Leukocyte-Endothelial Cell Interactions Evoked by Mast Cells

Cardiovascular Research 32 (1996) 699-708

Review Leukocyte-endothelial cell interactions evoked by mast cells

Paul Kubes a,* , D. Neil Granger ’ Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 a Immunology Research Group, University of Calgag Medicul Center, Culgaq, Alberta RN 4N1, Cmrud~~ ’ Department qf Physiology, Louisiana State University Medical Center. Shre~~epon. LA, USA

Received 12 December 1995; accepted 20 May 1996

Abstract

In this review we have summarized some of the evidence to support the view that mast cells play a critical role in leukocyte recruitment to sites of infktmmation. Initially, data using a pharmacological tool, compound 48/80, which directly activates mast cells, is reviewed, demonstratingthat this reagentcan induce the multi-step recruitment of leukocytes (rolling, adhesiotrand emi~tionl to sites of inflammation. The adhesive mechanisms and pro-inflammatory mediators implicated in mast cell-induced leukocyte recruitment are discussed.Additionally, data are presentedto implicate mast cells in delayed-type hypersensitivity reactions as they pertain to leukocyte recruitment. There is a growing body of evidence to suggest that mast cells also recruit leukocytes in IgE-indepandent leukocyte recruitment. Ischemia/reperfusion- and bacterial toxin- (Helicobacter pylori and Clostridium dificile) induced leukocyte recruitment is at least in part mast cell dependent.Future directions including preliminary work highlighting the role of nitric oxide as a modulator of mast cell function and subsequentleukocyte recruitment is also discussed.

Keywrds: Integrins: Selectins; Microcirculation; Neutrophils; Adhesion

1. htrduction adhere to the endothelium and ultimately emigrate out of the vasculature. This event is mediated by the integrins The recruitment of leukocytes from the mainstream of found on leukocytes and in the case of neutrophils the blood to the extravascular space is a key feature of the &-integrin (CD 11/CD 18). inflammatory response. It has been well established that There are three important assumptions made when con- leukocyte infiltration is a multi-step mechanism (Fig. 1) sidering this scheme for leukocyte recruitment. First this is which requires that leukocytes moving at very high speeds an interrelated cascade of events and therefore the rolling in the mainstream of biood make initial contact with the event is a necessary prerequisite to leukocyte adhesion and endothelial cells lining the vessel wall and roll along the subsequent emigration. Secondly, the rolling, adhesion and vessel at a greatly reduced velocity relative to flowing red emigration transpire primarily within the pastcapillary blood cells [ 1,2]. This initial leukocyte-endothelial cell venules and very rarely do investigators observe leuko- interaction is termed leukocyte rolling and is entirely cyte-endothelial cell interactions in other vessels. A final dependent upon the selectin family of adhesion molecules point that should be made is that histology has established [3,4], L-selectin is constitutively expressed on the surface the phenotype of adhering and emigrating cells as neu- of leukocytes and appears to be essential for the ability of trophils in the early event (first 8 h) and monocytes and these cells to initiate rolling in the microcirculation. Two lymphocytes at later time points (longer than 8 h). Al- other selectins, P-selectin (induced in minutes) and E- though it is tempting to conclude that the rolling cells selectin (3-6 h for maximal induction) expressed on acti- follow the same time frame as the adhering and emigrating vated em&helium also contribute significantly to the cells, to date it has been impossible to determine the rolling event. When activated, the rolling leukocytes firmly phenotype of the rolling population(s) of leukocytes.

* Corresponding author. Tel.: (+ l-403) 220-8558: fax (+ l-403) 283. 3028 Time for primary review 17 days.

Mast Cell Table 7 Activators of maht cells

IgE and specific antigen (via FceRI) complrmrilr c’omporlerlt,s C3a C5a Neuropeptides Substance P Vasoactive intestinal peptide Somatostatin Fig. 1. Leukocyte (in this example neutrophil) recruitment is a multi-step Calcitonin gene-related peptide event that is initiated by the selectina and manifested as a rolling Oxidarl ts Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 interaction. This event is a prerequisite to firm adhesion and emigration Superoxide via the &-integrins. The expression and/or activation of these adhesion Lipids molecules occurs via pro-inflammatory mediators released from the local Platelet-activating factor environment. The mast cell found in close apposition to the vessels is a Leukotrienes LTB,, LTC, likely source of these molecules. Prostaglandin D? E.~ogenous srcretugogues Bacterial toxins Therefore, the rather vague term “leukocyte” will be used Compound 48/X0 throughout this review. Calcium ionophore A23 I87 To date, a tremendous amount of literature has been Other,v Histamine, stem cell factor, bradykinin published on the adhesive mechanisms underlying leukocyte recruitment and the identity of a signal or multiple signals that are released directly from the injured site to initiate the multi-step recruitment. However, far less is In light of the tremendous interest in cytokines and other known about the source of these mediators. Mast cells are pro-inflammatory molecules in disease, it is somewhat found closely apposed to the vasculature in essentially all surprising that more attention has not been focused on the tissues and are exquisitely sensitive to very subtle changes role of mast cells as contributors to cardiovascular disease. in the surrounding milieu. Once activated, they are capable Table 2 summarizes some stimuli that have been docu- of releasing a myriad of inflammatory mediators and there- mented to activate mast cells (see reviews for more detail fore are certainly candidates as a first-line detector system [5-81). Although to date mast cell activation has been and initiator of the inflammatory response (Fig. 1). Mast associated with IgE activation, there is a growing body of cells release preformed cytokines, histamine, serotonin as evidence to suggest that mast cells are activated via oxi- well as newly synthesized platelet-activating factor, leuko- dants, anaphylatoxins, neuropeptides and bacterial prod- trienes and many other mediators (see reviews for detail ucts and are also exquisitely sensitive to physical stress, [5-81). Noteworthy for example is the fact that mast cells temperature changes and other physical disturbances [7,10]. prepackage TNFG in granules where it is readily available Clearly, increased levels of oxidants, anaphylatoxins, bac- for release [9]. Table 1 summarizes some of the mast terial toxins and neuropeptides in ischemia/reperfusion, cell-derived mediators that could impact directly upon the sepsis and anaphylaxis may affect the activity state of mast vasculature to either express endothelial adhesion cells and thereby mediate one of the key features of the molecules, activate leukocytes and/or induce both events. pathology associated with cardiovascular disease: leukocyte adhesion to endothelium.

