The gut-brain and gut-macrophage contribution to gastrointestinal dysfunction with systemic inflammation

The gastrointestinal tract is one of the main organs affected during systemic inflammation and disrupted gastrointestinal motility is a major clinical manifestation. Many studies have investigated the involvement of neuroimmune interactions in regulating colonic motility during localized colonic inflammation, i.e., colitis. However, little is known about how the enteric nervous system and intestinal macrophages contribute to dysregulated motility during systemic inflammation. Given that systemic inflammation commonly results from the innate immune response against bacterial infection, we mimicked bacterial infection by administering lipopolysaccharide (LPS) to rats and assessed colonic motility using ex vivo video imaging techniques. We utilized the Cx3cr1-Dtr rat model of transient depletion of macrophages to investigate the role of intestinal macrophages in regulating colonic motility during LPS infection. To investigate the role of inhibitory enteric neurotransmission on colonic motility following LPS, we applied the nitric oxide synthase inhibitor, N ω -nitro-L-arginine (NOLA). Our results confirmed an increase in colonic contraction frequency during LPS-induced systemic inflammation. However, neither the depletion of intestinal macrophages, nor the suppression of inhibitory enteric nervous system activity impacted colonic motility disruption during inflammation. This implies that the interplay be-tween the enteric nervous system and intestinal macrophages is nuanced, and complex, and further investigation is needed to clarify their joint roles in colonic motility.


Introduction
Gastrointestinal (GI) disorders like inflammatory bowel diseases, functional dyspepsia, and irritable bowel syndrome, arise from systemic inflammation due to the innate immune system failing to adequately combat invading pathogens (Corridoni et al., 2018;Burns et al., 2019).Systemic inflammation typically involves a proinflammatory cytokine spike, an increased number of immune cells in organs such as the liver and lungs (Bian et al., 2012;Vatankhah et al., 2015), as well as a rise in core body temperature (Mouihate et al., 2008).In addition to changes in inflammatory pathways, systemic inflammation can lead to myriad gut disturbances due to intestinal barrier disruption and increased GI permeability (Salim & Söderholm, 2011;Ramos & Papadakis, 2019).It is important to note that GI disorders pertaining to dysmotility account for more than 40 % of all GI disorders in the USA (Chen et al., 2017).It has also been extensively demonstrated that localized GI inflammation, such as in multiple preclinical models of colitis, disrupts GI motility (Ren et al., 2011;Bassotti et al., 2014;Mawe, 2015b;Chen et al., 2021).However, how broad innate immune activation stemming from a systemic infection affects GI motility is largely unknown.
Lipopolysaccharide (LPS) is derived from the cell wall of Gramnegative bacteria and is a useful tool to elicit broad innate immune activation in preclinical models.LPS is sensed by the gut wall mainly via toll-like receptors (TLR)s expressed by intestinal macrophages and enterochromaffin cells (Dheer et al., 2016;Ozge et al., 2021;Zheng et al., 2022).LPS has been used as an agent in both in vitro and in vivo studies to mimic different inflammatory disease conditions including brain inflammation (Wang et al., 2006;Zhao et al., 2019) and GI immune dynamics (Creely et al., 2007;Steimle et al., 2019;Ozge et al., 2021).Intraperitoneal administration of LPS typically leads to systemic inflammation in rodents as evidenced by elevated inflammatory cytokine levels in the blood (Bossù et al., 2012;Malik et al., 2023).However, it is still unknown if and how intestinal macrophages are involved in regulating colonic motility during LPS-induced bacterial infection.
Intestinal macrophages are a major immune cell population within the GI tract.This population comprises a diverse array of subtypes of macrophages located in various layers of the gut, each with different origins and functions (Margolis & Gershon, 2016;Margolis et al., 2016;Bain & Schridde, 2018).Intestinal macrophages are predominantly derived from circulating blood monocytes, via a monocyte chemoattractant protein-1(MCP-1/CCL2)-dependent pathway (Bain & Schridde, 2018).This subtype of macrophage expresses Ly6C + MHCII + CX3CR1 + surface markers.Although these macrophages reside in all layers of the gut, they are mainly located in the lamina propria, where the continuous turnover of macrophages in the interface is replenished by blood monocytes circulating between the gut lumen and the epithelium, as reviewed in (Yip et al., 2021).The main function of this macrophage subtype is to combat pathogens encountered at the gut lumen.In addition to intestinal macrophages, tissue-resident macrophages that express the surface markers Tim-4 + and CD4 + are derived from embryonic precursors (Shaw et al., 2018).These tissue-resident macrophages are located mainly within the muscularis externa, within the submucosal plexus and myenteric plexus.It has been reported that such tissue-resident macrophages are indispensable in regulating intestinal permeability and motility due to their proximity to enteric neurons (Phillips & Powley, 2012).We previously demonstrated that cluster of differentiation 163 (CD163) tissue-resident macrophages inhibit colonic motility in the rat, and that this action does not rely on the inhibitory neuronal input from the enteric nervous system to the colon in the absence of an inflammatory challenge (Yip et al., 2023).Other groups have shown that intestinal commensal bacteria help maintain enteric neuronal populations (including nitrergic neurons) by promoting colonic neurogenesis via TLR2 signaling (Yarandi et al., 2020).Thus, we hypothesized that both GI macrophages and the enteric nervous system neuronal populations interact to influence colonic motility during immune challenge.We aimed to test the relative contributions of these inputs in a model of intestinal macrophage depletion and by inhibiting enteric input with the nitric oxide synthase (NOS) inhibitor, Nω-nitro-Larginine (NOLA).The Cx3cr1-Dtr conditional rat model has previously been demonstrated to temporarily deplete intestinal macrophages upon administration of diphtheria toxin (DT) (Yip et al., 2023).In addition, microglial depletion with this model disrupts circadian rhythms and leads to reversible anorexia and weight loss (De Luca et al., 2019;Sominsky et al., 2021).Here, we utilized the Cx3cr1-Dtr rat model to examine the role of intestinal macrophages in regulating colonic motility.Specifically, we assessed the influence of these cell types on motility during LPS-induced systemic inflammation in the presence of NOLA.

