Deficiency of the Purinergic Receptor 2X7 Attenuates Nonalcoholic Steatohepatitis Induced by High-Fat Diet: Possible Role of the NLRP3 Inflammasome

Molecular mechanisms driving transition from simple steatosis to nonalcoholic steatohepatitis (NASH), a critical step in the progression of nonalcoholic fatty liver disease (NAFLD) to cirrhosis, are poorly defined. This study aimed at investigating the role of the purinergic receptor 2X7 (PR2X7), through the NLRP3 inflammasome, in the development of NASH. To this end, mice knockout for the Pr2x7 gene (Pr2x7−/−) and coeval wild-type (WT) mice were fed a high-fat diet (HFD) or normal-fat diet for 16 weeks. NAFLD grade and stage were lower in Pr2x7−/− than WT mice, and only 1/7 Pr2x7−/− animals showed evidence of NASH, as compared with 4/7 WT mice. Molecular markers of inflammation, oxidative stress, and fibrosis were markedly increased in WT-HFD mice, whereas no or significantly reduced increments were detected in Pr2x7−/− animals, which showed also decreased modulation of genes of lipid metabolism. Deletion of Pr2x7 gene was associated with blunted or abolished activation of NLRP3 inflammasome and expression of its components, which were induced in liver sinusoidal endothelial cells challenged with appropriate stimuli. These data show that Pr2x7 gene deletion protects mice from HFD-induced NASH, possibly through blunted activation of NLRP3 inflammasome, suggesting that PR2X7 and NLRP3 may represent novel therapeutic targets.


Introduction
Nonalcoholic fatty liver disease (NAFLD) has become the leading cause of chronic liver disease worldwide [1], as a result of the epidemics of obesity and type 2 diabetes. It encompasses a wide spectrum of disease conditions, from simple steatosis to nonalcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma [2]. Transition from steatosis to NASH is characterized by superimposition of inflammation and hepatocyte degeneration and death, ultimately leading to tissue fibrosis [3]. Lipotoxicity from nontriglyceride fatty acid metabolites is now recognized as a central mechanism driving hepatic inflammation and injury. When flux of free fatty acids (FFAs) from diet, adipose tissue lipolysis, and hepatic de novo lipogenesis is higher than the rate of FFA oxidation or incorporation into triglycerides for storage as lipid droplets or export as VLDL, exceeding FFAs can generate lipotoxic intermediates which may induce oxidative stress, endoplasmic reticulum stress, mitochondrial dysfunction, and overproduction of proinflammatory cytokines and adipokines [4,5]. However, the precise molecular mechanism triggering steatosis progression to NASH has not been elucidated yet.
A growing body of evidence indicates a major role for the purinergic system, particularly extracellular ATP (eATP) signaling through the purinergic receptor 2X 7 (PR2X 7 ), in several inflammatory and fibrotic disorders including NASH [6,7]. In detail, Das et al. showed that Pr2x 7 gene deletion was associated with protection from liver injury in two rodent models of NASH: the toxin-induced model, using coadministration of a high-fat diet (HFD) and a low-dose environmental toxin bromodichloromethane, and the diet-induced model, using a methionine choline-deficient (MCD) diet [8]. Likewise, Hoque et al. reported that mice knockout for Pr2x 7 gene (Pr2x 7 −/− ) as well as wild-type (WT) mice treated with the specific PR2X 7 antagonist A438079 exhibited significantly decreased acetaminophen hepatotoxicity [9]. Antagonism of PR2X 7 with A438079 and Brilliant blue G was also found to attenuate liver fibrosis induced by carbon tetrachloride in mice [10] and common bile duct ligation in rats [11], respectively.
