Ribozyme Structure

Structures have been obtained from crystals of a modified form of the genomic ribozyme (Ferre-D'Amare et al. 1998a; Ferre-D'Amare and Doudna 2000; Ke et al. 2004). No physical structure is available for the antigenomic ribozyme, but because the secondary structures of the genomic and antigenomic ri-bozymes are so similar, it is likely that the three-dimensional structures will be similar as well. For an in-depth discussion of the structure and its significance, readers are directed to papers from the Doudna lab (Ferre-D'Amare et al. 1998a; Ferre-D'Amare and Doudna 1999, 2000; Doherty and Doudna 2000; Ke et al. 2004). A partial overview will be provided here and select aspects will be discussed later in context of the catalytic mechanism.

Each ribozyme contains five duplexes or pairings labeled P1, P2, P3, P4 and P1.1. Extensive duplex formation and a nested pseudoknot arrangement of those duplex elements generates a compact structure. The importance of base pairing in all five duplex regions for in vitro cleavage activity was established by mutagenesis experiments in both ribozymes (Perrotta and Been 1991, 1993; Been et al. 1992; Thill et al. 1993; Wadkins et al. 1999; Nishikawa and Nishikawa 2000). An in vivo requirement for base pairing in P1 and P2 is also supported by the effect of mutations (Jeng et al. 1996). The lengths of P1, P3 and P1.1 (7 bp, 3 bp and 2 bp, respectively) are the same for both ribozymes and changes to their lengths might be expected to distort the structure of the active site. P1 and P3 show some sequence variation between the genomic and antigenomic ribozymes but those sequences do not vary amongst clinical isolates (Tanner 1995; Been and Wickham 1997; Wadkins and Been 2002). P1.1, an invariant and essential two base-pair duplex that forms a coaxial connector between P1 and P4 (Ferre-D'Amare et al. 1998a), was only identified upon solving the crystal structure of the genomic ribozyme, and was not a feature of the secondary structures as originally proposed (Perrotta and Been 1991). P1.1 is always CC paired with GG, and that combination appears optimal for activity (Wadkins et al. 1999; Nishikawa and Nishikawa 2000). P4 is a long, imperfect duplex that extends away from the core ribozyme. It shows more sequence variation than the rest of the ribozyme domain, both between the two ribozymes and amongst the clinical isolates. The tolerance to sequence variation in P4 was exploited for the structural work. For the structures, the P4 hairpin was largely replaced by a U1A protein binding site and the crystals grown as the ribonuclear protein complex (Ferre-D'Amare et al. 1998b). While P4 can be shortened without loss of in vitro ribozyme activity, the issue of whether the longer natural P4 hairpin sequence has a role in viral replication or ribozyme activity in vivo remains an interesting question. The 3' end of P2 defines the 3' boundary of the ribozyme domain. P2 would be predicted to play an important role in positioning Cyt75/76, the proposed catalytic cytosine, contained within the J4/2 joining segment. Alterations to the sequence and length of P2 can dramatically affect the extent of cleavage in vitro, which suggests that P2 may also function in directing the correct RNA folding (Perrotta and Been 1998; Perrotta et al. 1999a; Chadalavada et al. 2000, 2002).

The cleavage site is located at the 5' end of P1, but in the structure (Ferre-D'Amare et al. 1998a; Ke et al. 2004) this position is well buried in an active-site pocket between the two parallel coaxial stacks of P1-P1.1-P4 and P2-P3. The short sequence connecting P4 and P2 (J4/2) also forms part of that pocket and also contributes and positions Cyt75. No strict requirements for sequence 5' to the cleavage site have been identified (Perrotta and Been 1990, 1992; Shih and Been 1999). There are, however, sequence preferences. For example, a guanosine just 5' to thecleavagesiteslows cleavage activity roughly tenfold under our standard reaction conditions (Perrotta and Been 1992). In addition, sequence 5' to the cleavage site can interfere with ribozyme activity by participating in alternative pairings that disrupt the ribozyme structure (Perrotta and Been 1990, 1991; Chadalavada et al. 2000). Notably, cleavage site selection by the ribozyme does not appear to require the sequence 5' to the cleavage site to base pair with a particular sequence within the ribozyme domain. The most recent structures (Ke et al. 2004), which are of a precursor form of the genomic ribozyme, reveal a sharp bend in the backbone at the cleavage site that nearly reverses the direction of the RNA backbone. Two nucleotides just 5' of the cleavage site occupy a tight space between P1 and P3 without making specific base-base contacts.

Thestructures of the3' productformand themorerecentprecursor form of the ribozyme reveal changes that occur upon backbone cleavage and provide invaluable insights into the catalytic mechanism (Ferre-D'Amare et al. 1998a; Ke et al. 2004). To prevent self-cleavage, the precursor structure was solved using an inactive ribozyme with a uracil replacing Cyt75. Near the position of the scissile phosphate, at the bend in the substrate strand mentioned above, are located a bound divalent metal ion and the pyrimidine base at position 75 (Ke et al. 2004). While there do not appear to be major changes in the general architecture of the ribozyme upon cleavage, there are local conformational changes in the vicinity of the cleavage site. It appears that, following cleavage, the 5' fragment bearing the newly generated 2',3'-cyclic phosphate is released along with the metal ion, and the space between P1 and P3 that had been occupied by the 5' fragment narrows slightly. The nucleotide sugar of G1, at the newly generated 5' end, moves further into the active site pocket. Thus, the 5' hydroxyl oxygen of the product is displaced relative to its position when it functioned as the 5' bridging oxygen of the precursor. Very importantly, the base at position 75 has moved in the product relative to its location in the precursor structures. In the product, Cyt75 can form a hydrogen bond between its N3 (or O4) and the 5' hydroxyl group oxygen of G1, while in the precursor, the uracil base at position 75 is positioned higher (~2 A) in the active site and away from that same oxygen. As such, it is positioned closer to the predicted location of the 2'-OH group of the sugar at position -1, but only following a proposed rotation of that nucleotide that provides a more favorable geometry for the cleavage reaction. Knowing the site of action of Cyt75 is important for understanding details of the catalytic mechanism.

0 0

Post a comment