Guilty Until Proven Otherwise

See allHide authors and affiliations

Science  12 Sep 2003:
Vol. 301, Issue 5639, pp. 1487-1488
DOI: 10.1126/science.1089799

In October last year, the fruits of 7 years of hard labor were finally realized with the publication of an almost complete annotated genome sequence of the deadly human malaria parasite, Plasmodium falciparum (1). Several accompanying papers presented a comparative genome sequence of the rodent malaria parasite Plasmodium yoelii (2) and two detailed surveys of the proteins expressed during the P. falciparum life cycle (3, 4). Annotation of the P. falciparum genome predicts 5409 genes, 60% of which lack a known function or a homolog in any other organism. Now, less than a year later, two papers by Le Roch et al. (5) on page 1503 of this issue and by Bozdech et al. (6) in Public Library of Science Biology present gene expression microarray profiles (transcriptomes) of P. falciparum during the different stages of its complex life cycle.

The three principle stages of P. falciparum development take place in the liver cells and red blood cells of the human host and the tissues of the mosquito vector (see the figure). Le Roch et al. (5) examined the gene expression profiles of P. falciparum at nine different developmental time points. Seven profiles were obtained from highly synchronized cultures of the blood stage parasites and included free merozoites, which burst out of one host erythrocyte and invade another. These investigators also “profiled” mature gametocytes (the sexual stage transmitted to the mosquito) and sporozoites from the mosquito salivary glands (which invade liver cells when introduced into the bloodstream of the human host by the mosquito). In contrast, Bozdech et al. (6) concentrated their efforts on the erythrocytic stages of development only, examining the gene expression profiles of P. falciparum every hour for the duration of the 48-hour blood stage. The emphasis of both groups on bloodstage parasites is desirable because these forms cause malaria-associated pathology and are the target of the majority of vaccine and drug development programs.

Bar codes for a pathogen bar none.

The life cycle of the P. falciparum malaria parasite is complex, involving developmental stages in the liver cells and red blood cells of the human host and in the tissues of the mosquito vector (top). Different forms of the parasite invade, colonize, and reproduce within different cell types at different points in the life cycle. For example, merozoites invade human erythrocytes and sporozoites invade both mosquito salivary glands and human liver cells. The exquisite specificity of the different invasive forms and their ability to survive in very different cellular environments is reflected in a specific gene expression profile (transcriptome) for each life cycle stage (bottom). The transcriptome provides a “bar code” for each stage of the parasite's life cycle.


The two groups designed different microarrays based on the P. falciparum genome: Le Roch et al. used a custom-made, high-density oligonucleotide microarray chip made by Affymetrix; Bozdech et al. designed a microarray chip containing long (70-nucleotide) oligonucleotides. The microarrays themselves are not truly global (that is, they do not contain every single P. falciparum gene) because they are based on prepublication versions of the P. falciparum genome (a concrete benefit of the immediate data release policy). Nevertheless, the microarrays are comprehensive representing 5159 (5) and 4488 (6) genes, respectively. The Le Roch et al. microarray has a greater dynamic range, samples the genome more comprehensively, and contains 367,226 features on both DNA strands with an average probe periodicity of 150 base pairs. Interested readers should refer to the impressive Plasmodium genome resource PlasmoDB (7), which integrates genome organization, annotation, proteome surveys, as well as the comparative transcription profile data of the two new studies.

How are we to make sense of this huge amount of quality data, particularly as Le Roch et al. report that all but 885 genes of the P. falciparum genome are expressed by blood-stage parasites and that more than 60% of the expressed genes have no known function? The two groups performed statistically supported cluster analyses, grouping together genes with similar expression profiles and reaching remarkably similar conclusions for bloodstage parasites, which concur with earlier less extensive studies (8, 9). Reassuringly, genes predicted to encode proteins of related function have very similar or identical transitory expression profiles that describe a developmental cascade. Genes encoding proteins of the transcription apparatus or proteins that regulate the synthesis of other proteins are expressed in the early blood stages (ring, early trophozoite). Genes encoding proteins involved in DNA replication are expressed in the period leading up to mitosis (the late trophozoite stage), and those encoding proteins associated with invasion of new erythrocytes are expressed in the schizont stage as the daughter merozoites are formed. The transcription profiles of many hypothetical genes of unknown function cluster with those of annotated genes of known function. Thus, both groups posit “guilt-by-association”: The functions of hypothetical genes may be similar to those of annotated genes in the same cluster (10). Such associations need not be restricted to housekeeping genes—expanded pools of “guilty” stage-specific genes have been found in sporozoites, gametocytes, and merozoites and reflect the highly specialized nature of these forms.

