Aptamer Domains of Riboswitches

The lack of conservation observed among expression platforms reflects the degree to which they may be interchanged or adapted to control different processes. Li-gand-binding elements, however, tend to be evolutionarily more constrained compared with the expression platforms with which they interface. Unlike RNA elements that interact with protein factors, natural aptamers are prevented from significantly changing their recognition strategies simply because their metabolites targets are immutable, unable to co-evolve in step with their receptors. As genetic control elements, riboswitches are challenged with operating in an environment of enormous chemical complexity. Therefore, in order to avoid inappropriate genetic responses, their aptamer domains must be capable of discriminating with precision against compounds with related structures. These high levels of specificity exhibited by natural aptamers toward their ligands are often accompanied by impressive binding affinities, which in several instances extend into the nanomo-lar range (Table 8.1). Presumably, the avidity of a particular RNA-ligand interaction is tuned in order to respond appropriately over the physiologically relevant

Table 8.1 Known riboswitch classes

Riboswitch

Gene/operon

Aptamer

Kd (nmol/L)

References3

class

controlled

size (nt)

Coenzyme B12

E. coli btuB

202

300

Nahvi et al., 2002

TPP

E. coli thiM

78

30

Winkler et al., 2002a

Glycine

V. cholera gcvT

220b

30000

Mandal et al., 2004

FMN

B. subtilis ribD

121

5

Winkler et al., 2002b

SAM I

B. subtilis yitJ

118

4

Winkler et al., 2003

SAM II

A. tumefaciens metA

55

1000

Corbino et al., 2005

Guanine

B. subtilis xpt

66

5

Mandal et al., 2003

Adenine

B. subtilis ydhL

64

300

Mandal and Breaker, 2004

Lysine

B. subtilis lysC

169

1000

Sudarsan et al., 2003b

GlcN6P

B. subtilis glmS

150

200000

Winkler et al., 2004

a References pertain to the specific aptamer constructs used in determinations of Kd-values. b Size of the tandem motif containing two aptamers.

a References pertain to the specific aptamer constructs used in determinations of Kd-values. b Size of the tandem motif containing two aptamers.

concentration range for a given metabolite. It should be noted, however, that binding constants determined in vitro do not necessarily reflect the metabolite concentrations to which riboswitches respond in vivo. Analysis of an FMN-specific ribos-witch in Bacillus subtilis has revealed that the decision whether to terminate transcription is made before thermodynamic equilibrium is attained, indicating that kinetic parameters are likely to have a greater impact (Wickiser et al., 2005).

The impressive conservation of individual aptamer domains in no way implies that a particular motif represents the only solution by which RNA may recognize the respective ligand. Just as selections in vitro have generated multiple aptamers recognizing the same target molecule (Huang and Szostak, 2003; Sassanfar and Szostak, 1993; Sazani et al., 2004), nature also should have access to the numerous solutions in sequence space. An example of this has recently been provided with the identification of a second, distinct motif, occurring primarily in a-proteo-bacteria, that recognizes the coenzyme SAM (Corbino et al., 2005). This newly identified aptamer is significantly smaller and has a less complex secondary structure than the previously reported SAM-binding domain, which has been observed almost exclusively among Gram-positive organisms. Despite its unique architecture, the a-proteobacterial SAM aptamer binds its target with high affinity and discriminates quite effectively against structurally related analogs. The non-overlapping distribution of these two SAM-specific aptamer classes indicates that different bacterial subgroups can employ distinct recognition strategies for RNA-metabolite interactions.

Despite the impressive phylogenetic conservation exhibited by aptamer domains of a given riboswitch class, these motifs are subject to subtle degrees of variation. Among individual TPP aptamers, for example, an extraordinarily wide range of sizes of the P3 stem-loop is tolerated, presumably because this element has little impact on the global architecture of this motif (Rodionov et al., 2002; Sudarsan et al., 2003a). Similarly, helical elements commonly occurring at defined positions within a riboswitch class need not be present in all examples, as is the case with the RNA motif recognizing glycine (Barrick et al., 2004).

Although more drastic deviations from consensus structures are not commonly observed among known riboswitches, substantial change to an aptamer core is not unprecedented. For example, a variant of the AdoCbl-binding motif has been identified in which nearly one half of the consensus structure is absent (Nahvi et al., 2004). Despite the severity of the modification, the truncated motif nonetheless binds AdoCbl with high affinity. This abbreviated aptamer contains a short sequence element not found in larger versions and discriminates less efficiently against AdoCbl analogs varying at the adenosyl moiety, suggesting that different subdomains may interact with distinct ligand moieties in a modular fashion. These observations demonstrate that an aptamer can depart significantly from its consensus while maintaining specificity toward its cognate ligand. Conversely, aptamers also exist in which ligand specificities may be changed without substantial alterations to their global architectures. The first such demonstrations occurred in the laboratory, where artificial aptamers were subjected to further evolution in vitro in attempts to create new ligand specificities (Famulok, 1994; Souk-

up et al., 2000). For example, an arginine-selective aptamer, which was evolved from an RNA motif recognizing the related compound citrulline, differs from its parent at only three nucleotide positions (Famulok, 1994). Structural studies have revealed that the two aptamers share a similar overall architecture, but discriminate between the two ligands through subtle changes to the binding pocket effected by the nucleotide substitutions (Yang et al., 1996).

Nature has relied on a similar approach in recognizing the related metabolites guanine and adenine with distinct but closely related motifs. These two classes of natural aptamers, despite exhibiting preferential binding to their respective purine targets, possess nearly identical architectures. The features of these motifs, including detailed structural views that have emerged from the first X-ray crystallo-graphic studies of natural aptamers, are discussed in more detail below.

0 0

Post a comment