SUMMARYPla.smodium vivax malaria infeetions in non-immune individuals manifest as periodic clinical episodes of fever with chills and rigors known as paroxysms. We have demonstrated that in nonimmune patienls the period of paroxysm is associated with the transient presence of plasma factors which kill gametocytes. the in tra-erythrocy tic sexual stages ofthe malaria para.sitc which transmit the infection from humans to mosquito, rendering them non-infectious to mosquitoes. Gametocytc killing in paroxysm plasma is mediated by tumour necrosis faclor (TNF) acting in conjunction with other essential serum factor(s). Plasma TNF levels were elevated during a paroxysm. In semi-immune individuals from a P. r/ra-V-endemic area clinical symptoms of malaria are mild and the parasite killing factors are not induced during paroxysm. Scrum TNF levels were correspondingly lower in endemic palients during a paroxysm. Human peripheral blood mononuclear cells (PBMC) can be stimulated in vitro by extracts of P. viva.x blood stage parasites to produce TNF and associated parasite killing faclor(s), thus simulating in vitro the events that occur during a paroxysm, this being the release ofparasiteexo-aniigens by rupturing schizonts and the subsequent induction of PBMC lo produce TNF and other parasile-killing factors. We were able to show that convalescent serum from P. ['(t'«x semi-immune individuals block Ihe induction of TNF and parasite-killing factors by malaria antigens in viiro. presumably through anlibodies ihat neutralize parasite exo-aniigens. Thus, individuals living in malaria-endemic areas appear to acquire clinical immunity lo malaria by avoiding their induction during infeetion; we have .shown thai one such mechanism is the neutralization of parasite exo-anligens that induce the production of parasite killing factors. Keywords tumour necrosis factor malaria paroxysms gametocytes clinical immunity INTRODLiCTIONNaturally acquired immunity lo malaria in humans manifests as an anli-parasite immunity which results in a lowering of the parasilaemia, and as clinical or anti-disease imtiiunity which leads to decrea.sed clinical symptoms, or 'tolerance' lo the parasite [I,2J. The two types of immunity are dislincl. because in highly malaria-endemic areas of the world individuals may harbour high parasitaemias and yet be clinically well (1.3], Faclor{s)causingthec!inical disease in miilaria and immunity to it arc poorly understood, Cytokines have been implicated lo play a role in malarial infections, as mediators of immunopalhology [4][5][6][7][8][9][10] and In parasite killing [11][12][13][14][15][16]. We have studied natural human infections of P. vivax malaria in Sri Lanka. Acuic human malarial infections are characterized by paroxysms, clinical episodes of fever with chills and rigors lasting a few hours. Paroxysms are co-ordinated with the synchronous rupture and release ofthe asexual blood stage parasites from their host erythrocytes. Each paroxysm in a nonimmune individual begins with a sharp rise of temperature accompani...

Infection of CBA mice with Plasmodium berghei ANKA results in severe malaria, which is characterized by mortality 6 to 10 days after infection and is associated with alterations of the brain microcirculation. These alterations consist of (i) intravascular sequestration of monocytes, (ii) an increase in vascular permeability as documented by Evans blue diffusion, and (iii) microhemorrhages. This syndrome may be due to an increase of production of tumor necrosis factor alpha which upregulates the endothelial expression of ICAM-1 and thus leads to adhesion of CDlla/CD18 (LFA-l)-bearing cells. During severe malaria, we found an important sequestration of the CDlla-bearing polymorphonuclear neutrophil leukocytes (PMN) in the lung but not in the brain. Treatment with a monoclonal antibody (MAb) against PMN, which induces profound neutropenia, prevented mortality and Evans blue diffusion in the brain and the lung, while it unexpectedly increased the occurrence of microhemorrhages. The anti-PMN MAb abolished PMN sequestration in the lung and also partially decreased monocyte sequestration in the brain and the lung. Treatment with an anti-CD1la MAb also prevented mortality, Evans blue diffusion, and PMN and monocyte sequestration. This study shows that PMN contribute to the mortality and the microvascular lesions resulting from severe malaria. This may be due to their CDlla-dependent sequestration in the lung and also to their indirect influence on vascular permeability and the sequestration of monocytes.

The production of tumor necrosis factor alpha (TNF-␣), interleukin-1␤ (IL-1␤), and IL-6 and their pharmacomodulation were evaluated in a model of polymicrobial sepsis induced in mice by cecal ligation and puncture (CLP) and were compared with the effects of endotoxin (lipopolysaccharide [LPS]) treatment. LPS levels rose as early as 1 h after CLP and increased further after 2 and 21 h. TNF-␣ was detectable in serum, spleen, liver, and lungs during the first 4 h, with a peak 2 h after CLP. IL-1␤ was measurable in serum after 24 h, and levels increased significantly in spleen and liver 4 and 8 h after CLP. IL-6 levels increased significantly in serum throughout the first 16 h after CLP. These cytokines were detectable after LPS injection, with kinetics similar to those after CLP but at a significantly higher level. To cast more light on the differences between these two animal models of septic shock, we studied the effects of different reference drugs. Pretreatment with dexamethasone (DEX); ibuprofen (IBU), an inhibitor of cyclooxygenase; and N G-nitro-L-arginine, an inhibitor of nitric oxide synthase, significantly reduced survival, while chlorpromazine (CPZ) and TNF did not affect it. Only the antibiotics and pentoxifylline significantly increased survival in mice with CLP. However, CPZ and DEX protected the mice from LPS mortality. On inhibiting TNF-␣ with DEX, CPZ, or pentoxifylline, survival was reduced, unchanged, and increased, respectively, and on increasing TNF-␣ with IBU and TNF, survival was decreased or unchanged, respectively, suggesting that the modulation of this cytokine does not play a significant role in sepsis induced by CLP, unlike treatment with LPS. The negative effects of IBU and N G-nitro-L-arginine suggest a protective role by prostaglandins and nitric oxide in sepsis induced by CLP.

In this study, we investigated whether the recently identified lectin-like domain of tumor necrosis factor (TNF) is implicated in its biological activities on mammalian cells. To this end, a mouse TNF (mTNF) triple mutant, T104A-E106A-E109A mTNF (referred to hereafter as triple mTNF), lacking the lectin-like affinity of mTNF for specific oligosaccharides, was compared with the wild-type molecule for various TNF effects in vitro and in vivo. The triple mTNF displayed a 50-fold-reduced TNF receptor 2 (TNFR2)-mediated bioactivity but only a 5-fold-reduced TNFR1-mediated bioactivity in vitro. The specific activity of the triple mutant on L929 fibrosarcoma cells was slightly reduced compared with that of the wild type. We subsequently assessed the systemic toxicity of triple versus wild-type mTNF, since TNFR2 is partially implicated in this activity. The triple mTNF had a significantly reduced toxicity compared with that of wild-type mTNF in vivo. Moreover, we compared the effects of the triple and the wild-type mTNFs in TNFR1-mediated phenomena, such as (i) induction of tolerance towards a lethal mTNF dose and (ii) protective activity in cecal ligation and punctureinduced septic peritonitis. No significant differences between the mutant and wild-type forms were observed. In conclusion, these results indicate that triple mTNF, lacking TNF's lectin-like binding capacity, has reduced systemic toxicity but retains the tolerance-inducing and peritonitis-protective activities of wild-type mTNF.