Induction of HO-1 in tissue macrophages and monocytes in fatal falciparum malaria and sepsis

Background: As well as being inducible by haem, haemoxygenase-1 (HO-1) is also induced by interleukin-10 and an anti-inflammatory prostaglandin, 15d PGJ(2), the carbon monoxide thus produced mediating the anti-inflammatory effects of these molecules. The cellular distribution of HO-1, by immunohistochemistry, in brain, lung and liver in fatal falciparum malaria, and in sepsis, is reported. Methods: Wax sections were stained, at a 1:1000 dilution of primary antibody, for HO-1 in tissues collected during paediatric autopsies in Blantyre, Malawi. These comprised 37 acutely ill comatose patients, 32 of whom were diagnosed clinically as cerebral malaria and the other 5 as bacterial diseases with coma. Another 3 died unexpectedly from an alert state. Other control tissues were from Australian adults. Results: Apart from its presence in splenic red pulp macrophages and microhaemorrhages, staining for HO-1 was confined to intravascular monocytes and certain tissue macrophages. Of the 32 clinically diagnosed cerebral malaria cases, 11 (category A) cases had negligible histological change in the brain and absence of or scanty intravascular sequestration of parasitized erythrocytes. Of these 11 cases, eight proved at autopsy to have other pathological changes as well, and none of these eight showed HO-1 staining within the brain apart from isolated moderate staining in one case. Two of the three without another pathological diagnosis showed moderate staining of scattered monocytes in brain vessels. Six of these 11 (category A) cases exhibited strong lung staining, and the Kupffer cells of nine of them were intensely stained. Of the seven (category B) cases with no histological changes in the brain, but appreciable sequestered parasitised erythrocytes present, one was without staining, and the other six showed strongly staining, rare or scattered monocytes in cerebral vessels. All six lung sections not obscured by neutrophils showed strong staining of monocytes and alveolar macrophages, and all six available liver sections showed moderate or strong staining of Kupffer cells. Of the 14 (category C) cases, in which brains showed microhaemorrhages and intravascular mononuclear cell accumulations, plus sequestered parasitised erythrocytes, all exhibited strong monocyte HO-1 staining in cells forming accumulations and scattered singly within cerebral blood vessels. Eleven of the available and readable 13 lung sections showed strongly staining monocytes and alveolar macrophages, and one stained moderately. All of the 14 livers had strongly stained Kupffer cells. Of five cases of comatose culture-defined bacterial infection, three showed a scattering of stained monocytes in vessels within the brain parenchyma, three had stained cells in lung sections, and all five demonstrated moderately or strongly staining Kupffer cells. Brain sections from all three African controls, lung sections from two of them, and liver from one, showed no staining for HO-1, and other control lung and liver sections showed few, palely stained cells only. Australian-origin adult brains exhibited no staining, whether the patients had died from coronary artery disease or from non-infectious, non-cerebral conditions. Conclusions: Clinically diagnosed 'cerebral malaria' in children includes some cases in whom malaria is not the only diagnosis with the hindsight afforded by autopsy. In these patients there is widespread systemic inflammation, judged by HO-1 induction, at the time of death, but minimal intracerebral inflammation. In other cases with no pathological diagnosis except malaria, there is evidence of widespread inflammatory responses both in the brain and in other major organs. The relative contributions of intracerebral and systemic host inflammatory responses in the pathogenesis of coma and death in malaria deserve further investigation.