Proc. Natl. Acad. Sci. USA Vol. 96, pp. 173–178, January 1999 Evolution

Emergence of a dual-catalytic RNA with metal-specific cleavage and ligase activities: The spandrels of RNA evolution (2؅,5؅-link͞prebiotic͞manganese͞RNA editing͞ribozyme)

LAURA F. LANDWEBER* AND IRINA D. POKROVSKAYA

Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ 08544

Communicated by Thomas E. Shenk, Princeton University, Princeton, NJ, October 26, 1998 (received for review August 7, 1998)

ABSTRACT In vitro selection, or directed molecular evolu- triggers the switch from selected ligase to unselected self- tion, allows the isolation and amplification of rare sequences that cleaving RNA. The facile construction of a minimal RNA satisfy a functional-selection criterion. This technique can be ligase ribozyme from random sequences and the simple con- used to isolate novel ribozymes (RNA enzymes) from large pools version between two RNA metalloenzymes (9) by a single of random sequences. We used in vitro evolution to select a metal switch imply a surprising plasticity of RNA evolution. ribozyme that catalyzes a novel template-directed RNA ligation that requires surprisingly few nucleotides for catalytic activity. METHODS With the exception of two nucleotides, most of the ribozyme Nucleic Acids. The initial DNA pool was prepared as a contributes to a template, suggesting that it is a general prebiotic 132-base oligonucleotide (16 nucleotides of constant region ligase. More surprisingly, the catalytic core built from random- flanking 100 bases of random sequence) on a Millipore ized sequences actually contains a 7-nt manganese-dependent Expedite synthesizer, confirmed for even degeneracy by cycle self-cleavage motif originally discovered in the Tetrahymena sequencing, gel purified, and amplified by using large-scale group I intron. Further experiments revealed that we have PCR (1) with the 5Ј and 3Ј primers to extend the constant regions selected a dual-catalytic RNA from random sequences: the RNA (5Ј PCR primer including T7 promotor 5Ј-TTCTAATACGACT- Ј Ј promotes both cleavage at one site and ligation at another site, CACTATAGGCTCACACTTGTATAGATCTACT-3 ;3 con- Ј suggesting two conformations surrounding at least one divalent stant region PCR primer: 5 -(CUA)4CGGGATCCTAATGAC- Ј metal ion-binding site. Together, these results imply that similar 3 ). Pool RNA for each round was gel-purified and eluted by a catalytic RNA motifs can arise under fairly simple conditions crush-and-soak method after T7 transcription [with T7 RNA and that multiple catalytic structures, including bifunctional polymerase either purchased from Epicentre Technologies ligases, can evolve from very small preexisting parts. By breaking (Madison, WI) or prepared as described in ref. 11] from apart and joining different RNA strands, such ribozymes could phenol-extracted, ethanol-precipitated PCR templates. Dou- have led to the production of longer and more complex RNA ble-stranded amplified DNA was also gel-purified after the polymers in prebiotic evolution. first round of amplification. Approximately 20 nmol (1.5 mg) of pool RNA was used in the first round, 0.1 nmol was used in the second round, and 20 pmol was used in subsequent rounds. in vitro The selection of functional RNA molecules from random One of three substrate RNAs was used in each alternating sequence pools has successfully isolated far more classes of Ј selection round (RNA1, 5 biotin-(cat)4agGUACAAGUGUAG- ribozymes than have yet been found in nature. The classic Ј Ј AAAAUCAGUCUUUUUU-3 ; RNA2, 5 biotin-(cat)4ctcga- experiment (1) revealed an abundance of RNA ligase ribozymes gaaUUAUACAAGAAAGAUUACUCUUUUU-3Ј; RNA3, in a random sampling of 1015 unique sequences. The first exper- Ј Ј 5 ggga(cau)4agguaCAAGAAAGAUACAGUCUUUUU-3 ; iments to probe in vitro evolution of ribozymes searched for tag sequences are lowercase. RNA1 and RNA2 were con- improved or altered versions of the naturally occurring Tetrahy- structed on a Millipore Expedite synthesizer [with DNA mena group I ribozyme. For example, Green et al. (2) used in vitro sequence tags italicized and biotin-dT (Glen Research, Ster- evolution to restore and to improve activity of a pool of molecules ling, VA) at the 5Ј end] and purified by anion exchange HPLC based on the Tetrahymena group I intron, whereas Joyce and followed by gel purification. RNA3, used in largest quantity in colleagues subsequently isolated variants of the group I ribozyme the first round (1 mg, 90 nmol) was prepared by T7 transcrip- that performed subtly different tasks, such as cleavage of a DNA tion (11, 12) and gel purified. Other RNA substrates and small substrate (3) or acquisition of a novel dependency on a specific ribozymes were either prepared by T7 transcription of oligo- divalent metal cation (4). The ability of directed evolution and nucleotide or PCR templates or purchased from Yale Uni- ribozyme engineering to evoke secondary activities of group I versity (Keck Oligonucleotide Facility). variants such as cleavage of aminoacyl ester or amide bonds (5–7; In Vitro Selection Strategy. The selection protocol we followed reviewed in ref. 8) suggests that RNA has an intrinsic capacity to was similar to the one described in refs. 1 and 10. In the first seven adapt to novel contexts or substrates, although in these cases the rounds, 1 ␮M pool RNA and 4-fold excess substrate were acquisition of new catalytic activity was aided by careful design of incubated overnight at high concentrations of Kϩ and Mg2ϩ (400 ͞ ͞ the substrate to mimic the natural phosphodiester bond. mM KCl 60 mM MgCl2 30 mM Tris, pH 7.4) at 21–25°C to In this paper we report the in vitro selection and evolution facilitate binding and catalysis with the substrate usually tethered of a very small ribozyme that catalyzes the ligation of two to a solid support to prevent loss of pool RNA caused by substrate RNAs but also possesses the unexpected ability to precipitation at elevated Mg2ϩ concentrations (1). Mg2ϩ was undergo a separate self-cleavage reaction. Substitution of a always added last. Round 1 was performed with the substrate divalent metal ion in the absence of further sequence evolution RNA3 tethered to streptavidin agarose via a 5-fold excess of a biotinylated DNA oligonucleotide complementary to the se- The publication costs of this article were defrayed in part by page charge quence tag, and rounds 2, 4, and 5 were performed with biotin- payment. This article must therefore be hereby marked ‘‘advertisement’’ in ylated substrate directly attached to magnetic beads. Use of a accordance with 18 U.S.C. §1734 solely to indicate this fact. © 1999 by The National Academy of Sciences 0027-8424͞99͞96173-6$2.00͞0 *To whom reprint requests should be addressed. e-mail: lfl@princeton. PNAS is available online at www.pnas.org. edu.

