Abstract

   Gram-positive bacterial pneumonia is a significant cause of
   hospitalization and death. Shortage of a good experimental model and
   therapeutic targets hinders the cure of acute lung injury (ALI). This
   study has established a mouse model of ALI using Gram-positive bacteria
   Lactobacillus casie cell wall extracts (LCWE) and identified the key
   regulator NLRP3. We show that LCWE induces TNF, NF-κB signaling, and so
   on pathways. Similar to lipopolysaccharide (LPS), LCWE induces the
   infiltration of CD11b-positive cells and inflammation in lungs. LCWE
   also triggers inflammatory signaling through TLR2, different from LPS
   through TLR4. It suggests that cytokines amplify inflammation signaling
   relying on NLRP3 in LCWE-induced ALI. NLRP3 deletion disrupts
   inflammation, IL-1β cleavage, and the infiltration of neutrophils and
   macrophages in the injured lung. Our study highlights an animal ALI
   model for Gram-positive bacterial pneumonia and that NLRP3 is a key
   therapeutic target to prevent inflammation and lung damage in
   LCWE-induced ALI.

   Subject terms: Immunology, Molecular medicine
     __________________________________________________________________

   NLRP3 drives inflammatory signaling and cleaved IL-1β in a mouse model
   of acute lung injury induced by Lactobacillus casei cell wall extracts,
   highlighting its role as a therapeutic target in Gram-positive
   bacterial pneumonia.

Introduction

   Pneumonia is a common cause of ALI and acute respiratory distress
   syndrome (ARDS), leading to significant hospitalizations and deaths
   among adults^[52]1,[53]2. Pathogens of pneumonia include bacteria,
   viruses, mycobacteria, fungi, and parasites^[54]3–[55]6. Bacteria are
   the primary pathogens of pneumonia, categorized as community-acquired
   and hospital-acquired pneumonia^[56]3. One of the most common bacteria
   is Streptococcus pneumoniae^[57]1,[58]2, a Gram-positive bacterium. In
   addition, secondary bacterial infections play a critical role following
   viral infection. Gram-positive bacteria such as Staphylococcus aureus
   and S. pneumoniae were the main pathogens after severe influenza
   infection, historically causing morbidity and mortality^[59]7–[60]9.

   S. aureus activates inflammatory signaling pathways during lung
   infections. The nucleotide-binding oligomerization domain containing 2
   (Nod2) functions as a sensor to detect muramyl dipeptide (MDP), a
   common unit of peptidoglycan found in most Gram-positive
   bacteria^[61]10. Loss of Nod2 reduced lung inflammation and neutrophil
   recruitment induced by S. aureus^[62]11. Ahn et al. discovered that
   IL-16 contributes to S. aureus infection through tumor necrosis factor
   receptor (TNFR) in CD4^+ cells^[63]9. In mouse lungs and THP-1
   monocytes, the inflammasome protein NLRP3 is essential for activated
   inflammation^[64]12. Gram-positive bacteria activate the expression of
   inflammasome proteins like NLRC4, which reduces IL-17A-dependent
   neutrophil accumulation in Gram-positive pneumonia^[65]13. During
   infection, the bacterial cell wall is the first part to interact with
   host cells. However, it remains unknown whether and how the cell wall
   of Gram-positive bacteria contributes to lung inflammation and ALI.

   The murine model of ALI facilitates uncovering molecular mechanisms
   underlying ARDS^[66]14,[67]15. In recent years, LPS from Gram-negative
   bacteria has been used to induce inflammatory responses in animal
   models by inhalation administration, establishing an ALI mouse
   model^[68]16,[69]17. Lipoteichoic acid (LTA) has been mentioned in a
   few studies but is not widely used due to an unestablished model. In
   addition, global changes in gene expression and signaling pathways in
   ALI induced by Gram-positive and Gram-negative bacteria are yet to be
   determined. The Gram-positive bacteria Lactobacillus casie cell wall
   extract (LCWE) is widely used as an inducer of the autoimmune disease
   IgA vasculitis^[70]18. As cell wall components of L. casie are similar
   to the ones of the primary ALI pathogen S. aureus, we wonder if LCWE of
   L. casie can induce lung inflammation and acute lung injury to
   facilitate the identification of therapeutic target.

   In this study, we established a murine model of ALI induced by LCWE to
   uncover how Gram-positive bacteria induce lung inflammation and to
   determine the key therapy targets for ALI/ARDS. Our data suggested that
   LCWE induced inflammation signaling pathway through TLR2, and MYD88
   relays the signaling to activate NF-κB signaling pathways. Loss of
   NLRP3 disrupted inflammatory gene expression of IL6, TNFα, and
   pro-IL-1β in THP-1 cells and LCWE-induced ALI lung, highlighting NLRP3
   as the therapeutic target of Gram-positive bacterial pneumonia.

