Biochemical and Biophysical Research Communications 291, 925–931 (2002)

doi:10.1006/bbrc.2002.6521, available online at http://www.idealibrary.com on

Selection and Stabilization of the RNA Aptamersagainst the Human ImmunodeficiencyVirus Type-1 Nucleocapsid Protein

Se Jin Kim,*,† Mee Young Kim,* Jae Ho Lee,† Ji Chang You,†,1 and Sunjoo Jeong*,1,2

*Department of Molecular Biology, Dankook University, Seoul 140-714, Korea; and †Department of Pathology,Catholic University School of Medicine, Seoul 137-701, Korea

Received February 1, 2002

The nucleocapsid (NC) protein of the human immu-nodeficiency virus-1 (HIV-1) plays an important role inthe encapsidation of viral RNA and assembly of viralparticle. Since the NC protein is resistant for muta-tion, it might be an excellent target for the anti-viraltherapy. RNA aptamers that bind to the mature formof the NC protein were isolated from a RNA library.Surface plasmon resonance measurement and gel shiftassay showed that the RNA aptamers specifically bindto the NC protein with high affinity and compete forthe psi RNA binding to the NC protein. Mapping of theRNA aptamer showed at least two sites for the proteinbinding, suggesting a multiple and cooperative bind-ing by the NC to RNA. In addition, the circular form ofRNA avidly binds to the NC protein as a linear counterdoes. Stabilized RNA aptamer is expected to act asan inhibitor for the viral packaging. © 2002 Elsevier

Science (USA)

Key Words: HIV-1; nucleocapsid; RNA aptamer; cir-cular RNA; packaging.

The nucleocapsid (NC) protein of Human Immuno-deficiency Virus type-1 (HIV-1) plays critical roles invarious steps in viral life cycles, such as viral replica-tion, encapsidation and assembly (1). NC protein spe-cifically recognizes the viral genomic RNA via specificbinding to the viral psi (�) sequences, thus selectivelyencapsidates the viral RNA among many different cel-lular RNA pool (2–3). The �-sequence has been shownto form four stem-loop structures, which is denotedSL1 through SL4. Structural studies revealed the crit-ical interactions between the zinc finger motifs of theNC protein and specific nucleotides in the �-sequences

1 Both labs contributed equally to the paper.2 To whom correspondence should be addressed at Department of

Molecular Biology, Dankook University, Hannam-dong san 8,Yongsan-ku, Seoul 140-714, Republic of Korea. Fax: 82-2-793-0176.E-mail: sjsj@dankook.ac.kr.

925

the �-sequences is structured as a short RNA mole-cules and how it interacts with the NC proteins; how-ever, these studies have provided little informationon the general recognition pattern between whole�-sequences and the protein. In fact, the geneticmutagenesis studies indicate that all four of thestem-loops might be needed for the efficient encapsida-tion (3).

In vitro selection using a large pool of RNA moleculeshas identified high affinity RNA ligands for varioustarget proteins, which is generally named as aptamers(8–9). Thus, RNA aptamers for the NC protein couldprovide mechanistic insights for the protein-RNA in-teraction and, if any, common denominators for the NCprotein binding RNA molecules. Such selected RNAmolecules could also be developed as decoy moleculesthat would inhibit the specific interaction between theviral genomic RNA and the NC proteins. In fact, theNC protein is an extremely attractive target for anti-viral therapeutics, because it is known to have very lowor no mutation permissive nature unlike other viralproteins (10–12). A couple of RNA aptamers were re-ported either for the gag polyprotein (13) or for the NCprotein (14, 15). However, none of these studies weredone by using the fully mature form (55 amino acids) ofthe NC as a target for the selection. In addition, theyemployed short RNA molecules with 30–50 nucleotidesof random sequences that were likely to form only asingle prominent secondary structure with lower diver-sity. Thus, it would be advantageous to use longer sizeof the RNA molecules that might form diverse andstable structure with more than one binding sites forthe NC protein, because a multiple binding is likely tooccur during viral encapsidation (6–7).

