M1 Cholinergic Signaling Modulates Cytokine Levels and Splenocyte Sub-Phenotypes Following Cecal Ligation and Puncture

Abstract Background: The contribution of the central nervous system to sepsis pathobiology is incompletely understood. In previous studies, administration of endotoxin to mice decreased activity of the vagus anti-inflammatory reflex. Treatment with the centrally-acting M1/M4 muscarinic acetylcholine (ACh) receptor (M1/M4AChR) attenuated this endotoxin-mediated change. We hypothesize that decreased M1/M4AChR-mediated activity contributes to inflammation following cecal ligation and puncture (CLP), a mouse model of sepsis. Methods: Basal forebrain cholinergic activity (immunostaining), serum cytokine/chemokine levels (ELISA) and splenocyte subtypes (flow cytometry) were examined at baseline and following CLP in male C57BL/6 male mice. Rersults: At 48hrs. post-CLP, activity in basal forebrain cells expressing choline acetyltransferase (ChAT) was half of that observed at baseline. Lower activity was also noted in the hippocampus, which contains projections from ChAT-expressing basal forebrain neurons. Serum levels of TNFα, IL-1β, MIP-1α, IL-6, KC and G-CSF were higher post-CLP than at baseline. Post-CLP numbers of splenic macrophages and inflammatory monocytes, TNFa + and ILb + neutrophils and ILb + monocytes were higher than baseline while numbers of central Dendritic Cells (cDCs), CD4 + and CD8 + T cells were lower. When, following CLP, mice were treated with xanomeline, a central-acting M1AChR agonist, activity in basal forebrain ChAT-expressing neurons and in the hippocampus was significantly higher than in untreated animals. Post-CLP serum concentrations of TNFα, IL-1β, and MIP-1α, but not of IL-6, KC and G-CSF, were significantly lower in xanomline-treated mice than in untreated mice. Post-CLP numbers of splenic neutrophils, macrophages, inflammatory monocytes and TNFα + neutrophils also were lower in xanomeline-treated mice than in untreated animals. The effects of CLP on percentages of IL-1β + neutrophils, IL-1β + monocytes, cDCs, CD4 + T cells and CD8 + T cells were similar in xanomeline - treated and untreated post-CLP mice. Conclusion: Our findings indicate that M1/M4AChR-mediated responses modulate CLP-induced alterations in the distribution of some, but not all, leukocyte phenotypes and certain cytokines and chemokines.


Background
Sepsis is de ned as a life-threatening organ dysfunction caused by a dysregulated host response to infection (1).Studies have suggested that central cholinergic (ACh) dysfunction contributes to sepsis pathobiology (2)(3)(4).In particular, recent studies have demonstrated decreased activity in the basal forebrain cholinergic system of mice subjected to cecal ligation and puncture (CLP, the most commonly used animal model of sepsis (2,(5)(6)(7).This system, which is involved in altered executive functions in neurodegenerative disorders (8-10), also contributes to immune dysfunction (4,6,11,12), one of the de ning characteristics of sepsis (1).In particular, Zhai et al used optogenetic stimulation to activate basal forebrain neurons that express choline acetyltransferase (ChAT, the enzyme that catalyzes the ratelimiting step in Ach biosynthesis) at very early time points post-CLP (6).Optogenetic activation increased activity at several loci in the brainstem and decreased serum levels of TNFα and IL-6 via a mechanism that involved the vagus nerve and the spleen.Interestingly, Rosas-Ballina et.al. used xanomeline, a centrally acting M1/M4AChR agonist, to reverse the effects of lipopolysaccharide (LPS) on the relative abundance of several splenocyte subsets and on induced cytokine release by splenic cells (13).In the same study, xanomeline improved survival following CLP (13).Immune abnormalities are a key component of the dysregulated host response that causes organ dysfunction, the de ning characteristic of sepsis (1).These two studies led us to test the hypothesis that altered central M1/M4AChR-mediated activity contributes to CLP-induced effects on the release of a larger set of cytokines and chemokines and alters the relative distribution of splenocyte subtypes.

