Cholinergic macrophages promote the resolution of peritoneal inflammation

Significance Acetylcholine, originally recognized as a neurotransmitter, is increasingly acknowledged as a key immune modulator in various physiological and pathological conditions. This study unveils a population of cholinergic macrophages (Mϕs), which are crucial in resolving acute peritonitis. These Mϕs, expressing the enzyme choline acetyltransferase (ChAT) and synthesizing acetylcholine, are primarily monocyte-derived small peritoneal Mϕs, distinct from tissue-resident large peritoneal Mϕs. The selective deficiency of ChAT in Mϕs impedes efferocytosis and delays apoptotic neutrophils clearance, effects reversible by acetylcholine supplementation. These findings illuminate the pivotal role of Mϕ-mediated cholinergic regulation in inflammation resolution and enhance our understanding of the non-neuronal cholinergic system’s role in immune regulation.

receptor (TLR)-MyD88 signaling.Notably, this process was attenuated by the TRIF-dependent TLR signaling pathway.Selective deletion of Chat in Mϕs impeded the resolution of inflammation, a phenomenon that was reversed by ACh supplementation.These findings uncover a unique aspect of the non-neuronal cholinergic axis in the myeloid compartment, highlighting its critical role in controlling the resolution of inflammation.

Results
Emergence of ChAT-Expressing Mϕs during the Resolution Phase of Acute Peritonitis.ChAT expression plays multifaceted roles in regulating immune responses within the immune system (15)(16)(17).These recent findings highlight the need to delineate and clarify the heterogeneous population of ChAT-expressing immune cells and their functional dynamics within specific pathological contexts.To this end, we analyzed the expression pattern of ChAT in the peritoneal lavage of reporter mice that express enhanced green fluorescent protein (GFP) under the control of transcriptional regulatory elements of the Chat gene (Chat-GFP reporter mice) (22).Under resting conditions, B cells constituted the dominant GFP-positive population in peritoneal lavage (Fig. 1A, labeled "0 h").A similar pattern was observed in lipopolysaccharide (LPS; a TLR4 agonist)-induced sterile peritonitis (Fig. 1 A, Upper), albeit with a slight increase in GFP expression among Mϕs and T cells (15% in Mϕs and 10% in T cells at their peak; Fig. 1 A, Upper).In contrast, we observed a striking time-dependent increase in GFPexpressing Mϕs in response to Pam3CSK4 (Pam3), a TLR1/2 agonist (Fig. 1 A, Lower).Following Pam3 stimulation, GFP expression in Mϕs was evident as early as 4 h poststimulation, corresponding to the peak phase of leukocyte recruitment in acute peritonitis (SI Appendix, Fig. S1A).By day 3, Mϕs constituted approximately 50% of the GFP + cells (Fig. 1 A, Lower), with more than 30% of the intraperitoneal Mϕs exhibiting GFP expression (Fig. 1 B and C).At this time point, the inflammatory response was in the resolution phase, as evidenced by the significant decrease in neutrophils and monocytes (SI Appendix, Fig. S1A).Meanwhile, the elevation of GFP expression in T cells, B cells, and neutrophils was only marginal (Fig. 1 B and C and SI Appendix, Fig. S1B).The Pam3-induced Chat-GFP expression in resolution-phase Mϕs was consistent in experiments using various administration doses (SI Appendix, Fig. S1C).In comparison, even though LPS could induce similar peritonitis at the acute phase (SI Appendix, Fig. S1D), its impact on Chat-GFP expression in resolution-phase Mϕs was mild regardless of administration doses (SI Appendix, Fig. S1C).To further evaluate the expression of ChAT in Mϕs during bacterial infection, we adopted a self-resolving model of peritonitis induced by Escherichia coli or Staphylococcus aureus (23).The expression of ChAT Mϕs was significantly increased at the resolution phase (72 h) but not at the acute phase of either E. coli-induced peritonitis (4 h) or S. aureusinduced peritonitis (24 h) (Fig. 1D and SI Appendix, Fig. S1E).
To corroborate the results obtained with the Chat-GFP reporter system, we sorted both GFP-negative and GFP-positive peritoneal Mϕs (SI Appendix, Fig. S1F) and validated the Pam3-induced expression of ChAT at the mRNA (Fig. 1E) and protein (Fig. 1F) levels in these cells.Furthermore, we detected an average intracellular ACh level of ~20 µM per million Chat-GFP-positive peritoneal Mϕs, remarkably higher than that in their GFP-negative counterparts (Fig. 1G).Genetic deletion of the fourth and fifth exons of Chat, guided by Lyz2 cre (Chat fl/fl Lyz2 cre mice), led to an almost complete loss of ChAT protein expression in Pam3-stimulated peritoneal Mϕs (Fig. 1 H and I).In addition, we found an over 80% ACh reduction in the peritoneal lavage from Pam3-treated Chat fl/fl Lyz2 cre mice compared to Chat fl/fl mice (Fig. 1J), consistent with the findings that Mϕs were the main ChAT-expression cell population at this time point and the high intracellular ACh level in these cells.Collectively, our findings underscore the significant upregulation of ChAT and ACh synthesis in peritoneal Mϕs during the resolution phase of the Pam3-induced inflammatory response.

