Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (155K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (45)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Tang, J.
Right arrow Articles by Breaker, R. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, J.
Right arrow Articles by Breaker, R. R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Nucleic Acids Research Pages 4214-4221  

Mechanism for allosteric inhibition of an ATP-sensitive ribozyme
Introduction
Materials And Methods
   RNA and DNA molecules
   Ribozyme assays
   Binding assays
   Molecular modeling
Results And Discussion
   A minimal allosteric ribozyme construct
   A structural model for an allosteric ribozyme
   Aptamer structural conformation and interdomain connectivity
   Allosteric function and the length of stem IV
   Allosteric function and the length of stem II
   The ATP effector binds the enzyme-substrate complex
   Allosteric interactions in proteins and nucleic acids
Conclusions
Acknowledgements
References

Mechanism for allosteric inhibition of an ATP-sensitive ribozyme

Mechanism for allosteric inhibition of an ATP-sensitive ribozyme

Jin Tang and Ronald R. Breaker*

Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, CT 06520-8103, USA

Received May 22, 1998; Revised and Accepted July 13, 1998

ABSTRACT

We report the structural basis for the modulation of an ATP-sensitive ribozyme that was engineered by modular rational design. This allosteric ribozyme is composed of two independently functioning domains, one a receptor for ATP and the other a self-cleaving ribozyme. When fused in the appropriate fashion, the conjoined aptamer-ribozyme construct functions as an allosteric ribozyme that is inhibited in the presence of ATP. The aptamer domain remains conformationally heterogeneous in the absence of ATP, but folds into a distinct structure upon ligand binding. This ATP-induced conformational change causes a reduction in catalytic activity of the adjacent ribozyme domain due to steric interference between the aptamer and ribozyme tertiary structures. This mechanism for structural and functional modulation of nucleic acids is one of several possible mechanisms by which the function of ribozymes could be specifically controlled by small effector molecules.

INTRODUCTION

Many examples of allosteric behavior with protein enzymes have been reported in the years that followed the initial descriptions of this regulatory process by Monod and co-workers (1,2). Allosteric enzymes can exist as single subunit or multisubunit proteins that respond either positively or negatively to the presence of allosteric effector molecules. These effector molecules can be substrates, reaction products or metabolites that are entirely unrelated to the enzymatic function of the protein (3,4). Allosteric regulation of enzyme function operates by mechanisms that are fundamentally different from that of inhibitors that block the active sites of enzymes. A true allosteric effector binds to a site located apart from the active site and its influence on the activity of the enzyme is brought about by changes in protein conformation.

It is known that the catalytic activity of ribozymes can also be modulated by certain antibiotic compounds (5-8) and nucleobases (9) by directly binding to the catalytic domain of the ribozyme. Moreover, oligonucleotide binding and alternative base pairing interactions can contribute to the control of ribozymes, such as the self-cleaving hammerhead RNA (10,11). These studies with catalytic RNA provide support for the notion that the function of ribozymes might also be brought under the control of various small organic effector molecules through allosteric interactions. Recently, we used a modular rational design strategy to create an artificial ribozyme that is specifically deactivated in the presence of ATP (12). This ATP-sensitive ribozyme was created by fusing a pre-existing ATP-binding aptamer (13) to a hammerhead self-cleaving ribozyme (14). Like allosteric protein enzymes, this modified ribozyme binds an effector molecule at a site that is distal to the active site of the enzyme. Molecular modeling of the conjoined aptamer-ribozyme construct indicates that, when the two structural domains are fused in the appropriate fashion, ATP induces a conformational change in the aptamer domain that precludes formation of the active ribozyme structure. In this study, we provide a more detailed examination of the structural basis for allosteric modulation of this ATP-sensitive allosteric ribozyme.

MATERIALS AND METHODS

RNA and DNA molecules

DNA molecules and the 14 nt RNA substrate were synthesized (solid phase) by the Keck Biotechnology Resource Laboratory, Yale University, and were prepared as described previously (12). The 18 nt substrate RNA was synthesized and prepared by Ribozyme Pharmaceuticals Inc. Each conjoined aptamer-ribozyme construct was prepared by in vitro transcription (12) from double-stranded DNA templates that were produced by extending the DNA primer 5[prime]-d(TAATACGACTCACTATAG) on the corresponding antisense DNA template using reverse transcriptase (RT). RT extension reactions (64 µl) contained 240 pmol primer, 200 pmol template, 0.2 mM each dNTP and 10 U/µl SuperScript RT (BRL). Extension reactions were incubated at 37°C for 30 min in the buffer supplied by the manufacturer. When needed, RNA constructs were radiolabeled by introducing [[alpha]-32P]UTP into the in vitro transcription reaction. Synthetic RNA substrates were end-labeled with [[gamma]-32P]ATP and T4 polynucleotide kinase.