Stimulation of PR2X 7 by eATP results in rapid opening of a ligand-gated cation channel, followed by induction of a cytoplasmic pore via pannexin-1, which triggers K + efflux and allows danger signals to access the cytosol and activate the nucleotide-binding and oligomerization domain (NOD), leucine-rich repeat, and pyrin domain containing (NLRP) 3 inflammasome [12]. The NLRP3 inflammasome is expressed primarily in macrophages and dendritic cells [13], but also in nonhematopoietic tissues, including the liver at the level of hepatocytes, Kupffer cells, endothelial cells, stellate cells, and myofibroblasts [14]. This multiprotein platform is composed of an inflammasome sensor molecule, containing an N-terminal pyrin domain (PYD) and a caspase recruitment domain (CARD). Once activated, it oligomerizes and recruits the PYD-CARD adaptor protein apoptosisassociated speck-like protein containing a CARD (PYCARD or ASC) and the protease procaspase-1. Caspase-1 autoactivation results in the cleavage of prointerleukin-(IL-) 1β and IL-18 into their mature form, which are released [13]. In addition to mediating inflammation via canonical IL-1β and IL-18-dependent mechanisms, the NLRP3 inflammasome regulates cell death through noncanonical caspase-1dependent and independent pathways leading to pyroptosis and pyronecrosis, respectively [15].
We have previously shown that deletion of Pr2x 7 gene attenuated renal disease in mice fed a HFD and that this was associated with blunted upregulation of the NLRP3 inflammasome components NLRP3, ASC, procaspase-1, pro-IL-1β, and pro-IL-18 and reduced the formation of mature caspase-1 [16]. The observation that PR2X 7 was required for eATP-stimulated IL-1β release in Kupffer cells supports the hypothesis that also in the liver the effects of PR2X 7 stimulation are mainly mediated by NLRP3 activation [9]. Indeed, numerous experimental observations suggest an involvement of NLRP3 in transition from steatosis to NASH [14]. Treatment with the NLRP3 inflammasome blocker MCC950 attenuated liver inflammation and fibrosis in two mouse models of NASH, that is, the Alms1 mutant (foz/foz) mice fed with an atherogenic diet and C57BL/6 mice fed with an MCD diet [17]. Likewise, deletion of the NLRP3 gene protected from NASH induced by choline-deficient amino acid-defined diet [18], though another report found exacerbated MCD diet-induced hepatic steatosis and inflammation mediated by gut microbial dysbiosis [19]. Conversely, Nlrp3 knockin mice showed accelerated NASH [18,20], associated with marked hepatocyte pyroptosis [20].
This study aimed at investigating the effect of Pr2x 7 gene deletion on the development of experimental NASH induced by HFD in mice.

Methods
2.1. Design. The study design consisted of (a) in vivo studies, in which we assessed whether disruption of the Pr2x 7 gene attenuates NASH via blunted activation of the NLRP3 inflammasome, and (b) in vitro studies, in which we assessed whether PR2X 7 -dependent activation of the NLRP3 inflammasome occurs also at the level of resident liver cells. In the in vivo studies, adult (aged six weeks), male Pr2x 7 −/− and coeval WT mice were used. The Pr2x 7 −/− mice were of the two currently available strains, that is, the Pfizer strain, backcrossed onto a B6D2 background, which has a neomycin cassette (Neo) inserted into exon 13 [21], and the GlaxoS-mithKline strain, backcrossed onto a C57BL6 background, which has a LacZ gene and Neo inserted into exon 1 [22]. The Pr2x 7 −/− mice of Pfizer strain were purchased from Jackson Laboratory (Bar Harbor, ME), whereas those of the GlaxoSmithKline strain were kindly provided by Professor Francesco Di Virgilio (University of Ferrara, Ferrara, Italy). The reason for using the two strains is that splice variants with reduced function may escape