Reality check for malaria proteomics
© BioMed Central Ltd 2009
Published: 26 February 2009
Skip to main content
© BioMed Central Ltd 2009
Published: 26 February 2009
New studies highlight the wide diversity of post-translational protein modifications in the intra-erythrocytic stages of the malaria parasite, raising new avenues for inquiry.
The sequencing of the genome of Plasmodium falciparum in 2002 made possible high-throughput global analysis of the transcriptome [2–5]. Interpreted in the light of the limited previous work on the expression of individual proteins, these transcriptome analyses suggested that a significant fraction of the genome is regulated in a 'just-in-time' manner; that is, immediate translation (implicitly of bioactive proteins) of newly synthesized transcripts . The first proteomic studies emerged soon after, looking at large datasets from individual or multiple parasite life stages [6–12].
While proteomic studies confirmed the expression of many proteins as consistent with the 'just-in-time' hypothesis, they also found that a previously described disjunction of transcription and translation  was not the rarity suspected, but might represent a 'master strategy' by which quiescent stages of the parasite life cycle are pre-programmed for rapid developmental transitions - for example, when the cell-cycle-arrested gametocytes are transferred from the human bloodstream into the stomach of the mosquito vector. Here, induction of gametogenesis (see Figure 1) by mosquito-derived xanthurenic acid, and a fall in temperature of the bloodmeal, activates calcium- and protein-kinase-mediated pathways that control gamete formation . Transcripts for as many as 370 proteins expressed in the gamete or in the zygote (for example, the candidate vaccine targets P25 and P28), were found to be stabilized by a DDX6-class RNA helicase, DOZI . These mRNAs are translated within minutes following ingestion of infected blood into the mosquito's stomach.
There is a second (and reciprocal) life-stage transition when another cell-cycle-arrested form (the sporozoite) leaves the mosquito salivary gland and enters the liver of the human host to initiate infection (see Figure 1) but, interestingly, here there is less compelling evidence for translational control . It is somewhat surprising, therefore, that a growing body of evidence, exemplified by the study of Foth et al. , indicates that translational control can regulate differentiation of the rapidly replicating asexual stage of the parasite during its pathogenic development inside red blood cells.
Analysis of some 9,000 spots in the gels showed that the abundance of 278 proteins changed more than 1.4-fold between samples, the most extreme being the translation initiation factor eIF5a, which exhibited a 15-fold change. Detailed analysis including identification by mass spectrometry (MS) was achieved for 54 proteins, a small but significant return for the massive investment made when compared with previous less discriminatory approaches using multidimensional protein identification technology (MudPIT) or one-dimensional gel/liquid chromatography/MS technologies [6–11], methods that have identified many hundreds of proteins at individual life stages.
What the new data lack in quantity is, however, more than compensated for by the new information on protein abundance and isoform changes. Foth et al.  detected multiple isoforms for 50% of all the proteins identified. Different isoforms of equivalent mobility (Mr) were considered to be due to changes in phosphorylation. An increase in mobility between two isoforms was interpreted to be due to post-translational protein cleavage (or proteasomal degradation). One protein, enolase, was described in no less than seven different isoforms, of which two appeared to be truncated.
By comparing the proteomic data from these samples with previous transcriptomic data from comparable samples , Foth et al.  found that expression of some proteins or isoforms - for example, the chaperone protein HSP40 and four actin isoforms - were concordant with the 'just-in-time' synthesis model. Interestingly, peak protein abundance of another actin isoform was delayed following transcription, indicating regulated post-translational modification. The expression of yet other proteins, for example HSP60, was negatively correlated with their mRNA levels.
Where does this leave us? Reductionists can argue strongly that this paper  reinforces the concept that it is essential to treat each molecule and pathway separately and investigate each and every one in depth, whereas 'synthesizers' can emphasize that such global approaches have the potential, perhaps not fully realized in this work, to understand 'master regulatory mechanisms', which require consideration before examining individual pathways, each of which will be, by definition, unique. It will be interesting to see whether the application of systems approaches to data of this type will permit resolution of these questions at the global level.
Above all, Foth et al.  provide a healthy reality check as to the complexity of the molecular mechanisms regulating the development of this important parasite, which should caution the researcher against making assumptions as to the time and place of protein activity from transcriptome, or indeed proteome, analyses. Even the phenotypic analysis of genetic mutations may not provide unequivocal solutions to these questions . For those enjoying the 'thrill of the academic chase' there is clearly ample room for more exciting research. For those seeking to control this global scourge, an understanding of the fundamental yet multi-faceted mechanisms regulating parasite development may bring ways of interrupting the parasite's life cycle, or perhaps of generating new attenuated strains for therapy or transmission blockade.