Abstract
Subclinical circulating bacterial endotoxin LPS has been implicated as an important cofactor in the development and progression of nonalcoholic steatohepatitis, but the underlying mechanisms remain unclear. In this study, we demonstrated that 4-wk injection with superlow-dose LPS significantly promoted neutrophil infiltration and accelerated nonalcoholic steatohepatitis progression, including exacerbated macrovesicular steatosis, inflammation, and hepatocyte ballooning in high-fat diet–fed apolipoprotein E knockout mice. This effect could sustain for a month after stoppage of LPS injection. LPS also significantly increased numbers of apoptotic nuclei in hepatocytes and expressions of proapoptotic regulators. Moreover, LPS sustained the low-grade activation of p38 MAPK and inhibited the expression of the upstream MAPK phosphatase 7. By applying selective inhibitors, we demonstrated that the activation of p38 MAPKs is required for neutrophil migration induced by superlow-dose LPS in vitro. Together, these data suggest that superlow-dose LPS may sustain the low-grade activation of p38 MAPKs and neutrophil infiltration, leading to the exacerbation of steatohepatitis.
Introduction
Nonalcoholic fatty liver disease (NAFLD) is a condition ranging from simple lipid accumulation in the liver (steatosis) to steatosis combined with inflammation (nonalcoholic steatohepatitis [NASH]) (1). Steatosis alone is considered a relatively benign and reversible condition. The transition toward NASH represents a key step in pathogenesis, because it will set the stage for further damage to the liver, including fibrosis, cirrhosis, and liver cancer (2). The actual risk factors that drive hepatic inflammation during the progression to NASH remain largely unknown. Therefore, knowledge about events that induce steatohepatitis is of great importance for the diagnosis and treatment of NASH.
A growing body of experimental and clinical data suggests that subclinical low levels of blood circulating endotoxin, most likely derived from mucosal leakages, play an important part in low-grade chronic inflammatory disease such as nonalcoholic liver injury (3). LPS may be capable of stimulating inflammation, cytokine production, and accumulation of inflammatory cells within the liver (4). In fact, slightly elevated circulating endotoxemia has been reported to be associated with a rise in TNF-α gene expression in the hepatic tissue, which supports a role of subclinical superlow-dose endotoxemia in the development of steatohepatitis in obese individuals (5, 6). However, the pathophysiology of this phenomenon is not well understood. Neutrophils constitute the first line of defense against most microorganisms (7). Neutrophils actively recruited to liver sinusoids are a distinguishing feature of endotoxemia and sepsis (8). Their effector functions include phagocytosis, release of proteolytic enzymes, and regulation of the immune response (9). Although facilitating the elimination of invading organisms and their derived products, these functions can also cause severe tissue injury (10, 11). Despite these findings, the impact of superlow grade endotoxemia on liver neutrophil infiltration during NASH progression remains elusive. Therefore, we hypothesized that subclinical superlow-level circulating LPS might sustain a low-grade systemic inflammation and facilitate the recruitment of neutrophils to the liver tissue, thereby promoting progression of NASH.
To address this question, we used a high-fat diet (HFD)-fed apolipoprotein E knockout (ApoE−/−) mice model, which results in graded steatosis, inflammation, and fibrosis, closely resembling the pathophysiology of progressive NAFLD in humans (12). We studied the exacerbating effect of chronic superlow-dose LPS injection on liver histopathology. We also determined whether steatohepatitis may be sustained when LPS injection is terminated. At the cellular level, the effect of superlow-dose LPS on neutrophil migration was evaluated in vitro. We provide evidence that superlow-dose LPS induces elevated neutrophil migration in vitro as well as liver neutrophil infiltration during the progression of NASH, through sustaining the low-grade activation of MAPKs p38. Our data support a model of sustained chronic low-grade inflammation and exacerbation of NASH pathogenesis in mice that experience superlow-dose endotoxemia.
Materials and Methods
Reagents
Abs against phospho-p38 MAPK (Thr180/Tyr182), p38 MAPK, phospho-JNK (Thr183/Tyr185) MAPK, JNK, phospho-ERK (Thr202/Tyr204) MAPK, ERK, MAPK phosphatase 7 (MKP7), poly(ADP-ribose) polymerase, and active caspase-3, and HRP-conjugated secondary Abs were purchased from Cell Signaling Technology (Beverly, MA). Mouse mAb specific for GAPDH was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). PE-labeled anti-mouse Ly-6G Ab was provided by BioLegend (San Diego, CA). Fat stain kit was provided by Newcomer Supply (Middleton, WI). All other chemicals, unless otherwise specified, were from Sigma-Aldrich (St. Louis, MO).