Phenotypic Characteristics of ChAT-Expressing Peritoneal Mϕs.
To delineate the phenotypic characteristics of ChAT-expressing peritoneal immune cells, we performed InfinityFlow analysis, a multiparametric flow cytometry approach (24,25), to profile the expression patterns of 253 surface protein markers under various conditions (Fig. 2A).Seven primary peritoneal immune subtypes were identified: Mϕs, neutrophils, B cells, dendritic cells (DCs), natural killer (NK) cells, αβ T cells, and γδ T cells (Fig. 2B and SI Appendix, Fig. S2A and Table S1).These cell types could be further classified into 23 specific subpopulations based on unique phenotypic profiles (Fig. 2 B-D and SI Appendix, Fig. S2B and Dataset S1).Notably, LPS and Pam3 not only exhibited different capacities to induce ChAT expression in Mϕs at various doses (Figs. 1 A and B and 2C and SI Appendix, Fig. S1C), but rather activated differential Mϕ responses, characterized by the emergence of different Mϕ clusters (Fig. 2 D and E).
The peritoneal cavity hosts two distinct Mϕ subsets: tissueresident large peritoneal Mϕs (LaPMs) and smaller, monocytederived Mϕs (MDMs) termed small peritoneal Mϕs (SmPMs) (26,27).In contrast to homeostatic and LPS-induced conditions, Pam3 stimulation led to a significant accumulation of SmPMs, characterized by a smaller size and lower granularity, as indicated by reduced forward scatter (FSC) and side scatter (SSC) signals, respectively (26) (Fig. 2 E and F).Moreover, these Mϕs exhibited lower expression levels of typical LaPM markers such as CD11b, F4/80, CD24, CD9, CD73, and Tim-4 (28) and displayed increased expression of the adhesion molecules CD43, CD31, and CD11a.These characteristics of SmPMs were even more pronounced in the Pam3-induced GFP-positive Mϕs (Fig. 2 E and F), suggesting selectively enriched GFP expression in SmPMs.Conventional flow cytometry analysis confirmed that over 70% of peritoneal Mϕs at 72 h after various doses of Pam3 stimulation were classified as F4/80 lo Tim-4 − SmPMs, whereas less than 10% of these cells were found in homeostatic and LPS-treated conditions (Fig. 2G and SI Appendix, Fig. S2C).Furthermore, in these Pam3-induced SmPMs, more than 40% of the cells were GFP positive, in contrast to the low levels (≤15%) of GFP expression in both LaPMs and SmPMs under homeostatic and LPS-stimulated conditions (Fig. 2G).
Given the activation of MyD88 signaling by both LPS and Pam3, the relatively lower level of ChAT expression induced by LPS than by Pam3 was intriguing (Fig. 3 A and B).Considering that LPS-TLR4 stimulation activates both MyD88 and TRIF-dependent pathways (Fig. 3C), we hypothesized that the TRIF pathway might impede ChAT expression.As expected, Trif knockdown in BMDMs (Fig. 3L) elevated the LPS-induced ChAT expression to a level similar to that induced by Pam3 (Fig. 3M).These findings suggest that while TLR-MyD88-MAPK p38/ERK signaling promotes ChAT expression in MDMs, TRIF-dependent TLR signaling partially inhibits this process.

ChAT Expression in Mϕs Promotes the Resolution of Inflammation.
To elucidate the role of ChAT expression in Mϕs during the resolution of inflammation, we conducted scRNA-seq on peritoneal cell populations induced by Pam3 in Chat fl/fl Lyz2 cre cells and their control Chat fl/fl counterparts (Fig. 4A).The results showed that Chat deletion in Mϕs was associated with increased neutrophil counts (Fig. 4B and Dataset S4) and a notable downregulation of genes essential for Mϕ-mediated resolution of inflammation, such as Alox15, Timd4, F5, Ltbp1, Cxcl13, and Tgfb2 (30-32) (Fig. 4C Relative expression level (10 -3  and SI Appendix, Fig. S4A and Dataset S5).In Chat-deficient Mϕs, a significantly enhanced degree of active gene transcription was observed, as evidenced by greater levels of nascent (unspliced) mRNAs (SI Appendix, Fig. S4B), than that in their Chat-competent counterparts.Gene set enrichment analysis (GSEA) demonstrated enrichment of these mRNAs among the genes associated with innate immune responses (GO:0002218) (Fig. 4D), indicating a state of prolonged Mϕ activation.Concurrently, there was a profound reduction in the expression of genes associated with endocytosis (GO:0030100) and apoptotic cell clearance (GO:0043277), as well as in the expression of features typical of resolution-phase Mϕs (33), suggesting an impaired function for resolving inflammation (Fig. 4D).
