Small Efficient Hammerhead Ribozymes Philip Hendry, Trevor J. Lockett, and Maxine J. McCall 1. Introduction The hammerhead ribozyme was discovered as a self-cleavmg RNA molecule in certain plant vtroids and satellite RNAs (1). Shortly after its conserved features were defined (2,3), the hammerhead was shown to be able to act as a true enzyme, cleaving multiple substrates m a bimolecular reaction (4). The self-cleaving hammerhead can be divided in a number of ways into two, or even three, separate strands (45). The most useful form has almost all of the conserved nucleotides on the ribozyme strand, leaving minimal sequence requirements in the substrate strand. To be cleavable the substrate must possess the sequence 5’ UH (H is C, U, or A), where cleavage occurs to the 3’ stde of H (6) (Fig. 1). This particular configuration has been the para- digm for hammerhead ribozyme design since 1988. Here we describe varia- tions on this basic design, with the constant theme being to minimize the size of the ribozyme. The advantages of minimizing the size of the ribozyme are several-fold. An obvious advantage, for ribozymes which will be used as exogenously-supplied therapeutics, is that the cost of synthesis is reduced if the number of rtbonucle- otides is mimmized. A second advantage, for the exogenously-supplied ribozyme, is that delivery to cells may be aided if the ribozyme is small. An additional, and unexpected, advantage is that some minimized ribozymes cleave their substrates faster in vitro than the analogous, standard ribozymes do, although it is not yet known if this advantage is carried over to produce an enhanced effect in vivo. Two strategies for minimizing the size of the hammer- head ribozyme are described in this chapter. The first involves shortening the hybridizing arms of the ribozyme (which results m shortening helix I and/or From Methods m Molecular Me&we, Vol It Therapeutic Appkahons of Rlbozymes Edlted by K J Scanlon 0 Humana Press Inc , Totowa, NJ 1 Hendry, Lockeft, and McCall Helix III 1 Helix I Substrate 5’ N N N N N N N U HI7 N N N N NM 3’ 3’ r;r i il Ii Ii a,&, i.l I; il Ii i N 5’ A % A G Ribozyme G12 AG uA6 C-G s Helix II N-N N.N NON Loop II N N NN Fig. 1 Schematic representation of the hammerhead rtbozyme m complex with its complementary substrate. The ribozyme forms helix I with its 5’ arm and the substrate, helix III with its 3’ arm and the substrate, and helix II and loop 2 with the nucleotldes joming A, and G12 In the rtbozyme, all nucleotldes, except for the conserved C3 to A, and GU to 4, I, may be erther ribonucleotrdes or deoxyrlbonucleotldes The site of cleavage m the substrate is shown by the downward arrow, 3’ to HI7 (H = C, U or A) helix III in the ribozyme-substrate complex), whtle mamtaming maximal cleav- age rates. The second involves shortening, or completely elimmatmg, helix II in the ribozyme. The consequences of each of these modifications on the cleav- age activrties will be discussed. 2. Guidelines for Design The work we describe here is based largely on observations made using short substrates m vitro. To directly extrapolate these observations to the use of ribozymes m vivo IS difficult because of the unknown factors that operate wtthm a living organism. In particular, the ribozyme may be affected by the activities of RNA binding proteins; these proteins can either enhance or retard substrate binding and dissociation (7-9), may stabilize the ribozyme against degradation (10,11), or may direct the rrbozyme to specific compartments within the cell in a manner observed for some mRNAs (12). However, it 1s important to understand the basic processes of ribozyme cleavage and, by mak- ing some assumptions about the intracellular envtronment experienced by an RNA molecule, tt should be possible to project some of this m vitro experience mto therapeutic practice. Hammerhead Ribozymes 3 2.7. Minimiring Arm Lengths 2.1.1. Introduction The standard form of the hammerhead ribozyme, as used by most research- ers, is shown in Fig. 1. It has the conserved nucleotides C3 to A, and Gi2 to Ai5 i, 4 bp m helix II with a G C bp adjacent to A9 Glz, 4 nucleotides in loop II, and a variable number of nucleotides in the 5’ and 3’ arms which, on bmdmg to the substrate, form helix I and helix III, respectively. The goal here is to rede- sign this hammerhead ribozyme so that it contains as few bases