Methods
Experiments were conducted on 12-16-week-old C57Bl/6 mice (Jackson Labs, Bar Harbor ME).Animals were housed in a veterinarian -supervised facility and were acclimated for a minimum of one week prior to use.All studies were approved by the Feinstein Institute IACUC (2017-013 Term I & II) and conformed to ARRIVE guidelines.

Cecal Ligation and Puncture
CLP was performed under iso urane anesthesia using two 22 -gauge punctures as previously described (14).Animals were resuscitated with a subcutaneous (SQ) injection of 50mL/kg of sterile saline and received 0.5mg/kg of imipenim/cilastatin SQ at the end of surgery and at 24 post-procedure.Blood was obtained via cheek bleed and mice were then euthanized by decapitation.Organs, including brains, were harvested with brain tissues xed in 4% paraformaldehyde.

Xanomeline Administration
Xanomeline (5mg/kg in 0.2 ml saline) was administered intraperitoneally at the time of CLP and at 24and 47 hrs.post-CLP.Mice were euthanized one hour after the nal dose.We elected not to study animals that received vehicle only following CLP because multiple studies have demonstrated that injection of 0.2 ml of saline did not affect post-CLP parameters (7,15).

Measurements of Serum Cytokines
Serum was separated from blood by centrifugation, aliquoted and frozen at -80ºC until used.Levels of TNFα, IL-1β, IL-6, MIP-1α, KC, and G-CSF were determined using a multiplex ELISA (Eve Technologies, Calgary, Alberta, Canada).

Leukocyte Isolation
As previously described ( 16), spleens harvested post-euthanasia were immediately digested with DNAse (100µg/mL) and Collagenase A (1mg/mL) in complete media for 30 minutes at 37°C.Cells were passed through a 70µm lter and resuspended.Red blood cells were lysed, white cells were counted using a Countess II Automated Cell Counter (ThermoFisher, Waltham, MA) and spleen cells were analyzed using ow cytometry.A minimum of 2x10 6 events were analyzed for each sample.

Cytokine Production Assays
Single cell suspensions were stimulated with LPS (500ng/ml) for 3 hours in the presence of Brefeldin A as previously described (17).Concurrently prepared cell suspensions without stimulation served as controls for background production.

Flow Cytometry
Single-cell suspensions were stained for ow cytometric analysis with LIVE/DEAD xable viability dye (Life Technologies) and the following antibodies: CD90.2,CD8a, CD4, Ly6C, CD11c, Ly6G, MHCII, IL1β, and TNFα (antibody details -Suppl.Table 1).All ow cytometric analysis was performed on a BD LSR Fortessa 16-color cell analyzer and analyzed using FlowJo software version 10 (BD Bioscience, San Jose, CA).Gating strategies -Supp.Figure 1 Brain Harvesting, Preparation and Staining Brains were xed with 4% paraformaldehyde for 48 hours, immersed in 30% sucrose, embedded into optimal temperature cutting compound and sliced to yield 10 µmsections; sagittal sections were examined in studies of basal forebrain, coronal sections were used to evaluate staining in the hippocampus.Sections that contained basal forebrain were using the Allen Brain Atlas based on characteristic structure and location.Basal forebrain sections were co-immunostained with antibodies to choline acetytransferase (ChAT, goat anti-ChAT, 1:100, EMD Millipore Corp, Burlington VT),and c-Fos (rabbit polyclonal, 1:500, Cell Signaling, Danvers, MA).Secondary antibodies were donkey anti-goat conjugated to Alexa 488 for ChAT, donkey anti-rabbit conjugated to Alexa 594 for c-Fos, and donkey antimouse Alexa 488 for c-Fos).Immuno uorescence intensity was determined in 10 non-contiguous 40xpowered elds per slide, 1-2 slides/animals.

The fraction of activated cells was determined as
This number was multiplied by 100 to yield a percentage.