Animals
Specific pathogen-free male, C57BL/6J ApoE−/− mice were purchased from The Jackson Laboratory (Sacramento, CA) and used at 8–12 wk of age in the experiments described. Thirty-two mice were randomly divided into two groups, as follows: LPS group, injected i.p. with 4 ng/kg body weight (∼100 pg/mouse) purified LPS dissolved in endotoxin-free PBS; control group, injected i.p. with the same volume of PBS. Mice were injected i.p. every 3 d with either PBS or LPS. Upon acclimation with injection for 10 d, the mice were fed with an HFD (TD90221; Harlan Tek-lad, Madison, WI) together with i.p. injection of either LPS or PBS every 3 d for 4 wk. A batch of mice was harvested for analyses. To determine whether the systemic effects of LPS injection may last postinjection, another batch of mice was continually fed with an HFD, but without further LPS injection for an additional 4 wk. Animals were housed under a 12:12-h light/dark cycle and permitted ad libitum consumption of water and diet. Animal protocols were approved by the Animal Care and Use Committee of Virginia Polytechnic Institute and State University.
Biochemical analyses
The plasma and liver samples of each mouse were collected and stored at −80°C for further analysis. Plasma alanine aminotransferase, lactate dehydrogenase, free fatty acid (FFA), triglyceride (TG), and hepatic TG contents were determined using their respective detection commercial kits (Biovision, Milpitas, CA), in accordance with the manufacturer’s protocols. Myeloperoxidase (MPO) activities in liver extracts were quantified using Duoset ELISA development kit (R&D Systems, Minneapolis, MN). Briefly, 50–100 mg tissue was homogenized in 1 ml potassium phosphate buffer (50 mM, pH 6.0) on ice. Then homogenates were centrifuged at 12,000 × g for 10 min at 4°C. MPO activity and protein concentration (DC Protein Assay; Bio-Rad Laboratories, Hercules, CA) in the supernatant were assayed, according to the manufacturer’s instructions, respectively. The results are expressed as MPO ng/mg protein.
Histological assessment
Fresh liver tissue was embedded in Tissue Tek OCT, rapidly frozen in liquid nitrogen, and then stored at −80°C for preparation of frozen sections. Five-micrometer–cut sections were stained with H&E and Oil Red O for histological analysis and lipid under light microscopy (Olympus, Tokyo, Japan), respectively. The Pathology Committee of the NASH Clinical Research Network (13) provided guidance and recommendations for the NAFLD activity scoring by semiquantitatively evaluating the following histological features: steatosis (<5% = 0; 5–33% = 1; 33–66% = 2; >66% = 3); lobular inflammation (none = 0; <2 foci = 1; 2–4 foci = 2; >4 foci = 3); and hepatocellular ballooning (none = 0; few = 1; prominent = 2). All features were scored blindly based on at least five samples per group and five fields of vision in each sample.
Immunofluorescence and immunohistochemistry
Serial 5-μm sections were fixed with PBS containing 4% paraformaldehyde for 15 min. Slides were then incubated in serum-blocking solution for another 30 min and incubated overnight in anti-Ly6G Ab or anti-MOMA2 Ab, as specified in the figure legends. Following removal of the primary Ab, fluorescent secondary Abs were used for immunofluorescence. DAPI (Roche) was used for nuclear staining and observed through Inverted Fluorescence Microscopy (Olympus, Tokyo, Japan). Ly6G-expressing cells were then counted by ImageJ software (Research Services Branch, National Institutes of Health, Bethesda, MD).
Apoptotic cells in liver were detected by immunocytochemistry using Apo-BrdU-IHC in situ DNA fragmentation assay kit (Bio-Rad AbD Serotec, Raleigh, NC), according to the manufacturer’s protocols. Visualization of the reaction was performed using diaminobenzidine as the chromogenic substrate and methyl green as a counter stain. Cells were considered positive for apoptosis, as assessed by the following two criteria: a positive nuclear stain and morphologic evidence of apoptotic cells. Five fields were counted per slide. The apoptotic index was calculated as follows: number of apoptotic cells/(number of apoptotic cell + number of negative cells).