173 Downloaded by guest on September 25, 2021 174 Evolution: Landweber and Pokrovskaya Proc. Natl. Acad. Sci. USA 96 (1999)

solid support allowed specific retention of covalently ligated merase (Perkin–Elmer) by using the 3Ј sequence tag and a 5Ј ribozymes, and selective PCR with the tag sequence as primer PCR primer containing the T7 promotor fused to the first eight amplified only ligated sequences. Each round achieved Ͼ1000- nucleotides of the RNA 5Ј end (GGUGUGGU). fold enrichment of reacted RNA molecules. We introduced RNA Labeling. RNA molecules were either 32P-labeled at their different guide sequences as well as DNA sequence tags on the 3Ј termini with 3Ј-[␣-32P]dATP (NEN) and poly(A) polymerase substrate RNA in alternate rounds of selection to select for a (BRL) or 5Ј-end labeled either with T4 polynucleotide kinase general ligase (instead of one with restricted sequence require- (New England Biolabs) and [␥-32P]ATP or [␥-33P]ATP (NEN) or ments) and to increase the efficiency of the selection by prevent- by T7 transcription in the presence of [␣-32P]GTP and unlabeled ing sequence dependency on the substrate. ATP, CTP, and UTP (for RNase T2 digestion). T7 transcripts In round 1, pool RNA and substrate RNA3 were annealed were treated with phosphatase (calf intestinal alkaline phospha- and reacted on a 7.5-ml streptavidin agarose column (Pierce), tase, Boehringer Mannheim), phenol extracted, and precipitated and unligated pool RNA was released from the solid support before labeling with T4 kinase. in 25–100 mM NaCl, followed by elution of any reacted Ligation Reactions. Reactions were performed in 5-␮l vol- (ligated) RNA in 0–5 mM NaCl. Radiolabeled RNA3 was used umes (20 ␮l for time course reactions) with 1 ␮M template to monitor the presence of fractions containing covalently (ribozyme), 1 ␮M3Ј substrate (for trans-ligations), and trace ligated RNA. These were pooled and ethanol-precipitated amounts of 5Ј-end-labeled substrate in 400 mM KCl, 30 mM with glycogen as a carrier. Round 2 was performed similarly by buffer, 0.4 units͞␮l RNasin (Promega) at the indicated pH and using biotinylated RNA1 and 0.25 ml of M280 Streptavidin with the indicated divalent cation. Tris buffer was used for pH (Dynal, Great Neck, NY) magnetic beads. RNA collected from 7.5–9.0 and 2-(N-cyclohexylamino)ethanesulfonic acid (Ches) round 1 was affinity-purified on a second column (2.5 ml of buffer was used for pH 8.5–10.0 (both were obtained from streptavidin agarose) containing 2 nmol of a 20-base oligonu- Sigma). Metal salts [MgCl2 (99.995%), MnCl2 (99.99%), CaCl2 Ј ϩ cleotide complementary to the 3 constant region to remove (99.99 %), SrCl2 (99.995%), BaCl2 (99.999%), and CdCl2 excess substrate (oligo 20.99 in ref. 1). Retained RNA was (99.99ϩ%)] were purchased from Aldrich and of the highest reverse-transcribed (SuperScript RNase HϪ, BRL) directly on purity available. Solutions were prepared in diethyl pyrocar- the magnetic beads when biotinylated substrates were used. bonate-treated H2O (Research Genetics, Huntsville, AL) and In the first five rounds, 10–15 cycles of selective PCR (1) filtered through 0.22-␮m Millipore filters before use. To avoid with the tag sequence as primer were followed by up to 33 oxidation, 1 M stock solutions were stored at Ϫ20°C and mixed cycles of nested PCR until a product was strongly visible on an with buffer before use. All reactions were incubated at 22– ethidium bromide-stained agarose gel. Selective PCR (25–35 25°C for between 0 hours and 7 days. Reactions were stopped cycles) was used in later rounds when a product was readily by addition of an excess volume of EDTA in formamide visible, with only five cycles of nested PCR used to introduce loading buffer (95% deionized formamide͞10 mM EDTA͞ the T7 promotor. Nonspecific RNA binding to the streptavi- 0.02% xylene cyanol and bromphenol blue). din-coated beads (Pierce, Dynal) was minimal (Ͻ0.01%) in RNAse T2 Digestion. RNase T2 reactions were carried to most rounds. Progressively shorter incubation times, lower completion by incubation for 30 min at 37°C in 5 ␮l containing concentrations of Mg2ϩ, and mutagenic PCR were introduced 50 mM sodium acetate (pH 4.5), 2 mM EDTA, and 10–20 