In this study, we identified stable RNA aptamersthat bind to the mature form of NC protein by in vitroselection with longer size of RNA with 70 nucleotides ofrandom sequences. Binding affinity was measured by

(4–7). These studies suggested how a small portion of

0006-291X/02 $35.00© 2002 Elsevier Science (USA)All rights reserved.

Surface Plasmon Resonance (SPR) technique, demon-strating nanomolar range of Kd in most of the selectedaptamer molecules. Selected RNA was found to com-pete for the �-RNA binding to the NC protein, asexpected. Comparison of the selected sequences withother previously identified NC binding RNA revealsinteresting common sequence blocks with minor differ-ences. Multiple binding sites seem to be required forthe efficient binding, suggesting a cooperative bindingof the NC protein to the RNA. Selected RNA aptamerswere tailored to circular RNA for more stabilized formof RNA, which were also found to have high bindingaffinities to the NC protein. Therefore, stabilized RNAaptamer for the NC protein could provide a valuablelead molecule for the anti-HIV drug development.

MATERIALS AND METHODS

Preparation of the RNA library. The DNA library with 70 nucle-otides of random sequences was synthesized by the Midland Com-pany (Midland Certified Reagent Co.). Random region is flanked bydefined sequences, which include T7 promoter and a few restrictionenzyme sites for the in vitro transcription and the cloning purposes.The 5� and 3� defined sequences were 5�-CGGAATTCCGTA-ATACGACTCACTATAGGGGAGCTCGGTACC-3� and 5�-AAGCTT-TGCAGAGGATCCTT-3�, respectively. Polymerase chain reaction(PCR) was performed with the DNA random library (1.3 � 1013

molecules), along with 0.25 �M each of 5�- and 3�-primers, 3 mMMgCl2, 250 �M each of dNTPs, 1 U/�l of Taq DNA polymerase. Toreserve the abundance of original library, PCR was limited to tencycles, not to amplify skewed population of random DNA library.DNA library was converted to the RNA library by in vitro transcrip-tion reaction in 40 mM Tris (pH 7.5), 6 mM MgCl2, 2 mM spermidine,10 mM NaCl, 10 mM DTT, 0.5 mM each of rNTPs, 40 U of RNaseinhibitor, 50 U of T7 RNA polymerase for 2 h at 37°C. Template DNAwas removed by the DNase I digestion and the resulting RNA waspurified by phenol:chloroform extraction and ethanol precipitation.The RNA band of expected 110 nucleotides was cut from the 6%polyacrylamide/7 M urea gel and eluted from the gel by incubating in0.5 M ammonium acetate, 0.2% SDS, 1 mM EDTA for 4 h at 37°C.Purified RNA library was quantified by using an UV spectrophotom-eter.

In vitro selection procedure. Two independent selections wereperformed to select nucleocapsid binding RNA aptamers. Selection Iwas performed using 10 �g of the original RNA library in RNAbinding buffer (150 mM NaCl, 50 mM Tris-HCl (pH 7.5), 30 �MZnCl2 1 mM MgCl2, 10 mM DTT, 1 mg/ml BSA, 100 �g/ml tRNA, 40 uRNase inhibitor) with molar ratio of RNA to the protein as 10 to 1.Selection II was performed with similar binding buffer, except withincreased amount of BSA (10 mg/ml) and tRNA (250 �g/ml). Duringselection II cycles, concentration of the nucleocapsid protein wasreduced in later cycles to select high affinity RNA. RNA was brieflyheated to 65°C and cooled to room temperature to make stablestructures before selection steps. To prevent enrichment of the non-specific RNA binding to the GST portion of protein, counter-selectionwith the 5 �g of purified GST proteins was pre-performed in eachcycle. Unbound RNA for the GST protein was equilibrated with 1 �gof GST fused nucleocapsid protein (145 nM) for 30 min in roomtemperature, followed by incubating with 40 �l of Glutathione-Sepharose 4B. After 3–5 times of washing with the RNA bindingbuffer, binding RNA was eluted with 50 mM EDTA and phenol:chloroform extraction, followed by Sephadex G-50 chromatographyand ethanol precipitation. Nucleocapsid binding RNA pool was re-verse transcribed by 10 units of M-MuLV reverse transcriptase andamplified by 20 cycles of PCR. After confirming the PCR products by

2% agarose gel, DNA was transcribed to RNA by in vitro transcrip-tion as described above. The RNA band of the expected size waseluted from 6% polyacrylamide/7 M urea gel and used for the nextround of selection.

