INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

REPERFUSION OF ISCHEMIC MYOCARDIUM results in an inflammatory response characterized by rapid accumulation of neutrophils in the reperfused tissue (6, 7, 15, 21). Extravasation of neutrophils can be seen histologically as a phenomenon that potentially allows direct contact between cardiac myocytes and neutrophils (1, 15, 26). We have investigated the possibility that adherent neutrophils can damage cardiac myocytes and have obtained evidence in vitro for a Mac-1- and intercellular adhesion molecule (ICAM)-1-dependent process with associated intracellular oxidant stress and subsequent death of myocytes (10, 12, 28). Neutrophil adhesion to the myocyte is necessary, and the cytotoxic events can be prevented by pretreatment of the neutrophils with antibodies that block the adhesive function of Mac-1 (12). For this neutrophil-myocyte adhesion to occur, the myocytes must be stimulated with cytokines [interleukin (IL)-6, IL-1, or tumor necrosis factor (TNF)-] that induce synthesis and surface expression of ICAM-1 (28), and the neutrophils must be stimulated with chemotactic factors [IL-8, C5a, or platelet-activating factor (PAF)] (12, 18, 33) that activate Mac-1-dependent adhesion.

A potentially important characteristic of Mac-1-dependent neutrophil adhesion is that it greatly augments hydrogen peroxide (H₂O₂) production by chemotactically stimulated cells (24, 27). This has been shown with neutrophils adherent to endothelial cells and cardiac myocytes, and the quantities of reactive oxygen released are equivalent to those induced by stimulation of neutrophils with phorbol esters. Not only are reactive oxygen species potentially cytotoxic for target cells but also their accumulation locally may, in turn, enhance further neutrophil adhesion. For example, studies in vitro have shown that H₂O₂ treatment of previously unstimulated endothelial cells induces rapid ICAM-1- and CD18-dependent adherence of unstimulated neutrophils (19). Gasic et al. (14) observed a similar phenomenon with isolated vessels. They found that H₂O₂ in the absence of evidence for permanent endothelial cell injury can induce a transient, reversible adhesion of neutrophils to isolated vessels by mechanisms that depend on an intact endothelium and involve CD18 on the neutrophils.

In the present study, we have chosen to investigate a possible role for H₂O₂ in the adhesion of neutrophils to cardiac myocytes. Such a mechanism may result in recruitment of additional neutrophils by neutrophils whose production of reactive oxygen species is greatly augmented by Mac-1-dependent adherence to myocytes. Our results demonstrate that exposure of isolated cardiac myocytes to H₂O₂ induces a rapid increase in leukocyte function-associated antigen (LFA)-1 (CD11a/CD18)-dependent neutrophil adhesion, peaking within 15 min.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Materials. R15.7 (IgG1), an anti-CD18 monoclonal antibody (MAb) (12), R7.1 (IgG1), an anti-CD11a MAb (2, 3); and R6.5 and CA7 (IgG1), MAbs specific for human ICAM-1 domains 2 and 5 (25), were kindly provided by Dr. R. Rothlein (Boehringer Ingelheim Pharmaceutical, Ridgefield, CT). MY-904 (IgG1) (5), an anti-CD11b MAb, was obtained from Lilly (Indianapolis, IN). CL18/6 and CL18/1 (IgG1), anti-canine ICAM-1 MAbs, were produced in our laboratory (28).

Construction of full-length ICAM-1. The canine ICAM-1 cDNA sequence resides in three overlapping cDNA clones obtained from a lipopolysaccharide (LPS)-stimulated canine endothelial cell library (20). A full-length cDNA was constructed from these three clones by a series of subcloning and ligation. The first six nucleotides, AUGCCG, conserved among ICAM-1 of other known species, were added to the canine ICAM-1 sequence by PCR.

Construction of chimeric ICAM-1. Both human (31) and canine ICAM-1 cDNA were subcloned into pGEM4 vector. A Bgl I site was introduced into the canine ICAM-1 sequence at a position homologous to that of the unique Bgl I site in human cDNA by PCR. The unique Bgl I site, at ~17 codons from the carboxyl end of domain 2 in human cDNA, was used to construct chimeric ICAM-1 having a canine amino terminus with domains D1 and D2 and a human carboxyl terminus with domains D3-D5 (C1,2:H3-5) and vice versa (H1,2:C3-5).