Animals
All experiments were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes, with approval from the RMIT University Animal Ethics Committee (AEC project # 22448).As described previously, we generated a Cx3cr1-Dtr knock-in rat model on a Wistar Han background using CRISPR/Cas9 technology to conditionally deplete Cx3cr1-expressing cells including microglia and intestinal macrophages (De Luca et al., 2019;Yip et al., 2023).
In this study, we used female rats aged between 13 and 17 weeks.A previous study from our group suggested that females perform similarly to males in terms of microglial and weight responses to the DT (De Luca et al., 2019) so we elected to use females only for the current study to first establish mechanistic insight into how macrophages affect gut motility during systemic inflammation before proceeding to sexcomparison studies.Males will be studied in future investigations.Rats were kept under standard laboratory housing conditions, with a 12 h light cycle (7 am to 7 pm), an ambient temperature of 22 • C, with humidity between 40 and 60 %, and free access to water and standard rat chow.

Intestinal macrophage ablation and LPS treatment
We administered DT as two separate injections, 8 h apart, of 25 ng/g in sterile saline, subcutaneously (s.c.), according to our previous studies (De Luca et al., 2019;Yip et al., 2023).Our previous work has shown that depletion of microglia and monocytes is maximized at 48 h after DT (De Luca et al., 2019).Control rats were administered sterile saline (vehicle), it having been established in our previous work that the Dtr transgene has no measurable effect on rats without DT (De Luca et al., 2019).Thus, the rats were first injected with DT, followed at 48 h by 250 µg/kg LPS injection intraperitoneally (i.p.; E. coli., serotype 0.127: B8; L-3129; Sigma-Aldrich, St-Louis, MO, USA) to induce systemic inflammation (Fig. 1).Previous studies indicate that an LPS dose of 100-250 µg/kg is sufficient to induce an inflammatory response including fever and an increase in proinflammatory cytokines over a time course of 0-24 h (Kelly et al., 2018;Malik et al., 2023).Tissue collection was therefore performed 12 h after LPS injection, after the rats were deeply anesthetized with 20 mg/mL ketamine (Cenvet Australia, Lynbrook, VIC, Australia) and 5 mg/mL xylazine (Cenvet).All tissue collection experiments were completed between 9 am and 1 pm to avoid circadian rhythms affecting parameters measured.

Colon collection and wholemount tissue preparation for immunofluorescence
The colon dissection procedure for wholemount immunofluorescence labelling has been described previously (Yip et al., 2022).Briefly, the proximal colon from each animal was opened, stretched, pinned with the mucosa facing upwards, and submerged in 0.1 M phosphatebuffered saline (PBS) on a Petri dish lined with Sylgard (Sylgard Silicone Elastomer, Krayden Inc., Denver, CO, USA).The mucosa, submucosal plexus and circular muscle were then peeled away from the remaining colonic tissue under a dissecting microscope.To measure the distribution of intestinal macrophages across the tissue layers of the gut, an additional intact piece of proximal colon was dissected and processed for immunofluorescence.

