The formation of endosymbiotic membrane compartments: membrane identity markers and the regulation of vesicle trafficking

In symbiosis of plants and arbuscular mycorrhizal fungi as well as in rhizobium-legume symbiosis the microbes are hosted intracellularly, inside specialized membrane compartments of the host. These membrane compartments are morphologically different but similar in function, since they control the exchange of compounds between host and its microsymbiont thus forming a highly specialized symbiotic interface. These are the arbuscules, containing highly branched fungal hyphae, and organelle-like symbiosomes containing rhizobium bacteria. Recent studies have markedly extended our insight in the evolution of the signaling mechanism underlying the formation of these symbiotic interfaces. These studies strongly suggest that rhizobium co-opted the complete signaling mechanism (including lipo-oligosaccharides signal molecules) from the more ancient AM fungi symbiosis. Further, in plant species (Parasponia) where rhizobium nodulation evolved rather recent and independent from legumes, the same lipo-oligosaccharide receptor is essential for the formation of the rhizobium symbiotic interface as well as arbuscules. Therefore it seems likely that rhizobium also co-opted the cellular mechanism controlling arbuscule formation to form a rhizobium symbiotic interface. This would imply that even after co-evolution in legumes the key regulators involved in the formation of these interfaces are similar or even identical. In this thesis I have shown that rhizobium symbiosis shares with AM symbiosis molecular and cell biological mechanisms that control symbiotic interface formation. I identified a plant exocytotic pathway marked by two highly homologous vesicle associated membrane proteins (VAMP) that control the formation of the symbiotic interface in both symbioses. RNAi of these two Medicago VAMP genes did not affect non-symbiotic plant development nor nodule formation. However, it hampered the formation of cell wall free regions at infection threads, and therefore blocks symbiosome formation. Further arbuscule formation was blocked, whereas root colonization was not affected. By identifying these VAMPs as common symbiotic regulators in secretory vesicle trafficking, I postulated that during evolution of rhizobium symbiosis pre-existing cellular mechanisms of the AM fungal symbiosis have been co-opted. These findings also revealed a primary role of exocytosis in symbiosome formation and allowed to postulate the apoplastic nature of symbiosome. Using identity markers of endocytotic compartments of plant cell (early endosome and late endosome) such as small GTPases belonging to the Rab family and SNARE (soluble N-ethylmaleimide sensitive factor attachment protein receptor) proteins, I have shown that they never occur on symbiosome membranes at any stage of symbiosome formation and development. This makes untenable long-standing hypothesis that symbiosomes originate from endocytosis-like process and represent endocytic (vacuolar) compartments. Instead symbiosomes have an apoplastic nature. Although symbiosomes have an apoplastic nature they acquire the vacuolar marker MtRab7 when they reach an elongated stage. However, vacuolar SNAREs which execute fusion of membranes are not present on functional symbiosomes, but they do appear on symbiosome membranes at the onset of senescence when symbiosomes are turned into a lytic compartment. Therefore I postulate that the acquisition of Rab7 primes the symbiosomes for degradation by the host. By this the host has full control over its microsymbiont. The finding that rhizobium symbiosis has co-opted the signaling mechanism as well as cellular mechanism from AM fungi symbiosis to facilitate an intracellular life style, has major implications for strategies to transfer the nodule symbiosis to non-legume crops. This is a “dream” that is already about a century old. The AM fungal symbiosis is far more ancient than the rhizobial symbiosis. It is also wide spread in the plant kingdom and almost 80% of plant species can establish an AM symbiosis. This implies that plants which are able to interact with AM fungi contain in principle the genes that are necessary for the intracellular accommodation of rhizobium. So the question is no longer why the rhizobium-legume symbiosis is specific for legumes, but why non-legumes are not yet able to establish this symbiosis?