Secondary Structure for RNA Aptamers Binding to Guanine-Rich Sequence in the 5'-UTR RNA of N-Ras Oncogene

Guanine-rich sequences observed in the terminal segments of eukaryotic genomes form the four-stranded topology called G-quadruplexes.^1,2 They play a biological role in DNA telomere end and 5’-untranslated region (UTR) RNA in proximity to translational start sites. Especially, telomeres located at the ends of eukaryotic chromosomes play a crucial role in cellular aging and cancer.^3-7 In comparison with DNA quadruplexes, RNA quadruplexes have got less attention in spite of implication of their involvement at site of translational control. Guanine-rich sequences in the 5’-UTR of human genome, were identified.⁸ Sequence of 5’-GGGAGGGGCGGGUCUGGG-3’ containing four guanine-tracts in the 5’-UTR of oncogenic N-ras is thought to be one of G-quadruplex-forming elements in the 5’-UTRs of human genome.

To design the therapeutic RNAs inhibiting translation level of N-ras, we selected RNA aptamers interacting with the G-rich sequence in the 5’-UTR RNA of N-ras from an RNA pool containing 30 randomized nucleotides. Optimally predicted secondary structures of five selected RNA aptamers - 11-21-5, 11-21-30, 11-30-32, 11-30-36 and 11-48-2 - including primer sequence were found by the CLC RNA workbench ver. 4.2 program (www.clcbio.com). Their secondary structures were investigated with RNA structural probes such as RNase T1 which has specificity for a G in single-stranded region, RNase V1 specific for double strand and nuclease S1 specific for single strand,^9-12 which have been widely used for probing RNA structure in solution.^13-16 In this study, the determined secondary structures of five RNA aptamers were analyzed and compared. The new secondary structure model which can be commonly applied to them was proposed and characterized.

RNA molecules binding to the G-rich sequence in the 5’-UTR RNA which plays an important role in expression of N-ras, were selected from a pool (Fig. 1).^18,19 After 11th round of selection, the selected RNA aptamers were reverse-transcribed into cDNA, amplified by PCR and cloned into pGEM-T easy vector for sequece analysis. The sequences of the selected RNAs were shown in Fig. 2. The consensus sequences of the selected RNA aptamers were searched with CLUSTAL W (1.83) multiple sequence alignment. The consensus sequence GGGAUCCGCAUGCAAGCUUA were found between the selected RNAs.

Figure1.

(a) Guanine-rich sequence used for in vitro selection. (b) Scheme of experimental strategy for SELEX. RNA aptamers interacting with a G-rich sequence RNA were selected with G-rich sequence RNA-attached affinity column.

Figure2.

Sequences of the randomized region in selected RNAs. The consensus sequences are underlined in bold letters.

Secondary structures of selected RNA aptamers were optimally predicted with theoretical method like CLC RNA workbench ver. 4.2 program and supplemented with RNA structural probes. Phosphodiester bonds cleaved by RNA structural probes for analysis of secondary structures were shown in Table 1 and Fig. 3. Base G1 was conserved in the secondary structures of five RNA aptamers - 11-21-5, 11-21-30, 11-30-32, 11-48-2 and 11-30-36. Phosphodiester bonds of sequence G2-G-A-U-C6 in an RNA aptamer 11-30-36, C6-C7 in an RNA aptamer 11-30-32 and C7-G8 in two RNA aptamers - 11-30-32 and 11-48-2 were cleaved by RNase V1. Phosphodiester bonds of G33-G34 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2, G34-A35 in an RNA aptamer 11-30-32, C37-C38 in two RNA aptamers - 11-30-32 and 11-48-2, and C38-G39 in an RNA aptamer 11-30-36 were also cleaved by RNase V1. These cleavage results by double-strand specific RNase V1 suggested that two sequences G2GAUCC7 and G33GAUCC38 formed stable base pairs. Bases in this double strand region were well conserved in five RNA aptamers.

Figure3.

Secondary structure for RNA aptamers which bind to the G-rich sequence in the 5’-UTR RNA of N-ras. Squares indicate nucleotides on double helical regions that form base-pairs and circles indicate nucleotides on single-strand regions. Two nucleotides in squares show that one of them can occur on this position. Arrows indicate the sites cleaved by each enzyme.

Base G8 existed as a bulge in three RNA aptamers - 11-21-5, 11-21-30 and 11-30-36 because U32 was deleted. But this base was thought to pair with U32 to form G8:U32 base pair in two RNA aptamers - 11-30-32 and 11-48-2.

Phosphodiester bond of C9-A10 in an RNA aptamer 11-30-32 and two bonds of sequence A10-U-G12 in an RNA aptamer 11-48-2 were cleaved by RNase V1. And phosphodiester bonds of U30-G-U32 in an RNA aptamer 11-30-36 were also digested by RNase V1. These cleavage results by RNase V1 meant that two sequences C9AU11 and G29UG31 formed base pairs in four RNA aptamers - 11-21-5, 11-21-30, 11-30-36 and 11-48-2. But A10U11 was thought to exist as a bulge in an RNA aptamer 11-30-32 because the counterpart sequence G29U30 was deleted.