Wholemount immunofluorescence for neuronal populations and identification of intestinal macrophages
The myenteric plexus wholemount immunofluorescence protocol was employed as previously described for mouse tissues (Leembruggen et al., 2020).Here, immunofluorescence was performed on wholemount rat colonic tissue samples to assess for potential differences in intestinal macrophage populations and distribution after systemic immune challenge with LPS.Wholemount colonic tissue consisting of longitudinal muscle and myenteric plexus (LMMP) or intact colon were incubated at room temperature (RT) for 30 min in 0.01 % Triton X-100 (Sigma Aldrich, St Louis, MO, USA) with 10 % CAS-block TM (Invitrogen Australia, Mt-Waverley, VIC, Australia) to reduce non-specific binding of the antibodies.Then, tissues were incubated with two primary antisera: rabbit anti-ionized calcium binding adaptor molecule 1 (Iba1; FUJIFILM Wako Shibayagi, RRID: AB_839504; 1:400) and mouse antimajor histocompatibility (MHC) class II (Abcam, RRIB: AB_447796; 1:500) or mouse anti-cluster of differentiation 163 (CD163; Bio-Rad Laboratories, RRID: AB_2074558; 1:100) overnight, RT.After incubation, colonic tissues were washed with 0.1 M PBS (three washes of 10 min each).Secondary antisera corresponding to the host of primary antibodies (donkey anti-rabbit Alexa 647 (Jackson ImmunoResearch Laboratories, Inc., RRID: AB_2340572; 1:400) and donkey anti-mouse Alexa 488 (Abcam, RRID: AB_2732856; 1:400)) were applied to the samples and left for 2.5 h at RT on a shaker incubator.Colonic tissues were then mounted using fluorescence mounting medium (DAKO Australia Pty Ltd; Botany, NSW, Australia).Tissue samples were imaged using a confocal microscope (Nikon Confocal Microscope: A1; Version 4.10).A Z-series of images of LMMP sections (6.5 μm/step with total tissue thickness of approximately 60 μm) or intact colon sections (3 μm/ step with total tissue thickness of approximately 100 μm) was captured and saved in the ND2 file format.

Intestinal macrophage density and morphology
Z-series images of wholemount tissue were analysed using the Imaris software volume function (Imaris 64X 9.1.0;Bitplane AG, UK) to assess the cell density and morphology of intestinal macrophages.Z-series images of intact colon tissue were analysed using the Imaris software vantage function to examine the distribution of macrophages across intestinal layers by obtaining the Z-position of each cell.Three proximal colon areas of 0.25 mm 2 per sample were selected as regions of interest.Sphericity and cell density data for Iba1 and CD163 labelled cells were exported and analysed using GraphPad Prism software (version 9.0.1).

Ex vivo video imaging of colonic motility
We previously described the rat colonic motility protocol (Yip et al., 2022) as modified from mouse (Swaminathan et al., 2016).Briefly, the proximal to mid colon (i.e., a colon segment of 5-7 cm in length, measured from the caecum (oral) end) was placed into an organ bath chamber, which was connected to an in-flow reservoir containing Krebs solution via inlet tubes and was continuously superfused with Krebs solution bubbled with carbogen and maintained between 33-35 • C. The oral and anal end of the colon preparation were cannulated to the inlet tube and outlet tube, respectively, and secured using standard cotton sewing thread.Intraluminal pressure was maintained at 5.5-6.5 cm above the height of the colon segment as this pressure range was previously ascertained to produce consistent data in rat colonic segments being assessed for motility patterns (Yip et al., 2022).Video recordings of spontaneous colonic motility were then acquired using a Logitech camera (QuickCam Pro 4000; I-Tech, Ultimo, NSW, Australia) mounted directly above the organ bath at a standard distance of 10 cm.Each colon was given 30 min to equilibrate before we recorded four 15-min duration videos of spontaneous contractile activity under basal conditions.Subsequently, 100 mM NOLA (Sigma Aldrich) was added to the inflow reservoir.Colonic contractile activity was then recorded for an additional four 15-min videos.Following the NOLA application recordings, a final four 15-min videos were recorded and considered as the washout period.These final recordings also enabled the effect of removing the NOLA-mediated inhibition to be assessed and to ensure tissue remained viable throughout the full duration of the experiment.Swaminathan et al. (Swaminathan et al., 2016) provides a detailed explanation of heatmap generation, including step-by-step images for clarity.Briefly, our in-house MATLAB based software interface utilizes an edge detection algorithm to analyse the upper and lower borders of the gut segment being assessed for motility patterns (i.e., this approach enables the edges of the ex vivo tissue segment to be distinguished from the black base of the organ bath) from video recordings (50 frames per second, for a duration of 15 min) of spontaneous gut contractile activity.This allows changes in gut diameter to be estimated within the software.The software additionally assigns cooler hues to dilated regions and warmer hues to constricted regions of colon tissue samples to form pseudo-colored spatiotemporal maps.The resulting spatiotemporal maps depict the diameter of the colon at a point in time along its length.