Table 1 Mast cells release numerous classes of inflammatory mediators 2. Mast cells recruit leukocytes Vusoactice Adenosine, nitric oxide, prostaglandins Amines The fact that mast cells can indeed affect the vascula- Histamine, serotonin ture to recruit circulating leukocytes to sites of inflamma- Cytokines tion has been known for many years. Initial work [I l] IL-I-IL-6, IL-8, TNFcu, MIP-I [Y, MIP-1 p, IFNy, and others supporting the view that mast cell degranulation was suffi- Lipid mediators Leukotrienes, thromboxanes, platelet activating factor cient to induce leukocyte infiltration made use of the mast Enzymes/proteases cell degranulating agent, compound 48/80 (CMP 48/80). Chrymase/tryptase, cathespin G, p-glucoxidase, kinogenase, and others This agent, when injected into the skin of mice induced a Others rapid (i.e., within l-2 h) neutrophilic influx into skin. This Proteoglycans, oxidants was followed at 24 h by a mononuclear infiltrate suggest- This is not an exhaustive list. Numerous classes are highlighted and some ing that mast cell degranulation had the capacity to recruit examples are given. at least two types of leukocytes. Injection of mast cell granules into mouse skin elicited a similar cellular infil- dependent leukocyte rolling [13]. However. Thorlacius et trate as CMP 48/80, suggesting albeit indirectly that al. [ 171 reported that the CMP 48/80-induced, selectin-de- CMP 48/80 was functioning to degranulate mast cells to pendent leukocyte rolling was not inhibitable by a different recruit leukocytes. HI -receptor antagonist (mepyramine), in combination with Visualization of leukocyte migration out of the mi- an HZ-receptor antagonist (cimetidinel. This discrepancy crovasculature using intravital microscopy (see Ref. [ 121 remains unresolved: however, a recent study has demon- for review) has further elucidated the actions of CMP strated that mast cell-dependent leukocyte recruitment was 48/80 and revealed that mast cells can indeed contribute inhibitable with a third H ,-receptor antagonist (hydroxy- to the multi-step recruitment of leukocytes (rolling, adhe- zinc) in a Chstridium di&iciZe toxin A model of inflamma- sion and emigration) [13]. In addition to being able to tion (discussed later). Whether mepyramine has a lowet visualize the multi-step recruitment of leukocytes, one can efficacy than the other H ,-receptor antagonists in the rat Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 use an intravital stain (ruthenium red) that continuously vasculature remains unknown. Inhibitors of‘ other media- detects the degree of mast cell activation throughout the tors w:ere unremarkable in their ability to prevent the rise experiment [ 13,141. IJsing this technique it has been in mast cell-induced leukocyte rolling: inhibitors of leuko- demonstrated that CMP 48/80 caused rapid activation of trienes, PAF or serotonin did not reduce the increased mast ceils which preceded the induction of leukocyte-en- leukocyte rolling [ 13.253. dothelial cell interactions by approx. 1.5 min [ 131. Stabi- Further evidence to demonstrate that mast cells could lization of the mast cells with sodium cromoglycate or induce leukocyte recruitment was based on the observation ketotifen reduced both mast cell activation and leukocyte that immediately upon exteriorization of tis5uc for intravi- recruitment. Finally, chronically depleting mast cells of tal microscopy. there was a mast cell-dependent induction their granules in vivo prevented the subsequent recruitment of leukocyte rolling [21]. Sodium cromoglycate. a potent of leukocytes with CMP 48/80. These data as a whole inhibitor of histamine release from rat connective tissue strongly support the view that CMP 48/80-induced leuko- mast cells [26]. given prior to surgical manipulation (but cyte recruitment i5 mast cell dependent. not after) attenuated by 80% baseline leukocyte rolling [2 11. Moreover, diphenhydramine. the H ,-receptor antago- 2.1. Mast ~11s mediute leukocyte rolling nist. also significantly ( > .50%‘0)reduced baseline leukocyte rolling. These data collectively suggested that mast cell-de- Applying intravital microscopy to the rat mesentery, we rived histamine (and other unidentified agents), promoted [ 131 and others [ 15- 171 assessed whether mast cells could leukocyte rolling during surgical manipulation. The con contribute to both the initial rolling movement along the cept of surgically induced rolling is consistent with the length of the vessel wall and subsequent firm adhesion. observations that leukocyte rolling is very low in normal Superfusion of the rat mesentery with CMP 48/80 re- tissues including the intact bat wing ]27] arid that leuko- vealed a rapid increase in the number of rolling leukocytes cyte rolling increases rapidly at the time tot surgery [2X]. (relative to untreated preparations) that persisted for the Moreover, the fact that baseline roliing could be prevented next 60 min. The selectin responsible for the initial rapid by immunoneutralizing P-selectin [29] with %elective anti- rolling event along the length of vessels [ 1.21 is P-selectin, bodies or inhibiting both L-selectin and P-selectin with mobilized to the endothelial plasma membrane from fucose polymers such as fucoidin [30] raises the possibility Weibel-Palade bodies. P-Selectin supports leukocyte adhe- that the preparation-induced mast ceil activation poten- sion to monolayers of endothelium in static assay systems tially leads to P-selectin expression and increased baseline [ 181 and leukocyte rolling under shear conditions on P- rolling. The proximity of mast cells to the vasculature selectin-containing artificial membranes [ 191. In vivo data [16.31] in conjunction with their intense sensitivity to also support a role for P-selectin in the induction of physical stress, temperature changes and other physical leukocyte rolling inasmuch as exogenously administered disturbances that are likely to occur during surgical prepa- superoxide initiated leukocyte rolling, an event entirely ration [7,10] make mast cells a good candidate to mediate inhibited by a P-selectin antibody 1201. The degranulation baseline alterations in leukocyte-endothelinl cell interac- of perivascular mast cells was also sufficient to invoke tiona. leukocyte rolling [ 13,171, and the P-selectin MAb, PB 1.3, Animals pretreated with cromolyn (must cell-stabilized completely inhibited this event, supporting the view that preparations) were sensitive to histamme and responded to substances released from mast cells were capable of induc- this amine with prolonged leukocyte rolling that was en- ing endothelial P-selectin-associated leukocyte rolling. tirely inhibitable by either diphenhydramine or a P-selectin P-selectin can be mobilized to the endothelial cell sur- antibody [2 I]. In animals that had not been mast ccll-stabi- face by many different pro-inflammatory mediators includ- lized, there was little response to histamine It wa5 hypoth- ing histamine, LTC,, thrombin and oxidants [ 18,2 l-241. esized that histamine released from m;t~r cells during In our laboratory the CMP 48/80-induced leukocyte rolling surgical manipulation caused the vasculaturc to be tachy- appeared to be histamine dependent inasmuch as diphenhy- phylactic to subsequent histamine treatment. Very similar dramine (H ,-receptor antagonist) inhibited the mast cell- data were simultaneously published 1)~ +I\+iho et al. (321. who circumvented the problem of surgically induced mast cell degranulation by choosing those vessels wherein initial baseline leukocyte rolling was extremely low. Animals with a higher leukocyte rolling again failed to respond to histamine. Collectively, these data are consistent with the view that rat connective tissue mast cells are capable of recruiting rolling leukocytes via histamine. The mast cell- stabilized preparation revealed that other mediators including LTC, [33] and H,O, [34] but not PAF [21] or sero- g- _ CON anti- anti- DPH WEB tonin [25] induced leukocyte rolling in vivo. It should be d CD18 P-se1 2086 noted that it was impossible to use the mast cell-stabilized + COMPOUND 48/80 model to study (CMP 48/80-induced) mast cell-induced Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 leukocyte rolling and therefore the increase in CMP 48/80-induced leukocyte rolling was relatively subtle in TO HUVEC TO PLASTIC magnitude, but significant when compared to untreated I I I control animals [ 131.

