Abstract

   Acute liver failure is characterized by the rapid development of liver
   dysfunction and remarkably high mortality. Accumulating evidence
   suggests that soyasaponin possesses potential anti-inflammatory
   activities. Here, we aimed to investigate the potential role of
   soyasaponin II in acute liver failure and establish the underlying
   mechanism.

   Methods: Lipopolysaccharide/D-galactosamine (LPS/GalN) was employed to
   induce acute liver failure. We applied liquid chromatography and mass
   spectrometry (LC/MS) to characterize the changes of soyasaponin II
   levels in the cecal content and liver. Transcriptomics and proteomics
   analysis were used to evaluate the functional molecule mediated by
   soyasaponin II in macrophages.

   Results: LPS/GalN administration markedly decreased fecal and hepatic
   soyasaponin II levels. Soyasaponin II treatment protected mice against
   LPS/GalN induced acute liver injury. Additionally, soyasaponin II
   markedly diminished Y-Box Binding Protein 1 (YB-1) phosphorylation and
   nuclear translocation, Nlrp3 inflammasome priming, and interleukin 1β
   (Il-1β) production in macrophages. Phosphorylated YB-1 could activate
   Nlrp3 mRNA transcription by binding the promoter region. Finally,
   immunofluorescence analysis showed elevated p-YB-1 nuclear
   translocation in macrophages of acute liver failure patients compared
   to controls.

   Conclusion: Our data shows that soyasaponin II which serves as a novel
   inhibitor for YB-1 phosphorylation and Nlrp3 inflammasome priming could
   protect mice against LPS/GalN induced acute liver failure.

   Keywords: acute liver failure, soyasaponin II, Nlrp3-inflammasome, YB-1

Introduction

   Acute liver failure (ALF) is a severe and complicated clinical
   condition, and patients can deteriorate sharply when the innate immune
   system is amplified inappropriately [49]^1. High morbidity and
   mortality is associated with multiple etiologies including virus
   hepatitis, acetaminophen overdose, indeterminate elements, and Wilson
   disease (hepatolenticular degeneration) [50]^2^-[51]^3. Of note, the
   pathogenesis of ALF includes not only direct liver damage but
   immunologically mediated processes that are commonly triggered by
   various causes [52]^4^-[53]^5. However, effective strategies for
   identifying pharmacologic targets in order to ameliorate the excess
   inflammation in ALF are still needed.

   Recent prominent research efforts have advanced our understanding of
   the role of inflammasome in ALF [54]^6^-[55]^8. For example, Ilyas et
   al reported that Il-1β overproduction due to inflammasome activation in
   macrophages is recognized as the main contributor for ALF in
   preclinical model [56]^9. Nonetheless, the underlying molecule
   associated with the activation of inflammasome in ALF remains elusive.
   Because inflammasome as an intermediary agent contributes to such
   severe immunoreaction as previously reported [57]^10^-[58]^12, it is
   imperative to understand its elaborate regulatory mechanism.

   Mammalian Y-box binding protein 1 (YB-1) is a versatile protein which
   contains an evolutionarily conserved cold-shock domain (CSD), binds
   DNA/RNA sequences preferentially, and orchestrates transcription as
   either a repressor or activator [59]^13. Previous studies pointed to
   YB-1 as possibly relevant to the apoptotic pathway, DNA damage
   recognition, and gene repair, leading to variety of unfavorable
   outcomes such as sarcoma invasion, metastasis, and cancer drug
   resistance [60]^14^-[61]^16. It has also been reported that AKT
   phosphorylates YB-1 at Ser102 located in the cold shock domain, which
   affects the growth of breast cancer cells [62]^17. Notably, YB-1 also
   showed increased expression in certain inflammatory diseases
   [63]^18^-[64]^19, but its role in the innate immune system is still
   equivocal.

   Legumes in our diet contain many compounds, but soyasaponin A and B are
   the most prolific phytochemicals in soybeans [65]^20. It is
   increasingly feasible to take advantage of soyasaponins for their
   salutary and bioactive anti-virus, antioxidant, and anti-carcinogenic
   characteristics, but it is their anti-inflammatory properties that
   attracted our attention [66]^20^-[67]^21. This study aims to thoroughly
   explore the influence of a certain type of soyasaponin (soyasaponin II,
   SSII) on the immunological alterations during endotoxin induced acute
   liver failure, as well as to search for a pivotal molecule that is
   mechanically dependent on this nutrient in order to identify a
   potential clinical therapeutic target.

Materials and Methods

Animal model

   Eight- to ten-week-old male C57BL/6J mice were used. Mice were given
   oral administration of soyasaponin II (SSII; ChromaDex, 5 mg/kg
   dissolved in dimethyl sulfoxide (DMSO)) once a day for three
   consecutive days. Control animals received the equivalent volume of
   DMSO. On the third day, the mice received intraperitoneally
   administered D-galactosamine (GalN; Macklin, 700 mg/kg) in combination
   with Lipopolysaccharide (LPS, isotype O111:B4; Sigma-Aldrich, 10
   μg/kg). 6 h after the LPS/GalN injection, the mice were sacrificed for
   the following experiments. For acetaminophen overdose experiment, mice
   were pretreated with or without SSII for three days and received
   intraperitoneal injection with acetaminophen (300 mg/kg, dissolved in
   PBS). After 24 h, the mice were sacrificed for the following
   experiments. All experimental procedures were in accordance with the
   National Institutes of Health guidelines and were approved by the local
   Animal Care and Use Committee of the Southern Medical University.

