Enrichment of a LINE subfamily in a single chromosomal region in Peromyscus David H. Kass, 1 John A. Peppers, 2 Mary Maltbie, 2 Robert J. Baker 2 1 Department of Biology, Eastern Michigan University, Ypsilanti, Michigan 48197, USA 2 Department of Biological Sciences, Texas Tech University, Lubbock, Texas 79409, USA Received: 15 December 1997 / Accepted: 20 February 1998 LINEs are a class of transposable elements greater than 5 kb in length consisting of 10 4 –10 5 copies, most of which are hetero- geneously 5' truncated (Voliva et al. 1983). These elements are not randomly inserted in the genome, and generally correspond to the A-T rich chromosomal G bands, as well as accumulate on the sex chromosomes, as observed in the human (Koren- berg and Rykowski 1988), European house mouse (Boyle et al. 1990), and deer mouse (Wichman et al. 1992; Baker and Kass 1994). Sequence analysis of L1 elements suggests few original source or master genes generated the numerous L1 elements (Deininger et al. 1992). As mutations accumulate in the master gene, the more recently diverged species would share these mutations and yield concerted variations and hence subfamilies. LINE subfamilies have been described by restriction endonuclease cleavage patterns (Jubier-Maurin et al. 1985; Kass et al. 1992), sequences of 5' motifs (Padgett et al. 1988), diagnostic nucleotides (Jurka 1989), and stretches of sequence variability within an open reading frame (ORF; Pascale et al. 1990; Kass et al. 1992). Two highly divergent LINE (L1) subfamilies (L1Pm62 and L1Pm55) have been identi- fied that coexist in the genomes of Peromyscus, exhibiting differ- ing restriction fragment length variants (RFLVs) on Southern blots with DNA isolated from various Peromyscus species (Kass et al. 1992). Sequences of the elements (subcloned L1-hybridizing frag- ments from a P. maniculatus genomic library) representing these subfamilies demonstrate 29% divergence from each other within ORF-2, which is unusually high for intraspecific LINEs compared with mice and rats (Martin et al. 1985; Soares et al. 1985). Com- parisons of second and third position nucleotides by codon align- ment between L1Pm62 and L1Pm55 support the view that these subfamilies were derived from different source genes (Kass et al. 1992). A comparative sequence analysis of several Peromyscus LINEs depicts L1Pm62 and L1Pm55 as relatively ‘‘old’’, predat- ing the divergence of two recent LINE-1 lineages (Casavant et al. 1996). The lack of a conserved ORF additionally supports the ancient origin of the L1-62 and L1-55 subfamilies (Kass et al. 1992) and is indicative of retropositional inactivity during the course of speciation within this genus. However, the maintenance of conserved restriction sites that follow phylogenetic trends (Kass et al. 1992) suggests these subfamilies have more recent evolu- tionary histories. L1Pm55 appeared to be the oldest sequence of elements iso- lated from P. maniculatus and branches as an off-shoot from the other Peromyscus L1 sequences from a derived molecular phylogeny (Casavant et al. 1996). L1Pm55 was estimated at 100 copies per haploid genome by a genomic Southern blot (Kass et al. 1992), representing less than 1% of total LINE repeats. The same blot demonstrated a less intense EcoRI fragment in species of the P. leucopus (0.7 kb) and P. truei (1.7 kb) species groups, indicative of fewer copies or high sequence divergence. The use of alternative restriction enzymes for Southern blots also demonstrated concerted variations, primarily at the species group level, always with a more intense band in species of the P. manicu- latus species group. Surprisingly, this suggests a more recent ori- gin of this apparently retropositionally inactive subfamily. We therefore elected to analyze the chromosomal organization of L1Pm55 to gain insight into the evolutionary history of this enig- matic subfamily. Fluorescence in situ hybridization (FISH) as previously de- scribed (Wichman et al. 1992; Baker and Kass 1994) was used to analyze chromosomes of P. leucopus, P. maniculatus, and P. mela- notis. The L1 hybridizing probe (pDK55) is a 1.5-kb fragment subcloned from a P. maniculatus genomic library (Kass et al. 1992). Although two small regions within pDK55 (80 bp and a 27-bp GC-rich region) are not observed in L1Pm62 or L1 se- quences from other mammals (Kass et al. 1992), it is highly im- probable that the observed FISH patterns are due to homologies to these short DNA stretches. A BLAST search of the database (Altschul et al. 1990) identified no corresponding sequences to the 27-bp region. Various sequences located on different mouse and human chromosomes were identified that corresponded to the 80- bp region, and any significant sequence identities (80%) that were observed included only various 20-bp stretches of this region. One microgram of the clone (plasmid plus insert) was biotinylated by nick translation, and the pDK55 probe was hybridized in the pres- ence of sheared E. coli DNA. Hybridization under the conditions used (Baker and Kass 1994) should detect sequences greater than 70% in similarity. The hybridized probes were detected with fluo- rescein-conjugated avidin (Vector Labs), followed by two ampli- fications consisting of alternate treatments of biotinylated goat anti-avidin (Vector Labs). Digital enhancement of images and pseudo-G banding were accomplished by using software from On- cor Images. Repeated hybridizations for each species confirmed the results. The L1-55 subfamily is primarily located on a single pair of homologous chromosomes (Fig. 1) within P. maniculatus and P. melanotis, both of the P. maniculatus species group, and in P. leucopus. The distinguishable FISH pattern in P. leucopus further supports the subfamily classification of L1-55, as opposed to one or a few aberrant element(s) strictly within the P. maniculatus genome. The accompanying pseudo-G band spreads support the existence of this subfamily on the same chromosome in these species. The use of P. melanotis further denotes Chr 12 as the residence for L1-55, based on G banding patterns (Committee for the Standardization of Chromosomes of Peromyscus, 1994). L1-55 had not been observed by FISH in P. eremicus (data not shown), a member of a separate subgenus (Haplomyomys) of Peromyscus, either as a consequence of the level of divergence, low copy num- ber, or its absence. Several mechanisms for the non-random chromosomal ar- Correspondence to: D.H. Kass Mammalian Genome 9, 488–490 (1998). © Springer-Verlag New York Inc. 1998 Incorporating Mouse Genome