Statistical analysis
Statistical analyses were performed using GraphPad Prism software (version 9.0.1).Student's unpaired t-tests were used to compare the following anatomical and cellular parameters for saline-and LPS-treated rats: body weight, colon length, immune cell density, Z-position and sphericity.Functional parameters (i.e., the number of proximal colon contractions, contraction magnitude and resting gut diameter) were assessed per 15 min and were also analysed using Student's unpaired ttests.Statistical significance was assumed when p < 0.05.A repeated measures three-way analysis of variance (ANOVA) was used to compare the effect of LPS, macrophage depletion and NOLA on resting gut diameter, contraction magnitude and frequency in an additional cohort of animals.A repeated measures two-way analysis of variance (ANOVA) was used to compare resting gut diameter, contraction magnitude and contraction frequency of colons from saline-and LPS-injected rats under control conditions and with NOLA treatment.Tukey post hoc tests were used to identify where significant differences occurred in the case of a significant interaction.Data are presented as mean with maximum and minimum values.Sample sizes are included in the results section.

Systemic inflammation leads to body weight loss but does not affect colon length
As expected, 250 µg/kg LPS administered to rats reduced body weight by 12 h relative to saline-treated rats (t (13) = 7.84, p < 0.0001, n = 8, Fig. 2A).There was no change in colon length (Fig. 2B).

Systemic inflammation constricts resting gut diameter and increases colonic motility
Although there was no change in colon length, LPS significantly reduced resting proximal colon diameter (t (7) = 3.11, p = 0.02, Fig. 3A,  B).There were also more contractions in LPS-treated rat proximal colons than in saline controls (t (7) = 4.97, p = 0.002, Fig. 3A, C), albeit no change in contraction magnitude over the 15 min assessment period (Fig. 3D).Together these data indicate that LPS increases colonic motility, which may influence colonic transit time (Liang et al., 2005).

Colonic dysmotility during systemic inflammation
Since systemic inflammation due to LPS disrupted colonic motility, we next aimed to test whether this involved enteric neural regulation.In these experiments, we used NOLA to deplete nitric oxide (NO), the major inhibitory neurotransmitter in the colon.As expected, both control and LPS-treated colons had reduced resting gut diameter, increased contraction frequency, and reduced contraction magnitude when inhibitory input was inhibited (main effect of NOLA on resting gut diameter: F (1,13) = 15.86,p = 0.002, Fig. 3E; contraction frequency: F (1,13) = 437.2,p < 0.0001, Fig. 3F; contraction magnitude: F (1,13) = 60.96,p < 0.0001, Fig. 3G).However, there was no interaction between NOLA and LPS on the number of contractions or contraction magnitude, suggesting that LPS did not substantially change gut motility when NOdependent inhibitory activity is reduced.Notably, LPS-treated colons had a greater decrease in resting gut diameter in the presence of NOLA than controls (LPS by NOLA interaction: F (1,13) = 8.72, p = 0.011, Fig. 3E).

Systemic inflammation reduces macrophage numbers in the myenteric plexus
Our previous study suggested that CD163 tissue-resident macrophages play a crucial role in regulating colonic motility under unstimulated conditions (Yip et al., 2023).We therefore hypothesized that changes to these macrophages in response to LPS contribute to the observed LPS-induced changes in colonic motility.We first assessed whether LPS affected the density or morphology of immune cells within the myenteric plexus of the colon.We used Iba1 as a pan-macrophage marker and CD163 to label tissue-resident macrophages.There was no significant effect of LPS on intestinal macrophage density, either in terms of all macrophages or of CD163-labelled tissue-resident macrophages (Fig. 4A-D).However, the number of cells positive for Iba1 only (i.e., Iba1-positive but CD163-negative) was reduced by LPS.This finding suggests that LPS causes a reduction in macrophage infiltration or an increase in the turnover of infiltrating macrophages (t (8) = 3.03, p = 0.02, Fig. 4F).We also found that the sphericity of Iba1-only cells was lower in the colon of LPS-treated rats than in saline-treated.The presence of a more elongated morphology suggests that LPS resulted in stimulation of these macrophages (t (7) = 2.67, p = 0.03, Fig. 4I).