units in later rounds to allow refinement of the catalytic pool. of RNase T2 (BRL). RNA for this experiment was labeled at Beginning with round 6, biotinylated RNA⅐cDNA hybrids were the phosphate of the ligation junction by ligating a 10-base captured onto streptavidin-coated magnetic beads after reverse substrate RNA containing a 32P label at the ␣ phosphate of the transcription (because rogue RNA molecules that bind the 5Ј-triphosphate to the 22-nt RNA (Fig. 1c) and gel-purifying streptavidin support lose their affinity after reverse transcription the ligated product. Digestion products were analyzed on a and conversion to double-stranded molecules). These RNA⅐DNA 20% acrylamide, 8.3 M urea gel. hybrids were washed off the solid support while specifically Manganese-Dependent, Site-Specific Cleavage. RNA5 (15) retaining 5Ј-biotinylated duplexes in TE (10 mM Tris⅐HCl͞1mM was a gift from S. Kazakov (Somagenizo, Palo Alto, CA). ␮ EDTA) buffer. After DNA restriction-fragment and sequence Reaction volumes (10 l) containing 10 mM MnCl2, 100 mM analyses revealed the presence of a single class of closely related NaCl, and 50 mM Tris⅐acetate (pH 7.5) were incubated at 45°C molecules in round 7, the addition of mutagenic PCR (7 mM for 1 hr. Reactions were stopped by addition of an equal ͞ ͞ MgCl2 0.2 mM each purine nucleotide 1 mM each pyrimidine volume of formamide loading buffer, incubated for 15 min at ͞ nucleotide 0.5 mM MnCl2; see refs. 13 and 14) for mixtures of 37°C (to avoid nonspecific hydrolysis of RNA), and loaded approximately 30, 60, and 90 doublings (1) with 200,000-fold directly onto a 20% acrylamide, 8.3 M urea gel. dilutions between double sets of PCR introduced more variation. In each subsequent round, 30 cycles of nested PCR were per- RESULTS AND DISCUSSION formed with mutagenic conditions on 200,000-fold dilutions of Selection for a General Ligase. Inspired by the biological DNA recovered after selective PCR. The MgCl2 concentration process of RNA editing, which uses multiple cycles of cleavage, U during ligation was reduced to 10 mM in rounds 8–11 and 5 mM addition, and ligation to rewrite up to 90% of the coding- in rounds 12–15. Incubation time was reduced from overnight to sequence information in kinetoplastid mitochondrial transcripts 3 hr in round 10, to 30 min in round 11, to 15 min in round 12, (16), we designed a substrate RNA and a pool 5Ј constant region and to 2 min (after preincubation of RNA with substrate in the that contained a truncated NADH dehydrogenase subunit 7 absence of MgCl2) in rounds 13 and 14. All reactions were (ND7) guide RNA from Leishmania tarentolae and 18 nucleotides quenched by adding EDTA. Selective PCR products were cloned of the 5Ј editing domain in the ND7 mRNA, respectively (17). A into a UDG cloning vector (BRL). side activity of RNA editing that has been observed both in vivo Minimal Base-Pairing Requirements. One round of selec- (18) and in vitro (17) is the covalent attachment of the guide RNA tion was performed to determine the pairing requirements for to the mRNA molecule. We sought to test whether this activity ligation by ligating overnight 1 ␮M randomized hairpin RNA is an intrinsic property of the RNA molecules themselves. The (Fig. 1d) to 3-fold excess RNA substrate containing the 10-nt one unifying feature among all guide RNA molecules is the sequence 5Ј-pppGACUCACACU fused to the 22-nt sequence presence of a 3Ј oligo(U) tail (18); hence, we initially constructed tag from RNA2 in a 0.9-ml reaction volume (30 mM Tris⅐HCl, all 5Ј substrate molecules with a short (5–6 nt) oligo(U) tail. From ͞ ͞ ϫ 15 pH 8.5 0.4 M KCl 100 mM MgCl2). Product ligated to the a pool of 1.6 10 unique sequences containing 100 nucleotides 10-nt substrate (Ϸ1.2%) was visualized with autoradiography of random sequence flanked by constant regions, a single class of and isolated on a 12% polyacrylamide gel containing 8.3 M RNA molecules emerged after six cycles of selection for the urea, reverse transcribed, and amplified by rTth DNA poly- ability to ligate each of three tagged substrate RNAs to their 5Ј Downloaded by guest on September 25, 2021 Evolution: Landweber and Pokrovskaya Proc. Natl. Acad. Sci. USA 96 (1999) 175