2.2. Mast cells cause leukocyte adhesion

The second step in the inflammatory cascade is firm adhesion of leukocytes to the venular endothelium, a mechanism dependent on the expression of the &integrin

also termed CD1 l/CD18 [35,36]. CMP 48/80-induced MAST CELLS - + - + + + - + + mast cell degranulation elicited a significant increase in CMP 48/80 - - + + + + + + leukocyte adhesion (Fig. 2a). This adhesion was signifi- Fig. 2. The top panel demonstrates that in single postcapiliary venules cantly attenuated by CL26, a monoclonal antibody to compound 38/80 (CMP 48/80)-induced leukocyte adhesion that is CD18, as well as by the aforementioned reagents that inhibitable by an anti-CD18 antibody (CL26), an anti-P-selectin antibody prevent leukocyte rolling (P-selectin antibody and diphen- (PB i .3), an H, -receptor antagonist (diphenhydramine) or a PAF receptor hydramine) [ 13,171, observations consistent with the con- antagonist (WEB 2086). The lower panel demonstrates that CMP 48/80 alone or mast cells alone do not affect neutrophil adhesion to human cept that rolling is a necessary prerequisite for adhesion. umbilical vein endothelium but together a profound adhesion is noted that CL26 had no effect on the flux of rolling leukocytes, is inhibited by an anti-CD18 antibody (IB,) or WEB 2086. Mast cells suggesting that unlike adhesion which was entirely depen- also induced adhesion to fetal calf serum-coated plastic (right side of dent on the recruitment of rolling leukocytes, leukocyte bottom panel). Please see Ref. [ 141 for details. S P < 0.05 relative to rolling occurred independently of CD18 and the adhesive control. process. Fig. 2b demonstrates that adhesion of leukocytes to endothelium could also be induced by degranulating mast cells in vitro. Addition of CMP 48/80 alone or mast sequent synthesis of leukotrienes [37,38], did not affect the cells alone to a monolayer of endothelium or protein-coated adhesive response to CMP 48/80 [13]. By contrast, WEB plastic did not promote neutrophil adhesion to either sub- 2086, a PAF receptor antagonist, completely blocked the stratum. However, co-administration of CMP 48/80 and number of adherent leukocytes associated with CMP 48/80 mast cells induced neutrophil adhesion to both endothe- superfusion of the rat mesentery (Fig. 2a). The in vitro lium and protein-coated wells. The adhesion was entirely data confirmed the in vivo results; CMP 48/80-stimulated inhibited by anti-CD18 antibody. These in vitro data ad- mast cells induced leukocyte adhesion to both endothelial dress two important issues: (1) CMP 48/80 does not cells and protein-coated plastic, an event inhibitable by induce neutrophil-endothelial cell interactions directly (re- WEB 2086 (Fig. 2b). These data suggest that mast cell-de- quires mast cells) and (2) the endothelium is not necessary rived PAF contributes significantly to subsequent neu- for this adhesion as neutrophils can still be induced to trophil-endothelial cell adhesion, consistent with the fact adhere via mast cells to protein-coated plastic. that PAF is produced and released in significant quantities CD1 l/CD18 is constitutively expressed on the surface by mast cells. This is in part evidenced by the fact that the of leukocytes in an inactive state but can be rapidly intestine synthesizes 70% less PAF in mast cell-deficient activated in the presence of various chemotactic stimuli mice than in normal littermates [39]. Although these in including PAF and LTB, to induce and support adhesion vitro data do not exclude the possible contribution of [35,36]. It appears that synthesis of leukotrienes is not endothelial-derived PAF to mast cell-induced adhesion in critical to the induction of leukocyte adhesion observed vivo, they do suggest that activated mast cells can indeed with mast cell degranulation inasmuch as MK886, which induce leukocyte adhesion through the release of this blocks membrane translocation of Slipoxygenase and sub- phospholipid. P. K&es, N.D. Grcrnger/ Cardicwasculnr Reserirch 32 i I9961 6YP--708 703