Soyasaponin II analysis

   Cecal content and liver were collected and a liquid chromatography-mass
   spectrometry (LC-MS) system was used to analyze SSII levels. Water was
   added to the sample at 9-fold volume/weight and extracted
   ultrasonically for 10 min, then methanol was added and the samples were
   centrifuged at 13000 rpm for 10 min at 4 °C. The supernatant was
   concentrated by drying under nitrogen, and then dissolved into mobile
   phase for analysis by HPLC. Chromatographic separation was conducted on
   a Thermo Scientific Prelude SPLC system, and Thermo TSQ Vantage triple
   quadrupole mass spectrometer was used for detection. The
   chromatographic column was Hypersil gold 50*2.1 mm, 1.9 μm. Mobile
   phase A was comprised of a 0.1% formic acid water solution and mobile
   phase B was methanol. The gradient elution was carried out according to
   the following procedure: 0-1.25 min, 90% A and 10% B; 1.25-2.25 min,
   20% A and 80% B; 2.25-5.25 min, 5% A and 95% B; 5.25-6.25 min, 90% A
   and 10% B. Data acquisition and analysis were performed with
   TraceFinder^TM software version 3.3 sp1 (Thermo Fisher Scientific
   Corp., USA).

Transcriptome analysis

   The RNA from hepatic macrophages was extracted and RNA quality was
   detected by NanoDrop 2000. High-quality RNA were used for cDNA
   libraries construction and sequencing (Wuhan Bioacme Technologies
   Corp.). RNA sequencing libraries were generated using the Kapa RNA
   library prep kit for illumina with multiplexing primers, according to
   the manufacturer's protocol. Then, we performed sequencing on an
   Illumina NovaSeq sequencer to obtain raw data. The reference genome
   (GRCm38.94) was used for quality controlled raw data, and quantitation
   of gene expression were conducted. We employed R package
   (ClusterProfiler/enrichplot) to perform the enrichment analysis of the
   KEGG signaling pathway by GSEA (Gene Set Enrichment Analysis) method as
   described previously [68]^22^-[69]^23. Briefly, we first calculate
   log[2] fold change (DMSO/SSII) between the two groups of all genes
   expression. Then, GSEA was carried out to analyze the differences
   between two groups (DMSO/SSII). The gene sets were referring to KEGG
   (Kyoto Encyclopedia of Genes and Genomes) database. The ridgeplot will
   visualize expression distributions of core enriched genes for GSEA
   enriched categories. The parameters used in the pre-ranked GSEA were
   the classic enrichment statistic, with 1,000 gene set permutations, and
   a false discovery rate (FDR) <0.25 was recognized as statistically
   significant.

Proteomics Analyses

   Proteins of isolated hepatic macrophages in SSII treated or non-treated
   group (n=10) were mixed. The sample was prepared as describe previously
   [70]^24. In brief, according to the standard filter-aided sample
   preparation (FASP) protocol, the protein sample was mixed with 8 mol/L
   urea and centrifuged. After reduction by dithiothreitol (DTT),
   alkylation was performed with iodoacetamide (IAA). Then trypsin was
   added to digest the proteins into peptides. Data independent
   acquisition (DIA)-based quantitative proteomic analysis [71]^25 was
   performed by Orbitrap Fusion™ Tribrid™ connected with EASY-nLC1200
   (Thermo Scientific, Waltham, MA, USA). Triplicate analyses were
   performed for both groups (DMSO and SSII). Identified proteins with a P
   value<0.05 and a fold change (FC) greater than 1.5 compared to the
   control group were considered significantly dysregulated. GO annotation
   was implemented to explore the gene regulatory networks
   ([72]http://www.geneontology.org). Kyoto Encyclopedia of Genes and
   Genomes (KEGG, [73]http://www.genome.jp/kegg/) pathway enrichment
   analysis was performed to investigate the significant pathways enriched
   by the differentially expressed genes.

   For further details regarding the materials used, please refer to the
   [74]supplementary information.

