Receptor-Interacting Protein-2 Deficiency Delays Macrophage Migration and Increases Intracellular Infection during Peritoneal Dialysis-Associated Peritonitis

Peritonitis is a major complication of peritoneal dialysis (PD), a technique used in the treatment of patients with end-stage renal disease [1, 2]. Bacteria causing peritonitis may enter the peritoneal cavity through the catheter or via the catheter exit site, and elicit an inflammatory response [3]. Such an inflammatory response is initiated by the release of proinflammatory mediators and chemoattractants from activated peritoneal mesothelial cells and resident macrophages, resulting in the rapid recruitment of neutrophils and inflammatory monocytes from the bloodstream to the peritoneum [4, 5]. Although the release of such mediators is necessary for effective pathogen clearance, many mediators, such as TNF-α and TGF-β may also promote peritoneal membrane fibrosis that may ultimately lead to technique failure [6, 7].Since unchecked inflammatory responses can cause significant tissue damage, it is not surprising that the earliest immune activation events also program a resolution phase to minimize tissue injury and activate tissue repair. One key step in initiating the resolution of inflammation is the activation of the adaptive immune response by antigen-presenting cells. Concurrent to the influx of inflammatory leukocytes from the blood, peritoneal dendritic cells and macrophages migrate out of the cavity to the draining lymphatics by way of the omentum, whereby their interaction with naïve T cells results in the activation of pathogen-specific adaptive responses [8,9,10,11]. These responses are necessary to aid in both the clearance of pathogen and the removal of apoptotic neutrophils, helping return the site to steady-state conditions [12, 13].In an attempt to characterize molecules involved in linking early innate responses with adaptive responses that could potentially serve as therapeutic targets, we identified receptor-interacting protein-2 (RIP2) as a molecular biomarker of PD-associated peritonitis outcome. Specifically, we found that early upregulation of RIP2 expression in peritoneal CD14+ cells correlated with a favorable outcome, while failure to upregulate RIP2 correlated with the need for hospitalization, prolonged inflammatory responses and catheter removal in 36% of patients [14].RIP2, also known as RICK or CARDIAK, is a caspase-recruitment domain (CARD)-containing molecule that has been linked to signaling through the NF-ĸB pathway [15,16,17]. RIP2 is a member of the RIP family of proteins and contains a putative serine/threonine kinase domain in its N-terminus and a CARD domain in its C-terminus. In contrast to other protein kinases, RIP2 is regulated at the level of transcription and its expression is upregulated in immune cells upon activation with bacterial products [14].The function of RIP2 is unknown. Early data indicated a role for RIP2 downstream of various TNF receptor family members, including CD95 and p75, while subsequent data on RIP2-deficient mice implicated RIP2 in the activation of NF-ĸB and JNK downstream of pattern recognition receptors (PRRs; i.e. toll-like receptors (TLRs) and cytosolic NOD1 and NOD2) and the T-cell receptor [15,16,17,18,19]. However, more recent work with highly purified TLR and NOD ligands indicates that RIP2 is required for signaling from NOD1 and NOD2, implying that RIP2 is likely more important for the sensing of intracellular bacteria [20]. This is important when considering that Staphylococcus aureus, a common peritonitis-inducing pathogen, has recently been shown to exist as an intracellular pathogen [21,22,23]. It is thus possible that, in the absence of appropriate NOD signaling, such an infection could go undetected, resulting in significant morbidity.As a result of our observation that RIP2 is a useful biomarker for human peritonitis, we decided to examine in more depth the involvement of this molecule during the immune response to bacterial peritonitis. We report here that patients who do not upregulate RIP2 during PD-associated peritonitis (and thus have a protracted infection) have higher numbers of macrophages in the peritoneal cavity in the early phase of the disease. To dissect this observation at the mechanistic level, we used a mouse model of peritoneal inflammation. With this model, we report here that RIP2-deficiency did not prevent the recruitment of inflammatory neutrophils and monocytes from the blood, but affected the migration capacity of resident peritoneal macrophages, providing a potential explanation for the above-mentioned clinical observation. This defect was only apparent at