Ribozyme assays

Assays were conducted under single turnover conditions with ribozyme (400 nM) in excess over trace amounts (~5 nM) of 5[prime]-32P-labeled substrate. This concentration of ribozyme is above the Kd for the enzyme-substrate complex and was used to ensure that substrate is saturated with enzyme. Ribozyme, substrate and effector (if used) were pre-incubated in the presence of 50 mM Tris-HCl at 23°C for 15 min prior to initiation of each reaction by addition of MgCl2 (20 mM final concentration). Reaction products were separated by denaturing (8 M urea) 20% PAGE and visualized and quantitated using a PhosphorImager and ImageQuant software (Molecular Dynamics). Product yields were corrected for the amount of substrate that remained uncleaved after exhaustive incubation with an unmodified hammerhead ribozyme (12). Rate constants were derived by plotting the natural log of the fraction of substrate remaining versus time and establishing the negative slope of the resulting line. The values for each rate constant or activity ratio reported are an average of a minimum of two replicate assays, with each replicate differing by <50%.

Binding assays

RNA binding assays were conducted with agarose beads (4%) that were derivatized with either 5[prime]-AMP or 2[prime]-AMP via an 8 atom linker attached to the C8 carbon of adenine (Sigma). Beads (30 µl 30 mg/ml aqueous suspension, 1-5 nmol ligand/µl) were washed with a binding buffer composed of 50 mM Tris-HCl (pH 7.5 at 23°C) and 20 mM MgCl2 by repeatedly resuspending beads in binding buffer, centrifuging the mixture and removing the supernatant. Internally 32P-labeled aptamer-ribozyme fragments (1-10 pmol) were individually added in the absence of substrate to the binding buffer/agarose bead suspension and allowed to incubate at 23°C for 30 min. Fraction bound values for each construct were derived from the measurements of the specific activities of the agarose bead suspension and of the supernatant after pelleting the beads by centrifugation. Radioactivity was quantified by liquid scintillation counting.

Examination of the simultaneous binding of ribozyme and 14 nt substrate RNAs utilized 5[prime]-AMP-agarose resin that was prepared in an identical manner. Internally 32P-labeled aptamer-ribozyme construct (6 pmol) was bound to the matrix, washed with 3 × 30 µl binding buffer and resuspended in buffer containing 2 pmol unlabeled substrate RNA. Initial binding and subsequent retention of the labeled construct were assessed by determining the specific activities of the suspension and the supernatant upon centrifugation. Alternatively, unlabeled construct (6 pmol) was bound to the matrix, washed and resuspended in binding buffer containing 2 pmol 5[prime]-32P-labeled substrate. Binding of substrate RNA was also assessed by determining the specific activities of the suspension and the supernatant upon centrifugation. Elution of the constructs was achieved by addition of ATP at 1 mM final concentration.

Molecular modeling

Molecular models were produced on a Silicon Graphics workstation using MidasPlus (15) (Computer Graphics Laboratory, University of California, San Francisco). The structure of the hammerhead ribozyme (16) was modified such that stem I includes an additional 3 bp. For this purpose, the least squares fit method was used to superimpose a generic A-form RNA duplex (6 bp) upon stem I by overlapping the four phosphates intervening the terminal 3 bp of each element. One terminal base pair of the ribozyme and two terminal base pairs of the RNA duplex were deleted to generate a model with an extended A-form stem I. Using this same technique, the structure for the ATP aptamer that was derived by NMR (17) was superimposed on stem II of the modified hammerhead structure to generate the model for the construct IV-up. A model for the construct IV-down was generated by exchanging the connectivity of the two stems of the ATP aptamer.

RESULTS AND DISCUSSION

A minimal allosteric ribozyme construct

The secondary structure of the hammerhead ribozyme consists of a three stem junction (14) that for this study has been divided into separate substrate and ribozyme strands (Fig. 1a). Stems I and III are formed from the ribozyme-substrate complex, while stem II is entirely contained within the catalytic strand. In its original manifestation (12), the ATP-sensitive allosteric ribozyme was composed of a hammerhead ribozyme domain, the ATP-binding aptamer domain that was appended to stem II of the ribozyme and two hairpin domains that were used in a related study (18) to facilitate amplification by RT-PCR. This 80 nt RNA and several variant constructs display ATP-specific inhibition of the rate of RNA cleavage that equals or in some examples exceeds 100-fold (12,18).

Figure 1. Secondary structure and tertiary structure models for an ATP-sensitive allosteric ribozyme. (a) Sequence and secondary structure of the bimolecular ribozyme construct IV-up. Roman numerals designate the three stems of the hammerhead ribozyme domain (I-III) and one of the aptamer stems (IV). Stem II is shared by both structural domains. Bracket designates the base pairs added to some RNA constructs to extend the length of stem I. The arrowhead identifies the site of substrate cleavage and the enclosed nucleotides comprise the ATP-binding aptamer domain. Filled circles represent G-U wobble pairs. (b) Alternate depiction of the secondary structure of the engineered ribozyme that more accurately depicts the relative three-dimensional orientations of the hammerhead stem structures. The tertiary contacts between bases that form a pseudo A-form helix structure are denoted by open circles. The base pairs that comprise stems I and II nearest the catalytic core form a plane as depicted. (c) Ribbon model of the three-dimensional structure of the conjoined aptamer-ribozyme construct. Dark blue identifies the tertiary structure of the hammerhead ribozyme as reported by Scott et al. (12). Red identifies the tertiary structure of the ATP-binding aptamer as reported by Jiang et al. (12). Light blue depicts an A-form helix added to the hammerhead model to produce the extended stem I structure used in the constructs for this study.    a,b

