Intracellular mRNA cleavage by 3′ tRNase under the direction of 2′-O-methyl RNA heptamers

Abstract

Mammalian tRNA 3′ processing endoribonuclease (3′-tRNase) can cleave any RNA at any site under the direction of small guide RNA (sgRNA) in vitro. sgRNAs can be as short as heptamers, which are much smaller than small interfering RNAs of ∼21 nt. Together with such flexibility in substrate recognition, the ubiquity and the constitutive expression of 3′-tRNase have suggested that this enzyme can be utilized for specific cleavage of cellular RNAs by introducing appropriate sgRNAs into living cells. Here we demonstrated that the expression of chloramphenicol acetyltransferase can be downregulated by an appropriate sgRNA which is introduced into Madin–Darby canine kidney epithelial cells as an expression plasmid or a synthetic 2′-O-methyl RNA. We also showed that 2′-O-methyl RNA heptamers can attack luciferase mRNAs with a high specificity and induce 3′-tRNase-mediated knock-down of the mRNAs in 293 cells. Furthermore, the MTT cell viability assay suggested that an RNA heptamer can downregulate the endogenous Bcl-2 mRNA in Sarcoma 180 cells. This novel sgRNA/3′-tRNase strategy for destroying specific cellular RNAs may be utilized for therapeutic applications.

INTRODUCTION

In mammalian cells, tRNA 3′ processing endoribonuclease (3′-tRNase) is an essential enzyme to remove 3′ trailers from pre-tRNAs (1–7). This enzyme is unique in that it possesses the property of being able to cleave any RNA at any site under the direction of small guide RNA (sgRNA) in vitro like an RNA-induced silencing complex underlying RNA interference (RNAi) (8–12). For instance, 3′-tRNase works as the 4-nt-recognizing RNA cutter RNase 65 by forming a complex with a 3′-truncated tRNA (8). Partial human immunodeficiency type 1 (HIV-1) RNA targets can be cleaved site specifically by the enzyme when the targets form pre-tRNA-like structures with the aid of appropriate 5′-half-tRNAs (9). Surprisingly, the enzyme can recognize a lin-4–lin-14 RNA complex of Caenorhabditis elegans and cleave the lin-14 mRNA, although we do not know yet that this activity occurs in the cells (10). Together with such flexibility in substrate recognition, the ubiquity and the constitutive expression of 3′-tRNase have suggested that this enzyme can be utilized for specific cleavage of cellular RNAs by introducing appropriate sgRNAs into living cells.

RNA heptamers, when they form acceptor stem-like duplexes with their targets through base pairing, can also direct specific cleavages of target RNAs by 3′-tRNase in vitro as efficiently as the original 5′-half-tRNAs (11). In this case, however, the target sites are restricted to immediate downstream regions of stable hairpin structures resembling the T stem–loop. This is advantageous because a heptamer can direct efficient RNA cleavage with a specificity of an ∼12-nt sequence and not merely a 7-nt sequence due to the need for a stable hairpin structure (11).

The use of RNA heptamers to knock-down specific mRNAs has additional advantages over long RNAs such as conventional antisense RNAs (13), ribozymes (14) and small interfering RNAs (siRNAs) (12). First, heptamers are much easier and cheaper to synthesize than long ones. Secondly, cells can take up RNA heptamers relatively easily without any stimulating reagents. When we incubated human 293 cells in culture medium containing 10% fetal bovine serum (FBS) with fluorescein-labeled 2′-O-methyl heptamers for 8 h without cationic reagent, the fluorescence was detected in most of the cells (unpublished data).

Small RNAs (such as 5′-half-tRNAs and RNA heptamers) that can direct RNA cleavage by 3′-tRNase are referred to as sgRNAs. In this study, we demonstrate that intracellular target mRNAs can be specifically cleaved presumably by endogenous 3′-tRNase under the direction of appropriate sgRNAs which are introduced into the cells as expression plasmids or synthetic 2′-O-methyl RNAs.