Cloning and sequencing of the selected RNA aptamers. After 8(Selection II) to 10 (Selection I) rounds of the reiterated selections,the nucleocapsid binding RNA was converted to DNA and amplifiedby primers containing recognition sites for EcoRI (in 5� primer) andBamHI (in 3� primer). Using these restriction sites, DNA sequenceswere inserted into pUC19 and the recombinant DNA was trans-formed into Escherichia coli DH5�. Plasmid DNAs were preparedfrom 20 different colonies, and their sequences were determined withthe automatic sequencer (ABI prism).

GST pull-down binding assay. Binding affinity of the selectedRNA was measured by GST pull-down assay with the radiolabeledRNA and GST-NC protein. The preparation of radiolabeled RNA wasperformed as followings. Approximately 0.5–1 �g of DNA was incu-bated in transcription buffer (40 mM Tris (pH 7.5), 6 mM MgCl2, 2mM spermidine, 10 mM NaCl, 10 mM DTT, 500 �M each of rAGC-mix, 10 �M rUTP) with 40 �Ci [�-2P] UTP (800 Ci/mmole), 50 unit ofT7 RNA polymerase, 20 unit of RNase inhibitor at 37°C for 3 h.Reactions were then terminated by incubating with 4 units of theRQ1 DNase at 37°C for 15 min. After extracting with phenol/chloroform/isoamylalcohol, RNA was precipitated with ethanol, re-suspended and separated on a 6% polyacrylamide/7 M urea gel.Expected size of RNA was eluted, purified from the gel and itsradioactivity was measured by the scintillation counter. To measurebinding affinity of selected RNA, five nanomolar of the labeled orig-inal RNA or the selected RNA were incubated either with GST (190nM) or with GST-NC fusion protein (145 nM) in the RNA bindingbuffer for 30 min in room temperature as above. Unbound RNA wasremoved by washing the beads 3–4 times; bound RNA was elutedfrom the bead by phenol extraction, loaded to 6% polyacrylamide/7 Murea gel. The gel was dried and the bound RNA was directly visual-ized by autoradiography.

Biosensor assay. BIAcore 3000 was used for the Surface PlasmonResonance experiments. To attach the protein to the CM5 sensorchip, surface of the chip was pre-equilibrated with HEPES andactivated with 0.05 M of N-hydroxysuccinimide (NHS) and 0.2 M ofN-ethyl-N�-(dimethylaminopropyl) carbodiimide (EDC) by modify-ing carboxymethyl groups of dextran. After pre-activation, thenucleocapsid-GST fusion protein was injected to one of the flowcelland the GST protein to the other. After immobilization of the pro-teins, the chip surface was deactivated with 1 M ethanolaminehydrochloride, pH 8.5. After stabilizing the base line, various con-centrations (500 nM, 250 nM, 100 nM, 20 nM and 5 nM) of theoriginal RNA pool, the selected RNA pools and cloned RNA aptamerswere injected to measure KD values of these RNA. After each exper-iment, the sensor chip was regenerated with 125 �M NaCl and 6.25mM NaOH.

Competition binding assay. Selected RNA aptamer was radiola-beled by in vitro transcription in the presence of [�-32P]-UTP asdescribed above. Hundred picomolar of labeled RNA was mixed withvarious amounts of unlabeled competitor RNA (original RNA pooland selected SE8-6 RNA) in concentration from 10�8 to 10�6 M.Hundred nanomolar of nucleocapsid protein was incubated with themixed RNA in the binding buffer for 30 min at room temperature.Binding complex was loaded on a 5% native polyacrylamide gel,dried and analyzed by autoradiography.