inactivation in selected tissues from both the Pfizer [23] and the GlaxoSmithKline [24] strain. Mice were housed in a germ-free stabularium in accordance with the Principles of Laboratory Animal Care (NIH Publication number 85-23, revised 1985) and with national laws and received water and food ad libitum. The study protocol was approved by the locally appointed ethics committee. Mice from both genotypes were fed for 16 weeks either a HFD (DIO diet D12492, 60% of total calories from fat) or a normal-fat diet (NFD, DIO diet D12450B; 10% of total calories from fat), purchased from Research Diets (Mucedola, Settimo Milanese, Italy). The four groups consisted of 12-14 animals each, half of the Pfizer strain and half of the GlaxoSmithKline strain. At the end of the 16-week period, mice were anaesthetized with intraperitoneal ketamine (Imalgene®, 60 mg/kg body weight) and xylazine (Rompum®, 7.5 mg/kg body weight) and a longitudinal incision of the abdominal wall was performed, a blood sample was obtained, and the liver was removed and weighed. Then, a portion of liver tissue was immediately fixed by immersion in phosphate buffered 4% formaldehyde solution and processed for light microscopy examination, morphometrical evaluation, and immunohistochemistry. The remaining liver tissue was frozen in liquid nitrogen and used for extraction of proteins and total RNA [25]. In the in vitro studies, human liver sinusoidal endothelial cells (LSECs) were plated onto fibronectin-coated dishes and cultured in endothelial cell medium supplemented with 5% FBS, antibiotics, and endothelial cell growth supplement (ScienCell Research Laboratories, Carlsbad, CA), at 37°C in 95% air-5% CO 2 humidified atmosphere. Cells were then incubated for 21 hours with serum-free medium containing 100 ng/ml recombinant human tumor necrosis factor-(TNF-) α (PeproTech, Rocky Hill, NJ) followed by 0.2 mM 2 ′ (3 ′ )-O-(4-benzoylbenzoyl)ATP (BzATP, Sigma-Aldrich, HOMA-IR = homeostasis model assessment-insulin resistance; FFAs = free fatty acids; AST = aspartate transaminase; ALT = alanine transaminase; WT = wild type; Pr2x 7 −/− = knockout for purinergic receptor X2 7 gene; NFD = normal-fat diet; HFD = high-fat diet. * P < 0 001 versus the corresponding NFD-fed mice; † P < 0 001 versus WT mice.  Saint Louis, MO) for 45 min. Cell lysates were then collected [16]. This stimulus was chosen in order to mimic the sterile inflammation occurring in NASH, consistent with a very recent report showing a key role for TNF-α as a mediator of liver inflammation and fibrosis induced by constitutive NLRP3 inflammasome activation in myeloidderived cells [26].
2.2. Liver Morphology/Morphometry. Liver morphology was assessed based on the American Association for the Study of Liver Disease Guidelines [27], as previously reported [25]. Briefly, NAFLD was graded and staged in hematoxylin and eosin-stained sections. NAFLD grading was assessed based on the percentage of parenchyma involved by steatosis (grades 0 to 3 as follows: 0, no fat; 1, <33%; 2, 33-66%; and 3, >66%). The steatosis grade and the presence of inflammation, hepatocyte degeneration (acidophil or Councilman's bodies, ballooning, and Mallory's hyaline) or necrosis, and fibrosis were then considered for NAFLD staging (stages 1 to 4, with stages 3 and 4 corresponding to NASH). Subsequently, samples from mice with NAFLD stage 3 and 4 were graded and staged for NASH. NASH grading was accomplished based on the type of fat (macrovesicular, microvesicular, or mixed) and the extent of inflammation (scored 0 to 3 as follows: 0, no; 1, mild; 2, moderate; and 3, severe) and hepatocyte degeneration or necrosis. NASH staging was performed by assessing the extent and distribution of fibrosis in sections stained with Masson's trichrome.

Biochemistry and ELISA.
