Activation of the complement cascade with the generation of anaphylatoxins accompanies the inflammatory response elicited by acute myocardial ischemia and reperfusion. Although complement is activated in the interstitium during acute myocardial ischemia, we have studied mechanisms whereby complement might exacerbate ischemia by using a model employing intracoronary injection of C5a in nonischemic hearts. Intracoronary injection of complement component C5a induces transient myocardial ischemia, mediated through the production of the coronary vasoconstrictors thromboxane A2 and peptidoleukotrienes (LTC4, LTD4), and causes sequestration of polymorphonuclear leukocytes (PMN) in the coronary vascular bed. To further investigate the role of the PMN in the C5a-induced vasoconstriction, the left anterior descending coronary artery (LAD) in pigs was perfused at constant pressure and measurements of coronary blood flow, myocardial contractile function (sonomicrometry), arterial/coronary venous blood PMN count, and thromboxane B2 (TxB2) levels were performed. The myocardial response to intracoronary C5a (500 ng) was determined before, during, and after perfusion with blood depleted of PMNs using leukocyte filters (Sepacell R-500, Pall PL-100). In additional animals, the myocardial response to the PMN chemotactic agent, LTB4, and the effects of intracoronary C5a during constant flow perfusion were measured. Control intracoronary injection of C5a decreased flow (41% of baseline) and contractile function (39% of baseline), PMNs were trapped (5.1 X 10(3) cells/mu-l), and TxB2 concentration increased in coronary venous blood. The response to C5a during coronary perfusion with arterial blood depleted of PMNs with Sepacell or Pall filters (< 0.1 X 10(3) cells/mu-l) was greatly blunted, with flow and contractile function falling by less than 14 and 8%, respectively, from baseline, and release of TxB2 was greatly attenuated. However, the myocardial ischemia and TxB2 release remained depressed in response to C5a after removal of the filters and perfusion with either arterial blood containing normal levels of PMNs or stored arterial blood never exposed to filters. In contrast, the repeat C5a challenge resulted in equivalent myocardial extraction of PMNs, thus indicating a dissociation of PMN sequestration from the acute ischemic response and release of TxB2. In separate experiments, the intracoronary injection of LTB4 also resulted in a pronounced myocardial extraction of PMNs (8.6 x 10(3) cells/mu-l) greater than during C5a, but did not depress coronary flow or function. Perfusion at constant flow greatly diminished the ischemic response to C5a, indicating that vasoconstriction and resultant ischemia is the main cause of the contractile dysfunction. These data indicate that leukocyte filters inhibit the myocardial ischemia and release of TxB2 induced by C5a via mechanisms not related to PMN depletion. These filters may directly desensitize the myocardium or may induce the release of factors that interfere with the effects of C5a in our model. These data argue against a causal role of microvascular obstruction or PMN trapping in the acute myocardial ischemic response to intravascular activated complement C5a and suggest that the acute ischemic response may have PMN-independent components. However, the longterm contribution to injury of PMNs trapped by complement activation during ischemia requires further study.