Results

LCWE induces severe ALI

   To determine whether Gram-positive bacteria Lactobacillus casie cell
   wall extract (LCWE) induces ALI, we administered LCWE to mouse lungs
   and collected lung tissues for pathological analysis. LPS, derived from
   Gram-negative bacteria and known as an ALI inducer, was used for
   positive control, and PBS was applied for vehicle control. Hematoxylin
   and Eosin (H&E) staining showed that the alveoli were evenly
   distributed in the control lungs (Fig. [71]1A). Compared to the vehicle
   control groups, the LPS-treated groups have smaller alveoli. The lungs
   were infiltrated with more immune cells (Figs. [72]1A, [73]S1A),
   possibly causing the disappearance of the alveoli. These observations
   were consistent with previous reports of ALI^[74]14,[75]19,[76]20.
   Surprisingly, in the LCWE-treated groups, some lung regions were
   infiltrated with massive immune cells, and the alveoli disappeared
   completely, suggesting severe lung injury was induced by LCWE.
   Pathological analyses showed that the lung injury score induced by LCWE
   was comparable to that induced by LPS (Fig. [77]1B). As the marker
   genes of Toll/NF-κB signaling, total IL-1β and IL-10 faithfully reflect
   inflammation status and were used for a readout of
   inflammation^[78]21,[79]22. Both LPS and LCWE induced a significant
   increase in Il1b and Il10 mRNA (Fig. [80]1C). Altogether, LCWE induces
   comparable lung injury and inflammation to LPS. Subsequently, we
   compared the effects of LCWE and LPS on ALI at doses 1, 3, and 6 mg/kg
   (Fig. [81]S1B–D). Consistently, all three different doses (1, 3, and
   6 mg/kg) of LPS and LCWE induced severe ALI according to pathological
   analysis, and dramatic expression increase of IL-1β (Fig. [82]1D) and
   IL-10 proteins (Fig. [83]1E) in the injured lung tissues by
   immunostaining and both in the injured lung tissues and bronchoalveolar
   lavage fluid (BALF) by ELISA (Fig. [84]1F). These assays suggested that
   LCWE, the cell wall of Gram-positive bacteria, induces severe ALI, and
   inflammation in lung infection.

Fig. 1. Establishment of LCWE-induced ALI murine model.

   [85]Fig. 1
   [86]Open in a new tab

   A H&E staining of mouse lungs treated with PBS, 3 mg/kg of LPS or LCWE
   for 24 h. After treatment, animals were anesthetized with avertin, then
   lung tissues were collected after perfusion with formaldehyde. AS:
   Alveolus, BV: Blood Vessels, T: Trachea. B Pathological injury scores
   were collected in a blinded fashion by pathologists using H&E staining
   sections of mouse lung tissues after treatment with LPS or LCWE.
   Statistical analysis: One-way ANOVA test with p-value as indicated,
   n = 5. Error bar: SD. C Il1b and Il10 mRNA levels of lung tissues
   induced with PBS, LPS, or LCWE by RT-qPCR. Statistical analysis:
   One-way ANOVA test with p-value as indicated, n = 5. Error bar: SD. D,
   E IL-1β and IL-10 immunostaining of lung tissues with ALI induced with
   1, 3, 6 mg/kg of LPS or LCWE for 24 h. F IL-1β and IL-10 protein levels
   in both lung tissues and BALF were measured by ELISA. Statistical
   analysis T-test **p < 0.01, n = 4. Error bar: SD.

LCWE induces Toll/NF-κB and TNF signaling and promotes T cell differentiation
in ALI

   To understand the global changes of gene expression in our
   well-established LCWE- and LPS-induced ALI mouse models, we performed
   transcriptomic analyses on lung tissues with ALI (Fig. [87]2A, B). LPS
   altered the mRNA expression of 1884 genes, while LCWE did 5592 genes.
   The 897 upregulated genes by LPS were involved in TNF, Toll-like
   receptor, and NF-κB signaling pathways in KEGG pathway enrichment
   analysis (Fig. [88]S2-A). In contrast, besides TNF, the Toll-like
   receptor signaling pathways, the IL-17 pathway was specifically
   activated by LCWE within the top signaling pathways of the 2043 genes
   (Fig. [89]S2-B). Since both LCWE and LPS induce ALI, we analyzed the
   792 overlapping upregulated genes to identify commonly activated
   signaling pathways during ALI. The confirmed top signaling pathways
   were TNF, Toll-like receptor, and NF-κB pathways (Fig. [90]2C, D). In
   addition, LCWE activated T cell differentiation into Th1, Th2, and Th17
   subsets (Fig. [91]2E). The GO (Gene Ontology) analysis of
   differentially expressed genes indicated that the overlapped biological
   processes included leukocyte migration, immune effector functions,
   inflammatory response, and toll-like receptor binding (Fig. [92]S3).
   Pathways related to calcium, cAMP signaling were downregulated by both
   LPS and LCWE, LCWE alone and LPS alone (Fig. [93]S4). We further
   verified the mRNA expression of NF-κB pathway genes including Nfkbia,
   Myd88, Tnfa, and Il6 in ALI by quantitative PCR. All these genes were
   significantly upregulated by both LPS and LCWE stimulation
   (Fig. [94]3). The expression of CD14, the cofactor of TLR4, was
   significantly activated by LPS, consistent with the previous
   report^[95]23. The interferon pathway genes Stat6, Irf7, T cell
   activation regulator gene Saa3, and chemokine genes Cxcl2 and Cxcl3
   were strongly activated by LCWE and LPS (Fig. [96]3), suggesting that
   LCWE may induce infiltration of immune cells.