RNA extraction and real-time PCR
Total RNA was extracted from liver tissues using TRIzol reagent, and reverse transcription of mRNA was performed using the SuperScript reverse-transcriptase reagent kit (Applied Biosystems, Foster City, CA), according to the manufacturer’s recommendations. Quantitative PCR was performed using a real-time PCR system (Applied Biosystems CFX96), and reactions were performed using SYBR Green Master Mix (Bio-Rad) with gene-specific primers. To normalize expression data, GAPDH was used as an internal control gene.
Neutrophil preparation and chemotaxis assay
Male C57BL/6J mice were purchased from The Jackson Laboratory and used at 6 wk old. The mice were sacrificed by cervical dislocation. The femurs were removed, and the femoral bone marrow was harvested by flushing with 5 ml PBS buffer. The erythrocytes were removed using hypotonic shock, and the leukocyte population was resuspended in PBS buffer. The purity of isolated neutrophils was routinely >95%, as assessed by light-microscopic analysis of the cells stained with Diff-Quick (Wako Pure Chemical Industries, Osaka, Japan), and >95% viable, as assessed by a trypan blue exclusion test. Harvested neutrophils were pretreated with 20 μM SB203580 (p38 inhibitor) or 20 μM SP600125 (JNK inhibitor) for 2 h, and then treated with or without 100 pg/ml LPS and incubated for 24 h.
Chemotaxis assays were performed using HTS Transwell-96 plates with 3-μm pores (Corning, Lowell, MA). After LPS stimulation, mouse neutrophils were added into the upper wells (2 × 105 cells/well), and PBS or 100 ng/ml recombinant mouse chemokine CXCL2/MIP-2 (R&D Systems) was added to the bottom chamber. The plates were incubated at 37°C for 30 min. Because MIP-2 induces chemotaxis of neutrophils, but not monocytes, only neutrophils will migrate into the matrigel (14). After removing the upper wells, migrated cells were stained with Wright-Giemsa and counted by light microscopy after coding the samples. For each experiment, the number of neutrophils in five fields (×200) was counted and averaged at each incremental level.
Western blot
Liver tissues or neutrophil lysates containing 30 μg total protein were loaded onto 10% SDS polyacrylamide gels. After 90 min of electrophoresis, the proteins were transferred onto a polyvinylidene difluoride membrane in ice for 2 h. The membrane was blocked with 5% skim milk, which dissolved in trihydroxymethyl aminomethane buffer salt plus Tween 20 (TBST). The membrane was next incubated at 4°C overnight with primary Abs. Secondary Abs conjugated with HRP were incubated for 2 h at room temperature. Signals were detected by ECL (Thermo Fisher Scientific, Waltham, MA). GAPDH was used for normalization. The density of the specific bands was quantified using ImageJ software.
Statistics
The data are expressed as mean ± SEM. Statistical differences were analyzed by Student's t test using SPSS version 18.0 (SPSS, Chicago, IL). Differences were considered as statistically significant if p < 0.05.
Results
Subclinical superlow-dose LPS exacerbates lipid accumulation in liver and plasma
Although recent studies have suggested potential connection between low-grade endotoexmia and chronic inflammation, dosages of LPS injection in animal models tend to be in the mg/kg range (15–17), which are significantly higher than the clinical concentrations (∼ng/kg) routinely observed in mice and humans with chronic conditions (18, 19). To mimic the pathophysiological effects of low-grade endotoxemia, the dosages of LPS used in this study were much lower than previous studies. We set up the i.p. injection of ApoE-deficient mice with either PBS or LPS (4 ng/kg body weight), as the i.p. route was known to cause low-level circulating endotoxemia (20, 21). Mice were fed with an HFD and continued with the injection of PBS or LPS every 3 d for 4 wk. Fifty percent of the animals in each group were sacrificed, and the rest continued on the same diet but without injection for another 4 wk, to further examine the potential lasting memory effects of low-grade endotoxemia. Liver pathology, neutrophil infiltration, and potential molecular mechanisms in vivo and in vitro were systematically evaluated.