   c

For the current study we synthesized a similar construct, termed IV-up, wherein the extraneous hairpin domains were eliminated (Fig. 1a). This minimized 59 nt construct retains sensitivity to the effector and operates with a kobs of 0.20 or 0.011/min in the absence or presence of 1 mM ATP, respectively (Fig. 2). Although the presence of the hairpin domains in the original construct contributed 5-10-fold towards ATP-induced inactivation of the engineered ribozyme, the results obtained with the minimized construct IV-up confirm that the structures responsible for the allosteric response are largely confined to the aptamer and ribozyme domains. The inhibition constant (Ki) for ATP with the IV-up construct is ~2 µM (data not shown), a value that is similar to the Ki derived from data reported previously for a larger version of this same engineered ribozyme (12). These Ki values closely approximate the Kd for ATP and the independent aptamer domain (13), suggesting that no significant reciprocal effect of hammerhead structure on ATP affinity is present. This is not surprising if the stability of the properly folded ATP aptamer far exceeds the stability of the active hammerhead configuration. In this case, the corresponding effect of the hammerhead structure on ATP binding affinity of the adjacent aptamer would be negligible.


Figure 2. Inhibition of the engineered ribozyme IV-up (400 nM) by ATP (1 mM). Reaction conditions were as described in Materials and Methods.

A structural model for an allosteric ribozyme

A reorganized secondary structure model for the conjoined aptamer-hammerhead construct is depicted in Figure 1b. This model incorporates information obtained from the X-ray (16,19-21) and FRET (22) analyses of the hammerhead tertiary structure. The three helices of the hammerhead ribozyme adopt a Y-shaped conformation. The two Watson-Crick base pairing elements that form stems I and II originate in the same horizontal plane relative to the catalytic core (Fig. 1b) and their helices propagate roughly parallel to each other in the active ribozyme conformation (22-24). For our original design efforts (12), we took advantage of the fact that stem II stability (25,26) and the precise positioning of stems I and II were critical for maximal catalytic function.

The atomic resolution models of the hammerhead (16,19-21) and the ATP-binding aptamer (17,27) are now available. When the structures of these two independently folding domains are joined to represent the allosteric ribozyme IV-up (Fig. 1c), a mechanism becomes apparent for ATP-induced inhibition of ribozyme function that involves the mutually exclusive formation of aptamer and ribozyme domains. A schematic representation (Fig. 3) of the tertiary structure of the ATP aptamer bound with ligand and fused to a portion of the hammerhead was produced using structural information derived from the three-dimensional models for both RNA components and using parameters for a typical A-form RNA helix (28,29). Previously (12), we proposed that ATP-induced allosteric inhibition results from disruption of the active structure of the hammerhead ribozyme upon formation of the ATP-aptamer complex in the adjoined ligand-binding domain. Both models (Figs 1c and 3) indicate that, upon ATP binding, stem IV of the aptamer domain displaces stem I from its location in the active conformation of the hammerhead domain, thereby inhibiting ribozyme function.


Figure 3. Schematic representation of the mechanism of allosteric regulation of a ribozyme. (a) Side view. Relative orientations of the stems (cylinders; I-IV) and ATP-binding domain of the conjoined aptamer-ribozyme complex IV-up. The dimensions (Å) for various structural elements and features, and the angle formed by the helix axis of stems II and IV are indicated. ATP designates the location of ligand binding. (b) Top view. Projected orientations of stem IV relative to stem I for constructs that include 2-7 bp in stem II. Construct IV-up includes 4 bp in stem II and is designated as 4 in this graphic. Note that for each base pair addition or subtraction in stem II, an estimated 32.7° rotation of stem IV relative to stem I occurs.

Table 1. RNA binding to a 5[prime]-AMP-derivatized agarose matrix
a    Construct Fraction bound
  ATP-40-1* 0.64
IV-up 0.50
IV-down 0.53
b Stemlength Fraction bound
II IV
  2 0.08 -
3 0.23 0.17
4 0.28 0.49
5 0.64 0.34
6 0.65 0.64
7 0.82 0.64
For stem II variants, ligand binding was determined with series 2 constructs. (a) comparison of binding efficiencies of constructs IV-up, IV-down and the ATP-binding aptamer ATP-40-1*; (b) assessment of the ATP-binding potential of different series of aptamer-ribozyme constructs.

Aptamer structural conformation and interdomain connectivity

Evidence for a significant change in the conformational state of the ATP-binding aptamer is provided by both chemical probing studies (13) and by NMR analysis (17,27). It is known that the tertiary structure of the aptamer domain remains heterogeneous in the absence of ATP, but conforms to a single highly ordered tertiary structure upon ligand binding. The two stems (II and IV) of the occupied aptamer form an ~109° angle relative to each other, an arrangement that is probably only rarely sampled by the aptamer in the absence of ATP. According to our structural models, IV-up has the appropriate interdomain arrangement to produce a steric clash between stems I and IV upon ATP-induced stem re-alignment. This steric overlap between stems I and IV is ~20% when ATP is bound.