The resolution interval (R i ) is widely applied to quantify resolution speed (23,34), which is defined as the time period between T max (the time point when neutrophil numbers reach their maximum) and T 50 (the time point corresponding to approximately 50% neutrophil reduction).Chat fl/fl Lyz2 cre mice exhibited a prolonged R i (R i ; ~18 h and 39 h in the Chat fl/fl and Chat fl/fl Lyz2 cre mice, respectively) and a delayed reduction in exudate neutrophils (Fig. 4E).The efferocytosis of apoptotic neutrophils by Mϕs was reduced by half in Chat fl/fl Lyz2 cre mice at 72 h (Fig. 4F), resulting in a twofold increase in the number of peritoneal apoptotic neutrophils compared to that in Chat fl/fl mice (Fig. 4G).In contrast, Chat deficiency in B cells, the other main population of ChATexpressing cells under Pam3 stimulation (Fig. 1A), had only a marginal, if any, impact on the neutrophil numbers (Fig. 4H), efferocytosis (Fig. 4I), and peritoneal apoptotic neutrophil counts (Fig. 4J).Collectively, these findings underscore the critical role of ChAT expression in peritoneal Mϕs, rather than B cells, in facilitating inflammation resolution.

ACh Enhances Mϕ Phagocytosis through Nicotinic ACh
Receptors.To investigate the role of Mϕ-derived ACh in the resolution of inflammation, we administered ACh intraperitoneally to Chat fl/fl Lyz2 cre mice following Pam3 treatment (Fig. 5A).Remarkably, ACh supplementation significantly reduced the R i from 35 h to 18 h, effectively normalizing both the decline of neutrophils (Fig. 5B) and the apoptotic neutrophils count in Chat fl/fl Lyz2 cre mice to those in ChAT-competent mice (Fig. 5C).These findings prompted an investigation into the specific receptors mediating the effect of ACh on Mϕ phagocytosis.MDMs were found to express a range of ACh receptors (AChRs), specifically including the M3 and M4 subunits of muscarinic AChRs and the α2, α5, α9, β1, and β2 subunits of nicotinic AChRs (SI Appendix, Fig. S5 A and B and Dataset S6).The intraperitoneal administration of the nicotinic receptor inhibitor mecamylamine (Meca; Fig. 5D), but not the muscarinic receptor antagonist atropine (Atp), led to an increase in the R i from 23 h to 44 h and a delayed reduction in exudate neutrophils (Fig. 5E), diminished efferocytosis (Fig. 5F), and elevated apoptotic neutrophil counts (Fig. 5G).Additionally, ACh significantly enhanced the phagocytosis of latex bead-rabbit IgG-FITC complexes by both murine BMDMs and human blood MDMs, indicating that ACh generally enhanced Mϕ phagocytic activity (Fig. 5H).
Interestingly, despite the removal of the fourth and fifth exons from the Chat transcripts and the absence of ChAT protein expression in the Mϕs from the Chat fl/fl Lyz2 cre mice, an unexpected increase in the number of mRNA reads corresponding to the remaining exons was observed in both the scRNA-seq and bulk RNA-seq data (Fig. 5I and SI Appendix, Fig. S5 C and D and Dataset S6).Moreover, the addition of ACh to murine BMDM cultures decreased Chat-GFP expression in these cells (Fig. 5J).Therefore, these data suggest that environmental ACh negatively affects ChAT expression in Mϕs, serving as a negative feedback mechanism to fine-tune Mϕ functions in regulating inflammation process.
In the present study, our findings reveal a unique role of Mϕs as cholinergic immune cells that emerge in the resolution phase of acute inflammation.Derived from peripheral monocytes, these Mϕs up-regulate ChAT expression via the TLR-MyD88-MAPK p38/ERK signaling pathway.Importantly, the absence of Chat in Mϕs, rather than in B lymphocytes, led to delayed resolution of inflammation.This finding highlights the critical role of Mϕderived ACh in orchestrating the resolution of inflammation.
