The American Journal of Surgery (2015) 209, 493-497
Midwest Surgical Association
Phosphatidylcholine and the intestinal mucus layer: in vitro efficacy against Clostridium difficile-associated polymorphonuclear neutrophil activation Alicia Olson, M.D., Lawrence N. Diebel, M.D.*, David M. Liberati, M.S. Department of Surgery, 6C University Health Center, 4201 Saint Antoine, Detroit, MI 48201, USA
KEYWORDS: Phosphatidylcholine; PMN activation; Clostridium difficile; Chemotaxis
Abstract BACKGROUND: Phosphatidylcholine (PC), an important component of intestinal mucus, protects against Clostridium difficile toxin-induced intestinal barrier injury in vitro. Polymorphonuclear neutrophil (PMN) activation may contribute to intestinal injury and systemic toxicity in patients with C. difficile-associated disease. We therefore hypothesized that the intestinal barrier function against C. difficile toxin by exogenous PC would ameliorate PMN activation. METHODS: Intestinal epithelial cell (IEC) monolayers were cocultured with C. difficile toxin A and/ or exogenous PC. Naı¨ve PMNs were cocultured with IEC culture supernatants and PMN activation, and chemotactic potential determined. RESULTS: PC treatment of IEC abrogated the enhanced PMN activation and chemotactic potential following toxin A exposure (P , .001). CONCLUSIONS: Exogenous PC ameliorated PMN activation from IECs exposed to C. difficile toxin. Administration of exogenous PC may be a useful adjunctive treatment in severely ill or immunocompromised patients with C. difficile-associated disease. Ó 2015 Elsevier Inc. All rights reserved.
Clostridium difficile-associated disease (CDAD) has increased in incidence and severity especially since 2000. The increase in disease severity may in part relate to the emergence of epidemic strains with enhanced virulence properties.1 However, host proinflammatory and humoral immune responses are also important in limiting disease severity and risk of mortality with CDAD.2 Study was supported by departmental funding only. Presented at the Midwest Surgical Meeting, August 3–6, 2014. * Corresponding author. Tel.: 11-313-577-5314; fax: 11-313-5775310. E-mail address: [email protected] Manuscript received July 31, 2014; revised manuscript October 7, 2014 0002-9610/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2014.10.012
It is now recognized that new therapies are needed for the treatment of CDAD, especially in patients with severe or recurrent disease. In this regard, the colonic mucus layer has been demonstrated to be an important component of the innate barrier that prevents colonic bacteria from invading the mucosa and causing inflammation. In a previous study, we demonstrated that the mucus layer acts in a synergistic fashion with secretory IgA (the main antibody in intestinal mucosal secretions) to protect against C. difficile toxininduced intestinal injury.3 Phosphatidylcholine (PC) is an important component of the mucus layer and it has intrinsic anti-inflammatory properties.4 Decreased PC content in the intestinal mucus
494 has been demonstrated in patients with chronic ulcerative colitis (CUC).5 This finding lead to the clinically effective use of a delayed release formulation of PC in recently published trials in CUC.5 In a prior in vitro study, we demonstrate that exogenous PC supplementation protects against intestinal barrier injury after C. difficile toxin exposure.6 This was evident even in experiments with an intact mucus layer. Intestinal epithelial cell (IEC) neutrophil interactions are important in mucosal inflammation.7 In our prior study, there was a 3-fold to 4-fold decrease in the proinflammatory cytokines tumor necrosis factor alpha and interleukin 6 (IL-6) with IECs supplemented with PC and exposed to C. difficile toxin A.6 This may be important as proinflammatory signaling from the gut is a trigger for neutrophil trafficking to the gut, which may increase intestinal barrier injury. Modulation of neutrophil–epithelial cell interaction may be a novel approach in the treatment of severe forms of CDAD. We therefore hypothesized that exogenous PC supplementation would ameliorate polymorphonuclear neutrophil (PMN) inflammatory responses and chemotactic potential. This was studied in an in vitro model.
Patients and Methods Intestinal epithelial cells HT29 cells were obtained from American Type Culture Collection and routinely cultured with Dulbecco’s Modified Eagle Medium containing 10% fetal bovine serum, 4.5 g/L glucose, and gentamicin in an atmosphere of 5% CO2 at 37 C. Cells (5 ! 105) were seeded on the apical surface of a polycarbonate membrane (3.0-mm pore size) (Transwell; Corning Costar Core, Cambridge, MA) in a 2chamber cell culture system and allowed to form polarized monolayers. Monolayer integrity was monitored by serial measurement of the transepithelial electrical resistance with a Millicell electrical resistance meter (Millipore Corp., Bedford, MA).
The American Journal of Surgery, Vol 209, No 3, March 2015 in this study to extend these findings to PMN responses in this model.
Polymorphonuclear neutrophil isolation Venous whole blood was collected from random healthy donors in vacuum tubes containing ethyenediamine tetra-acetic acid. PMNs were isolated by first incubating the whole blood with 6% dextran for 45 minutes at 4 C. The leukocyte-rich supernatant was aspirated and layered on top of Histopaque 1077 (Sigma, St. Louis, MO) and centrifuged at 1,300 rpm (400g) for 30 minutes at 4 C. A 25-second red blood cell lysis was performed on the pellet and isotonicity restored by adding 2 mL of 3.4% NaCl. The cells were then diluted in phosphate buffered saline (PBS) and centrifuged for 10 minutes at 1,300 rpm at 4 C. The pellet was washed and gently resuspended in PBS at a concentration of 1 ! 106 cells/mL and used immediately.
