Apical Loop Internal Loop Interactions

The presence of purine-purine "pairs" closing the loop of RNA hairpins giving rise to kissing complexes both in naturally interacting RNA species (the DIS element of HIV-1) and in selected RNA motifs (aptamers to TAR or to tRNAPhe) suggested that this might be a rule for generating stable loop-loop interactions. Indeed the substitution of a GC pair by a G,A combination at the top of the stem of the hairpin RNA I' stabilized its interaction with RNA II', two stem-loops derived from the structure involved in plasmid Col E1 replication (Duconge et al., 2000). However, the rational design according to these rules of a kissing hairpin potentially able to interact with the SL1 structure of the HCV mRNA (i.e. a hairpin with a loop complementary to that of SL1 framed by G and A residues) led to a complex of weak stability. Therefore, not every loop is prone to kissing complex formation.

This is related to both the size and the sequence of the loop. The most common loop size for kissing complexes is 6-7 nucleotides (Brunel et al., 2002). When a larger loop is involved only part of it contributes to loop-loop helix formation. This is actually driven by the distance that can be spanned by a single ribose phos-phodiester unit for connecting the 3' end of a stem to the 5' end of the loop-loop helix, stacked on the 3' strand of the stem, through the major groove of the "kiss" (Haasnoot et al., 1986).

RNA aptamers selected against the SL1 hairpin within a library of 1013 candidates with a 40-nt random window led, after 10 rounds of selection, to aptamers containing 6-10 contiguous nucleotides complementary to the apical part of SL1 (Aldaz-Carroll et al., 2002). This might allow the formation of a helix comprising the entire 6-nt loop and the top part of the stem, of the target hairpin assuming the opening of the double-stranded region. The consensus motif was invariably located in a loop of the aptamers. Interestingly the selected sequences could be ranked in two different classes. Hairpins in which the consensus motif is displayed in the apical loop led to mediocre ligands characterized by a K > 1 mmol/L, confirming the result obtained with the rationally designed kissing aptamer (see above).

The largest family of anti-SL1 aptamers actually corresponds to imperfect hairpins with a prominent internal loop complementary to the apical loop of the target, thus generating apical loop-internal loop (ALIL) complexes (Fig. 7.4) as demonstrated by footprinting and mutational analysis for the strongest ligand, the aptamer 5-39 (Aldaz-Carroll et al., 2002). As previously discussed for the kissing RNA aptamer R06 targeted to TAR, tertiary interactions contribute to the stability of the ALIL complex in addition to the formation of the loop-loop helix.

Fig. 7.4 Aptamer giving rise to an apical loop-internal loop (ALIL) complex. The consensus sequence (orange) of such an aptamer is invariably located in an internal loop complementary to the apex of the target hairpin (thin lines in the right scheme), giving rise to a loop-loop helix through base pairing (blue dots). Note that the loop-loop helix may involve only part of the internal loop.

First, the deletion of the top part of the aptamer 5 -39 (above the internal loop) abolishes its binding to SL1 and secondly an antisense RNA 9-mer identical to the internal loop sequence displays a Kd about 20-fold higher than that of the intact aptamer. In both cases the molecules retained the base-pairing properties of the parent aptamer. In contrast, the deletion of the apical loop or its substitution by a non-nucleotide linker did not prevent the binding of the 5-39 aptamer to SL1 (Aldaz-Carroll et al., 2002). Interestingly the upper and lower stems flanking the internal loop of the aptamers selected against SL1 displayed a very high G,C content: the upper stem of the aptamer 5 -39 is composed of four GC pairs whereas five out of the six pairs of the lower stem adjacent to the internal loop are also GC ones.

Such ALIL-forming aptamers have been independently selected in our laboratory against three different targets: namely the hairpins corresponding to domain IV (Aldaz-Carroll et al., 2002) and to the top part of domain II of the internal ribosome entry site (IRES) of the hepatitis C virus mRNA (Da Rocha Gomes et al., 2004). These aptamers shared the following properties which consequently can be considered as characteristics:

x An internal loop containing a sequence complementary to the apical loop of the target hairpin. The loop-loop helix contained potentially 6-9 base pairs. x A few nucleotides (two or three) connect the sequence, giving rise to the loop-loop helix on either the lower or the upper stem of the aptamer. On the target a single nucleotide constitutes a link between the stem and the loop-loop helix. This is at variance from the kissing complexes for which no nucleotide is required in the connector, even though the same number of base pairs (6-7) are formed in the loop-loop association. • G,C-rich double-stranded stems on both 5' and 3' sides of the internal loop.

Kikuchi et al. (2005) isolated aptamers targeted to domain IIId of the HCV IRES which fulfilled several of the above criteria. The strongest ligands showed five bases, complementary to part of the apical loop of the domain IIId, in a 6-nt-long internal loop. The authors demonstrated the formation of an ALIL complex using RNase footprinting. Recently a study aiming at identifying the determi nants of the interaction between tRNA and the tyrST box antiterminator of Bacillus subtilis led to the unanticipated selection of aptamers that generate ALIL associations (Fauzi et al., 2005). In this complex a seven base pair loop-loop helix could be formed. The internal loop involved in the interaction is flanked by two GC-rich stems: seven out of the eight base pairs are GC pairs. The internal loop is connected to the stems by one or three residues depending on the formation of a GU pair at the bottom of the upper stem. Therefore this fulfills the criteria of ALIL complexes.

Interestingly, natural RNA-RNA complexes are known that do engage ALIL interactions. A first example is constituted by the bicoid (bcd) mRNA that codes for a protein involved in the development pattern of Drosophila melanogaster (Ferran-don et al., 1997). cis-Acting sequences have been identified within the 3'-UTR of bcdmRNA that are responsible for the dimerization of the message. These regions contain two complementary sequences AAGCCC and GGGCUU located in an apical loop and an internal loop, respectively. Indeed, dimers can be formed by these elements that involve two ALIL interactions. The B. subtilis bacteriophage phi29 provides a second example of such interactions. The packaging of the phage DNA involves the virus-encoded pRNA that shows the propensity to generate an hexamer (Guo et al., 1998). The packaging unit results from an ALIL association giving rise to "hand-by-arm" interaction. Therefore ALIL complexes constitute a motif for RNA-RNA recognition and assembly.

0 0

Post a comment