Systemic inflammation causes displacement of immune cells
The reduced number and lower sphericity of colonic macrophages in the proximal colon following LPS is in contrast with previous literature reporting an increase in immune cell populations in the kidney after i.p.LPS (Patel et al., 2019).In that study, Patel and colleagues suggested that during an immune challenge, tissue-specific infiltration of immune cells occurs to resolve the inflammation.We therefore hypothesized that intestinal macrophages in the myenteric plexus could be recruited to the epithelial layer of the gut to combat pathogens in the lumen.This could account for why we observed fewer intestinal macrophages labelled for Iba1-only following LPS.To test this idea, we performed immunohistochemistry on intact colonic tissue instead of the LMMP alone and examined the Z-position of intestinal macrophages corresponding to different structural layers in the GI wall (i.e., from the lamina propria, circular muscle, to the longitudinal muscle).Because it has been established that MHCII-positive macrophages are the predominant subtype recruited to inflamed sites to resolve inflammation, we immunolabelled these preparations for MHCII (Stables et al., 2011;Louwe et al., 2021).
In general, MHCII-expressing cells were localized in the circular muscle layer (see data from representative animals in Fig. 5A and B).In contrast, Iba1-positive only cells were mostly located within the myenteric plexus.The relative Z-position between MHCII-positive and Iba1-positive macrophages changed significantly in response to LPS treatment.Following LPS, the localisation of MHCII-positive and Iba1positive macrophages was restricted and the distribution profiles of the cell types showed increased overlap within the same layer of the colon (t (12) = 2.74, p = 0.02, n = 7, Fig. 5C).These data suggest that intestinal macrophage subtypes are displaced under LPS-induced inflammatory conditions.
Colonic motility was increased in response to LPS, but this increase did not appear to be dependent upon inhibitory myenteric nervous system input (i.e., the application of NOLA had no additional effects).However, in the presence of LPS, we observed fewer infiltrating macrophages within the myenteric plexus and a shift in the localization of MHCII and Iba1 cells within the gut wall.Thus, we next examined if the intestinal macrophages were essential for regulating colonic motility during systemic inflammation.To test this, we depleted intestinal macrophages in the Cx3cr1-Dtr rat model and assessed the rats for changes in motility patterns.

Macrophage depletion attenuates weight loss induced by LPS
As expected (Malik et al., 2023), macrophage depletion led to acute weight loss (DT by LPS interaction: F (1,25) = 7.14, p = 0.013, Fig. 6A) but this was not exacerbated by LPS.Similar to our findings in response to LPS alone (Fig. 1B), macrophage depletion had no additional effect on colon length.

Colonic dysmotility during systemic inflammation and the role of intestinal macrophages
In a second cohort, LPS reduced resting colonic diameter in rats, however this effect persisted when intestinal macrophages were removed (main effect of LPS on resting gut diameter: F (1,27) = 8.77, p = 0.0063, Fig. 7A).Likewise, the number of colonic contractions per 15 min increased with LPS treatment and this effect remained unchanged when macrophages were depleted (main effect of LPS on contraction frequency: F (1,27) = 38.8,p < 0.0001, Fig. 7B).In this experiment, we identified that LPS caused a significant reduction in contraction magnitude (main effect of LPS on contraction magnitude: F (1,27) = 10.41,p = 0.003, Fig. 7C), however, macrophage depletion had no effect on this parameter.
We next tested whether the inhibitory neuronal input interacts with the immune milieu in the gut during inflammation to contribute to these changes.Here we used the Cx3cr1-Dtr rat model to deplete intestinal macrophages and assessed for changes in gut contractile parameters.NOLA reduced gut diameter irrespective of macrophage depletion, there being statistically significant main effects of LPS (F (1,27) = 7.93, p = 0.009) and NOLA (F (1,27) = 19.22,p = 0.0002), but no significant interaction between the treatments (Fig. 8B), supporting our earlier finding that there is not an additive effect of LPS and NOLA on gut diameter changes.With a depletion of macrophages and the neural inhibitory input, the contraction frequency was enhanced, indicating that nitric oxide is important for constraining contractility.This high frequency of contractility in the absence of macrophages and inhibitory neuronal input was curtailed by LPS.This finding suggests that additional inputs (i.e., additional to macrophages and the inhibitory neuronal inputs) regulate motility under inflammatory conditions.As such, there was a three-way interaction between LPS, macrophage depletion and NOLA on contraction frequency (F (1,27) = 7.693, p = 0.01, Fig. 8C).
Although both neural and immune inputs contributed to a decrease in contraction magnitude, neither appeared to be crucial for the effect of LPS-induced systemic inflammation on this parameter.Removal of macrophages and reducing inhibitory neuronal input did not rescue the LPS-induced reduction contraction magnitude.There was a significant interaction between the effects of macrophage depletion and NOLA (F (1,27) = 9.47, p = 0.0048, Fig. 8D) and a significant main effect of LPS (F (1,27) = 9.38, p = 0.005) but no post hoc differences.These findings suggest that neither intestinal macrophages nor NO-induced neurotransmission are essential for the LPS-induced reduction in contraction magnitude.