2.2. Mast ceils induce leukocyte emigration The third selectin, E-selectin, may also induce leukocyte rolling [43,44] in inflamed vessels. However, it is It is interesting to note that in our study with CMP unlikely that E-selectin plays a role in the first 60 min of 48/80 we saw only a subtle increase in leukocyte emigra- reperfusion inasmuch as induction of this molecule on the tion at 30-60 min of CMP 48/80 superfusion [13]. The surface of endothelium is thought to require at least 4 h of lack of emigration of adherent ceils may be related to the reperfusion. Moreover, only a very small number of ves- fact that the emigration may require longer periods of time sels demonstrated E-selectin expression ( 15-200/c of all than 60 min. Indeed, a leukocytic infiltrate was first noted vessels) in postischemic coronary vasculature 1451.Finally, at 60- 120 min following injection of CMP 48/80 into the the leukocyte roiling in the feline intravimi microscopy skin of mice and reached peak values only at 8 h. Alterna- study was abolished by fucoidin. yet fucoidin has been tively. Raud et al. [15] demonstrated that superfusion of shown not to bind to E-seiectin [3,46]. Fucoidin, a heavily Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 the hamster cheek pouch with CMP 48/80 caused leuko- sulfated poiysaccharide. may interfere with the ability of cytes to leave the vasculature within the first hour and leukocytes to interact with other potential adhesive mecha- migrate towards degranulating mast ceils. In this particular nisms. For example, fucoidin may compete for binding tissue mast cells localized around arterioles and leukocytes sites with sulfate-containing proteogiycans on the surface emigrated out of venules in response to CMP 48/80 and of vascular endothelium. Although the possibility that oriented towards the periarteriolar mast ceils. The pattern these molecules contribute to rolling is purely speculative. of mast ceil localization around arterioles in the hamster various other sulfated molecules have been shown to cheekpouch, however, does not appear to be characteristic strongly interfere with leukocyte rolling in viva 147-491 of all tissues; mast cells appear to be randomly distributed and leukocyte adhesion in vitro [50]. throughout the rat mesentery. This may explain the differ- Acute treatment with a monocionai antibody (MoAb ence in leukocyte emigration between different tissues. IB,), which immunoneutralized the P-subunit (CD181 of the CD I 1/CD 18 adhesion giycoprotein complex, pre-- vented reperfusion-induced increase in myeioperoxidase 3. Mast cells contribute to leukocyte recruitment in levels in numerous tissues. implicating a CD i g-dependent inflammation mechanism of neutrophil recruitment during l/R [5 i-531. 3.1. Isdwmirr / rep$u.sion (I / RI Additionally, MoAb IB, completely prevented the reperfusion-induced increase in neutrophil adhesion as visualized There has been a tremendous amount of work com- by intravital microscopy [54]. These results suggested that pleted on the adhesion molecules involved in the recruit- the adhesion was entirely mediated by chr CD18 giyco- ment of leukocytes into postischemic vessels. Following protein complex. Although neutrophii emigration was also the induction of ischemia/reperfusion in for example the completely prevented, whether the emigration was depen- cat mesentery. a very dramatic increase in neutrophil dent on CD18 or simply on the inability of the neutrophiis rolling. adhesion and ultimately neutrophil emigration into to adequately adhere remains unclear. A clinically relevant the tissue is observed ]40,41]. Using intravital microscopy, observation is that administration of MoAB lB, at 60 min it was observed that the L- and P-selectin-binding carbo- of reperfusion almost immediately reversed the adhesion hydrate, fucoidin, essentially abolished ( > 90%) leukocyte of neutrophils to postcapillary venuies [55]. Clearly the roiling in postischemic vessels [41]. Antibodies against neutrophii-endothelial ceil interaction could be interrupted. either P-selectin (PB1.3) or L-selectin (DREG 200) re- Similar results have been observed in the rat mcsentery duced the number of rolling leukocytes by approx. 60% at and furthered by the availability of antibodies against other both 10 and 60 min of reperfusion. Leukocyte rolling was adhesion molecules [56]. Antibodies against either CD18 not decreased further in animals given both anti-L-selectin (MoAb CL26I. the CD1 1b subunit of CD 1 I /CD 18 adhe- and anti-P-selectin antibody; a 60% reduction in rolling sion complex (MoAb 1B6c) or against endnthelial ICAM- 1 was still observed. The tack of additive effect of tandem (MoAb lA29), a Ii&and for CD1 8. also significantly re- antibody therapy suggests either that P-selectin and L- duced reperfusion-induced neutrophii adhesion and emi- selectin pathways worked in concert, i.e., as counter- gration. The ICAM-I data were recently confirmed in the ligands. or that they mediate different components of cat model of ischemia/reperfusion using the monoctonai leukocyte roiling in a sequential manner so that one is antibody RR 1/ I [57]. dependent upon the other 1421. Moreover, based on the L- There is a growing body of evidence IO suggest that and P-selectin antibody studies and the fucoidin study, mast cells may be involved in the multi-step leukocyte these data may suggest that there exists an L- and P-selec- recruitment associated with I/R. First. there is evidence to tin-independent fucoidin-inhibitable roiling pathway dur- suggest that mast ceils become activated during ing reperfusion. Although each of the reagents was used at ischemia/reperfusion. Boros et al. [5X] have reported in- optimal concentrations it is also possible that fucoidin creased histamine release from the postischemic intestine. simply blocks P-selectin and L-selectin more effectively Although the cellular source of the histamine was not than the antibodies. identified. mast ceils are a primary source of this pro-inflammatory mediator in the small bowel [7]. Although 0 UNTREATED m CROMOLYN histamine is not thought to be a chemotactic molecule per * se. it may participate in the recruitment of leukocytes by causing the expression of endothelial P-selectin [ 18.2 I]. To the best of our knowledge, the role of anti-histamines on I/R-induced leukocyte recruitment has not been docu- mented. Although histamine could be released by numerous other cells, only rat mucosal mast cells release a CON ISC 5 10 60 protease termed rat mast cell protease II (RMCPII). In- REPERFUSION(min) E creased plasma protease II levels (an index of mast cell 1 i ;; - degranulation) have been reported following intestinal I/R [59]. These data more specifically support the contention Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 q15- that mast cells degranulate in postischemic tissues. Re- it 9 I lo- cently, Kurose et al. [60] have visually assessed mast cell integrity at 30 min reperfusion of postischemic mesentery and noted a significant increase in degranulated mast cells CON ISC 10 60 that coincide closely with leukocyte adhesion. However, B 0.6 - REPERFUSION(min) whether the mast cell degranulation was a cause or effect of I/R-induced leukocyte recruitment was unclear. 