Statistics
We used one-way ANOVA corrected with Tukey's multiple comparison test to identify statistical signi cance, with p < 0.05.

Results
Basal forebrain cholinergic activity is lower following CLP than at baseline.Previous studies have demonstrated a link between changes in serum cytokine levels and a loss of basal forebrain cholinergic activity (6, 13).We therefore sought to examine the effects of CLP on the activity of neurons expressing ChAT in this region of the brain.Our studies revealed that, at 48 hrs.post-CLP, basal forebrain co-localization of ChAT and c-Fos, a marker for recent neural depolarization, was signi cantly lower than that observed at baseline (T 0 ) (Fig. 1A,B).At the same post CLP time point, co-localization of ChAT and c-Fos in mice treated with the M1/M4AChR agonist xanomeline was signi cantly higher than that observed in untreated animals, as well as in unoperated controls (Fig. 1A, B).These ndings suggest that the CLP-induced decrease in basal forebrain cholinergic activity is driven, at least in part, by a difference in M1/M4AChR-mediated activity.
Post-CLP activity in the hippocampus is lower than baseline and is altered by the M1/M4AChR agonist xanomeline.
Muscarinic AChRs that are activated by ACh-secreting basal forebrain neurons.These receptors have been identi ed in several different regions of the brain, including but not limited to the cortex, thalamus, hippocampus, neostriatum and basal forebrain (18,19).Xanomeline directly activates M1 and M4AChRs (20).M1 receptors are predominantly post-synaptic and increase neuronal activity while pre-synaptic M4 receptors limit ACh release.Thus, a xanomeline-induced increase in neuronal activation likely results from M1 stimulation while a decrease should re ect M4-mediated effects.We therefore used immunostaining with an antibody to c-Fos to compare activity at T 0 with the effects of CLP and CLP + xanomeline on activity in several brain regions.Baseline expression of c-Fos was most pronounced in the hippocampus, a nding consistent with previous reports (8).At 48hrs post-CLP c-Fos expression in this region was signi cantly lower than at T 0 (Fig. 2A,B).In contrast, c-Fos expression following CLP + xanomeline was signi cantly higher than those observed in untreated mice and could not be distinguished from activity observed at T 0 .Therefore, our ndings suggest that CLP-induced decreases in hippocampal neuronal activity result, in part, from an attenuation of M1AChR stimulation.
Reduced central M1AChR-mediated activity contributes to CLP-induced elevation in serum levels of some, but not all, cytokines.Previous studies demonstrated that serum concentrations of TNFα and IL-6 were increased by LPS administration (13).Levels were reduced in mice pre-treated with the centrally acting M1/M4AChR agonist xanomeline; this effect was not present in animals receiving saline (13).Post-CLP levels of these two cytokines were also higher than at T 0 ; optogenetic stimulation of ChAT-expressing basal forebrain neurons lowered TNFα and IL-6 levels (6).We therefore examined the effects of xanomeline administration on CLP-induced elevations of serum levels of TNFα, IL-6 and several other cytokines and chemokines.Data are presented in Fig. 3.At 48 hrs.post-CLP, serum levels of TNFα, IL-1β, MIP-1α, IL-6, KC and G-CSF were higher than at T 0 (Fig. 3A,B).Following treatment with xanomeline, serum TNFα, IL-1β and MIP-1α levels were lower than those observed following CLP alone and were indistinguishable from levels seen prior to CLP (Fig. 3A).In contrast, levels of IL-6, KC and G-CSF following CLP + xanomeline were not statistically distinguishable from levels observed following CLP alone (Fig. 3B).These ndings indicate that M1AChR-mediated responses modulate serum levels of some, but not all, cytokines.Central M1/M4AChR-mediated responses modulate CLP-induced differences in the distribution of splenocyte subgroups.
Previous work has demonstrated that the effects of LPS/in ammation on the relative proportions and activation states of both innate and adaptive immune cells are in uenced by cholinergic neural input (4, 11-13, 16, 17, 21, 22).To examine the contribution of M1/M4AChR-mediated signaling to CLP-induced differences in immune cell phenotypes we harvested splenic cells at 48 hrs.post CLP from mice treated with either vehicle or xanomeline.Using ow cytometry we examined the abundance of these cells and their ability to mount a cytokine response to ex vivo LPS stimulation.At 48 hrs.post-CLP, the absolute numbers of splenic macrophages (CD11b + /Ly6G − /MHCII − /CD64 + /Ly6C mid ) and in ammatory monocytes (CD11b + /Ly6G − /MHCII − /CD64 − /Ly6C hi ) were both higher than at baseline (Fig. 4A).When post-CLP mice were treated with xanomeline, the numbers of neutrophils (CD11b + /Ly6G + ), macrophages and in ammatory monocytes were lower than following CLP alone and could not be distinguished from numbers observed at baseline (Fig. 4A).Post-CLP differences in the number of central dendritic cells (cDCs, CD11c + /MHCII + ) and of CD4 + (CD90 + /CD4 + ) and CD8 + (CD90 + /CD8 + ) T cells were lower than baseline (Fig. 4B).However, xanomeline did not affect these observed differences, suggesting that the effects of CLP on cDCs and T cell numbers are M1/M4 AChR -independent.These data are consistent with previous work demonstrating that CLP profoundly affected the numbers of splenic T cells (23) and dendritic cells (24).
We also examined how CLP affected ex vivo responses of splenic innate immune cells to stimulation with LPS.At 48hrs post-CLP, the numbers of neutrophils expressing TNFα or IL-1β and the number of monocytes expressing IL-1β was higher than at baseline (Fig. 4C).Following CLP, the percentage of responsive cells was also higher than at T 0 (Fig. 4D).With the addition of xanomeline, the numbers of cytokine expressing cells were signi cantly lower and was indistinguishable from baseline (Fig. 4C).The percentage of neutrophils expressing TNFα followed suit.However, while post-CLP treatment with xanomeline was associated with markedly reduced numbers of IL-1β -expressing neutrophils and monocytes, the percentage of cells expressing this cytokine was unaffected (Fig. 4D).These ndings indicate that M1/M4AChR-mediated responses contribute to the increased number of neutrophils and monocytes and that may also affect TNFα expression by these cells.However, LPS stimulated expression of IL-1β appears to be M1/M4AChR independent.