L cell transfection. Canine ICAM-1 and canine-human chimeric ICAM-1-expressing L cell lines were generated using the cationic lipid N-[1-(2,3-dioloyloxy)propyl]-N,N,N-trimethylammonium methylsulfate (Boehringer Mannheim Biochemicals, Indianapolis, IN) transfection by following the manufacturer's instructions. Cells were allowed to recover in fresh medium for 72 h after the transfection and were then selected under 600 µg/ml Geniticin (G418). Cells with high ICAM-1 expression were identified by flow cytometry analysis after single-colony picks with a cloning cylinder (Bellco, Vineland, NJ). ICAM-1-transfected L cells were maintained in a selection medium that consisted of 600 µg/ml G418. Untransfected L cells were maintained in RPMI 1640 without G418.

Flow cytometry. Flow cytometry study was performed to analyze the L cell surface expression of ICAM-1. Briefly, ICAM-1-transfected L cells were trypsinized and resuspended at a concentration of 1 × 10⁶ cells per 90 µl in Dulbecco's phosphate-buffered balanced solution (DPBS; GIBCO-BRL). Ten microliters of first antibody at 200 ng/ml were added to the cells and incubated for 15 min at room temperature. Cells were washed three times with DPBS and resuspended in 100 µl of a 1:20 dilution of purified FITC-conjugated goat anti-mouse IgG light- and heavy-chain antisera (Zymed) in DPBS and incubated at room temperature in the dark for 15 min. Mock-transfected L cells and isotype-matched IgG1 first antibody were used as negative controls. After being incubated with antibody, cells were washed three times with DPBS and fixed in 1% paraformaldehyde. The samples were analyzed in a Becton Dickinson FACscan.

Preparation of L cell monolayers. L cell transfectants were trypsinized with trypsin-EDTA. For 25-mm round coverslips pretreated with 0.1% gelatin, 0.5 × 10⁶ trypsinized L cells in 0.5 ml of culture medium were added and sedimented for 4 h at 37°C, 5% CO₂. Two milliliters of culture medium were then added, and the cells were grown to confluency within 2-3 days. Only confluent monolayers were used in the adhesion assay.

Isolation of canine polymorphonuclear neutrophils. Neutrophils were isolated from healthy mongrel dogs as described previously (30).

Canine neutrophil adherence to L cell transfectants. L cells or transfected L cells were plated onto gelatin (0.1%)-pretreated 25-mm coverslips and allowed to become visually confluent. A visual static adhesion assay was described in detail previously (29). Leukocytes were allowed to settle onto monolayers of L cells for a contact time of 500 s under static conditions. The number of neutrophils in contact with the monolayer was counted, and the chamber was inverted for an additional 500 s to allow the unattached leukocytes to fall off. The leukocytes remaining attached were counted, and this number was expressed as a percentage of the total number of cells initially contacting the monolayer. In studies designed to evaluate the involvement of ₂-integrins or ICAM-1 in neutrophil adhesion, cells were preincubated as follows: the coverslip with L cell transfectants was treated with anti-ICAM-1 MAbs at a concentration of 20 µg/ml in 1 ml of PBS for 30 min at room temperature and mounted in the adhesion chamber directly. Neutrophils were incubated with antibodies specific for integrin subunits at two to four times the saturating concentration at room temperature for 30 min. Chemotactic stimuli of 1% zymosan A (Sigma, St. Louis, MO)-activated serum [ZAS; prepared as previously described (12)] for canine neutrophils were added immediately before the cell mixture was injected into the adhesion chamber.

Isolation of canine myocytes. Healthy mongrel dogs weighing 10-15 kg were anesthetized with pentobarbital sodium. The heart was removed through the left lateral chest under sterile conditions and immediately placed in ice-cold saline. The aorta was then cannulated using a tubing adapter suitable for the individual heart, and the procedure for obtaining isolated myocytes was followed exactly as described previously (12).