Results

ALF reduced intestinal and hepatic soyasaponin II levels in mice

   Soyasaponin II (SSII) is a bioactive material extracted from leguminous
   plants, which has a wide range of health-promoting effects. In order to
   explore the biological role of SSII during LPS/GaIN induced acute liver
   failure pathogenesis, we first performed LC/MS analysis to compare the
   level of SSII between LPS/GaIN administrated mice and control mice. As
   shown in Figure [75]1, SSII levels in the cecal content and liver were
   significantly lower in the LPS/GaIN group than the control group. As
   known, the presence of soyasaponins in vivo is dependent on diet
   supplementation with soybean foods. In order to rule out the
   possibility that different food intake contributed to the different
   levels of soyasaponin II in mice between control with LPS/GaIN group,
   we monitored food consumption after LPS/GaIN treatment. As shown in
   [76]Figure S1, we found that the difference of food consumption was not
   statistically significant. Thus, these results also manifested that the
   difference of SSII between LPS/GalN treated and control group is not
   due to the food intake and we speculated that SSII may play an
   important role in LPS/GaIN induced acute liver failure.

Figure 1.

   [77]Figure 1
   [78]Open in a new tab

   Acute liver failure occurrence reduced intestinal and hepatic
   soyasaponin II levels in mice. C57BL/6J mice were intraperitoneally
   administrated with D-galactosamine (700 mg/kg) in combination with
   lipopolysaccharide (10 μg/kg). Cecal contents and liver were collected
   after LPS/GalN treatment for 6 h. (A) Soyasaponin II level in cecal
   content and the representative chromatogram (n=5). (B) Soyasaponin II
   levels in liver and the representative chromatogram (n=3 for
   LPS/GalN-treated and n=4 for PBS-treated mice). *p<0.05.

Soyasaponin II pretreatment mitigated ALF

   To further investigate the function of SSII in LPS/GalN induced ALF,
   mice received pretreatment with or without SSII to compare their
   response. We firstly found soyasaponin II levels were higher in the
   liver and cecal content after soyasaponin II administration ([79]Figure
   S2A-2B), suggesting soyasaponin II could be uptake by liver.
   Furthermore, after LPS/GalN, compared to the control mice, plasma
   alanine aminotransferase (ALT) levels in SSII pretreated mice were
   remarkably decreased (Figure [80]2A). Liver morphology, HE staining,
   and TUNEL staining confirmed the ALT results (Figure [81]2B-D). Local
   inflammatory response in the liver was indicated by increased mRNA
   levels of inflammatory cytokines, including Il-1β, Il-6, Cxcl-2,
   Cxcl-10, Ccl-2, Ccl-4, Ccl-7 (Figure [82]2E, [83]Figure S2C), the mRNA
   levels of these inflammatory factors were reduced in SSII-pretreated
   mice compared to control mice. Then, we detected CD11b expression level
   by immunohistochemistry to assess monocyte and neutrophil recruitment
   in the liver. As expected, we found that LPS/GaIN treatment
   significantly increased the number of CD11b+ cells in the liver,
   whereas SSII pretreatment reduced CD11b+ cells recruitment (Figure
   [84]2F). Systematic inflammatory response was demonstrated by elevated
   levels of serum Il-1β (Figure [85]2G). Together, these results,
   consistent with previous studies [86]^26, demonstrate that SSII
   pretreament exhibited protective effects against LPS/GalN-induced ALF.

Figure 2.

   [87]Figure 2
   [88]Open in a new tab

   Soyasaponin II pretreatment mitigated acute liver failure. After
   pretreatment with or without soyasaponin II (5 mg/kg) for 3 days, mice
   were intraperitoneally administered D-galactosamine (700 mg/kg) in
   combination with lipopolysaccharide (10 μg/kg) for 6 h. (A) Effect of
   SS II (5 mg/kg, i.g) pretreatment on ALT levels (n=14-15 for
   LPS/GalN-treated and n=4 for PBS-treated mice). (B) Representative
   photographs of livers. (C) HE staining and (D) TUNEL staining of liver
   samples. The number of apoptotic cells was quantified (n=5 for
   LPS/GalN-treated and n=4 for PBS-treated mice). (E) qPCR analyses of
   hepatic Il-1β, Il-6, Cxcl-2, Ccl-2 gene expression. (n=8 for LPS/GalN
   treated and n=4 for PBS treated mice). (F) Immunohistochemical staining
   for CD11b in liver sections (n=5-7 for LPS/GalN-treated and n=3 for
   PBS-treated mice). (G) Level of serum Il-1β (n=8-9 for LPS/GalN treated
   and n=4 for PBS treated mice). Scale bars: 100 μm. *p<0.05.

   We post-treated SSII 1 hour after LPS/GalN to check whether
   post-treatment could alleviate acute liver injury in vivo.
   Unfortunately, plasma alanine aminotransferase (ALT) levels in SSII
   post-treated mice exhibited decrease trend without significance
   compared with control mice, furthermore hepatic HE staining confirmed
   the ALT results ([89]Figure S3A-B). These data indicated post-treatment
   may be less efficient than pre-treatment. Additionally, SSII
   administration could not attenuate acetaminophen (APAP) induced liver
   damage as evidenced by similar plasma ALT level and hepatic histologic
   changes between two groups ([90]Figure S3C-D), suggesting SSII
   pretreament did not protect APAP induced acute liver failure.