FIG.1. (a) Possible secondary structure of one of the ligase ribozymes selected from round 13, shown ligated to the substrate guide RNA (gRNA). The 5Ј and 3Ј constant regions of the pool are shaded (and sequences derived from the 5Ј constant region are shown in blue throughout Fig. 1). The 3Ј-U6 tail (underlined and in green) and the 5Ј biotin of the guide RNA substrate are marked. (CAT)4, a DNA sequence tag, and other sequences in red are arrow indicates the 3Ј-terminal base needed for efficient ligation. Dashed lines ء not needed. The core region of the ribozyme is boxed in the center. The and arrows indicate sites of allowed breakage for trans-ligation or truncation experiments. U–A pairs in magenta are weakly supported by sequence covariation. Substitutions in black increase activity, and mutations in light gray abolish activity, following a grayscale. The small arrow indicates attack of the 2Ј-hyrdoxyl of the 3Ј-U6 tail on the 5Ј-triphosphate of the G at the end of the RNA. (b) Very small cis-ligation with the same substrate as in a, Ϫ4 Ϫ1 Ϫ4 Ϫ1 kobs ϭ 3.4 ϫ 10 min .(c) Very small trans-ligation with a 10-nt 3Ј substrate and a 29-nt ribozyme (boxed), kobs ϭ 5.4 ϫ 10 min at 25°C and 30 ϫ 10Ϫ4 minϪ1 at 45°C. (d and e) Two very simple catalytic motifs still ligate with up to 10% (d) and 4% (e) the efficiency of the ligation in b or c, demonstrating remarkable versatility of the small ligase motif. Randomized nucleotides (N) are not intended to be base paired.

end. Covalent attachment to the substrate marks the pool RNA Because of the difficulty of sampling such a large sequence molecules with either biotin or a specific sequence tag (1), and the space [fewer than 1016 molecules of a possible 4100 Ͼ1060], the strategy of recovering biotin-tagged ribozymes as RNA⅐cDNA ribozymes selected initially were expected to be suboptimal hybrids avoided recovery of single-stranded RNAs that bind only representatives of a class of catalytically active molecules (1). streptavidin. Seven more rounds of in vitro selection were performed by using All ribozymes that survived the selection contained a se- heavy mutagenic PCR on the pool of ribozymes recovered after quence within the random region that can anneal to the 5Ј the seventh round of selection. This improved the reaction by constant region of the pool molecules as well as to multiple reducing the required magnesium concentration and reaction substrate RNAs, forming a long imperfect stem, which posi- time. tions the 3Ј terminus of the oligo(U) tail of the substrate RNA Alignment of 51 clones revealed a sequence consensus for near the 5Ј-triphosphate of the pool RNA (Fig. 1a). The 5Ј-end this ribozyme with only a conserved 29-nt core region (Fig. 2). ligation catalyzed by this RNA involves nucleophilic attack by By assaying different clones, we verified that the extent of base the small substrate RNA on the ␣ phosphate of a 5Ј- pairing between ribozyme and substrate in the conserved triphosphate (1). We found that this ligation requires the region contributes to the rate of ligation. In addition, most presence of a 5Ј-triphosphate, or activation of the ␣ phosphate sequence covariation allowing G⅐U pairs preserves pairing to by pyrophosphate, because activity is lost with phosphatase- the substrate in the core region. treated or kinase-treated RNA (not shown) that contains Remarkably, this ligase ribozyme has the fewest sequence either a 5Ј-hydroxyl or 5Ј-monophosphate, respectively. requirements of any selected ribozyme, with only two nucleotides

FIG. 2. Alignment of 12 representative functional ribozyme clones from a total of 51 such clones. A dash indicates identity with the top sequence. A ь indicates a deletion. The most conserved region is boxed. The constant regions are shaded in the reference sequence. Downloaded by guest on September 25, 2021 176 Evolution: Landweber and Pokrovskaya Proc. Natl. Acad. Sci. USA 96 (1999)