Circulization of the selected RNA aptamers. Cloned aptamerDNA was linearized at the 3� of RNA by BamHI and converted to theRNA by in vitro transcription reaction as described above. TemplateDNA was removed by the DNase I digestion and the resulting RNAwas purified by phenol:chloroform extraction and ethanol precipita-tion. After electrophoresis in 6% polyacrylamide/7 M urea gel, RNAband of the expected 110 nucleotides was eluted from the gel and

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

926

quantified by using an UV spectrophotometer. To generate a circularRNA, 100 pM of RNA transcripts was annealed to 400 pM of bridgingDNA oligo (5�-GGTACCGAGCTCCCCAAGGSTCCTCTGCAAAG-CTT-3�) in 100 mM NaCl and the mixture was heated and cooled toroom temperature to make RNA-DNA hybrids. After removing saltby ethanol precipitation, the hybrid mixture was incubated with 30mM Tris-HCl (pH 7.8), 10 mM MgCl2, 10 mM DTT, 10 mM ATP, 20U of T4 DNA ligase for 8–12 h at room temperature. Ligationmixture was purified by phenol:chloroform extraction and ethanolprecipitation and separated in 8% polyacrylamide/7 M urea gel. Thecircular RNA band was cut and eluted from the gel by incubating in0.5 M ammonium acetate, 0.2% SDS, 1 mM EDTA for 4 h in 37°C.Purified circular RNA was quantified by using an UV spectropho-tometer.

RNA footprinting assay. In order to map the binding sites of thenucleocapsid within the selected SE8-6 and SE8-13 aptamers, RNAfootprinting assays were performed. In vitro transcribed RNA wastreated with calf intestine phosphatase for 1 h at 37°C, extractedwith phenol and precipitated with ethanol. RNA was labeled with[�-32P] ATP and T4 polynucleotide kinase, electrophoresed on 6%polyacrylamide/7 M Urea gel and eluted from the gel. Labeled RNAwas allowed to fold into tertiary structure by heating at 65°C for 5min and slow cooling to room temperature. Various amounts (0, 10,50, 100, 500 nM) of protein was incubated with the end-labeled RNAin the RNA binding buffer without BSA and tRNA for 10 min at37°C. RNase T1 was added to the mixture and then was incubated at37°C for 10 min. After extraction and precipitation, RNA was loadedon 15% polyacrylamide/7 M Urea gel. The gel was dried and exposedon X-ray film for autoradiography.

RESULTS AND DISCUSSION

Selection of the RNA Aptamers for the NucloecapsidProtein

In order to select for the high affinity RNA aptamersfor the NC protein, the RNA library that could formstable and diverse structures was bound to the GST fusedmature form of NC protein. Excess amount of GST pro-tein itself was preincubated with the same RNA to re-move GST binding non-specific RNA. After repeating se-lection cycles several times, enrichment of the NCbinding RNA was confirmed by the GST pull-down assayas shown in Fig. 1A. Selected RNA bound specifically tothe NC protein (N) but not to the GST (G) even after 5rounds of selection (SE5), while the original RNA pool(Ori) was shown to bind neither to the GST (G) nor to theNC (N). After repeating selection cycles up to 8 to 10times with slight modification of the procedure, variousRNA aptamer clones were isolated and their binding tothe target protein were determined (Fig. 1B). One of thecloned aptamers (8–13) showed the highest binding tothe NC protein, in comparison to the aptamers 8–6 or8–10 as shown in Fig. 1B.

Specificity of the aptamer binding to the NC proteinwas examined by the competition mobility shift assay(Fig. 1C). Binding of the 8–6 RNA to the NC proteinwas decreased as increasing amount of the cold com-petitor 8–6 RNA was included in the binding reaction(lanes 8–12). However, excess amount of the unlabeledoriginal RNA pool (lanes 3–7) did not compete for thebinding, suggesting that the selected RNA aptamerspecifically bind to the NC protein.

