An Overexpression Screen of Toxoplasma gondii Rab-GTPases Reveals Distinct Transport Routes to the Micronemes

The basic organisation of the endomembrane system is conserved in all eukaryotes and comparative genome analyses provides compelling evidence that the endomembrane system of the last common eukaryotic ancestor (LCEA) is complex with many genes required for regulated traffic being present. Although apicomplexan parasites, causative agents of severe human and animal diseases, appear to have only a basic set of trafficking factors such as Rab-GTPases, they evolved unique secretory organelles (micronemes, rhoptries and dense granules) that are sequentially secreted during invasion of the host cell. In order to define the secretory pathway of apicomplexans, we performed an overexpression screen of Rabs in Toxoplasma gondii and identified Rab5A and Rab5C as important regulators of traffic to micronemes and rhoptries. Intriguingly, we found that not all microneme proteins traffic depends on functional Rab5A and Rab5C, indicating the existence of redundant microneme targeting pathways. Using two-colour super-resolution stimulated emission depletion (STED) we verified distinct localisations of independent microneme proteins and demonstrate that micronemal organelles are organised in distinct subsets or subcompartments. Our results suggest that apicomplexan parasites modify classical regulators of the endocytic system to carryout essential parasite-specific roles in the biogenesis of their unique secretory organelles.

. Overview of Apicomplexan Rabs • Table S2. Summary of Rabs in T.gondii • Table S3. List of Primers used in this study • File S1.xls Overview and accession numbers of all sequences of Rabs used in the phylogenetic analysis

Result and Discussion of localisation analysis of Rabs
In order to identify Rabs that are involved in vesicular transport to the unique secretory organelles of Toxoplasma gondii we performed an overexpression screen, where each Rab has been fused to ddFKBP-myc [1], that allows rapid and tuneable regulation of protein levels.
While overexpression phenotypes can be expected at high concentrations of the inducer Shield-1, localisation analysis of the protein of interest can be performed at low concentrations and in some cases even in absence of inducer. Using this approach we found that all analysed Rabs are localised to the early-late secretory system of the parasite, but not to the apical secretory organelles.
We found that Rab1B and 18 localise predominantly to organelles of the early secretory pathway, the ER and the Golgi ( Figure 1B, Figure S4B,F) as demonstrated by co-localisation with the marker proteins TgERD2 and GRASP [2]. However Rab 2 shows a tendency of accumulation within the ER/Golgi region ( Figure 1B, Figure S4C), but the colocalisation is not as clear as it is for Rab1B and 18. This could suggest that Rab1B and Rab2 have a similar role in the transport of vesicles between the ER and the Golgi, as observed in other eukaryotes [3,4], but a different function within apicomplexans for Rab2 cannot be ruled out. Rab18 has been identified in several eukaryotic lineages, indicating that this Rab was present in the LCEA [5]. However, in contrast to Rab1 and 2 it has been lost in several species, including some apicomplexan parasites, where it is missing in case of Cryptosporidium, Theileria and Babesia [6]. Rab18 has been implicated in diverse roles, including ER-Golgi-traffic [7,8], formation of lipid droplets [9,10] or regulation of secretion in neuroendocrine cells [11], indicating that this protein does not show a strict functional conservation.
Rab4 can be identified in diverse eukaryotic lineages, but have been lost on several occasions [12]. Similarly, in apicomplexans Rab4 is present in Toxoplasma and Cryptosporidium but absent in Plasmodium and Theileria [6]. Rab4 had been first characterised in human cells and shown to be essential for endocytosis and the formation of early endosomes [13]. A similar role for Rab4 has been demonstrated in diverse eukaryotes, including protozoan parasites such as Trypanosoma brucei, where it plays an important role in endocytic recycling [14]. In contrast we found that T.gondii Rab4 is almost exclusively localised to the Golgi ( Figure 1B Figure S4D). On some occasions we also found a partial co-localisation of Rab4 with proM2AP [15], a marker for what we call, endosomal-like compartments ( Figure S4D). Post-Golgi organelles, where TgVP1, proM2AP and TgCPL are localised are often assumed to be early or late endosomes, lysosome or other organelles like VAC or plant-like vacuole. Since no evidence exist to prove this, we prefere to call these compartments endosomal-like compartments (ELCs).
A recent phylogenetic analysis suggests that apicomplexan Rab1A defines a unique paralog shared by alveolates [16] and our phylogenetic analysis supports this view ( Figure 1A, Figure   S2). When we analysed parasites expressing a ddFKBPmyc-tagged version of Rab1A, we found this protein concentrated within the post-Golgi region as indicated by co-localisation with the Golgi marker GRASP-RFP and ELC marker proM2AP and TgVP1. Due to this highly dynamic location of Rab1A we cannot define a compartment, where exactly this GTPase is localised. (Figure S4A).
Rab7 has been previously localised to the ELCs in T. gondii [17,18]. In our study we confirmed partial co-localisation of Rab7 with ELCs (proM2AP and TgVP1) ( Figure   1B, Figure S4E) , consistent with a conserved role of Rab7 in trafficking to endosomes [19] Three Rab5-GTPases can be identified in the genome of apicomplexan parasites ( Figure 1A).
While Rab5A and Rab5C appear to be derived from a lineage specific gene duplication event ( Figure 1A, Figure S2), Rab5B belongs to a unique class that is only conserved in apicomplexan parasites. Interestingly this protein lacks the typical prenylation motif at the Cterminus. Instead a potential myristoylation motif at the N-terminus can be identified ( Figure   S1). Therefore we tagged Rab5B C-terminally with ddFKBPHA for localisation studies and found that this protein showed a concentration at ELCs ( Figure 1B, Figure S5D) and to a lesser extent at the surface of the parasite, possibly the inner membrane complex (IMC) ( Figure S5D). Consistent with earlier studies [20], we identified Rab5A at ELCs and found an identical location for Rab5C ( Figure 1B, Figure S5A,B,C), indicating a role of the two Rabs in the organisation and function of the ELCs.
Although the analysis of the localisation of Rabs is a first step to obtain information about their individual function, we would like to mention that without a more detailed analysis (as presented for Rab5A and Rab5C), we cannot draw any firm conclusions about their role.
Although we found that overexpression of Rab2, Rab4 and Rab5B is not tolerated by the parasite, we were unable to identify a clear phenotypic consequence (apart from parasite death). Similarly we were unable to pinpoint the consequence of dominant active expression of Rab7(G18E) (see Figure S7). Furthermore, we do not suggest that a particular Rab is not essential in case no overexpression phenotype can be observed (as in case of Rab18). Future studies using conditional knockout technologies should be employed to further characterise this important protein family.