Evidence on the chromosomal location of centromeric DNA in Plasmodium falciparum from etoposide-mediated topoisomerase-II cleavage

Results

To identify the sites of accumulation of topoisomerase-II on chromosomes of P. falciparum, cultures of infected human red blood cells (6–10% parasitemia) were incubated with 100 μM etoposide for various times as indicated (Materials and Methods and legends to Fig. 1 –4). All experiments were performed by using the haploid asexual stages of P. falciparum 3D7, the genome project strain. Parasite chromosomal DNA was isolated and fractionated by pulsed-field gel electrophoresis (PFGE). After Southern blotting and hybridization with cloned probes, we observed that chromosome cleavage had occurred and were able to determine the sizes of the resulting fragments. The data from chromosomes 1 (0.70 Mb) and 5 (1.4 Mb) illustrate the type of results obtained (Figs. 1 –4). They show that topoisomerase-II activity is concentrated at a single locus on each chromosome and that the sizes of the cleavage products can be used to infer the location of centromeric DNA.

With the chromosome 1-specific probes PfGCAP and C1-B, we identified major products of 480 and 220 kb, respectively (Fig. 1), tentatively mapping the cleavage site(s) to a 9-kb “gene-free” region of the chromosome that contains a prominent 2.4-kb AT-rich domain similar to the type of element previously suggested as a candidate centromere (12, 13). Using a probe (C1-A) derived from sequences immediately adjacent to this element, we found that both of these fragments could be detected (Fig. 1). This hybridization pattern indicates that, although etoposide-mediated topoisomerase-II cleavage is restricted to a specific locus, the susceptible area is sufficiently broad that it covers each side of the C1-A sequence. As a result, the C1-A probe hybridizes to both the 480- and the 220-kb products. By implication, therefore, topoisomerase-II cleavage in P. falciparum must be a regional phenomenon.

To assess this finding further, and to determine the extent of the topisomerase-II activity, we restriction-digested genomic DNA from etoposide-treated parasites and examined the cleavage pattern in the vicinity of this putative centromeric region (Fig. 2), revealing the presence of several major bands, indicative of a region of cleavage that stretched over 8–10 kb, including the AT-rich domain and the gene corresponding to the C1-A probe. The observation that topoisomerase-II activity is not restricted to the immediate confines of the AT-rich domain implies that the skewed base content of this region does not, by itself, confer susceptibility to cleavage. When these blots were rehybridized with PfGCAP, a gene situated 50 kb downstream of the putative centromere, no cleavage products were detected (Fig. 2), confirming that topoisomerase-II activity is restricted to sites immediately adjacent to the AT-rich domain. The locations of topoisomerase-II cleavage sites within mammalian centromeres have been mapped to a resolution of ≈50–150 kb (25–27).

For chromosome 5, we used probes C5-A and C5-B, which were derived from DNA sequences located on either side of a similar AT-rich domain, and probe C5-C, which was derived from an ORF ≈150 kb from this element. On PFGE blots, we detected products of 500 and 1,000 kb, respectively, with probes C5-A and C5-C, delineating the cleavage sites to the region between these two sequences (Fig. 3). With probe C5-B, we detected both fragments, implying that topisomerase-II cleavage had occurred on each side of the corresponding genomic DNA sequence. Consistent with this finding, when restriction-digested genomic DNA was analyzed (Fig. 4), we observed that topoisomerase-II activity was spread over a 10- to 12-kb region encompassing the C5-B sequence and the 2.3-kb AT-rich domain, with at least four major cleavage sites.

The cleavage patterns produced by all 14 P. falciparum chromosomes after etoposide treatment were determined by Southern analysis of PFGE gels (available from the author upon request). The probes used and the proposed locations of centromeric DNA for each chromosome are summarized in Fig. 5. Topoisomerase-II activity could be mapped to the vicinity of single AT-rich domains with a sharply defined size limit of 2.3–2.5 kb (Fig. 6and data available from the author upon request). These domains exhibit a minimal GC content (<3%), are present in only one copy per chromosome, and are located in gene-free regions that vary from 6 to 12 kb. We could find no common themes to centromere-proximal genes in terms of their orientation, temporal expression, or putative function (www.plasmodb.org). Despite their high AT content, the domains are distinct from introns and other untranslated regions of the P. falciparum genome. They lack the characteristic long stretches of polyA, polyT, and A/T dinucleotide or trinucleotide repeats. We also examined each of the AT-rich domains for other repetitive arrays or for conserved elements (28). A series of repeat motifs of various lengths, copy number, and sequence were identified (Fig. 6). However, apart from the nucleotide content and tightly restricted size range of the domains, we could detect no features that were extensively conserved between chromosomes. In the case of chromosome 10, the data indicate that the topoisomerase-II cleavage site is 1.1 Mb from the left-hand telomere (Fig. 5 and information available from the author upon request), although there is not an apparent AT-rich domain at this site or elsewhere on the chromosome (www.plasmodb.org). Whether this exception represents a functional variant or reflects a gap in sequence assembly remains to be determined.