As shown in Fig. 1A, Oil Red O staining clearly indicated that the HFD feeding induced lipid accumulation in the liver, especially in hepatocytes. In line with a previous report with chronic injection of LPS (22), we documented that 4-wk injection with superlow-dose LPS exacerbated HFD-induced lipid accumulation in the liver (Fig. 1B). The enhanced liver TG accumulation in the LPS group was transient and not sustained after the stoppage of LPS injection at the 8-wk time point. There was no significant difference in plasma TG content between the two groups (Fig. 1C). The plasma FFA levels were significantly elevated in the LPS-injected group at the 4-wk time point as compared with the PBS group. Strikingly, the significantly elevated plasma FFA levels in the LPS-injected group were sustained even 1 mo after the stoppage of LPS injection at the 8-wk time points (Fig. 1D).
Chronic injection of superlow-dose LPS did not affect the body weight gain in HFD-fed ApoE−/− mice, accompanied by no significant difference in liver weights (percentage of body weight) during the experimental period (Table I). Compared with the PBS control group, the plasma levels of lactate dehydrogenase were transiently increased in the LPS group at wk 4 as compared with the PBS group. The plasma alanine aminotransferase activities in the LPS group were also higher than that of control group, but without significant difference.
. | LPS . | Control . | p Value . |
---|---|---|---|
Initial body weight (g) | 22.04 ± 0.29 | 22.26 ± 0.41 | 0.669 |
4 wk | |||
Body weight (g) | 30.22 ± 1.59 | 32.34 ± 1.05 | 0.282 |
Liver/body weight (%) | 6.00 ± 0.10 | 5.66 ± 0.26 | 0.290 |
Plasma LDH (U/L) | 23.25 ± 2.07 | 10.93 ± 0.90 | <0.001 |
Plasma ALT (U/L) | 38.01 ± 3.27 | 31.52 ± 1.56 | 0.102 |
8 wk | |||
Body weight (g) | 35.33 ± 1.25 | 35.34 ± 1.84 | 0.995 |
Liver/body weight (%) | 5.97 ± 0.13 | 5.79 ± 0.21 | 0.456 |
Plasma LDH (U/L) | 16.38 ± 5.45 | 13.94 ± 3.73 | 0.315 |
Plasma ALT (U/L) | 43.22 ± 2.72 | 41.55 ± 2.13 | 0.304 |
. | LPS . | Control . | p Value . |
---|---|---|---|
Initial body weight (g) | 22.04 ± 0.29 | 22.26 ± 0.41 | 0.669 |
4 wk | |||
Body weight (g) | 30.22 ± 1.59 | 32.34 ± 1.05 | 0.282 |
Liver/body weight (%) | 6.00 ± 0.10 | 5.66 ± 0.26 | 0.290 |
Plasma LDH (U/L) | 23.25 ± 2.07 | 10.93 ± 0.90 | <0.001 |
Plasma ALT (U/L) | 38.01 ± 3.27 | 31.52 ± 1.56 | 0.102 |
8 wk | |||
Body weight (g) | 35.33 ± 1.25 | 35.34 ± 1.84 | 0.995 |
Liver/body weight (%) | 5.97 ± 0.13 | 5.79 ± 0.21 | 0.456 |
Plasma LDH (U/L) | 16.38 ± 5.45 | 13.94 ± 3.73 | 0.315 |
Plasma ALT (U/L) | 43.22 ± 2.72 | 41.55 ± 2.13 | 0.304 |
Values represent mean ± SEM; n = 16 for initial body weight and n = 8 for the other parameters.
ALT, alanine aminotransferase; LDH, lactate dehydrogenase.
Superlow-dose LPS initiates sustained chronic inflammation and exacerbates liver steatohepatitis
To determine the effect of superlow-dose LPS on hepatic morphology, we performed H&E staining of the liver sections. In general, the histopathological features required for a diagnosis of NASH in humans include macrovascular steatosis (hepatocyte fat accumulation), lobular inflammation around hepatocytes and hepatic sinusoids, and hepatocyte ballooning (23). At wk 4, mice injected with PBS did not show significant inflammation (Fig. 2A, 2B). In contrast, ApoE−/− mice injected with superlow-dose LPS developed more severe hepatic inflammation compared with controls, as reflected by increased inflammatory cell infiltration and significantly elevated expression of inflammatory markers, such as TNF-α, MCP-1, and IL-6 (Fig. 2B). Strikingly, the significant elevation of liver inflammatory mediators was remarkably sustained, 1 mo after the stoppage of LPS injection at the 8-wk measurement point, suggesting an inflammatory memory effect following the chronic injection of superlow-dose LPS (Fig. 2B).