High Binding Affinity and Specific Inhibitory Activityof the RNA Aptamer

To quantify the binding affinities of the various RNAaptamers, the Surface Plasmon Resonance (SPR) tech-nique was utilized. The SPR profile in Fig. 2A clearly

FIG. 1. Isolation and enrichment of high affinity RNA aptamersto the HIV nucleocapsid protein by in vitro selection. (A) Bindingassay was performed with the original RNA library (Ori), the se-lected RNA pools from 5th cycles (SE5). Each of RNA was labeledwith [�-32P] UTP and 5 nM of RNA was incubated either with GSTprotein (G, 190 nM) or with GST-nucleocapsid protein (N, 150 nM).After washing out nonspecific binding RNA, bound RNA was elutedfrom the beads, separated in 6% polyacrylamide/7 M urea gel anddirectly visualized by the autoradiography. Selected RNA pools dem-onstrated high affinity binding to the nucleocapsid protein (N) ascompared to the GST (G). One tenth of the labeled input RNA wasloaded as a control (Input). (B) Binding affinities were measuredwith the various RNA aptamer clones from the eighth selection cycle(SE8). Names of the aptamers are as follows: 8–6, 8–10, and 8–13.Experimental details are the same as in A. (C) Specificity of the RNAaptamer 8–6 to the nucleocapsid protein was determined by the gelshift competition experiment. Hundred picomolar of the labeled 8–6RNA was incubated with 100 nM of the nucleocapsid protein in thebinding buffer for 30 min and loaded onto 5% polyacrylamide gel.Bound RNA migrated slowly, as indicated with the arrow. In eachsample, cold unlabeled RNA was included as a cold competitor. Lane1, no protein; lane 2, no competitor; lanes 3–8, 10 nM, 50 nM, 100nM, 500 nM, 1000 nM of original RNA library competitor (Ori),respectively; lanes 8–12, 10 nM, 50 nM, 100 nM, 500 nM, 1000 nMof SE8–6 RNA aptamer competitor (8–6), respectively.

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

927

demonstrates that the selected RNA aptamers 8–13and 8–6 bind strongly to the NC protein, as comparedto the RNA pools when the same concentrations ofRNA were injected. By injecting various concentrationsof RNA in kinetic condition, dissociation constants ofRNA could be obtained. As summarized in Table 1,binding affinities of the selected RNA pool and the

cloned aptamers were at least 100–500 fold increasedas compared to the original RNA library. Most of theselected aptamers have subnanomolar range of Kd,some of which has high binding affinity as the NCcognate target RNA, the viral genomic � sequence,does. Aptamer 8–13 that was shown to have high bind-ing affinity by the GST pull-down assay (Fig. 1A) alsodemonstrated the highest binding affinity (Kd � 4.9 �10�10 M) to the NC protein. Such binding affinity is ashigh as that of the NC cognate �-RNA (Kd � 7.3 �10�10 M).

Such binding affinity is higher, or at least no less,than the affinity of RNA molecules that have beenpreviously selected against either gag polyprotein orimmature form (71 amino acids) of NC protein (13–15).In addition, we used the mature form of NC protein (55amino acids) and a solution based binding protocol,such as the GST pull-down method, rather than atypical membrane-based nitrocellulose filter bindingmethod. Therefore, RNA aptamer selected in this studyis likely act as a high affinity inhibitor for the fullyprocessed mature form of NC protein during viral en-capsidation.

Since binding affinities of the RNA aptamers arehigh, it is expected that they might inhibit the �-RNAbinding to the NC protein. To test this, �-RNA wasradiolabeled and incubated with the NC protein, andcompeted with increasing amount of the aptamer 8–6RNA (lanes 3–4, Fig. 2B). �-RNA binding was com-pletely abolished when the excess amount of theaptamer 8–6 was included as an inhibitor (lane 4), butcontrol tRNA did not compete for �-RNA binding to theNC (lanes 5–6), suggesting a specific inhibition of theaptamer RNA. Such inhibitory effect of the aptamerRNA was comparable to that of �-RNA (lanes 7–8).