Discussion

We have shown here that treatment of asexual stage P. falciparum parasites with the topoisomerase-II inhibitor etoposide results in chromosomal cleavage at single loci. With the availability of a completed genome sequence (12) and the fact that parasite chromosomes are of a size range that is amenable to analysis using PFGE-based approaches, we have been able to accurately map these lesions. The resulting data identify the major sites where catalytically active topisomerase-II accumulates on P. falciparum chromosomes. Topoisomerase-II plays a central role in chromosome biology and has been implicated in the segregation process in organisms ranging from yeast (18, 29, 30) to vertebrates (31–33). After replication, sister chromatids remain attached, partly through strand catenation at centromeres (22), as cells enter mitosis. The premitotic accumulation of topoisomerase-II at functional centromeres is a major regulator of sister-chromatid cohesion and is essential for ordered segregation (19). Evidence for this centromeric sequestration of enzyme activity includes the observation that etoposide-mediated cleavage sites in human chromosomes can be mapped to within the α-satellite arrays (23, 25–27). It has even been suggested that these topoisomerase-II-binding sites could function as epigenetic markers for kinetochore assembly (3). Using etoposide, we have also found (J.M.K., unpublished observation) that topoisomerase-II activity in chromosomes of the parasitic protozoan Trypanosoma cruzi maps specifically to regions that we had shown to be essential for mitotic stability (34). Intriguingly, the cleavage sites in P. falciparum that we have mapped here encompass the extremely AT-rich elements of 2.3–2.5 kb, which are present once per chromosome and which have been proposed as candidate centromeres (12, 13).

These results, based on sites of topoisomerase-II accumulation, provide experimental evidence on the location of centromeres in P. falciparum. The findings suggest the presence of a class of compact centromeric DNA that differs, in terms of size and organization, from both the “regional” and point centromeres described previously in other organisms. These unusual properties may also be a conserved feature of centromeres in the Plasmodium genus. In Plasmodium vivax, although the genome assembly is not yet complete, sharply defined AT-rich domains similar to, and syntenic with, those in P. falciparum can be detected on contigs from chromosomes 4, 6, and 14 (www.plasmodb.org). The precise role of the centromeric DNA sequence in chromosome segregation remains controversial and is a focus of considerable research (3, 9, 35). In Plasmodium species, the lack of experimental data on the location and organization of centromeric DNA has limited any meaningful progress in this area. For instance, it remains to be established how centromeres are recognized by the segregation machinery, which proteins are involved in mediating the process, and whether centromeres are even associated with areas of heterochromatin. In the case of the AT-rich domains, we propose that the DNA sequence per se probably does not play direct a role in segregation, because there is a lack of evident interchromosomal conservation. Rather, we suggest that features or properties associated with the skewed nucleotide content and the sharply defined size range, which are conserved between chromosomes, are more likely to be determinants of centromere function. These AT-rich domains could be involved, directly or indirectly in interaction with specific proteins or complexes, or be implicated in some other aspect of higher-order chromatin structure.

The demonstration that topisomerase-II accumulates on P. falciparum chromosomes at sites adjacent to the AT-rich domains now provides a more robust framework for identifying and functionally characterizing other determinants of centromere function. This finding could have important implications, because the unusual nature of centromeric DNA in P. falciparum, combined with the apparent absence of genes for most of the conventional core centromeric proteins, suggests that some aspects of the segregation process may be distinct from those in the human host. In addition, there could also be differences in the factors that mediate topisomerase-II accumulation at centromeres or that regulate enzyme activity. For example, in Xenopus, conjugation of the ubiqutin-related peptide SUMO-2 to topoisomerase-II by the SUMO ligase PIASy is thought to be a requirement for proper centromeric function and chromosome segregation (33). There is no obvious equivalent of PIASy in the P. falciparum genome database. Finally, our results suggest that the biochemical mapping approach, based on etoposide-mediated topoisomerase-II cleavage, can be applied to other protozoa. This is an extensive group of organisms, frequently of medical importance, where information on the location and organization of centromeric DNA is almost universally lacking.

Materials and Methods