The model for the allosteric function of IV-up predicts that a reorganization of the interdomain connectivity will result in an altered response to ATP. To test this, we synthesized a new construct that retains the necessary sequence composition for the independent function of aptamer and ribozyme domains, but which carries these domains in an altered orientation relative to IV-up (Fig. 4a). This reorganized construct, termed IV-down, is modeled to form an active aptamer complex in which stem IV projects away from stem I (Fig. 4b), thereby allowing the two functional domains to operate without interference. As predicted, IV-down remains fully active, both in the absence and presence of ATP, suggesting that proper orientation of the domains is necessary for allosteric function.

Figure 4. Secondary (a) and tertiary structure (b) models for the ATP-insensitive construct IV-down. Details are as described in the legend for Figure 1.    a

   b

ATP binding by these and other constructs was assessed using an agarose matrix derivatized with 5[prime]-AMP. Matrix binding with the minimal aptamer ATP-40-1 (13) was compared with that for all constructs used in this study. We find that ~64% of a preparation of an ATP-40-1 variant (ATP-40-1*) binds to the agarose matrix that is derivatized with 5[prime] AMP (Table 1a). The asterisk identifies this RNA as a variant of the original ATP-binding aptamer described previously (13). This variant aptamer, which shows improved affinity for ATP, differs from the original aptamer by the replacement of a U-A base pair with a C-G base pair at the end of stem IV nearest to the ATP-binding core. Likewise, both IV-up and IV-down retain the ability to bind ATP, indicating that the loss of allosteric control with IV-down is not due to a loss of effector molecule binding. Without exception, incubation of all RNA samples with an agarose matrix similarly derivatized with the isomeric compound 2[prime]-AMP gave <2% binding, in accordance with the reported ligand specificity of this aptamer.

Allosteric function and the length of stem IV

We have used the structural models depicted in Figures 1c and 3 to predict the catalytic performance of several new ribozyme designs. For example, the length of stem IV should be a critical determinant of the ATP sensitivity of our allosteric constructs. To an extent, longer stem IV elements should produce greater interference between stems IV and I, resulting in more pronounced ribozyme inhibition. Using a variant of IV-up as a starting point, a series of constructs that carry stem IV elements of different lengths was synthesized and tested for ATP-dependent inhibition (Fig. 5a). Catalytic activity is unresponsive to ATP when stem IV is reduced to 3 nt, while progressively greater ATP sensitivity is seen with increasing stem lengths up to 7 bp (Fig. 5b). Each construct demonstrates near equal kobs values in the absence of ATP (0.3/min), while maximum ATP-dependent inhibition was observed with a stem IV element composed of 7 bp (kobs 0.021/min).

Although these results are consistent with the proposed mechanism for allosteric function, it is not clear whether the observed variation in ATP sensitivity is due to increased steric interactions between the two domains or whether the effects of stem length are due to possible variability in affinity of the constructs for ATP. Most constructs that carry modified stem IV elements show substantial levels of binding to the 5[prime]-AMP-agarose matrix that are comparable with the minimal ATP aptamer (Table 1b). However, the fraction of RNA that binds to the 5[prime]-AMP-derivatized matrix is significantly lower for several constructs that carry the shortest stem IV structures. These results highlight the possibility that certain constructs bind the 5[prime]-AMP-modified matrix with a lower affinity or that the RNAs may adopt structures that preclude binding to the derivatized matrix. We can conclude that the length of stem IV is critical for allosteric performance. However, it cannot be concluded from these data alone whether stem IV participates in the proposed mechanism of allosteric function or whether ATP binding is indeed affected by stem length and that another mechanism for ATP-dependent modulation might be utilized.

Allosteric function and the length of stem II

An alternative approach to the examination of the allosteric mechanism makes use of modifications to stem II, the shared stem that bridges the aptamer and ribozyme domains. We have observed that the addition of 3 bp to stem II completely eliminates ATP sensitivity in a related allosteric ribozyme (12). It is apparent from the structural models that both the addition or removal of base pairs should eliminate allosteric function of the construct if the proposed mechanism for allosteric function is correct. If stem IV is disrupting the formation of the active conformation of the hammerhead ribozyme as shown in Figure 3a, then increasing or decreasing the number of base pairs that form stem II will change the orientation of stem IV relative to stem I in two ways. First, adding or deleting stem II base pairs will raise or lower the ATP-binding domain, respectively, relative to the catalytic domain. Second, and perhaps most importantly, changing the number of base pairs will rotate stem IV around the screw axis formed by stem II (Fig. 3b).