Fig. 2. LCWE activates the inflammatory signaling pathways and promotes
T-helper cell differentiation in ALI.

   [97]Fig. 2
   [98]Open in a new tab

   A, B Volcano plots of the differentially regulated genes of the lungs
   by LPS (A) or LCWE (B), based on the data from RNA-seq analysis of the
   regulated genes in ALI induced by LPS or LCWE. C Similar to (A, B),
   Venn diagram plots of the upregulated genes in ALI induced by LPS
   and/or LCWE. D, E Similar to (A, B), KEGG analysis of the upregulated
   genes in ALI induced by LPS and LCWE (D), by LCWE only (E).

Fig. 3. Genes in NF-κB and chemokine signaling pathways were activated by
LCWE in ALI.

   [99]Fig. 3
   [100]Open in a new tab

   A–C RT-qPCR validation of gene expression in NF-κB pathway (A),
   Jak/Stat pathway (B), and Chemokine signaling pathway (C) of ALI
   induced by LPS or LCWE. One-way ANOVA test with p-value indicated,
   n = 5. Error bar: SD.

Massive CD11b positive cells infiltrate the lung tissues with ALI induced by
LCWE

   Based on the above analyses, Toll/NF-κB and TNF signaling pathways are
   turned on during ALI by both LCWE and LPS. The activation of these
   pathways results in the infiltration of many immune cells into the
   pathological loci. To understand the cells migrating to the ALI loci,
   we performed immune infiltration analysis using our RNA-seq data with
   the CIBERSORT algorithm (Fig. [101]4A). Among the immune cells, M1
   macrophages and the activated CD4 memory T cells increased in the
   injured tissues in response to LCWE stimulation. To identify the types
   of immune cells migrating in response to inflammatory stimulation in
   injured lung tissues, we first stained the lung tissues with anti-IL-1β
   antibodies to confirm inflammation in ALI. After LPS or LCWE
   stimulation, IL-1β positive cells increased, indicating inflammation in
   ALI was induced by LPS or LCWE (Fig. [102]4B). To know further the
   migrating cell type in the injury tissues, we used antibodies against
   CD11b to identify macrophages, monocytes, and neutrophils
   (Fig. [103]4B). Both LPS and LCWE inductions increased CD11b positive
   cell infiltration in the injured lungs, albeit a few CD4 positive cells
   appeared in the injured lung induced by LCWE (Fig. [104]4B).

Fig. 4. Immune infiltration analysis and staining suggested that immune cells
migrate to lung tissues in either LPS- or LCWE-induced ALI.

   [105]Fig. 4
   [106]Open in a new tab

   A Proportion of immune cells in ALI lung tissues treated with 3 mg/kg
   of LPS or LCWE, and the analysis was based on the differentially
   regulated genes from RNA-seq analysis. B IL-1β, CD11b, and CD4 levels
   were shown by DAB staining in the lung tissues with ALI induced by LPS
   or LCWE.

LCWE activates NF-κB and TNF signaling through TLR2, different from LPS
through TLR4

   Generally, monocytes patrol the bloodstream and migrate into the
   injured tissues in response to inflammatory stimuli. Eventually, the
   inflammatory monocytes differentiate into macrophages to help the host
   clear injured tissues and microbes^[107]24–[108]27. Moreover, Toll-like
   receptors (TLRs) are important mediators of inflammatory
   pathways^[109]24,[110]26,[111]28. Hence, to understand how macrophages
   are activated and migrated to injured tissue, we knocked down TLR2 or
   TLR4 in THP-1 monocytes (THP-1) and induced the cells with LCWE or LPS
   (Fig. [112]5A, B). The results showed that TLR2 mRNA silencing
   significantly disrupted LCWE-induced, but did not affect LPS-induced,
   proinflammatory cytokines IL6 and TNFA expression, suggesting that TLR2
   is a specific LCWE receptor to activate monocyte-mediated inflammation
   (Fig. [113]5A). In contrast, when TLR4 was silenced by shRNA
   (Fig. [114]5B), LPS-induced IL6 and TNFA mRNA expressions were
   significantly and almost completely downregulated, while shTLR4 did not
   affect LCWE induction. It confirmed the well-known notion that TLR4 is
   the LPS receptor (Fig. [115]5B). The general proinflammatory cytokine
   IL1B expression in mRNA and protein levels was jointly regulated by
   TLR2 and TLR4 (Fig. [116]5A–C). Since these cytokines are the main
   factors amplifying inflammation in ALI, we concluded that LCWE induces
   ALI through TLR2, unlike LPS, which acts through TLR4.

