Macrophage microbicidal mechanisms in vivo: reactive nitrogen versus oxygen intermediates in the killing of intracellular visceral Leishmania donovani

Abstract

To determine the relative contributions of respiratory burst-derived reactive oxygen intermediates (ROI) versus reactive nitrogen intermediates (RNI) to macrophage-mediated intracellular host defense, mice genetically deficient in these mechanisms were challenged with Leishmania donovani, a protozoan that selectively parasitizes visceral tissue macrophages. During the early stage of liver infection at wk 2, both respiratory burst-deficient gp91(phox)-/- (X-linked chronic granulomatous disease [X-CGD]) mice and inducible nitric oxide synthase (iNOS) knockout (KO) mice displayed comparably increased susceptibility. Thereafter, infection was unrestrained in mice lacking iNOS but was fully controlled in X-CGD mice. Mononuclear cell influx into infected liver foci in X-CGD and iNOS KO mice was also overtly impaired at wk 2. However, granuloma assembly in parasitized tissue eventually developed in both hosts but with divergent effects: mature granulomas were functionally active (leishmanicidal) in X-CGD mice but inert in iNOS-deficient animals. These results suggest that (a) ROI and RNI probably act together in the early stage of intracellular infection to regulate both tissue recruitment of mononuclear inflammatory cells and the initial extent of microbial replication, (b) RNI alone are necessary and sufficient for eventual control of visceral infection, and (c) although mature granulomas have traditionally been associated with control of such infections, these structures fail to limit intracellular parasite replication in the absence of iNOS.

Figure 1

Course of L. donovani liver infection in (A) X-CGD (○) and control C57BL/6 mice (•), and in (B) iNOS KO mice (○) and wild-type controls (•). Results are mean ± SEM values for 8–15 mice at each time point from two to four experiments. Asterisk, significantly higher (P < 0.05) than value in control mice.

Figure 3

Histologic response in livers of infected iNOS KO mice and wild-type controls. At wk 2, granulomas are widespread and developed in controls (A) but absent at infected foci (arrows) in iNOS KO mice (B). At wk 4, granulomas are well-established or empty in control mice (C and E); in iNOS KO mice (D and F), granulomas are developing but heavily parasitized. At wk 8, inflammatory reaction in controls has involuted (G); in iNOS KO mice (H), granulomas are well-formed (mature) but contain numerous replicating amastigotes (see also Fig. 4). Original magnification: A, C, and D, ×200; B and E–H, ×315.

Figure 4

The ineffective granuloma. Photomicrograph shows well-developed tissue structure in 8-wk-infected iNOS KO liver containing a striking load of intracellular amastigotes. Original magnification: ×500.

Figure 2

Histologic response to L. donovani infection in the liver in X-CGD versus C57BL/6 mice. 2 wk after challenge, granulomas are developing at infected foci (arrows) in C57BL/6 controls (A); in X-CGD mice (B), there is little or no reaction at well-parasitized foci (arrows). At wk 4, granulomas in C57BL/6 mice (C) in this field are mature and largely parasite free; (D) shows developing granulomas in X-CGD mice. At wk 8, (E) shows receding granulomas in C57BL/6 mice, whereas (F) illustrates intense, mature reaction in X-CGD mice. Original magnification: A and B, ×315; C–F, ×200.