TABLE 1

Binding Parameters of Various RNA as Measuredby the Surface Plasmon Resonance

RNA

Kd (M)

Mean SD

Original pool 2.0 � 10�7 —Selected pool 1.2 � 10�8 —Aptamer 10–2 2.4 � 10�9 1.1Aptamer 8–6 1.4 � 10�9 5.0Aptamer 10–3 8.4 � 10�10 —Aptamer 8–13 4.9 � 10�10 2.0Psi (�) 7.3 � 10�10 6.1

Note. Various concentrations of RNA (5, 10, 25, 50, 100, 250 nM)were injected into the flow cells that were previously immobilizedeither with the GST or with the GST-nucleocapsid proteins. Aftersubtracting nonspecific binding to the GST protein, the dissociationconstants to the nucleocapsid protein were determined by the BIAevaluation program. At least three different measurements wereperformed and their mean values and standard deviation (SD) of Kdare presented.

FIG. 2. (A) Binding profiles of various RNA using Surface PlasmonResonance (SPR). Various RNA molecules were synthesized by the invitro transcription and the same concentration (250 nM) of RNA mole-cules were sequentially injected into the BIAcore. Order of the bindingcurves is as follows: 1, Aptamer 8–13; 2, aptamer 8–6; 3, SE10 pool; 4,original RNA library pool. (B) Specific inhibition of the aptamer 8–6 forthe �-RNA binding to the NC protein. �-RNA was radiolabeled with[�-32P] UTP and 75 pM of RNA was incubated with 3500 nM of His-NCprotein along with cold competitors as follows: lane 1, no protein; lane 2,no competitor; lanes 3–4, 350 nM, 3500 nM of RNA aptamer 8–6,respectively; lanes 5–6, 350 nM, 3500 nM of tRNA, respectively; lanes7–8, 350 nM, 3500 nM of �-RNA, respectively.

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

928

These data strongly suggest that the selected RNAaptamers are not only a high affinity binder to theprotein, but also a specific competitor for the viral RNAbinding to the NC protein.

Multiple Binding Patterns of the NucleocapsidProtein to the Aptamers

We next determined the binding motif of the RNAaptamer for the NC protein. Longer size of the RNAlibrary we used here is likely to form more complexstructure than small RNA does, which would provide

more than one binding sites for the target protein.Binding sites of the selected RNA aptamers to the NCprotein were examined to characterize the binding pat-tern and minimal binding motif on the RNA. Two dif-ferent aptamers (8–6 and 8–13) that were found tohave high binding affinity to the protein (see Table 1and Fig. 2) were chosen for the mapping study.

Result of the RNA footprinting analysis with the 8–6RNA aptamer is shown in Fig. 3, which clearly demon-strates at least two distinct binding sites for the NCprotein. Similar results were obtained for the 8–13RNA aptamer (data not shown). Based on this study,binding sites were mapped and their predicted second-ary structures were presented in Fig. 4. In each RNAaptamer, at least two different NC binding motifs seemto be present in the 110-mer size of RNA, which weredenoted as the Site I and the Site II (Figs. 4A and 4B).Since multiple stem-loops in the �-sequences mightbind better to the NC than single stem-loop does (3, 6,7), it is likely that the NC protein binds to viral RNA asmultimers. Such a multiple binding pattern of the NCprotein is also observed in RNA aptamer binding asshown in here. Thus, it is presumed that longer size ofRNA with at least two binding motifs might be moreeffective and tighter binding molecule for the NC pro-tein.

RNA mapping study suggests that all of the NCbinding motifs have similar stem-loop structures witha GC rich stem and a GU rich loop. Comparison of the

FIG. 3. RNA footprinting analysis of the aptamer 8–6. End-labeled RNA was incubated with increasing amount of the NC pro-tein and exposed to RNase T1 to map the NC binding sites on theRNA as described under Materials and Methods. Lane 1, alkalinehydrolysis; lane 2, RNase T1 treatment; lane 3, No treatment; lane 4,without NC; lane 5, 5 nM of NC; lane 6, 10 nM of NC; lane 7, 50 nMof NC; lane 8, 100 nM NC protein. Two NC binding motifs wereindicated in the figure as I and II.