Canine neutrophil adherence to canine myocytes. Isolated canine myocytes were suspended in medium A at a concentration of 50,000/ml as previously described (12). Neutrophils and myocytes were coincubated in a volume of 0.4 ml and a neutrophil-to-myocyte ratio of 10:1 for various periods of time as indicated at 37°C. The cells were resuspended, and a small aliquot was transferred to a microscope slide for examination under phase-contrast or differential interference-contrast optics. For each preparation, the number of neutrophils adherent to each of 200 myocytes was counted. Myocytes were incubated in the presence of cytokines, including human recombinant IL-1 (Genzyme, Boston, MA) and human recombinant IL-6 (rIL-6; Genzyme) for 3 h at 37°C before the 2-min exposure to H₂O₂ (Sigma). The H₂O₂ applied to myocytes was neutralized by adding catalase (Sigma) for 1 min before the addition of neutrophils. In experiments with stimulated neutrophils, ZAS was added immediately before the neutrophil suspension was mixed with the suspension of myocytes. MAbs were added to the neutrophil-myocyte suspension at the beginning of incubation and were present throughout the incubation period.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

H₂O₂ induces adherence of canine neutrophils to cytokine-activated cardiac myocytes. We have shown previously that adhesion of canine neutrophils to cytokine-stimulated cardiac myocytes occurs only if a chemotactic stimulus such as ZAS, PAF (12), or IL-8 (18) is added to the cell mixture. Peak adhesion occurs within 60 min. In the present experiments with rIL-6-stimulated myocytes treated for 2 min with H₂O₂, neutrophil adhesion occurred without added chemotactic stimulus (Fig. 1A). This effect of H₂O₂ was limited to the cardiac myocytes, because catalase was added after the 2-min incubation before the addition of neutrophils. In contrast to results without H₂O₂ exposure, peak adhesion occurred within 15-20 min, and adhesion was maximal and then began to wane (Fig. 1B). When a chemotactic stimulus was added to the cell mixture, the early kinetics of adhesion were unchanged, but the duration of peak adhesion was extended to at least 60 min (Fig. 1B). H₂O₂ exposure of neutrophils did not induce adhesion to IL-6-stimulated cardiac myocytes (Fig. 2A), and addition of a chemotactic stimulus resulted in the same slow kinetics of adhesion seen in earlier studies with no significant increase in adhesion at 15 min of contact between leukocytes and myocytes.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effect of H2O2 on adhesiveness of cytokine-stimulated myocytes for canine neutrophils. A: isolated canine myocytes were exposed to H2O2 for 2 min after 3 h of stimulation with recombinant interleukin (rIL)-6. Catalase was then added for 1 min, and cells were washed with PBS. Isolated canine neutrophils were then mixed with myocytes, and adhesion was assessed at 20 min. Control conditions without H2O2 pretreatment were evaluated for comparison; n = 3 experiments. B: isolated canine myocytes were exposed to H2O2 (15 mM) for 2 min after 3 h of stimulation with rIL-6. Catalase was then added for 1 min, and cells were washed with PBS. Isolated canine neutrophils were then mixed with myocytes, and adhesion was assessed at various times up to 60 min. , Adhesion of previously unstimulated neutrophils; , adhesion in presence of 1% zymosan-activated serum (ZAS); n = 4 experiments. *P < 0.05.

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Effects of H2O2 were evident only with cytokine-stimulated cardiac myocytes. A: neutrophils were pretreated with 15 mM H2O2 for 2 min, catalase was added for 1 min, and cells were washed in PBS. These cells were then mixed with cytokine-stimulated (3 h) canine cardiac myocytes in absence () or presence (+) of 1% ZAS for 15 min. B: Isolated myocytes were stimulated for 3 h with rIL-6, washed, and exposed to unstimulated neutrophils. Isolated but unstimulated cardiac myocytes were exposed to H2O2 for 2 min, followed by catalase for 1 min, and were then washed in PBS. Adhesion of unstimulated neutrophils at 15 min was low. Adhesion at 15 min was also very low when unstimulated myocytes were treated with H2O2. When cytokine-stimulated myocytes were exposed to H2O2, adhesion of unstimulated neutrophils was significantly higher (* P < 0.01; n = 3 experiments).