MATERIALS AND METHODS

Plasmid constructions

A synthetic double-stranded DNA containing a tRNA^Arg promoter, an RNA polymerase III terminator and an effector sequence (sgCAT, sgCATM or antiCAT) was cloned between the EcoRI and HindIII sites of pBluescript SK+ (Stratagene) to generate the effector expression plasmids psgCAT, psgCATM or pantiCAT. The synthetic sense strand DNA sequences for psgCAT and pantiCAT are 5′-AATTCGTACGTCGCAGGAATTC (ctcaacc for psgCATM) TGGCGCAATGGATAACGCGGCAAG (gaatg for psgCATM) CTACGGAT CAGAAGATTCCAGGTTCGACTCCTGGCTGGCTCGTTTTA-3′ and 5′-AATTCGTACGTCGCACTCAACCTGGCGCAATGGGGAATTCCGGATGAGCATTCATCAGAGATTCCAGGTTCGACTCCTGGCTGGCTCGTTTTA-3′, respec tively. The regions complementary to the CAT mRNA are underlined.

To construct the 5′-modified luciferase expression plasmids p5LucWT, p5LucM1 and p5LucM2, we inserted synthetic double-stranded DNAs containing the DNA sequences WT, M1 and M2, respectively, into the NcoI site (immediately before the initiation codon) of the pGL3-control vector (Promega). The 3′-modified luciferase expression plasmids p3LucWT, p3LucM1 and p3LucM2 were generated by cloning the synthetic double-stranded DNAs containing the DNA sequences WT, M1 and M2, respectively, into the XbaI site (immediately after the stop codon) of the pGL3-control vector. The synthetic sense DNA containing the WT sequence (indicated by underlining) is 5′-TCAAAGCTTCTAGACCATGGCC (aa in M2) AGGTTCGACTCCTGGCTG (c in M1) GC (a in M1) TCGGTGCCATGGTCTAGAAGCTTTGA-3′.

2′-O-methyl RNA synthesis for in vivo assays

The following 5′-phosphorylated 2′-O-methyl sgRNAs were synthesized by Nippon Bioservice: sgCAT, 5′-pGGAAUUCUGGCGCAAUGGAUAACGCGGCAAG-3′; sgCATM, 5′-pCUCAACCUGGCGCAAUGGAUAACGCGGAAUG-3′; antiCAT, 5′-pGGAAUUCCGGAUGAGCAUUCAUCAG-3′; CAT7, 5′-pGGAAUUC-3′; Hep1, 5′-pGGGCCAG-3′; Hep2, 5′-pGAUCGAG-3′; and Bclhep, 5′-pGGGGGCA-3′.

DNA transfection, CAT assays and luciferase assays

The Madin–Darby canine kidney epithelial (MDCK) cells or 293 cells were plated at a density of 1 × 10⁵ cells/ml on 12-well (for CAT assay) or 24-well (for luciferase assay) dishes in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS. After incubation at 37°C for 24 h, the cells were co-transfected with a reporter (CAT or luciferase) expression plasmid, an effector expression plasmid or MeRNA, and the pSV-β-galactosidase control vector (Promega) using Lipofectamine 2000 (Invitrogen). The cells were further cultured at 37°C for 48 h unless specified. The CAT protein amounts were measured using a CAT enzyme-linked immunosorbent assay (ELISA) kit (Roche) (15). The luciferase activity was quantitated with a PicaGene kit (Toyo-inks) using the luminescence reader BLR201 (Aloka) (15). The CAT activity and the luciferase activity were normalized against the amount of the β-galactosidase activity.

RNA isolation and northern blot analysis

Total cellular RNA was prepared 24 h after transfection by the acid guanidinium thiocyanate–phenol–chloroform extraction method (16), and northern blot analysis was performed as described (17). The intensity of RNA bands was quantitated using the software ImageQuant (Molecular Dynamics), and the reporter mRNA levels were normalized against the internal control mRNA levels.

MTT cell viability assays

Sarcoma 180 cells were plated at a density of 2 × 10⁵ cells/ml on 24-well dishes in 0.4 ml of DMEM supplemented with 10% FBS. After incubation at 37°C for 24 h, the cells were transfected with the MeRNA heptamer Hepbcl using Lipofectamine 2000 and cultured in the absence or presence of human recombinant hepatocyte growth factor (HGF) (a generous gift from T. Ishii) for a further 72 h. After the incubation, the MTT assay was performed as described (18).