as possible in the hybridizing arms without compromising the cleavage ability or specificity of the ribozyme. There are a number of steps m defining the hybridizing arms of a hammer- head ribozyme. First, the target site within the RNA of interest must be chosen. Good targets have the UH cleavage site located within an accessible region of the RNA, so that the ribozyme 1s readily able to hybridize to the site (see Note 1). Second, the number of base pairs to be formed between the ribozyme and sub- strate, or the extent of complementarity, must be chosen. The aspects to con- sider here are catalytic activity and specificity. Ribozymes that form a large number of base pairs with the substrate are unlikely to turnover. Smaller ribozymes with relatively short hybridizing arms are able to turnover rapidly and therefore have the potential for high catalytic activity. Therefore, the extent of complementarity should be such that the ribozyme-substrate complex formed is (relatively) stable under the conditions of the experiment, and this typically requires 1 l-l 7 bp. For specificity, the number of bp formed between the ribozyme and substrate should be large enough to make the target sequence unique, but not so large that imperfectly matched substrates form stable com- plexes (13). Statistically, about 13 nucleotides are required to uniquely define a particular site m a mRNA pool in a mammalian cell (see Note 2). Finally, once the target site and the number of nucleotides to be bound has been decided, the disposition of the nucleotides about the cleavage site must be determined. By far, the most common arrangement has been to target an equal number of nucle- otides on either side of Hi7. A ribozyme of this design we call a symmetric ribozyme. We have recently shown that this design does not produce the most rapid cleavage rates m vitro. 2.1.2. The Optimum Length for Helix I We observed that the length of helix I in the ribozyme-substrate complex has a very profound effect on the cleavage rate constant for that complex (14). In a number of systems we have varied the length of helix I, both by varymg the number of nucleotides to the 3’ side of the cleavage site of the substrates and by changing the length of the 5’ hybridizing arms of the nbozymes. All Hendry, Lockett, and McCall Length of Helix I (bp) Fig 2 Dependence of rate constants on length of helix II, with helix III constant at 10 bp substrates with varying numbers of nucleotrdes to the 3’ side of the cleavage site and 10 nucleotides on the 5’ side are cleaved by their cognate rtbozymes; I Kr RA- 101 10; A Kr RB-IO/lo; 0 TAT RB-lO/lO. Reactions condlttons; 10 mA4MgCl,, 37°C pH 7 13 From ref. 14 these experiments were performed under ribozyme excess conditions with the rrbozyme-substrate complex fully formed prior to initiation of the reaction. The cleavage rate constants for substrates with varying numbers of nucleottdes 3’ of Hi7 by three different symmetrrc (10 + 10) ribozymes are shown m Fig, 2. Cleavage rate constants of 2 1 -mer substrates by ribozymes with varying lengths of hybridizing arms are given m Table 1. Together these data demonstrate that the optimum length for helix I m a hammerhead rrbozyme 1s about 5 or 6 nucle- otides whether the length of the hehx 1s hmited by the length of the substrate or ribozyme. Ribozymes wtth longer 5’ arms are potentially limited m their acttv- tty by slow cleavage rates. 2.1.3. The Optimum Length for Helix III To determine whether there was an optimum length for helix III, the cleav- age rate constants were compared for ribozyme-substrate pairs with optimum, or near optimum, helix I lengths and either 10 or 6 bp rn helix III. The variation in rate constant observed is shown m Table 2. The effect is quite small, twofold Hammerhead Ribozymes 5 Table 1 Cleavage Rate Constants for 21-mer Substrates by Various Cognate Ribozymes Substrate Rlbozyme III/P TAT S21-lo/lob TAT RA- 1 O/l Ob 10110 TAT S21-IO/10 TAT RA- 1 O/5 5110 TAT S21-lO/lO TAT RA-5110 1015 Kr s21-lO/lO Kr RA-lO/lO lO/lO Kr S21-lO/lO Kr RA-6110 1016 k&mm 0 63” 0.09 -I- 0.01 lo+ 1 0.10 6.7 Condltlons; pH 7 13,37”C, 10 mMMgC12 ONumber of bp m hehces III and I, respectively bThe nomenclature for the substrates and rlbozymes 1s as follows The sequences of the substrate molecules are taken from naturally-occurrmg mRNAs and are identified by their on- gin, The TAT series