The identification of a large number of proteins in complex biological samples was made feasible by the application of mass spectrometry to generate peptide fingerprints to be compared with peptides predicted from genome databases. Two general approaches can be adopted in proteomic studies: protein samples can be first separated through one or bidimensional gel electrophoresis and successively submitted to protease digestion or, in alternative, protein mixtures can be directly treated with proteases and the generated peptides successively separated using liquid chromatography (LC) (for review see ref ).
However, one of the key challenges in the study of the proteome of endocellular parasites, as the case of Plasmodium, is to obtain a sufficient material with a low level of contamination from host cell. This limits the analysis of particular stages (e.g. hepatic or sexual stages) or subcellular fractions. To overcome these problems selection methods such as magnetic beads coated with specific monoclonal antibody were successfully used to purify invasive mosquito stage (sporozoite) . Male and female gametocytes were, instead, separated by flow cytometry using transgenic lines expressing the green fluorescent protein (GFP) under the control of sex-specific promoters .
Data on both rodent  and human [8, 9] malaria parasite proteome of invasive, replicative or sexual stages were obtained and showed a high degree of specialization between the different developmental stages, revealed by the presence of a high number of stage-specific proteins.
Further, the complete proteome analysis of human erythrocyte  led to the identification of 181 unique proteins half of which reside in the plasma membrane and half in the cytoplasm, most of them were also categorized according to their function. This might provides a basis for the study of metabolic and structural changes induced by malaria infection, thus contributing to unravel the complex host-parasite relationships.
Combined transcriptomic and proteomic data allowed also to uncover regulatory mechanisms of gene expression. It was observed for a number of genes involved in gametocyte differentiation that transcriptional and translational processes are not coupled but occur at different stages of sexual development, suggesting for them a posttranscriptional control of gene expression . This confirms bioinformatic analysis which predicts from whole genome annotation an abundance of proteins modulating mRNA decay and translation rates  that might indicate that protein levels in malaria parasite are significantly controlled through mRNA processing and translation .
Proteomics of subcellular parasite compartments
Malaria parasite has the peculiar ability to invade highly specialized red blood cells, and to develop within a parasitophorous vacuole, which represents the interface between parasite and the host cell environment. The infected erythrocytes have no intracellular organelles and are devoid of protein synthesis and trafficking. In order to ensure its own survival Plasmodium remodels the host cell by exporting proteins into the cytoplasm and plasma membrane of the erythrocyte and generating new subcellular compartments within the infected erythrocytes which allow the parasite to establish mechanisms of immune evasion and create new permeation pathways for nutrient uptake. A complex tubovescicular membrane network is formed  that extends from the parasitophorous vacuole throughout the erythrocyte cytoplasm which most probably associates with flattened vesicular structures beneath the erythrocyte membrane, a Golgi like compartments called Maurer’s clefts, involved in translocation of parasite-encoded proteins to the erythrocyte surface (see ref  for review). The infected erythrocytes acquire adhesive properties and sequester in capillary of internal organs. This allows the parasite to escape passage to the spleen but causes severe pathology to the host . Proteome analysis of subcellular components of Plasmodium is at the early days also for the difficulties encountered in their purification. Components of rhoptries, apical secretory organelles of merozoites, most probably involved in parasitophorous vacuole biogenesis, have been identified through proteomic combined with bioinformatic approaches . Among proteins predicted to be located in this compartment are proteases possibly involved in processing merozoite proteins targeted to the rhoptries, enzymes of lipid metabolism that might be implicated in the establishment of the vacuolar membrane, as well as, proteins known to localize to the vacuolar membrane, finding that supports the involvement of rhoptry components in the genesis of parasitophorous vacuole.
Two complementary studies have been reported: the one on Maurer’s cleft proteomics  that shed new light on the important biological functions of this parasite-derived compartment. This study confirms that Maurer’s cleft have characteristics of a secretory compartment addressing parasite proteins to the red cell surface but also suggests that these structures are not involved only in protein trafficking as indicated by the presence of enzymes, and proteins implicated in signal transduction. The other study  applied high throughput proteomics to identify antigens on the surface of infected erythrocyte using a multidimensional protein identification technology (MudPIT), a two-dimensional liquid chromatography coupled with tandem mass spectrometry. The advantage of this approach is its ability to analyse complex protein mixture that are difficult to resolve by gel-based protein separation systems, such as membrane proteins. 36 candidates were selected with the support of bioinformatics for the presence of a predicted signal peptide and/or of transmembrane domain(s). The surface location was confirmed by immunolocalization experiments for two of the selected proteins.