Discussion
Elevated serum cytokine levels following an in ammatory stimulus in part re ect decreased activity in the vagus-mediated anti-in ammatory re ex (25).Among the affected components of this pathway are cholinergic neurons in basal forebrain and splenocytes that release cytokines (11).The central M1/M4AChR agonist xanomeline reduced cytokine release (13).Studies with CLP, the most commonlyused animal model of sepsis (7), indicated that reduced activity in basal forebrain cholinergic neurons contribute to early in ammation and that restored M1/M4AChR-mediated activity in the brain reduced mortality (6, 13).Relative to measurements at baseline (T 0 ), data from 48 hrs.post CLP demonstrated 1) lower activity in basal forebrain cholinergic neurons and in hippocampal cells, 2) higher serum levels of TNFα, IL-1β, MIP-1α, IL-6, KC and G-CSF, 3) a larger number of splenic macrophages and in ammatory monocytes, but fewer splenic dendritic cells, CD4 + and CD8 + T cells and 4) a higher percentage of neutrophils that responded to ex vivo stimulation by secreting TNFα, and of both neutrophils and monocytes that responded by elaborating IL-1β.When CLP was followed by administration of the M1/M4AChR agonist xanomeline, differences with T 0 in neuronal activity, serum levels of TNFα, IL-1β and MIP-1α, and number of macrophages and in ammatory monocytes were not present.However, concentrations of IL-6, KC and G-CSF and percentages on cytokine-expressing neutrophils and monocytes were higher than baseline and indistinguishable from measurement made in post-CLP mice not receiving the drug.These ndings indicate that decreased M1/M4AChR-mediated activity contributes to some, but not all, of the changes in cytokines levels and leukocyte abundance induced by CLP.
Acute encephalopathy/delirium and cognitive dysfunction in long-term survivors, both well-documented and well-described, are the generally recognized manifestations of sepsis-induced dysfunction in the brain (26-28).However, the contribution of the brain to the pathogenesis of organ dysfunction, the de ning characteristic of sepsis, has been under-appreciated.The ability of complex organisms to respond to constant uctuations in either the internal or external environment depends on communication and coordinated activity between cells and organs that are not in direct physical contact.Some years ago, Godin and Buchman proposed that the dysregulated host response of sepsis compromises these interactions, causing an "uncoupling of biologic oscillators" that reduced biologic variability and ultimately contributed to organ dysfunction (29).Organ-to-organ interactions in humans are mediated by the immune, endocrine and neuronal systems.While sepsis-induced dysfunction in the rst two has been extensively examined (30,31), the brain has been less-well scrutinized.Nerves can disseminate signals far more rapidly than the other two systems and thus may be the " rst responders" to perturbations.Indeed, neuronal activity contributes to both endocrine and immune responses, and these responses are compromised in sepsis (15,31,32).Thus, it is logical to postulate that neuronal dysfunction is a primary determinant of sepsis-induced organ dysfunction.The data presented here support that hypothesis.
The ability of xanomeline to affect some, but not all, of the differences in cytokine expression and leukocyte abundance/activity is important.Previous work has identi ed CLP-induced circulating levels of TNFα, IL-1β and IL-6 that are higher than T 0 levels.Much of those data were generated while exploring the vagus-mediated anti-in ammatory re ex, a neural pathway by which central M1AChR-expressing neurons alter cytokine expression and in ammatory cell abundance in the spleen (6,12,13).The more distal portions of this pathway involve several different neurotransmitters (22,(33)(34)(35).In effect, this pathway has been described as a "brake on in ammation" whose therapeutic potential is currently being explored (36, 37).The ndings reported here suggest that this hypothesis be viewed with a degree of caution.Pretreatment with xanomeline attenuated LPS-induced elevations of TNFα, IL-6, IFNγ and IL12p70 while optogenetic activation of basal forebrain cholinergic neurons has a similar effect on increases in TNFα, IL-6 and IL-10 levels at a very early post-CLP timepoint (6, 13).However, in our studies, which examined mice 48 hrs.post-CLP, levels of IL-6, as well as the chemokine KC and the neutrophil growth factor G-CSF, were not affected by xanomeline.These ndings suggest that some aspects of the in ammatory re ex do not involve an M1/M4-mediated ACh pathway.Perhaps other muscarinic or nicotinic isoforms contribute (38)(39)(40)(41).Alternatively, the CLP-induced difference in IL-6 may be mediated by a noncholinergic neural pathway.We have demonstrated CLP-induced decreases in the orexinergic system of the hypothalamus (15).Glutaminergic and GABAergic systems have been implicated in neuroin ammation (42).Further investigation is clearly indicated.
