It’s all about perception : nod factor perception inside nodules of Medicago truncatula

Legumes are unique in that they are able to establish a mutual symbiotic interaction with nitrogen fixing soil bacteria generally referred to as rhizobia. This interaction starts off in the root epidermis where the bacterial signal molecule, the Nod factor, is perceived by the plant (Nod factor signaling). This recognition sets in motion a series of responses leading to the formation of a root nodule. This organ is specifically created to host the bacteria in an intracellular manner. The rhizobia develop into a mature, nitrogen fixing state in these infected cells. The rhizobia are surrounded by a host membrane and this forms a cell wall free symbiotic interface, that allows the exchange of nutrients between the symbionts. The comparison of the mechanism controlling symbiotic interface formation in rhizobium symbiosis and the much more common symbiosis with arbusucal micorrhizal (AM) fungi strongly suggest that rhizobia co-opted parts of the AM symbiosis. The signalling as well as cellular processes controlling symbiotic interface formation in the ancient AM symbiosis have been recruited by the rhizobium nodule symbiosis. In this thesis I present the results of my research on the role of and mechanisms controling Nod factor signalling on symbiotic interface formation in nodules of the model legume Medicago truncatula (Medicago).

To study the role of Nod factor signaling in the formation of a symbiotic interface it was essential to define a fate map for Medicago root nodules. The formation of a nodule starts, after Nod factor perception in the epidermis, with divisions in the cortex, pericycle and endodermis. Devisions in the cortex start from the inner most layer and spread outwards. When cortical layer 4 (C4) and C5 have already divided a few times, mitotic activity is induced in C3 and ultimately C2 divides. After the first periclinal divisions have been induced in C3, cells derived from C4 and C5 stop dividing and form about 8 cell layers. C3 continues to divide and ultimately forms the nodule meristem. The infection thread that developed in the root hairs has to penetrate the primordium before the first periclinal divisions occur in C3. The cells derived from C4 and C5 are then infected by bacteria from the infection thread in the primordium. After the establishment of a meristem this meristem adds new cells to the nodule, which are gradually infected when they leave the meristem and enter the infection zone. The fate map shows that formation of the symbiotic interface occurs in two ways: first in cells of the primordium and later in the infection zone in daughter cells derived from the meristem.

For the infection of daughter cells derived from the meristem the Nod factor receptors (NFP and LYK3 in Medicago) are needed, most likely to perceive the Nod factor. When we knocked down one of the receptors in a nodule specific manner release of bacteria is hampered (NFP) or massive unwalled droplets were observed (LYK3) indicating most likely slow release. This shows that the receptors are needed for the formation of the symbiotic interface in the nodule. In line with this function in the infection zone we found the receptors to accumulate in the nodule apex. The receptors accumulate in a narrow zone of two cell layers that is the border between the merisitem and the infection zone. The receptors accumulate to a zone markedly narrower than the zone where the promoters are activity. In this narrow zone the receptors accumulate at the cell periphery, most likely the plasma membrane. Outside this layer we observed accumulation of the receptors in the vacuoles suggesting degradation. In cells with LYK3 at the plasma membrane LYK3 occurs only in 35% of the cells at the membrane surrounding infection threads. The removal of LYK3 from the plasma membrane appears to be first completed at the membrane surrounding the infection thread. As the receptors are expressed in a broader zone, post-transcriptional mechanisms limit their accumulation at the plasma membrane. Ectopic expression experiments show that broader expression, and most likely accumulation, could induce defence responses (NFP) and reduces infection (LYK3). Therefore the accumulation of the receptors needs to be limited and the receptors are only allowed to accumulate in a narrow zone where Nod factor perception could take place.

As both receptors accumulate in the same cell layers we tested whether they form heteromeric complexes. It was already proposed at their discovery that NFP and LYK3 should form a complex based on their phenotype and the lack of an active kinase domain in NFP. Also most receptors form complexes to modulate their (kinase) activity. We show that a heteromeric complex is also formed in the proper biological context in nodules. The cell death responses that are induced when the receptors are co-expressed in heterologous systems are avoided in Medicago nodules. Also homomeric complexes containing LYK3 were observed in nodules. This homomeric complex is formed either as an addition to the heteromeric complex with NFP or the complex with NFP contains multiple LYK3 molecules.

As receptors often function in larger complexes we performed immunoprecipitation coupled with mass spectrometry to detect possible interactors of LYK3. With this we detected several proteins interactors of LYK3 including a EF1α, Dynamin-2B, PR10, fructose-bisphosphate aldolase and 14-3-3 f2 protein. Some of these proteins have a function in the regulation of vesicle transport and the cytoskeleton. Remarkable, other known or putative interactors of LYK3 (NFP, PUB1, SYMREM1 and FLOT4) were not detected. To test why we could not find these interactors we performed co-localization experiments. These experiments show that the interaction with SYMREM1 and the putative interaction with FLOT4 could take place as both proteins co-localize with LYK3. SYMREM1 accumulates at the membrane surrounding the infection threads and at the symbiosomes. As LYK3 also, although transient, accumulates at the membrane surrounding the infection threads, LYK3 and SYREM1 co-localize there. This shows that these interactors could form a complex with LYK3, although the complex is most likely very transient. This transient nature makes the complexes difficult to detect.

Because co-localization is a prerequisite for two proteins to interact we studied the expression of the Medicago genome. We used laser capture micro-dissection experiment to isolate RNA from different zones of the nodules and measured the differential expression of the Medicago genome. This experiments provides us with a digital in situ experiment for Medicago nodules. The data show that the genes from the Nod factor signaling cascade do not show the same expression pattern. The Nod factor receptor mRNAs of NFP and LYK3 are enriched in the meristem and distal infection zone. This is in line with our localization studies where the receptors accumulate at the border between meristem and infection zone. Most genes from the Nod factor signaling cascade show no differential expression at all (expressed equal in the entire nodule), where in situ or promoter GUS studies show enrichment of all these genes in the meristem/ infection zone. Further, some show remarkable behaviour. DMI1 and IPD3, for instance, are enriched in infected cells. These data could point to a new function for these proteins which is not related to the known signaling cascade. Also a survey on the differential expression of other LysM domain containing proteins show that there are more candidates that may have an important role in Nod factor perception. These LysM domain containing proteins are also enriched in the mersitem and/or infection zone and could thus function as a co-receptor for NFP or LYK3.

Nod factor perception is the key step in the progress of the rhizobium – legume symbiosis. Not only in the early steps in the root epidermis, but also in the nodule. Because of the developmental gradient in the indeterminate Medicago nodule this nodule is a perfect biological system to study the cell biology of the symbiosis. The differential expression analysis and the nodule fate map are important tools for these studies.