Fig. 5. LCWE and LPS induce inflammation in THP-1 monocytes through TLR2 and
TLR4, respectively.

   [117]Fig. 5
   [118]Open in a new tab

   A–C The changes in the mRNA level of IL6, TNFA, and IL1B by RT-qPCR (A,
   B) and the changes of pro-IL-1β protein level by Immunoblotting (C) in
   human THP-1 monocytes in response to LPS or LCWE after knocking down
   TLR2 or TLR4. Statistical analysis: two-way ANOVA test with p-value as
   indicated. Error bar: SD. D–G After knocking out MYD88 (D, E) or TNFR1
   (F, G), the pro-IL-1β protein level by Immunoblotting and the mRNA
   level changes of IL6, TNFA, and IL1B by RT-qPCR in THP-1 monocytes in
   response to LPS or LCWE stimulation. LPS:100 ng/mL, LCWE: 500 ng/mL.
   Statistical analysis: two-way ANOVA test with p-value as indicated.
   Error bar: SD.

   MYD88 is the canonical adaptor of TLR-mediated signaling
   pathways^[119]26. To verify the essential role of MYD88 in LCWE-induced
   ALI, we knocked out MYD88 in THP-1 monocytes and induced cells with LPS
   or LCWE. The results showed that MYD88 ablation blocked LPS or
   LCWE-activated proinflammatory cytokine expression, including IL6,
   TNFA, and IL1B (Fig. [120]5D, E). Similarly, when we knocked down TNFR1
   (Fig. [121]5F, G), both IL6 and IL1B levels were significantly
   downregulated in response to LCWE or LPS stimulation, suggesting that
   TNFα is an amplifying factor of inflammatory signaling in ALI induced
   by LCWE or LPS.

NLRP3 inflammasome contributes to inflammation induced by LCWE

   Overactive inflammation leads to the formation of
   inflammasomes^[122]29–[123]32. To investigate whether the NLRP3
   inflammasome is involved in LCWE-induced ALI, we knocked out the sensor
   protein NLRP3 with CRISPR-Cas9 lentivirus in THP-1 cells and induced
   cells with LCWE (Fig. [124]6A). Interestingly, quantitative PCR results
   indicated that the loss of NLRP3 significantly blocked the mRNA
   expression of IL6, TNFA, and IL1B promoted by LCWE, similar to LPS. At
   the protein level, pro-IL-1β induction by LCWE or LPS was significantly
   disrupted by NLRP3 knockout (Fig. [125]6B). Both LCWE and LPS promoted
   IL-1β mRNA expressions in either THP-1 macrophages or primary BMDM,
   suggesting macrophages are of importance for LCWE-induced inflammation
   (Fig. [126]S5A–D). We prepared BMDM cells from Wild-type (WT) or NLRP3
   knockout (Nlrp3^−/−) mice and treated cells with LPS or LCWE. Similarly
   to those in THP-1 monocytes, in BMDM macrophages, the Il6, Tnfa, and
   Il1b mRNA expressions were significantly upregulated by LCWE or LPS,
   but blocked by Nlrp3 genetically knocked out (Fig. [127]6C).
   Immunoblotting assays confirmed that pro-IL-1β was significantly
   increased by Nlrp3 ablation in response to LCWE or LPS (Fig. [128]6D),
   suggesting processing of cleaved IL-1β was inhibited. We further wanted
   to know the Nlrp3 change in the mouse lungs treated with LCWE or LPS.
   Expectedly, Nlrp3 protein levels were significantly upregulated by
   LCWE, similar to LPS (Fig. [129]6E). Since Nlrp3 inflammasome
   activation leads to cleaved IL-1β release, we detected both pro-IL-1β
   and cleaved IL-1β changes. LCWE decreased the level of pro-IL-1β
   proteins and promoted the production of cleaved-IL-1β (Fig. [130]6F).
   However, when Nlrp3 was knocked out, pro-IL-1β proteins increased but
   cleaved IL-1β proteins decreased in response to LCWE induction compared
   to wild-type animals (Fig. [131]6F). These data together showed that
   NLRP3 is an important mediator in the inflammation associated with ALI
   induced by LCWE.

Fig. 6. NLRP3 ablation disrupted the expression of inflammatory genes and the
process of cleaved IL-1β in THP-1 monocytes and damaged lungs with ALI in
response to LPS or LCWE.