RNA synthesis for in vitro analysis

The target RNA SPH2 was synthesized with T7 RNA polymerase (Takara Shuzo) from a synthetic double-stranded DNA containing a T7 promoter (11). The transcription reaction was carried out in the presence or absence of [α-³²P]UTP (Amersham Pharmacia Biotech) under the conditions specified by the manufacturer (Takara Shuzo). The synthesized SPH2 was gel purified before assays.

Chemically modified 5′-phosphorylated RNA heptamers, 2′-O-methyl RNA (MeRNA), phosphorothioate RNA (SRNA), DNA and phosphorothioate DNA (SDNA) together with unmodified RNA were synthesized with a DNA/RNA synthesizer and purified by high-performance liquid chromatography before various assays. The synthesis of these heptamers was carried out by Nippon Bioservice. 5′-Biotinylated SPH2 was synthesized and purified by Xeragon.

BIACORE kinetic assays

To obtain dissociation constants (K_d) of SPH2–heptamer complexes, we determined association (k_a) and dissociation (k_d) rate constants by using surface plasmon resonance biosensors of a BIACORE3000 instrument. Kinetic assays were performed basically according to the instructions supplied by BIACORE Inc. About 2700 resonance units (RU) of 5′-end-biotinylated SPH2 were non-covalently immobilized on flow cell 2 of Sensor Chip SA, which contained streptavidin covalently immobilized on its carboxymethylated dextran matrix. Several different concentrations (100–2000 nM) of unmodified or chemically modified RNA heptamers were tested for kinetic constants. Each sample in a buffer containing 10 mM Tris–HCl pH 7.5 and 3.2 mM MgCl₂ was injected at 20 µl/min through flow cell 1 (as reference) and flow cell 2 in the instrument equilibrated at 25°C. After a 120 s injection, both flow cells were washed with the same buffer without heptamers to monitor duplex dissociation. The surface of Sensor Chip SA was regenerated by washing with 6 M urea before each assay.

BIACORE sensorgrams obtained through the kinetic assays were analyzed with the computer software, BIAevaluation version 3.0 (BIACORE Inc.), to calculate k_a and k_d values for each heptamer. The data were subjected to global fitting based on a one-to-one binding model, and the fittings were sufficiently good judging from small χ² values (0.13–0.41), which represent deviations from fitted curves.

Kinetic analysis for 3′-tRNase cleavage

The 3′-tRNase cleavage of a complex of SPH2 with each heptamer was examined at various concentrations of SPH2. A reaction mixture (6 µl) contained 10 mM Tris–HCl pH 7.5, 1.5 mM dithiothreitol, 3.2 mM MgCl₂, 10 µM RNA heptamer and 0.17–3.3 µM SPH2. After pre-incubation at 25°C for 10 min, the reactions were started by adding the pig 3′-tRNase fraction (1–5 ng) after glycerol gradient centrifugation (4), and continued at 25°C for 5 min. The reaction products were resolved on a 10% polyacrylamide–8 M urea gel and then quantitated with a PhosphorImager (Molecular Dynamics). Actual concentrations of SPH2–heptamer complexes were calculated using the K_d value from the BIACORE analysis. K_m and V_max values were obtained from the best-fit line on a Lineweaver–Burk plot.

5′-End-labeling of heptamers and nuclease stability assays

The 5′-phosphates of unmodified or chemically modified RNA heptamers were removed with bacterial alkaline phosphatase (Takara Shuzo). After the reaction, the heptamers were extracted with phenol, precipitated with ethanol, and redissolved in water. Each heptamer was 5′-end-labeled with [γ-³²P]ATP (Amersham Pharmacia Biotech) using T4 polynucleotide kinase (Takara Shuzo), and purified on a 20% polyacrylamide–8 M urea gel.

The ³²P-labeled heptamers (1 pmol) were assayed for nuclease stability by incubating them with 10 µl of DMEM containing 10% FBS at 37°C for various time periods. After the incubation, the products were analyzed on a 20% polyacrylamide–8 M urea gel, and quantitated with a PhosphorImager (Molecular Dynamics).