Soyasaponin II ameliorated canonical Nlrp3 inflammasome-associated innate
immune response

   We next explored the underlying mechanism for the protective effects of
   SSII. Macrophages were found to play a pivotal role in ALF development
   [91]^27^-[92]^29. To screen for the key molecule mediated by SSII, we
   performed transcriptome analysis of the macrophages isolated from the
   liver after LPS/GaIN exposure treatment with or without soyasaponin II.
   PCA analysis showed that the gene-expression profile of soyasaponin
   II-stimulated hepatic macrophages had significant differences when
   compared to controls after LPS/GaIN treatment (Figure [93]3A,
   [94]Figure S4A). GSEA (Gene Set Enrichment Analysis) highlighted
   NOD-like receptor signaling as a top ten enriched Kyoto Encyclopedia of
   Genes and Genomes (KEGG) pathway that decreased visibly after LPS/GaIN
   exposure treatment in the soyasaponin II group (Figure [95]3B,
   [96]Figure S4B). We next selected NOD-like receptor signaling
   pathway-related genes for hub gene analysis and found that inflammasome
   related genes, such as Nlrp3 and Il-1β, were mainly influenced upon
   SSII administration (Figure [97]3C).

Figure 3.

   [98]Figure 3
   [99]Open in a new tab

   Soyasaponin II ameliorated canonical Nlrp3 inflammasome-associated
   innate immune response. After pretreatment with or without soyasaponin
   II (5 mg/kg) for 3 days, mice were intraperitoneally administered
   D-galactosamine (700 mg/kg) in combination with lipopolysaccharide (10
   μg/kg) for 6 h. Hepatic macrophages were isolated, followed by
   transcriptome analysis. (A) Scatter plots of PCA of hepatic macrophage
   gene expression profile (n=3). (B) KEGG signaling pathway by GSEA
   (n=3). (C) Hub gene analysis of NOD-like receptor signaling pathway by
   CytoHubba (n=3). (D) Mature BMDMs were stimulated with DPBS or LPS for
   2 h, with or without SSII (5 μg/ml) co-incubation. The relative mRNA
   levels of Nlrp3 and Il-1β (n=4). (E) Representative western blot
   analysis for Nlrp3, caspase-1 (precursor and spliced variant) and
   pro-Il-1β from BMDMs cell lysis solution and the quantification results
   (n=4). (F) Representative immunofluorescence images of ASC in BMDMs,
   and the fluorescence quantification results. Red represents ASC; Blue
   represents DAPI (n=3). (G) Cleaved Il-1β released in BMDMs culture
   supernatant was quantified by ELISA (n=10 for LPS-treated and n=8 for
   PBS-treated cells) and Il-1β levels from isolated liver macrophages
   (n=4-5 for LPS/GalN-treated and n=3 for PBS-treated mice). Scale bars:
   20 μm. *p<0.05.

   Accumulating evidence described the pivotal role of the Nlrp3
   inflammasome in ALF. Consequently, based on the above transcriptome
   results, we speculated that SSII may have an antagonistic effect on
   Nlrp3 inflammasome activity. Bone marrow-derived macrophages (BMDMs)
   were challenged with LPS in the absence or presence of SSII. As
   expected, Nlrp3 and Il-1β displayed decreased gene expression levels in
   the SSII group compared to the control group (Figure [100]3D). Next, we
   compared the responsiveness of Nlrp3 inflammasome protein levels to
   LPS-stimulated BMDMs in both groups. The Nlrp3, caspase-1 (p20), and
   pro-Il-1β expression after LPS stimulated significantly decreased when
   incubated with SSII (Figure [101]3E). SSII group indicated less ASC
   accumulation compared with SSII non-treated group, indicating SSII also
   attenuated Nlrp3 inflammasome activation (Figure [102]3F). Given the
   fact that Il-1β serves as a key contributor to ALF, we measured cleaved
   Il-1β in BMDMs supernatant by ELISA. We confirmed that SSII
   substantially blocked the release of mature Il-1β (Figure [103]3G).
   Decreased hepatic macrophage Il-1β protein levels in mice after SSII
   treatment were also found (Figure [104]3G). Taken together, these
   findings corroborate the theory that SSII decreases the immune reaction
   by abolishing Nlrp3 inflammasome component priming, and that this
   inhibitory activity is believed to be actuated by a transcriptional
   program.