Figure 5. The effect of stem IV length on allosteric function. (a) Sequences of different stem IV domains that were appended to a IV-up ribozyme variant. These ribozyme variants included the stem II structure 5[prime]-GGCUG-(aptamer)-CAGCC-3[prime] (Fig. 6a, series 2, structure 5) in place of the original stem II element present in IV-up (Fig. 1a). (b) Plot of stem IV length versus ATP-induced inhibition of RNA cleavage activity. Fold inhibition values were derived from the ratio of cleavage rate constants in the absence versus the presence of 1 mM ATP. The values for each activity ratio reported are an average of a minimum of two replicate assays, with each replicate differing by <50%. The dashed line indicates the bar height (ratio = 1) expected in the absence of allosteric inhibition.

To investigate the effect of stem II length and sequence on allosteric function, we synthesized two series of constructs in which the aptamer and ribozyme domains are joined by different stem II elements (Fig. 6a). As described earlier, construct IV-up displays an ~17-fold rate reduction upon addition of ATP (Figs 2 and 6b, series 1). According to the average rotation angle ([phiv]) for an A-form RNA helix, each base pair rotates ~33° relative to the preceding base pair (28,29). Therefore, each base pair addition or deletion will rotate stem IV ~33° counter clockwise or clockwise respectively around the screw axis of stem II (Fig. 3b). This repositioning of stem IV is expected to reduce or eliminate the steric clash between the competing RNA domains. As predicted, 1 bp insertions or deletions in stem II reduce ATP sensitivity, while larger changes to its length completely eliminate allosteric function.


Figure 6. The effect of stem II length and sequence on allosteric function. (a) Sequences of different stem II domains used to join the ATP aptamer and hammerhead ribozyme. Stem II elements are organized into two series that survey identical lengths, but which differ in sequence composition. Construct IV-up is represented in series 1 with a stem II length of 4 bp. (b) Plots of stem II length versus ATP-induced inhibition of ribozyme function for series 1 and series 2 elements. The dashed line indicates the bar height expected in the absence of allosteric inhibition.

The same pattern of allosteric modulation observed with series 1 RNA constructs is also observed when stem II elements of different sequence are used (Fig. 6b, series 2). This finding confirms that the number of base pairs present in stem II is a critical determinant of allosteric function. However, it is also apparent that the specific sequence of the stem contributes to the precise level of allosteric performance. It is well established that individual base pairs contribute differently to the helical twist of base-paired structures (29). Therefore, both the insertion or deletion of base pairs and the alteration of base pair identities are expected to produce changes in orientation of stem IV around the screw axis of stem II.

The initial base pairs of stems I and II nearest the catalytic core lie on the same plane (Fig. 1b). As a result, the difference in the distance between the distal ends of these stems will lie between 0 and 2.8n Å, where n is the difference in the number of base pairs between stems I and II. With IV-up, for example, the terminal base pairs of stem I that reside furthest from the catalytic core are expected to lie immediately adjacent to the core of the aptamer domain (Fig. 1b and c, Fig. 3). This may not be true, however, for the constructs that include 6 or 7 nt in stem II (Fig. 6a), where the aptamer domain may lie beyond the terminal nucleotides of stem I. With these constructs, the loss of ATP sensitivity could conceivably be due to the increase in the length of stem II, rather than rotational repositioning of stem IV.

To separate the effects of length versus rotational changes, we synthesized and assayed a construct that includes a stem I structure composed of 12 bp, as opposed to the original eight (Fig. 1a). The use of this revised stem I domain ensures that the appended aptamer domain will not extend beyond the terminus of stem I. In this new format, we find that an aptamer-ribozyme construct that carries 6 bp in stem II also displays no inhibition with ATP (data not shown), consistent with the hypothesis that the loss of ATP-dependent inhibition is due to rotational rather than translational repositioning of stem IV. The results obtained from examination of constructs carrying modified stem II elements further support the proposed mechanism for allosteric function, wherein precise positioning of stem IV is critical for allosteric performance.

With one exception, all constructs display catalytic rate constants in the absence of ATP that range between 0.2 and 0.6/min. The construct that carries the shortest stem II element (2 nt) suffers a 10-fold reduction in rate (kobs 0.020/min), likely due to the effect of this weakened stem on the catalytic structure of the adjoining ribozyme (data not shown). When assayed for ATP-binding ability, this same construct demonstrates poor binding to the 5[prime]-AMP-agarose matrix (Table 1b), raising doubt as to the ability of this construct to bind ATP in solution. Although this particular construct may not respond to ATP due to the disruption of the aptamer domain, all other constructs show some affinity for the immobilized ligand. Therefore, loss of ATP binding potential alone cannot explain the observed loss of allosteric control with the remaining variants. Interestingly, although a stem II length of four is optimal for allosteric inhibition of IV-up, we find that a related construct binds poorly to the agarose matrix, indicating that the column binding assay may not be a perfect reflection of effector binding in the solution phase.

Using a similar approach, we examined the ATP sensitivity of IV-down variants that also carry stem II modifications. We speculated that the aptamer orientation of IV-down (Fig. 4) might be repositioned to re-establish ATP-dependent inhibition simply by introducing additional base pairs into stem II in order to rotate stem IV around the screw axis of the bridging stem. To investigate this possibility, a series of stem II variants based on the IV-down construct were synthesized and assayed in the presence and absence of ATP. The parent IV-down construct was modified to include 12 bp in stem I and 8 bp in stem IV. The extensions added to stems I and IV were intended to offset the effects of the inherent rise in the aptamer domain relative to the ribozyme core (stem I) and to project the interfering helix farther towards the ribozyme domain (stem IV). The latter modification is expected to be particularly important because as steric interference occurs farther away from the catalytic core of the ribozyme, greater displacement of stem I is necessary to create a comparable disruption of the catalytic core and its function.