Blood samples obtained from experimental animals were analyzed for fasting levels of glucose, with the aid of an automated colorimetric instrument (Glucocard™ SM, A. Menarini Diagnostics, Florence, Italy); cholesterol, triglycerides, aspartate transaminase (AST), and alanine transaminase (ALT), by standard chemical methods (VITROS5.1 FS Chemistry System, Ortho-Clinical Diagnostics, Rochester, NY); free fatty acids (FFAs), using the NEFA C kit (Wako, Osaka, Japan); and insulin, by enzyme immunoassay (Ultrasensitive Mouse Insulin ELISA kit, Mercodia AB, Uppsala, Sweden). The homeostasis model assessmentinsulin resistance (HOMA-IR) index was then calculated from fasting glucose and insulin [16,25]. Hepatic levels of eATP were assessed colorimetrically in liver tissue homogenates using the ATP assay kit (ab83355, Abcam, Cambridge, UK) as per manufacturer's instructions.

Immunohistochemistry and Western Blot Analysis.
Liver content and distribution of the advanced glycation end product (AGE) N ε -(carboxymethyl)lysine (CML) and the 47 kDa cytosolic subunit of neutrophil NADPH neutrophil cytosol factor 1 (NCF1, also known as p47phox) were assessed by immunohistochemistry, together with protein expression of PR2X 7 and NLRP3. The primary and secondary antibodies  Table 1 available online at https://doi.org/10.1155/2017/8962458. Staining was analyzed using an image analysis system (Optimas™6.5, Bioscan, Washington DC) at a fixed color threshold in 20 random fields of the liver at a final magnification of 400x, and results were expressed as the mean percentage of field area occupied by the specific stain [16,25]. Levels of caspase-1 in the liver and LSECs were evaluated by Western blot analysis and normalized to the levels of β-actin. The antibodies used are reported in Supplemental Table 1. Bands were detected by an enzymatic chemiluminescence kit (Immobilon Western, Millipore, Billerica, MA) and quantified by scanning densitometry using a GS-670 Imaging Densitometer (Bio-Rad Laboratories, Hercules, CA) [16]. Scientific, Monza, Italy) following the standard protocol.

Grading and Staging of NAFLD and NASH.
Steatosis was detected only in HFD-fed mice from both genotypes, though it was of higher grade in WT than in Pr2x 7 −/− mice (Figure 1(a)). In detail, NAFLD grading showed that three out of 7 WT-HFD mice fell in grade 3 steatosis, whereas four out of 7 Pr2x 7 −/− -HFD mice were assigned to grade 1 steatosis (Figure 1(b)). Likewise, NAFLD grading showed that, of the WT-HFD mice, three had stage 2 (predominantly microvesicular steatosis with mild lobular inflammation), three had stage 3 (mixed steatosis with lobular inflammation and ballooning degeneration), and one had stage 4 (mixed steatosis with lobular inflammation and ballooning degeneration and macrovesicular steatosis, associated with portal and lobular inflammation, Councilman's bodies, ballooning degeneration, Mallory's hyaline, and fibrosis) NAFLD. Conversely, only one of the Pr2x 7 −/− -HFD animals fell in stage 3 NAFLD, whereas the remaining six mice did not fulfill NASH criteria, as one had stage 2 and five had stage 1 (simple steatosis) NAFLD (Figure 1(c)). Of the four WT-HFD animals classified as having NASH, three showed grade 2 (moderate) and one grade 3 (severe or florid) NASH, whereas one exhibited stage 2 (zone 3 portal/periportal, perivenular/ centrolobular, perisinusoidal/pericellular fibrosis; focal or extensive) and three stage 1 (i.e., as stage 2, except portal fibrosis) fibrosis (Figures 2(a) and 2(b)). The only one Pr2x 7 −/− mouse fulfilling NASH criteria was graded 1 (mild) for NASH and staged 1 for fibrosis (Figures 2(c) and 2(d)).