ICAM-1 and LFA-1 mediate H₂O₂-induced adherence of neutrophils to myocytes. The potential role of ICAM-1 in neutrophil adhesion induced by H₂O₂ treatment of stimulated myocytes was analyzed using the MAbs CL18/1 and CL18/6. The binding characteristics of these two antibodies were evaluated on L cells expressing full-length canine ICAM-1 (Fig. 3). CL18/1 and CL18/6 did not bind to L cells expressing human ICAM-1 (data not shown). The binding specificities of these two antibodies were further characterized by using two chimeric ICAM-1 constructs, C1,2:H3-5 and H1,2:C3-5. These chimeras were transfected into L cells, and surface expression of chimeric ICAM-1 was detected by the MAbs CA7 and R6.5 recognizing human ICAM-1 domains 5 and 2, respectively (Fig. 4). CL18/1 bound to L cells expressing H1,2:C3-5 but not C1,2:H3-5, and CL18/6 bound to C1,2:H3-5 but not H1,2:C3-5.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Binding of monoclonal antibodies (MAbs) CL18/6 (A) and CL18/1 (B) to canine intercellular adhesion molecule (ICAM)-1. L cells transfected with canine ICAM-1 were analyzed by flow cytometry after indirect immunofluorescence staining with MAb CL18/1 (20 µg/ml) and CL18/6 (20 µg/ml). Background binding of an Ig isotype-matched control is indicated (dotted lines).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Different binding specificities for MAbs CL18/1 and CL18/6. L cells transfected with chimeric ICAM-1 with canine domains 1 and 2 and human domains 3-5 (C1,2:H3-5; A) or human domains 1 and 2 and canine domains 3-5 (H1,2:C3-5; B) were analyzed by flow cytometry after indirect immunofluorescence staining with anti-ICAM-1 antibodies indicated. R6.5 binds to human domain 2, and CA7 binds to human domain 5. These MAbs were used to distinguish the ICAM-1 chimeras. Background binding of IgG1 is indicated (dotted lines). Data indicate that CL18/1 recognizes only H1,2:C3-5 and that CL18/6 recognizes only C1,2:H3-5.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effects of MAbs CL18/1 and CL18/6 on canine neutrophil adhesion to canine ICAM-1. L cells (L-mock, open bars) or L cells transfected with canine ICAM-1 (L-ICAM-1, filled bars) were grown into confluent monolayers and placed in static adhesion chambers. Isolated canine neutrophils were stimulated with ZAS (1% vol) immediately before cells were injected into adhesion chamber. When MAbs were present, they were incubated with L cells for 30 min before adhesion assay, and they were retained with cells throughout time allowed for adhesion. *P < 0.01 compared with stimulus condition without MAbs (none); n = 5 experiments.

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Effects of R15.7 (anti-CD18), CL18/6 (anti-ICAM-1), CL18/6 (anti-ICAM-1, nonblocking) MAbs on H2O2-induced adherence. Isolated canine cardiac myocytes were stimulated for 3 h with rIL-6 and then treated for 2 min with 15 mM H2O2 (hatched bars). Catalase was then added for 1 min, and cells were washed in PBS. Antibodies were retained throughout adhesion period. Control condition in absence of H2O2 pretreatment (filled bars) was assessed for comparison. *P < 0.01, compared with stimulus conditions without MAbs (none); n = 4 experiments.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effects of anti-CD11a on neutrophil adherence to H2O2-treated cardiac myocytes. Isolated canine cardiac myocytes were stimulated with rIL-6 for 3 h, exposed to H2O2 (15 mM) for 2 min followed by catalase for 1 min, and washed in PBS. Isolated, unstimulated canine neutrophils were mixed with myocytes, and adhesion was assessed under phase-contrast microscopy at 5, 15, 30, and 60 min. A: addition of MAb R7.1 (anti-CD11a, 20 µg/ml) to neutrophil mixture () for 15 min before adhesion assay prevented transient peak in neutrophil adhesion (data representative of 4 separate experiments). Antibody MY-904 (anti-CD11b) had a small effect on this peak adhesion (). , No MAb added. B: when 1% ZAS was included with neutrophils, adhesion was sustained (). Addition of MAb R7.1 (anti-CD11a, 20 µg/ml) prevented rapid peak of adhesion but failed to alter delayed peak of adhesion (), which we previously showed to be dependent on CD11b/CD18 (Mac-1). For comparison, a control condition without H2O2 pretreatment was evaluated (). Data are representative of 4 separate experiments.

