Structure of the chromosome VII centromere region in Neurospora crassa: degenerate transposons and simple repeats

Abstract

DNA from the centromere region of linkage group (LG) VII of Neurospora crassa was cloned previously from a yeast artificial chromosome library and was found to be atypical of Neurospora DNA in both composition (AT rich) and complexity (repetitive). We have determined the DNA sequence of a small portion (approximately 16.1 kb) of this region and have identified a cluster of three new retrotransposon-like elements as well as degenerate fragments from the 3' end of Tad, a previously identified LINE-like retrotransposon. This region contains a novel full-length but nonmobile copia-like element, designated Tcen, that is only associated with centromere regions. Adjacent DNA contains portions of a gypsy-like element designated Tgl1. A third new element, Tgl2, shows similarity to the Ty3 transposon of Saccharomyces cerevisiae. All three of these elements appear to be degenerate, containing predominantly transition mutations suggestive of the repeat-induced point mutation (RIP) process. Three new simple DNA repeats have also been identified in the LG VII centromere region. While Tcen elements map exclusively to centromere regions by restriction fragment length polymorphism analysis, the defective Tad elements appear to occur most frequently within centromeres but are also found at other loci including telomeres. The characteristics and arrangement of these elements are similar to those seen in the Drosophila centromere, but the relative abundance of each class of repeats, as well as the sequence degeneracy of the transposon-like elements, is unique to Neurospora. These results suggest that the Neurospora centromere is heterochromatic and regional in character, more similar to centromeres of Drosophila than to those of most single-cell yeasts.

FIG. 1

Schematic diagram illustrating the structure of the centromere of LG VII and the region that was subcloned and sequenced. (A) The centromeric interval between the qa-2 and met-7 genes on LG VII and the YAC clones that cover this region. YAC 10-8-H is internally deleted (dotted line) but physical mapping demonstrates greater than 80 kb of colinearity with the left end of YAC 19-5-B, in the region of overlap with YAC 12-10-H (11). (B) The two overlapping regions that were subcloned from YACs by plasmid rescue (10). (C) The arrangement of transposons, transposon fragments, and simple sequence repeats within the subcloned region shown in panel B at the same scale.

FIG. 2

A theoretical translation of the composite of the three interrupted Tgl1 fragments denoted in Fig. 1C. Amino acid sequences of the pol region of three fungal Gypsy-like transposons, Ty3 (25), Maggy (20), and Grasshopper (17), were aligned by using the CLUSTAL algorithm (27), and the putative Tgl1 amino acids were subsequently aligned by hand. Raw sequence data was used for the translation of Tgl1, but some frameshifts were included to improve the alignment. Periods in the sequence refer to translational stop codons, and dashes indicate spaces inserted by the algorithm, or in the Tgl1 case, by hand. Regions of identity are boxed with open boxes if all four residues are identical or with shaded boxes to indicate three of four identities. Possible progenitor amino acids that can be derived from the nucleotide sequence, if one GC to AT transition is posited, are bracketed. The dTad1 and dTad2 insertion sites are indicated by triangles.

FIG. 3

Defective Tad elements found elsewhere in the genome of N. crassa. (A) A schematic representation of the de novo insertion event of the Tad1-1 LINE-like transposon into the dTad3 element residing near the centromere of LG III (7). (B) Alignment of the 3′ ends of dTad3 and Tad1-1. Identities are indicated by solid lines, and transition mutations are shown as double dots. The alignment demonstrates substantial colinearity of the sequence and nearly 90% sequence similarity if changes due to RIP are taken into account (see Table 1). (C) A schematic representation of a subtelomeric dTad element found approximately 200 bp internal to the TTAGGG telomere repeats of LG IV R, derived from previously published sequence from the telomere-specific clone lambda-11R (59).

FIG. 4

Amino acid alignment of Tcen with the RNase H region of four known copia-like transposons. (A) Sequences were aligned as in Fig. 2, but a GC consensus sequence of Tcen was first derived from a comparison of the nucleotide sequences of the ends of three different copies of Tcen. If any of the three sequences contained a G or C rather than an A or T, this was used for the consensus sequence (to correct for RIP-induced transitions). Theoretical translation of the consensus results in an open reading frame with substantial similarity to the corresponding regions of the copia-like elements shown (Tnt1 [21], BARE-1 [42], Osser [38]). (B) Alignment of LTR end sequences from Tcen, two other unlinked Tcen elements (R5 and R10), three classes of transposons from S. cerevisiae (Ty1, Ty3, and tau), and retrovirus Moloney murine leukemia virus (MoMLV) (see reference 4). TSDs flanking Tcen are underlined, TG and CA end elements are boxed, and terminal duplications (inverted or direct) are indicated by arrows.

FIG. 5

Alignment of the critical regions of the various domains of the pol region of known retroelements and Tcen. Alignment of the protease, integrase, and reverse transcriptase regions was as described previously (4). The homologies within each domain of the Tcen element were found in the expected regions, and the order of the homologous domains was similar to that within known copia-like transposons. Amino acids shared by all four elements are boxed and indicated above the aligned sequence; regions of similarity are also boxed and shaded. No GC consensus in the translation of Tcen was used in this alignment.

FIG. 6

Distribution of Tcen and Tad homologous sequences across the LG VII centromere region. (A) Southern blot of genomic DNAs from yeast strains that carry the four YACs shown in Fig. 1. The lanes contained DNA from strains containing YAC 12-10-H (lanes 1), YAC 15-6-H (lanes 2), YAC 19-5-B (lanes 3), and YAC 24-3-G (lanes 4). Genomic DNA was isolated and digested with either BglII, SalI, or HindIII restriction enzyme. The DNAs were fractionated and blotted as described in Material and Methods and were sequentially probed with (i) the hph gene (a YAC-specific end probe), (ii) a Tcen LTR-specific probe, and (iii) an internal Tad probe. Multiple overlapping bands can be observed in the autoradiograms from the Tcen and Tad probings. (B) A schematic diagram showing the probable location of Tcen and Tad clusters in the LG VII centromere region. Autoradiograms were scanned, and the panel was created with Adobe Photoshop 4.0.

FIG. 7

Analysis of simple repeated DNAs from the LG VII centromere region. (A) The nucleotide sequence of the rescued centromeric region containing three classes of simple repetitive DNAs. Each repeat is shown underlined, and the individual Tsp and Sma satellite subunits are additionally delineated by arrows. (B) Southern blot of Neurospora genomic DNA, probed sequentially with labeled DNA segments containing the Sma and Tsp repeats. DNA was isolated, digested with either AseI or SspI restriction enzyme, and fractionated and blotted as described in Materials and Methods. The Sma repeat is repeated at a relatively high copy number within the Neurospora genome. The Tsp repeat also appears to be repeated but at a much lower copy number. The images were processed as described for Fig. 6.