Two basic regions of NCp7 are sufficient for conformational conversion of HIV-1 dimerization initiation site from kissing-loop dimer to extended-duplex dimer.

Nucleocapsid (NC) protein possesses nucleotide-annealing activities, which are used in various processes in retroviral life cycle. As conserved characters, the NC proteins have one or two zinc fingers of CX(2)CX(4)HX(4)C motif surrounded by basic amino acid sequences. Requirement of the zinc fingers for the annealing activities of NC protein remains controversial. In this study, we focused the requirement in the process of maturation of dimeric viral RNA. Discrimination between immature and mature dimers of synthetic RNA corresponding to the dimerization initiation site of human immunodeficiency virus type 1 (HIV-1) genomic RNA was performed based on their Mg(2+)-dependent stability in gel electrophoreses and on their distinct signal pattern from NMR analysis of imino protons. Chaperoning activity of the HIV-1 NC protein, NCp7, and its fragments for maturation of dimeric RNA was investigated using these experimental systems. We found that the two basic regions flanking the N-terminal zinc finger of NCp7, which are connected by two glycine residues instead of the zinc finger, were sufficient, although about 10 times the amounts of peptide were needed in comparison with intact NCp7. Further, it was found that the amount of basic residues rather than the amino acid sequence itself is important for the activity. The zinc fingers may involve the binding affinity and/or such a possible specific binding of NCp7 to dimerization initiation site dimer that leads to the maturation reaction.

Nucleocapsid (NC) protein possesses nucleotide-annealing activities, which are used in various processes in retroviral life cycle. As conserved characters, the NC proteins have one or two zinc fingers of CX 2 CX 4 HX 4 C motif surrounded by basic amino acid sequences. Requirement of the zinc fingers for the annealing activities of NC protein remains controversial. In this study, we focused the requirement in the process of maturation of dimeric viral RNA. Discrimination between immature and mature dimers of synthetic RNA corresponding to the dimerization initiation site of human immunodeficiency virus type 1 (HIV-1) genomic RNA was performed based on their Mg 2؉ -dependent stability in gel electrophoreses and on their distinct signal pattern from NMR analysis of imino protons. Chaperoning activity of the HIV-1 NC protein, NCp7, and its fragments for maturation of dimeric RNA was investigated using these experimental systems. We found that the two basic regions flanking the N-terminal zinc finger of NCp7, which are connected by two glycine residues instead of the zinc finger, were sufficient, although about 10 times the amounts of peptide were needed in comparison with intact NCp7. Further, it was found that the amount of basic residues rather than the amino acid sequence itself is important for the activity. The zinc fingers may involve the binding affinity and/or such a possible specific binding of NCp7 to dimerization initiation site dimer that leads to the maturation reaction.
Nucleocapsid (NC) 1 protein is a component in retroviral particles and takes various functional roles in retrovirus life cycle, involving encapsidation of the genomic RNA, maturation of the dimeric RNA, annealing of tRNA onto primer binding site (PBS) of the viral RNA, protection of reverse transcriptase (RT) pausing, and strand transfer during reverse transcription (for review, see Refs. 1 and 2 and references therein). These functions are associated with the specific binding and the annealing activities to viral RNA. The NC proteins of all retroviruses except those of spumaviruses have one or two zinc fingers of CX 2 CX 4 HX 4 C motif surrounded by basic amino acid sequences. Three-dimensional structures of NC proteins in several retroviruses are known (3)(4)(5)(6)(7)(8)(9)(10). The crystal structure of complex of the NC protein and packaging signal of the genomic RNA in the human immunodeficiency virus type 1 (HIV-1) demonstrated two zinc finger knuckles bind to G-rich loop of the RNA in a stem-loop structure and N-terminal basic sequence forms a 3 10 -helix and binds to major groove of stem (5). Alternations and deletions of the zinc fingers impaired the specific RNA binding activity (11)(12)(13) and packaging of the viral RNA (11,14), and extinguished viral infectivity (11,14), but did not much affect nonspecific RNA binding activity (13,15) and the annealing activities such as annealing of complementary DNA and RNA (16), tRNA-PBS annealing (16 -19), viral RNA dimerization (17)(18)(19), strand transfer, (20), and reduction of RT pausing (21). On the other hand, mutations and deletions of the basic regions flanking the zinc fingers impaired the annealing activities (16 -22), RNA binding activity (12,15,17,23,24), virion formation (15), and viral infectivity (22, 24 -27). In contrast, Remy et al. (28) disagreed the dispensability of the zinc fingers for the tRNA-PBS annealing activity of NC protein. Feng et al. (29) also reported on the requirement of residues within the N-terminal zinc finger, not the zinc finger structure itself, for maturation reaction of dimeric viral RNA. Rong et al. (30) reported on functional difference of RNA complexes annealed by NC protein and its mutants. They showed that NC protein devoid of the zinc fingers could anneal tRNA onto the viral RNA, but the resultant primer-template complex was not elongated by RT to the full-length negative-strand of strongstop DNA, unlike the primer-template complex yielded by wildtype NC protein.
Maturation of dimeric viral RNA is one of the NC proteininvolved reactions in retrovirus life cycles. Maturation of virion particle occurring after release from host cell is executed through cleavage of gag protein to generate NC protein by viral protease. In this maturation event, the dimeric RNA genome in virion particle is converted into a heat-stable dimer (31,32). In vitro experiments showed that 5Ј leader sequences of viral RNA genome form two types of dimer with different stability, named as loose and tight dimers (33)(34)(35)(36)(37). Loose dimer is spontaneously * This work was supported by Research for the Future Program Grant JSPS-RFTF97L00503 from the Japan Society for the Promotion of Science and in part by a grant-in-aid for high technology research from the Japanese Ministry of Education, Science, Sports and Culture. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  formed at physiological temperature and converted into tight dimer by incubation with NC protein (35)(36)(37). The sequence required for the spontaneous dimerization of the genomic RNA in HIV-1 was identified and named as the dimerization initiation site (DIS), which can form a stem-loop structure with a self-complementary sequence in the loop and a bulge in the stem (38,39). The DIS was also shown to be necessary and sufficient for exhibiting the two-step dimerization involving the HIV-1 NC protein, NCp7 (37,40). In this reaction, Zn 2ϩ is not required (40). The two types of secondary structure of the DIS dimer were predicted (38,39). In the first one, base pairs are mostly within monomer to form a stem-loop structure and inter-monomer base pairs are formed only at the self-complementary six nucleotides in the loop (Fig. 1A, kissing-loop dimer). In the second one, base pairs are entirely between two monomers (Fig. 1A, extended-duplex dimer). These two different topologies of the DIS dimer were confirmed by unambiguous discrimination between intra-and intermolecular nuclear Overhauser effect in NMR measurements of an equimolar mixture of 15 N-labeled and non-labeled DIS molecules (41) and also displayed in the three-dimensional structures of the two types of the DIS dimer determined by 1 H NMR spectroscopy (42)(43)(44) and x-ray crystallography (45).
In the present study, we investigated whether the zinc fingers are required for the maturation reaction of the DIS dimer. The previously established experimental system were clearly able to discriminate the two types of the DIS dimer (40,41) and thus reveal the annealing activity of NCp7. We show here that the zinc fingers are dispensable and the basic regions surrounding the N-terminal zinc finger are sufficient for the maturation reaction of the DIS dimer.