Previous work also indicated that pre-treatment with xanomeline attenuated LPS induced increases in splenic dendritic cell numbers ( 13), a nding consistent with our observations post-CLP.However, administering the drug prior to LPS challenge increased numbers of splenic T cells and did not affect neutrophil or monocyte/macrophage abundance (13).In contrast, we found that, while the abundance of neutrophils, macrophages, and in ammatory monocytes was higher post-CLP than at baseline, levels following the addition of xanomeline were signi cantly lower than in untreated mice and were indistinguishable from baseline.The differing effects of xanomeline on cytokines may well re ect the timing of administration -treatment before LPS administration but after CLP.However, the discrepancies in numbers of splenic neutrophils, monocytes/macrophages and T cells suggest that mechanisms underlying some of the effects of CLP and LPS differ from each other.Importantly, CLP increased numbers of TNFa-and IL-1b-expressing neutrophils and IL-1b-expressing monocytes, differences that were eliminated when xanomeline was added.However, the percentages of cytokine producing cells was not altered by CLP + xanomeline.Thus, xanomeline did not appear to in uence the effects of CLP on TNFa/IL-1β expression by innate immune cells, but rather simply reduced the total number of cells present.
That CLP and administration of LPS induce distinct cytokine and immune phenotypes also has implications that may affect our understanding of the pathobiology of sepsis.The data presented here and results generated by other investigators suggest that both CLP and human sepsis evoke unique immune pro les.Some of these changes can be considered "pro-in ammatory" (43), others "antiin ammatory" (30,44) but the overall response de es easy classi cation (45).Indeed, the work here demonstrated that CLP-induced changes in serum levels of two quintessential "pro-in ammatory" cytokines -TNFα and IL-6 -respond differently to M1/M4AChR stimulation.Further, while elevated levels of splenic macrophages and monocytes point to an enhanced immune response, decreased numbers of dendritic cells and of CD4 + and CD8 + T cells suggest immune suppression.These discrepant ndings suggest that CLP, and perhaps sepsis, involve complex and likely unique mechanisms that are not easily characterized.
This study has limitations and is subject to certain caveats.We examined responses only at 48hrs.post-CLP.While previous work suggests that this approach captures most post-CLP pathobiology (14,15), the dynamic nature of both CLP and sepsis mandates investigation of other time points.
Perhaps the most important limitation lies in attempts to model sepsis, a distinctly human disorder, using animals.While it is the most commonly used animal model of sepsis, the de ciencies of CLP are welldocumented and indisputable.The concerns include differences in gene expression and in organ-speci c pathophysiologic responses to in ammation (46).Importantly, lab mice are immunologically naïve while humans are exposed to an antigenically rich environment from an early age on.These discrepancies may have limited translation of approaches to treatment; therapy that has worked in post-CLP mice has not improved outcome in human sepsis.However, when the focus is directed towards organ dysfunction, the de ning clinical characteristic of human sepsis (1), some of the distinctions blur.In ammation in mice causes hypothermia, bradycardia and hypopnea; in humans the response is characterized by fever, tachycardia and tachypnea.But both mice and men respond to in ammation with alterations in metabolism, cardiac function and respiration.Recent modi cations have improved the delity of the model to the disorder (47); these include our demonstration that induction of a broad memory T cell repertoire in mice prior to CLP increased organ dysfunction and reduced survival to levels closer to those observed in human sepsis (43).Further research is required to identify pathobiology and to enhance clinical relevance.