directly implicated in a nontemplate role. The only known smaller ribozyme promotes cleavage rather than ligation (15) and is derived from an RNA hairpin found in the Tetrahymena group I intron (19). The ligation catalyzed by this ribozyme thus differs from the ligations catalyzed by other RNA ligase ribozymes (10) by requiring a G–G juxtaposition at the site of ligation and very few selected nucleotides for activity (Fig. 1). We constructed a minimal version of this RNA-ligase ri- bozyme. Deletion of sequences 3Ј of the boxed conserved region (Fig. 1a) led to a 2-fold increase in activity; however, deletion of Ј some sequences 5 of the boxed region produced a slight decrease FIG. 4. Ribonuclease T2 digestion to determine regiospecificity of in activity, indicating that some of these bases may contribute bond formation. Lane 2 is the RNase T2 digestion product of the modestly to catalysis. The smaller ribozyme contains only sub- ligation in Fig. 1c (see Methods). Lane 1 is the RNase T2 digestion strate nucleotides and a central 27-nt core derived from the product of a ligated RNA of identical sequence, but with a labeled 3Ј,5Ј originally selected ribozyme. This represents the smallest such linkage generated by T4 DNA ligase and a complementary DNA splint ribozyme isolated by in vitro selection, with only two potentially (42). Digestion products were analyzed on a 20% denaturing acryl- amide gel. Lanes 3 and 4 are a time course of alkali hydrolysis at 5 and unpaired nucleotides required for catalysis (Fig. 1b). 10 min, respectively, of the ligated product digested in lane 1. Ligation in Trans. The ribozyme also functions efficiently in trans (Fig. 1c) but without turnover because of the slow We examined the dependency of the ligation on time, Mg2ϩ dissociation of the enzyme–substrate complex (consisting of concentration, and pH (Fig. 5). The reaction exhibits approxi- three annealed strands that together form the catalytic RNA mately linear dependence on Mg2ϩ concentration below 50 mM ϩ system). The pseudo-first-order rate constant kobs for this and a saturable Mg2 binding site with half-maximal activity ϫ Ϫ4 Ϫ1 reaction was 2.7 10 min for the time course shown in Fig. above 50 mM over a pH range of 7.0 to 9.0 (data not shown; ref. 3. The 29-nt ‘‘ribozyme’’ is essentially as active as one con- 27). The logarithm of k increases linearly with a slope of Ј obs taining the full 5 end. To our surprise, we discovered that a approximately 1 as a function of pH, indicating a single critical general ligase ribozyme stripped of its template either up- deprotonation step—probably at the 2Ј hydroxyl. The pH plateau stream or downstream of the ligation site still ligates several above pH 9.0 is consistent with an attacking 2Ј hydroxyl (Fig. 5; sequences, albeit with approximately 4% the original effi- pH Ͼ9.5 leads to degradation, deprotonation, and helix disrup- ciency (Fig. 1 d and e). An alignment of 36 successful ligase tion and so could not be validated). Deletion analysis mapped the clones derived from a single round of selection (as in Fig. 1d) 3Ј and 5Ј terminal positions required for activity to a small exhibited no marked sequence consensus, suggesting that the conserved core, which provides mostly a template. Only two catalytic structures are versatile. In fact, the number of nucleotides are directly implicated in a nontemplate role: the G Watson–Crick or G⅐U base pairs found in the randomized opposite the ligation site and a single bulged pyrimidine (Fig. 1). region of these sequences was not significantly greater than Mutational Analysis and Rate Enhancement. Many single ϭ expected for any pair of random hexamers (nobserved 2–3 bp; point mutations (summarized in Fig. 1a) abolish activity below 3 n ϭ 6( ⁄8) ϭ 2.25 expected at random). our threshold of detection. For example, deletion of the single Creation of 2؅–5؅ Linkages. The regioselectivity of bond bulged U or introduction of any purine base opposite (and formation is exclusively 2Ј,5Ј-phosphodiester linkages as as- potentially paired with) this U results in complete loss of sayed by resistance of 2Ј,5Ј linkages to ribonuclease T2 (Fig. 4). activity, as does mutation of the G opposite the cleavage site This is consistent with the absence of extended base pairing at to any other base or mutation of the adjacent U⅐G base pair to the ligation junction, as a Watson–Crick duplex favors 3Ј,5Ј a C-G base pair. On the other hand, conversion of this U⅐G pair linkages (20) but considerably destabilizes 2Ј,5Ј linkages (20– toaU-A (Fig. 1a) leads to a 4- to 10-fold increase in both cis- Ј Ј ϫ Ϫ4⅐ Ϫ1 22). Although nonbiological in extant systems, 2 ,5 linkages and trans-ligation (kobs approx. 6 10 min ). are favored in nonenzymatic oligoribonucleotide synthesis We estimated the rate enhancement of our ribozyme- because of the increased nucleophilicity of the 2Ј-hydroxyl catalyzed ligation by comparing it against single and double toward activated phosphate esters (23). However, recent ex- point mutants that lack the bulged U and͞or contain a C-G periments demonstrate that both 2Ј,5Ј linkages and mixed pair instead of a G–G juxtaposition at the ligation site. This linkages are active in template-directed oligomerization of gave an upper boundary of kobs for the background ligation of mononucleotides (24) and oligonucleotides (25–26), and this 3 ϫ 10Ϫ8 minϪ1, a rate enhancement of the catalyzed reaction has strengthened their possible role in primitive biochemistry. greater than 20,000-fold. Our ability to measure the back-