early time points, indicating that macrophage emigration to the draining lymphatics was delayed but not inhibited. In addition, we found that RIP2-deficient peritoneal macrophages were more permissive to intracellular infection with S. aureus, suggesting that, in the absence of RIP2 expression, resident macrophages could become reservoirs for bacteria during episodes of peritonitis.Patients were recruited from the London Health Sciences Centre Peritoneal Dialysis Unit. Two liters of long-dwell PD effluent from day 1 of PD-associated peritonitis were collected from 19 patients on automated or cycler PD using either Baxter Dianeal or Icodextrin solutions, over the course of 30 months. The demographic data from these 19 patients are representative of those in our PD program and PD patients in general, as reported previously [14]. Diagnosis of PD-associated peritonitis was based on the presence of abdominal pain, a cloudy PD effluent with a leukocyte count above 100 × 106/l of effluent, and a positive effluent culture. The processing of effluent samples has been previously reported [14]. Patient participation in the study was voluntary and research protocols were approved by the Office of Research Ethics at the University of Western Ontario. Peritonitis outcome was defined based on the need for hospitalization; patients not requiring hospitalization were considered to have conventional peritonitis while patients requiring hospitalization within the first 5 days of infection onset were considered to have protracted peritonitis.RIP2+/– mice were obtained from Dr. R. Flavell (Yale University) and bred inhouse to generate homozygous wild-type (RIP2+/+) and RIP2-deficient (RIP2–/–) mice. PCR genotyping and Western blotting confirmed RIP2 deficiency [Fairhead et al., manuscript submitted]. Five to ten mice were used per experimental group, and experiments were repeated at least twice. All mice were housed under conventional conditions, and experimental protocols were approved by the University of Western Ontario Animal Care Committee.Commercial preparations of lipopolysaccharide (LPS; Escherichia coli) and peptidoglycan (PGN; S. aureus) were obtained from Sigma-Aldrich (Oakville, Ont., Canada). Muramyl dipeptide (MDP) was obtained from InvivoGen (Cedarlane Laboratories Ltd., Hornby, Ont., Canada). The preparation of Staphylococcus epidermidis cell-free supernatant (SES) was performed as described previously [24]. Lyophilized stocks of SES were stored at –70°C and resuspended in culture-grade sterile 1× PBS (Sigma-Aldrich).Eight- to twelve-week-old RIP2+/+ and RIP2–/– mice were injected intraperitoneally (i.p.) with 0.5 ml SES or sterile 1× PBS as a control. At 3, 6 and 24 h after injection, peritoneal lavages were performed with 2 ml cold sterile 1× PBS. Lavages were centrifuged at 300 g for 5 min, and supernatants stored at –70°C for cytokine detection. The total number of peritoneal cells in the pellets was determined using a hemacytometer and the various leukocyte populations in the lavages were documented by flow cytometry.Peritoneal cells were incubated with heat-inactivated normal mouse serum before staining with the following directly labelled anti-mouse monoclonal antibodies (mAbs): anti-CD11b (BD Biosciences, Mississauga, Ont., Canada), anti-F4/80 (Serotec, Raleigh, N.C., USA), anti-CD49d, anti-TLR2 and anti-TLR4 (eBiosciences, San Diego, Calif., USA). Cells were analyzed by flow cytometry (Becton Dickinson, Mountain View, Calif., USA). Resident peritoneal macrophages were identified based on high expression of CD11b and F4/80 [25]. Analyses were performed using FlowJo computer software (Treestar, Ashland, Oreg., USA).Murine IL-6, IP-10/CXCL10, KC/CXCL1, CCL1 (R&D Systems, Minneapolis, Minn., USA), MCP-1/CCL2, TNFα and IFNγ (BD Biosciences) were quantified using commercially available ELISA kits. Supernatants were assayed in triplicate and graphed as mean ± SEM.Peritoneal macrophages were isolated by plastic adherence and stimulated ex vivowith LPS (100 ng/ml), PGN (10 µg/ml), MDP (10 µg/ml), or MDP (10 µg/ml) and LPS (2 ng/ml) and plated on 6.5-mm fibrinogen-coated inserts (25 µg/ml, 5 µm pore size) of a 24-well transwell plate (Corning Costar, Fisher Scientific, Ottawa, Ont., Canada) at 37°C. After 4 h, the inserts were washed, fixed with methanol and stained with Gill No. 3 hematoxylin (Sigma-Aldrich). The mean number of migrated cells per cubic centimeter was determined from the counts of five independent fields of view from assays performed in duplicate.Peritoneal macrophages were isolated by plastic adherence, washed and resuspended in antibiotic-free medium and incubated with S. aureus at a multiplicity of infection (MOI) of 10, as described previously [26]. Briefly, peritoneal macrophages and S. aureus were brought into contact by centrifugation and incubated at 37°C. After 2 h, the plates were centrifuged and the supernatants were isolated and stored at –20°C for cytokine analysis. The cells were then washed with fresh medium containing 20 µg/ml gentamycin and incubated for an additional 2 h, after which the cells were washed 2 times with 1× PBS and the intracellular bacteria were released with 0.02% Triton X-100. The number of surviving intracellular bacteria (CFU/ml) was counted by plating serial dilutions in triplicate on TSB agar plates.Statistical comparisons were done either by the Mann-Whitney nonparametric test or the unpaired Student’s t test in cases when data were normally distributed using GraphPad Prism software (GraphPad Software Inc., San Diego, Calif., USA) and a p value of <0.05 was considered statistically significant.We have recently reported that PD patients experiencing episodes of peritonitis can be divided in 2 groups: those who upregulate RIP2 expression early during infection, and those who fail to upregulate RIP2. More importantly, patients showing an early upregulation of RIP2 expression have conventional peritonitis with favorable clinical outcome, whereas those who fail to upregulate RIP2 have protracted peritonitis characterized by the need for hospitalization, longer duration of infection, and a high frequency of catheter loss [14]. These 2 groups, so-called conventional versus protracted peritonitis, were further examined. In patients for whom RIP2 levels and cellular subsets were determined, we did not detect significant differences between these 2 outcome groups in terms of total cellularity (fig. 1a). However, we did find that patients with protracted peritonitis had on average a significantly greater number of macrophages (detected by expression of CD14 on their surface) in their cavities on the first day of infection (fig. 1b; p = 0.0343; Mann-Whitney). Furthermore, we observed that there was an inverse trend (although not significant) between the number of peritoneal macrophages on day 1 of infection and RIP2 mRNA expression levels (fig. 1c).The above-mentioned clinical observation prompted us to examine the role of RIP2 in the earliest phase of an acute inflammatory response using a model of acute peritonitis in RIP2-knockout mice. First, we confirmed that there was no significant variation in the peritoneal leukocyte populations of RIP2-deficient mice. In comparison to wild-type (WT) littermates, RIP2–/– mice had on average a similar number of total peritoneal cells (fig. 2a). However, upon examination of the different subsets, we found that RIP2–/– mice had slightly fewer resident macrophages (8.61 × 105 for RIP2+/+ vs. 6.53 × 105 for RIP2–/–; p = 0.046). No differences for neutrophil or monocyte counts were detected (fig. 2b).Once baseline cellularity in these mice had been determined, we induced peritoneal inflammation. This was done using a well-characterized model of acute peritoneal inflammation induced by i.p. administration of a preparation of SES. In this model, we compared the leukocyte infiltrates and cytokine production in RIP2+/+ versus RIP2–/– mice. Consistent with our human data, we found no significant differences in the total number of peritoneal leukocytes present at 3, 6, or 24 h after injection (fig. 3a), although there appeared to be a trend towards higher cellularity in the RIP2-deficient mice after 3 h. In addition, the total number of infiltrating neutrophils and inflammatory monocytes recruited from the bloodstream was the same in RIP2+/+ and RIP2–/– mice (fig. 3b). This response was specific to the SES, as no leukocyte infiltrate was detected in response to PBS alone (fig. 3a, b, left panels).Consistent with this finding, we failed to see a difference between RIP2+/+ and RIP2–/– mice in the amount of proinflammatory cytokines and chemokines detected in the peritoneal exudates in response to SES injection. In comparison to WT, RIP2-deficient mice produced the same amount of IL-6 and MCP-1 at 3 h after injection (fig. 4a). There was, however, a slight trend for a higher level of IL-6 in the cavities of RIP2+/+ mice at 6 h after injection (fig. 4a; not significant), but at this time point the levels were very low and returning to baseline. We failed to detect significant levels of TNFα, KC, CCL1 or IP-10 in peritoneal exudates from either RIP2+/+ or RIP2 knockout mice (data not shown).To confirm the previous findings, we stimulated peritoneal leukocytes ex vivo with SES for 1, 3, 6, 16 and 24 h (fig. 4b). In accordance with our in vivo data, we failed to see any