RESULTS AND DISCUSSION

Downregulation of CAT expression by an sgRNA expressed under the control of the tRNA^Arg promoter

First of all, we investigated whether sgRNAs can direct specific mRNA cleavage in living cells by introducing an expression plasmid encoding a 5′-half-tRNA. We tested the CAT mRNA (GenBank accession No. X65321) for 3′-tRNase-specific cleavage in MDCK cells. Because the sgRNA sgCAT36 has been shown to direct CAT mRNA cleavage in vitro after the 222nd nucleotide U from the initiation codon (10), we constructed the pBluescript-based plasmid psgCAT that produces the transcript containing the sgCAT36 sequence under the control of the tRNA^Arg promoter (Fig. 1A). We also generated two control plasmids, psgCATM and pantiCAT, which produce the transcripts containing 7-nt substitutions in the CAT mRNA-binding regions and a conventional 25-nt antisense sequence, respectively.

Figure 1.

sgRNA-directed CAT mRNA knock-down in MDCK cells. (A) A plausible secondary structure of the CAT mRNA–sgCAT complex. An arrow indicates the 3′-tRNase cleavage site expected from the in vitro assay (10). The nucleotide numbers of the mRNA start at the first letter of the initiation codon. (B) Relative CAT protein levels in MDCK cells. The cells were co-transfected with pcDNA3/CAT (0.2 µg/well) and an effector expression plasmid (2 or 4 µg/well) or a 2′-O-methyl RNA (1 or 2 µM). pBluescript SK+ or 2′-O-methyl CAT7, which did not direct CAT mRNA cleavage in vitro, was used as a control. Data are the means ± SD of six independent experiments. (C) Northern blot analysis for CAT mRNA. Total RNA was extracted from the cells co-transfected under the same conditions as above. The level of GAPDH mRNA was used as an internal control.

After co-transfection of MDCK cells with the CAT expression vector pcDNA3/CAT (Invitrogen) and one of the effector plasmids, we measured the amount of CAT protein in cell lysates. The CAT protein levels were significantly reduced in a dose-dependent manner, and the degree of reduction was higher than that in pantiCAT-transfected cells, while the transfection with psgCATM did not reduce the CAT protein level (Fig. 1B). The CAT mRNA levels were also decreased in accordance with the protein levels (Fig. 1C). By northern blot analysis, we confirmed that the transcripts containing sgCAT, sgCATM and antiCAT sequences were equivalently expressed in the cells (data not shown). In spite of less stable binding to the mRNA, the downregulation by the sgRNA was more efficient than that by the conventional antisense RNA. It is likely that this is because the mRNA can be cleaved by 3′-tRNase under the direction of the sgRNA and subsequently degraded beyond the conventional antisense effect. Similar effects of the co-transfection were observed in 293 cells (data not shown). These results suggest that our sgRNA/3′-tRNase RNA cleavage method really works in living cells.

3′-tRNase RNA cleavage in vitro under the direction of chemically modified sgRNAs

When we introduce synthetic RNA heptamers into living cells, we need to make the heptamers more stable than natural RNAs to avoid their degradation by cellular nucleases. Various chemically modified RNA heptamers such as MeRNA, SRNA, DNA and SDNA together with unmodified RNA were tested for their efficiency to direct the specific RNA cleavage and for nuclease stability. The 3′-half portion of human tRNA^Arg SPH2 was used as an RNA target, which is well characterized and is known to be cleaved efficiently by 3′-tRNase under the direction of the unmodified RNA heptamer (11). The efficiency was examined by measuring K_d values for heptamer–target complexes and V_max/K_m values for 3′-tRNase cleavage, and the stability was investigated by estimating their half-lives in culture media.

The K_d values for MeRNA and SRNA were smaller and slightly larger, respectively, than that for unmodified RNA, and the K_d values for DNA and SDNA were much more elevated (Table 1). The V_max/K_m values differed depending on heptamers, and that for MeRNA was second best behind unmodified RNA (Table 1). MeRNA and SDNA were much more stable against nucleases than unmodified RNA, SRNA and DNA (Table 1). These results suggest that the MeRNA heptamer would greatly exceed the others in guiding 3′-tRNase RNA cleavage in living cells.

Table 1. Kinetic parameters and guiding efficiency of modified heptamers.

Heptamer pGGGCCAG	kaa (1/Ms)×104	kda (1/s) ×10–3	Kd (µM)	Kmb (µM)	Vmaxb (pmol/min)	Relative Vmax/Km/Kd	t1/2c (h)	Relative t1/2× (Vmax/Km/Kd)
RNA	1.8	8.4	0.47	0.39	1.9	100	<0.5	<19
MeRNA	2.0	7.7	0.39	0.50	1.8	89	∼3	∼100
SRNA	1.8	8.8	0.49	0.66	1.8	53	<0.5	<10
DNA	1.9	14.7	0.77	1.08	1.8	21	<0.5	<4
SDNA	1.9	16.2	0.85	3.87	1.8	6	∼1.3	∼3

^aThe k_a and k_d values for the interactions between the target SPH2 and heptamers (11) were obtained by global fitting of BIACORE data.

