Legionella pneumophila achieved unhappy fame as the pathogen of legionnaires’ disease in humans. Nonetheless the bacteria’s preferred hosts are not Homo sapiens but single-celled amoebae. After being taken in by amoebae, these widely occurring bacteria form a membrane-enclosed compartment in the host cell, the “Legionella-containing vacuole”. This compartment changes from a closely fitting vacuole to a spacious one in which the bacteria multiply. The remarkable aspect of their passage through the amoebae is that the Legionella avoid being broken down by the host cell and leave it again unharmed. This achievement is closely connected with one of the Legionella’s transport systems, which is called “Icm/Dot”. However, the way in which the Icm/Dot system affects the host cell’s metabolism was largely unclear up to now.

The metabolism of phosphoinositides can be considered as a possible candidate because the phosphoinositides represent important signalling lipids in the formation of membrane-enclosed compartments. This was also suspected by Hubert Hilbi, SNF-Professor (funded by the Swiss National Science Foundation) (1). So he was not particularly surprised to discover that amoebae of the species Dictyostelium discoideum with mutations in their phosphoinositide metabolism responded in a special way to attack by Legionella. To gain a better understanding of the mechanism, he and his colleagues planned a series of experiments that was also intended to elucidate the direct connection between phosphoinositides and the Icm/Dot system. The results of the studies were published in the May issue of the scientific journal PLoS Pathogens (2).

Disadvantageous and useful host cell enzymes

In this work, Hilbi’s group discovered that the transition from tight-fitting to spacious vacuoles is not necessary for the multiplication of the Legionella. An important role in this is played by phosphatidylinositol-3 (PI-3) kinases, i.e. enzymes that attach a phosphate group to phosphoinositide substrates. Functional PI-3-kinases hinder the intracellular multiplication of Legionella, while at the same time they cause the more rapid uptake and more efficient breakdown of Legionella with a defective Icm/Dot system.

A vacuole measuring about two microns (green) and containing Legionella in Dictyostelium amoebae. It is apparent that the amoebae’s PI(4)P lipid (blue) is a constituent of the vacuoles. (Photo: Hubert Hilbi) large

However, amoebae also have molecules that the Legionella harness for their own purposes: the host cell lipid phosphatidylinositol-4-phosphate (PI(4)P) binds onto the vacuoles containing the Legionella the protein SidC which is transported by the Icm/Dot system, thus acting as a membrane anchor for the bacteria. This enables the vacuoles to interact with other membranes. In addition, since PI(4)P is a typical feature of the amoebae’s excretion pathway, the discovery is further evidence that this lipid possibly helps the Legionella to leave the host cell.

A necessary factor, but not a sufficient one

Although the Hilbi group researchers carried out certain experiments only in Dictyostelium amoebae, it is very likely that their insights are also valid for humans. This is because individual results, such as the occurrence of PI(4)P on vacuoles containing Legionella, were also transferable to macrophages, which are human host cells for Legionella. The intracellular multiplication mechanism is very similar in amoebae and in macrophages, and in this respect PI(4)P appears to represent a necessary factor for a successful attack by Legionella, although in Hilbi’s opinion not a sufficient one.

Talking of ‘successful’: although the ETH scientist hopes that his results will also help in the longer term to improve the fight against legionnaires’ disease, he points out that humans are themselves endangered mainly by inappropriately designed and badly maintained hot water systems. In certain circumstances it is possible for Legionella to reproduce so well in these technical water systems that humans can become infected, although this really constitutes an infection “dead end” for the Legionella.