Together, collective disease scores that combine the evaluation of steatosis, inflammatory neutrophil infiltration, and hepatocyte ballooning reveal a significant elevation of disease scores in the liver of LPS-injected mice at the 4-wk time point. The exacerbated liver pathology can still be detected 1 mo after the stoppage of LPS injection at the 8-wk time point (Fig. 2C).
Sustained neutrophil infiltration in liver of mice injected with superlow-dose LPS
Neutrophils and monocytes represent one of the most prominent components of the innate immune system (9). To further explore the hypothesis that neutrophils are activated in response to LPS stimulation, we assessed the activation marker of neutrophils, Ly6G, in liver tissues using immunofluorescence (Fig. 3A). The Ly6G-positive cells were significantly elevated in LPS-treated mice as compared with control mice (Fig. 3B). In line with this, biochemical analysis revealed that the injection of superlow-dose LPS caused a 2.7-fold increase of liver MPO levels in HFD-fed ApoE−/− mice as compared with PBS-injected ones (Fig. 3C). Significantly, the elevation of liver neutrophils and MPO levels was sustained after 1 mo after the stoppage of LPS injection at the 8-wk time points. To further address whether the increased neutrophil infiltration may be due to elevated levels of neutrophil chemoattractants, we performed ELISA analyses of MIP2 and KC. Indeed, we found significantly higher levels of plasma KC and MIP2 as well as higher levels of liver MIP2 at the 8-wk time point, long after the withdrawal of LPS injection, further supporting the lasting memory effects of subclinical low-dose endotoxemia (Fig. 3D). Furthermore, we also observed elevated levels of macrophages within the liver tissues injected with subclinical dose endotoxin both at the 4-wk and the 8-wk time points (Fig. 3E). Together, our data suggest that superlow-dose LPS challenge may establish sustained memory inflammatory state in mice, leading to hepatic inflammation.
Elevated cellular apoptosis in liver tissues from mice injected with superlow-dose LPS
Neutrophil-derived MPO has powerful proapoptotic effects, partly attributable to the generation of reactive oxygen species (24). Given sustained neutrophil infiltration and elevated liver MPO levels in mice preconditioned with superlow-dose LPS, we next tested the hypothesis that superlow-dose LPS challenge may cause a sustained apoptotic response in liver. Indeed, we observed a significantly increased number of apoptotic nuclei within liver hepatocytes with classical cubical shapes in LPS-injected mice as compared with control mice (Fig. 4A, 4B). Such effect was not only apparent in mice harvested at the 4-wk time point, but also in mice harvested 1 mo after the stoppage of LPS injection at the 8-wk time point.
We then determined the gene expression levels of several apoptotic markers such as membrane apoptotic receptor Fas and its ligand FasL, intracellular proapoptotic regulator Bax, and anti-apoptotic regulator Bcl-2. We observed a lasting and sustained elevation of Fas, FasL, and Bax mRNAs in liver of mice injected with superlow-dose LPS at both the 4-wk and 8-wk time point (Fig. 4C). Our immunoblotting data demonstrated that HFD-fed ApoE−/− mice injected with superlow-dose LPS had elevated levels of active caspase-3, and increased cleavage of poly(ADP-ribose) polymerase, as compared with vehicle-treated mice, both at the 4-wk and 8-wk time points (Fig. 4D). These results reveal that superlow-dose LPS conditioning may lead to a lasting memory effect of liver cell apoptosis.
Superlow-dose LPS initiates sustained activation of the p38 MAPK signaling axis in liver, through reducing the negative feedback phosphatase MKP7
MAPKs play essential roles during the processes of neutrophil migration as well as cellular apoptosis (25). Next, we tested whether superlow-dose LPS injection may cause sustained activation of MAPKs in liver tissues. Indeed, Western-blotting analyses revealed that there were significantly elevated levels of phosphorylated p38 and JNK in liver tissues from mice injected with superlow-dose LPS (Fig. 5). Intriguingly, the memory effect of superlow-dose LPS injection on the phosphorylation of p38 and JNK was still apparent 1 mo after the stoppage of LPS injection.