Modified IV-down constructs that carry stem II elements composed of 8, 9 or 10 bp were surveyed for evidence of ATP-dependent inhibition. Based on the average [phiv] value of 33° for an RNA helix, these stem II lengths are expected to cause the repositioning of stem IV back towards stem I. Unfortunately, we observed no ATP sensitivity with these new constructs (data not shown). It is possible that subtle variations in RNA structural conformation, of particular concern in the newly enlarged RNA elements of IV-down variants, cause significant errors to be made in predicting the precise positioning of important structural elements. Therefore, the lack of responsiveness to ATP with these constructs may be due to the inherent limitations of the modular rational design process.

The ATP effector binds the enzyme-substrate complex

Inhibition of this class of allosteric ribozymes by ATP requires a functional aptamer domain. This was concluded from the observation that ATP-dependent inhibition is observed only at concentrations above the Kd for the aptamer-ligand complex and that single mutations within the aptamer that eliminate ATP binding also eliminate the allosteric response (12). According to the proposed mechanism, the ATP-aptamer complex interferes with formation of the active enzyme-substrate (E-S) complex. ATP binding is not expected to interfere with the early steps of the ribozyme kinetic pathway, including the binding of substrate to form a pre-catalytic E-S complex.

We examined the nature of the inhibited allosteric complex using both unlabeled and radiolabeled portions of a variant of the IV-up construct. The conjoined aptamer-ribozyme portion of this construct binds and is retained by a matrix derivatized with 5[prime]-AMP, both in the presence and absence of substrate RNA. This demonstrates that formation of the aptamer-ligand complex is not disrupted by substrate. Moreover, substrate molecules do not bind 5[prime]-AMP-agarose unless the matrix has been pretreated with the aptamer-ribozyme fragment. In each case, the addition of free ATP to the matrix induces the release of the aptamer-ribozyme and substrate molecules (data not shown). These results indicate that effector binding does not preclude formation of the E-S complex and are consistent with a mechanism for allosteric function that involves formation of an inactive ATP . E-S complex where ATP is properly docked into the aptamer domain.

Allosteric interactions in proteins and nucleic acids

Many allosteric proteins are composed of subunits and the allosteric effects are mediated through assembly or rearrangement of the protein complex. The original models developed to describe allosteric interactions focus on this general mechanism of enzyme control (30,31). However, it is clear that allosteric behavior can be mediated through effector-triggered restructuring of a single polypeptide or nucleic acid chain. For example, allosteric regulation of DNA binding by the trp repressor protein involves tertiary structure changes, triggered by tryptophan binding, that do not result in changes in the quaternary structure of this dimeric protein complex (4,32). Likewise, the class of ATP-sensitive allosteric ribozymes described in this report also appear to undergo tertiary structure rearrangement upon allosteric modulation.

The tertiary structure change to the trp repressor protein complex brought about by effector binding induces a 200-fold increase in DNA-binding affinity (33). This allosteric effect is similar in magnitude to the 180-fold decrease in the catalytic activity of our most sensitive allosteric ribozyme (12). It has been suggested that the widespread use of quaternary changes for allosteric protein regulation is due to the ease of bringing about structural changes between separate peptides as opposed to inducing structural changes within the domains of a single polypeptide chain (4). In each case, the key requirement for efficient operation of these molecular switches is that the thermodynamic and kinetic parameters must be favorable for switch activation and for maximization of the `on' and `off' states of the function of interest. For allosteric inhibition, the active ribozyme structure needs to be favored in the absence of effector, while the active structure must be precluded by a more favorable alternative effector-RNA complex. The speed of interconversion between states and the extent to which function is modulated will be dependent upon various and sometimes subtle kinetic and thermodynamic parameters. Perhaps, a greater on/off ratio can be achieved more easily by inducing changes in the quaternary structures of proteins, which may explain why this general mechanism occurs frequently among natural allosteric enzymes. Similar mechanisms for quaternary structure reconfiguration for allosteric nucleic acids can also be envisioned.

CONCLUSIONS

RNAs whose function can be triggered by the binding of proteins are already known to exist in nature, where functional modulation likely serves important roles in cellular operations. In this report, we describe a unique mechanism whereby an aptamer-ribozyme hybrid functions with true allosteric behavior as originally defined by Monod and co-workers (1,2). Molecular modeling using the three-dimensional structures of the ATP-bound aptamer domain joined with the ribozyme domain reveals a spatial overlap between two hairpin structural elements, suggesting that steric interference between the two domains may inhibit ribozyme function. Biochemical data are consistent with this proposed mechanism, which involves the mutually exclusive formation of aptamer and ribozyme domains. This `helix bumping' mechanism is only one of a number of different molecular mechanisms that could act to sense and respond to small effector molecules.

