Malaria var gene expression: Keeping up with the neighbors

Abstract

Plasmodium falciparum PfEMP1 is a malaria virulence protein whose expression is epigenetically regulated. The parasite’s ability to express exclusively only one of the sixty var genes that encode PfEMP1 is essential for disease pathogenesis. Two recent papers identify key molecular players in determining whether a var gene is active or silenced.

A fascinating aspect of Plasmodium falciparum virulence is the parasite’s ability to express variant antigens in a mutually exclusive manner enabling it to elude antibodies developed by the host. PfEMP1, the most important variant surface antigen exposed on the surface of infected erythrocytes, is responsible for cytoadherence of infected erythrocytes to the vascular endothelium and is encoded by a family of approximately 60 var genes. The sequential, exclusive expression of a single var prolongs the infection cycle within the human host, and PfEMP1 is linked to the lethal complications of malaria infection including cerebral malaria and anemia. Although it is known that the activation of an expressed var gene and the silencing of the others are epigenetically regulated, the molecular basis of the mutually exclusive expression of var genes has been an enigma.

Two recent papers by Zhang et al and Volz et al (Volz, 2012; Zhang et al., 2011), illuminate the critical roles of subnuclear localization and actin in the silencing and activation of var genes. Silenced var genes are marked with the conserved H3K9me3 heterochromatin mark (Lopez-Rubio et al., 2007). Heterochromatin is found in the nuclear periphery and silenced var genes localize to 3–10 distinct puncta corresponding to heterochromatin regions. The histone deacetylase PfSIR2 localizes to the nuclear periphery within telomeric clusters, as well as var gene promoters, and is important for maintenance of var silencing (Duraisingh et al., 2005; Freitas-Junior et al., 2005). The active var is also peripherally localized within the nucleus, but its 5′ region is marked with gene activation marks H3K4me2, H3K3me3, H3K9ac as well as the variant histone H2A.Z (Petter et al., 2011).

The conserved var intron is important for silencing of var genes, but the mechanism is not understood (Deitsch et al., 2001). Zhang et al find that the intron encompasses a sequence that is sufficient for co-localization of an episome to the nuclear periphery with silenced var genes (Zhang et al., 2011). This activity maps to a conserved var intron motif iNPE (or intron nuclear protein-binding element) that was used to identify associated proteins using iNPE pulldowns. Among the associated proteins are actin and Pf11_0091, a member of the plant-like AP2 transcription factor family recently identified in the Apicomplexa (Balaji et al., 2005). The Pf11_0091 AP2 DNA binding domain is able to interact with the iNPE and the key nucleotides are similar to the motif independently identified by the Llinas group using Protein Binding Microarray technology (Campbell et al., 2010). Pf11_0091 is the second AP2 protein implicated in regulation of var gene silencing (Flueck et al., 2010). PFF0200c (or SIP2) binds a motif SPE2 found in heterochromatin regions upstream of subtelomeric var genes and in telomere-associated repeats, but it does not have an obvious role in transcriptional regulation as would be expected for AP2 proteins. Whether Pf11_0091 regulates activity of the var intron promoter or whether it acts to regulate heterochromatin structure, as suggested for PFF0200c (Flueck et al., 2010), is not yet clear, but the prediction is that Pf11_0091 will be essential for maintenance of silenced var loci in heterochromatin regions of the nucleus.

Zhang et al demonstrate a critical role for actin in repositioning of var loci. Actin does not directly bind the iNPE motif, although actin chromatin immunoprecipitation studies confirm that actin is part of the protein complex that interacts with the var intron. Perturbation of normal actin activity with jasplakinolide, an actin stabilizer, results in movement from the heterochromatin clusters and activation of previously silent var genes, while cytochalasin, an inhibitor of actin filament assembly and disassembly, has no effect. These data are consistent with actin being critical for transport of the var intron and its associated proteins from heterochromatin to the active site within the nucleus.

Active var mRNA peaks in ring stages, but the active locus needs to retain the euchromatin epigenetic marks, including the general gene activation mark H3K4me3, that allow access of the transcriptional machinery. Volz et al implicate a dedicated methyltransferase, PfSET10 (PFL1010c), for this process. PfSET10 is one of four methyltransferases with predicted H3K4me3 activity, and Volz et al report its unique localization to single nuclear region that colocalizes with the active var expression locus, suggesting a specialized function. The binding specificity of the PfSET10 PHD domain for naked H3, suggests that this methyltransferase binds to a freshly deposited histone within a nucleosome to mark it as the active var. PfSET10 is expressed in trophozoites and schizonts, after peak var mRNA levels, consistent with a role in the maintenance of the “poised” (i.e. ready to go) state of the expressed var gene locus.

Proteomics studies of proteins associated with iNPE and PfSET10 also identify a variety of RNA binding proteins. The importance and specificity of these proteins for var expression are not yet known, but given the still mysterious but ubiquitous sterile transcripts driven by the bidirectional promoter within the intron of var genes (Epp et al., 2009), one can speculate a role for these RNA binding proteins in regulation of noncoding RNA involved in maintenance of the heterochromatin structure of the silenced var loci or regulation of the stability of the expressed var mRNA.

A lingering question is how and when these specialized nuclear structures are assembled and maintained. Early studies indicated that the critical processes that govern var gene activation and silencing occur during S phase, when chromatin is reorganized during the processes of DNA replication, histone deposition, and nuclear division (Deitsch et al., 2001). Inheritance of nuclear spatial structure may also be another mechanism by which the epigenetic information governing var gene expression is bestowed during the process of schizogeny. Actin was also identified by Volz et al as co-precipitating with PfSET10, so actin may be important for recruitment, maintenance or tethering the chromatin complexes in the active var expression site.

We now have an improved understanding of var gene activation and silencing, but many questions remain. While var localization within heterochromatin clusters appears necessary for silencing, it is not clear how a silenced var locus transitions from heterochromatin to a euchromatin nuclear region competent for var expression. Other modifying enzymes such as lysine demethylases should be in play, but the identity and location of these other factors remains a mystery. It is possible that var loci are active by default unless they are sequestered into heterochromatin, but the unique localization of PfSET10 argues that there are dedicated effectors that act in a specialized transcriptional zone to achieve variant gene expression. Stay tuned for further insights into the var variations.