Neuhaus, Jean-Marc

2007, Zava, Olivier, Neuhaus, Jean-Marc

Plant cells contain two or even three types of vacuoles: the lytic, the (seed) protein storage and vegetative storage vacuoles. Soluble vacuolar proteins are sorted through the secretory pathway to these vacuoles by three different routes, depending on different types of Vacuolar Sorting Determinants (VSD) and involving several types of receptors and vesicles. The AtRMR1 protein has been identified in cellular structures associated with the seed storage vacuole pathway (Jiang et al. 2000). Based on its localisation and homology to a known vacuolar receptor, it has been hypothesised to be a receptor protein for the C-terminal type of VSD (CtVSD) involved in sorting to the storage vacuole. The genome of Arabidopsis thaliana contains 5 genes homologous to AtRMR1. The main goal of this study was to test the involvement of AtRMR1 in vacuolar sorting and to test the specificity of the different AtRMR proteins for different known CtVSDs. To test the involvement of AtRMR proteins in vacuolar sorting we studied the effects of manipulations of these genes on targeting of vacuolar reporter proteins. I used two different models: A. thaliana leaf protoplasts and whole plants of insertional mutants from the SALK Institute collection. The protoplast model allowed me to study in vivo the effects on vacuolar sorting of versions of AtRMR1 with loss of function deletions or modified functions. I obtained interesting results with two of these constructs: one lacking the luminal VSD-binding domain (RMRΔlum) and one consisting of this soluble luminal domain retained in the ER by the addition of a HDEL peptide signal (RMRlumER). When overexpressed with GFP fused to the ssVSD of barley aleurain, the RMRΔlum construct showed a different pattern of fluorescence compared with the control: in RMRΔlum, a significant proportion of protoplasts showed a dot-like fluorescent pattern, whereas the fluorescence was more often found in the central vacuole or ER in the control. When I used either the CtVSD of tobacco chitinase or a short version of the barley aleurain ssVSD, I didn’t see a different fluorescent pattern with or without the dominant negative construct. To better estimate the effects of the dominant negative constructs, the level of secretion of enzymatic reporters was investigated. The reporters were α-amylase fused either to the CtVSD of tobacco chitinase, the CtVSD of barley lectin or the ssVSD of sweet potato sporamin. For these three different reporters, overexpression of RMRΔlum induced an increased secretion of the reporter, whereas a simple fractionation showed that the RMRlumER construct provoked their accumulation in microsomal compartments, presumably ER. Two other constructs, where either the luminal domain of AtRMR1 was redirected to the PM (RMRlumPM) or the soluble cytosolic domain was overexpressed (RMRcyt) didn’t have any effect on the sorting of enzymatic reporters, suggesting that the transmembrane and cytosolic domains are both important for the VSD binding and that the cytosolic domain alone is not able to interfere with the vacuolar sorting machinery. I completed these results with a study of vacuolar sorting in gene knockout (KO) mutants. I carried out observations on whole plants containing single KO mutations for AtRMR1, AtRMR3 and AtRMR4 genes, where the fluorescent reporters GFPchi or Aleu143GFP had been introduced by crossing or by agro-infiltration. In these plants, the fluorescence pattern of vacuolar reporters showed a striking difference compared with reporter plants: they mostly appeared in punctate, peripheral structures and tended to accumulate in the corners of the cells. Taken together these results give new insights in the receptor-mediated protein vacuolar sorting: (1) AtRMR1 is important for both the CtVSD and ssVSD pathways, (2) three single KO for three AtRMR genes show similar impairment in vacuolar sorting. There are at least two possible explanations that help to define a new model for RMR function: first, RMR could be a “general purpose” receptor that discriminates as early as in the ER/Golgi the proteins to be sorted to plant vacuoles (CtVSD and ssVSD proteins), whereas receptors of the VSR family would act more specifically in a later intermediate sorting compartment; second, AtRMR action could be regulated through the formation of receptor complexes, as at least three of them seem to be needed simultaneously for a proper sorting of vacuolar reporters.

2006, Okmeni Nguemelieu, Jeannine, Neuhaus, Jean-Marc

Fusion of the Green Fluorescence protein (GFP) to propeptides of different vacuolar proteins like barley aleurain and tobacco chitinase allowed to visualize two different vacuolar compartments with different sizes in different tissues. These propeptides contain vacuolar sorting determinant (VSD) of two different types: sequence-specific (aleurain) and C-terminal (chitinase). These VSDs are supposed to be recognized by receptors such as VSRs and RMRs. VSRs are supposed to mediate protein sorting to lytic vacuoles, while RMRs are supposed to mediate protein sorting to storage vacuoles. Partial cDNA sequences for these vacuolar sorting receptors were cloned into a geminivirus silencing vector, and introduced by biolistics into transgenic Arabidopsis plants expressing either Aleu-GFP or GFP-chi to visualize effects of gene silencing. The inactivation of the subfamily AtVSR3 in Aleu-GFP transgenic plants caused the absence of the GFP in the large central vacuole in epidermal cells (which are lytic vacuoles) of rosette leaves, while GFP appeared in small compartments which can be ER or Prevacuolar compartments (PVC). Silencing of subfamilies AtVSR 1and 2 did not affect strongly GFP distribution in cells. Seeds from these plants were not able to germinate, and scanning electron micrographs showed that seed coat cells were no more hexagonal and miss their columella compared to Wild type seeds. Unexpectedly, silencing of RMRs in Aleu-GFP plants lead to the secretion of GFP from mesophyll cells. In GFP-chi plants, RMRs silencing also lead to the secretion of the GFP into the extracellular space in mesophyll cells .In these plants, silencing of the VSR subfamilies did not affect the GFP fluorescent in epidermal cell vacuoles. Therefore we confirmed that VSRs and specially the subfamily 3 is the best candidate for sorting of proteins with sequence- specific VSDs in leaves while RMRs seem to be involved in the sorting in both pathways. Also interesting is the used of reverse genetic to study RMRs. This technic was used because of symptoms obtained with germinivirus. Using in situ hybridization, I have detected VSRs receptors in leaves and in roots of Arabiopsis thaliana plants. These results showed that AtVSR 1 and 5 mRNA were the most transcribed in leaves and in root. Finally, it seems that direct interaction between VSRs and RMRs is necessary to sort proteins to lytic vacuoles.