The malaria sporozoite, the parasite stage transmitted by the mosquito, is delivered into the dermis and differentiates in the liver. Motile sporozoites can invade host cells by disrupting their plasma membrane and migrating through them (termed cell traversal), or by forming a parasite-cell junction and settling inside an intracellular vacuole (termed cell infection). Traversal of liver cells, observed for sporozoites in vivo, is thought to activate the sporozoite for infection of a final hepatocyte. Here, using Plasmodium berghei, we show that cell traversal is important in the host dermis for preventing sporozoite destruction by phagocytes and arrest by nonphagocytic cells. We also show that cell infection is a pathway that is masked, rather than activated, by cell traversal. We propose that the cell traversal activity of the sporozoite must be turned on for progression to the liver parenchyma, where it must be switched off for infection of a final hepatocyte.

Sporozoites, the invasive form of malaria parasites transmitted by mosquitoes, are quiescent while in the insect salivary glands. Sporozoites only differentiate inside of the hepatocytes of the mammalian host. We show that sporozoite latency is an active process controlled by a eukaryotic initiation factor-2α (eIF2α) kinase (IK2) and a phosphatase. IK2 activity is dominant in salivary gland sporozoites, leading to an inhibition of translation and accumulation of stalled mRNAs into granules. When sporozoites are injected into the mammalian host, an eIF2α phosphatase removes the PO4 from eIF2α-P, and the repression of translation is alleviated to permit their transformation into liver stages. In IK2 knockout sporozoites, eIF2α is not phosphorylated and the parasites transform prematurely into liver stages and lose their infectivity. Thus, to complete their life cycle, Plasmodium sporozoites exploit the mechanism that regulates stress responses in eukaryotic cells.

We describe here an efficient method for conditional gene inactivation in malaria parasites that uses the Flp/FRT site-specific recombination system of yeast. The method, developed in Plasmodium berghei, consists of inserting FRT sites in the chromosomal locus of interest in a parasite clone expressing the Flp recombinase via a developmental stage-specific promoter. Using promoters active in mosquito midgut sporozoites or salivary gland sporozoites to drive expression of Flp or its thermolabile variant, FlpL, we show that excision of the DNA flanked by FRT sites occurs efficiently at the stage of interest and at undetectable levels in prior stages. We applied this technique to conditionally silence MSP1, a gene essential for merozoite invasion of erythrocytes. Silencing MSP1 in sporozoites impaired subsequent merozoite formation in the liver. Therefore, MSP1 plays a dual role in the parasite life cycle, acting both in liver and erythrocytic parasite stages.

SummaryInoculation of Leishmania (L.) spp. promastigotes in the dermis of mammals by blood-feeding sand flies can be accompanied by the rapid recruitment of neutrophils, inflammatory monocytes and dendritic cells. Despite the presence of these lytic leucocytes, parasitism is efficiently established. We show here that Leishmania donovani promastigotes are targeted to two different compartments in neutrophils. The compartments harbouring either damaged or non-damaged parasites were characterized at the electron microscopy (EM) level using the glucose 6-phosphatase cytochemistry and endosomephagosome fusion assays. One involves the contribution of lysosomes leading to the formation of highly lytic compartments where parasites are rapidly degraded. The other is lysosome-independent and involves the contribution of a compartment displaying some features of the endoplasmic reticulum (ER) where parasites are protected from degradation. Using genetically modified parasites, we show that the promastigote surface lipophosphoglycan (LPG) is required to inhibit lysosome fusion and maintain parasites in neutrophil compartments displaying ER features. L. donovani-harbouring neutrophils that eventually enter apoptosis can be phagocytosed by macrophages enabling the stealth entry of parasites into their final replicative host cells. Thus, the ability of L. donovani to avoid trafficking into lysosomesderived compartments in short-lived neutrophils constitutes a key process for the subsequent establishment of long-term parasitism.