Evidence for PAF receptor activation in H₂O₂-induced neutrophil adhesion. Current evidence indicates that ₂-integrins on unstimulated neutrophils reside in a low-avidity (affinity) state, and to function efficiently as adhesion molecules the neutrophils must be stimulated (e.g., with chemotactic factors). Previous studies showed that H₂O₂ treatment of endothelial cells would cause PAF receptor stimulation (19). The PAF-receptor antagonist WEB-2086 inhibited the H₂O₂-induced neutrophil adhesion to cardiac myocytes in a dose-dependent manner (Fig. 8). WEB-2086 also inhibited the rapid adhesion seen when ZAS-stimulated neutrophils contacted H₂O₂-treated, cytokine-stimulated cardiac myocytes, but it had no effect on the slowly developing adhesion seen when ZAS-stimulated neutrophils contacted cytokine-stimulated myocytes (Fig. 9). Another PAF-receptor antagonist, SDZ-64-412 (16), also inhibited the rapid adhesion in a dose-dependent fashion (e.g., 75% inhibition at 50 µM) when the ZAS-stimulated neutrophils contacted H₂O₂-treated, cytokine-stimulated cardiac myocytes (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 8.   Effect of WEB-2086 on H2O2-induced adhesiveness of cardiac myocytes. Isolated canine cardiac myocytes were stimulated with rIL-6 for 3 h, exposed to H2O2 (15 mM) for 2 min followed by catalase for 1 min, and washed in PBS. Isolated, unstimulated canine neutrophils were mixed with myocytes, and adhesion was assessed at 15 min in presence of various concentrations of WEB-2086 ([WEB-2086]). Control condition in absence of H2O2 pretreatment and WEB-2086 evaluated for comparison resulted in 0.29 ± 0.07 neutrophils per myocyte; n = 3 experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Effect of WEB-2086 on H2O2-induced adhesiveness of cardiac myocytes to ZAS-stimulated neutrophils. Isolated canine neutrophils were stimulated with ZAS (1% vol) and mixed with isolated canine myocytes in presence of WEB-2086 (5 µM). Adhesion was assessed under phase-contrast microscopy for up to 60 min. A: isolated canine myocytes were stimulated with rIL-6 for 3 h, exposed to H2O2 (15 mM) for 2 min followed by catalase for 1 min, and washed in PBS in presence () or absence () of WEB-2086. B: isolated canine myocytes were stimulated with rIL-6 for 3 h with no H2O2 treatment in presence () or absence () of WEB-2086. Data are representative of 4 separate experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results of the experiments in the current study show for the first time that canine neutrophils can utilize the ₂-integrin LFA-1 (CD11a/CD18) to adhere to canine cardiac myocytes. This conclusion is supported by the following observations. 1) ICAM-1 (CD54), a well-known ligand for LFA-1, is expressed on cardiac myocytes after exposure of the myocytes to rIL-6, IL-1, and TNF- (12, 33). In the current studies, neutrophil adhesion to myocytes was demonstrable only when myocytes were stimulated to express ICAM-1. 2) Canine neutrophils adhered to recombinant chimeric ICAM-1 containing canine ICAM-1 domains 1 and 2 (the region shown to support human LFA-1-dependent adhesion to human ICAM-1), and this adhesion was inhibited by R7.1 (anti-CD11a) and CL18/6 (an MAb that maps to domains 1 and 2 of canine ICAM-1). 3) These antibodies blocked adhesion of canine neutrophils to canine cardiac myocytes. Thus canine neutrophils can use LFA-1 to adhere to ICAM-1, and they can use this mechanism to attach to cardiac myocytes.

Our results also show that stimulation of myocytes with cytokines is not sufficient to promote LFA-1-dependent neutrophil adhesion. Although the presence of ICAM-1 on the parenchymal cell surface is necessary, stimulation of the neutrophil is also required. Addition of chemotactic factors to the culture medium promotes adhesion (12, 33), and, as we have shown, exposure of myocytes to H₂O₂ apparently generates a stimulus sufficient to promote neutrophil attachment. The molecular nature of this stimulus is obscure, but our finding that two different PAF-receptor antagonists, WEB-2086 and SDZ-64-412 (16), inhibit neutrophil adhesion is consistent with activation through the PAF receptor. Lewis et al. (19) showed that H₂O₂ treatment of endothelial cells generates oxidized phospholipids apparently capable of stimulating neutrophils via the PAF receptor to transiently increase their adhesion to endothelial cells. Such a mechanism appears to function at the surface of cardiac myocytes. We recently reported that activation of canine neutrophils through the PAF receptor will stimulate LFA-1-dependent adhesion to canine pulmonary interstitial fibroblasts in vitro (4).