   [132]Fig. 6
   [133]Open in a new tab

   A, B After NLRP3 ablation in the human THP-1 monocyte cell line, the
   mRNA levels of IL6, TNFA, and IL1B (A) and the protein levels of
   pro-IL-1β (B) in response to LPS and LCWE were measured with qPCR and
   WB. Statistical analysis: two-way ANOVA test with p-value as indicated.
   Error bar: SD. C, D After Nlrp3 knockout in the mouse primary BMDMs,
   the mRNA levels of IL6, TNFA, and IL1B (C), and the protein levels of
   pro-IL-1β (D) were measured. Statistical analysis: two-way ANOVA test
   with p-value as indicated. Error bar: SD. E The protein level changes
   of Nlrp3 induced by LPS or LCWE in the mouse pooled ALI lung samples
   (n = 5) were analyzed by immunoblotting. One-way ANOVA test with
   p-value as indicated. Error bar: SD. F Before or after Nlrp3 knockout,
   the protein level changes of Nlrp3 and the pro- and cleaved- IL-1β
   levels induced by LCWE in the mouse pooled ALI lungs samples (n = 5)
   were analyzed by immunoblotting. Statistical analysis: two-way ANOVA
   test with p-value as indicated. Error bar: SD.

Nlrp3 ablation disrupts inflammation and ALI induced by LCWE

   To support the above hypothesis, we continued investigating the NLRP3
   deletion effect on the downstream cytokines. We treated Nlrp3^−/− mice
   with LCWE, using wild-type animals as controls. We found that the mRNA
   levels of Il6, Tnfa, and Il1b were all promoted in response to LCWE,
   and Nlrp3 deletion significantly decreased them (Fig. [134]7A).
   Subsequently, we stained the lung tissues with anti-IL-1β and
   anti-IL-10, and repeatedly found that IL-1β and IL-10 increased upon
   LCWE stimulation. Importantly, Nlrp3 loss disrupted IL-1β and IL-10
   expression induced by LCWE, suggesting the promoted inflammation was
   blocked by Nlrp3 deletion (Fig. [135]7B). To know Nlrp3 effects on cell
   migration to ALI tissues, we immunostained the mouse lungs with CD11b
   and CD14 antibodies. It showed that the migrated cells were mainly
   CD11b positive and partially CD14 positive. The infiltration of these
   immune cells in the lung was blocked by Nlrp3 deletion.

Fig. 7. The Nlrp3 deletion blocked inflammation, and macrophage infiltration
in ALI induced by LCWE.

   [136]Fig. 7
   [137]Open in a new tab

   A The WT or Nlrp3^−/− mice lungs were treated with LCWE at 3 mg/kg, the
   mRNA levels of Il6, Tnfa, and Il1b were measured by RT-qPCR (n = 5).
   Statistical analysis: two-way ANOVA test with p-value as indicated.
   Error bar: SD. B The above mouse lungs were stained with the specific
   antibodies against IL-1β, CD11b, CD14, IL-10, and NLRP3. C The WB
   results of Nlrp3, Stat3, NF-κB p-p65, NF-κB p65, pro-IL-1β,
   cleaved-IL-1β in WT or Nlrp3^-/- mice treated with LCWE. * indicates
   the specific band.

Discussion

   In this study, we successfully established a murine model of ALI
   induced with Gram-positive bacteria cell wall extract (Figs. [138]1,
   [139]S1). As a technical control, 3 mg/kg of LPS was often used to
   induce ALI for the study of Gram-negative bacterial pneumonia. In our
   model, 1 to 6 mg/kg of LCWE induced most mice with comparable
   inflammation to LPS, and we chose 3 mg/kg of LCWE for the ALI model and
   80% of mice appeared ALI. Following damage to lung parenchyma or
   vasculature, both ALI and ARDS have features of rapid respiratory
   failure^[140]33. Patients with ARDS often die from respiratory
   failure^[141]34,[142]35, and another challenge for the therapy of this
   severe disease ARDS is tissue heterogeneity^[143]36. Thus, better
   understanding of the cause and mechanism of subphenotypes of ARDS is
   essential for precision medicine^[144]34. Studies on ALI caused by
   Gram-positive bacteria will shed light on the target therapy of ARDS.
   Even though a couple of studies with Gram-positive bacteria-induced ALI
   have unveiled signaling pathways and key factors in the process, a lack
   of understanding of differentially regulated genes and a
   well-established model impedes the identification of new targets of
   ALI. In addition, the cell wall extracts of Gram-positive bacteria have
   not been used to establish a murine ALI model. Based on our knowledge,
   this is the first report of an experimental murine ALI model induced
   with gram-positive cell wall extract. Hence, our study is important for
   the clinic to find new therapy for ALI/ARDS.