The first step of Plasmodium development in vertebrates is the transformation of the sporozoite, the parasite stage injected by the mosquito in the skin, into merozoites, the stage that invades erythrocytes and initiates the disease. The current view is that, in mammals, this stage conversion occurs only inside hepatocytes. Here, we document the transformation of sporozoites of rodentinfecting Plasmodium into merozoites in the skin of mice. After mosquito bite, ∼50% of the parasites remain in the skin, and at 24 h ∼10% are developing in the epidermis and the dermis, as well as in the immunoprivileged hair follicles where they can survive for weeks. The parasite developmental pathway in skin cells, although frequently abortive, leads to the generation of merozoites that are infective to erythrocytes and are released via merosomes, as typically observed in the liver. Therefore, during malaria in rodents, the skin is not just the route to the liver but is also the final destination for many inoculated parasites, where they can differentiate into merozoites and possibly persist.intravital imaging | Plasmodium | schizogony M alarial infection starts with the inoculation of Plasmodium sporozoites by mosquitoes probing the vertebrate skin for blood. The highly motile sporozoites eventually invade host target cells where they differentiate and divide into numerous merozoites, the parasite form that invades erythrocytes and initiates the pathogenic phase of malarial infection. The host cell type in which sporozoites transform into merozoites, however, differs between Plasmodium species. In species that infect birds, sporozoites differentiate inside macrophages primarily in the skin but also in the spleen, liver, and bone marrow (1). In species that infect mammals, sporozoites are known to differentiate only inside hepatocytes in the liver (2-4).The first demonstration that sporozoites of mammal-infecting Plasmodium species develop inside hepatocytes was made in 1948 after i.v. inoculation of sporozoites of P. cynomologi into rhesus monkeys (2). In addition to reporting fully mature parasites inside hepatocytes, the authors also documented the persistence of immature and dormant forms of the parasite in the liver several months after the initial inoculation, which they proposed to be the cause of relapses (5), and were later called hypnozoites (6). Subsequent work indicated that sporozoites of species that infect humans (7) also undergo complete development inside hepatocytes.Since these early studies, P. berghei and the related P. yoelii species, which infect rodents, have been used as practical and safe models for studying the pre-erythrocytic phase of malaria. These parasites were shown to differentiate in the liver of laboratory rodents (8), and the P. berghei/rodent system was used to demonstrate that the majority of sporozoites were inoculated by mosquitoes in the skin rather than directly into the blood circulation (9), as traditionally assumed. More recently, the generation of fluorescent P. berghei parasites, alo...

Within a period of 2 1/2 years, Bordetella bronchiseptica was isolated four times from a 79-year-old woman with bronchopneumonia. We have demonstrated by pulsed-field gel electrophoresis that this infection was related to contact with infected rabbits. The initial human B. bronchiseptica isolate had a phenotype characteristic of usual B. bronchiseptica clinical isolates; it produced toxin and adhesins, such as adenylate cyclasehemolysin, filamentous hemagglutinin, and pertactin, and was able to induce lethality in a murine respiratory model. By contrast, although the three successive human isolates produced adhesins, they did not express adenylate cyclase-hemolysin and were unable to induce lethality. This implies that adenylate cyclase-hemolysin is required to induce lethality. We suggest that B. bronchiseptica may persist in the host, with expression of adenylate cyclase-hemolysin being essential for the initiation of infection and expression of adhesins being essential for persistence.

Although France has had a vaccination program for 40 years, since 1990, an increase in whooping cough cases with parent-infant transmission has been observed. This study prospectively assessed the frequency of Bordetella pertussis infection in adults who consulted general practitioners for a persistent cough without an evident diagnosis. Among 217 patients, 70 (32%) confirmed whooping cough cases were identified. One case was culture positive, 36 were polymerase chain reaction positive, and 40 had increases or decreases of > or =2-fold in anti-pertussis toxin IgG titer between serum samples collected during the acute and convalescent phases. The median duration of cough in confirmed cases was 49 days (range, 13-123 days). Of the patients, 60% reported vaccination, and 33% reported whooping cough in infancy. Pertussis should be considered for diagnosis of acute and chronic cough in adults. Future studies should evaluate the public health interest of booster doses of pertussis vaccine in adults.

Bordetella pertussis, the agent of whooping cough, can invade and survive in several types of eukaryotic cell, including CHO, HeLa 229, and HEp-2 cells and macrophages. In this study, we analyzed bacterial invasiveness in nonrespiratory human HeLa epithelial cells and human HTE and HAE0 tracheal epithelial cells. Invasion assays and transmission electron microscopy analysis showed that B. pertussis strains invaded and survived, without multiplying, in HTE or HAE0 cells. This phenomenon was bvg regulated, but invasive properties differed between B. pertussis strains and isolates and the B. pertussis reference strain. Studies with B. pertussis mutant strains demonstrated that filamentous hemagglutinin, the major adhesin, was involved in the invasion of human tracheal epithelial cells by bacteria but not in that of HeLa cells. Fimbriae and pertussis toxin were not found to be involved. However, we found that the production of adenylate cyclase-hemolysin prevents the invasion of HeLa and HTE cells by B. pertussis because an adenylate cyclase-hemolysin-deficient mutant was found to be more invasive than the parental strain. The effect of adenylate cyclase-hemolysin was mediated by an increase in the cyclic AMP concentration in the cells. Pertactin (PRN), an adhesin, significantly inhibited the invasion of HTE cells by bacteria, probably via its interaction with adenylate cyclase-hemolysin. Isolates producing different PRNs were taken up similarly, indicating that the differences in the sequences of the PRNs produced by these isolates do not affect invasion. We concluded that filamentous hemagglutinin production favored invasion of human tracheal cells but that adenylate cyclase-hemolysin and PRN production significantly inhibited this process.