL-6, TNF-α, and Th17 cell–related master regulator transcription factor and unique cytokines (i.e., RORγt and IL-17F) in homozygous pigs compared with those in wild-type pigs, and these results were similar to those of spontaneous dermatitis developed in homozygous pigs. T-bet was significantly upregulated, whereas GATA-3 showed no difference between the two groups. These findings were inconsistent with those of spontaneous dermatitis developed in homozygous pigs. The expression levels of IFN-γ and IL-4 were significantly upregulated (Fig. 6D, 6E). IL-1β plays a key role in contact hypersensitivity (38). Thus, a significant increase in expression indicates the successful induction of DTH response in pigs. T-bet and IL-5 were both upregulated, which was consistent with the study that Th1 immune response mediated the DNCB-induced contact dermatitis (39). In summary, adult NLRP3 R259W homo pigs without spontaneous dermic inflammation showed enhanced DTH responses than wild-type pigs.
Generation of NLRP3 R259W hetero piglets by natural mating
A total of 23 piglets were born after 114 d of gestation by mating three NLRP3 R259W homo boars (i.e., 903303, 903309, and 903403) with five wild-type sows (i.e., two Tibetan and three Bama minipigs). The ear tissues of the 23 piglets were collected for genomic DNA extraction and genotyping. The Sanger sequencing results showed that all piglets carried the NLRP3 R259W hetero mutation (Fig. 7A, Supplemental Fig. 3). All 23 heterozygous piglets were live-born, and the nine live-born piglets were weak at birth and died within 2 d (Fig. 7B, 7C, Supplemental Table I). Of the nine short-lived heterozygous piglets, tremor (n = 6), hind leg weakness (n = 1), shock (n = 9), severe contracture (n = 3), and impaired vision (n = 2) were observed. The symptoms of these heterozygous piglets were similar to their parental generation, but the severity and the probability of occurrence were reduced compared with the homozygous ones (Fig. 7B, Supplemental Table I). The remaining 14 heterozygotes survived well and showed normal body weight (Fig. 7C, 7E).
The tissues of the heart, liver, spleen, lung, kidney, and small intestine of all the nine spontaneously inflamed heterozygous piglets were collected and analyzed by H&E staining (Fig. 7D). The signs of fibrosis and inflammatory cell infiltration were observed in myocardial tissue from inflamed heterozygous piglets, which were further supported by the evidence from Masson staining (Fig. 7D, Supplemental Fig. 3A). The H&E staining of liver showed hepatocellular ballooning and the lobular inflammation of heterozygous piglets, and the hepatocytes were loosely arranged. In addition, the number of F4/80+ granulocytes and macrophages and the MPO+ neutrophils in the livers of heterozygous piglets increased (Fig. 7D, Supplemental Fig. 3B). Decreased white pulp cell density in the spleen of heterozygous piglets was found, and the periarterial lymphoid sheath of white pulp appeared “moth-eaten,” which was similar to that observed in the spleen of homozygous piglets. Moreover, hyperemia, emphysema, and inflammatory cell infiltration appeared in the lung of spontaneously inflamed heterozygous piglets. The renal H&E staining results showed severe inflammatory cell infiltration, and decreased glomerular and tubular cells in the deep cortex, and these results were similar to the amyloid lesions observed in the kidneys of human patients. Furthermore, severe inflammation and structure disruptions occurred in the small intestines of short-lived heterozygous piglets (Fig. 7D).
The blood biochemical tests of the six 3 mo old heterozygous piglets revealed myocardial injury, hepatorenal dysfunction, and steady-state imbalance but not as pronounced as that observed in their homozygous parents. Compared with wild-type pigs, the heterozygous pigs had significantly increased AST/ALT, CREA, HDL-C, total cholesterol, and triglyceride by 1.53, 1.43, 1.79, 1.59, and 2.31 times, respectively (Supplemental Fig. 2C). In addition, the erythrocyte-related indices, including mean RBC volume, hematocrit, and RBC distribution width–SD were significantly increased, whereas the mean erythrocyte hemoglobin concentration was significantly decreased compared with wild-type pigs (Supplemental Fig. 2D).
