Partitioning of Large and Minichromosomes in Trypanosoma brucei

See allHide authors and affiliations

Science  25 Apr 1997:
Vol. 276, Issue 5312, pp. 611-614
DOI: 10.1126/science.276.5312.611


The Trypanosoma brucei nuclear genome contains about 100 minichromosomes of between 50 to 150 kilobases and about 20 chromosomes of 0.2 to 6 megabase pairs. Minichromosomes contain nontranscribed copies of variant surface glycoprotein (VSG) genes and are thought to expand the VSG gene pool. Varying VSG expression allows the parasite to avoid elimination by the host immune system. The mechanism of inheritance of T. brucei chromosomes was investigated by in situ hybridization in combination with immunofluorescence. The minichromosome population segregated with precision, by association with the central intranuclear mitotic spindle. However, their positional dynamics differed from that of the large chromosomes, which were partitioned by kinetochore microtubules.

Trypanosoma brucei is a flagellated protozoan parasite that separated from the eukaryotic lineage very early in the evolution of eukaryotes (1,2). The extracellular parasite survives in the bloodstream of the host by periodically changing its VSG coat, a process known as antigenic variation (3). VSG genes, which are the only known open reading frames on the 100 or so minichromosomes of T. brucei, are transcriptionally inactive (4-8). Minichromosomes are thought to increase the repertoire of VSG genes, which can be transposed to expression sites on the larger chromosomes (9).

In eukaryotic cells, chromosomes are typically segregated by association with a bipolar mitotic spindle. This ensures an almost perfect mechanism to faithfully segregate chromosomes during cell division. Although there is an intranuclear mitotic spindle in T. brucei, indirect evidence argues against a microtubule-dependent segregation mechanism for at least part of the genome. Electron microscopic (EM) studies of dividing T. brucei nuclei indicate that the mitotic spindle contains insufficient microtubules to provide a conventional centromere-microtubule interaction for all of the more than 100 chromosomes (2, 10). Rather, the chromatin of dividing T. brucei cells seems to be closely associated with the nuclear envelope (10,11). In a more recent study, in situ hybridization was used to demonstrate a peripheral localization of minichromosomes in the nucleus of interphase cells of trypanosomes (12).

Given genetic evidence of mitotic stability (13) yet uncertainty over the segregation mechanism, we addressed the question of where chromosomes of procyclic T. brucei are located during the cell cycle and how their segregation is achieved. To visualize spindle microtubules, we used a monoclonal antibody (KMX) specific for β-tubulin (14, 15). Fluorescence in situ hybridization (FISH) was used to visualize DNA segments of individual chromosomes (16, 17). To locate minichromosomes, we used a 177–base pair (bp) repeat sequence (MC177) as a probe (18). This sequence is specific for minichromosomes and has been used previously to study their distribution in interphase cells (12). To visualize a DNA segment on one of the larger chromosomes, we chose the 5Sribosomal DNA (R5S) because this gene exists as a linear array of hundreds of tandem repeats on one of the chromosomes larger than 4 megabase pairs (Mbp) (19, 20). The segregation pattern observed for the chromosome carrying the 5S genes is likely to be representative of the large chromosomes of T. brucei in general. In support of this hypothesis, very similar results were obtained with genomic DNA clones covering about 150 kilobase pairs of the tubulin locus located in the center of a chromosome of about 1 Mbp (21). The small size of even the largest chromosomes of T. brucei and the size and position of the target DNA sequences used for FISH exclude trailing effects of those loci in relation to potential centromeres, which have not yet been identified.

Although chromosomes in trypanosomes do not visibly condense, probably as a result of their unusual histone composition (22), it is relatively easy to follow the progression of mitosis using nuclear elongation and segregation of the mitochondrial kinetoplast as markers (10, 23). Elongation of the nucleus indicates the onset of mitosis. The nucleolus does not disperse during mitosis but elongates and acquires a dumbbell shape before it splits into two entities preceding karyokinesis.

During interphase the minichromosomes were located in small clusters distributed asymmetrically around the periphery of the nucleus (Fig. 1, A and D). As the cells progressed toward M phase, a reorganization occurred. During the transition from G2 to M phase, the small minichromosome groups started to aggregate into larger and fewer clusters. Once the cell had clearly entered mitosis, indicated by the appearance of a small spindle, the minichromosomes congressed into one mass in the center of the nucleus (Fig. 1B). The mitotic spindle of trypanosomes consists of a central array of densely packed microtubules plus peripheral microtubules; the latter might represent pole-to-kinetochore microtubules because they terminate in electron-dense structures resembling the kinetochores described in higher eukaryotes (10, 11). After the establishment of the spindle, the minichromosomal DNA split into two equal-sized clusters that subsequently moved to the poles of the spindle (Fig. 1, C and D). As the central spindle elongated, the minichromosomes remained at their polar position (Fig. 1E). Late in mitosis the poles of the spindle were close to the nuclear envelope with the minichromosomes still attached. After the disassembly of the spindle, shortly before nuclear division, the minichromosomes congregated close to the nuclear envelope. This distribution pattern could still be observed in cells during and after cytokinesis; only later, during S and G2 phases, did the clusters disintegrate into smaller units (Fig. 1A, lower cell). This explained the nonrandom spatial distribution of minichromosomes in G1cells and, to a lesser extent, in S phase (12).

Figure 1

The distribution of minichromosomes during the cell division cycle. (A) In interphase, minichromosomes were asymmetrically distributed near the nuclear periphery (see also right cell in D). (B) After formation of a spindle, minichromosomes congregated in the nuclear center. (C toE) During progression of mitosis, the minichromosomes separated into two entities and relocated to the poles of the spindle. The minichromosomal signal is shown in red, the antibody to tubulin in green, and the total DNA in blue. The third and fourth frame of each row represent the merged signals and the phase-contrast image, respectively. The kinetoplasts as markers for cell cycle progression are labeled by arrows (first row only). Bar, 10 μm.