Phosphodiester bond of G12-C13 in an RNA aptamer 11-30-36 was cleaved by RNase T1, the same bond in an RNA aptamer 11-30-32 was cleaved by both RNase T1 and RNase V1, and the same bond in an RNA aptamer 11-48-2 was cleaved by both RNase T1 and nuclease S1. Bond of A27-C28 in an RNA aptamer 11-30-36 was cleaved by RNase V1. These results suggested that base G12 paired with U28 to form G12:U28 base pair in two RNA aptamers - 11-21-5 and 11-21-30, and with C28 to form G12:C28 base pair in an RNA aptamer 11-30-32. And base C13 paired with G27 to form C13:G27 base pair in three RNA aptamers - 11-21-5, 11-21-30 and 11-30-32. Bases A14 and C26 existed as bulges in these three RNA aptamers. But two sequences G12CA14 and C26GU28 were thought to exist as single strand regions in two RNA aptamers - 11-30-36 and 11-48-2.

Phosphodiester bond of A15-G16 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by RNase V1 and bond of G16-C17 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by RNase T1. Bond of G23-C24 in two RNA aptamers - 11-30-36 and 11-48-2 was weakly digested by RNase T1. So sequence A15GC17 was thought to pair with G23CC25 in two RNA aptamers - 11-30-36 and 11-48-2, with G23CU25 in an RNA aptamer 11-21-5, with G23UC25 in an RNA aptamer 11-21-30, and with A23CU25 in an RNA aptamer 11-30-32 to form the double strand region. But base pair C17:G(A)23 was thought to be unstable according to the result of accessibility of RNase T1 to this base pair.

Phosphodiester bond of U18-U19 in two RNA aptamers - 11-30-32 and 11-30-36 was cleaved by both RNase V1 and nuclease S1, bond of U19-A20 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by both RNase V1 and nuclease S1, bond of A20-C21 in two RNA aptamers - 11-30-32 and 11-30-36 was cleaved by both RNase V1 and nuclease S1, and bond of C21-A22 in an RNA aptamer 11-48-2 was cleaved by nuclease S1. Seeing that phosphodiester bonds in U18UANN22 were cleaved by single-strand specific nuclease S1, this sequence U18UANN22 was thought to be located in single strand loop. And it was thought that bases in this loop must be stacked from intramolecular interaction because they were also susceptible to double-strand specific RNase V1. The size of the single strand loop was composed of four or five nucleotides. The single strand loop of an RNA aptamer 11-30-32 had sequence of UUAC, the loop of an RNA aptamer 11-21-30 had sequence of UUAG, the loop of two RNA aptamers - 11-21-5 and 11-30-36 had sequence of UUACU, and the loop of an RNA aptamer 11-48-2 had sequence of UUACA.

Phosphodiester bond of G39-C40 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by RNase T1, bond of C40-A41 in two RNA aptamers - 11-30-32 and 11-30-36 was cleaved by both RNase V1 and nuclease S1, bond of A41-U42 in two RNA aptamers - 11-30-32 and 11-30-36 was cleaved by both RNase V1 and nuclease S1, bond of U42-G43 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by both RNase V1 and nuclease S1, bond of G43-C44 in three RNA aptamers - 11-30-32, 11-30-36 and 11-48-2 was cleaved by both RNase T1 and nuclease S1, and bond of G47-C48 in an RNA aptamer 11-48-2 was cleaved by RNase T1. So sequence G39CAUGCAAGCUU50 was thought to be a single strand region. And the bases in this single strand region must be stacked from intramolecular interaction because they were also susceptible to RNase V1.

From above cleavage results with theoretically predicted model, the generalized secondary structure for RNA aptamers which bind to the G-rich sequence in the 5’-UTR RNA of N-ras was proposed in Fig. 3. It had a central long double strand region flanked by single strand region at both end sides. We anticipate that this model can be employed to design the candidates of therapeutic agents inhibiting translation level of N-ras oncogene. Footprinting assay will be applied to get the information for contact points between G-rich RNA ligand and aptamer RNA. And binding assay will also be performed to know whether aptamer RNA does not bind to UTR of other genes but UTR of N-ras.

In summary, we proposed the secondary structure model for RNA aptamers which bind to the G-rich sequence in the 5’-UTR RNA of N-ras. The model was composed of a central long double strand region flanked by single strand region at both end sides. The double strand region had an internal single-strand region and bulges. The single strand loop in the right side was composed of four or five nucleotides. The bases in this model were considerably conserved. This secondary structure model is expected to be used to give the information necessary to get higher order structure with X-ray crystallography or NMR (nuclear magnetic resonance) spectroscopy.