Markers of Liver Inflammation, Fibrosis, and Lipid
Metabolism. The liver content of CML (Figures 3(a) and  3(b)) and NCF1 (Figures 3(c) and 3(d)) was markedly increased in HFD-fed WT and, to a significantly lesser extent, Pr2x 7 −/− mice. Transcripts of the markers of inflammation and endoplasmic reticulum and carbonyl stress Ccl2, Tnfa, Cxcr3, Adgre1, Ddit3, and Ager, but not Ifng, increased significantly in WT-HFD mice, but not in Pr2x 7 −/− -HFD animals ( Table 2). Likewise, the mRNA expression levels of the extracellular matrix components Fn1 and Col1a1 increased significantly in WT, but not Pr2x 7 −/− mice on a HFD, whereas those Tgfβ increased to a much lesser extent in Pr2x 7 −/− than in WT animals ( Table 2). In response to the HFD, marked changes were observed in the gene expression level of several transcription factors and enzyme of lipid metabolism. In particular, transcript expression of Srebf1 and, to a lesser extent, of Ppara, Pparg, and Nr1h3 increased in WT-HFD mice, but not or significantly less markedly in Pr2x 7 −/− -HFD animals. In response to the HFD, also transcripts of Acaca, Fasn, and Cpt1 increased only or significantly more in WT, as compared with Pr2x 7 −/− mice. In contrast, no significant effect of HFD or Pr2x 7 −/− gene deletion was observed in the mRNA levels of Nr1h2, Acox1, Hmgcr, and Mttp (Table 2).

Liver Expression and Activation of the PR2X 7 /NLRP3
Axis. Staining for PR2X 7 was observed at both the hepatocyte and sinusoidal level in HFD-fed WT mice, whereas it was only faint in those receiving a NFD (Figures 4(a) and 4(b)). Likewise, liver mRNA expression of Pr2x 7 was significantly higher in HFD-versus NFD-fed WT animals, as assessed by RT-PCR (Figure 4(c)). Levels of eATP increased significantly in HFD-fed WT and to a lesser extent Pr2x 7 −/− mice (Figure 4(d)). Also, NLRP3 protein expression was markedly increased in the liver of WT-HFD mice, as shown by the de novo appearance of a strong positive staining for NLRP3 in cells lining the hepatic sinusoids and bile ducts and in infiltrating inflammatory cells, but not in hepatocytes. Conversely, virtually no NLRP3 staining was observed in PR2X 7 −/− -HFD animals (Figures 5(a) and 5(d)). Consistently, Nlrp3 mRNA levels were significantly increased in the liver of WT, but not Pr2x 7 −/− mice fed with a HFD (Figure 5(e)). Similar changes were observed in the hepatic gene expression of Pycard, Casp1, and Il1β, which were increased in WT, but no or significantly less in Pr2x 7 a HFD (Figures 6(a), 6(b), and 6(c)). Finally, Western blot analysis showed that the 20 kDa subunit (active caspase-1) was significantly increased in response to the HFD in WT, but not in Pr2x 7 −/− mice (Figures 6(d) and 6(e)).

Cell Culture Experiments.
Exposure of LSECs to TNFα + Bz-ATP resulted in a marked upregulation of the gene expression levels of Ccl2, Pr2x 7 , Nlrp3, Casp1, and Il1β (Figure 7(a)) and in the activation of caspase-1, as shown by the appearance of a 20 kDa band at Western blot analysis (Figure 7(b)).

Discussion
This study provides the experimental evidence that Pr2x 7 gene deletion results in attenuation of NASH induced by a HFD, a well-established model of the human metabolic syndrome, as well as in blunted expression and activation of the NLRP3 inflammasome, which may ultimately mediate protection from liver injury. The protective effect of Pr2x 7 gene deletion toward NASH is attested by the decreased inflammation, hepatocyte degeneration, and fibrosis at histological examination and the reduced or abolished increment in the protein and/or gene expression of several proinflammatory, prooxidant, and profibrotic markers. These results are consistent with a previous report showing that Pr2x 7 gene deletion attenuates hepatic inflammation and fibrosis in different animal models of NASH [8] and also with the findings that lack or blockade of PR2X 7 prevents other experimentally induced liver diseases [9][10][11].
