Intracellular trafficking of the Coxiella burnetii lipopolysaccharide

Abstract

Coxiella burnetii is known to survive in monocytes ⁄macrophages and is the causative agent of Q fever. It undergoes an irreversible transition from virulent (phase I) to avirulent (phase II) forms that resembles smooth-to-rough transition in Enterobacteria [1]. C. burnetii inhibits phagosome conversion in professional phagocytes, while the phase II variants are eliminated within a phagolysosome [1]. The nature of bacterial components that inhibit phagosome–lysosome fusion has to be established. There are several indications that the C. burnetii lipopolysaccharide (LPS) could be a candidate molecule as it has long been considered a major determinant of virulence of the bacterium [2]. In addition, it plays an important role in the phase variation of the bacterium [1]. LPS I expressed in the phase I bacteria is of smooth (S) type and contains an O-polysaccharide chain [3] in contrast with LPS II present in the phase II bacteria that is of deeprough (R) type [3]. In LPS I, there are considerable amounts of two unique sugars, virenose and dihydrohydroxystreptose [3], in its O-polysaccharide chain, mainly in terminal positions. In this study, we elucidated the intracellular trafficking of Nine Mile LPS I and LPS II in murine bone marrow-derived macrophages (BMDM). The intracellular distribution of both LPSs with several markers of endocytic compartments was studied. We found that intracellular trafficking of LPS I was considerably similar to that of LPS II. BMDM were incubated with LPS I (1 mg ⁄L) and LPS II (1 mg ⁄L) [3] for 1 h at 4 C, followed by different periods of chase at 37 C. The co-localisation of LPSs with EEA1 (Early Endosome Antigen-1), a marker of early endosomes, and with Lamp-1 (Lysosomal Associated Membrane Protein-1), a marker of late endosomes and lysosomes, was determined by the quantification of immunofluorescence staining after the confocal microscopy. About 22% of LPS II co-localised with EEA-1 after 10 min; this percentage reached a maximum at 60 min and decreased thereafter (Fig. 1a). LPS II co-localised with Lamp-1. Indeed, the percentage of LPS II that co-localised with Lamp-1 was 33.6 ± 9% after 10 min, increased to 36.2 ± 3% at 30 min, reached a maximum of 66.5 ± 11% at 60 min, and decreased thereafter (Fig. 1b). It is evident that more than 40% of the internalised LPS II did not co-localise with the endosomal markers, suggesting that a considerable part of LPS II did not transit through early endosomes and the degradative pathway. Taken together, these data indicate that LPS II was rapidly internalised in early endosomes and transported thereafter to late endosomes and ⁄ or lysosomes. This time course is similar to that of fluorescein-conjugated albumin [4] and Shigella flexneri LPS [5]. The kinetics of LPS II trafficking differs from that of LPS of Brucella abortus as LPS II was localised in late compartments in 60 min while at least 5 h were required for the Brucella LPS [2]. In addition, when LPS II used the endosomal route, its trafficking was quite similar to that of the phase II C. burnetii [1]. LPS I co-localised with EEA1, and the percentage of co-localisation significantly decreased at 120 min (Fig. 1c). LPS I co-localised strongly with Lamp-1 (Fig. 1d). Indeed, the percentage of LPS I that co-localised with Lamp-1 was 33 ± 7% after 10 min, increased and reached a maximum of 60 ± 8% at 120 min, and decreased thereafter (Fig. 1d). Again, an important portion (40%) of LPS I did not co-localise with the endosomal markers, demonstrating that a considerable part of this LPS did not transit through both early endosomes and the degradative pathway. Trafficking of LPS I is similar to fluoresceinconjugated albumin [4], B. abortus LPS [2] and Corresponding author and reprint requests: E. Ghigo, URMITE CNRS UMR 6236 – IRD 3R198, Faculte de Medecine, 27 Bld J. Moulin, 13385 Marseille Cedex 05, France E-mail: eric.ghigo@univmed.fr