^bThe kinetic parameters were determined as described (11). Each measurement was from the averages of three trials with an SD of 5–12%. The maximum velocity per nanogram of pig 3′-tRNase fraction after glycerol gradient centrifugation (2) is shown.

^cThe ³²P-labeled heptamers (1 pmol) were assayed for nuclease stability by incubating them with DMEM containing 10% FBS at 37°C for 0.5, 1.5 or 3.0 h. Half-lives were estimated from these data.

Knock-down of the cellular CAT mRNA by a 2′-O-methyl sgRNA

Before examining the efficacy of MeRNA heptamers in the cells, we tested 2′-O-methyl 5′-half-tRNAs for their guiding ability. We co-transfected MDCK cells with pcDNA3/CAT and synthetic 2′-O-methyl sgCAT, sgCATM or antiCAT. The CAT protein levels decreased up to ∼45% in a dose-dependent manner in 48 h in the presence of 2′-O-methyl sgCAT (Fig. 1B). While 2′-O-methyl sgCATM did not show a meaningful effect, 2′-O-methyl antiCAT decreased CAT protein levels up to ∼70%. This decrease would be not due to an RNAi effect but to a simple antisense effect, because MeRNA cannot induce RNAi (19). The CAT mRNA level was reduced to ∼70% in the presence of 2′-O-methyl sgCAT (Fig. 1C). Similar effects of the co-transfection were observed in 293 cells (data not shown). These results suggest that a 2′-O-methyl 5′-half-tRNA can direct specific mRNA cleavage and downregulate the protein level in mammalian cells.

2′-O-methyl heptamers downregulate the luciferase expression

We next challenged targeting modified firefly luciferase mRNAs (GenBank accession no. U47296) using synthetic MeRNA heptamers in the cells. Based upon the luciferase expression vector pGL3-control, we constructed its six derivatives, each of which can express a luciferase mRNA containing one of three different target sequences (WT, M1 and M2) of 3′-tRNase (11) in either the 5′- or 3′-untranslated region (UTR, Fig. 2A). These target sequences are based on the well-characterized SPH2. The 5′-modified luciferase mRNAs from the resultant plasmids p5LucWT, p5LucM1 and p5LucM2 have single 7-nt sequences complementary to the 2′-O-methyl heptamers Hep1, Hep2 and Hep1, respectively (Fig. 2A). In the same fashion, the 3′-modified luciferase transcripts from the plasmids p3LucWT, p3LucM1 and p3LucM2 contain single heptamer-binding sites. Because the WT and M1 sequences can fold to form 5-bp stem–loops, the complexes of Hep1–WT and Hep2–M1 can form co-axially stacked perfect 12-bp stem–loops. Hep1–M2 can form a similar stem–loop with two mismatches due to the presence of 2-nt substitutions in M2. Hep1–M1, Hep2–WT and Hep2–M2 can also form similar stem–loops containing two, two and four mismatches, respectively. Judging from the previous observations in in vitro 3′-tRNase assays (10,11), only Hep1–WT and Hep2–M1 complexes could be cleaved efficiently by endogenous 3′-tRNase.

Figure 2.

2′-O-Methyl heptamer-guided cleavage of modified luciferase mRNAs in 293 cells. (A) The 5′- or 3′-modified luciferase mRNA and secondary structures of the luciferase mRNA–heptamer complexes. The heptamers are shown in lower case letters. Arrows denote the potential 3′-tRNase cleavage sites. (B) Relative luciferase activities. The cells were co-transfected with 1 µg/well of p5LucWT, p5LucM1, p5LucM2, p3LucWT, p3LucM1 or p3LucM2 together with 1 µM CAT7 (as a control), Hep1 or Hep2. (C) Heptamer dose-dependent downregulation of the luciferase activity from p5LucWT (1 µg/well). (D) Time course of the luciferase activity. 293 cells were co-transfected with 1 µg/well of p5LucWT and 1 µM heptamer, and incubated for the indicated time periods. (E) Northern blot analysis for luciferase mRNA. The GAPDH mRNA level was used as an internal control. Data are the means ± SD of six independent experiments.