Interestingly, also hepatic fat accumulation was lower in Pr2x 7 −/− than in WT mice fed with a HFD, a finding attributable to a favorable effect of Pr2x 7 gene knockout on the  expression of transcription factors and enzymes of lipid metabolism. In particular, the upregulation of those involved in de novo lipogenesis, such as Srebf1, Ppara, Pparg, Nr1h3, Acaca, and Fasn, was significantly blunted in Pr2x 7 −/− mice, as compared with WT animals. These data support the concept that the protection afforded by Pr2x 7 gene deletion was related not only to reduction of the proinflammatory and profibrotic response to the HFD but also to decreased lipotoxicity driving hepatocyte injury and release of danger signals such as eATP, which activate the innate immune system. This interpretation is consistent with the lower levels of eATP detected in the liver of Pr2x 7 −/− versus WT mice on a HFD and also with the finding that FFAs (and glucose) are capable of upregulating Pr2x 7 [29], thus implying that the activation of purinergic signaling may occur at an early stage in HFD-fed mice, prior to the occurrence of inflammationdependent hepatocyte degeneration and death.
The attenuation of NAFLD/NASH observed in PR2x7 −/− mice could not be attributed to a better metabolic profile. In fact, a reduced liver injury was observed not only in animals of the GlaxoSmithKline strain, which appeared to be partly protected from metabolic dysfunction induced by the HFD feeding, but also in those of the Pfizer strain, which did not show any improvement in metabolic parameters. The different metabolic responses observed in the two strains are in keeping with previous studies reporting that expression of splice variants which escape inactivation and maintain function, though to a reduced extent, may occur in selected tissue of Pr2x 7 −/− mice from both strains [23,24]. In addition, the finding that mice of the Pfizer strain developed the characteristic features of the metabolic syndrome, including obesity, dyslipidemia, hyperglycemia, and insulin resistance, to the same extent as their corresponding WT animals, is consistent with a previous report by Sun et al. In fact, these authors showed no change in metabolic phenotype (and adipose tissue inflammasome activation) in mice from this strain, though they did not referred these unexpected results to splice variants of Pr2x 7 transcripts [30].
The association of Pr2x 7 gene deletion with blunted hepatic activation of the NLRP3 inflammasome points to a major role of purinergic signaling through PR2X 7 in triggering of this process in the liver of HFD-fed mice. All the three not mutually exclusive models proposed for inflammasome activation might have been responsible for the activation of NLRP3 inflammasome in these animals [31]. In fact, dead hepatocytes may have released eATP [9] and uric acid [32], thus inducing pore formation via PR2X 7 and frustrated phagocytosis with consequent lysosomal disintegration and release of proteases, respectively [31]. In addition, multiple signals may have caused generation of reactive oxygen species (ROS) via NADPH oxidase or mitochondrial sources, which is thought to represent a common final pathway leading to inflammasome activation [31]. These signals include glucose [33], saturated FFAs such as palmitate [34], the lipotoxic FFA metabolite ceramide [35], and modified lipoproteins such as oxidized LDLs, which may also act via formation of cholesterol crystals [36]. In addition, Chatterjee et al. recently showed that, in Kupffer cells, PR2X 7 mediates NADPH oxidase activation by upregulating the expression of the p47 phox subunit and p47 phox binding to the membrane subunit, gp91 phox [37], consistent with our finding of a reduced liver expression of NCF1 (p47 phox) in Pr2x 7 −/− animals. Other possible mechanisms implicated in the PR2X 7 -mediated activation of the NLRP3 inflammasone include the induction of endoplasmic reticulum stress [38] and the impairment of autophagy [39], which have been implicated in the development of NASH [8,40]. Finally, the release of microvesicles from fat-laden hepatocytes may   Figure 8: Potential mechanisms involved in PR2X 7 -mediated expression/activation of the NLRP3 inflammasome. CML = N ε -(carboxymethyl)lysine; DAMPs = damage-associated molecular patterns; eATP = extracellular ATP; ER = endoplasmic reticulum; FFAs = free fatty acids; IL-1β = interleukin-1β; LDLs = low density lipoprotein; MVs = microvesicles; NADPH ox = NADPH oxidase; NF-κB = nuclear factor-κB; NLRP3 = nucleotide-binding and oligomerization domain (NOD), leucine-rich repeat and pyrin domain containing 3; PR2X 7 = purinergic receptor 2X 7 ; PYCARD = the PYD-CARD adaptor protein apoptosis-associated speck-like protein containing a caspase recruitment domain; RAGE = receptor for advanced glycation end products; ROS = reactive oxygen species; TLR = toll-like receptor; TNF-α = tumor necrosis factor-α. promote activation of NLRP3 in surrounding hepatocytes via a paracrine mechanism [41].