FIG.3. (Left) Time course of representative trans-ligation shown in Fig. 1c with 10-nt 3Ј substrate; Lower band, labeled 22-nt guide RNA band; Ϫ4 Ϫ1 upper band, 32-nt ligated product band. kobs ϭ 2.7 ϫ 10 min .(Right) Time course of randomized cis-ligation shown in Fig. 1e. M, marker lane; lower band, labeled 41-nt partially randomized RNA; upper band, ligated 63-nt product. Downloaded by guest on September 25, 2021 Evolution: Landweber and Pokrovskaya Proc. Natl. Acad. Sci. USA 96 (1999) 177

ground reaction containing the C-G pair may be hampered by the preferential hydrolysis of 2Ј,5Ј linkages in a paired duplex (22); however, we estimated the same upper boundary for the single mutant lacking only the bulged U, which preserves the rest of the ligation site. This reaction also is 3,000-fold faster than a template-dependent uncatalyzed ligation with a similar Ј ϫ Ϫ8 Ϫ1 3 -terminal U-A base pair (kobs 8.5 10 min ; ref. 20). Ligation of Hexauridylate. Because the guide RNA substrate in the original experiments contains an oligo(U) sequence at its 3Ј end, we tested a 6-nt substrate, 5Ј,3Ј-(UUUUUU). Even this small RNA is a substrate for both cis- and trans-ligation (Fig. 6), albeit with 0.5% the efficiency of the 22-nt substrate shown in Fig. 1 c and e. As this sequence is small enough to be synthesized by mineral-catalyzed RNA condensation (28), this result suggests the possibility that we have uncovered an RNA ligation reaction with general prebiotic utility. Manganese Dependency. Surprisingly, the catalytic core built from random sequences actually contains the very small manganese-dependent cleavage motif formed during autocy- FIG. 6. Ligation of 5Ј,3Ј-(UUUUUU) (lower band) to a 10-nt clization of the Tetrahymena group I intron (refs. 15 and 19; substrate. Upper band is the 16-nt product. Lane 1 was incubated Fig. 7). The trinucleotide UUU in a complex with GAAA (the without substrate and lane M is a marker; lane 2, 3-day incubation; lane smallest known catalytic RNA system) acts as a template 3, 6-day incubation. catalyst (15, 29) to promote cleavage between the G and the A (15). The Roles of RNA and the Metal. The simplicity of the ligation Further experiments revealed that we had in fact selected a depicted in Fig. 1 d and e suggests that this motif can be dual-catalytic RNA from random sequences. This RNA pro- generalized even further. The GNRA tetraloops at the ends may motes both cleavage at one site (Fig. 7) and ligation at another act as terminal clamps, thus reducing the need for base pairing to (Fig. 1), with possibly two conformations surrounding a diva- juxtapose substrates along the ribozyme template. The role of lent-metal ion binding pocket (or pockets). We therefore divalent metal ions is particularly relevant, as they are generally tested the metal-ion dependence of the original ligation: at pH thought to have been important prebiotic catalysts (31). Evoking 7.0, the ligation proceeds Ϸ10-fold faster with Mn2ϩ than with the notion of a ‘‘Cheshire catalyst’’ (32), in which an RNA scaffold Mg2ϩ (Fig. 5), but at higher pH the required concentration of mainly positions a catalytic metal ion (9), we suggest that the core Mn2ϩ leads to RNA degradation. nucleotides—most notably the bulged pyrimidine (Fig. 1)—serve We subsequently confirmed that the selected ligase also un- two roles: (i) to bring together the 2Ј and 5Ј termini in a 2ϩ dergoes self-cleavage in the presence of Mn (Fig. 7) with kobs configuration that resembles the transition state for ligation (33) ϭ 0.01 minϪ1 for self-cleavage of RNA5 (15), compared with and (ii) to properly position the metal ion(s) in a favorable Ϫ1 ϭ ϫ 0.0036 min for cleavage of GAAApCp (30) and kobs 5–8 geometry, whereas the conserved flanking nucleotides (or tetra- 10Ϫ4 minϪ1 for cleavage of the 29-nt core ribozyme (template loop) provide intrinsic binding energy that leads to the closed strand in Fig. 1c) in the presence of either 5–50 ng͞␮l poly(U), 0.5 state required for catalysis (34). nM–5 ␮M5Ј,3Ј-UUUUUU, or 0.5 ␮M 22-nt substrate RNA A Case for RNA Preadaptation. The self-cleaving GAAA͞ shown in Figs. 1 c and e as template catalyst. In addition, the UUU motif arose in this case as an unintended consequence ligation shown in Fig. 1c proceeds remarkably well in optimal of selection for ligation at an independent site. In other words, ϭ Ϫ1 cleavage buffer conditions, with kobs 0.003 min at 45°C. it appears as a ‘‘spandrel’’ (35–36), an offshoot or structural Other Divalent Metal Ions. Cd2ϩ also supports G2AAA byproduct of the design of an RNA ligase ribozyme from cleavage (15) but supports ligation only weakly. Ca2ϩ,Sr2ϩ, random sequences. A spandrel has been defined as ‘‘any and Ba2ϩ also catalyze the ligation. Overall the ligation rates geometric configuration of space inevitably left over as a (Mn2ϩ Ͼ Mg2ϩ Ͼ Ca2ϩ Ͼ Sr2ϩ ϭ Ba2ϩ) vary inversely with the consequence of other architectural decisions’’ (36). Because 2 pKa of their bound water molecules and follow the same order the ability to cleave within G AAA clearly did not emerge as as hammerhead cleavage and nonenzymatic ligation (20), an adaptation itself during the selection experiment—as it was implicating a metal hydroxide or directly coordinated metal neither selected for nor feasible in the magnesium-containing ion in the catalytic mechanism. buffer—this additional activity presumably results from the