   Using the established LCWE-induced ALI model, our transcriptomic
   analyses highlighted that signaling pathways of Toll-like receptor,
   NF-κB, and TNF are significantly activated in LCWE-induced ALI
   (Figs. [145]2, [146]S2). Expression levels of the key genes which
   include Nfkbia, Tnfa, Il6, and CD14 in these pathways were verified by
   quantitative PCR (Fig. [147]3). As a result, lung cells develop
   inflammation and immunity to clear pathogens and protect themselves
   from damage^[148]37. Mice with stimulation of LCWE have focal coronary
   arteritis, and TLR2 ablation reduces severe symptoms in Kawasaki
   disease^[149]38. In monocytes, we showed that LCWE triggers
   inflammatory signaling through TLR2, different from that LPS depends on
   TLR4. Subsequently, the activated signaling pathways converge to MYD88.
   These results were consistent with previous reports in
   macrophages^[150]38. In addition, we found that the down-regulation of
   the TLR receptor (TNFR1) impaired inflammatory cytokine expression
   (Fig. [151]5F, [152]G), indicating the amplification role of TNF
   signaling in the inflammatory process in ALI. Hence, our detailed
   studies will provide insights for ALI induced by Gram-positive and
   negative bacteria components.

   The key regulator in ALI induced by Gram-positive bacteria has been
   kept debate. In the study, using our well-established LCWE-induced cell
   model, we indicated that NLRP3 promotes LCWE-induced inflammation in
   monocytes, macrophages, and mouse lung and it is the essential
   regulator in LCWE-induced inflammation (Figs. [153]6 and [154]7).
   Strong inflammation in lung parenchyma and immune cells develops
   cytokine storm syndrome, which is mediated by inflammasomes and
   accompanied by cell death^[155]22,[156]39,[157]40. Inflammasomes are
   protein complexes activated by endogenous and exogenous pathogen
   stimuli^[158]41, and the NLRP3 inflammasome is formed upon LPS
   stimulation in ALI^[159]21,[160]42. Hence, targeting NLRP3 inflammasome
   is an essential way to avoid damage from ALI induced by LPS^[161]39.
   However, due to the complexity of causes for the formation of NLRP3
   inflammasome, more knowledge of the role of NLRP3 inflammasome in
   LCWE-induced ALI is very important to identify new therapy targets. In
   the mouse lung with ALI, the elevation of Nlrp3 and cleaved-IL-1β upon
   LCWE was disrupted by Nlrp3 deletion (Fig. [162]6E, [163]F), indicating
   that pyroptosis is possibly induced by LCWE and inhibited by loss of
   function of NLRP3 inflammasome.

   Our data suggested that targeting Nlrp3 reduces LCWE-induced ALI,
   possibly by reduction of cytokines IL-6, TNFα, and IL-1β
   (Fig. [164]7A). It is known that these cytokines account for the immune
   microenvironment in injured tissues^[165]43, attracting most leukocytes
   to the inflammatory loci^[166]44. As shown by CD11b and CD14, both
   neutrophils and monocytes/macrophages infiltrated to the injured lungs
   induced by LCWE but decreased after Nlrp3 loss of function
   (Fig. [167]7B). In addition, NF-κB p-p65 was downregulated when Nlrp3
   was deleted in response to LCWE stimulation (Figs. [168]7C, [169]S6),
   suggesting that amplification of NF-κB signaling was regulated by Nlrp3
   inflammasome in LCWE-induced ALI. The decrease of cleaved-IL-1β
   indicated that pyroptosis in injured lung cells was constrained by the
   loss of Nlrp3 upon LCWE stimulation (Figs. [170]7C, [171]S6). Hence,
   our study highlighted the possibility of targeting the process of
   mature IL-1β, and pyroptosis is a potential strategy to prevent
   Gram-positive bacteria-induced lung damage.

   Taken together, using the cell wall extracts of Gram-positive bacteria,
   we established an LCWE-induced ALI murine model (Fig. [172]8). Unlike
   LPS, which depends on TLR4, LCWE-induced ALI depends on TLR2. Both LCWE
   and LPS depend on Myd88 to activate NF-κB signaling pathways and to
   induce the expression of cytokines such as IL6, TNFα, and IL-1β. These
   cytokines amplify NF-κB signaling and the NLRP3 inflammasome,
   attracting immune cells to the injured tissues and accounting for ALI
   and cell death. Hence, targeting NLRP3 could be an efficient approach
   for the treatment of ALI and ARDS induced by cell wall extracts of
   Gram-positive bacteria.

Fig. 8. Model of LCWE-induced ALI.

   [173]Fig. 8
   [174]Open in a new tab

   In macrophage cells, LCWE activates inflammatory signaling pathways
   through TLR2, unlike LPS through TLR4. MYD88 functions as the adaptor
   for the TLR receptors of both LCWE and LPS, activating the NF-κB
   signaling pathway and the expression of inflammatory cytokines
   including IL6, TNFα, pro-IL-1β, and NLRP3. These cytokines further
   increase inflammatory signaling, resulting in an inflammatory storm and
   ALI.

Methods

Ethics approval

   This study was approved by the Biomedical Ethics and Animal Committee
   at the Anhui Medical University (LLSC20241345). We have complied with
   all relevant ethical regulations for animal use.