The DTH test was also performed on two of the 3 mo old heterozygous piglets (i.e., NLRP3 F1-2 and NLRP3 F1-3) and one age-matched wild-type piglet. The heterozygous piglets showed more intense inflammation at 24 h post application of 5% DNCB solution in the back with skin contraction and redness compared with wild-type piglets (Fig. 7F).
Some of the disadvantages of the use of pig models include cost, limited reproducibility, and inconvenience of experiment operation, but pig models are ideal immune disease models because their immune system closely resembles that of humans (>80%) on the basis of the analyzed parameters compared with that of mice (<10%) (40). In this study, the phenotype of early mortality, poor growth, and systemic inflammation in our NLRP3 R259W homo pig model exhibits many similarities to that of patients with CAPS and mouse models (6, 12, 41). Symptoms such as urticaria, edema, articular signs, shock, impaired vision, myocardial injury, hepatorenal dysfunction, steady-state imbalance, thrombocytopenia, and neurologic manifestations are observed in the pig model and similar to those observed in human patients with CAPS. Interestingly, NLRP3 R259W homo pigs can survive to adulthood and obtain offspring through natural mating, which is difficult to achieve in the mouse model. The difference in survival rate and the intensity of inflammatory response between homozygous and heterozygous piglets also suggest the dose effect of the NLRP3 gain-of-function mutation.
Pigs have multiplex genetic backgrounds and are housed in standard agricultural housing rather than specific pathogen-free conditions, which makes the microbiomes and immune response of pigs close to the real physiological functions of human. As such, like human patients living in the real world, the phenotype of the pig model is controlled by genetic factors and affected by environmental factors and different genetic backgrounds. Although theoretically homozygous piglets should come from the same cell clone, the survival rate of piglets born from different sows is significantly different. In the past years, increasing evidence has shown that in humans carrying the NLRP3 mutation in the NACHT domain, some are asymptomatic, whereas some show CAPS symptoms (42–44). Moreover, the disease severity varies even among people carry similar NLRP3 mutations. Other factors account for the final disease phenotype besides genetic factors, for which our present pig model can be well applied for future investigation.
In summary, our pig models with precise NLRP3 point mutation are expected to mimic human autoinflammatory syndromes more accurately than mice, and can be used in preclinical studies and to investigate the mechanism of NLRP3 inflammasome and screen anti-inflammatory drugs.
We thank Henan Chuangyuan Biotechnology Co. Ltd. for assistance with SCNT, animal feeding, and care. We also thank all the members in the laboratory of Prof. L. Lai.
This work was supported by the National Key Research and Development Program of China (2017YFA0105103, 2018YFA0507300), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA16030503, XDB29030303), the National Natural Science Foundation of China (81830049), the National Science and Technology Major Project in New Varieties Cultivation of Transgenic Organisms (2014ZX0801011B), the National Program on Key Basic Research Project of China (973 Program) (2011CBA01001), the Key Research and Development Program of Guangzhou Regenerative Medicine and Health Guangdong Laboratory (2018GZR110104004), the Youth Innovation Promotion Association of the Chinese Academy of Sciences (2019347), the Science and Technology Planning Project of Guangdong Province, China (2019A1515012090, 2017B020231001, 2017A050501059, 2017B030314056, 2014B020225003), the Science and Technology Program of Guangzhou, China (201704030034, 202007030003, 202002030382), and the Research Unit of Generation of Large Animal Disease Models, Chinese Academy of Medical Sciences (2019-I2M-5-025).
The online version of this article contains supplemental material.
Abbreviations used in this article:
cryopyrin-associated periodic syndrome
clustered regularly interspaced short palindromic repeats
familial cold autoinflammatory syndrome
- NLRP3 R259W hetero
NLRP3 R259W heterozygous
- NLRP3 R259W homo
NLRP3 R259W homozygous
porcine fetal fibroblast
somatic cell nuclear transfer
The authors have no financial conflicts of interest.