Discussion
Our study demonstrates that systemic inflammation induced by LPS disrupts ex vivo colonic motility.This complements previous studies that showed that LPS reduces smooth muscle contractility in duodenum and ileum in rabbits and rats, respectively (Gonzalo et al., 2011;Mikawa et al., 2015).Although there was no obvious infiltration of immune cells across gut layers, some Iba1 and macrophages were displaced from the level of the myenteric plexus towards the epithelial layer in response to LPS.We used the macrophage-depleted Cx3cr1-Dtr rat model in combination with the NOS inhibitor, NOLA, to examine the involvement of intestinal macrophages and inhibitory enteric input.Our results revealed that the inflammation-associated dysregulation in motility still takes place even when both the intestinal macrophages and the inhibitory neural input from the enteric nervous system are disrupted.We previously demonstrated a novel role for intestinal macrophages in regulating colonic motility when inhibitory input from the enteric nervous system is reduced (Yip et al., 2023).Here we provide evidence that mechanisms regulating motility may differ under conditions of inflammation.
Nitric oxide is the major inhibitory neurotransmitter in the myenteric plexus and is indispensable in regulating muscle tone in intestinal motility (Barthó & Lefebvre, 1995).Conventionally, GI motility is thought to be regulated through excitatory (cholinergic) neurotransmission (Parkman et al., 1999;Law et al., 2001;Spencer, 2001).However, a recent study suggested that colonic migrating motor complexes (CMMCs) originate from the blockade of the inhibitory nitrergic cyclic  guanosine monophosphate (cGMP)-dependent pathway (Koh et al., 2022).Based on this proposed mechanism it been suggested that the nitrergic pathway is a key regulator of colonic motility (Koh et al., 2022).It is still unknown whether infiltrating inflammatory macrophages contribute to dysregulated intestinal motility during systemic inflammation.
Previous preclinical investigations of intestinal inflammation, such as studies using dextran sodium sulphate (DSS) or trinitrobenzene sulfonic acid (TNBS)-induced colitis models provide a general consensus that GI motility is supressed during localized inflammation (Cheon et al., 2012;Mawe, 2015a).This local modulation of motility is mainly mediated through the T-helper cell 1 (Th1) and Th2-induced cytokine signalling cascades (Akiho et al., 2011).In particular, it is reported that Th1-induced cytokines (such as interleukin (IL)-1β, interferon (IFN)-γ, and tumor necrosis factor (TNF)) reduce smooth muscle contraction via the nuclear factor κB (NF-κB)-dependent downregulation of L-type Ca 2+ channels (Ihara et al., 2009) in response to stimulation of TLR4 (Mukherjee et al., 2009).Systemic effects of LPS treatment can also cause Th1-mediated cytokine production that spikes during the acute response to LPS (Malik et al., 2023).It is notable, then, that we observed the opposite effect, where a more global systemic inflammation induced via i.p. administration of LPS led to increased (not decreased) colonic motility in rats.These findings suggest that the cytokine response would not be sufficient to explain the motility changes we see with LPS in our rats and that additional mechanisms are at play.TLR4 plays a critical role in mediating LPS-induced inflammation (Kuzmich et al., 2017).It is worth noting that not only is TLR4 expressed on intestinal macrophages, but it is also expressed on different cellular subtypes within the enteric nervous system (Caputi et al., 2017).It has been reported that in the absence of TLR4 (in TLR4-/-mice), the morphology of the glial cell population is significantly changed, with the processes becoming substantially more ramified and with increased S100 calcium-binding protein B (S100b) expression (Caputi et al., 2017).The number of neuronal NOS expressing neurons is also decreased in TLR4-/-mice, suggesting that TLR4 is indispensable in maintaining healthy enteric nervous system architecture (Caputi et al., 2017).More importantly, GI motility is delayed in TLR4-/-mice.This delay is thought to be due to reduced carbachol-induced contractility and a redistribution of purinergic receptors (Caputi et al., 2017).Another study also reported delayed GI motility in TLR4-/-mice via a mechanism involving the adaptor molecule, MyD88, a crucial component in the TLR4 signalling pathway (Anitha et al., 2012).Interactions between microbial products and TLR4 have also been shown to regulate GI motility.For example, antibiotic treatment specifically affects in vivo GI transit in WT mice, but not in TLR4-/-mice (Anitha et al., 2012), suggesting TLR4 is crucial for motility regulation under at least some conditions.Taken together, these reports indicate that TLR4 signalling (in cells other than intestinal macrophages and NO-dependent inhibitory enteric nervous system neurons) could be a potential contributing factor in regulating GI motility during systemic inflammation.In fact, previous reports have demonstrated that both enteric glial cells and mast cells express TLR4 (Barajon et al., 2009;Yang et al., 2019).