Experimental design Confluent HT29 and HT29-MTX IECs were first established in a 2-chamber cell culture system. Clostridium difficile toxin A (50 mg/mL) was cocultured with the IEC for 6 hours. In a subset of experiments, phosphatidylcholine (PC) (100 mM) was cocultured with the IECs for 1 hour before exposure to C. difficile toxin A. Basal chamber supernatants from HT29 and HT29-MTX monolayers exposed to toxin A and PC as described above were collected and cocultured with naı¨ve PMNs freshly isolated from healthy volunteers, and PMN activation was indexed by CD11b expression (MFI), superoxide anion generation (O22) after addition of N-formylmethionyl-leucyl-phenylalanine (fMLP) (activation), and percent elastase release. Chemotaxis of PMNs was analyzed following incubation of IEC culture supernatants (with or without C. difficile and PC) with PMNs in a chemotaxis system.
CD11b adhesion molecule quantification HT29 methotrexate cells The HT29 MTX cell line was isolated from human HT29 colon carcinoma cells through growth adaptation to methotrexate (MTX) as described by Olson et al.6 Briefly, exposure of HT29 cells to 1027 M MTX led to differentiation into a homogeneous monolayer of polarized goblet cells, which secrete mucins of gastric immunoreactivity. After stabilization of the growth curve, cells adapted to 1027 M MTX after passage 8 were reverted back to drugfree medium and subsequently cultured in the absence of MTX. The growth curve remained unchanged as compared with the MTX-treated cells. In our previous study, exogenous PC protected the intestinal barrier against C. difficile toxin A in both mucus-producing and non–mucusproducing HT29 clones. Thus, we used both HT29 clones
Mean receptor density on the PMN plasma membrane was quantified by flow cytometry with a fluorescent (phycoerthrin)-labeled monoclonal antibody directed against the CD11b receptor (BD Biosciences Pharmingen, San Diego, CA). Briefly, PMNs were cocultured with HT29 or HT29-MTX supernatants for 60 minutes at 37 C and then isolated by centrifugation and subsequently incubated with a 1:100 dilution of the anti-CD11b antibody for 30 minutes at 4 C in the dark. Control samples were stained separately with a mouse isotype control antibody to assess background antibody binding. The PMNs were then washed with cold PBS containing 0.5 mM glucose and .1% gelatin and fixed with 1% paraformaldehyde in PBS. The data are reported as mean fluorescence intensity, reflecting the mean CD11b receptor density on the PMN cell surface.
A. Olson et al.
Phosphatidylcholine and the intestinal mucus layer
Superoxide assay Superoxide generation by PMNs was measured by superoxide dismutase-inhibitable cytochrome c reduction. PMNs exposed to supernatants collected from HT29 and HT29-MTX cells cocultured with C. difficile toxin A 6 PC were added to microtiter wells. Some PMNs were preincubated with LTB4 (1 mmol/mL) (Sigma) for 5, 10, 30, or 60 minutes at 37 C before being added to microtiter plates. All wells also contained cytochrome c (80 mmol/ mL) and blank wells contained superoxide dismutase at 15 mg/mL to arrive at a total volume of 150 mL. fMLP (1 mmol/mL) (Sigma) was now added to the experimental wells to initiate the activation of the PMNs. Immediately after the addition of fMLP, the plate was placed in a plate reader and absorbance was measured at 550 nm every 20 seconds for 5 minutes. The rate of superoxide anion production (Vmax) was then determined by the slope of the absorbance curve using GraphPad Prism 4. The data recorded were the endpoint or maximal amount of superoxide anion produced over the 5-minute period and is expressed as nanomoles per milligram protein.
Elastase assay PMN elastase release was measured by incubation of HT29 and HT29-MTX cell supernatants with the elastase substrate methoxy-succinyl (MeO-Succ)-Ala-Ala-Pro-Val (AAPV)-para-nitroanilide (PNA). PMNs were exposed to HT29 and HT29-MTX culture supernatants collected after coculture with C. difficile toxin A and PC for 6 hours and 1 hour, respectively. An additional sample of PMNs was treated with Triton X-100 (.1%) for quantification of the total PMN elastase. All cells were then centrifuged and the supernatants incubated for 1 hour at 37 C in N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid/NaCl buffer (pH 7.5) containing 1% MeO-Succ-AAPV-PNA. Cleavage of the substrate with release of PNA was measured at 405 nm in duplicate wells. Control wells containing the elastase inhibitor MeO-Succ-AAPV-chloromethyl ketone were subtracted from the above samples and data are expressed as percentage of total cell elastase.
Chemotaxis assay Chemotaxis of PMNs was assessed using the ChemoTx System purchased from NeuroProbe, Inc. (Gaithesburg, MD). Supernatants from the treated HT29 and HT29-MTX cell cultures were placed in the bottom layer of the chemotaxis system (300 mL). A total of 50 mL of purified PMNs at a concentration of 1 ! 106 cells/mL in Roswell Park Memorial Institute medium was added to the top layer of the chemotaxis membrane and allowed to incubate for 2 hours at 37 C. After incubation, PMN chemotaxis into the basal chamber was quantitated using light microscopy. Briefly, 10 mL of the basal chamber supernatants were
495
added to 90 mL of trypan blue and a 10 mL aliquot added to a hemocytometer and total cell numbers were quantitated using light microscopy.
Cytokine analysis by enzyme-linked immunosorbent assay Basal supernatants of HT29 and HT29-MTX cells in the presence and absence of PC for 1 hour and toxin A for up to 6 hours were collected at the end of the experiments. IL-8 was quantitated in the supernatant samples using a solid phase sandwich enzyme-linked immunosorbent assay. These immunoassay kits are commercially available and were used according to the manufacturer’s directions (Cytoscreen; Biosource International, Camerillo, CA). The minimal detectable levels of IL-8 with this kit are 2 pg/mL.