1 0.5 - To directly address the issue of whether the degranu- B lated mast cells contributed to leukocyte recruitment in El 8 1:: - I/R. mast cell stabilizers that prevented the rise in pro- gd 0.2 - t tease II levels were used. In animals pretreated with these % 0.1 - reagents, a reduction in myeloperoxidase activity (index of B 0.0 UNTREATED CROYOLYN neutrophil infiltration) was noted in the postischemic intes- REPERFUSION (60 min) tine [59]. To determine which phase of the leukocyte recruitment was affected by mast cells in I/R, intravital Fig. 3. Reperfusion of the cat mesentry induces leukocyte rolling (top panel), adhesion (middle panel) and emigration (lower panel). Each of microscopy was performed and leukocyte rolling, adhesion these parameters was at least in part attenuated by the mast cell stabilizer and emigration were assessed in the cat postischemic sodium cromoglycate. Emigration was measured as the number of cells mesenteric microvasculature [40]. Fig. 3 demonstrates that that appeared in the extravascular space over the 60 min reperfusion between I and 5 min of reperfusion, there was a very period. See Ref. [42] for details. _ P < 0.05 relative to control. ’ P < 0.05 dramatic rise in leukocyte rolling which persisted for the relative to untreated group. next 60 min. In cromolyn pretreated animals, the flux of rolling leukocytes increased in the very early (5 min) reperfusion period but by 10 min the number of rolling in mast cell degranulation. Although the source of the leukocytes was significantly reduced, an event that per- oxidants remains unknown, the fact that Boros et al. [62] sisted for the next 60 min. Cromolyn also reduced PMN demonstrated that allopurinol blocked by 87% the his- adhesion by 50% but almost entirely inhibited leukocyte tamine release from postischemic gut suggests an impor- emigration out of postcapillary venules. These data suggest tant role for the oxidant-generating enzyme xanthine oxi- an important contribution of interstitial mast cells as medi- dase in I/R-induced mast cell activation. ators of the multi-step recruitment of PMNs from blood to Another mediator known to activate mast cells is ana- postischemic tissues and raise the possibility that mast phylatoxin (C3a and C5a). In fact in the heart it has been cells may set up a chemotactic gradient to draw leukocytes shown that anaphylatoxins cause histamine release from out of the vasculature. cardiac mast cells [63,64]. Since C3a and C5a have been If mast cell degranulation causes leukocyte infiltration shown to play an important role in ischemia/reperfusion into postischemic tissue, then some factor(s) must be acti- of the heart [65], by inference it is tempting to speculate vating the mast cells. The increased flux of oxidants that these anaphylatoxins produced at the time of reperfu- previously described at the onset of reperfusion [61] may sion can in addition to their direct effects on the vascula- be responsible for mast cell activation inasmuch as super- ture also stimulate mast cells to release histamine and oxide is known to activate these cells [3l]. In order to test other pro-inflammatory mediators, thereby potentially con- the hypothesis that oxidants are involved in I/R-induced tributing to leukocyte recruitment and reperfusion injury. mast cell degranulation, some animals were pretreated with Direct evidence to support this view is presently not superoxide dismutase and catalase [59]. This protocol pre- available. Additionally, human mast cells release a number vented the release of protease II from mast cells and the of neutral proteases including tryptase which have the subsequent leukocyte infiltration, suggesting that indeed capacity to generate C3a anaphylatoxin from human C3 in superoxide and/or hydrogen peroxide were instrumental vitro [66]. This can conceivably function as an amplifica- tion system in numerous inflammatory conditions includ- tation of impaired leukocyte influx still awaits confirma- ing ischemia/reperfusion. tion. In similar experiments, Subramaniam et al. [72] rc- Finally, Goldman et al. [67] have reported that mast ported that oxazolone sensitized animals challenged with cells may play a role in infiltration of neutrophils into the oxazolone showed a significant mononuclear cell infiltrate lung and subsequent pulmonary injury associated with at 24 h that was reduced by 53.8% in mutant mice lacking remote (hind limb) I/R. Mast cell-deficient mice showed P-selectin. Interestingly, late phase responses were not an attenuation of sequestered neutrophils and a reduction inhibited in normal mice treated with a P-selectin antibody in lung edema in this model. Clearly, the role of mast cells [73]. Meanwhile, mutant mice lacking E-selectin had no may be far-reaching if products released from a postis- leukocytic impairment in late phase responses: however, chemic site entered the systemic circulation and activated addition of a P-selectin antibody significantly reduced mast cells at a distant site where these cells caused distant neutrophil infiltrate in this model [73]. These data raise a Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 organ injury. Alternatively. mast cells at the postischemic role for selectins in the recruitment of leukocytes in de- site could release factors (histamine, serotonin, cytokines. layed-type hypersensitivity reactions. However. it is likely platelet-activating factor as well as proteases that generate that some redundancy in these chronic models may exist. anaphylatoxins) that would affect distant organs. Regard- i.e., leukocyte recruitment may be dependent upon more less of the site of activated mast cells, the importance of than one selectin. Whether mast cell-derived mediators the observation that mast cell-deficient animals have re- contribute directly to inducing expression of endothelial duced distant organ injury may be directly related to the selectins in these models is an important. :15 yet undeter- pathogenesis of multiple organ failure. mined issue.

