The paper in brief

  • Plasmodium falciparum is devious. It uses 60 different var genes to express slightly different versions of one protein, PfEMP1, on the surface of the host's infected erythrocytes (red blood cells).

  • Moreover, the parasite expresses one var gene at a time, making it much harder for the immune system to recognize infected erythrocytes than if there were just one var gene.

  • On page 223 of this issue, Jiang et al.1 show that the gene pfSETvs silences the expression of the remaining 59 var genes at any one timeFootnote 1.

  • The protein product of pfSETvs modifies var genes through a H3K36me3 mark — that is, by adding three methyl (me) groups to lysine amino-acid residue (K) 36 of the histone (H) 3 protein associated with these genes.

  • When the authors deleted pfSETvs, almost all 60 of the var genes were expressed simultaneously in a single parasite, and the proteins they encode made their way to the surface of infected erythrocytes (Fig. 1).

    Figure 1: Regulation of var-gene silencing.
    figure 1

    a, In wild-type Plasmodium falciparum, the protein PfSETvs (not shown) adds the H3K36me3 mark to chromatin containing all but one var genes, thereby silencing their expression. Consequently, a single version of identity-determining PfEMP1 — the protein product of var genes — is expressed on the surface of an infected erythrocyte. Long non-coding RNA is expressed in the antisense direction only in the active var gene. b, Jiang et al.1 find that PfSETvs loss results in the simultaneous expression of all var genes, and so the infected erythrocyte displays several varieties of PfEMP1.

Unusual use of a mark

Swaminathan Venkatesh & Jerry L. Workman

Switching identity to evade immune detection is a common trick. What is surprising is Jiang and colleagues' finding that P. falciparum uses the H3K36me3 mark in an uncommon way2 to silence its identity-determining var genes.

In multicellular organisms, gene-silencing mechanisms work by reorganizing chromatin (complexes of DNA and associated proteins) into a tight, inflexible structure, to diminish access to the DNA. In plants and animals, methylation of specific lysine residues on histones (H3K9, H3K27 and H4K20) is crucial for engaging proteins that form this repressive structure3.

But not all methylation marks silence gene expression. H3K4 and H3K36 residues, for example, are methylated during gene transcription, maintaining the transcriptional competence of the chromatin template. Specifically, H3K36me2 and H3K36me3 are selectively enriched in coding DNA regions, functioning to preserve chromatin structure and so prevent initiation of transcription at inappropriate regions4. It was unexpected, therefore, when Jiang et al. found that P. falciparum uses H3K36me3 to coat not just the coding regions but also the promoter sequence of most var genes, thereby blocking their transcription.

Interestingly, experimental manipulations in yeast that mis-target the methyltransferase protein Set2, and so H3K36me3, to gene promoters repress transcription5. This raises two questions. Do Plasmodium parasites use a similar histone methyltransferase protein to add H3K36me3 to the var genes? And if so, what leads to its unusual localization to var-gene promoters in Plasmodium?

Jiang and co-authors' answers to these questions reveal previously unknown facets of parasite biology. It turns out that PfSETvs — the histone methyltransferase that functions in P. falciparum — shows sequence similarity to a fly protein involved in activating transcription. Whereas there is considerable uncertainty about whether the fly protein targets the H3K4 or H3K36 residues, the authors convincingly show that PfSETvs occupies the silent var genes and adds the H3K36me3 mark only at early stages of parasite infection. Intriguingly, in P. falciparum, PfSETvs is responsible for the addition of H3K36me3 to promoters and coding regions of only var genes and members of other variant-gene families that carry this mark. However, the methyltransferase responsible for adding H3K36me3 to other genes is unidentified.

What is the advantage of using H3K36me3 for gene silencing, instead of other marks that have evolutionarily conserved silencing activity? The answer might lie in the easy reversibility of the methylated and unmethylated states to allow var-gene switching. But first it is necessary to know how a single, specific var gene is turned on while all the others are silenced. Jiang and co-workers' analysis suggests that a long non-coding RNA (lncRNA) generated from the transcription of an active var gene in the opposite (antisense) direction removes PfSETvs from this gene's promoter, allowing initiation of its transcription (Fig. 1). Identifying parasite proteins that interact with H3K36me3 might clarify the exact mechanism of H3K36me3-mediated silencing. It might also clarify how the parasite differentiates H3K36me3-enriched active coding regions from the H3K36me3-enriched silent var-gene promotors to target the silencing complexes.

Repertoire unveiled

Mats Wahlgren & Maria Teresa Bejarano

The subject of Jiang and colleagues' work, PfEMP1, is a key target of immunity. P. falciparum expresses this adhesive protein on the surface of infected human erythrocytes to sequester itself within blood vessels and, thus, avoid destruction in the spleen. Therefore, specific antibodies that protect humans against severe malaria target PfEMP1 to overcome the obstruction to the blood flow caused by the parasite-infected erythrocytes6.

To evade immunity, P. falciparum varies the proteins it expresses on the surface of the infected host erythrocytes. Even in persistent malaria infections, the protein variants that appear later are distinct from those of the parental parasite in terms of their antigenic determinants — the triggers for an immune response. This antigenic variation reflects a fundamental element of parasitism7. PfEMP1 belongs to one of several families of variant proteins that are expressed on the surface of erythrocytes infected with P. falciparum8,9,10.

This is a powerful parasitic defence strategy, and immunity develops slowly in patients with malaria. That is because antibodies to a single PfEMP1 variant block sequestration only of the parasites expressing that variant, and their cross-reactivity with other PfEMP1 variants is limited. Consequently, anti-PfEMP1 antibodies of different specificities are needed to protect against the glut of the protein's variants that develop in an infected individual, especially in the most vulnerable — children and pregnant women.

Could P. falciparum lacking PfSETvs — which Jiang et al. find expresses the whole repertoire of PfEMP1-encoding var genes — be used for vaccination? A vaccine based on this mutant could allow the generation of a full repertoire of antibodies to protect against malaria, including the severe forms of the infection.

Human vaccines against bacteria and viruses are often based on killed, live attenuated or inactivated microorganisms. For P. falciparum, a unicellular organism, advances in the development of live whole-cell vaccines against malaria have mainly come from studies of pre-erythrocytic stages of the parasite's life cycle11, although vaccination with its blood stages has also been tried12. Moreover, Babesia bovis, a parasite related to P. falciparum that infects cattle, is used in a live vaccine in several countries and protects the animals against severe forms of the disease. So a vaccine based on the whole, blood-stage, PfSETvs-deficient parasite could potentially be developed and, to improve its efficiency, be combined with a vaccine based on a parasite form that is maturing in its mosquito vector11.

Parasites expressing the complete repertoire of variant genes do not appear spontaneously in nature nor during in vitro growth. PfSETvs-mediated silencing therefore seems robust. Still, PfSETvs might not be the only protein involved in regulating variant-gene families. Indeed, P. falciparum often loses the capacity to activate and express genes encoding PfEMP1 in vitro, generating parasites that would not be expected to survive in a human host. Although the present study implicates antisense lncRNA in activating var genes, to target the variant genes of P. falciparum with drugs and vaccines, the mechanisms that initiate and regulate their activation must be explored further.