Statistical analysis An analysis of variance with a post hoc Tukey test was used to analyze the data. Statistical significance was inferred at P values of less than .001. All data are expressed as mean 6 SD.
Results The data in Table 1 show the effects on PMN activation after exposure to IEC supernatants 6 C. difficile toxin A. There was nearly a 2.5! increase in CD11b expression in IEC cells exposed to toxin A. This was reduced to baseline levels with either an intact mucus layer or by PC supplementation. Neutrophil respiratory burst as indexed by superoxide anion production was increased 3! in IEC exposed to toxin A versus PMN cocultured with IEC. PC supplementation reduced superoxide anion generation by IEC after exposure to toxin A. However, these levels remained above baseline levels. Elastase, a protease in PMN intracellular granules, was increased nearly 3-fold in IEC after exposure to toxin A versus controls. PC reduced elastase in both non–mucus-containing and mucus-containing IEC. However, elastase remained above baseline levels, similar to the results noted with superoxide anion generation. IL-8 release, a potent chemotactic factor released by IEC, and neutrophil chemotaxis data are shown in Table 2. IL-8 was increased 8! versus control after IEC exposure to toxin A (without mucus). There was roughly a 50% reduction in IL-8 release with mucus versus non–mucusproducing IEC after exposure to toxin A. Supplemental PC and the native mucus layer decreased IL8 production in a similar fashion. Neutrophil chemotaxis was significantly greater in response to culture supernatants from the nonmucus IEC after exposure to toxin A. Either the native mucus layer or supplemented PC reduced PMN chemotaxis to at or near
496
The American Journal of Surgery, Vol 209, No 3, March 2015
Table 1
PMN activation quantified after exposure to HT29 and HT29-MTX supernatants cocultured with exogenous PC and/or toxin A Elastase (%)
PMN PMN PMN PMN PMN PMN PMN PMN
alone 1 PC (cell free) 1 HT29 supernatant 1 HT29 1 toxin A supernatant 1 HT29 1 PC 1 toxin A supernatant 1 HT29-MTX supernatant 1 HT29-MTX 1 toxin A supernatant 1 HT29-MTX 1 PC 1 toxin A supernatant
8.3 8.0 9.8 27.6 10.4 10.2 15.7 11.6
6 6 6 6 6 6 6 6
Superoxide anion (Nmol/well)
.8 1.0 1.2 2.3* .7†,‡ 1.1 1.6*,† 1.3
6.8 6.6 10.0 32.4 13.5 10.7 19.3 16.1
6 6 6 6 6 6 6 6
CD11b (MFI)
.6 1.1 2.2 3.4* 2.1†,‡ 1.3 2.0*,† 1.4
99.5 91.3 108.2 268.4 115.6 111.4 128.9 120.5
6 6 6 6 6 6 6 6
2.5 4.4 4.3 7.9* 5.6†,‡ 3.5 4.8*,† 3.5*,†
Results are expressed as mean 6 SD, n 5 5 for each group. PC 5 phosphatidylcholine; PMN 5 polymorphonuclear neutrophil; SD 5 standard deviation. *P , .001 vs HT29 supernatant. † P , .001 vs HT29 1 toxin A. ‡ P , .001 vs HT29-MTX 1 toxin A.
baseline values. In experiments using C. difficile toxin B, PMN activation and chemotaxis results were similar to those found with exposure using toxin A (results not shown).
animal and clinical studies with C. difficile infection show significant neutrophil infiltration that varies with disease severity.10,11 In addition, systemic neutrophil activation is evidenced by findings of a significantly elevated peripheral blood neutrophil count and bandemia. Although neutrophil infiltration is beneficial during intestinal inflammation, it may be pathological if prolonged or exaggerated. A previous study has indicated that toxin A directly stimulated human neutrophils at relatively high concentrations and toxin B had no direct stimulatory effect at all.11 Thus, the mechanism for neutrophil activation by C. difficile toxin in vivo remains unclear. In this prior study, Linevsky et al11 demonstrated that human monocytes release significant amounts of the neutrophil chemoattractant IL8 following C. difficile toxin A or B exposure. Subsequent monocyte activation was then found to initiate neutrophil activation in vitro likely because of the production of proinflammatory cytokines. IECs are an important part of the innate immune barrier defense and secrete a variety of factors that ‘‘orchestrate’’ the immunoinflammatory responses at mucosal surfaces.7 In our previous studies, we have shown that HT29 colonic epithelial cells produce tumor necrosis factor alpha and
Comments Clostridium difficile is an increasingly important hospital-associated pathogen. Its clinical spectrum ranges from asymptomatic colonization to fulminant colitis and death. The variable clinical manifestations likely reflect C. difficile virulence factors as well as the immune and inflammatory response of the host.1,2 The major virulence factors include toxin A and toxin B, a binary toxin, and surface layer proteins.8 The importance of toxin levels produced by the organism has been noted with the emergence of the hypervirulent B1/NAP1/027 strain. However, in other studies markers of intestinal inflammation rather than bacterial factors have been found to correlate better with clinical outcome in patients with CDAD.2,8 Neutrophil recruitment is an essential component in the pathogenesis of C. difficile toxin-induced injury.9 Both
Table 2 Basal cytokine expression and PMN chemotaxis quantified in HT29 and HT29-MTX supernatants collected after cell monolayer exposure to exogenous PC and/or C. difficile toxin A Basal IL-8 (pg/mL) HT29 control HT29 1 toxin A supernatant HT29 1 PC 1 toxin A supernatant HT29-MTX supernatant HT29-MTX 1 toxin A supernatant HT29-MTX 1 PC 1 toxin A supernatant
3.9 31.9 6.8 4.8 16.6 4.4
6 6 6 6 6 6
.3 1.9* 1.1*,† .8 1.0*,† 1.1†,‡
Results are expressed as mean 6 SD, n 5 5 for each group. PC 5 phosphatidylcholine; PMN 5 polymorphonuclear neutrophil; SD 5 standard deviation. *P , .001 vs HT29. † P , .001 vs HT29 1 toxin A. ‡ P , .001 vs HT29 1 toxin A 1 PC. x P , .001 vs HT29-MTX 1 toxin A.