We co-transfected 293 cells with one of these plasmids together with one of the heptamers. After incubation for 48 h, the cells were harvested and luciferase activities were measured. The unrelated heptamer CAT7 was used as a control. As expected, significant reduction of the luciferase activity was observed only in the cells where the Hep1–WT or Hep2–M1 complex can be formed in either the 5′- or 3′-UTR of the luciferase mRNA (Fig. 2B). The cellular luciferase activity was not significantly affected by heptamers that form imperfect stem–loop complexes with the mRNA (Fig. 2B). Hep1 started to downregulate the cellular luciferase activity from the 5′-WT-modified mRNA at the concentration of 0.2 µM, and continued to reduce the activity up to ∼35% at 2 µM (Fig. 2C). The downregulation by 1 µM Hep1 was observed in 12 h and maintained for at least 48 h (Fig. 2D). We confirmed that the luciferase mRNA levels were reduced in the cells co-transfected with p5LucWT and Hep1, and with p5LucM1 and Hep2 to ∼55 and ∼60%, respectively (Fig. 2E). These results suggest that the MeRNA heptamers can attack target mRNAs with a high specificity and induce 3′-tRNase-mediated cleavage of the mRNA in living cells.

An RNA heptamer knocks-down the endogenous Bcl-2 mRNA

Furthermore, in order to examine whether an RNA heptamer can direct endogenous mRNA cleavage by 3′-tRNase, we selected the Bcl-2 mRNA as a target. It is known that downregulation of Bcl-2 induces apoptosis (20,21) and that human recombinant HGF induces the apoptosis signaling pathway in Sarcoma 180 cells (18). We transfected Sarcoma 180 cells with the 2′-O-methyl heptamer Bclhep, which is designed to bind immediately downstream of a potential T-stem–loop-like hairpin structure, to the 126th to 132nd nucleotide from the initiation codon of the Bcl-2 mRNA (GenBank accession No. M16506) and to induce 3′-tRNase cleavage. The introduction of Bclhep caused a reduction in the viability of Sarcoma 180 cells in the presence or absence of HGF up to ∼50% (Fig. 3). This observation supports the efficacy of the heptamer/3′-tRNase strategy in the cells.

Figure 3.

Apoptosis induction by an RNA heptamer. Sarcoma 180 cells were transfected with 1 µM of the 2′-O-methyl heptamer Bclhep or Hep2 (as a control), and incubated with the indicated amount of HGF. Suppression of cell viability was determined by the MTT assay. Data are the means ± SD of six independent experiments.

Heptamer potentiality for therapeutic agents

Here we showed that intracellular target mRNAs can be specifically downregulated under the direction of appropriate sgRNAs which are introduced into the cells as expression plasmids or synthetic 2′-O-methyl RNAs. Although our data suggest that the observed downregulation is attributed to specific mRNA cleavage by endogenous 3′-tRNase, further experiments would be needed to corroborate this mechanism. Experiments to see if the sgRNA-directed downregulation can be canceled after knocking-down the 3′-tRNase expression by siRNAs or ribozymes would be very useful.

The number of the total human genes has been estimated to be ∼32 000 (22), while the estimated average length of human mRNAs excluding poly(A) tails is ∼2500 nt (23). From these values, the whole human mRNA sequence space can be calculated to be ∼80 million nucleotides. In theory, the frequency of appearance of potential heptamer-directed cleavage sites is 1/4¹², or once every ∼17 million nucleotides, resulting in the appearance of roughly five potential mRNA cleavage sites per heptamer. In the experiment of in vitro 3′-tRNase cleavage of the CAT mRNA, we have not detected cleavage at about four-fifths of the selected potential target sites probably due to tight RNA folding to refuse access to sgRNAs (data not shown). Therefore, we can expect that one specific heptamer may be able to direct cellular mRNA cleavage at one unique site among the ∼80 million nucleotide sequence space. Because 4⁷, or 16 384, heptamers of different sequences can be generated, about half of the human mRNAs could be targeted under the direction of heptamers. Hopefully, a significant number of the heptamers could be selected as therapeutic agents.