DNA Target Site Recognition by LlLtrB Intron RNPs

Thus far, the most detailed studies of DNA target site recognition have been carried out for the LLLtrB intron. Figure 5a shows a model of the target site interactions for the LLLtrB homing endonuclease, identified by DNA foot-printing, modification-interference, and missing-base analysis (Singh and Lambowitz 2001). Figure 5b shows the information content of different positions in the target site and intron RNA (Zhong et al. 2003; Perutka et al. 2004). Critical bases in the target site were also identified by extensive mutagenesis and selection experiments (Guo et al. 2000; Mohr et al. 2000; Zhong and Lambowitz 2003).

Biochemical analysis showed that Ll.LtrB RNPs bind DNA non-specifical-ly in a rapid diffusion-limited process and then search for DNA target sites, presumably via a facilitated diffusion mechanism like those used by site-specific DNA-binding proteins (Aizawa et al. 2003). Figure 5 shows that most of the interactions with the IEP are in the distal 5'-exon region of the DNA target site (positions -23 to -14), with one set of critical bases, T-23, G-21, and A-20, being recognized via major groove interactions (Singh and Lambowitz 2001). The recognition of these bases maybe the initial step in DNA target site recognition. Ethylnitrosourea modification-interference experiments showed that the IEP also makes functionally important phosphate-backbone contacts along one face of the helixbetween positions -24 and -13. The IEP crosses the minor groove between -17 and -13, and there are preferences for specific bases in this region (positions -17,-15, and -14), but it is not clear if these reflect direct base recognition or indirect readout of DNA structure. Notably, mutations at critical nucleotide residues in the distal 5'-exon region (positions -23, -21, -20, and -15) strongly inhibit reverse splicing into double-stranded but not otherwise identical single-stranded DNA target sites, implying that their

Fig. 5. Model and information content analysis of the L. lactis Ll.LtrB intron DNA target site, a Target site model summarizing critical residues identified by DNA footprinting, modification-interference and missing-base analysis, reproduced from Singh and Lambowitz (2001). The Ll.LtrB target site is represented simply as a B-form helix; its actual conformation in the complex is unknown. Critical bases recognized by the IEP are colored red, and critical phosphates are colored dark and light purple, indicating strong and weak interference, respectively, by ethylnitrosurea modification. Blue indicates bases protected

of the DNA target site is indicated by a bracket above. The intron-insertion site (IS) and bottom-strand cleavage site (CS) are indicated by arrows, b Information content analysis of DNA target site and intron RNA positions involved in target site recognition. The analysis is based on 88 random insertion sites in the E. coli genome, obtained using an Ll.LtrB intron with randomized EBS and 6 sequences (Zhong et al. 2003). The logo shows the information content (bits) for each position represented by the size of the letter (http://weblogo. berkeley.edu/)

recognition by the IEP is required primarily for local DNA unwinding (Zhong and Lambowitz 2003).

The intron RNA's EBS2 and EBS1/S sequences are located in two separate stem-loop structures (denoted DIdl and DId3(ii) in the Ll.LtrB intron; Figs. la,c, and 4). Base pairing of these sequences to the IBS and 6' sequences in the target DNA may occur concomitantly with and help drive local DNA unwinding (Singh and Lambowitz 2001). Detailed analysis showed that the DNA positions recognized by base pairing are -12 to -8 (IBS2), -6 to -1 (IBS1), and +1 and +2 (6'; Perutka et al. 2004). The complementary EBS2, EBS1 and 6 positions in the intron RNA are numbered according to the DNA target site position with which they pair. Although nucleotide residues at intron RNA positions -13,-7 and +4 could also potentially base pair with the DNA target site, the U at RNA position -13 instead pairs with the opposite G in the intron RNA to form the top of the DIdl stem, and the G and U at RNA positions -7 and +4, respectively, pair with each other to form the top of DId3(ii) stem (see Fig. 4, top). Additionally, an A-residue is favored at the intron RNA +3 position regardless of whether it can base pair with the target site, suggesting that the contribution of a favorable loop structure can override any additional contribution from base pairing at this position. The information content analysis and mutagenesis experiments show that base pairing at some positions is more important than others and that there are preferences for specific base pairs at certain positions (Guo et al. 2000; Mohr et al. 2000; Zhong et al. 2003; Perutka et al. 2004). For example, positions -12 and -6 at the beginning of IBS2 and IBS1 show strong preferences for GC or CG base pairs, possibly to anchor the ends of these duplexes (see Fig. 5 and references cited above).

Bottom-strand cleavage occurs after reverse splicing and requires the same IEP and base-pairing interactions that are required for reverse splicing, as well as additional interactions between the IEP and the 3' exon. Experiments using RNPs containing lariat RNA whose 3' OH had been blocked by per-iodate oxidation showed that reverse splicing per se is not required for second-strand cleavage (Aizawa et al. 2003). For the LLLtrB intron, the most critical IEP interaction with the 3' exon is at T+5 (Fig. 5). Mutation of this base has no effect on reverse splicing but strongly inhibits second-strand cleavage (Mohr et al. 2000). The bases flanking T+5 appear to make smaller contributions (Fig. 5b). Although interactions between the IEP and the 3' exon are not required for DNA unwinding, the locally unwound region extends into the 3' exon and includes T+5, perhaps facilitating its recognition for bottom-strand cleavage (Singh and Lambowitz 2001).

The En domain of the IEP also interacts at the bottom-strand cleavage site between positions +9 and +10, but there are no strong sequence requirements in this region, and single phosphate-backbone modifications are not inhibi tory in ethylnitrosourea modification-interference experiments (Singh and Lambowitz 2001). Thus, positioning of the En active site for second-strand cleavage appears to be determined mainly by the recognition of T+5 and, to a lesser extent, neighboring bases, with IEP interactions around the cleavage site being relatively weak and non-specific. This also appears to be the case for the Bombyx mori non-LTR-retrotransposon R2Bm, which uses an analogous TPRT mechanism in which a separate En domain of the RT cleaves the DNA target site to generate the primer for reverse transcription (Christensen and Eickbush 2004). As in that case, weak, non-specific binding at the cleavage site may facilitate transfer of the cleaved 3' end to the RT active site for initiation of TPRT.

0 0

Post a comment