Soyasaponin II blocked YB-1 phosphorylation in macrophages

   Based on the potential interaction between soyasaponin II and Nlrp3
   inflammasome, we sought to further decipher this molecular pathway by
   employing mass spectrometry-based proteomics analysis. A panoramic view
   by unsupervised hierarchical clustering analysis of the relative
   quantitative data showed that the composition and abundance of proteins
   in liver macrophages in ALF mice pre-conditioned with soyasaponin II,
   had greatly changed compared with controls (Figure [105]4A). Volcano
   Plot analysis showed that noteworthy proteins subdivided into two major
   categories based on the enrichment of differential proteins (Figure
   [106]4B). Next, we conducted gene ontology (GO) enrichment analyses to
   interpret the idiographic function of the discrepant protein sequences
   biologically. According to the -Log[10] (FDR) and the observed gene
   count, proteins belonging to Biological Process, Cellular Component and
   Molecular Function showed differences between SSII treated and
   non-treated groups ([107]Figure S5A). As expected, according to (KEGG)
   pathway, there was obvious alteration in protein catabolism (such as
   valine, leucine, and isoleucine degradation, etc.) processes,
   reflecting the differences of the liver macrophages between the two
   groups ([108]Figure S5B). Two proteins exhibited significantly
   different trend: PRPF19 (Reactome database identifier: R-MMU-5420907)
   which presented the highest fold change, and Ddx39B (Reactome database
   identifier: R-MMU-8849139) which also showed significant alteration
   were noted. We ultimately took a crack at deciphering the discrepancy
   by looking for the senior protein which regulates these two proteins.
   Surprisingly, by searching the Reactome database and using the
   protein-protein interaction (PPI) network, we found these two proteins
   were potentially interconnected with YB-1 (Figure [109]4C), suggesting
   that YB-1 may be responsible for driving the different pathophysiologic
   phenotype between control and SSII groups.

Figure 4.

   [110]Figure 4
   [111]Open in a new tab

   Soyasaponin II blocked YB-1 activation in macrophages. (A, B) Heat-Map
   and volcano plots of liver macrophages proteins with protein cluster
   analysis of differentially regulated proteins in DMSO and SSII groups
   from LPS/GaIN treated animals. Plot of SSII/DMSO ratios (Log[2]FC)
   versus -Log[10]FDR of liver macrophage proteins: green dots represent
   downregulated proteins, red dots represent differentially up-regulated
   proteins. (C) Protein-protein interaction (PPI) network diagram of
   discrepant proteins connected with YB-1 from the two groups. (D) Mature
   BMDMs were stimulated with PBS or LPS for 2 h, with or without
   co-incubation of SSII (5 μg/ml), and cells harvested after activation
   with ATP (5mM) for 30min. Representative western blot analysis for
   phosphorylated YB-1 and total YB-1 in cell lysis solution is displayed,
   as well as the protein quantification (n=4). (E) Left panel: PBS or LPS
   primed BMDMs with (n=3) or without (n=3) co-incubation with SSII (5
   μg/ml) in petri dishes for 2 h. Immunofluorescence staining with p-YB-1
   antibody and the fluorescence quantification. Right panel: Macrophages
   isolated from LPS/GaIN treated mouse liver with (n=3) or without (n=3)
   SSII pretreatment for three days. Immunofluorescence staining with
   p-YB-1 antibody and the fluorescence quantification. (F) BMDMs were
   stimulated with PBS or LPS for 15 min, with or without co-incubation of
   SSII (5 μg/ml). Representative western blot for phosphorylated AKT in
   BMDMs and the quantification results (n=3). (G) BMDMs was pretreated
   with LY294002 (10 μM) for 30 min, then were stimulated with LPS
   co-treatment with or without soyasaponin II. Representative western
   blot for phosphorylated YB-1 and the quantification results (n=3).
   Scale bars: 10 μm. *p<0.05.

   YB-1 is a transcription factor which can be activated by
   phosphorylation, and YB-1 nuclear translocation plays an important role
   in diseases development [112]^30. Consistent with previous studies
   [113]^31, we found that YB-1 primarily expresses in the cytoplasm under
   normal status ([114]Figure S6A). We then reasoned that YB-1
   phosphorylation regulates YB-1 nuclear translocation. To verify this
   hypothesis, we transfected pcDNA-YB-1 mutant (Ser^102 to Glu^102,
   p-YB-1 overexpression) plasmid or pcDNA-YB-1 plasmid (control) to
   macrophages, followed by LPS treatment. We first confirmed that p-YB-1
   was overexpressed in pcDNA-YB-1 mutant (Ser^102 to Glu^102) plasmid
   transfected cells ([115]Figure S6B), and as shown in [116]Figure S6C,
   we found that YB-1 nuclear translocation was significantly increased in
   pcDNA-YB-1 glutamic acid mutant plasmid group compared to control
   group. These results suggested that YB-1 phosphorylation can directly
   promote YB-1 nuclear translocation. We next attempted to monitor
   whether phosphorylated YB-1 levels were influenced by SSII. Western
   blotting revealed that the level of phosphorylated YB-1 was markedly
   elevated in BMDMs after LPS administration, whereas SSII significantly
   mitigated p-YB-1 compared to controls (Figure [117]4D). To further
   detail the inhibitory effects of SSII on phosphorylated YB-1 levels in
   nucleus, we performed immunofluorescence staining for both BMDMs and
   mice liver macrophages. As presented in Figure [118]4E, LPS challenge
   promoted p-YB-1 nuclear translocation, but treatment with SSII largely
   prevented this critical protein from phosphorylating and transportation
   into the nucleus. Emerging evidence has indicated that AKT
   phosphorylates YB-1 at Ser102 in the cold shock domain [119]^17. Here,
   studies were conducted to determine whether soyasaponin II regulates
   YB-1 phosphorylation (Ser102) through mediating AKT activation or not.
   As shown in Figure [120]4F, AKT was activated upon LPS challenge
   compared to control group and soyasaponin II treatment markedly
   downregulated AKT phosphorylation in BMDMs. Furthermore, BMDMs were
   pretreated with LY294002 which is an inhibitor of AKT signaling, and
   then were stimulated with LPS co-treated with or without soyasaponin
   II. As shown in Figure [121]4G, we found that LY294002 markedly
   decreased YB-1 phosphorylation level whereas soyasaponin II treatment
   did not further reduce YB1 phosphorylation in BMDMs. Therefore, we
   concluded that soyasaponin II reduced YB-1 phosphorylation after LPS
   challenge may be mediated by AKT signaling.