FIG. 4. Predicted secondary structures of the NC binding motifs.(A) Two binding motifs of aptamer 8–6 are presented as the Site Iand II. (B) Two binding motifs of the aptamer 8–13.

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

929

NC binding sites with previously defined sites fromother studies revealed a couple of interesting features(13–16). First, stretches of G or U are frequently foundin NC binding RNA or oligonucleotides. Our study alsosuggests that the G or U rich sequence block is afavorite binding motif for the NC protein, however, A isalso frequently found in the loop structures in ourstudy. Secondly, most of the NC binding sites werepredicted to form the stem-loop structures with rela-tively stable stems with 5 to 6 complementary basepairs and 4 to 6 single-stranded loops. Similar struc-tures were found not only in some selected sequencesfrom other studies (13–16) but also in �-sequence ofHIV-1 genomic RNA (4–7). In fact, the �-sequence isalso composed of four stem-loops with stable stemstructures. In spite of such similarities and differencesof the NC binding motifs, it is noteworthy that variousbut no redundant sequences are still selected by differ-ent selection procedures. Thus, it appears that morediverse structures and sequences need to be discoveredto fully understand the binding characteristics of NC toRNA as well as to the �-sequences of HIV-1 genomicRNA.

High Affinity Binding of the Circular RNA

To enhance the therapeutic value of the selectedRNA molecules as an inhibitory molecule in vivo, ithas to be prevented from the degradation by stabi-lizing its structure. Because the ends of the RNAmolecules might be the most vulnerable sites to thenucleases, we attempted to stabilize it by joining theends of the linear molecules. Circular RNA was pre-pared by hybridizing it with the bridging oligonucle-otides that could bring the ends of the molecule,followed by the intra-molecular ligation of the RNAends (17).

Binding parameters of the circular RNA aptamerswere measured by the Surface Plasmon Resonancetechnique as described above. As shown in Table 2,Kd values of the circular RNA were about nanomolar

range, which is as high as the linear RNA does. Allthree different aptamers (10–2, 10 –3 and 8 –13)showed nanomolar binding affinity with similar dis-sociation and association rates. It suggests that thestabilization of the aptamer RNA structures by ligat-ing ends of the RNA molecules can be achieved,without interfering with the binding motif. Such ap-proach could be applied to generate stable aptamerRNA for various purposes. The selection and stabili-zation of RNA aptamer presented in this paper mightcast a light toward the development of anti-HIVtherapeutics based on the NC protein as a targetmolecule (18).

ACKNOWLEDGMENT

This study was financially supported by a research fund from theKorean Ministry of Science and Technology KISTEP grants(M10015000021) to J.C.Y. and S.J.

REFERENCES

1. Rein, A., Henderson, L. E., and Levin, J. G. (1998) Nucleic acidchaperone activity of retroviral nucleocapsid proteins: Signifi-cance for viral replication. TIBS 23, 297–301.

2. Berkowiz, D. B., Luban, J., and Goff, S. P. (1993) Specific bindingof human immunodeficiency virus type 1 gag polyprotein andnucleocapsid protein to viral RNAs detected by RNA mobilityshift assay. J. Virol. 67, 7190–7200.

3. Clever, J., Sassetti, C., and Parslow, G. P. (1995) RNA secondarystructure and binding sites for gag gene products in the 5�packaging signal of human immunodeficiency virus type 1. J. Vi-rol. 69, 2101–2109.

4. De Guzman, R., Wu, Z. R., Stalling, C. C., Pappalardo, L., Borer,P. N., and Summers, M. F. (1998) Structure of the HIV-1 nucleo-capsid protein bound to the SL3 �-RNA recognition element.Science 279, 384–388.

5. Vuilleumier, C. Bombarda, E., Morellet, N., Gerard, D., Roques,B. P., and Mely, Y. (1999) nucleic acid sequence determination bythe HIV-1 nucleocapsid protein NCp7: A fluorescence study.Biochemistry 38, 16816–16825.

