Genome-wide characterization of Tetrahymena thermophila chromosome breakage sites. II. Physical and genetic mapping

Abstract

The chromosomes of the macronuclear (expressed) genome of Tetrahymena thermophila are generated by developmental fragmentation of the five micronuclear (germline) chromosomes. This fragmentation is site specific, directed by a conserved chromosome breakage sequence (Cbs element). An accompanying article in this issue reports the development of a successful scheme for the genome-wide cloning and identification of functional chromosome breakage sites. This article reports the physical and genetic characterization of 30 functional chromosome breakage junctions. Unique sequence tags and physical sizes were obtained for the pair of macronuclear chromosomes generated by fragmentation at each Cbs. Cbs-associated polymorphisms were used to genetically map 11 junctions to micronuclear linkage groups and macronuclear coassortment groups. Two pairs of junctions showed statistically significant similarity of the sequences flanking the Cbs, suggestive of relatively recent duplications of entire Cbs junctions during Tetrahymena genome evolution. Two macronuclear chromosomes that lose at least one end in an age-related manner were also identified. The whole-genome shotgun sequencing of the Tetrahymena macronucleus has recently been completed at The Institute for Genome Research (TIGR). By providing unique sequence from natural ends of macronuclear chromosomes, Cbs junctions will provide useful sequence tags for relating macro- and micronuclear genetic, physical, and whole-genome sequence maps.

Figure 1.—

DNA sequences immediately flanking functional chromosome breakage elements. (A) Alignment of G and C nucleotides. A's and T's in the flanking sequence (88.1% of the nucleotides) have been replaced with spaces. Underlined capital A's indicate substitutions in the Cbs element sequence. Note that just the locations of C and G nucleotides are sufficient to demonstrate that no two of the 33 junctions have identical sequence. (B) Nucleotide distributions at each of the 25 positions on either side of the Cbs element (C-rich strand, as shown at the top of A). These tabulations failed to show any statistically significant deviation from the null hypothesis that any given nucleotide has the same probability of occurrence (0.440 for A and T and 0.060 for G and C) at any position within 25 bp adjacent to the edges of the Cbs element (see text).

Figure 2.—

Sequence similarity between Cbs junctions 1R-2 and 1R-6. The two sequences were aligned using the unfiltered Blastn program at the NCBI site. No. 1 positions correspond to the first matching base pair. The expected value associated with the match is 4 × 10⁻⁴³. The percentage sequence identity is 83% (172/206 aligned positions), which is significantly higher (probability chi-square ≪0.001) than the 43% expected by chance alone, given the 92% A-+-T composition in the matching region. (Since the Cbs element has presumably been under selection pressure, its sequence was excluded from the chi-square analysis, although its inclusion makes a negligible probability difference.) Note that the sequence similarity occurs on both sides of the (underlined) Cbs element.

Figure 3.—

Two MAC chromosomes hybridize to labeled DNA probe from cloned inserts with functional Cbs junctions. Southern blots of three CHEF gels (pulsed-field conditions: 30 hr, 120°, 6 V/cm, 1× TAE, and 14°) of whole-cell DNA of inbred strains B (left lanes) and C3 (right lanes) were separately probed with labeled DNA from Cbs inserts 1L-3, 4L-2, and 5-7. MAC chromosome sizes are given in kilobases. No B-C3 size polymorphisms have been detected in any of the MAC chromosomes identified in this article, although one such polymorphism is known (Wong et al. 2000).

Figure 4.—

Age-related loss of the end of Cbs 4L-3G MAC chromosome. PCR amplification was primed with the Tel primer and a Cbs 4L-3G specific primer whose 5′-end is located 220 bp from the edge of the Cbs element. Template DNA, lanes 1 and 9: inbred strain B, <30 fissions old; lanes 2 and 10: inbred strain C3, <30 fissions old; lanes 3 and 4: SB2379P and SB2393P (meiotic segregant panel), <30 fissions old; lanes 5 and 6: CU378 and CU385 (nullisomic panel), >150 fissions old; lanes 7 and 8: SB1819 and SB1820 (terminal assortant panel) >500 fissions old. The expected band is indicated with an arrow; its presence or absence is indicated by + or − under each lane. Slightly reduced primer annealing stringency was used to be sure that the absence of the band was not the result of a failed PCR reaction.

Figure 5.—

Assortment patterns of CBSPs. Terminal assortants are written vertically and omit the “SB” prefix. The name of each coassortment group is prefixed by “cag.” Assortment data for CBSPs are shown in boldface type. B and C represent the allele (B and C3, respectively) to which each genetic locus has assorted in each terminal assortant. CBSPs that assorted independently of all previously mapped loci were assigned to a new coassortment group of their own. Underlined letters call attention to putative MAC recombinational events (see Wickert et al. 2000).