3. I. I. IKE-mediated hypersensitiLi5 reactions It is well established that mast cells are important 4. Bacterial toxins mediate leukocyte adhesion through cellular mediators of immediate hypersensitivity reactions mast cell activation and that they likely contribute significantly to the delayed In the next section of this review. we will focus on two recruitment of leukocytes during late phase reactions. The relatively new models of inflammation that may directly latter contention is based on experimental data that illus- reflect human disease. In these models there is clear trate that challenge with anti-IgE caused a profound leuko- evidence that microcirculation and more specifically mast cytic influx into skin reaching maximal levels at 6-12 h cell-induced leukocyte recruitment contribute to the pathol- after antigen challenge in normal mice [68]. This leukocyte ogy of bacterial toxin-induced tissue in.jur1 infiltration, however, was virtually undetectable in geneti- cally engineered, mast cell-deficient W/W ” mice [68]. 4. I. Helicohacter j?ylori toxins The leukocytic infiltrate observed at 6 h in normal mice challenged with anti-IgE could be reduced by approx. 50% In recent years. considerable attention has been devoted with anti-TNF antiserum, suggesting that this cytokine to the role of Helicobacter pylori in the pathogenesis of contributes significantly to the delayed leukocyte infiltra- chronic gastritis and gastric ulceration. This microorgan- tion in this allergen-induced mast cell degranulation sys- ism, when ingested by human volunteers. produces gastri- tem. Since IgE-dependent activation of human foreskin tis, while eradication of H. pylori in ulcer patients leads to mast cells has been shown to induce the expression (at 6 h) resolution of pathology. There is a growing body of evi- of E-selectin 1691, an adhesion molecule on endothelial dence that links bacterial infection to infiltrating neu- cells that induces leukocyte rolling [70], a proposed sce- trophils in the gastric mucosa wherein the leukocytes cause nario may be that IgE-dependent mast cell degranulation tissue injury. This view is supported by reports demonstrat- stimulates leukocyte-endothelial cell interactions via TNF ing (1) a direct relationship between the extent of H. and E-selectin. pylori-induced neutrophil infiltration and the severity of To date leukocyte recruitment and the role of TNFcu. gastric mucosal injury. and (2) the t&r that H. pyhri E-selectin or any of the other selectins has not been promotes the adhesion of isolated humax] neutrophils to examined in late phase reactions using intravital video monolayers of human umbilical endothehai cells [74-761. microscopy. primarily because of the difficulties with The inflammatory potential of Ei. l$or! has also been maintaining viable preparations for prolonged periods of demonstrated using intravital microscopic techniques to time. However, numerous studies have recently been pub- monitor leukocyte-endothelial cell adhesion in mcsenteric lished using gene knockout mice exposed to various late venules [77]. Exposure of rat mesentery to a water extract phase reactions. The use of mutant mice lacking L-selectin of H. pylori (HPE) leads to the recruitment of adherent has revealed reduced edema formation in delayed-type leukocytes. which subsequently emigrate into the adjacent hypersensitivity reactions [7 1I. Cellular infiltrates were not interstitial compartment. The HP&induced leukocyte ad- determined in these experiments. Although these data im- herence and emigration is accompanied by the formation ply a role for L-selectin as a mediator of tissue dysfunction of platelet-leukocyte aggregates, mast cell degranulation, in mast cell-dependent models of inflammation, documen- and enhanced albumin leakage in mescnteric vcnules. Comparable inflammatory responses were not elicited by duced by H. pylori in the gastric lumen which gain access exposure of the rat mesentery to a water extract of Es- to the mucosal interstitium via a disrupted epithelial barrier cherichia coli. (caused by stress and/or hyperacidity) promotes mast cell The mast cell degranulation induced by HPE occurred degranulation. Products of mast cell activation and degran- as early as 10 min after exposure to the extract and could ulation such as PAF appear to contribute to the inflamma- be largely prevented by prior treatment with ketotifen tory responses elicited by HPE by diminishing the barrier (Table 3). The mast cell stabilizer had no effect on the function of endothelial cells, which facilitates leukocyte HPE-induced recruitment of adherent leukocytes and emigration as well as the extravasation of adherent leuko- platelet-leukocyte aggregation; however, it significantly cytes. These events, in turn, facilitate the formation of attenuated the normally associated leukocyte emigration gastric erosions and ulcers that are now known to accom- and albumin leakage responses. In addition, ketotifen treat- pany an H. pylori infection [77]. ment prevented the fall in venular shear rate that was Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 usually elicited by HPE (Table 3). These observations 4.2. Clostridium difficile toxin A indicate that the mast cell degranulation induced by HPE contributes significantly to the accompanying changes in Antibiotic therapy can be associated with the develop- venular hemodynamics, leukocyte emigration and albumin ment of a chronic inflammatory condition affecting the leakage but not the earlier event, leukocyte adhesion. colon (i.e., pseudomembranous colitis). The principal etio- The mast cell-derived substance that mediates the ob- logic agent responsible for this iatrogenic form of entero- served responses to HPE remains undefined; however, colitis is C. dificile, a Gram-negative anaerobic bacillus some evidence supports a potential role for platelet activat- that produces two protein exotoxins: toxin A and toxin B. ing factor (PAF). The PAF receptor antagonist WEB 2086 Animal studies have revealed that toxin A (TX-A) mediates appears to be as effective as ketotifen in blunting the the excess fluid secretion and inflammation associated HPE-induced alterations in leukocyte emigration, albumin with experimental C. diflcile enterocolitis. Mast cells have leakage, and shear rate in rat mesenteric venules, while been implicated in the pathobiology of TX-A-induced mu- exerting no influence on the mast cell degranulation or cosal dysfunction, with mast cell degranulation observed in leukocyte adherence responses caused by HPE (Table 3). gut mucosa within 15 min after TX-A exposure and an The observation that both ketotifen and WEB 2086 were attenuation of leukocyte infiltration and tissue necrosis effective in blunting albumin leakage within 10 min after noted in animals pretreated with ketotifen [78,79]. HPE exposure, a time that was not associated with en- Exposure of rat mesentery to TX-A results in an in- hanced leukocyte adherence or emigration, suggests that a creased adherence and emigration of leukocytes, platelet- component of the albumin leakage response that is medi- leukocyte aggregation, and enhanced albumin leakage in ated by mast cell-derived PAF occurs independently of postcapillary venules and the degranulation of perivenular leukocyte recruitment. mast cells [80]. The recruitment of leukocytes elicited by Collectively, the responses elicited by HPE in the TX-A appears to result from adhesive interactions mediated mesenteric microcirculation would suggest that toxins pro- by P-selectin and ICAM- on endothelial cells and