In addition to enteric neurons, enteric glial cells have also been shown to regulate GI motility (Grubišić et al., 2018;Wang et al., 2022).Eliminating enteric glial cells specifically leads to disrupted GI motility in mice, although this occurs in a sex and age-dependent manner.Studies have shown that enteric glial cells are crucial in regulating GI motility in females and in aged mice (Rao et al., 2017;Woods et al., 2022).Glial disruption using transgenic mice with glial-specific expression of influenza virus haemagglutinin (HA) increased numbers of choline acetyltransferase (ChAT) expressing neurons and reduced NOS expressing myenteric neurons, indicating that enteric glial cells can regulate neurochemical coding in the intestine (Aubé et al., 2006).Enteric glial cells are also engaged in crosstalk with enteric neurons via purinergic and cholinergic pathways to regulate intestinal motor circuitry (Ahmadzai and Gulbransen, 2021).Following inflammation, it has been illustrated that enteric glial cells can modulate a phenotypic switch of intestinal macrophages to protect against visceral hypersensitivity via activity mediated by the glial connexin-43 and macrophage colony-stimulating factor (Grubišić et al., 2020).Therefore, enteric glial cells provide a possible candidate as a regulator of colonic motility in systemic inflammation.Further investigation of the interplay between enteric glial cells, enteric neurons and intestinal macrophages may provide us with a better understanding of how motility is regulated under inflammation.
Although infiltration of immune cells is a hallmark of tissue inflammation (Arango Duque & Descoteaux, 2014), we did not observe such a phenomenon in LPS-treated rats in the current study.This finding is in agreement with a previous study that used intracolonic administration of LPS but did not observe mucosal damage or infiltration of immune cells in the target tissue (Im et al., 2012).It is worth noting however, that Im and colleagues did report an increase in inflammatory cytokine levels and intestinal fluid secretion (Im et al., 2012).This work suggests other immune cells in the GI tract apart from intestinal macrophages could play a part in regulating cytokine levels or motility during inflammation.We note that mast cells exert a critical role in response to intestinal inflammation (De Winter et al., 2012).For example, infiltration of mast cells has been implicated in inflammatory GI disorders such as inflammatory bowel disease and achalasia (Hamilton et al., 2014;Nelson et al., 2020).IL-4 and IL-13, two of the major cytokines secreted by mast cells, are also elevated in an experimental model of colitis (West et al., 1996;Fuss & Strober, 2008;McLeod et al., 2015).Additionally, mast cells activate afferent nerves to release neurotransmitters while these neurotransmitters in turn enhance mast cell activity (Sharkey & Kroese, 2001;De Winter et al., 2012).These findings suggest that mast cells are involved in bidirectional communication with neurons in the GI tract.Therefore, mast cells could also contribute to regulating colonic motility under LPS-induced inflammation,  This study provides important insights into how colonic motility is regulated in the absence of an intact inhibitory enteric nervous system input and intestinal macrophages during systemic inflammation.Limitations of the study include the broad impact of NOLA that will inhibit the activity of a range of neurons in the enteric nervous system due to, for example, postsynaptic influences resulting from the dampening of activity of NOS expressing neurons, rather than a specific subset of neurons.In addition, the efficacy of NOLA to completely eradicate any impact of NO in our experimental system is not known.It is also important to acknowledge the sheer complexity of the response to inflammation in the GI tract whereby the expression of a multitude of genes is altered and a range of implications are present for gut function and the general health of the host organism.We also note that the current findings are based on ex vivo studies alone and that the response of the entire organism in vivo is likely to be far more complex and may yield further insights as to potential interacting mechanisms influencing GI motility in the presence of inflammatory stimuli.We therefore emphasise that future studies should include an in vivo component to further assess the role of enteric inhibitory neuronal input during systemic inflammation.Our data nonetheless show that the LPS-stimulated changes in GI motility are not reversed by our treatment and suggest that while the enteric nervous system and intestinal macrophages are prominent regulators of GI motility and the immune response, these two components are not the major contributors of dysmotility during systemic inflammation induced by LPS.Future investigation into the underlying causes of motility disorders in inflammatory conditions should focus on the interplay between enteric glial cells, neurons and mast cells as well as macrophage-glial interactions.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 2 .
Fig. 2. LPS reduces body weight but does not impact colon length.At 12 h following lipopolysaccharide (LPS) injection A) body weights for LPS-treated rats were reduced relative to saline-treated rats.B) There was no difference in colon length.Data are displayed as mean with maximum and minimum values.Student's unpaired t-tests: **** p ≤ 0.0001; n = 5-8 per group.