ChAT expression and ACh synthesis in immune cells have been well documented in lymphocytes, including T cells and B cells, and to a lesser extent, in NK cells (15)(16)(17).Recently, ChAT expression has also been discovered in the myeloid compartment of the immune system.A subset (approximately 4%) of adipose-resident Mϕs could up-regulate ChAT expression for adaptive thermogenesis upon acute cold exposure (18).However, the role of ACh-synthesizing myeloid cells in immune responses and their impact on inflammation and immunity remain largely unexplored.Our study addresses this gap by revealing a notable proportion of peritoneal Mϕs expressing ChAT in the resolution phase of TLR ligands-induced peritonitis and bacterial infections.Unlike adipose-resident Mϕs, peritoneal resident LaPMs exhibit limited ChAT upregulation in response to these stimuli.In contrast, approximately 40% of the SmPMs displayed robust ChAT expression in response to Pam3 stimulation, and similar percentages were observed in conditions induced by TLR ligands and bacteria that activate MyD88 signaling.This inducible ChAT expression kinetics in SmPMs contrasts with the relatively stable ChAT expression in peritoneal B cells.These SmPMs, which have a distinct transcriptional profile independent of Gata6 (42,43) and lower expression levels of typical LaPM markers such as F4/80, CD11b, and Tim-4, are proposed to constitute a group of heterogeneous, recruited MDMs.Consistent with this notion, BMDMs but not isolated peritoneal LaPMs up-regulated ChAT expression in response to multiple TLR signaling pathways.These findings demonstrate the intrinsic capacity of specific Mϕ subsets to synthesize ACh, thus expanding our understanding of the non-neuronal cholinergic system within the immune landscape.
The role of ACh in the immune system is multifaceted and appears paradoxical.While immune cell-derived ACh has been shown to protect mice against sepsis and reduce the severity of experimental autoimmune encephalomyelitis, this molecule also enhances inflammation against viral infections and boosts immune responses to intestinal pathogens (15,17,21,44).This complexity reflects the diverse roles of ACh in immune modulation and the variety of neurotransmitter receptors expressed by immune cells.Our study found that deleting Chat in Mϕs led to a delay in inflammation resolution, which was reversed by ACh supplementation.This underscores the importance of Mϕ-derived ACh in the inflammation resolution process.Interestingly, the conditional Chat knockout prominently altered the transcriptional landscape of Mϕs, suggesting an autocrine and/or paracrine mechanism of action.One key finding is that the absence of Chat in Mϕs impaired the efferocytosis of apoptotic neutrophils, thereby slowing the reduction in the number of exudate neutrophils.Moreover, the administration of nicotinic receptor inhibitors further suggested the importance of ACh-nicotinic receptor interaction in efferocytosis, as it led to reduced clearance of apoptotic neutrophils by Mϕs.Intriguingly, despite B cells being a notable ChAT-expressing population within the peritoneal cavity, Chat deficiency in B cells did not significantly alter the resolution process.These findings highlight the cell type-specific roles of ACh in immune regulation.Taken together, these insights into the role of Mϕ-derived ACh in immune regulation not only unravel the complexities of cholinergic signaling but may also open broad avenues for therapeutic interventions.
In immune cells, the regulation of ChAT expression is complex and diverse, often linked to cellular activation.ChAT expression in T cells is modulated by TCR engagement and cytokines such as IL-21, and in B cells, it is triggered by antigen receptor activation or MyD88-dependent signaling (13,16,21).Our study extends this understanding to peritoneal Mϕs and demonstrates that ChAT expression in peritoneal Mϕs is regulated via a MyD88-dependent pathway involving MAPK p38 and ERK signaling.Interestingly, although LPS effectively activates the TLR-MyD88 pathway, it only marginally induces ChAT expression in Mϕs.This difference could be attributed to the engagement of the TRIF-dependent TLR signaling pathway by LPS, which may partially counteract ChAT expression.Moreover, our tests of a range of neurotransmitters and cytokines did not reveal significant ChAT expression in BMDMs, suggesting a highly selective induction mechanism in Mϕs.
Our study identified distinct ChAT expression patterns in SmPMs and LaPMs.SmPMs but not LaPMs are the predominant peritoneal Mϕ population that express ChAT in response to TLR agonists and bacterial infection.This divergence suggests differing functional roles and activation thresholds among Mϕ subsets, an aspect that merits further investigation.Intriguingly, we observed a paradoxical increase in Chat gene transcription following ChAT protein depletion coupled with a reduction in ChAT expression upon exposure to elevated extrinsic ACh.This phenomenon suggests the presence of a sophisticated negative feedback mechanism involved in ChAT regulation, highlighting the complexity and adaptability of the cholinergic system during the immune response.Future research aimed at deciphering the molecular process regulating ChAT expression across lymphoid and myeloid cell populations could offer crucial insights into the diverse and context-dependent regulation of immune-neural cross talk.