There are some surprises too: The expression profiles of some genes that would be predicted to cluster together do not. Examples include some of the invasion-related proteins of the apical organelles (micronemes, rhoptries, dense granules), which are common to both sporozoites and merozoites. Although certain invasion-related proteins inevitably will be specific for either sporozoites or merozoites, one would predict some commonality in expression profiles. This data gap may be due in part to the inaccessible nature of certain P. falciparum life-cycle stages such as the oocyst in the mosquito gut. Sporozoites and their organelles are formed in oocysts, perhaps explaining the failure to detect certain transcripts in the mature salivary gland sporozoite stage that Le Roch et al. sampled (5). Laudably, these investigators had the foresight to include many P. yoelii genes on their microarray. The easier isolation of less accessible life-cycle stages of P. yoelli from its mouse and mosquito hosts should enable this problem to be resolved in the future (11).

Both groups studied blood-stage parasites, which exhibit clonal antigenic variation encoded by var genes. Surprisingly, the majority of var gene transcription was only observed for a single cluster of three var genes located in an internal region of P. falciparum chromosome 4 (5). Most var genes (35 of 59) are located on the subtelomeres of chromosome 4, prompting Le Roch et al. to propose that to be stably expressed these genes may need to be relocated to internal chromosomal regions through recombination (5). However, this finding may reflect hierarchical expression in the absence of selective pressures. Indeed, independent data from different strains of P. falciparum show that antigenic variation usually involves expression of subtelomeric var genes (12, 13) and that expression switching occurs in situ (14).

Looking to the future, the rigorously produced data sets of Le Roch et al. (5) and Bozdech et al. (6) will be analyzed in even greater detail to seek common regulatory regions in expression clusters. Clusters will be correlated with the locations of genes in the genome for evidence of epigenetic regulation of transcription (although the initial data suggest that physical clustering is limited). New analyses will be performed with additional life-cycle stages so that we can begin to understand the parasite's response to environmental stressors (therapeutic drugs, oxidative stress) and the normal environmental fluctuations (temperature, pH) experienced by the parasite throughout its life cycle. Another rich source of information will be parasite mutants that fail to undergo specific developmental phases, such as gametocyte production, enabling specific transcription programs to be elucidated.

Most importantly, how can we translate this bounty of information into the desperately needed vaccines and drugs that are the motivation for applying high-throughput techniques to a disease that principally plagues rural Africa? An intriguing possibility that stems directly from the Bozdech et al. data (6) might be to interrupt the blood-stage transcriptional cascade of Plasmodium early on by interfering with an as yet unidentified critical transcriptional regulator. The two studies confirm that genes encoding proteins that are targets for drug development (such as proteases and metabolic pathway components) are expressed in blood-stage parasites. Le Roch and colleagues also pinpoint genes expressed in both sporozoites and gametocytes that may prove to be new targets for drug and vaccine development. Stage-specific genes of invasive forms of Plasmodium might be exploited for vaccine potential. Exploration of the thesis that coordinated gene expression denotes a common purpose for the encoded proteins may yield new drug targets.

Le Roch et al. and Bozdech et al. make a significant contribution to our knowledge of Plasmodium gene expression. Yet, the majority of the proteins predicted by the malaria parasite genome still need to have functions assigned to them. We must continue to improve the functional genomics of malaria parasites so that the momentum generated by these two studies is consolidated. Furthermore, in the absence of funding increases, tough choices must be made to identify and pursue the most promising drug and vaccine targets.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
View Abstract

Stay Connected to Science

Navigate This Article