Membrane microdomain proteomics, future perspectives
The traditional view that the plasma membrane is a uniform lipid bilayer containing randomly distributed membrane proteins has given way to a more complex model in which glycosphingolipids associate with cholesterol to form organized “liquid ordered” structures or “lipid rafts” . The physical properties of these microdomains impose barriers to the free diffusion of membrane proteins. They act as molecular sieves admitting certain proteins while excluding others. The ability to dynamically compartmentalize membrane proteins allows synchronizing a wide array of cellular responses linked to signal transduction, transcytosis, cholesterol transport or internalization of pathogens . In addition, evidence is growing that lipid rafts are present not only at the plasma membrane but also internally where they could assist trafficking between membrane compartments.
Microdomains, most probably present in all cells, have been also identified at the erythrocytes membrane and at membrane compartments generated by malaria parasite. As detailed below, lipid rafts are involved in the establishment and maintenance of malaria parasite infection.
It was shown that host proteins associated to lipid rafts can be drawn into the P. falciparum vacuole during invasion and intraerythrocytic growth . Raft-cholesterol depletion from infected red blood cells destabilizes the vacuolar membrane and causes parasite expulsion from host cell  indicating that the presence of these microdomains is critical for maintaining parasitophorous vacuole. When raft-cholesterol depletion is performed on uninfected red blood cells, a reduction of invasion by 80-90% was observed , suggesting that parasite entry might occur via lipid rafts.
A number of parasite proteins have been identified as stably or transiently associated to lipid rafts [36, 37] but a complete analysis of raft protein composition in malaria parasite have not been undertaken. This would be of help to elucidate the raft-dependent functions in the course of parasite infection and possibly the interplay between host and parasite components.
A limited number of characterized raft proteins is available to date (examples are given in refs  and ), although the isolation of microdomain components exploiting their physico-chemical features (i.e. insolubility in nonionic detergents and low buoyant density) is a well established procedure. Our group isolated proteins associated with lipid rafts from P. berghei in order to provide candidate components of this crucial membrane compartment through a comprensive mass spectrometry analysis. Here we present a MALDI-MS analysis aimed at identifying in the complex mixture of raft components PbSEP1, a P. berghei protein dinamically associated to this membrane compartment. This protein belongs to a family (SEP) of small exported proteins (13-16 Kda), conserved within Plasmodium genus. SEPs are located at the membrane of parasitophorous vacuole and/or at the Maurer’s clefts . They are characterised by the presence of a signal peptide, a hydrophobyc region and a highly charged region at the carboxy-terminus. PbSEP 1 was identified in rhoptry proteome of free merozoites  while localizes at the vacuolar membrane early after invasion , suggesting a role for this protein in the genesis of this parasite compartment.
Lipid rafts were isolated, by buoyant density-gradient centrifugation, from purified P. berghei parasites with an intact vacuolar membrane (Fig. 3A). An immunoblot of the collected fractions probed with PbSEP 1 is shown in Fig. 3B. Raft-associated proteins were separated by one dimension gel electrophoresis and stained with Comassie Blue (Fig. 3C). MALDIMS analysis of a gel area containing proteins ranging from 13 to 16 KDa confirmed the presence of peptides derived from PbSEP 1 (Fig. 3D). This analysis indicated that integral membrane proteins which are minor components of this specialized proteome, as the case of PbSEP1, can be identified with the applied procedures and that a more comprehensive analysis of raft proteome by mass spectrometry is feasible. We expect that plasmodial rafts will contain proteins involved in the generic structural and functional tasks of membrane microdomains along with a subset of proteins unique to the parasite which might constitute potential targets for intervention.
The authors wish to thank the Italian Ministero della Salute and the European Commission (FP6, Network of Excellence BioMalPar, contract N° LSHP CT-2004 503578) for finantial support.
Received on 5 September 2005.
Accepted on 21 September 2005.