In summary, our research identified a significant population of ChAT-expressing Mϕs that are crucial for resolving inflammation.These findings enhance our understanding of the intricate interplay between cholinergic signaling and Mϕ-mediated inflammatory responses, enriching the broader landscape of the involvement of the non-neuronal cholinergic system.A better understanding of these cholinergic Mϕs may illuminate different pathways for therapeutic exploration, potentially revolutionizing treatment strategies for various immune-related conditions.

Culture of BMDMs, Peritoneal Resident Mϕs (PMs), and Human MDMs.
For BMDMs culture, tibia and femur bone were harvested bilaterally and bone marrow was flushed out using a syringe filled with phosphate-buffered saline (PBS) containing 1% fetal bovine serum (FBS; Cat# 10270106; Thermo Fisher Scientific) and 2 mM EDTA.Bone marrow cells were plated in Iscove's Modified Dulbecco's Medium (Cat# 12200036; Thermo Fisher Scientific) supplemented with 10% FBS and 10 ng/mL M-CSF (Cat# 51112-MNAH; Sino Biological) for 7 d in order to obtain differentiated BMDMs.For stimulation, BMDMs obtained at the end of the 7-d culture were treated with indicated stimulation for an additional 24 h prior to cell harvesting.
For peritoneal resident Mϕs culture, we implemented a protocol to isolate and culture these cells.Briefly, after killing the mice, the abdominal area was disinfected with 70% ethanol.A midline incision was made, and 10 mL of PBS was injected into the peritoneal cavity.The peritoneal lavage cells were collected, pelleted by centrifugation, and resuspended in RPMI 1640 medium (Cat# C11875500BT; Thermo Fisher Scientific) supplemented with 10% FBS.After a 1 h incubation at 37 °C, the cells were washed five times with 1 mL of warm PBS to remove nonadherent cells, followed by a 24 h resting period.The cells were stimulated with indicated stimulation for 24 h prior to cell harvesting, and GFP expression was assessed by flow cytometry.The detailed activation conditions are listed in SI Appendix, Table S2.
For human MDMs culture, monocytes were isolated from peripheral blood and cultured as previously described (45,46).Briefly, the buffy coats were collected from the blood of healthy individual donors to the Guangzhou Blood Center (Guangzhou, China).Primary blood mononuclear cells were isolated by Ficoll density gradient centrifugation at 450 g.CD14 + monocytes were then purified using magnetic beads (Cat# 130-050-201; Miltenyi Biotec).For Mϕ differentiation, 8 × 10 5 monocytes were cultured per well in 24-well plates (Cat# P24-1.5H-N;Cellvis), using RPMI 1640 medium supplemented with 10% FBS and 100 ng/mL of human recombinant M-CSF (Cat# 11792-HNAH; Sino Biological).Seven days later, the cells were replenished with fresh media without M-CSF and subjected to phagocytic activity testing as described below.
Bacteria.E. coli (ATCC 25922) or S. aureus (ATCC 29213) was grown in Tryptic Soy Broth (Cat#211825; BD Bioscience) and harvested at mid-log phase and washed twice with sterile saline before inoculation into mouse peritoneal cavity.
Flow Cytometry.For the analysis of BMDMs, single-cell suspensions were obtained by incubating BMDMs in 5 mM EDTA at 37 °C for 5 min and then resuspended in FACS buffer (1% FBS and 2 mM EDTA in PBS) and stained with antibodies against CD45, CD11b, and F4/80.For the peritoneal immune cell analysis, monoclonal antibodies against CD45, CD11b, Ly6G, Ly6C, F4/80, CD3, and B220 were used.After washing three times, cells were resuspended in 200 µL of FACS buffer.Flow cytometry acquisition was performed on a Cytoflex S flow cytometer (Beckman Coulter).Finally, the data were analyzed using FlowJo software (TreeStar).Detailed descriptions of the antibodies used for flow cytometry are provided in SI Appendix, Table S3.
RT-PCR.The total RNA of BMDMs was extracted by TRIzol reagent (Cat# 15596018; Thermo Fisher Scientific) according to the manufacturer's protocol.Total RNA was used to generate cDNA with All-in-One RT Master Mix (Cat# G492; Applied Biological Material).Real-time qPCR was performed using the SYBR Green qPCR Master Mix (Cat# B21203; Bimake).All reactions were conducted using a LightCycler 480 Instrument (Roche) in triplicate.For detection of AChRs by RT-PCR, cDNA was obtained as described above.Bands of expected sizes from electrophoresis gels were used to retrieve DNA for sequencing verification of the amplicon identities.The sequences of primers used are listed in SI Appendix, Table S4.
Small Interfering RNA (siRNA) Transfection.siRNAs (GenePharma) were mixed with jetPRIME transfection reagent (Cat# 101000046; Polyplus-transfection) according to the manufacturer's instructions and added to the BMDM culture at a final concentration of 20 nM.After incubation with the siRNA complex for 12 h, BMDMs were stimulated with LPS or Pam3 for 24 h as indicated in the figure legends.Details of the siRNA sequences are listed in SI Appendix, Table S5.