Chemotaxis (cell count) 2,625 6,250 3,500 2,525 3,150 2,250
6 6 6 6 6 6
120 202 150* 105 100*,† 82†,x
A. Olson et al.
Phosphatidylcholine and the intestinal mucus layer
IL-6 in response to C. difficile toxin and this response was modulated by an intact mucus layer or by the addition of exogenous PC.3,6 Exogenous PC was effective in this study likely as ‘‘reinforcement’’ to the mucus layer but also by direct effect of PC on proinflammatory signaling in IECs. In our preparation, there is no PC in the culture media; thus the results reflect only exogenous sources. These levels may vary in the clinical scenario. In this study, we demonstrate that exogenous PC downregulates neutrophil activation and trafficking after exposure to culture supernatants from IEC exposure to C. difficile toxin. This is likely because of a reduction in proinflammatory signaling from IEC cells treated with PC and exposed to C. difficile toxin. There is increasing evidence that exogenous PC is helpful in resolution of inflammation in patients with CUC. Studies in patients with ulcerative colitis have shown that a modified PC preparation is necessary to achieve effective release of PC in distal ileum.5 This delayed release formulation avoids early intestinal absorption of the supplemented PC, thus allowing effective delivery into the colon. Current PC formulas may even be efficacious when instilled locally via a loop ileostomy in patients treated with a colonic salvage therapy as proposed by Neal et al.12 We believe that this study supports our ongoing hypothesis that ‘‘reinforcement’’ of the mucus layer may be a useful adjunct in patients with severe forms of CDI. However, further studies in animal models of CDAD are warranted.
References 1. Vedantam G, Clark A, Chu M, et al. Clostridium difficile infectiond toxins and non-toxin virulence factors, and their contributions to disease establishment and host response. Gut Microbes 2012;3: 121–34. 2. Solomon K, Martin AJ, O’Donoghue C, et al. Mortality in patients with Clostridium difficile infection correlates with host proinflammatory and humoral immune responses. J Med Microbiol 2013;62:1453–60. 3. Olson A, Diebel LN, Liberati DM. Effect of host defenses on Clostridium difficile toxin-induced intestinal barrier injury. J Trauma Acute Care Surg 2013;74:983–9. 4. Ehehalt R, Braun A, Karner M, et al. Phosphatidylcholine as a constituent in the colonic mucosal barrierdphysiological and clinical relevance. Biochim Biophys Acta 2010;1801:983–93. 5. Karner M, Kocjan A, Stein J, et al. First multicenter study of modified release phosphatidylcholine ‘‘LT-02’’ in ulcerative colitis: a randomized, placebo-controlled trial in mesalazine-refractory courses. Am J Gastroenterol 2014;109:1041–51. 6. Olson A, Diebel LN, Liberati DM. Exogenous phosphatidylcholine supplementation improves intestinal barrier defense against Clos-
7. 8.
9. 10. 11.
12.
497
tridium difficile toxin. J Trauma Acute Care Surg 2014;77:570–5; discussion, 576. Kinnebrew MA, Pamer EG. Innate immune signaling in defense against intestinal microbes. Immunol Rev 2012;245:113–31. El Feghaly RE, Stauber JL, Deych E, et al. Markers of intestinal inflammation, not bacterial burden, correlate with clinical outcomes in Clostridium difficile infection. Clin Infect Dis 2013;56:1713–21. Fournier BM, Parkos CA. The role of neutrophils during intestinal inflammation. Mucosal Immunol 2012;5:354–66. Price AB, Davies DR. Pseudomembranous colitis. J Clin Path 1977;30:1–12. Linevsky JK, Pothoulakis C, Keates S, et al. IL-8 release and neutrophil activation by Clostridium difficile toxin-exposed human monocytes. Am J Physiol 1997;273:G1333–40. Neal MD, Alverdy JC, Simmons RL, et al. Diverting loop ileostomy and colonic lavage: an alternative to total abdominal colectomy for the treatment of severe complicated Clostridium difficile associated disease. Ann Surg 2011;254:423–7.
Discussion Dr Anthony J. Senagore (Saginaw, MI). Do you have any data assessing PMN activation status due to C. diff toxin without the presence of the colonic cells, ie, is this directly antiinflammatory on the PMN’s or is it actually mucosal enhancement? Second, do you have any clinical model data demonstrating either reduced inflammatory infiltrates into the bowel wall as a result of C. diff toxin and the enhancement of the mucosal barrier? Third question, do you have any data on glutathione levels? Because Phosphatidylcholine is a precursor molecule for glutathione, and that may be the active agent. And, finally, do you have any data on enhancement of the mucosal layer in other models that increase the risk for C. diff infection, like chronic PPI exposure? Dr Olson: Toxin A actually is able to activate PMN’s but that’s at super physiologic levels. Toxin B has not been shown to activate PMN’s on its own at all. Secondly, do we have any animal model data or any other clinical models describing this reduced inflammatory cell infiltrate? And that’s where we are proceeding in our next steps. We don’t have an animal model that we work with at the current time. And, number three, do you have any data on glutathione as a mechanism? Many of the injuries that are seen are oxidant based, so that may be where all this is coming from. Finally, we do have data from our laboratory demonstrating that H2 blockers decrease intestinal epithelial cell mucus layers. It has been shown that patients using PPI therapy are more at risk for C. diff activation, but this is because of direct cellular effects. It’s not directly related to the mucus.