Conclusions
In summary, our ndings indicate that, at 48hrs.post-CLP 1. activity in hippocampal neurons and in basal forebrain neurons expressing ChAT is lower than at baseline.Lower hippocampal activity re ected, at least in part, reduced activity in neurons that respond to M1/M4 AChR-mediated stimulation.2. circulating levels of TNFα, IL-1β, MIP-1α, IL-6, KC and G-CSF were higher than at baseline.The levels of TNFα, IL-1β and MIP-1α were driven, in part, by lower levels of M1/M4AChR stimulation.In contrast, levels of IL-6, KC and G-CSF were not affected by differences in M1/M4-mediated activity.
3. the numbers of splenic macrophages and in ammatory monocytes were higher than at baseline and a greater percentage of neutrophils and monocytes responded to an ex vivo in ammatory stimulus.This increase resulted, in part, from lower levels of M1/M4 AChR-mediated stimulation.4. a greater fraction of splenic neutrophils and monocytes responded to ex vivo stimulation by elaborating cytokines.This difference was not driven by lower activation of M1/M4AChRs.
5. the spleen contained fewer central dendritic cells, and fewer CD4 + and CD8 + T cells.This difference did not re ect altered M1/M4AChR-mediated activity.
This study re-a rms that altered activity in the brain contributes to CLP-mediated immune responses.Our ndings support the hypothesis that altered activity in the brain constitutes one component of the "dysregulated host response" that is a de ning characteristic of sepsis.

Figures
Figures

Figure 1 Effects
Figure 1

Figure 2 Effects
Figure 2