UPLC-MS/MS.GFP-negative and GFP-positive peritoneal Mϕs sorted from FACS were washed three times with PBS containing 0.1% pyridostigmine bromide.Subsequently, cell pellets were snap-frozen in liquid nitrogen and stored at −80 °C until analysis.To detect ACh in peritoneal lavage fluid from Pam3-treated (72 h) Chat fl/fl and Chat fl/fl Lyz2 cre mice, peritoneal cavities were lavaged with 1 mL of sterile PBS containing 0.1% pyridostigmine bromide.Peritoneal lavage fluid was collected and centrifuged (380 g, 10 min) to remove cells.The remaining supernatant is centrifuged at high speed (12,000 g, 10 min).Following centrifugation, supernatants were snap-frozen in liquid nitrogen and stored at −80 °C until analysis.Samples were spiked with Tolbutamide internal standard (Cat# 64-77-7; Sigma-Aldrich) and analyzed by LC-tandem MS on an AB 6500Plus mass spectrometer (AB Sciex, United States) equipped with an AB EXIONLC AD instrument (AB Sciex, United States).
ACh and AChR Inhibitor Treatment.For ACh treatment, Chat fl/fl and Chat fl/fl Lyz2 cre mice were injected intraperitoneally with Pam3, followed by intraperitoneal injection of ACh (5 mg/kg) 0.5 h later, and continued daily for the duration of the experiment.For AChR inhibitor treatment, wild-type B6 mice were injected intraperitoneally with Pam3, followed by intraperitoneal injection of muscarinic receptor inhibitor atropine (Atp; 1 mg/kg) or nicotinic receptor inhibitor mecamylamine (Meca; 1 mg/kg) 0.5 h later, and continued daily for the duration of the experiment.At designated points, peritoneal exudate was collected by lavaging with 5 mL PBS.The exudate cells were subsequently used for flow cytometric analyses.

Measurement of Neutrophil Apoptosis.
To determine neutrophil apoptosis, peritoneal lavage cells were labeled with annexin V and an anti-Ly6G antibody.The annexin V + Ly6G + neutrophil population was determined by flow cytometry.
Phagocytosis Assays.For phagocytosis assays, cells were plated to 90% confluence in 24-well glass bottom plates and pretreated with cholinesterase inhibitor pyridostigmine bromide (Cat# HY-B0207A; MedChemExpress) and ACh (Cat# HY-B0282; MedChemExpress) for 24 h.Subsequently, cells were assessed for phagocytic activity using the Phagocytosis Assay Kit (Cat# 500290; Cayman Chemical).Briefly, BMDMs or human MDMs were labeled with 1 μM Cell Trace Far Red (Cat# C34564; Thermo Fisher Scientific).After incubation for 20 min at 37 °C, cells were washed and then the latex beads conjugated to rabbit IgG-FITC (1:400 dilution) were added to the culture in the absence or presence of ACh.During live imaging, the cells were incubated in a stage-top incubator (Tokai Hit) at 37 °C and 5% CO 2 .Images were then obtained continuously every 20 min for 2 h using an inverted microscope (Nikon, ECLIPSE Ti2) equipped with the spinning disk unit (Yokogawa, CSU-W1) and sCMOS camera (Hamamatsu, ORCA-Fusion BT).
RNA-Seq.BMDMs from Chat fl/fl and Chat fl/fl Lyz2 cre mice were treated with or without LPS, Pam3, or FSL-1 for 24 h, and then washed twice with PBS.The total RNA of treated BMDMs was extracted by TRIzol reagent according to the manufacturer's protocol.Poly-A mRNA was extracted from total RNA using Oligo-dT beads in a NEBNext Poly(A) RNA Magnetic Isolation Module (New England Biolabs) and a RiboZero Magnetic Gold Kit (Epicentre), and RNA-Seq libraries were prepared using KAPA Stranded RNA-Seq Library Prep Kit (Illumina) according to the manufacturer's protocols.Libraries were paired-end sequenced on a NovaSeq sequencer (Illumina).The reads were aligned to the mm10 mouse genome using Salmon, and the aligned data were analyzed using R software.The analysis was conducted following the Bioconductor RNA-seq workflow and differential gene expression was analyzed using the R package DESeq2.

Fig. 1 .