are from the TAT gene of HIV-l, and the Kr senes are from the Krdppel gene of Drosophzla melanogaster Rlbozymes are denoted by an R followmg the identifying prefix, and substrates by the letter S and a number which indicates the number of nucleotldes m the substrate There are three versions of hammerhead rlbozyme used m this chapter, and they are denoted as nbozymes A, B and C Rlbozymes A (RA) are composed solely of RNA (with the exceptlon of the 3’ nucleotlde), rlbozymes B (RB) possess DNA m the arms that hybridize to the substrate (with the exception of nucleotldes 15 1 and 15 2 which remam as RNA [Fig. l]), and rlbozymes C (RC) are the same as rlbozymes B except that their helix II and loop II are also composed of DNA. The number of nucleotldes m the hybridlzmg arms (for nbozymes) or on each side of Cl7 (for substrates) are added to the name, with the first number referrmg the 5’ side and the second to the 3’ side For example, TAT I&4-5/10 IS an all-RNA rlbozyme with 5 nucleotldes m its 5’ arm and 10 nucleotldes m its 3’ arm, TAT S21-lO/lO 1s a 21-nucleotlde substrate with 10 nucleotldes each side of C,,, and TAT RA-5/10 and TAT S21-lO/lO form a complex with 10 bp m helix III and 5 bp m helix I (III/I = 10/5) CAt pH 8 00 Table 2 Effect of Length of Helix Ill on Cleavage Rate Constants S17-10/6 S13-616 Ribozyme III/I k,b,/min III/I k&mm TAT RB-lO/lO 1 O/6 1.8 f 0.4 616 0.92 f 0 1 Kr RA-lO/lO 1016 4.8 f 0 6 616 3.4 f 1.0 Kr RB-lO/lO 1016 3 2 f 0.3 616 17+04 Condltlons 10 mMMgQ, 37”C, pH 7 13 at most, with the more efficient cleavage occurring in ribozyme-substrate pairs possessing the longer helices III. Although we have not pursued this further, there is no reason to suspect that cleavage efficiency will be impaired rf the ribozymes are able to form even longer helices III with their substrates. How- 6 Hendry, Lockett, and McCall ever, there is the danger that excessrvely long hybridizing arms are able to form folded, stable structures that prevent substrate binding. 2.1.4. Multiple Turnover The observations above relate to reaction condttrons m which the rrbozyme is m excess of the substrate, and the rrbozyme and substrate were preannealed before mrtration of the reaction by addition of Mg2+. Under these condmons, substrate binding and product drssociatton have no effect on the observed cleav- age rate constants, However, under multiple turnover conditions, a desirable situation for therapeutic uses, the rates of substrate binding and product disso- ciation must also be considered. The rate of substrate association 1s difftcult to predict for large substrates and will be largely dependent on the structure of the RNA m that region (15). On the other hand, for a given sequence, the rate of dissociation of the cleavage product 1s expected to consistently decrease with increasing length (15,16). Given that the optimum length for helix I is around 5 or 6 bp, and that duplexes of this length usually drssocrate quite rapidly, It 1s the length of helix III that is most crucial m this respect. The rate constant for dissociation of either cleavage product in vitro may be readily esttmated by a number of techniques (see Note 3). As a rough guide, m condmons like that encountered m biological systems, (pH 7.0, 37°C 100 mM NaCl), the rate constants for drssociation of the helix III (after cleavage) are hkely to approach that observed for the cleavage step when the length of helix III is m the range 5 -9 bp, depending on the sequence. 2.1.5. Summary The most efficient hammerhead rtbozymes have 5 or 6 nucleotrdes in their 5’ hybridizing arms, so that they may form a helix I of 5 or 6 bp in complex with their substrates. On the other hand, the ribozymes should have a minimum of 5 or 6 nucleotides m then 3’ hybridrzmg arms, so that they may form a helix 3 of at least 5 or 6 bp m the complex. No diminution of cleavage rate constant under rrbozyme excess condmons 1s expected for ribozymes with longer 3’ hybrrdiz- mg arms, but the turnover rate under substrate excess conditions will be adversely affected by excessrvely long helices III. 2.2. Minimizing or Eliminating Helix II 2 2.1. Introduction Apart from the hybridizing arms, the other region of the hammerhead rrbozyme which may be reduced m size, or even eliminated, IS helix II and loop II. When hehx II and loop II are completely eliminated, so that the rrbozyme consists of