Midwest Surgical Association
Phosphatidylcholine and the intestinal mucus layer: in vitro efficacy against Clostridium difficile-associated polymorphonuclear neutrophil activation Alicia Olson, M.D., Lawrence N. Diebel, M.D.*, David M. Liberati, M.S. Department of Surgery, 6C University Health Center, 4201 Saint Antoine, Detroit, MI 48201, USA
KEYWORDS: Phosphatidylcholine; PMN activation; Clostridium difficile; Chemotaxis
Abstract BACKGROUND: Phosphatidylcholine (PC), an important component of intestinal mucus, protects against Clostridium difficile toxin-induced intestinal barrier injury in vitro. Polymorphonuclear neutrophil (PMN) activation may contribute to intestinal injury and systemic toxicity in patients with C. difficile-associated disease. We therefore hypothesized that the intestinal barrier function against C. difficile toxin by exogenous PC would ameliorate PMN activation. METHODS: Intestinal epithelial cell (IEC) monolayers were cocultured with C. difficile toxin A and/ or exogenous PC. Naı¨ve PMNs were cocultured with IEC culture supernatants and PMN activation, and chemotactic potential determined. RESULTS: PC treatment of IEC abrogated the enhanced PMN activation and chemotactic potential following toxin A exposure (P , .001). CONCLUSIONS: Exogenous PC ameliorated PMN activation from IECs exposed to C. difficile toxin. Administration of exogenous PC may be a useful adjunctive treatment in severely ill or immunocompromised patients with C. difficile-associated disease. Ó 2015 Elsevier Inc. All rights reserved.
Clostridium difficile-associated disease (CDAD) has increased in incidence and severity especially since 2000. The increase in disease severity may in part relate to the emergence of epidemic strains with enhanced virulence properties.1 However, host proinflammatory and humoral immune responses are also important in limiting disease severity and risk of mortality with CDAD.2 Study was supported by departmental funding only. Presented at the Midwest Surgical Meeting, August 3–6, 2014. * Corresponding author. Tel.: 11-313-577-5314; fax: 11-313-5775310. E-mail address: [email protected] Manuscript received July 31, 2014; revised manuscript October 7, 2014 0002-9610/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2014.10.012
It is now recognized that new therapies are needed for the treatment of CDAD, especially in patients with severe or recurrent disease. In this regard, the colonic mucus layer has been demonstrated to be an important component of the innate barrier that prevents colonic bacteria from invading the mucosa and causing inflammation. In a previous study, we demonstrated that the mucus layer acts in a synergistic fashion with secretory IgA (the main antibody in intestinal mucosal secretions) to protect against C. difficile toxininduced intestinal injury.3 Phosphatidylcholine (PC) is an important component of the mucus layer and it has intrinsic anti-inflammatory properties.4 Decreased PC content in the intestinal mucus
494 has been demonstrated in patients with chronic ulcerative colitis (CUC).5 This finding lead to the clinically effective use of a delayed release formulation of PC in recently published trials in CUC.5 In a prior in vitro study, we demonstrate that exogenous PC supplementation protects against intestinal barrier injury after C. difficile toxin exposure.6 This was evident even in experiments with an intact mucus layer. Intestinal epithelial cell (IEC) neutrophil interactions are important in mucosal inflammation.7 In our prior study, there was a 3-fold to 4-fold decrease in the proinflammatory cytokines tumor necrosis factor alpha and interleukin 6 (IL-6) with IECs supplemented with PC and exposed to C. difficile toxin A.6 This may be important as proinflammatory signaling from the gut is a trigger for neutrophil trafficking to the gut, which may increase intestinal barrier injury. Modulation of neutrophil–epithelial cell interaction may be a novel approach in the treatment of severe forms of CDAD. We therefore hypothesized that exogenous PC supplementation would ameliorate polymorphonuclear neutrophil (PMN) inflammatory responses and chemotactic potential. This was studied in an in vitro model.
Patients and Methods Intestinal epithelial cells HT29 cells were obtained from American Type Culture Collection and routinely cultured with Dulbecco’s Modified Eagle Medium containing 10% fetal bovine serum, 4.5 g/L glucose, and gentamicin in an atmosphere of 5% CO2 at 37 C. Cells (5 ! 105) were seeded on the apical surface of a polycarbonate membrane (3.0-mm pore size) (Transwell; Corning Costar Core, Cambridge, MA) in a 2chamber cell culture system and allowed to form polarized monolayers. Monolayer integrity was monitored by serial measurement of the transepithelial electrical resistance with a Millicell electrical resistance meter (Millipore Corp., Bedford, MA).
The American Journal of Surgery, Vol 209, No 3, March 2015 in this study to extend these findings to PMN responses in this model.
Polymorphonuclear neutrophil isolation Venous whole blood was collected from random healthy donors in vacuum tubes containing ethyenediamine tetra-acetic acid. PMNs were isolated by first incubating the whole blood with 6% dextran for 45 minutes at 4 C. The leukocyte-rich supernatant was aspirated and layered on top of Histopaque 1077 (Sigma, St. Louis, MO) and centrifuged at 1,300 rpm (400g) for 30 minutes at 4 C. A 25-second red blood cell lysis was performed on the pellet and isotonicity restored by adding 2 mL of 3.4% NaCl. The cells were then diluted in phosphate buffered saline (PBS) and centrifuged for 10 minutes at 1,300 rpm at 4 C. The pellet was washed and gently resuspended in PBS at a concentration of 1 ! 106 cells/mL and used immediately.