Cell culture and antibodies

   THP-1 cells were cultured in RPMI 1640 media containing 10% FBS, 1%
   penicillin/streptomycin, and 1% L-glutamine. THP-1 macrophages were
   obtained from the induced THP-1 with 100 ng/mL of PMA (Abmole, #M4647).
   On day 0, BMDMs were prepared from the femur and tibia of C57BL/6 mice
   and were cultured in DMEM medium with 30% L929 conditioned media and
   10% FBS (Complete Conditioned Medium) as described^[175]45. On day 3,
   the media was gently replaced with fresh Complete Conditioned Medium.
   After another 3 days (day 6), BMDMs were fully differentiated, and
   reseeded into six-well plates with standard DMEM medium for treatments.
   For induction, BMDM cells and THP-1 macrophages were treated with
   100 ng/mL of LPS or 500 ng/mL of LCWE for 24 hours. Anti-CD11b
   (#GB11058, DAB, 1:600), anti-GAPDH (#AC22027016, 1:1000), and
   anti-Actin (#AC220730003, 1:1000) are from Servicebio. Anti-CD14
   (#60253-1-Ig, DAB, 1:400) and anti-hIL-1β (#66737, DAB, 1:400) are from
   Proteintech. The antibodies from other suppliers include anti-CD4
   (Abcam, #ab183685, DAB, 1:1000), anti-NLRP3 (Cell Signaling, #15101,
   WB, 1:1000), anti-NF-kB p65 (Cell Signaling, #8242, WB,1:1000),
   anti-NF-kB p-p65 (Santa Cruz, #sc-136548, WB, 1:1000) anti-NLRP3
   (Servicebio, #[176]GB114320, DAB, 1:300), anti-MYD88 (Wanleibio,
   #WL02494, 1:1000), anti-cleaved IL-1β (Affinity, #AF4006, WB, 1:1000)
   and anti-IL-10 (HUABIO, #ER65464, DAB, 1:200).

Mouse and ALI model

   The C57BL/6 wild-type and Nlrp3^−/− mice were maintained in the animal
   center at the Anhui Medical University. To establish the mouse model of
   acute lung injury (ALI), LPS (Sigma, #L2880) or LCWE (Prepared in the
   study) was administered into the lungs of 6-week-old male mice from the
   trachea at 1, 3, and 6 mg/kg after the mice were anesthetized with
   pentobarbital sodium at a dose of 3 mg/kg. After 24 h, the mice were
   sacrificed, and bronchoalveolar lavage fluid (BALF) and lung tissues
   were collected for RT-qPCR, immunoblotting, and ELISA assays. For
   tissue preparation of immunostaining, the mice were anesthetized with
   avertin and followed by perfusion. In detail, the thoraxes were opened,
   and the right auricle was cut. A perfusion needle was inserted into the
   left ventricle from the apex of the heart, and precooled PBS was
   injected until the liver turned pale.

Lung injury scoring

   Lung injury was assessed according to the pathological manifestation of
   each section, including four items: the thickness of the alveolar wall,
   lung edema, neutrophil infiltration, and hemorrhage. Three pathologists
   blindly scored each item in six random fields with scores ranging from
   0 to 4: normal (0), mild (1), moderate (2), moderate severe (3), and
   severe (4). The lung injury score of each sample is the sum of the
   average scores of each item.

ELISA

   Lung tissue was minced in ELISA lysate buffer. After centrifugation of
   8000 × g at 4 °C for 5 min, the supernatant was collected for assay,
   and the protein concentration of the supernatant was measured by BCA
   assay. 200 μL of BALF was applied directly for assay. The
   concentrations of IL-1β (IL-1β ELISA kit: #KE10003, Proteintech) and
   IL-10 (IL-10 ELISA kit: #KE10103, Proteintech) were determined
   according to the manufacturer’s instructions.

Lactobacillus casei culture and preparation of LCWE

   Lactobacillus casei (ATCC11578) was purchased from Xinyang Zhongjian
   Metrology Biotechnology, China, and was cultured in MRS media, which
   were prepared at 1:1 (V/V) with the Part A and B media and autoclaved
   separately. Part A included glucose (20 g/L) and magnesium sulfate
   heptahydrate (0.58 g/L). Part B contained peptone (10 g/L), beef
   extract (10 g/L), yeast extract (5 g/L), dipotassium hydrogen phosphate
   (1.5 g/L), Tween-80 (1 mL/L), diammonium citrate (2 g/L), sodium
   acetate (3 g/L), and manganese sulfate monohydrate (0.25 g/L). The pH
   was adjusted to 6.2–6.6. LCWE was prepared as described with minor
   modifications^[177]46. In detail, the bacterial cells were incubated in
   the MRS media without shaking for 48 h at 37 °C and were collected at
   10,000 × g for 10 minutes at 4 °C. The bacterial pellet was suspended
   in PBS (twice of the packed cell volume) and 4% SDS (4 times of the
   packed cell volume) sequentially and kept at room temperature for 18 h.
   Subsequently, it was thoroughly washed 8 times with PBS through
   centrifugation to eliminate SDS, confirmed by the absence of
   precipitation with 2 M KCl. The PBS-washed pellet was sequentially
   incubated with 250 μg/mL of RNase A, DNase A, and Trypsin (SparkJade).
   The resulting cell wall preparation was extensively washed with PBS and
   sonicated for 2 h at the concentration of 0.25 g of wet pellet/mL PBS
   at maximum power. The resulting cell wall fragments were spun at
   12,000 rpm for 20 min to collect the supernatant followed by
   centrifugation at 40,000 × g for 1 h at 4 °C. The concentration of
   carbohydrates in the cell wall extract was determined using the Phenol
   Sulfuric Acid method^[178]46.