significant difference in the production of IL-6, MCP-1 or from RIP2+/+ and RIP2–/– peritoneal these that RIP2 not a role in the early activation and recruitment of inflammatory leukocytes in response to SES to an influx of neutrophils and an of resident peritoneal macrophages to the draining lymphatics also early inflammatory events specific populations were analyzed during peritonitis, we found that in comparison to RIP2-deficient mice had a significantly greater number of peritoneal macrophages RIP2+/+ mice at 3 h after injection (fig. p = compared to the total number of resident peritoneal macrophages present in the peritoneal cavity under steady-state conditions, injection of SES induced a significant in the total number of resident macrophages in but not in RIP2-deficient mice (fig. p However, such a difference at time 24 suggesting that macrophage migration was delayed and not confirm this finding, we purified peritoneal macrophages and stimulated the cells ex vivo with LPS (100 ng/ml), PGN (10 µg/ml), MDP (10 µg/ml) or MDP (10 µg/ml) a low of LPS (2 The stimulated cells were then to migrate fibrinogen-coated transwell inserts for 4 h at and the total number of cells that had migrated to the of the was No significant difference in migration capacity of RIP2+/+ and RIP2–/– macrophages in response to LPS or PGN was detected (fig. However, in response to the RIP2-deficient macrophages were significantly in their to migrate the inserts in comparison to = a defect that was with the of LPS (fig. it was possible that the delayed emigration of resident peritoneal macrophages in RIP2–/– mice was to to SES we analyzed the of these cells to upregulate TLR expression in response to SES. We found that both and RIP2-deficient macrophages upregulated and expression to the same after 24 h (fig. indicating that both were This that RIP2–/– mice are to SES RIP2 signaling has been linked to the activation of the NF-ĸB and NF-ĸB the expression of many receptors and it is possible that many migration could be affected in RIP2–/– mice. Although the peritoneal macrophage emigration is recent have indicated a role for the molecules and the in this We did not find differences in the expression of on and RIP2-deficient macrophages (data not However, of on resident peritoneal macrophages was affected by RIP2 of expression was apparent as early as 3 h after SES injection in both and by 24 h (fig. results not but the of expression was more in RIP2-deficient resident peritoneal macrophages at 3 h (fig. p = This difference was not apparent at time indicating that, macrophage expression of at time has recently been shown that RIP2 downstream of NOD1 and activation [20]. To that our RIP2-deficient mice were in their to to signaling, we stimulated RIP2+/+ and RIP2–/– peritoneal leukocytes ex vivo with µg/ml MDP for 1, 3, 6, 16 and 24 h. MDP only in an in IL-6 production above baseline in mice but not in RIP2-deficient mice (fig. with the that RIP2 is required for that RIP2 downstream of the cytosolic pattern recognition receptors NOD1 and NOD2, and that RIP2–/– mice have clearance of we RIP2-deficient resident peritoneal macrophages be more to intracellular To do that, we used a model based on S. This pathogen, S. is a major cause of peritonitis in patients on PD and is considered an intracellular pathogen in of infection [21,22,23]. we purified peritoneal macrophages from both RIP2+/+ and RIP2–/– mice and incubated with S. aureus at an of for 2 h. gentamycin was to and the intracellular bacteria were released by the The number of intracellular bacteria from RIP2-deficient macrophages was significantly higher the number from macrophages (fig. p = with the that RIP2-deficient macrophages were effective at intracellular This was by cytokine data showing that RIP2-deficient macrophages produced IL-6 (fig. and TNFα (fig. in response to intracellular S. aureus, indicating that their responses into S. aureus were this we the role of RIP2 in the of an acute inflammatory response using a model of peritonitis that PD-associated peritonitis [24]. In we found that RIP2 deficiency not the early recruitment of inflammatory leukocytes or the production of the proinflammatory cytokine IL-6 or the monocyte the absence of RIP2 expression the of resident peritoneal macrophages from the peritoneal and is with proinflammatory response and intracellular of S. these a mechanistic for our previous report that RIP2 is a biomarker for PD-associated peritonitis for the observation that patients with PD-associated peritonitis who fail to upregulate RIP2 expression in their peritoneal CD14+ cells containing macrophages and dendritic during the early phase of the disease had a