6. Amarasinghe, G. K., De Guzman, R. N., Turner, R. B., Chancel-lor, K. J. Wu, Z. R., and Summers, M. F. (2000) NMR structureof the HIV-1 nucleocapsid protein bound to stem-loop SL2 of the�-RNA packaging signal. Implications for genome recognition.J. Mol. Biol. 301, 491–511.

7. Maki, A. H., Ozarowski, A., Misra, A., Urbaneja, M. A., andCasas-Finet, J. R. (2001) Phosphorescence and optically detectedmagnetic resonance of HIV-1 nucleocapsid protein complexeswith stem-loop sequences of the genomic �-recognition element.Biochemistry 40, 1403–1412.

8. Baskerville, S., Frank, D., and Ellington, A. D. (1998) Directedevolutionary description of natural RNA structures. In RNAStructure and Function (Simons, R., and Grunberg-Manago, M.,Eds.), pp. 203–251, Cold Spring Harbor Laboratory Press.

9. Hermann, T., and Patel, D. J. (2000) Adaptive recognition bynucleic acid aptamers. Science 287, 820–825.

10. Covey, S. N. (1986) Amino acid sequence homology in gag regionof reverse transcribing elements and the coat protein gene ofcauliflower mosaic virus. Nucleic Acids Res. 14, 623–633.

11. Freed, E. O. (1998) HIV-1 gag proteins: Diverse functions in thevirus life cycle. Virology 251, 1–15.

TABLE 2

Binding Parameters of the Circular Formof the Selected RNA Aptamers

ka (1/Ms) kd (1/s) KA (1/M) KD (M)

Circular 10–3 5.44 � 105 7.97 � 10�4 6.82 � 108 1.47 � 10�9

Circular 10–2 4.50 � 105 6.73 � 10�4 6.69 � 108 1.49 � 10�9

Circular 8–13 2.67 � 105 2.47 � 10�4 1.08 � 109 9.25 � 10�10

Note. Various concentrations of RNA (5, 10, 25, 50, 100, 250 nM)were injected into the flow cells that were previously immobilizedeither with the GST or with the GST-nucleocapsid proteins. Aftersubtracting nonspecific binding to the GST protein, the dissociationconstants to the nucleocapsid protein were determined by the BIAevaluation program.

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

930

12. Aldovini, A., and Young, R. (1993) Mutations of RNA and proteinsequences involved in human immunodeficiency virus type 1packaging result in production of noninfectious virus. J. Virol.64, 1920–1926.

13. Lochrie, M. A., Waugh, S., Pratt, D. G., Clever, J., Parslow, T. G.,and Polisky, B. (1997) In vitro selection of RNAs that bind to thehuman immunodeficiency virus type-1 gag polyprotein. NucleicAcids Res. 25, 2902–2910.

14. Allen, P., Collins, B., Brown, D., Hostsmsky, Z., and Gold, L.(1996) A Specific RNA structural motif mediates high affinitybinding by the HIV-I nucleocapsid protein (NCp7). Virology 255,306–315.

15. Berglund, J. A., Charpentier, B., and Rosbash, M. (1997) A high

affinity binding site for HIV-1 nucleocapsid protein. Nucleic Ac-ids Res. 25, 1042–1049.

16. Fisher, R. J., Rein, A., Fivash, M., Urbaneja, M. A., Casas-Finet,J. R., Medaglia, M., and Henderson, L. E. (1998) Sequence-specific binding of human immunodeficiency virus type 1 nucleo-capsid protein to short oligonucleotides. J. Virol. 72, 1902–1909.

17. Chen, C. Y., and Sarnow, P. (1995) Initiation of protein synthesisby the eukaryotic translational apparatus on circular RNAs.Science 268, 415–417.

18. Clever, J. L., Taplitz, R. A., Lochrie, M. A., Polisky, B., andParslow, T. G. (2000) A heterologous, high-affinity RNA ligandfor human immunodeficiency virus gag protein has RNA pack-aging activity. J. Virol. 74, 541–546.

Vol. 291, No. 4, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

931