As previously shown at the kidney level, also the expression of NLRP3 inflammasome components was blunted in liver tissue of HFD-fed Pr2x 7 −/− mice, thus suggesting that stimulation of PR2X 7 is involved also in priming the inflammasome through activation of nuclear factor-κB (NF-κB) via multiple mechanisms. Activation of NADPH oxidase by PR2X 7 -mediated increase of TNF-α levels through autophagy-linked inflammation may play a major role in this respect [8,37]. Moreover, the decreased levels of CML and Ager and the improved lipid metabolism detected in the liver of these mice suggest that PR2X 7 also modulates the activation of NF-κB induced by carbonyl stress [42] and FFAs [43] via oxidantdependent mechanisms. Activation of procaspase-1 and release of mature IL-1β induced by PR2X 7 -mediated inflammasome activation might also be important in sustaining NF-κB activation via induction of TNF-α [44]. Figure 8 summarizes the potential mechanisms involved in PR2X7-mediated expression/activation of the NLRP3 inflammasome.
The finding that both the expression and the activation of the NLRP3 inflammasome were blunted in Pr2x 7 −/− mice supports the hypothesis that this effect of Pr2x 7 gene deletion was a major mechanism mediating protection from NASH. This interpretation is consistent with previous observations showing that Nlrp3 gene deletion [18] or NLRP3 blockade [17] attenuates hepatic inflammation and fibrosis in animal models of NASH, and also with the evidence that these strategies are effective in attenuating other experimentally induced liver diseases [14].
A debated issue is the contribution of resident cells to inflammasome activation and inflammasome-mediated tissue injury in inflammatory and fibrotic disorders. Indeed, in the liver of WT-HFD mice, the NLRP3 inflammasome was expressed in both resident and nonresident cells. In addition, inflammasome activation was induced in LSECs challenged with TNF-α + BzATP, that is, a primer and an activator of this multiprotein platform. These data are consistent with the concept that also resident cells are involved, as previously shown in the kidneys of these animals [16]. Csak et al. found that both bone marrow-(BM-) derived and non-BM-derived cells contributed to NLRP3 inflammasome activation in a myeloid differentiation primary response gene 88-(MyD88-) dependent manner in NASH induced by MCD diet, though only BM-derived cell-specific MyD88deficiency attenuated liver injury [45]. Conversely, Wree et al. reported that acceleration of NASH was lower in BM-specific than in global Nl3p3 knockin mice, thus pointing to an important contribution of resident cells to liver injury [20].
A limitation of this study is the use of LSECs only to assess whether PR2X 7 -dependent activation of the NLRP3 inflammasome occurs also at the level of resident liver cells. However, our aim was to test this hypothesis in a nonimmune cell type of the liver other than hepatocytes, which did not show any NLRP3 protein expression at IHC.

Conclusion
These data demonstrate a major contribution of PR2X 7 , possibly via activation of the NLRP3 inflammasome, in hepatic inflammation and injury driving transition from steatosis to NASH in the context of NAFLD. This study identifies P2X 7 R and NLRP3 as novel therapeutic targets for liver disease associated with metabolic disorders.