The segregation of large chromosomes, as deduced by the 5Sribosomal gene cluster localization, followed a different pattern (Figs. 2 and 3). Based on the analysis of many cells, the two dots representing the G1-phase diploid and, after S phase, the tetraploid chromosome complement did not exhibit a preferential localization within the nucleus (Fig. 2A). After DNA replication the R5S signals were still visible as two single dots because the sister chromosomes had not yet segregated (Fig. 2B). In early mitosis, when the minichromosomes were still congregated in the center of the nucleus, the R5S signals occupied a position near the nuclear periphery (Fig. 2B). Once the minichromosomes were positioned at the spindle poles, the R5S signals trailed behind and relatively closer to the midpoint of the spindle (Fig. 2C), suggesting different velocities of polar movement. As mitosis progressed, the R5S dots were frequently found at the outer periphery of the central spindle (Figs. 2, C and D, and 3, A and B), consistent with the position of kinetochore-like structures in spindles of dividing nuclei at these stages seen in EM images (11, 24). In late anaphase the R5S dots eventually moved closer to the poles but never overlapped with the minichromosomal locations (Figs. 2D and 3C).

Figure 2

Minichromosomes (red) and large chromosomes (yellow) exhibit different positional dynamics during mitosis. (A) During interphase the R5S signal was located in a central position within the nucleus, whereas minichromosomes were close to the nuclear envelope. (B) On the onset of mitosis minichromosomes congregated in the center of the nucleus, whereas the 5S signals were near the periphery of the nucleus. (C and D) As mitosis progressed, minichromosomes became localized at the spindle poles and the 5S dots were closer to the center of the spindle. The approximate position of the spindle corresponds to the black exclusion zone between the DNA (blue) and by the dark structure inside the nucleus in the corresponding phase-contrast images, which is caused by the spindle and the persistent nucleolus. Bar, 10 μm.

Figure 3

The localization of a large chromosome during mitosis. (A and B) In early stages of anaphase the 5S ribosomal gene signal (yellow) was located outside the central spindle (green). (C) In late anaphase, shortly before karyokinesis, the dots were much closer to the spindle poles. Phase-contrast images are shown on the far right. Bar, 10 μm.

To demonstrate the dependence of minichromosomal segregation on an intact mitotic spindle, we treated trypanosomes with the drug rhizoxin, which disassembles microtubules (25). A concentration of 5 to 20 nM rhizoxin affects the integrity of the spindle but leaves other microtubule-containing structures, such as the subpellicular cytoskeleton and the axoneme of the flagellum, largely unaffected. Treatment of cells for 4 hours with 5 nM rhizoxin resulted in the formation of aberrant spindle morphology in mitotic cells (Fig.4A). Minichromosomes in these cells still associated with the malformed spindles but failed to segregate to the poles. Instead, they often formed rod-shaped structures along bundles of microtubules. After treatment with 10 nM rhizoxin for 4 hours a spindle was no longer visible, but in some cells small tubulin-containing structures could be detected at a position corresponding to the poles of the former spindle (Fig. 4B). Treatment with 20 nM rhizoxin prevented any reorganization of minichromosomes in cells that, according to the position of their kinetoplasts, should have entered mitosis (Fig. 4C).

Figure 4

The effect of the anti-microtubule drug rhizoxin on the segregation of minichromosomes (color designation as in Fig. 1). (A) Rhizoxin (5 nM) still allowed a small spindle to form but prevented the polar organization of minichromosomes; they remained distributed alongside the entire spindle. (B) At 10 nM the drug prevented spindle formation but left two small structures. Minichromosomes colocalized with these structures, which were interpreted as spindle pole remnants. (C) Rhizoxin (20 nM) inhibited spindle formation completely and prevented minichromosomes from reorganizing in the nuclear center at the onset of mitosis. The interkinetoplast distance (the two DAPI-stained dots) indicated that the cell should have entered mitosis by this time. Bar, 10 μm.

We propose the following model for the mechanism of chromosome segregation in T. brucei. Minichromosomes congregate in the center of the nucleus at the onset of mitosis. This aggregation may, or may not, be preceded by a condensation of the chromosomes. After association with the emerging central spindle, they separate into two clusters that move to opposite spindle poles. They remain at the spindle poles during spindle elongation until they are in close proximity to the nuclear envelope. Their asymmetrical distribution within the nucleus is maintained after spindle disassembly until late S phase when they are distributed almost randomly near the nuclear envelope. As the ploidy is uncertain and individual minichromosomes cannot be visualized owing to the lack of large enough specific target DNA sequences, it is not clear whether there is faithful segregation of each minichromosome. However, detailed microscopic analysis of many cells (>100) showed that the minichromosomal clusters segregated on the spindle and inherited by each daughter cell were of equivalent size, indicating a precise segregation mechanism.

The existence of a highly coordinated segregation mechanism for minichromosomes suggests that they play an important role in the biology of this parasite. In addition, owing to their small size, minichromosomes may serve as an excellent model for the study of mitotic segregation, particularly with respect to the evolution of DNA partition mechanisms. The diploid large chromosomes, as exemplified by the chromosome harboring the 5S ribosomal gene, are likely segregated by peripheral pole-to-kinetochore microtubules. There is, however, an intriguing discrepancy between the number of large chromosomes, estimated to be at least 20 for the diploid set (4, 21), and the number of kinetochore-like structures, estimated to be approximately 10 (11,24).

  • * To whom correspondence should be addressed. E-mail: klaus.ersfeld{at}


View Abstract

Stay Connected to Science

Navigate This Article