significantly clinical of macrophages and dendritic cells from the site of injury to the is a key step in the from innate to adaptive to the activation of T cells and subsequent of the pathogen through cellular The that RIP2 deficiency such an emigration may be considered an that RIP2 an important role in linking innate with adaptive However, RIP2 not to be for the of the innate inflammatory response no differences between and RIP2-deficient mice in terms of leukocyte recruitment and production were data also that significant TLR can signaling to RIP2 deficiency we failed to see delayed migration in response to or to and the of LPS migration The specific and involved in the response may be as by the differences between our results and those reported by et These implicated RIP2 in the of production by mesothelial cells in response to the i.p. administration of the NOD1 In we that recruitment of neutrophils and monocytes into the peritoneal cavity was not when using in vivo administration of bacterial as our ex vivo migration data the that RIP2 deficiency macrophage in response to activation it not out the that is in The that expression is significantly in the early inflammatory response of RIP2-deficient mice may be linked to the delayed migration observed in the in vivo In response to activation resident peritoneal macrophages migrate to the by way of the Although the this is molecules and proteins and its have been implicated in the and subsequent the peritoneal membrane In a role in this as by the that peritoneal mesothelial cells upregulate the expression of its in response to activation and that is a of the that in the peritoneal which are within the that the peritoneal cavity with the of the In addition, may also and migration by its on the of and by the expression of to the expression following of RIP2-deficient macrophages is In T activation to of by rapid upregulation at time the expression is a of then can that RIP2-deficient macrophages were more activated may that RIP2 could the of for through of RIP2 on NF-ĸB and In dendritic expression was shown to with in a that is by the this has also been shown to RIP2 kinase and although highly a potential role for RIP2 in the of are to other such as chemokines and receptors are regulated by signaling independent of TLR early data RIP2 in signaling downstream of the recent indicates that RIP2 downstream of activation [20]. NOD1 and are cytosolic pattern recognition receptors implicated in the a number of intracellular including and Although in many it is an pathogen, S. aureus is also as an important intracellular pathogen of different including and mesothelial cells S. aureus could be to the of of NOD The data here confirm this RIP2-deficient peritoneal macrophages were more permissive to intracellular infection with S. This also correlated with production of the inflammatory cytokines IL-6 and TNFα indicating that, in the absence of RIP2 expression, peritoneal macrophages could become significant reservoirs for bacterial This is important as peritoneal macrophages from PD patients with peritonitis were shown to have a intracellular and intracellular with S. aureus and are with the production of which in infection and here have clinical as are important in PD-associated peritonitis, and S. aureus in are with significant and [1, Specifically, the point that the absence of RIP2 upregulation in the of a PD-associated peritonitis by these be considered a normal conditions, the early immune response to S. aureus is by the of these on Such an interaction to upregulation of RIP2 expression as we have previously reported [14]. RIP2 then to intracellular S. aureus that have the and the clearance of S. aureus in its intracellular phase of In the of RIP2 upregulation in the of infection by S. aureus the of pathogen It is important to the are that the of patients to upregulate RIP2 in response to our data that failure to upregulate RIP2 expression during PD-associated peritonitis the clearance of intracellular as as the migration of antigen-presenting cells into the draining Such could the of an adaptive immune resulting in protracted In addition, these can also lead to inflammation as previously documented in experimental of fibrosis and and on the of the here not only a mechanistic for our previous report that failure to upregulate RIP2 expression with protracted PD-associated peritonitis but also RIP2 as a therapeutic for the of inflammatory responses to Dr. for in the in migration and the of the for and work was by the Research from the of the of Health and by from the of and the at the London Health Sciences a Research in

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.