two strands with free ends at the conserved nucleotides Ag and Glz, very slow cleavage of the substrate is observed (17,18). When Ag and Gr2 are Hammerhead Ribozymes 7 linked, either by nucleotides or non-nucleotide chains, slow cleavage is observed with linkers containing just 13 atoms, whereas more reasonable rates are observed with linkers containing 25 atoms (I 7). These results demonstrate that helix II is not essential for cleavage activity. However, the nature of the linker which replaces helix II greatly affects the cleavage activity of the small ribozyme (17,19). Experiments to determine rate constants for the cleavage of short substrates by ribozymes that have a truncated helix II have shown that the number of base pairs m helix II can be reduced to two without loss of cleavage activity, relative to the standard ribozyme with a four base-pan helix II (20). A further reduction to one base pan m helix II results in a lo-fold loss of activity, and ehmmation of helix II (where A9 and Gla are connected by a 4-nucleotide loop) results m around a loo-fold loss of activity, relative to the standard ribozyme (20-22). The short substrates with which these experiments have been carried out were about 13 ribonucleottdes m length. When the substrate length, increases, for example to 2 1 nt, with a concomitant increase m the lengths of hehces I and III, the cleavage rates of ribozymes with one or no base pairs in helix II increases relative to that of the standard ribozyme. Furthermore longer RNA transcripts are cleaved in vitro by these small ribozymes faster than they are cleaved by standard ribozymes (22). It is not known if the same relative rates of cleavage by the small and standard ribozymes also occur in viva. In order to distinguish these small ribozymes from the standard ribozymes, we define a ribozyme with one base pair m helix II as a mmiribozyme, and a ribozyme with no helix II as a minizyme. 2.2.2. Miniribozymes A miniribozyme with nucleotides of sequence 5’rGUUUUC joining A9 and Gt2 is shown m Fig. 3A. Maximal cleavage activity is conferred on the mmiribozyme when the single base-pair that replaces helix II is a G.C in the orientatron shown in the figure. The optimal sequence for the loop connecting the G and C has not been determined, but a flexible loop of 4 ribouridmes gives good activity, whereas a loop of three ribouridines has about 70% the effi- crency of the 4-ribouridme loop in cleaving short substrates (unpubhshed data). A loop of sequence 5’ UUUG had a cleavage rate constant about 10% that of the parent ribozyme (21). The mmnibozyme may consist solely of ribonucleotides, or it may be syn- thesized with a mixture of deoxyribonucleotides and ribonucleotides. A simple configuration for the DNA/RNA hybrid has DNA in all positions except for C3-A9 and Gi2-Ai5 2. The inclusion of DNA in the molecule reduces the costs of synthesis, and may also give a degree of protection against degradation in human serum (23) or in cells (24). 8 Hendry, Lockett, and McCall 3’ N N N N N PJ2A,u N N N N N N 5’ A cq~ A G & AGuA C-G 9 u u Miniribozyme uu 3’ N N N N N f&Am N N N N N N 5’ A kG A Gu AG”A T Gg Minizyme TTT Fig. 3. Schematic representatton of a mmmbozyme and a mmizyme Formally, the mmmbozyme has single G.C base pair replacing helix II, and the minizyme has no helix II The sequences of nucleotides shown joining As and G12 in the miniribozyme and mmizyme confer good cleavage activities on these molecules. The numbers of nucleotides m the 5’ and 3’ arms of the mmnibozyme and mmizyme may be larger than indicated here, without dtmnushing cleavage activity m vitro. All nucleotides, except for the conserved CJ to A9 and G12 to A15 1, may be either ribonucleotides or deoxyri- bonucleottdes. DNA-contammg mimribozymes with d(GTTTTC) m place of helix II and loop II, and with DNA hybridizing arms, cleave 13-mer substrates at approx 40% of the rate of analogous ribozymes with DNA in the hybridizing arms and m helix II (Table 3). In these examples, both the mmn-ibozymes and ribozymes form heltces I and III each of 6 base pairs m complex with the substrate. The same mimribozymes cleave 21-mer substrates about twofold faster than the analogous ribozymes (Table 3), and here they form a helix I and a helix III each of 10 base pairs m complex with the longer substrate. Of more relevance to biological applications, an all-RNA