Abstract
Interactions with target membrane molecules and the resultant internalization into host cells are, in general, the first and foremost steps for subsequent host responses and pathogenesis of infectious micro-organisms. For pathogenic bacteria belonging to both spotted fever and typhus subgroups of Rickettsia species, which are known for dependence on the nutrient-rich environment of the host cytoplasm and fastidious growth requirements, target cell invasion is absolutely critical for subsequent intracellular replication and intercellular spread. Rickettsia rickettsii, the causative agent of Rocky Mountain spotted fever, primarily affects the endothelial cell lining of small- and medium-sized vessels leading to disseminated intracellular infection of the vasculature and, consequently, characteristic pathological features during human infections correspond to vascular inflammation, damage and dysfunction (Walker, 2007).
Initial studies using Rickettsia prowazekii, the aetiological agent of epidemic typhus, revealed that inactivation of bacteria or alterations in the host cell cytoskeleton adversely affect rickettsial uptake into cultured human umbilical vein-derived endothelial cells, suggesting that the viability of both the host cell and adherent rickettsiae is an essential prerequisite for successful invasion via induced phagocytosis (Walker, 1984). Subsequent investigations by electron microscopy further indicated that rickettsial entry into Vero cells is almost instantaneous (occurring within 3 min of bacterium–host cell contact) and rickettsiae are able to quickly escape from the phagosome into the cytoplasm, presumably prior to phagolysosomal fusion (Teysseire et al., 1995), and likely via a phospholipase activity (Whitworth et al., 2005). An important role for host actin polymerization in rickettsial internalization of non-phagocytic cells has also been suggested. Analysis of proteins known to govern actin dynamics has identified recruitment of the Arp2/3 complex to the sites of entry foci during Rickettsia conorii invasion of Vero cells, and potential coordinated involvement of GTPase Cdc42, phosphoinositide 3-kinase, c-Src and other protein tyrosine kinase activities in the activation of the Arp2/3 complex (Martinez & Cossart, 2004).
Conserved across all eukaryotes, the p38 group of mitogen-activated protein (MAP) kinases is implicated in many physiological processes, including cell growth, differentiation, apoptosis and immune response. A unique feature for activation of the p38 superfamily of enzymes is the requirement for dual phosphorylation of the TXY motif in the activation loop of the catalytic domain, where phosphorylation of tyrosine (Y) usually precedes that of threonine (T) in a two-step reaction. Our laboratory has demonstrated that in vitro infection of cultured human endothelial cells with spotted fever rickettsiae (R. rickettsii and R. conorii) triggers the activation of the p38 signalling module, the specific inhibition of which results in downregulation of infection-induced expression of chemokines [interleukin-8 (CXCL8) and monocyte chemoattractant protein-1 (CCL-2)], and downregulation of the inducible isoform of regulatory bifunctional enzyme cyclooxygenase-2 (Rydkina et al., 2005, 2006). In this report, we have further investigated the possibility of an interrelationship between p38 MAP kinase activation and R. rickettsii invasion of host endothelial cells.
Primary cultures of endothelial cells isolated from the umbilical vein of freshly collected human umbilical cords were established and maintained according to previously described procedures (Sahni et al., 1998). Vero C1008 cells were obtained from American Type Culture Collection and maintained in culture as recommended by the supplier. The Sheila Smith strain of R. rickettsii was grown in Vero cells and purified on sucrose-Renografin density gradients (Sahni et al., 1998; Rydkina et al., 2005). The pyridinyl imidazole SB 203580 was used as a selective inhibitor of p38 MAP kinase. This compound does not inhibit other MAP kinase homologues c-Jun-N-Kinase (JNK) and p42 MAP kinase, but has previously been shown to specifically inhibit R. rickettsii-induced activation of p38 (Rydkina et al., 2005). SB 202474, a structurally similar compound with no inhibitory activity towards p38 activity, was used as a negative control in all experiments. To determine the extent of rickettsial invasion into host cells, we employed three independent approaches: immunofluorescent staining of rickettsiae followed by quantification of the percentage of cells infected and the number of intracellular bacteria (Sahni et al., 1998), plaque formation assay (Rydkina et al., 2005) and citrate synthase gene (gltA)-based real-time PCR (Rydkina et al., 2007). Plaque formation assays were carried out using Vero cell monolayers, whereas indirect immunofluorescence and quantitative PCR-based enumeration was conducted using human endothelial cells. The durations for inhibitor treatment as well as R. rickettsii infection for both cell types were identical; each experimental condition was repeated to achieve a minimum of three replicates, and results obtained were subjected to statistical evaluation as described previously (Rydkina et al., 2005, 2007).
First, we compared the level of infection in endothelial cells grown to confluence on sterile coverslips and incubated for 3 h with R. rickettsii in the presence and absence of 10 µM SB 203580 (SB 80) and 10 µM SB 202474 (SB 74) in the culture medium. Indirect immunofluorescent staining using an anti-R. rickettsii antiserum and a rhodamine-conjugated compatible secondary antibody revealed that in the presence of specific p38 inhibition via SB 80 the number of adherent/intracellular rickettsiae was comparatively lower than in those endothelial cells infected in culture medium alone or in the presence of SB 74, an inactive structural analogue of SB 80 (Fig. 1). The extent of plaque formation on confluent monolayers of Vero cells was next determined under similar experimental conditions. After 3 h exposure to viable R. rickettsii, which was followed by thorough washing and layering of the cell monolayer with 0.5 % agarose in culture medium, visibly distinct plaques resulted due to infection-induced lysis of host cells on day 7 post-infection. Again, cell treatment with SB 74 during the process of initial contact and invasion had no effect on the ability to induce plaque formation. Presence of SB 80 during infection, however, led to a significant decrease (36±7 %) in the number of rickettsial plaques (P=0.017, when compared to corresponding untreated controls and SB 74-treated cells) (Fig. 2). Next, the number of intracellular rickettsiae in endothelial cells infected in the presence of p38 inhibition was quantified by a real-time PCR-based approach using a rickettsial gltA-specific primer pair and a TaqMan probe to determine the mean copy number of gltA in relation to that of the housekeeping gene GAPDH. Again, treatment of endothelial cells with SB 80 significantly reduced the invasion of R. rickettsii (38±12 % inhibition in comparison to cells infected without any treatment; P=0.025). Infection in the presence of SB 74, a negative control for SB 80, had no significant effect on rickettsial invasion when compared to controls (17±13 % inhibition; P=0.21) (Fig. 2). Thus, when directly compared to infection in the presence of SB 74, SB 80 caused an approximately 25 % decrease in the invasion of R. rickettsii. Taken together, these results suggest that p38 MAP kinase activation in response to infection plays an important role in facilitating host cell invasion by R. rickettsii.