The repression of Nlrp3 inflammasome by soyasaponin II is dependent on YB-1

   Since SSII can impact both YB-1 and Nlrp3 inflammasome priming, we next
   investigated the molecular correlation between YB-1 and Nlrp3
   inflammasome. First, we designed YB-1 siRNA and confirmed the
   repression effect ([122]Figure S7). YB-1 mRNA knockdown markedly
   lowered the mRNA levels of Nlrp3 and Il-1β in BMDMs after exposure to
   LPS (Figure [123]5A). Moreover, as shown in Figure [124]5B, we found
   higher mRNA levels of Nlrp3 and Il1β in the BMDMs transfected with
   pcDNA3.1-YB-1 glutamic acid mutant plasmid (phosphor-mimetic mutant of
   YB-1) compared with pcDNA3.1-YB-1 plasmid. Furthermore, destabilization
   of YB-1 mRNA showed decreased protein levels of Nlrp3, cleaved
   caspase-1, and pro-Il-1β (Figure [125]5C). Also, secreted Il-1β in
   BMDMs was markedly diminished in the YB-1 siRNA group after LPS
   challenge (Figure [126]5D). These experiments indicate that YB-1
   signaling attenuates Nlrp3 inflammasome priming and the release of
   pro-inflammatory cytokine Il-1β.

Figure 5.

   [127]Figure 5
   [128]Open in a new tab

   The repression of Nlrp3 inflammasome by soyasaponin II was dependent on
   YB-1. (A) BMDMs were transfected with YB-1 si-RNA plasmid or negative
   control (NC) plasmid for 36 h, then incubation with LPS or PBS for 2 h.
   The relative mRNA levels of Nlrp3 (n=12 for LPS-treated and n=4 for
   PBS-treated cells) and Il-1β (n=12-13 for LPS-treated and n=8-9 for
   PBS-treated cells). (B) BMDMs were transfected with pcDNA-YB-1 mutant
   (Ser102 to Glu102, p-YB-1 overexpression) plasmid or pcDNA-YB-1
   (control) plasmid for 36 h, followed by LPS treatment for 2 h. The
   relative mRNA levels of Nlrp3 and Il-1β mRNAs (n=8). (C) BMDMs were
   transfected with YB-1 si-RNA plasmid or negative control (NC) plasmid
   for 36 h, then incubation with LPS or PBS for 2 h. Western blot of
   Nlrp3, pro-caspase-1, caspase-1 p20 and pro-Il-1β, as well as the
   quantification results (n=3). (D) Cleaved Il-1β released in cell
   supernatant were quantified by ELISA (PBS group n=4, LPS group n=20).
   (E) LPS-primed BMDMs were co-incubated with or without SSII for 2 h,
   transfected with or without YB-1 si-RNA for 36 h, and the relative
   expression mRNA levels of Nlrp3 (n=5 for YB-1 si-RNA-treated and n=6-10
   for negative control (NC) plasmid -treated cells) and Il-1β (n=7). (F)
   Western blot of Nlrp3, pro-caspase-1, caspase-1 p20 and pro-Il-1β, as
   well as the quantification results (n=5). (G) Cleaved Il-1β released in
   cell supernatant were quantified by ELISA (n=8). *p<0.05.

   Our data shows that both SSII and YB-1 knockdown could repress Nlrp3
   inflammasome. To further link these molecules, we set up three
   independent groups: LPS, LPS + SSII, and LPS + SSII + YB-1 siRNA.
   Figure [129]5E-G shows that in the SSII supplemented microenvironment,
   Nlrp3 inflammasome priming was sharply decreased compared with the LPS
   group, but when co-treated with YB-1 siRNA, the degree of canonical
   Nlrp3 inflammasome activation was equivalent to the group without
   siRNA. These data demonstrate that SSII suppression of the
   LPS-stimulated Nlrp3 inflammasome priming may depend on YB-1 in
   macrophages.

