The early observation that amino-terminal sequences rich in hydroxylated and basic amino acids reminiscent of mitochondrial targeting signals were present in the conceptually translated products of the E. histolytica genes encoding mitochondrial-marker proteins Cpn60 and PNT provided the first indication that a presequence-dependent protein import pathway might operate in the then-hypothetical mitosomes (Clark and Roger 1995). Since then, putative amino-terminal mitosome targeting presequences have been identified in a range of mitosomal proteins from E. histolytica, E. cuniculi, C. parvum and G. intestinalis (Bakatselou et al. 2000; Katinka et al. 2001; Nixon et al. 2002; Riordan et al. 2003; Slapeta and Keithly 2004; Tachezy et al. 2001; Tovar et al. 2003). Detection of putative mitosome targeting presequences has relied mostly on bioinformatic algorithms trained to recognise mitochondr-ial targeting signals from model eukaryotes such as yeasts, plants and mammals. Their identification is therefore tentative and much dependent on the stringency used for each search. Given the extent of reductive evolution observed in mitosome-containing eukaryotes, it is much a matter of luck as of individual judgement whether or not a particular amino acid sequence turns out to be functional. As such, the functionality of each putative targeting presequence must be tested individually.
In the absence of an in vitro mitosome protein-import system, all efforts to test mitosome targeting presequence functionality have centred on suitable deletion experiments and the engineering of chimaeric fusion proteins to target reporter proteins into mitosomes in vivo. Further, the cross-functionality of mitosomal, hydrogenosomal and mitochondrial targeting signals has also been tested as a means of obtaining functional evidence in support of the evolutionary relatedness of these organelles. The first evidence for the functional requirement of a mitosome targeting presequence and the functional conservation of mitosomal and mitochondrial protein import came from experiments in E. histolytica where deletion of amino acids 2-15 from Cpn60 led to the accumulation of recombinant protein in the cytosol, a mutant phenotype that could be reversed by supplying the mutated protein with a mitochondr-ial targeting signal from Trypansoma cruzi mtHsp70 (Tovar et al. 1999). Since then, the putative organelle targeting signals from Cryptosporidium Cpn60, IscS, IscU and mtHsp70 as well as those from Giardia IscU and ferredoxin fused to green fluorescent protein (GFP) have all been shown to target the reporter protein into mitosomes and/or into yeast or mammalian mitochondria (LaGier et al. 2003; Regoes et al. 2005; Riordan et al. 2003; Slapeta and Keithly 2004). Moreover, recombinant Giardia ferredoxin and IscU proteins tagged with the viral haemaglutinin epitope or with GFP have also been shown to be imported in a presequence-dependent fashion into mitosomes, trichomonad hydrogenosomes and mammalian mitochondria using immunofluorescence microscopy and cell fractionation experiments (Dolezal et al. 2005; Regoes et al. 2005). That organelle targeting is abrogated by deletion of these amino-terminal targeting signals has demonstrated their absolute requirement for protein import into mitosomes.
A defining characteristic of mitochondrial and hydrogensomal protein import is the proteolytic removal of amino-terminal targeting presequences upon organelle import. In Giardia, two IscU bands have been observed by western blotting in the organelle fraction that differ in size by about 2 kDa, a minor band of 17 kDa (precursor protein) and a major band of around 15 kDa (mature protein), consistent with the removal of the IscU amino-terminal presequence. The cleavage of IscU and ferredoxin targeting signals has also been observed in recombinant parasites that overexpress GFP fusion proteins containing the corresponding targeting presequences at the amino terminus. In these cells, precursor GFP proteins accumulate in the cytosol, whereas mature proteins are found exclusively in the organellar fraction (Regoes et al. 2005). Such accumulation of precursor protein in the cell cytosol demonstrated that mitosome import, like mitochondrial and hydro-gensomal import, is an active and saturable process. Further evidence for the proteolytic processing of giardial IscU presequence comes from in vitro experiments in which radiolabelled IscU was incubated with mitochondrial and hydrogensomal lysates (Dolezal et al. 2005). In these experiments, the targeting presequence is cleaved off in a time-dependent manner and is blocked by addition of EDTA, an inhibitor of metaloproteases. Taken together, these data provide substantial evidence for the existence of a presequence-dependent protein-import pathway in mitosomes which is functionally conserved with those of mitochondria and hydrogensomes.
Was this article helpful?