Clustering and protein dynamics of Drosophila melanogaster telomeres

Abstract

Telomeres are obligatory chromosomal landmarks that demarcate the ends of linear chromosomes to distinguish them from broken ends and can also serve to organize the genome. In both budding and fission yeast, they cluster at the periphery of the nucleus, potentially to establish a compartment of silent chromatin. To gain insight into telomere organization in higher organisms, we investigated their distribution in interphase nuclei of Drosophila melanogaster. We focused on the syncytial blastoderm, an excellent developmental stage for live imaging due to the synchronous division of the nuclei at this time. We followed the EGFP-labeled telomeric protein HOAP in vivo and found that the 16 telomeres yield four to six foci per nucleus, indicative of clustering. Furthermore, we confirmed clustering in other somatic tissues. Importantly, we observed that HOAP signal intensity in the clusters increases in interphase, potentially due to loading of HOAP to newly replicated telomeres. To determine the rules governing clustering, we used in vivo imaging and fluorescence in situ hybridization to test several predictions. First, we inspected mutant embryos that develop as haploids and found that clustering is not mediated by associations between homologs. Second, we probed specifically for a telomere of novel sequence and found strong evidence against DNA sequence identity and homology as critical factors. Third, we ruled out predominance of intrachromosomal interactions by marking both ends of a chromosome. Based on these results, we propose that clustering is independent of sequence and is likely maintained by an as yet undetermined factor.

Keywords: Drosophila telomeres; nuclear organization; telomere; telomere clustering; telomere dynamics.

Figure 1

The distribution of telomere foci numbers. (A) Schematic of a side view of an embryo with a “zoomed-in” section for the confocal images in B, where it is viewed from the top. (B) Representative images from a Z stack taken in EGFP-HOAP embryos, with approximate positions along the vertical axis indicated on the schematic representation on the left. Schematic shows a polarized syncytial nucleus: black circles are centromeres, green circles are telomeres, and black and blue lines connecting the two are the chromosomal arms, with the two colors representing the two homologs. Right, images are confocal sections from a Z stack, with a zoom-in of the area in the red square and EGFP-HOAP signal in white. (C) Image from a confocal Z stack of embryos in mitosis, with a zoom-in of the area in the red square. H2Av-RFP is in red, and EGFP-HOAP is in white. (D) Histogram of telomere foci numbers from 100 nuclei in the syncytial blastoderm (from 13 embryos). Mean of the distribution is shown. (E) Histogram of telomere foci distribution from 40 nuclei (from three larval brains). Red line is a trace of bar plot values; mean of the distribution is shown. Bars for all, 5 μm.

Figure 2

Telomere signal throughout the cell cycle. Panels represent a time-lapse series in a syncytial blastoderm embryo. Each time point, labeled with its corresponding time (from tp-0min to tp-18min), shows three panels: top, merged image of EGFP-HOAP (green) and H2Av-mRFP (red); middle, EGFP-HOAP signal alone (white); and bottom, zoom-in of the EGFP signal from the area in the red square from the middle panel. Cell cycle stages deduced from chromatin characteristics are shown above the merged image. Green arrowheads within panels for tp-0min point to two telomeric foci. White ovals label two anaphase figures (tp-8min and tp-10min). Yellow arrows point out the telomeric foci that allow one to observe the increase of telomeric signal over time (from tp-12min to tp-18min).

Figure 3

Time-lapse series analysis of EGFP signal. (A) Embryo 1 intensity distributions. (Top) Panels used for measurements are shown for each time point with zoom-in below. Time points (tp) are indicated within the top panel. (Bottom) Ranking graph: data for each time point were ranked and represented with different shades of gray as indicated in the legend underneath the images for each time point. y-axis: intensity is normalized to the smallest intensity value from the whole series (set to 1). Box plots graph: intensity distribution for each time point with pink circles showing the median value. x-axis: time point. y-axis: intensity, actual values [in thousands of arbitrary densitometric units (ADU)]. The boxes contain data between the first and third quartiles, while whiskers extend to minimum and maximum. (B) Zoomed-in images and ranking graph for embryo 2, presented as in A.

Figure 4

Test of homolog pairing as a determinant of clustering. (A) Schematic for hypothesis 1, homolog pairing. (B) EGFP-HOAP foci in wt and a haploid embryo; left, confocal projection, with a zoom-in; right, schematic of chromosomes for wt (top) and ms(3)k81 mutant (bottom) as indicated. Bar, 5 μm. (C) A histogram of telomere count distribution from wt (same as the one shown in Figure 1), represented by gray bars and red line, and from ms(3)k81, represented by white bars and blue line (44 nuclei). Means are shown within the graph.

Figure 5

Test of sequence homology as a determinant of clustering. (A) Schematic for hypothesis 2, sequence homology. (B) Pairing of LacO telomeres. Left, confocal projection of LacO FISH, with two zoom-ins, LacO probe in white, DAPI in blue. Right, genotype and schematic of chromosomes for LacO3R-TD. Telomere 3R is replaced by the LacO array, and LacO is visualized with a LacO probe (yellow star). Sometimes we observed more than two foci in a nucleus (see circled cluster of three foci in second zoom-in). We assume that those were instances where the replicated sister telomeres separated from each other. (C) Histogram of distance between the two LacO signals (n = 85 nuclei, from three embryos). Distance has been normalized by dividing by the average size of the nuclear diameter. (D) Interaction of LacO telomere with retrotransposon telomeres. Left, confocal projections of LacO and HeT-A FISH, with a zoom-in, HeT-A in red, LacO in cyan, and DAPI in blue. Right, genotypes and schematics of chromosomes for LacO3R-TD (top) and LacO-int (bottom). Bars for all, 5 μm.

Figure 6

Test of intrachromosomal interaction as a determinant of clustering. (A) Schematic for hypothesis 3, intrachromosomal interaction. (B) Pairing of chromosome 3 telomeres. Left, confocal projection of LacO FISH, with a zoom-in, LacO probe in white, DAPI in blue. Right, genotype and schematic of chromosomes: telomere 3R is replaced by a LacO array and telomere 3L is marked by a subtelomeric LacO array; both are visualized with a LacO probe (yellow stars). Bar, 5 μm. (C) A histogram of distance between the two LacO signals (n = 65 nuclei, from four embryos). Distance has been normalized as in Figure 5B.

Figure 7

Candidate proteins that might mediate telomere clustering. (A) HeT-A FISH on mutant embryos, with genotypes listed on the left. Each is a confocal projection with a zoom-in, HetT-A in white, DAPI in blue. Bar for all, 5 μm. (B) Box plot of the foci-number distributions for each genotype. Pink circle: median of distribution. x-axis, genotypes: w¹¹¹⁸ (n = 80 nuclei/8 embryos), mre11^58S (n = 50 nuclei/5 embryos), and Ku80^KO (n = 50 nuclei/5 embryos). y-axis: number of foci per nucleus. (C) Histogram of EGFP-HOAP foci distribution for wt embryos, represented by gray bars and red line (n = 20 nuclei/1 embryo), and for koi embryos, represented by white bars and blue line (n = 45 nuclei/5 embryos).