YB-1 binds the promoter region of Nlrp3 to upregulate transcriptional
activity

   We further explored the detailed molecular mechanism by which YB-1
   stimulates Nlrp3 at the transcription level. We speculate that YB-1
   protein directly binds the promoter region of Nlrp3 and regulates its
   expression. To confirm this, the Dual-Luciferase Report Gene System was
   used. Figure [130]6A-[131]6B showed that pcDNA3.1-YB-1 plasmid
   dramatically increased the luciferase activity of pGL3-Nlrp3 (-2000bp
   to +100bp promoter region) compared to pcDNA3-basic plasmid in
   macrophage. In addition, the luciferase activity of pGL3-Nlrp3 promoter
   reporter plasmid was increased upon LPS challenge compared to control
   group, and soyasaponin II treatment markedly suppressed the luciferase
   activity (Figure [132]6C). Furthermore, as shown in Figure [133]6D,
   ChIP sequencing results confirmed that p-YB-1 could bind to the
   promoter region of Nlrp3 gene in macrophage under LPS stimulation.
   These results suggest that YB-1, specifically p-YB-1 can interact with
   Nlrp3 promoter to upregulate transcription.

Figure 6.

   [134]Figure 6
   [135]Open in a new tab

   YB-1 binds the promoter region of Nlrp3 to upregulate transcriptional
   activity. (A) The sketch map of plasmid. (B) Raw264.7 cells were
   transfected with pGL3-promoter reporter plasmid and pcDNA3.1-YB-1
   plasmid or pcDNA3-basic plasmid for 36 h, followed by LPS treatment for
   2 h. Relative luciferase activity in each group (n=4). (C) Raw264.7
   cells were transfected with pGL3-promoter reporter plasmid and
   pcDNA3.1-YB1 plasmid for 36 hours. Then Raw264.7 cells were treated
   with or without SSII for 2 h in presence of LPS. Relative luciferase
   activity in each group (n=5-7 for LPS-treated and n=5 for PBS-treated
   cells). (D) The chromatin from Raw264.7 cells that were treated with or
   without LPS (500 ng/ml) for 2 h was collcted for the
   immunoprecipitation by using p-YB-1 antibody. DNA was extracted and
   sequenced. Integrative Genomics Viewer (IGV) at locus containing Nlrp3
   was shown. *p<0.05.

Phosphorylated YB-1 is enriched in the macrophage nuclei of ALF patients

   Our above data clearly showed that p-YB-1 may be a key regulator in
   murine liver failure, we finally linked evidence from mouse models to
   data from ALF patients. Total YB-1 was mainly located in cytosol in
   control liver under normal circumstance ([136]Figure S8), while
   immunofluorescence staining of liver sections from ALF patients showed
   statistically significant increases in p-YB-1 in the nuclei of CD64
   marked macrophages compared with controls (Figure [137]7A-B). This
   confirmed the presence of increased p-YB-1 in the liver macrophages of
   ALF patients.

Figure 7.

   [138]Figure 7
   [139]Open in a new tab

   Phosphorylated YB-1 was enriched in the macrophage nuclei of ALF
   patients. (A, B) Macrophages in human liver paraffin section marked by
   CD64 antibody, and p-YB-1 expression were detected by
   immunofluorescence, and the p-YB-1 nuclear translocation positive
   macrophages were quantified (control n=5; ALF n=11). (C) Working model:
   LPS/GalN challenge decreased fecal and hepatic soyasaponin II levels.
   Soyasaponin II repressed YB-1 phosphorylation and nuclear
   translocation, p-YB-1 could translocate into nuclear and bind the
   promoter of Nlrp3, to promote Nlrp3 inflammasome priming, Il-1β
   overproduction and liver damage. Scale bars: 10 μm. *p<0.05.

Discussion

   ALF is characterized as a rapidly deteriorating disease with different
   etiologies causing overwhelming hepatocyte death including systemic
   inflammatory response syndrome [140]^1^-[141]^3. Clinical information
   and tissue samples collected and dissected by US Acute Liver Failure
   Study Group indicate that focusing on inflammasome may lead to some
   therapeutic discoveries for treatment of ALF [142]^6. In this regard,
   new treatment methods should be established pressingly. Epidemiological
   investigations show that soybean products contain essential amino
   acids, unsaturated fatty acids, dietary fiber, minerals and a variety
   of phytochemicals, which are considered to be benefit for human and
   animals [143]^32. Although there are no direct clinic evidences to show
   eating more soybeans could protect from ALF, it is reported that black
   soybean could modulate cholesterol metabolism and mitigate oxidative
   damage during non-alcoholic fatty liver disease progression [144]^33.
   Besides, the preclinical studies also demonstrated that administration
   of soyasaponins, extracted from soybeans, effectively protected mice
   from alcohol induced acute hepatotoxicity [145]^34. These conclusions
   provide strong support for soyasaponins research in acute or chronic
   liver diseases, but it remains need to be confirmed in clinical trial.
   Our research investigated the anti-inflammation role of SSII and its
   protective mechanisms in LPS/GaIN induced ALF. With regards to clinical
   therapy, another objective of this study was to discover an
   intermediary mediator so as to interpret the integrated signal pathway
   in acutely inflamed liver disease. It has been reported that Nlrp3
   inflammasome complex plays an important role in ALF [146]^35^-[147]^37.
   Our data further shows that soyasaponin II protects mice against
   LPS/GaIN-induced acute liver injury by depressing Nlrp3 inflammasome
   priming to reduce Il-1β release. This finding provides a new treatment
   method for acute liver injury in the preclinical model, but more
   clinical research is needed to investigate the role of soyasaponin II
   in treating ALF. Whether or not soyasaponins can be absorbed by the
   intestinal tract has been a focus of debate. Early studies found few
   soyasaponins uptake by the circulatory system, and soyasaponins are not
   detected in the blood or urine of hosts who have been exposed to a
   soybean diet [148]^38^-[149]^40. It was generally believed that
   soyasaponins were hydrolyzed into metabolites by the gut microbiota and
   then excreted in the feces. However, other recent studies found that
   soyasaponins could be detected in human and rat serum using
   high-performance liquid chromatography-tandem mass spectrometry
   (HPLC-MS/MS) and that the bioavailability of group B soyasaponins was
   better than group A [150]^41^-[151]^42. We also successfully detected
   soyasaponin II in the liver with LC/MS. In conclusion, it is reasonable
   to speculate that a certain amount of soyasaponin II can be absorbed
   into liver after soyasaponin II treatment and exert an
   anti-inflammatory effect by suppressing Nlrp3 inflammasome priming in
   ALF progression.