In earlier studies we showed that adherent canine neutrophils will release substantial amounts of reactive oxygen (27). This is true of neutrophils attached to protein-coated, plastic endothelial cells (27) and cardiac myocytes (12, 13). With regard to cardiac myocytes, this release is measurable in two distinct compartments. Initially, reactive oxygen is detected intracellularly within both neutrophils and the myocytes to which they adhere (13); later, reactive oxygen is detected in the extracellular environment (12). The latter phenomenon is quantitatively equivalent to stimulation of neutrophils with phorbol esters (12, 23, 27). Thus adherent neutrophils are potential sources of H₂O₂ in the microenvironment of the cardiac myocyte surface. It remains to be shown that such H₂O₂ could initiate the sequence of events that we generated in vitro by addition of H₂O₂, but the hypothesis that adherent neutrophils can recruit additional neutrophils to adhere to cardiac myocytes is of potential interest. Our results indicate that adhesion to myocytes through such a mechanism would occur maximally within 15 min, would involve LFA-1 adhesion to ICAM-1, and, when combined with exogenous chemotactic stimulation, would result in transition from LFA-1- to Mac-1-dependent adhesion.

Our recent studies in a canine model of myocardial reperfusion do not negate this hypothesis. After 1 h of ischemia, neutrophils localize in the previously ischemic myocardium during the first 3-4 h of reperfusion, with most localization occurring within the first 2 h. Intravascular accumulation of neutrophils is quite marked early, followed by substantial emigration of neutrophils at 3 h of reperfusion (9, 15). At this time, many neutrophils appear to be in contact with cardiac myocytes. C5a is present in the cardiac lymph draining the reperfused tissue in sufficient quantities to activate neutrophil adhesion and locomotion, and it peaks within the first 2 h after reperfusion (8). ICAM-1 is expressed by the cardiac myocytes within the first few hours of reperfusion (11, 17, 22, 32). We have shown that adding canine C5a in vitro will activate Mac-1- and/or ICAM-1-dependent adhesion of neutrophils to cardiac myocytes, a condition promoting massive release of H₂O₂.

The functional role that LFA-1-dependent adhesion could play remains to be defined, but two considerations arise from our observations. The first relates to the kinetics of adhesion. Our observations in vitro indicate that the rate of neutrophil adhesion to myocytes is rather slow if the chemotactic stimulus is simply added to the mixture of cells, peaking in ~1 h. In contrast, with H₂O₂-treated myocytes, the rate of adhesion is significantly faster, peaking within 15 min. The second consideration relates to neutrophil motility. As indicated above, we have recently observed that stimulation of canine neutrophils through the PAF receptor increases adhesion to pulmonary fibroblasts. One striking characteristic of this adhesion is that these adherent neutrophils are immobile (4). This is in contrast to the effects of recombinant canine IL-8, which promotes migration of neutrophils over the fibroblast surface. The LFA-1-dependent adherent neutrophils on cardiac myocytes appear to be immobile as well. We have previously shown that Mac-1-dependent adherent neutrophils form a compartmented space at the interface with the myocyte that is sufficiently tight to exclude molecules the size of superoxide dismutase and catalase (13). Nathan et al. (22) observed that chemotactically stimulated adherent and randomly migrating neutrophils did not release measurable H₂O₂, but once they began to spread and become immobile, they released large quantities. If immobile adhesion precedes secretory activity at the myocyte surface, then the consequences of H₂O₂ stimulation may be enhancement of the transition of emigrated neutrophils into secretory cells.

The results in this report simply provide evidence that exposing canine cardiac myocytes to H₂O₂ will activate a sequence of events that leads to rapid LFA-1-dependent adhesion of neutrophils by activating the PAF receptor. If exogenous chemotactic stimulation is absent (a condition that would only occur in the artificial in vitro setting), then the LFA-1-dependent adhesion is transient and detachment of neutrophils from the myocytes occurs. If exogenous chemotactic factors are present (e.g., C5a), then there is a transition from LFA-1- to Mac-1-dependent adhesion and prolonged attachment of the neutrophil to the myocyte.