Experimental design Confluent HT29 and HT29-MTX IECs were first established in a 2-chamber cell culture system. Clostridium difficile toxin A (50 mg/mL) was cocultured with the IEC for 6 hours. In a subset of experiments, phosphatidylcholine (PC) (100 mM) was cocultured with the IECs for 1 hour before exposure to C. difficile toxin A. Basal chamber supernatants from HT29 and HT29-MTX monolayers exposed to toxin A and PC as described above were collected and cocultured with naı¨ve PMNs freshly isolated from healthy volunteers, and PMN activation was indexed by CD11b expression (MFI), superoxide anion generation (O22) after addition of N-formylmethionyl-leucyl-phenylalanine (fMLP) (activation), and percent elastase release. Chemotaxis of PMNs was analyzed following incubation of IEC culture supernatants (with or without C. difficile and PC) with PMNs in a chemotaxis system.
CD11b adhesion molecule quantification HT29 methotrexate cells The HT29 MTX cell line was isolated from human HT29 colon carcinoma cells through growth adaptation to methotrexate (MTX) as described by Olson et al.6 Briefly, exposure of HT29 cells to 1027 M MTX led to differentiation into a homogeneous monolayer of polarized goblet cells, which secrete mucins of gastric immunoreactivity. After stabilization of the growth curve, cells adapted to 1027 M MTX after passage 8 were reverted back to drugfree medium and subsequently cultured in the absence of MTX. The growth curve remained unchanged as compared with the MTX-treated cells. In our previous study, exogenous PC protected the intestinal barrier against C. difficile toxin A in both mucus-producing and non–mucusproducing HT29 clones. Thus, we used both HT29 clones
Mean receptor density on the PMN plasma membrane was quantified by flow cytometry with a fluorescent (phycoerthrin)-labeled monoclonal antibody directed against the CD11b receptor (BD Biosciences Pharmingen, San Diego, CA). Briefly, PMNs were cocultured with HT29 or HT29-MTX supernatants for 60 minutes at 37 C and then isolated by centrifugation and subsequently incubated with a 1:100 dilution of the anti-CD11b antibody for 30 minutes at 4 C in the dark. Control samples were stained separately with a mouse isotype control antibody to assess background antibody binding. The PMNs were then washed with cold PBS containing 0.5 mM glucose and .1% gelatin and fixed with 1% paraformaldehyde in PBS. The data are reported as mean fluorescence intensity, reflecting the mean CD11b receptor density on the PMN cell surface.
A. Olson et al.
Phosphatidylcholine and the intestinal mucus layer
Superoxide assay Superoxide generation by PMNs was measured by superoxide dismutase-inhibitable cytochrome c reduction. PMNs exposed to supernatants collected from HT29 and HT29-MTX cells cocultured with C. difficile toxin A 6 PC were added to microtiter wells. Some PMNs were preincubated with LTB4 (1 mmol/mL) (Sigma) for 5, 10, 30, or 60 minutes at 37 C before being added to microtiter plates. All wells also contained cytochrome c (80 mmol/ mL) and blank wells contained superoxide dismutase at 15 mg/mL to arrive at a total volume of 150 mL. fMLP (1 mmol/mL) (Sigma) was now added to the experimental wells to initiate the activation of the PMNs. Immediately after the addition of fMLP, the plate was placed in a plate reader and absorbance was measured at 550 nm every 20 seconds for 5 minutes. The rate of superoxide anion production (Vmax) was then determined by the slope of the absorbance curve using GraphPad Prism 4. The data recorded were the endpoint or maximal amount of superoxide anion produced over the 5-minute period and is expressed as nanomoles per milligram protein.
Elastase assay PMN elastase release was measured by incubation of HT29 and HT29-MTX cell supernatants with the elastase substrate methoxy-succinyl (MeO-Succ)-Ala-Ala-Pro-Val (AAPV)-para-nitroanilide (PNA). PMNs were exposed to HT29 and HT29-MTX culture supernatants collected after coculture with C. difficile toxin A and PC for 6 hours and 1 hour, respectively. An additional sample of PMNs was treated with Triton X-100 (.1%) for quantification of the total PMN elastase. All cells were then centrifuged and the supernatants incubated for 1 hour at 37 C in N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid/NaCl buffer (pH 7.5) containing 1% MeO-Succ-AAPV-PNA. Cleavage of the substrate with release of PNA was measured at 405 nm in duplicate wells. Control wells containing the elastase inhibitor MeO-Succ-AAPV-chloromethyl ketone were subtracted from the above samples and data are expressed as percentage of total cell elastase.
Chemotaxis assay Chemotaxis of PMNs was assessed using the ChemoTx System purchased from NeuroProbe, Inc. (Gaithesburg, MD). Supernatants from the treated HT29 and HT29-MTX cell cultures were placed in the bottom layer of the chemotaxis system (300 mL). A total of 50 mL of purified PMNs at a concentration of 1 ! 106 cells/mL in Roswell Park Memorial Institute medium was added to the top layer of the chemotaxis membrane and allowed to incubate for 2 hours at 37 C. After incubation, PMN chemotaxis into the basal chamber was quantitated using light microscopy. Briefly, 10 mL of the basal chamber supernatants were
495
added to 90 mL of trypan blue and a 10 mL aliquot added to a hemocytometer and total cell numbers were quantitated using light microscopy.
Cytokine analysis by enzyme-linked immunosorbent assay Basal supernatants of HT29 and HT29-MTX cells in the presence and absence of PC for 1 hour and toxin A for up to 6 hours were collected at the end of the experiments. IL-8 was quantitated in the supernatant samples using a solid phase sandwich enzyme-linked immunosorbent assay. These immunoassay kits are commercially available and were used according to the manufacturer’s directions (Cytoscreen; Biosource International, Camerillo, CA). The minimal detectable levels of IL-8 with this kit are 2 pg/mL.
Statistical analysis An analysis of variance with a post hoc Tukey test was used to analyze the data. Statistical significance was inferred at P values of less than .001. All data are expressed as mean 6 SD.