Fig. 1.Upregulation of ChAT expression in Mϕs during the resolution of inflammation.(A) Constitution of the ChAT-expressing (GFP + ) population in peritoneal lavages at the indicated time points after intraperitoneal administration of either saline (indicated as 0 h), LPS, or Pam3.(B and C) Representative histogram (B) and percentages (C) of ChAT-expressing cells in Mϕs, B, and T cells.(D) Mice were inoculated with 10 6 CFU of either E. coli or S. aureus by intraperitoneal injection, and percentages of ChAT-expressing cells in Mϕs were analyzed by flow cytometry.(E and F) ChAT expression within GFP − and GFP + peritoneal Mϕs was assessed via qPCR (E) and immunoblotting (F).(G) ACh production was assessed by UPLC-MS/MS in GFP − and GFP + peritoneal Mϕs isolated from Pam3-treated (72 h) ChAT-GFP mice.(H and I) ChAT expression in Pam3-treated (72 h) peritoneal Mϕs from Chat fl/fl and Chat fl/fl Lyz2 cre mice was determined by qPCR (H) and immunoblotting (I).(J) The concentration of ACh in peritoneal lavage fluid from Pam3-treated (72 h) Chat fl/fl and Chat fl/fl Lyz2 cre mice was determined by UPLC-MS/MS.Data are from three independent experiments.Error bars indicate mean ± SEM.Statistics: (C and D) Two-way ANOVA corrected by Šidák's test.(E, G, H, and J) Student's t test.*P < 0.05, **P < 0.01, and ***P < 0.001.E. coli, Escherichia coli; S. aureus, Staphylococcus aureus.

Fig. 2 .
Fig. 2. InfinityFlow analysis reveals the ChAT enrichment in F4/80 lo Tim-4 − small peritoneal Mϕs.(A) Scheme of the InfinityFlow analysis on the cell surface markers.(B) Cellular discrimination was achieved through t-SNE dimensionality reduction, with clustering via FlowSOM and ConsensusClusterPlus based on the 93 discriminating markers defined in SI Appendix, Table S1.(C) The frequencies of GFP − and GFP + peritoneal leukocytes at 72 h posttreatment.(D) Subclustering of peritoneal Mϕs based on uniform manifold approximation and projection (UMAP), revealing distinct phenotypic cluster profiles detailed in Dataset S1. (E) UMAP of peritoneal Mϕs from control, LPS-, or Pam3-treated mice.(F) Normalized expression levels of various markers on peritoneal Mϕs, determined by InfinityFlow.(G) Percentages of F4/80 hi Tim-4 + LaPMs and F4/80 lo Tim-4 − SmPMs (Upper) and the proportions of ChAT expression in these cells (Lower) were analyzed by conventional flow cytometry in control, LPS-, or Pam3-treated mice.Representative results of three independent experiments are shown.

BFig. 3 .
Fig. 3. Induction of ChAT expression in BMDMs via MyD88-p38/ERK-dependent TLR signaling.(A) Representative flow cytometry results of GFP-expressing cells in resident peritoneal Mϕs (PMs) or BMDMs from ChAT-GFP mice stimulated with specified TLR ligands.(B) Immunoblotting of ChAT expression in BMDMs from wild-type B6 mice following specified treatments for 24 h.(C) A schematic diagram of TLR signaling and the inhibitors used in this study.(D) UMAP of scRNA-seq data obtained from peritoneal lavage cells (Left), and the frequencies of immune cell clusters (Right) within the unsorted, GFP + and GFP − Mϕs were assessed.(E) Gene set enrichment analysis (GSEA) of TLR2 signaling (R-MMU-168179), chemical synapses (R-MMU-112315), signaling to ERKS (R-MMU-198753), and ERK/ MAPK targets (R-MMU-187687) activities.The top portion of the plot shows the running enrichment score as the analysis walks down the ranked list.The bottom portion of the plot shows where the member genes of the gene set appear in the ranked list.The normalized enrichment score (NES) and the nominal P values are shown.(F-J) BMDMs were treated with the MyD88 inhibitor TJ-M2010-5 (30 μM; F), the p38 inhibitor SB 203580 (25 μM; G), the ERK inhibitor U0126 (25 μM; H), the JNK inhibitor SP600125 (10 μM; I) or in combination with SB 203580 and U0126 (J) for 24 h.Subsequently, the GFP expression levels were assessed by flow cytometry.(K) BMDMs were treated with LPS or Pam3 for 24 h following transfection with si-Myd88-1, si-Myd88-2, si-p38-1, si-p38-2, si-Erk2-1, si-Erk2-2, or nonspecific siRNA control (si-NC).Subsequently, the GFP expression levels were assessed by flow cytometry.(L) BMDMs were treated with Pam3 for 24 h following transfection with si-Trif-1, si-Trif-2, or si-NC.The relative expression of Trif RNA was determined using qPCR.(M) The expression levels of GFP in BMDMs under indicated treatments were assessed by flow cytometry.Error bars indicate mean ± SEM.Statistics: (F-K and M) Two-way ANOVA corrected by Šidák's test.(L) One-way ANOVA corrected by Dunnett's test.*P < 0.05, **P < 0.01, and ***P < 0.001.Med, Medium.