The studies reported herein represent our early attempts to introduce novel kinetic features into existing functional polynucleotides and to further develop a basis for engineering additional allosteric ribozymes. Similar molecular modeling exercises could be used to provide a conceptual framework for constructing new allosteric ribozymes with various effector specificities and catalytic activities, assuming that sufficient structural data for the component modules already exists. Ribozyme engineering techniques could be employed in a general fashion to add complexity to the existing set of ribozyme functions. By generating additional examples of allosteric ribozymes, we expect to identify and establish new mechanisms by which separate RNA domains can interact in a dynamic fashion. Already, we have demonstrated that similar aptamer-hammerhead constructs can be engineered to undergo ATP-dependent allosteric enhancement, presumably by a mechanism that involves stabilization of an RNA helix (12). In addition, we have used this same mechanism to create a hammerhead ribozyme that undergoes allosteric activation by several organic compounds, including flavine mononucleotide and theophylline (G. A. Soukup and R. R. Breaker, unpublished data). When fully developed, ligand-induced stabilization of localized structures may have greater utility for allosteric ribozyme engineering than is offered by the helix bumping mechanism described in this report.

Nucleic acids have considerable potential for catalytic activity and for receptor-ligand formation. It is now apparent that these two functions can be coupled to create new allosteric ribozymes or even deoxyribozymes that are regulated by small effector molecules. With these design strategies, it is possible to assess whether enzymes made of RNA and DNA could participate in more complicated catalytic networks, in which allosteric effects might provide tight regulation of ribozyme kinetic parameters. Recent growth in the understanding and manipulation of nucleic acids (34-36) will undoubtedly foster additional and more sophisticated attempts to create catalytic polynucleotides with desirable kinetic features, perhaps for use as controllable therapeutic enzymes or for industrial applications. We expect that both modular rational design and in vitro selection (37-39) will be generally applicable for the construction of novel allosteric ribozymes and deoxyribozymes (40,41). Further developments in our understanding of aptamer and ribozyme function promise to make routine the purposeful introduction of effector-induced catalytic control into nucleic acid catalysts for the construction of novel allosteric ribozymes.

ACKNOWLEDGEMENTS

We thank G. A. Soukup for generating the molecular models and he and other members of the Breaker laboratory for helpful discussions. We also thank F. M. Richards and members of the Yale Center for Structural Biology for their assistance. This work was supported by a Young Investigator Award to R.R.B. from the Arnold and Mabel Beckman Foundation.

REFERENCES

1. Monod,J. and Jacob,F. (1961) Cold Spring Harbor Symp. Quant. Biol., 26, 389-401.

2. Monod,J., Changeux,J.-P. and Jacob,F. (1963) J. Mol. Biol., 6, 306-329.

3. Kurganov,B.I. (1978) Allosteric Enzymes. John Wiley & Sons Ltd, New York, NY.

4. Perutz,M. (1994) Mechanisms of Cooperativity and Allosteric Regulation in Proteins. Cambridge University Press, New York, NY.

5. von Ahsen,U., Davies,J. and Schroeder,R. (1991) Nature, 353, 368-370. MEDLINE Abstract

6. Wank,H., Rogers,J., Davies,J. and Schroder,R. (1994) J. Mol. Biol., 236, 1001-1010. MEDLINE Abstract

7. Clouet-d'Orval,B., Stage,T.K. and Uhlenbeck,O.C. (1995) Biochemistry, 34, 11186-11190.

8. Olive,J.E., De Abreu,D.M., Rastogi,T., Anderson,A.A., Mittermaier,A.K., Beattie,T.L. and Collins,R.A. (1995). EMBO J., 14, 3247-3251. MEDLINE Abstract

9. Peracchi,A., Beigelman,L., Usman,N. and Herschlag,D. (1996) Proc. Natl Acad. Sci. USA, 93, 11522-11527. MEDLINE Abstract

10. Porta,H. and Lizardi,P.M. (1995) Biotechnology, 13, 161-164.

11. Miller,W.A. and Silver,S.L. (1991) Nucleic Acids Res., 19, 5313-5320. MEDLINE Abstract

12. Tang,J. and Breaker,R.R. (1997) Chem. Biol., 4, 453-459. MEDLINE Abstract

13. Sassanfar,M. and Szostak, J.W. (1993) Nature, 364, 550-553. MEDLINE Abstract

14. Forster,A.C. and Symons,R.H. (1987) Cell, 49, 211-220. MEDLINE Abstract

15. Ferrin,T.E., Huang,C.C., Jarvis,L.E. and Langridge,R. (1988) J. Mol. Graphics, 6, 13-27.19. Scott,W.G., Finch,J.T. and Klug,A. (1995) Cell, 81, 991-1002.