Results The data in Table 1 show the effects on PMN activation after exposure to IEC supernatants 6 C. difficile toxin A. There was nearly a 2.5! increase in CD11b expression in IEC cells exposed to toxin A. This was reduced to baseline levels with either an intact mucus layer or by PC supplementation. Neutrophil respiratory burst as indexed by superoxide anion production was increased 3! in IEC exposed to toxin A versus PMN cocultured with IEC. PC supplementation reduced superoxide anion generation by IEC after exposure to toxin A. However, these levels remained above baseline levels. Elastase, a protease in PMN intracellular granules, was increased nearly 3-fold in IEC after exposure to toxin A versus controls. PC reduced elastase in both non–mucus-containing and mucus-containing IEC. However, elastase remained above baseline levels, similar to the results noted with superoxide anion generation. IL-8 release, a potent chemotactic factor released by IEC, and neutrophil chemotaxis data are shown in Table 2. IL-8 was increased 8! versus control after IEC exposure to toxin A (without mucus). There was roughly a 50% reduction in IL-8 release with mucus versus non–mucusproducing IEC after exposure to toxin A. Supplemental PC and the native mucus layer decreased IL8 production in a similar fashion. Neutrophil chemotaxis was significantly greater in response to culture supernatants from the nonmucus IEC after exposure to toxin A. Either the native mucus layer or supplemented PC reduced PMN chemotaxis to at or near
496
The American Journal of Surgery, Vol 209, No 3, March 2015
Table 1
PMN activation quantified after exposure to HT29 and HT29-MTX supernatants cocultured with exogenous PC and/or toxin A Elastase (%)
PMN PMN PMN PMN PMN PMN PMN PMN
alone 1 PC (cell free) 1 HT29 supernatant 1 HT29 1 toxin A supernatant 1 HT29 1 PC 1 toxin A supernatant 1 HT29-MTX supernatant 1 HT29-MTX 1 toxin A supernatant 1 HT29-MTX 1 PC 1 toxin A supernatant
8.3 8.0 9.8 27.6 10.4 10.2 15.7 11.6
6 6 6 6 6 6 6 6
Superoxide anion (Nmol/well)
.8 1.0 1.2 2.3* .7†,‡ 1.1 1.6*,† 1.3
6.8 6.6 10.0 32.4 13.5 10.7 19.3 16.1
6 6 6 6 6 6 6 6
CD11b (MFI)
.6 1.1 2.2 3.4* 2.1†,‡ 1.3 2.0*,† 1.4
99.5 91.3 108.2 268.4 115.6 111.4 128.9 120.5
6 6 6 6 6 6 6 6
2.5 4.4 4.3 7.9* 5.6†,‡ 3.5 4.8*,† 3.5*,†
Results are expressed as mean 6 SD, n 5 5 for each group. PC 5 phosphatidylcholine; PMN 5 polymorphonuclear neutrophil; SD 5 standard deviation. *P , .001 vs HT29 supernatant. † P , .001 vs HT29 1 toxin A. ‡ P , .001 vs HT29-MTX 1 toxin A.
baseline values. In experiments using C. difficile toxin B, PMN activation and chemotaxis results were similar to those found with exposure using toxin A (results not shown).
animal and clinical studies with C. difficile infection show significant neutrophil infiltration that varies with disease severity.10,11 In addition, systemic neutrophil activation is evidenced by findings of a significantly elevated peripheral blood neutrophil count and bandemia. Although neutrophil infiltration is beneficial during intestinal inflammation, it may be pathological if prolonged or exaggerated. A previous study has indicated that toxin A directly stimulated human neutrophils at relatively high concentrations and toxin B had no direct stimulatory effect at all.11 Thus, the mechanism for neutrophil activation by C. difficile toxin in vivo remains unclear. In this prior study, Linevsky et al11 demonstrated that human monocytes release significant amounts of the neutrophil chemoattractant IL8 following C. difficile toxin A or B exposure. Subsequent monocyte activation was then found to initiate neutrophil activation in vitro likely because of the production of proinflammatory cytokines. IECs are an important part of the innate immune barrier defense and secrete a variety of factors that ‘‘orchestrate’’ the immunoinflammatory responses at mucosal surfaces.7 In our previous studies, we have shown that HT29 colonic epithelial cells produce tumor necrosis factor alpha and
Comments Clostridium difficile is an increasingly important hospital-associated pathogen. Its clinical spectrum ranges from asymptomatic colonization to fulminant colitis and death. The variable clinical manifestations likely reflect C. difficile virulence factors as well as the immune and inflammatory response of the host.1,2 The major virulence factors include toxin A and toxin B, a binary toxin, and surface layer proteins.8 The importance of toxin levels produced by the organism has been noted with the emergence of the hypervirulent B1/NAP1/027 strain. However, in other studies markers of intestinal inflammation rather than bacterial factors have been found to correlate better with clinical outcome in patients with CDAD.2,8 Neutrophil recruitment is an essential component in the pathogenesis of C. difficile toxin-induced injury.9 Both
Table 2 Basal cytokine expression and PMN chemotaxis quantified in HT29 and HT29-MTX supernatants collected after cell monolayer exposure to exogenous PC and/or C. difficile toxin A Basal IL-8 (pg/mL) HT29 control HT29 1 toxin A supernatant HT29 1 PC 1 toxin A supernatant HT29-MTX supernatant HT29-MTX 1 toxin A supernatant HT29-MTX 1 PC 1 toxin A supernatant
3.9 31.9 6.8 4.8 16.6 4.4
6 6 6 6 6 6
.3 1.9* 1.1*,† .8 1.0*,† 1.1†,‡
Results are expressed as mean 6 SD, n 5 5 for each group. PC 5 phosphatidylcholine; PMN 5 polymorphonuclear neutrophil; SD 5 standard deviation. *P , .001 vs HT29. † P , .001 vs HT29 1 toxin A. ‡ P , .001 vs HT29 1 toxin A 1 PC. x P , .001 vs HT29-MTX 1 toxin A.