mmmbozyme cleaves an 809-nt RNA mol- ecule in vitro much faster than does an analogous all-RNA ribozyme (Fig. 4), Hammerhead Ribozymes Table 3 Cleavage Rate Constants for 13.Mer and Pi-Mer Substrates by DNA-Containing Miniribozymes (Mgttttc) and Ribozymes (FE) Substrate k&mm Mgttttc RC-lo/10 Ratio Mgttttc/RC TAT S 13 -6J6 0 175+0003 0.43 5 0 08 0.41 TAT S21-lo/10 0.9 Ik 0.2 0.45 * 0 09 20 Kr S13-6/6 0.59 f 0.09 1.6 2 0.3 0 37 Kr S21-lO/lO 3.0 zk 0.2 1.34 zk 0.08 22 Condltlons 10 mh4 MgCl,, 37”C, 50 mh4 Trls-HCl, pH 7 13 Sequences of nbozymes, with upper-case letters representing nbonucleotldes, and lower- case letters representing deoxyribonucleotides, are as follows. TAT Mgttttc: 5’ gtcctaggctCUGAUGAgtttttcGAAACttcctgga. TAT RC 5’ gtcctaggctCUGAUGAgtccttttggacGAAACttcctgga Kr Mgttttc 5’ ctccagtgtgCUGAUGAgttttcGAAACtcgcaaat Kr RC 5’ ctccagtgtgCUGAUGAgtccttttggacGAAACtcgcaaat and shows good activity over a much wider temperature range than does the ribozyme (Fig. 5). Thus, at least m the examples studied to date, minirlbozymes seem to be superior to full-size hammerheads m cleaving long transcripts in vitro. The relative rates at which mimribozymes cleave 13-mer and 21 -mer swb- strates indicates that unlike full-stzed ribozymes, mimrlbozymes are not hm- dered in their cleavage activity when forming extensive base pan-mg with the substrate, particularly with respect to helix I. This gives an advantage to mmiribozymes over ribozymes in that, for a target sequence of defined length, the distribution of nucleotides on either side of the cleavage site is not restricted in any way, whereas, for optimal cleavage by the ribozyme, the number of nucleotides in the 5’ arm is restricted to 5 or 6 (see Subheading 2.1.2.). There- fore the miniribozyme may be made with hybridizing arms of equal length, which may assist turnover. 2.2.3. Minizymes A minizyme is a hammerhead ribozyme in which helix II and loop II have been replaced by a short linker that contains no Watson-Crick base pairs (25’. Minizymes have been made with linkers consisting of nucleotides (20,25,26), or of short polymers of ethyleneglycol and phosphopropanedlol (17,19). The rates of cleavage by mimzymes increase as the number of atoms m the chain lmking Ag and G12 increases, with optimal rates being achieved with 25-31 atoms in the chain (27). For all-nucleotide linkers, this corresponds to 4 or 5 nucleotides conferring optimal activity. IO Hendry, Lockett, and McCall 60 time (hr) Fig 4. Rates of cleavage of an 809-nucleotlde interleukm-2 transcript by mterleukm-2 mimrlbozyme and ribozyme, at 37”C, 50 mA4 Tris-HCl, pH 8.0, 10 mA4 MgC12, with no heat pretreatment. The minmbozyme has the sequence 5’ r(GUUUUC) in place of helix II and loop II. Both mmiribozyme and ribozyme are made of RNA, and have 8 nucleotides in the 5’ hybridizing arms and 6 nucleotides in the 3’ arms Cleavage of the transcript occurs 82 nucleotldes from the 5’ end. The first mmizymes synthesized had linkers of sequence d(TTTT) or r(UUUU) (25). These minizymes were about loo-fold less-active than analo- gous standard ribozymes m cleaving short substrates of about 13 nucleotides (20-22). Like the mimrlbozymes however, symmetric (10 + 10) mmizymes actually improve in cleavage activity when cleaving 21-mer substrates, such that the mimzymes are typically only IO-fold less active than analogous ribozymes (22). Thus it appears that mmizymes, like miniribozymes, and unlike ribozymes, are not inhtbtted in their cleavage rates by the formation of long (>6 bp) hehces I. Against a 428-nucleotide RNA target derived from the HIV- 1 TAT coding sequence, minizymes with these lmkers cleaved faster than did the analogous ribozymes (22). Apparently, the small size and/or flexibility enjoyed by the ribozyme variants that lack or possess substantially truncated hehces II gives them some advantage over full sized ribozymes at the cleavage of long RNA transcripts. Recently, a mimzyme with the lmker of sequence d(GTTTT) has been described (271, and it is shown in Fig. 3. The inclusion of the G at the 5’ posi- [...]