Engineered allosteric ribozymes that respond to specific divalent metal ions

Abstract

In vitro selection was used to isolate five classes of allosteric hammerhead ribozymes that are triggered by binding to certain divalent metal ion effectors. Each of these ribozyme classes are similarly activated by Mn2+, Fe2+, Co2+, Ni2+, Zn2+ and Cd2+, but their allosteric binding sites reject other divalent metals such as Mg2+, Ca2+ and Sr2+. Through a more comprehensive survey of cations, it was determined that some metal ions (Be2+, Fe3+, Al3+, Ru2+ and Dy2+) are extraordinarily disruptive to the RNA structure and function. Two classes of RNAs examined in greater detail make use of conserved nucleotides within the large internal bulges to form critical structures for allosteric function. One of these classes exhibits a metal-dependent increase in rate constant that indicates a requirement for the binding of two cation effectors. Additional findings suggest that, although complex allosteric functions can be exhibited by small RNAs, larger RNA molecules will probably be required to form binding pockets that are uniquely selective for individual cation effectors.

Figure 1

Allosteric selection of allosteric hammerhead ribozymes that respond to cations. (A) RNA construct used for in vitro selection of allosteric ribozymes. The hammerhead ribozyme core is identical to that described previously (21). A 40 nt random-sequence domain (N₄₀) replaces the majority of stem II of the hammerhead ribozyme core. The site of ribozyme cleavage is indicated by the arrowhead. (B) Scheme for the isolation of cation-dependent ribozymes by allosteric selection. Each population of ribozyme variants is prepared by transcription in vitro (1), subjected to negative (2) and positive (3) selections, and the cleaved RNAs that are enriched for allosteric function are amplified by RT–PCR (4). (C) Progress of the allosteric selection process. Values for signal and background reflect the fraction of RNA cleaved at step 3 in the presence or absence of the cation effector mixture, respectively. A ratio of 1 is expected if the addition of the cation mixture has no effect on ribozyme activity.

Figure 2

Five distinct classes of allosteric ribozymes from G11 that respond to specific divalent metal ions. Clones are classified by the similarity in nucleotide sequence.

Figure 3

The dependence of ribozyme activity on the concentration of divalent metal ion effector. Plots of the logarithm of the rate constant for ribozyme cleavage versus the logarithm of the concentration of divalent metal effector for the class II (A) and class IV (B) allosteric ribozymes. Each point represents a k_obs value calculated by conducting a time course for ribozyme activity and determining the negative slope of the line from a plot of the natural logarithm of the fraction of RNA remaining uncleaved versus time (min⁻¹). For the class II ribozyme, the dotted lines in the case of Cd²⁺, Co²⁺ and Ni²⁺ represent the expected plot for an ideal ribozyme with a k_max of 3 min⁻¹, a k_min of 4.8 × 10⁻⁵ min⁻¹ and two critical metal-binding sites each with an apparent K_D of 130 μM, as determined by a variation of the Henderson–Hasselbach equation. The apparent K_D, k_max and k_min parameters plotted for Mn²⁺ are 100 μM, 1.5 min⁻¹ and 6.7 × 10⁻⁵ min⁻¹ and for Zn²⁺ are 220 μM, 3 min⁻¹and 4.8 × 10⁻⁵ min⁻¹, respectively. For the class IV ribozyme, dotted lines in the case of Cd²⁺, Co²⁺ and Ni²⁺ represent the expected plots for and ideal ribozyme with an apparent K_D of 1 mM, a k_max of 1 min⁻¹ and one critical metal-binding site. The apparent K_D, k_max and k_min parameters plotted for Zn²⁺ are 5 × 10⁻⁴, 0.01 min⁻¹ and 2 × 10⁻⁴ min⁻¹ and for Mn²⁺ are 0.07, 0.7 min⁻¹ and 3.5 × 10⁻⁴ min⁻¹, respectively.

Figure 4

Rapid switching properties of cation-dependent allosteric ribozymes. Class II (A) and class IV (B) ribozymes were assayed for allosteric function at 23 and 50°C. The fraction of precursor RNAs remaining uncleaved at various times is plotted, wherein 1 mM CoCl₂ was added to the assay after 10 min (dashed line).

Figure 5

Class II ribozyme artificial phylogeny and predicted secondary structure. (A) Nucleotide sequence of the N₄₀ domain of the class II ribozyme parent is shown along with specific mutations exhibited by 12 variants recovered after additional mutagenesis and selection was completed. (B) Sequence and secondary structure model of the parental class II ribozyme. (C) Ribozyme assays conducted under standard conditions with 1 mM CoCl₂ at 23°C for 5 min with the parental (Par.) class II ribozyme, or on mutants carrying either disruptive (M1) or compensatory (M2) nucleotide changes (boxed). Open and filled arrowheads identify bands corresponding to 5′-³²P-labeled full-length precursor (Pre) or the 5′-cleavage product (Clv), respectively.

Figure 6

Class IV ribozyme artificial phylogeny and predicted secondary structure. (A) Primary sequence of the metal-binding domain is shown for ribozymes isolated from reselection on initial populations where 18% degeneracy was introduced into the metal-binding domain. (B) Sequence and secondary structure model of the parental class IV ribozyme. (C) Standard ribozyme assays conducted on parental (Par.) or mutant (M1, M2) ribozymes containing either disruptive (M1, M3) or compensatory (M2, M4) mutations (boxed). Open and closed arrowheads indicated trace radiolabeled full-length precursor or cleaved product, respectively.

Figure 7

The effects of various metal cations on class II ribozyme function. (A) Representative PAGE (denaturing 10%) separation of RNAs resulting from the incubation of trace amounts of 5′-³²P-labeled class II ribozymes in the presence of 100 μM Be²⁺, La²⁺ or Cu²⁺ under standard reaction conditions at 23°C for 5 min. Reactions were conducted in the absence (−) or presence (+) of 100 μM CoCl₂. Open and filled arrowheads identify bands corresponding to 5′-³²P-labeled full-length precursor (Pre) or the 5′-cleavage product (Clv), respectively. Bracket identifies bands corresponding to uncleaved precursor RNAs whose gel mobility is altered, presumably due to strong interactions between RNA and the cation being tested. The results for Be²⁺, La²⁺ or Cu²⁺ are typical of that for cations that are disruptive, inhibitory and non-activating, respectively. (B) Summary of the various metal cations tested and their effects on class II ribozyme activity in the absence and presence of 100 μM CoCl₂. Results from metal assays with Class II ribozyme are shown. Unless otherwise indicated, the metal form tested is the 2⁺ ionic species.

Figure 8

Some physicochemical characteristics of the metal ions used in this study. Metal ions are classified according to (A) ionic radius (B) pK_a of metal-bound water (C) hydration enthalpy and (D) ‘hard’ versus ‘soft’ character. Red and black circles reflect cations that are active or inactive, respectively, as allosteric effectors for the class II ribozyme. Plots of physical parameters were generated from parameters published elsewhere (35).