EFig. 4 .
Fig. 4. Role of ChAT-expressing peritoneal Mϕs in resolving peritoneal inflammation.(A) Scheme of the scRNA-seq analysis on peritoneal lavage cells from i.p.Pam3-treated Chat fl/fl and Chat fl/fl Lyz2 cre mice.(B) UMAP of scRNA-seq data obtained from peritoneal lavage cells (Left), and the frequencies of immune cell clusters (Right).(C) Volcano plots comparing differentially expressed genes in Mϕs between WT (Chat fl/fl ) and Chat-KO (Chat fl/fl Lyz2 cre ) mice.(D) GSEA of innate immune response (GO:0002218), endocytosis (GO:0030100), apoptotic cell clearance (GO:0043277), and resolution phase Mϕ activity (33).The top portion of the plot shows the running enrichment score as the analysis walks down the ranked list.The bottom portion of the plot shows where the member genes of the gene set appear in the ranked list.The normalized enrichment score (NES) and the nominal P values are shown.(E) Neutrophils were enumerated at different time points after i.p.Pam3 treatment in Chat fl/fl and Chat fl/fl Lyz2 cre mice.T max (time point when neutrophil numbers reach maximum), T 50 (time point corresponding to ~50% neutrophil reduction), and R i (resolution interval, the interval between T max and T 50 ).(F and G) Activity of efferocytosis (F4/80 + Ly6G + cell number; F) and the number of remaining apoptotic neutrophils (Annexin V + Ly6G + ; G) were assessed by flow cytometry at day 3 post-Pam3 stimulation in Chat fl/fl and Chat fl/fl Lyz2 cre mice.(H-J) The neutrophil number (H), efferocytosis activity (I), and the level of remaining apoptotic neutrophils (J) were assessed by flow cytometry at day 3 post-Pam3 stimulation in Chat fl/fl and Chat fl/fl Cd19 cre mice.Error bars indicate mean ± SEM.Statistics: (E) Two-way ANOVA corrected by Šidák's test.(F-J) Student's t test.*P < 0.05, **P < 0.01, and ***P < 0.001.WT, wild type; KO, knockout; i.p., intraperitoneal injection.
Chat fl/fl Chat fl/fl Lyz2 Cre E J Chat fl/fl Chat fl/fl Lyz2 Cre Chat fl/fl Lyz2 Cre + ACh D Chat fl/fl Lyz2 Cre + ACh Time (h)

44 h
Chat fl/fl Chat fl/fl Lyz2 r o p h i l s M a s t c e l l s Chat fl/fl Lyz2 Cre Chat fl/fl

Fig. 5 .
Fig. 5. ACh enhances Mϕ phagocytosis through nicotinic receptors.(A) Pam3 was injected i.p. to Chat fl/fl , Chat fl/fl Lyz2 cre mice, with or without daily supplementation of ACh at a dose of 5 mg/kg.(B).Neutrophils were enumerated at different time points following specified treatments.(C) The number of apoptotic neutrophils was assessed by flow cytometry at day 3 post-Pam3 stimulation in Chat fl/fl and Chat fl/fl Lyz2 cre mice.(D) Wild-type B6 mice were injected i.p. with Pam3, followed by daily administration of muscarinic receptor inhibitor atropine (Atp; 1 mg/kg) or nicotinic receptor inhibitor mecamylamine (Meca; 1 mg/kg).(E) Neutrophils were enumerated at different time points following specified treatments.(F and G) Activity of efferocytosis (F), and the number of remaining apoptotic neutrophils (G) were determined by flow cytometry.(H) The impact of ACh (1 μM) on the uptake of latex beads-rabbit IgG-FITC complex (green) in murine BMDMs (red; Upper) or human MDMs (red; Lower) was assessed.The data are presented as the percentage of latex beads-rabbit IgG-FITC-positive cells relative to the total cell count.(Scale bar, 100 μm.) (I) Dot plots of scaled Chat expression (scRNA-Seq) in indicated cell populations from Chat fl/fl and Chat fl/fl Lyz2 cre peritoneal lavage.(J) BMDMs from ChAT-GFP mice were treated with Pam3, or in combination with ACh (10 μM), for 24 h and subsequently evaluated for GFP expression.Error bars indicate mean ± SEM.Statistics: (B and E) Two-way ANOVA corrected by Šidák's test.(C, F, G, and J) One-way ANOVA corrected by Tukey's test.(H) Student's t test.*P < 0.05, **P < 0.01, and ***P < 0.001.Med, Medium; i.p., intraperitoneal injection.