Table 3 Role of mast cells in Helicobacrer pylori-induced microvascular dysfunction Control HPE HPE+ Ketotifen HPE&WEB 2086 Adherent 4.8& 1.2 19.6*2.6’ 14.1* 2.8 J 12.0+ 2.5 a leukocytes (per 100 pm venules)

Emigrated 1.2+ 0.6 11.2fl.7 a 5.1 _+ 0.8 a 5.0+ 0.8 * leukocytes (per field)

Platelet-leukocytes 0.2+ 0.2 7.4 * 0.9 a 6.8* 1.4” 6.4* 1.0 a aggregates (per 5 min)

Degranulated 3.3& 1.7 71.7+4.4 a 23.3f 4.4 a 51.0110.7 b mast cells (8) total

Albumin leakage 8.6+ 5.2 70.4 * 5.5 * 39.7+ 6.1 h 35.1* 4.3h Venular shear 560 It16 376 +9” 533 It18 522 +34 rate (s-’ )

* Denotes significance relative to controls; h denotes significance relative to HPE alone; HPE denotes Helicobacrer pylori extract. P. Kubes, N.D. Granger/ CardioL,ascular Research 32 (1996) 699-708 70’7

CD1 1/CD 18 on leukocytes, since monoclonal antibodies 5. Future directions (MAbs) directed against these adhesion molecules markedly blunt TX-A-induced leukocyte adhesion. The Regulating mast cell reactivity could conceivably be- P-selectin MAb also attenuated TX-A-induced platelet- come a viable approach to modulating feulcocyte influx leukocyte aggregation. Sialyl-Lewis x (SLe”), a counter- and the inflammatory process. One potential (but certainly receptor on leukocytes for P-selectin, is also effective in not exclusive) endogenous modulator of mast cell reactiv- reducing TX-A-induced leukocyte adhesion to both en- ity is nitric oxide (NO). Salvemini et al. 1811reported that dothelial cells and platelets. All of the reagents (MAbs and NO donors including nitroprusside decreased the amount SLe”) that reduced TX-A-induced leukocyte adhesion were of histamine released by mast cells. Moreover, this group also effective in blunting the enhanced albumin leakage demonstrated that L-NAME augmented the release of his- normally elicited by the toxin. tamine from mast cells. Hogaboam et al. 1825reported that Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 The leukocyte adhesion and albumin leakage caused by NO inhibited PAF production from mast cells. Recent data TX-A are affected by agents that either stabilize mast cells from our laboratory [25] demonstrate that NO donors or interfere with histamine action. Lodoxamide, a mast cell attenuate mast cell-dependent leukocyte recruitment in vivo stabilizer, effectively prevents the mast cell degranulation and in vitro. CMP 48/80-induced leukocyte rolling and in mt mesentery that is elicited by TX-A exposure. A adhesion were almost entirely prevented by the NO donor. consequence of this treatment is an attenuation of leuko- spermine-NO. Moreover, spermine-NO was as effective as cyte recruitment and a corresponding reduction in albumin WEB 2086 at preventing neutrophil adhesion in the in leakage. Similar protective effects are noted following vitro assay system described in Fig. I b. An obvious direc- treatment with either diamine oxidase (histaminase) or an tion is to begin to assess the role of NO in mast celi-medi- HI -receptor antagonist. The H ,-receptor antagonist was ated models of disease such as anapbytaxis, asthma and also effective in reducing the formation of platelet-leuko- I/R. Indeed, recent work from Kurose et al. 1603 has cyte aggregates. These observations suggest that histamine demonstrated that three different NO donors inhibit I/R- represents an important mast cell-derived mediator of Tx- induced mast cell degranulation and the resultant leukocyte A-induced inflammation. infiltration. Overall, the reported observations concerning Tx-A- Finally. removal of endogenous NO activates mast ceils mediated alterations in the mesenteric microvasculature are in vivo. Studies from our laboratory demonstrated that consistent with a mechanism wherein TX-A rapidly in- connective tissue mast cells located in close proximity to duces mast cell degranulation, leading to release and accu- the mesenteric vasculature became activated following NO mulation of histamine in the perivenular compartment. The synthesis inhibition [31]. In another study mast cell-derived mast cell-derived histamine appears to mediate at least part protease II levels from mucosal mast cells were shown to of the leukocyte-endothelial and platelet-leukocyte adhe- be elevated following systemic administration of L-NAME sion by engaging H,-receptors on endothelial cells and [83], suggesting that both connective tissue mast cells platelets, respectively, to increase the expression of P- (from mesentery) and mucosal mast cells (intestinal mu- selectin. Activation of the leukocyte adhesion glycoprotein cosa) may be under physiologic control from the continu- CD1 1/CD1 8 and its subsequent interaction with constitu- ous release of endogenous NO. These results were ex- tively expressed ICAM- also contributes to the leukocyte tended by two groups; exposure of tissue to L-NAME recruitment elicited by C. djfJi:cile toxin A (Fig. 4). caused detectable increase in oxidative stress within mast cells in vivo [84,85]. Moreover, various mast cell stabilizers prevented the L-NAME-induced leukocyte recruitment 1 Closhidium diflicile suggesting that mast cells played an important role in leukocyte adhesion following NO synthesis inhibition ]3 11. Determining whether the reduction in NO production reported in ischemia/reperfusion, atherosclerosis and other pathologic states activates mast cells and thereby recruits leukocytes to affected sites is warranted and will be an P-selectin expression important future direction.