... with those of the plant catalytic domains, which are classified as either the hammerhead or hairpin ribozymes Moreover, although there are a number of examples of such subvn-al pathogens in plants, HDV is the only example of its kmd found in mammals Becauseof the unique properties of HDV, its catalytic domains might be better suited for the development of m truns-cleaving rtbozymes as human therapeutic. .. constramts m the sequence of helix 1 This was notably true m the first base pair followmg the *GUC All four nt were active m this position Note that a 10 bp helix 1 was designed, but a range of helix 1 nbozyme-substrate lengths were assayed m order to optimize the length of helix 1 By making a ribozyme capable of forming a range of lengths of helix 1 up to 10 bp, the range of hehx 1 lengths could be... about 4 bands faster than the T7 transcribed RNA of the same size Even though purificatton of the transcript was done on 8 M urea gels, we often saw sequence dependent variations in mobility of RNA In order to be certain of the identification of the RNA bands, it was therefore necessary to identify the correct band by direct RNA sequencing (12,13) The number of Cs m the substrate RNA transcript can be used... a sequence requirement of BN*GUC where the * is the site of cleavage (Fig 1) The nucleotide B is G, U, or C but not A Sequence searches were done for a number of systems,including HIV- 1, to identify sequences containing BN*GUC for use as possible target sites (2,5) Using HIV-I as an example, ribozymes were made to a number of potential targets and cleavage efficiency of the ribozymes to these targets... helix II of 4 base pairs The mimzyme, ribozyme, and vartous control oligonucletides were transfected at concentrations of 5, 10, and 20 l.&! into PBMN cells, for a period of 6-8 h, at which time the cells were stimulated to express IL2 by addition of PHA The levels of IL2 secreted into the supfmatant after 16 h were measured using both bio- and ELISA assays The mimzyme inhibited the production of interleukin-2... radloactivlty by Cherenkov counting Plcomoles (pmoles) of C nt m the sample was calculated* pmoles of C m sample = $1 in isolated transcript x [pmoles CTP added/(@ CTP added x decay factor)] (1) Pmoles of C m the sample was converted to pmoles RNA using the number of C residues in a given RNA transcript: pmoles of RNA transcript = pmoles C in sample/(moles C/mole of RNA) (2) Yields varied with different templates,... (1988) Cleavage ofspecific sites ofRNA by designed ribozymes FEBS Lett 239,285-288 6 Haseloff, J and Gerlach, W L (1988) Simple RNA enzymes with new and highly specific endoribonuclease activities Nature 334, 585-591 7 Tsuchihashi, Z., Khosla, M., and Herschlag, D (1993) Protein enhancement of hammerhead ribozyme catalysis Science 262,99-102 8 Bertrand, E L and Rossi, J J (1994) Facilitation of hammerhead... a Ribozyme DNA Template Sequence Start of transcription 3’ATTATGC TGAGTGATAT”CCCNNNNNNNNNNTCTTNNNNTGG TCTC TTTGTGTGCAACACCATATAATGGACCATS Transcription was done as above to make the desired hairpin rtbozyme We have tested activity of a number of ribozymes both with and without the S’GGG sequence prior to the sequence of the ribozyme, and have found the addition of GGG to the 5’ ribozyme termmus has... reaction proceeded without depletion of the 50 nt RNA component, and therefore was catalytic It had true Michaelis-Menten kmetics allowmg determmation of KM, bat, energy of activation, Mg2+ dependence, and the pH optima 1.2 The Two-Dimensional Structure of the Hairpin Ribozyme We determined the two-dimensional structure of the hairpin by making an extensive collection of site-specific mutants in both the... biologically relevant (3,4) The properties of HDV are remmiscent of certain subvlral plant pathogens, the virolds and the virusolds These properties include the small, singlestranded, RNA-only genomes, the apparent rolling-circle mechanism of From Methods in Molecular Edited by Medune, K J Scanlon Vol 11 Therapeutic 0 Humana 29 Press Applications Inc , Totowa, of Rlbozymes NJ 30 Tanner replication, and . number of nucleotides to the 3’ side of the cleavage site of the substrates and by changing the length of the 5’ hybridizing arms of the nbozymes. All Hendry, Lockett, and McCall Length of Helix. the size of the ribozyme. The advantages of minimizing the size of the ribozyme are several-fold. An obvious advantage, for ribozymes which will be used as exogenously-supplied therapeutics,. bonucleottdes. DNA-contammg mimribozymes with d(GTTTTC) m place of helix II and loop II, and with DNA hybridizing arms, cleave 13-mer substrates at approx 40% of the rate of analogous ribozymes with DNA