   It is reported that certain types of intestinal microbiota including
   Lactobacillus, Bifidobacteria, Escherichia and Bacteroides can degrade
   saponins by synthetizing β-glucosidase and/or other glycosides
   metabolizing enzymes [152]^43. Hence, we hypothesized that changes in
   the composition and function of the gut microbiota may lead to the
   altered levels of soyasaponin II in LPS/GaIN induced ALF mice. In order
   to verify the hypothesis, we performed metagenomics of cecal contents
   to analyze gut microbiota in the control and LPS/GalN groups. There was
   a change in the intestinal microbial composition of control group
   compared with the LPS/GaIN group ([153]Figure S9A). Then, we conducted
   a functional analysis of the identified gut microbiota using CAZy
   (Carbohydrate-Active enZYmes Database) mapping. We found a difference
   in glycosyl transferases which mediates the metabolic processing of
   glycosides compound [154]^44, suggesting the difference in the function
   of the gut microbiota to metabolize glycosides between control and
   LPS/GaIN group ([155]Figure S9B). Our data preliminarily concluded that
   the reduction of intestinal and hepatic soyasaponin II levels in
   LPS/GalN treated mice may be partially resulted from altered gut
   microbial composition and function. Future research is needed to
   clarify the specific mechanisms.

   Empirical studies have linked YB-1 to the epithelial-to-mesenchymal
   transition (EMT) that is associated with breast carcinoma
   aggressiveness, suggesting YB-1 signaling could mediate the metastasis
   of epithelial malignancies. Apart from its oncogenic role, our work
   shows that YB-1 induces inflammation by binding to the transcriptional
   initial region of Nlrp3. Although we found YB-1 could also mediate
   Il-1β mRNA level in macrophages, there was no difference of Il-1β
   promoter luciferase activity between pcDNA3.1-YB-1 plasmid compared to
   pcDNA3.1-basic plasmid (data not shown), indicating the altered gene
   expression of Il-1β observed in YB-1 treated group may be mediated by
   indirect manner, for example, YB-1 could interact with other important
   transcription factor in the nuclear and promoter Il-1β transcription.
   Future work is needed to specify the detailed mechanism. However, our
   current results strongly suggest that YB-1 could function as potential
   therapy for treating inflammsome-mediated inflammatory diseases. It has
   been reported that heterozygous YB-1 knockout reduced hepatic
   chemokine/chemokine expression and diminished granulocyte infiltration
   into the liver, and these YB-1 knockout mice were protected from LPS
   induced mortality compared to WT mice [156]^45. The role of inflammsome
   has been studied in many types of liver diseases, and it is thought to
   be a primary contributor to hepatocyte injury, immunological cell
   motivation, and hepatic inflammation. Further investigation is needed
   to determine whether YB-1 has a similar effect on other inflammasomes
   such as NLRP1, NLRC4, IPAF, AIM2, etc. Additionally, our data revealed
   that soyasaponin II is a novel inhibitor for YB-1 activation, implying
   that it may be beneficial for other pathologic processes mediated by
   YB-1.

   Macrophages orchestrate multiple missions in immune system, but the
   excessive activation of immune cells may give rise to the tissue
   injury. Based on our findings related to the inflammation-suppressive
   effects of SSII, coupled with the ability of this soy compound to
   manipulate key transcription factor, we theorize that the inflammatory
   response of YB-1 in macrophages is mitigated by SSII in ALF (Figure
   [157]7C). Our results confirm the molecular immunological role of SSII,
   and we believe SSII effect on YB-1 could be clinically applicable for
   treating hepatic diseases.

Supplementary Material

   Supplementary materials and methods, figures, and tables.
   [158]Click here for additional data file.^ (1.5MB, pdf)

Acknowledgments