References

[I] Von Andrian UH. Chambers JD, McEvoy LN. et al. Proc Nat1 Acad Sci USA 1991:88:7538. [2] Ley K. Gaehtgens P, Fennie C. Singer MS. et xl. Blood Fig. 4. A schematic describing the mechanisms underlying Clostridinm 1991;77:2553. cliflkilr-induced pathogenesis. See text for details. [3] Bevilacqua MP. Nelson RM. J Clin Invest I 90 ‘i.9 i :17’). [4] Springer TA. Cell 1994:76:301. (471 Arfors KE. Ley K. J Lab Clin Med 1993:121:301. [s] Benyon RC, Imai T, Abe T, Befus D. Int Arch Allergy Appl [48] Ley K, Cerrito M, Arfors KE. Am J Physiol 1991:260:Hl667. Immunol 1991;94:218. [49] Tangelder GJ. Arfors KE. Blood I991 ;77: 1565. [6] Galli SJ. New Engl J Med 1993:328:257. [50] Cecconi 0, Nelson RM, Roberts WG. et al. J Biol Chem [7] Crowe SE, Perdue MH. Gastroenterology 1992; 103: 1075. 1994269: 15060. [8] Metcalfe DD, Costa PR, Burd JJ. In: Gallin JI, Goldstein IM, [5 I] Simpson PJ, Todd RF, Fantone JC. et al. J Clio Invest 19X8:8 1:624. Snyderman R (Editors), Inflammation. Basic Principles and Clinical [52] Hernandez LA. Grisham MB. Twohig B, et al. Am J Physiol Correlates. New York: Raven Press, 1992;709. l987:253:H699. [9] Gordon JR, Galli SJ. Nature 1994;346:274. [53] Ma X-L, Tsao PS. Lefer AM. J Clin Invest 1991:88:1237. [IO] Wasserman SI. Arthritis Rheum. 1984;27:841. [54] Granger DN, Benoit JN, Suzuki M. Grisham MB. Am J Physiol [II] Tannenbaum S, Oertel W, Henderson H. Kaliner M. J Immunol 1989:257:G683. 1980; 125:325. [55] Suzuki M. lnauen W. Kvietys PR, et al. Am J Physiol [12] Granger DN, Kubes P. J Leukocyte Biol 1994;55:662. 1989:257:H 1740. [13] Gaboury JP, Johnston B, Niu X-F, Kubes P. J Immunol [56] Kurose I. Anderson DC, Miyasaka M, et al. Circ Res 1994;74:336. Downloaded from https://academic.oup.com/cardiovascres/article/32/4/699/264552 by guest on 25 September 2021 1995; 154:804. [57] Kubes P, Kurose I, Granger DN. Am J Physiol 1994;267:H93 1. [14] Lagunoff D. J Histochem Cytochem 1972;20:938. [58] Bores M. Kaszaki J, Nagy S. Circ Shock 1991;35:174. [15] Raud J, Lindbom L, Dahlen S-E, Hedqvist P. Am J Pathol 1591 Kanwar S, Kubes P. Am J Physiol 1994;267:G3 16. 1989;134:161. [60] Kurose I. Wolf R, Grisham MB, Granger DN. Circ Rea 1994:74:376. [16] Raud J, Dahlen S-E, Smedegard G, Hedqvist P. Acta Physiol Stand [61] Blum H, Summers JJ. Schnall MD, et al. Ann Surg 1986;204:83. 1989;135:95. [62] Bores M, Kdszaki J, Nagy S. Eur Surg Res 1989;21:297. [17] Thorlacius H, Raud J, Rosengren-Beezley S, et al. Biochem Biophys [63] Johnson AR, Hugli TE, Muller-Eberhard HJ. .l Immunol Res Commun 1994;203:1043. 1974;28: 1067. [I81 Geng J-G, Bevilacqua MP, Moore KL, et al. Nature 1990:343:757. [64] Hachfeld de1 Balro U. Levi R, Polley MJ. Proc Nat1 Acad Sci USA [19] Lawrence MB, Springer TA. Cell 1991;65:859. 1985:82:886. [20] Gaboury J, Anderson DC, Kubes P. Am J Physiol 1994:266:H637. [65] Weisman HF, Bartow T, Leppo MK, et al. Science 1993;249:146. [21] Kubes P, Kanwar S. J Immunol 1994;152:3570. [66] Schwartz LB. Kawahara MS. Hugli TE, et al. J Immunol [22] Jones DA, Abbassi 0, McIntire LV. McEver RP, Smith CW. 1983;130:1891. Biophys. J. 1993;65: 1560. [67] Goldman G. Welbourn R, Klausner JM, et al. Surgery 1992; 1 12:578. [23] Zimmerman GA. Prescott SM, McIntyre TM. Immunol. Today [68] Wershil BK, Wang ZS. Gordon JR. Galli SJ. J Clin Invest 1992;13:93. 199 I :87:446. [24] Zimmerman GA, Prescott SM, McIntyre TM. In: Fritsch J, Miller- [69] Klein LM, Lavker RM, Matis WL, Murphy GF. Proc Nat1 Acad Sci Peddinghaus R (Editors), Advances in Rheumatology and Inflamma- USA 1989;86:8972. tion. Base]: Eular Pub]., 1992; 49. [70] Abbassi 0, Kishimoto TK, McIntire LV, Anderson DC. Smith CW. [25] Gaboury JP, Niu X-F, Kubes P. Circulation 1996;93:318. J Clin Invest 1993;92:2719. [26] Voyta JC, Via DP, Butlerfield CE, Zetter BR. J Cell Biol [71] Tedder TF, Steeber DA, Pizcuetta P. J Exp Med 1995;181:2259. 1984;99:2034. [72] Subramaniam M. Saffaripour S, Watson SR, et al. J Exp Med [27] Mayrovitz HN, Tuma RF, Wiedeman MP. Microvasc Res 1995;181:2277. 1980;20:264. [73] Labow MA, Norton CR, Rumberger JM. et al. Immunity 1995; 1:709. [28] Fiebig E, Ley K, Arfors KE. Int J Microcirc Clin Exp 1991;10:127. [74] Cover TL. Blaser MJ. Helicohucter p$ori and gastroduodenal dis- [29] Dore M, Korthuis RJ, Granger DN. Entman ML, Smith CW. Blood ease. Annu Rev Med 1992;43: 135- 145. 1993;82:1308. [75] Wallace JL. Possible mechanisms and mediators of gastritis associ- [30] Lindbom L, Xie X, Raud J, Hedqvist P. Acta Physiol Stand ated with Helicobacter pvlori infection. Stand J Gastroenterol 1992;146:415. 199 1;26:65-70. [31] Kubes P, Kanwar S, Niu X-F, Gaboury J. FASEB J 1993;7: 1293. [76] Yoshida N, Granger DN, Evans DJ, et al. Mechanisms involved in [32] Asako H, Kurose I, Wolf R, et al. J Clin Invest 1994;93:1508. Helicobncter pylori induced inflammation. Gastroenterology [33] Kanwar S, Johnston B, Kubes P. Circ Res 1995;77:879. 1994: 107:70-79. [34] Johnston B, Kanwar S, Kubes P. Am J Physiol 1996; in press. [77] Kurose I, Granger DN, Evans DJ, et al. Helicobacterpylori induced [35] Tonnesen MG. J Invest Dermatol 1989;93:53s. microvascular protein leakage in rats: role of neutrophils, mast cells [36] Tonnesen MG, Anderson DC, Springer TA, et al. J CIin Invest and platelets. Gastroenterology 1994;107:70-79. 1989;83:637. [78] Triadafilpoulos GC, Pothoulakis C, O’Brian M, LaMont JT. Differ- [37] Gillard J, Ford-Hutchinson AW. Chan C, et al. Can J Physiol ential effects of Chtridium di@cik toxin A and B on rabbit ileum. Pharmacol 1989;67:456. Gastroenterology 1987;92:273-279, 1381 Rouzer CA, Ford-Hutchinson AW, Morton HE, Gillard JW. J Biol [79] Pothoulakis C, Karmeli F, Kelly CP, et al. Ketotifen inhibits Chem 1990;265: 1436. Clostridium difSicile toxin A induced enteritis in rat ileum. Gastroen- [39] Wallace JL. In: Saito K, Hanahan, DJ (Editors), Platelet-Activating terology 1993; 105:701-707. Factor and Diseases. Tokyo: International Medical Pulishers, 1989; [SO] Kurose 1. Pothoulakis C, LaMont JT, et al. Clostridiunt d@cile 153. toxin A-induced microvascular dysfunction: role of histamine. J Clin [40] Kanwar S, Kubes P. Microcirculation 1994; l(3): 175. Invest 1994;94:1919-1926. [41] Kubes P, Jutila M, Payne D. J Clin Invest 1995;95:2510. [Sl] Salvemini D, Masini E, Pistelli A. Mannaioni PF, Vane J. J Cardio- [42] Lawrence MB, Bainton DF, Springer TA. Immunity 1994; 1: 137. vast Pharmacol 1991;17(supp1 3):SZSS. [43] Lawrence MB, Springer TA. J Immunol 1993; 15 1:6338. I821 Hogaboam CM, Befus AD, Wallace JL. J Immunol 1993;151:3767. [44] Kishimoto TK, Warnock RA, Jutila MA, et al. Blood 1991;78:805. 1831 Kanwar S, Wallace JL. Befus D, Kubes P. Am J Physiol [45] Albettine KH, Weyrich AS, Ma X, et al. J Leukocyte Biol 1994;266:G222. 1994:55:557. [84] Kurose 1. Wolf R, Grisham MB, et al. Circ Res 1995;76:30. [46] Nelson RM, Dolich S, Aruffo A, Cecconi 0, Bevilacqua MP. J Clin [85] Suematsu M, Tamatani T, Delano FA, et al. Am J Physiol Invest 1993;91:1157. 1994;266:H2410.