Based on systems analyses of analogous systems, one of the potential hypotheses for maintaining sustained memory and activation is the removal of negative feedbacks. In the context of MAPK activation, the negative feedback that dampens its sustained activation is the MAPK phosphatases (26). Indeed, we observed that superlow-dose LPS potently reduced the protein levels of MKP7, a selective p38/JNK MAPK phosphatase (27). The suppression of MKP7 was apparent 1 mo after the stoppage of superlow-dose LPS injection. Our data suggest that superlow-dose LPS may sustain a memory state of MAPK activation in liver tissues through the removal of negative modulator MKP7.
Sustained activation of p38 MAPKs is involved in enhanced neutrophil migration induced by superlow-dose LPS in vitro
Consistent with in vivo results, we found that p38 MAPKs were activated in isolated neutrophils after challenges with superlow-dose LPS for 24 h (Fig. 6C). To identify whether the enhanced neutrophil migration may be due to, or at least partly, the sustained activation of p38 MAPKs by superlow-dose LPS, we investigated the effects of the p38 MAPK inhibitors on neutrophil migration in vitro. As expected, superlow-dose LPS exposure induced a significant migration of neutrophils toward the chemoattractant MIP-2 (Fig. 6A, 6B). Application of either SB203580 (p38 inhibitor) or SP600125 (JNK inhibitor) potently inhibited the chemotactic activity of neutrophils challenged with superlow-dose LPS (Fig. 6A, 6B).
To further evaluate whether neutrophil may retain an activated memory state in vivo in mice preconditioned with superlow-dose LPS, we examined the CD14 levels of circulating neutrophils in HFD-fed ApoE−/− mice preconditioned with either PBS or superlow-dose LPS. ApoE−/− mice fed with HFD were injected with either PBS or LPS for 1 mo. Following the stoppage of injection, mice were continually fed an HFD for an additional month. Blood and bone marrow were harvested, and neutrophil activation status was evaluated by flow cytometry through the surface expression levels of CD14. Such ex vivo identification of activated neutrophil may serve as important evidence of in vivo programming of neutrophils by superlow-dose LPS. As shown in Fig. 6D, HFD-fed mice preconditioned with superlow-dose LPS retain the activated neutrophil memory state, as reflected in significantly higher numbers of activated neutrophils with CD14 expression. Collectively, our data suggest a novel paradigm of low-grade inflammation dynamics (Fig. 7).
Discussion
In this report, we provided evidence that reveals a memory state of low-grade inflammation in liver tissues of HFD-fed ApoE−/− mice chronically injected with superlow-dose LPS. Superlow-dose LPS may sustain and exacerbate the low-grade inflammatory state via sustained p38 MAPK signaling circuit due to the removal of negative modulator MKP7, subsequently enhanced leukocyte infiltration, MPO release, and hepatocyte apoptosis. These mutually propagating events may lead to the exacerbation of NASH progression.
Our study shed light on a novel proinflammatory circuitry in animals challenged with subclinical superlow-dose endotoxin, a phenomenon increasingly recognized in the biomedical field (28). Although empirical knowledge reckons that subclinical superlow-dose LPS may be capable of inducing liver inflammation, almost all existing studies have used significantly higher dosages of LPS that may bear little pathophysiological relevance in the context of low-grade chronic inflammatory disease (15–17). Most of these studies conclude a transient inflammatory effect triggered by LPS, followed by a tolerant anti-inflammatory state (29). Therefore, these studies could not reconcile the sustained low-grade inflammation seen in clinical settings. Our previous in vitro study reveals an important paradigm of priming and tolerance in innate leukocytes challenged with varying dosages of LPS (30). Instead of causing a transient inflammatory response in monocytes treated with a higher-dose LPS, as commonly reported in the literature (31), a sustained nonresolving low-grade inflammatory state can be established in cells treated with a subclinical superlow-dose LPS (32, 33). Capitalizing on this new concept, we extend our mechanistic in vitro studies in vivo, and demonstrated in this work that chronic injection of superlow-dose LPS can sustain a memory low-grade proinflammatory state in liver tissues of HFD-fed ApoE−/− mice. We used the well-established steatosis-prone model of ApoE-deficient mice, as previous studies suggest that hyperlipidemia is the critical first hit and an essential driver for the development of steatosis and atherosclerosis (34, 35), and that the low-grade circulatory inflammatory components are subsequent facilitators not capable of initiating the development of steatosis or atherosclerosis in the absence of hyperlipidemia. We employed the subclinical endotoxemia model with chronic injection of superlow-dose endotoxin, compatible with circulating levels of endotoxemia reported in humans and experimental animals with chronic inflammatory diseases, such as atherosclerosis and diabetes (36–39). Our data clearly support the role of circulating chronic low-grade inflammatory agent endotoxemia during the exacerbation of liver steatosis through elevating systemic low-grade inflammation, sustaining low-grade nonresolving p38 activation, and facilitating liver tissue leukocyte infiltration.