ϩ FIG. 5. Biochemistry of ligation reaction. (Left) pH dependence and (Center and Right) effect of replacing Mg2 with a series of divalent cations 2ϩ 2ϩ 2ϩ using ribozyme/substrate complex in Fig. 1c.(Center) kobs as a function of Mg (छ) and Mn (ᮀ) at pH 7.0. (Right) kobs as a function of Mg (ᮀ), Ca2ϩ (छ), Cd2ϩ (E), Sr2ϩ (‚), and Ba2ϩ (crossed squares) at pH 9.0. Downloaded by guest on September 25, 2021 178 Evolution: Landweber and Pokrovskaya Proc. Natl. Acad. Sci. USA 96 (1999)

tion employs a different leaving group, however, than does kinetoplastid editing. These observations do hint at another possible role of the poly(U) tails on kinetoplastid guide RNAs: they could assist in the cleavage at frequent G2AAA editing sites. This reaction might have played a more significant role early in the history of RNA editing. Cleavage of RNA at specific sites has been suggested as a mechanism for generating abrupt translation termination sig- nals (15); however, the presence of two metal-dependent activities in a single RNA molecule confers the ability to modulate biochemical activity in response to environmental cues. A single metal ion switch thus adds to the list of possible mechanisms that would have been available for control of gene expression in an RNA world.