Chemotaxis (cell count) 2,625 6,250 3,500 2,525 3,150 2,250
6 6 6 6 6 6
120 202 150* 105 100*,† 82†,x
A. Olson et al.
Phosphatidylcholine and the intestinal mucus layer
IL-6 in response to C. difficile toxin and this response was modulated by an intact mucus layer or by the addition of exogenous PC.3,6 Exogenous PC was effective in this study likely as ‘‘reinforcement’’ to the mucus layer but also by direct effect of PC on proinflammatory signaling in IECs. In our preparation, there is no PC in the culture media; thus the results reflect only exogenous sources. These levels may vary in the clinical scenario. In this study, we demonstrate that exogenous PC downregulates neutrophil activation and trafficking after exposure to culture supernatants from IEC exposure to C. difficile toxin. This is likely because of a reduction in proinflammatory signaling from IEC cells treated with PC and exposed to C. difficile toxin. There is increasing evidence that exogenous PC is helpful in resolution of inflammation in patients with CUC. Studies in patients with ulcerative colitis have shown that a modified PC preparation is necessary to achieve effective release of PC in distal ileum.5 This delayed release formulation avoids early intestinal absorption of the supplemented PC, thus allowing effective delivery into the colon. Current PC formulas may even be efficacious when instilled locally via a loop ileostomy in patients treated with a colonic salvage therapy as proposed by Neal et al.12 We believe that this study supports our ongoing hypothesis that ‘‘reinforcement’’ of the mucus layer may be a useful adjunct in patients with severe forms of CDI. However, further studies in animal models of CDAD are warranted.
References 1. Vedantam G, Clark A, Chu M, et al. Clostridium difficile infectiond toxins and non-toxin virulence factors, and their contributions to disease establishment and host response. Gut Microbes 2012;3: 121–34. 2. Solomon K, Martin AJ, O’Donoghue C, et al. Mortality in patients with Clostridium difficile infection correlates with host proinflammatory and humoral immune responses. J Med Microbiol 2013;62:1453–60. 3. Olson A, Diebel LN, Liberati DM. Effect of host defenses on Clostridium difficile toxin-induced intestinal barrier injury. J Trauma Acute Care Surg 2013;74:983–9. 4. Ehehalt R, Braun A, Karner M, et al. Phosphatidylcholine as a constituent in the colonic mucosal barrierdphysiological and clinical relevance. Biochim Biophys Acta 2010;1801:983–93. 5. Karner M, Kocjan A, Stein J, et al. First multicenter study of modified release phosphatidylcholine ‘‘LT-02’’ in ulcerative colitis: a randomized, placebo-controlled trial in mesalazine-refractory courses. Am J Gastroenterol 2014;109:1041–51. 6. Olson A, Diebel LN, Liberati DM. Exogenous phosphatidylcholine supplementation improves intestinal barrier defense against Clos-
7. 8.
9. 10. 11.
12.
497
tridium difficile toxin. J Trauma Acute Care Surg 2014;77:570–5; discussion, 576. Kinnebrew MA, Pamer EG. Innate immune signaling in defense against intestinal microbes. Immunol Rev 2012;245:113–31. El Feghaly RE, Stauber JL, Deych E, et al. Markers of intestinal inflammation, not bacterial burden, correlate with clinical outcomes in Clostridium difficile infection. Clin Infect Dis 2013;56:1713–21. Fournier BM, Parkos CA. The role of neutrophils during intestinal inflammation. Mucosal Immunol 2012;5:354–66. Price AB, Davies DR. Pseudomembranous colitis. J Clin Path 1977;30:1–12. Linevsky JK, Pothoulakis C, Keates S, et al. IL-8 release and neutrophil activation by Clostridium difficile toxin-exposed human monocytes. Am J Physiol 1997;273:G1333–40. Neal MD, Alverdy JC, Simmons RL, et al. Diverting loop ileostomy and colonic lavage: an alternative to total abdominal colectomy for the treatment of severe complicated Clostridium difficile associated disease. Ann Surg 2011;254:423–7.
Discussion Dr Anthony J. Senagore (Saginaw, MI). Do you have any data assessing PMN activation status due to C. diff toxin without the presence of the colonic cells, ie, is this directly antiinflammatory on the PMN’s or is it actually mucosal enhancement? Second, do you have any clinical model data demonstrating either reduced inflammatory infiltrates into the bowel wall as a result of C. diff toxin and the enhancement of the mucosal barrier? Third question, do you have any data on glutathione levels? Because Phosphatidylcholine is a precursor molecule for glutathione, and that may be the active agent. And, finally, do you have any data on enhancement of the mucosal layer in other models that increase the risk for C. diff infection, like chronic PPI exposure? Dr Olson: Toxin A actually is able to activate PMN’s but that’s at super physiologic levels. Toxin B has not been shown to activate PMN’s on its own at all. Secondly, do we have any animal model data or any other clinical models describing this reduced inflammatory cell infiltrate? And that’s where we are proceeding in our next steps. We don’t have an animal model that we work with at the current time. And, number three, do you have any data on glutathione as a mechanism? Many of the injuries that are seen are oxidant based, so that may be where all this is coming from. Finally, we do have data from our laboratory demonstrating that H2 blockers decrease intestinal epithelial cell mucus layers. It has been shown that patients using PPI therapy are more at risk for C. diff activation, but this is because of direct cellular effects. It’s not directly related to the mucus.