It is also interesting to note that subclinical endotoxemia not only sustained chronic low-grade inflammation, but also elevated tissue levels of FFA. Unlike conjugated TG that preferentially affects cellular metabolism, FFA have been shown to have additional important roles in inflammation. Due to their small and soluble nature, FFA serve as de facto inflammatory mediators through the activation of key inflammatory pathways that include NF-κB, protein kinase C, and others in both inflammatory cells and hepatocytes (40–42). In turn, these inflammatory processes may further facilitate the generation of FFA and perpetuate the nonresolving low-grade inflammation, even after the withdrawal of subclinical endotoxemia.
Our data suggest a fundamental biochemical mechanism responsible for the sustained memory in liver tissues, in that superlow-dose LPS may establish sustained activation of p38 MAPKs through the removal of the negative modulator MKP7. It was previously reported that MKP7 shuttles between the nucleus and the cytoplasm and specifically suppresses the activation of p38 instead of ERK (43, 44). In this study, we observed in vivo that decreased MKP7 expression is accompanied by sustained p38 MAPK activation in liver of mice chronically injected with superlow-dose LPS. Our complementary in vitro studies with isolated neutrophils further confirmed this notion.
Functionally, this study reveals an integrated circuit that couples the sustained activation of p38 MAPK signaling circuit with sustained recruitment and sequestration of neutrophils. There is increasing appreciation for the role of elevated neutrophil recruitment in chronic liver disease (9). Neutrophil recruitment is an elaborate process in which circulating neutrophils are attracted to the site of inflammation by following chemokine trails laid out by microbes or host cells (11). MAPKs such as p38 have been implicated in the enhanced neutrophil chemotaxis (25, 45). Our current study provides evidence that supports a critical role of sustained MAPK signaling in enhanced neutrophil migration following challenges with superlow-dose LPS both in vivo and in vitro.
The activation of cellular p38 MAPK circuit is also responsible for the release of neutrophil MPO, an additional potent neutrophil-recruiting mediator (46). Furthermore, MPO-derived oxidants may diffuse into hepatocytes and trigger intracellular oxidative stress. Moreover, neutrophils can express Fas ligand and kill hepatocytes through an apoptosis-induced mechanism (47). In this work, we demonstrated that liver tissues from mice chronically injected with superlow-dose LPS exhibit sustained elevation of several inflammatory and apoptotic genes, even long after the stoppage of LPS injection. Our study provides functional evidence that supports the innate memory inflammatory state in vivo (Fig. 7).
Due to the heterogeneous nature of NAFLD, it is unlikely that NASH pathogenesis is dependent upon any one single factor. Currently, the “multiple parallel hits” theory reflects more precisely the pathophysiological factors behind the progression from steatosis to NASH (48). The present study suggests that subclinical circulating superlow-dose endotoxin may be involved in sustaining the inflammatory responses, sustaining leukocyte infiltration, and contributing to subsequent hepatocyte apoptosis and steatohepatitis. We realize that other unidentified factors may also be affected by subclinical circulating endotoxemia and further sustain the systemic inflammation and subsequent liver damage. In addition to activated inflammatory mediators such as cytokines, chemokine, and FFA, sustained inflammatory environment may also alter the overall landscape of gut microbiota, which then may further favor the mucosal leakages and endotoxemia. Future studies are warranted to systemically examine the microbiota dynamics during the initiation and maintenance of nonresolving low-grade inflammation and related complications.
In summary, our present study reveals a novel circuit triggered by circulating subclinical superlow-dose LPS that involves sustained low-grade activation of p38, systemic low-grade inflammation, and leukocyte infiltration into liver tissues (Fig. 7). Coupled with high-fat diet, this secondary systemic low-grade inflammation may be critically involved in exacerbating the pathogenesis of liver steatosis.
Footnotes
This work was supported by National Institutes of Health Grant R01 HL135835 (to L.L.), National Natural Science Foundation Grants 81372994 and 81172655, and China Scholarship Council Grant 201308440005 (to H.G.).
References
Disclosures
The authors have no financial conflicts of interest.