ϩ We thank Sergei Kazakov for advice, Anne Torjussen, Caroline FIG.7. Mn2 -dependent RNA cleavage. The complex of UUU and Fichtenberg, Monika Pfunder, and David Kutzler for assistance, Rob UGAAA undergoes site specific Mn-dependent cleavage at the arrow- Knight, Drew Ronneberg, and two anonymous reviewers for comments, head (15). The U is not required in the substrate but was present in the and Jack Szostak and members of his laboratory for suggestions and RNA original sequence (19) and, surprisingly, the entire 8-nt motif shown here oligonucleotides. This research was supported by National Science Foun- appears in our selected ligase, just opposite the ligation site (Figs. 1 a–c dation Grants MCB-9520253 and MCB-9604377 and a Burroughs Well- and e). The lines represent optional RNA sequences, and the circle ϩ ϩ come Fund New Investigator Award in Molecular Parasitology to L.F.L. represents at least one binding site for a divalent ion—Mn2 or Cd2 to ϩ promote cleavage—and presumably Mg2 or other metals to promote 1. Bartel, D. P. & Szostak, J. W. (1993) Science 261, 1411–1418. ligation. Lanes M and 1 contain standards; lane 2 is unreacted 3Ј-end 2. Green, R., Ellington, A. D. & Szostak, J. W. (1990) Nature (London) 347, labeled 22-nt positive control synthetic RNA5 (15), and lane 3 is RNA5 406–408. digested to an 18-nt product. Lanes 4–6 represent Mn2ϩ cleavage of the 3. Beaudry, A. A. & Joyce, G. F. (1992) Science 257, 635–641. 29-nt ribozyme strand shown in Fig. 1c to a 9-nt shorter product in the 4. Lehman, N. & Joyce, J. F. (1993) Nature (London) 361, 182–185. presence of Mn2ϩ and 0.05, 0.5, and 5 ng/␮l poly(U). 5. Piccirilli, J. A., McConnell, T. S., Zaug, A. J., Noller, H. F. & Cech, T. R. (1992) Science 256, 1420–1424. 6. Dai, X., De Mesmaeker, A. & Joyce, G. F. (1995) Science 267, 236–240. need to construct a binding site for one or more divalent metal 7. Joyce, G. F., Dai, X. & De Mesmaeker, A. (1996) Science 272, 18–19. ions. In vitro evolution therefore provides an unambiguous 8. Landweber, L. F., Simon, P. J. & Wagner, T. A. (1998) Bioscience 48, example of preadaptation, in the sense that the product of 94–103. evolution was ultimately suited for a catalytic function other 9. Pyle, A. M. (1993) Science 261, 709–714. 10. Ekland, E. H., Szostak, J. W. & Bartel, D. P. (1995) Science 269, 364–370. than the one for which it was selected. It could thus be co-opted 11. Zawadzki, V. & Gross, H. J. (1991) Nucleic Acids Res. 19, 1948. to specialize in its new role, such as cleavage, which would 12. Milligan, J. F. & Uhlenbeck, O. C. (1989) Methods Enzymol. 180, 51–62. make this feature an exaptation (37). In a separate selection 13. Caldwell, R. C. & Joyce, G. F. (1992) PCR Meth. Applic. 2, 28–33. experiment, Faulhammer and Famulok (38) unveiled a DNA 14. Fromant, M., Blanquet, S. & Plateau, P. (1995) Anal. Biochem. 224, 347–353. enzyme that was preadapted for calcium binding. The pool 15. Kazakov, S. & Altman, S. (1992) Proc. Natl. Acad. Sci. USA 89, 7939–7943. molecules had never been exposed to calcium during the 16. Landweber, L. F. & Gilbert, W. (1993) Nature (London) 363, 179–182. selection; however, the optimal solution for magnesium- 17. Blum, B. & Simpson, L. (1992) Proc. Natl. Acad. Sci. USA 89, 11944–11948. dependent cleavage of RNA coincidentally produced a good 18. Blum, B., Sturm, N. R., Simpson, A. M. & Simpson, L. (1991) Cell 65, solution to the problem of calcium-dependent cleavage as well, 543–550. 19. Dange, V., Van Atta, R. B. & Hecht, S. M. (1990) Science 248, 585–588. as though the geometry of binding the metal hydroxide 20. Rohatgi, R., Bartel, D. P. & Szostak, J. W. (1996) J. Am. Chem. Soc. 118, ‘‘anticipated’’ the best fit for binding calcium. 3340–3344. Conclusions. The results presented here suggest that bifunc- 21. Usher, D. A. (1972) Nat. New Biol. 235, 207–208. tional catalytic RNA motifs can arise under fairly simple condi- 22. Usher, D. A. & McHale, A. H. (1976) Proc. Natl. Acad. Sci. USA 73, 1149–1153. tions and that multiple catalytic structures, including ligases, can 23. Lohrmann, R. & Orgel, L. E. (1978) Tetrahedron 34, 853–855. evolve from basic preexisting building blocks. By extension, we 24. Prakash, T. P., Roberts, C. & Switzer, C. (1997) Angew. Chem. Int. Ed. Engl. predict that the combinatorial shuffling of catalytic modules 36, 1522–1523. (such as metal binding sites) into random sequence pools will 25. Ertem, G. & Ferris, J. P. (1996) Nature (London) 379, 238–240. greatly increase the density and representation of nucleic acid 26. Sawai, H., Totsuka, S., Yamamoto, K. & Ozaki, H. (1998) Nucleic Acids Res. 26, 2995–3000. catalysts and that this will lead to a tremendous acceleration of in 27. Rohatgi, R., Bartel, D. P. & Szostak, J. W. (1996) J. Am. Chem. Soc. 118, vitro evolution. Along similar lines, Burke and Willis (39) recently 3332–3339. introduced recombination between RNA libraries to build 28. Ding, P., Kawamura, K. & Ferris, J. P. (1996) Origins Life Evol. Biosphere aptamers (RNA ligands) that can bind two specific targets. Our 26, 151–171. 29. Orgel, L. E. (1986) J. Theor. Biol. 123, 127–149. finding that RNA structures as simple as a few unpaired nucle- 30. Bombard, S., Kozelka, J., Favre, A. & Chottard, J.-C. (1998) Eur. J. Bio- otides in a short helix may possess two catalytic activities suggests chem. 252, 25–35. that similar metalloribozymes (i) arise surprisingly easily and 31. Joyce, G. F. & Orgel, L. E. (1993) in The RNA World, eds. Gesteland, R. F. would undoubtedly have been present in prebiotic evolution (as & Atkins, J. F. (Cold Spring Harbor Lab. Press, Plainview, NY) pp. 1–25. ii 32. Yarus, M. (1993) FASEB J. 7, 31–39. well as at high frequency in random pools) and ( ) could have 33. Harada, K. & Orgel, L. E. (1993) Proc. Natl. Acad. Sci. USA 90, 1576–1579. contributed to the generation of RNA diversity by breaking apart 34. Hertel, K. J., Peracchi, A., Uhlenbeck, O. C. & Herschlag, D. (1997) Proc. and joining new sequence combinations. Natl. Acad. Sci. USA 94, 8497–8502. Processes such as kinetoplastid RNA editing that may be 35. Gould, S. J. & Lewontin, R. C. (1979) Proc. R. Soc. London B 205, 581–598. primitive (40) require repeated cycles of both cleavage and 36. Gould, S. J. (1997) Proc. Natl. Acad. Sci. USA 94, 10750–10755. 37. Gould, S. J. & Vrba, E. (1982) Paleobiology 8, 4–15. ligation (41). These experiments demonstrate that a set of RNA 38. Faulhammer, D. & Famulok, M. (1996) Angew. Chem. Int. Ed. Engl. 35, molecules based initially on a Leishmania tarentolae guide RNA– 2837–2841. mRNA combination are capable of performing both reactions. 39. Burke, D. H. & Willis, J. H. (1998) RNA 4, 1165–1175. The full-length ND7 guide RNA is even a substrate for the ligase 40. Landweber, L. F. & Gilbert, W. (1994) Proc. Natl. Acad. Sci. USA 91, ϭ ϫ Ϫ5 Ϫ1 918–921. ribozyme (kobs 1.7 10 min ; however, kobs increases by a 41. Kable M. L., Seiwert, S. D., Heidmann, S. & Stuart, K. (1996) Science 273, factor of 2 by extending the ‘‘anchor’’ region of pairing between 1189–1195. ribozyme strand and guide RNA). The ribozyme-catalyzed liga- 42. Moore, M. J. & Sharp, P. A. (1992) Science 256, 992–997. Downloaded by guest on September 25, 2021