Fig. 3 .
Fig. 3. Lipopolysaccharide (LPS) increases colonic motility relative to saline-treated rats but this response persists when inhibitory neural input is dampened.A) Representative heatmaps for LPS and saline treated rats under basal and Nω-nitro-L-arginine (NOLA) treatment.LPS reduces resting proximal colon diameter and increases colonic motility.B) Resting diameter of the proximal colon in preparations from saline-treated controls and LPS-treated rats.C) Contraction frequency per 15 min in the proximal colon.D) Contraction magnitude of the proximal colon.E) Resting gut diameter in the proximal colon of control and LPS-treated rats under basal (control) conditions and after NOLA.F) Number of contractions per 15 min.G) Contraction magnitude.Data are displayed as mean with maximum and minimum values.B-D) Student's unpaired t-tests: * p ≤ 0.05, ** p ≤ 0.01; n = 5 per group.E-G) Repeated measures two-way ANOVA.# x $: interaction between LPS and NOLA, #: main effect of LPS, $ main effect of NOLA.Where separated by commas, these are main effects only; "x" indicates a significant interaction.n = 7-8 per group.

Fig. 4 .
Fig. 4. Lipopolysaccharide (LPS) decreases the density and sphericity of ionized calcium-binding adapter molecule 1 (Iba1)-positive (infiltrating) macrophages.Proximal colon myenteric plexus from A) saline-treated and B) LPS-treated rats immunolabelled with cluster of differentiation 163 (CD163; green), Iba1 (red) and Hu (purple), showing enteric neuronal ganglia within the myenteric plexus layer.C-D) LPS does not affect the number of C) macrophages overall (Iba1-positive) or D) resident macrophages (CD163-positive).LPS did not alter the sphericity of E) CD163 resident macrophages.F) Colons from LPS-treated rats had fewer Iba1-positive only macrophages and G-I) reduced Iba1-positive cell sphericity.G) Heatmap of sphericity of Iba1-positive macrophages in saline-treated and H) LPS-treated colons.I) Sphericity of Iba1-positive cells.Data are displayed as mean with maximum and minimum values.* Student's unpaired t-tests: * p ≤ 0.05.Scale bars = 100 µm.Colour bar on sphericity heatmap = 0.2-0.9sphericity.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5 .
Fig. 5. Lipopolysaccharide (LPS) causes displacement of major histocompatibility complex (MHC)II and ionized calcium binding adaptor molecule 1 (Iba1) intestinal macrophages.A) Z-position of MHCII and Iba1 macrophages in a representative saline-treated rat.B) Z-position of MHCII and Iba1 macrophages in a representative LPS-treated rat.Representative rats shown due to individual variation in terms of the Z-position of the MHCII and Iba1 cell populations.C) The difference in the Zposition between MHCII and Iba1 macrophages in saline and LPS treated colon.D) Schematic illustration of the displacement of Iba1 macrophages from myenteric plexus to circular muscle layer during inflammation.Data are displayed as mean with maximum and minimum values.* Student's unpaired t-tests p ≤ 0.05, **** p ≤ 0.0001.N = 6 per group.

Fig. 6 .
Fig. 6.Macrophage depletion does not exacerbate lipopolysaccharide (LPS)-induced weight loss or affect colon length.A) Both macrophage depletion and LPS led to significantly reduced body weights but B) did not affect colon length.Data are displayed as mean ± SEM in panel A and mean with maximum and minimum values in panel B. ** p < 0.01, **** p < 0.0001 indicate statistically significant differences from all other groups using Tukey post hoc tests after a significant interaction between LPS and macrophage depletion with a two-way ANOVA; n = 7-8 per group.

Fig. 7 .
Fig. 7. Intestinal macrophages are not essential for the reduced resting gut diameter, contraction magnitude or increased contraction frequency that occurred due to systemic inflammation.A) Resting gut diameter.B) Number of contractions per 15 min C) Contraction magnitude of rat colons at baseline with saline or lipopolysaccharide (LPS) treatment.Data are displayed as mean with maximum and minimum values.Two-way ANOVA: # main effect of LPS; n = 8 per group.DT: diphtheria toxin in Cx3cr1-Dtr rats to deplete macrophages.

Fig. 8 .
Fig.8.Systemic inflammation with lipopolysaccharide (LPS) alters colonic contraction frequency in the absence of the intestinal macrophage and enteric nervous system input to the colon.A) Representative spatiotemporal heatmaps of control and macrophage-depleted rats with saline or LPS treatment under control conditions and with Nω-nitro-L-arginine (NOLA).B) Resting gut diameter.C) Number of contractions per 15 min.D) Contraction magnitude at baseline and after NOLA with saline or LPS treatment.Data are displayed as mean with maximum and minimum values.Repeated measures three-way ANOVA: # main effect of LPS, $ main effect of NOLA, & main effect of macrophage depletion.Where separated by commas, these are main effects only; "x" indicates a significant interaction.*** p < 0.001, significantly different from all other groups with Tukey post hoc tests after LPS by macrophage depletion by NOLA interaction; n = 7-8 per group.DT: diphtheria toxin in Cx3cr1-Dtr rats to deplete macrophages.