Immunohistology and immunoblotting

   Hematoxylin & Eosin staining and immunostaining were performed on the
   paraffin sections of lung tissues as previously described^[179]47. For
   immunoblotting, the dounced tissues or the cultured cells were washed
   with ice-cold PBS for three times. Subsequently, the samples were lysed
   on ice with RIPA buffer (50 mM Tris pH 8.0, 150 mM NaCl, 0.05% SDS,
   0.5% DOC, 1% NP-40) supplemented with protease inhibitors (Abmole). The
   protein concentration was determined using a BCA assay (Vazyme). The
   protein lysates were loaded on SDS-PAGE and transferred to the
   nitrocellulose membrane (Merck, #IPVH00010). The blot was blocked with
   5% non-fat milk in TBST and incubated with the primary antibodies in 2%
   BSA in TBST overnight at 4 °C. After incubation with the HRP-linked
   secondary antibodies (Abbart) for 1 h at room temperature, the bands
   were detected with ECL and scanned with a SHST ChemiDoc Imager
   (Shenhua, China).

pLKO-shRNA and CRISPR-Cas9 vectors, lentivirus preparation and infection

   pLKO.1 from Addgene was used for shRNA knockdown. LentiCRISPR-V2 was
   used for RNA-guide gene editing. For lentivirus packaging, the pLKO.1
   or LentiCRISPR-V2 plasmids with targeting sequences, together with the
   package plasmids of VSV-g and Δ8.9 were transfected into 293T cells
   using PEI (Yeasen, #40816ES02/03). Lentiviruses were collected after
   transfection for 48 and 72 h. For infection, THP-1 cells were cultured
   with 50% virus (original virus: fresh media = 1:1) and 8 μg/ml of
   polybrene.

RT-PCR and qPCR

   Total RNA was extracted using TRIzol (Ambion, #400008). The RNA
   concentration was measured using SH-NanoOne (Shenhua, China). cDNA was
   synthesized using MonScript RTIII All-in-One Mix with dsDNase (Monad,
   #MR05101). Quantitative PCR was performed using MonAmp qPCR Mix reagent
   (Monad, #00007547-130712) with a LightCycle real-time PCR system
   (Roche). GAPDH was used for normalization.

RNA-seq and bioinformatics analysis

   RNA-seq and bioinformatics analysis were performed at the Sequencing
   and Bioinformatics Center at the Anhui Medical University. In detail,
   mRNA was isolated using Oligo-dT magnetic beads (Vazyme, #N401) from
   the total RNA. cDNA was synthesized using SuperScript II (Invitrogen,
   #18064014). The cDNA library was prepared using Tn5 DNA Library Prep
   Kit (Transgen, #KP101-11). The cDNA library was applied for PE150
   sequencing. The quality of Fastq data was controlled by Fastqc and
   analyzed using the RNA-seq pipeline^[180]48. The relative expression
   levels were determined with featureCounts^[181]49, and the
   differentially regulated genes were called by DEseq2 program in the
   pipeline with fold-change >2 and padj < 0.05. Bioinformatics analyses
   including GO, KEGG, and CIBERSORT were analyzed using R studio.

Statistical and reproducibility

   Statistical analysis was performed using the two-tailed unpaired
   Student’s t-test provided by Microsoft Excel and one-way or two-way
   ANOVA test combined with Tukey’s HSD test in Python (Spyder IDE). A P
   value of <0.05 was considered significant. All experiments were
   performed at least three times independently to ensure reproducibility,
   with consistent results observed across replicates. For experiments
   lacking statistical analysis, reproducibility was assessed by repeating
   the experiments under the same conditions to confirm consistency in
   outcomes.

Supplementary information

   [182]Supplementary Information^ (3.3MB, docx)
   [183]Supplementary Data 1^ (74.2KB, xlsx)
   [184]42003_2025_7462_MOESM3_ESM.docx^ (14.8KB, docx)

   Description of Additional Supplementary Files
   [185]nr-reporting-summary^ (77.7KB, pdf)

Acknowledgements