17. Jiang,F., Kumar,R.A., Jones,R.A. and Patel,D.J. (1996) Nature, 382, 183-186. MEDLINE Abstract

18. Tang,J. and Breaker,R.R. (1997) RNA, 3, 914-925. MEDLINE Abstract

16. Pley,H.W., Flaherty,K.M. and McKay,D.B. (1994) Nature, 372, 68-74. MEDLINE Abstract

20. Scott,W.G., Murray,J.B., Arnold,J.R.P., Stoddard,B.L. and Klug,A. (1997) Science, 274, 2065-2069.

21. Feig,A.L., Scott,W.G., and Uhlenbeck,O.C. (1998) Science, 279, 81-84. MEDLINE Abstract

22. Tuschl,T., Gohlke,C., Jovin,T.M., Westhof,E. and Eckstein,F. (1994) Science, 266, 785-789. MEDLINE Abstract

23. Sigurdsson,S.T., Tuschl,T. and Eckstein,F. (1995) RNA, 1, 575-583. MEDLINE Abstract

24. Bassi,G.S., Mollegaard,N.-E., Murchie,A.I.H., von Kitzing,E. and Lilley,D.M.J. (1995) Nat. Struct. Biol., 2, 45-55. MEDLINE Abstract

25. Tuschl,T. and Eckstein,F. (1993) Proc. Natl Acad. Sci. USA, 90, 6991-6994. MEDLINE Abstract

26. Long,D.M. and Uhlenbeck,O.C. (1994) Proc. Natl Acad. Sci. USA, 91, 6977-6981. MEDLINE Abstract

27. Dieckmann,T., Suzuki,E., Nakamura,G.K. and Feigon,J. (1996) RNA, 2, 628-640. MEDLINE Abstract

28. Arnott,S., Hukins,D.W.L. and Dover,S.D. (1972) Biochem. Biophys. Res. Commun., 48, 1392-1399. MEDLINE Abstract

29. Saenger,W. (1984) Principles of Nucleic Acid Structure. Springer-Verlag, New York, NY, pp. 220-241.

30. Monod,J., Wyman,J. and Changeux,J.P. (1965) J. Mol. Biol., 208, 307-325.

31. Koshland,D.E., Nemethy,G. and Filmer,D. (1966) Biochemistry, 5, 365-385. MEDLINE Abstract

32. Marmorstein,R.Q. and Sigler,P.B. (1989) Nucleic Acids Mol. Biol., 3, 56-78.

33. Carey, J. (1988) Proc. Natl Acad. Sci. USA, 85, 975-979. MEDLINE Abstract

34. Pyle,A.M. and Green,J.B. (1995) Curr. Biol., 5, 303-310.

35. Draper,D.E. (1996) Trends Biochem. Sci., 21, 145-149. MEDLINE Abstract

36. Westhof,E., Masquida,B. and Jaeger,L. (1996) Fold. Des., 1, 78-88.

37. Williams,K.P. and Bartel,D.P. (1996) Nucleic Acids Mol. Biol., 10, 367-381.

38. Osborne,S.E. and Ellington,A.D. (1997) Chem. Rev., 97, 349-370.

39. Breaker,R.R. (1997) Chem. Rev., 97, 371-390.

40. Breaker,R.R. (1997) Nat. Biotechnol., 15, 427-431. MEDLINE Abstract

41. Breaker,R.R. (1997) Curr. Opin. Chem. Biol., 1, 26-31. MEDLINE Abstract

*To whom correspondence should be addressed. Tel: +1 203 432 9389; Fax: +1 203 432 5713; Email: ronald.breaker@yale.edu

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 29 Aug 1998
Copyright©Oxford University Press, 1998.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nucleic Acids ResHome page
J. J. Tabor, M. Levy, and A. D. Ellington
Deoxyribozymes that recode sequence information.
Nucleic Acids Res., January 1, 2006; 34(8): 2166 - 2172.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
S. Cho, J.-E. Kim, B.-R. Lee, J.-H. Kim, and B.-G. Kim
Bis-aptazyme sensors for hepatitis C virus replicase and helicase without blank signal
Nucleic Acids Res., November 27, 2005; 33(20): e177 - e177.
[Abstract] [Full Text] [PDF]

Home page
 J. Gerontol. A Biol. Sci. Med. Sci.Home page
C. C. Deocaris, S. C. Kaul, K. Taira, and R. Wadhwa
Emerging Technologies: Trendy RNA Tools for Aging Research
J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2004; 59(8): B771 - B783.
[Abstract] [Full Text] [PDF]

Home page
 RNAHome page
S. K. SILVERMAN
Rube Goldberg goes (ribo)nuclear? Molecular switches and sensors made from RNA
RNA, April 1, 2003; 9(4): 377 - 383.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
A. Kertsburg and G. A. Soukup
A versatile communication module for controlling RNA folding and catalysis
Nucleic Acids Res., November 1, 2002; 30(21): 4599 - 4606.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
A. M. Jose, G. A. Soukup, and R. R. Breaker
Cooperative binding of effectors by an allosteric ribozyme
Nucleic Acids Res., April 1, 2001; 29(7): 1631 - 1637.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
G. A. Soukup and R. R. Breaker
Engineering precision RNA molecular switches
PNAS, March 30, 1999; 96(7): 3584 - 3589.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (155K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (45)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Tang, J.
Right arrow Articles by Breaker, R. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, J.
Right arrow Articles by Breaker, R. R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?