EXPERIMENTAL PROCEDURES
Preparation of RNAs and Peptides-RNA oligonucleotides were synthesized either chemically by a phosphoramidite method with an automatic DNA/RNA synthesizer, Expedite model 8909 (PerSeptive Biosystems Inc.) or enzymatically by an in vitro transcription method with AmpliScribe T7 transcription kits (Epicentre Technologies Co.). For chemically synthesized RNAs, protection groups were removed with ammonia and tetra-n-butylammonium fluoride. Purification with polyacrylamide gel electrophoresis (PAGE) using 30 ϫ 40-cm glass plates (Nihon Eido Co. Ltd., Tokyo, Japan) under denaturing conditions, and extensive desalting by ultrafiltration (Centricon YM-3, Amicon Inc.) were carried out. Lyophilized samples of synthetic nucleocapsid protein NCp7 (LAV strain 72-amino acid sequence) and its related peptides were purchased from Peptide Institute Inc. (Osaka, Japan) and Sawady Technology Co. Ltd. (Tokyo, Japan), respectively.
Assay for Conformational Conversion of RNA Dimers-Assay was performed in the following way. First, 12 M DIS39 in 4 l of water was heated at 95°C for 5 min and chilled on ice for 5 min, and then 4 l of 2ϫ PN buffer (1ϫ PN buffer contains 10 mM sodium phosphate (pH 7.0) and 50 mM NaCl) was added. In the case of heterodimer of DIS39K1 and DIS39K2, 12 M amounts of each RNA in 2 l of water were heated separately at 95°C for 5 min and chilled on ice for 5 min. After 2 l of 2ϫ PN buffer was added to each RNA solution, both solutions were mixed. Second, various concentrations of NCp7 or its related peptides in 12 l of 1ϫ PN buffer were added to the RNA solutions. The molar ratio of peptide to RNA was 1:1 through 10:1. In the case of experiments without peptide, 12 l of 1ϫ PN buffer was added. Third, the mixture solutions were incubated at 37°C for 2.5 or 20 h and then treated with phenol/chloroform solution, regardless of the presence or absence of peptide. Ten l of the aqueous layer containing RNA was collected and mixed with 10 l of loading buffer containing glycerol and dyes. Then, the solution was divided into two parts, which were separately analyzed by electrophoreses through non-denaturing polyacrylamide gels (10% or 15%) in TBM buffer (89 mM Tris, 89 mM borate, 1 mM MgCl 2 ) and in TBE buffer (89 mM Tris, 89 mM borate, 2 mM EDTA), respectively, at room temperature. After electrophoresis, the gels were stained with Red Stain (Bio-Rad) and RNA was visualized by a ultraviolet illuminator (FAS-III, Toyobo Co. Ltd., Osaka, Japan).
NMR Measurements-DIS39 in 1ϫ PN buffer was concentrated to 0.06 -0.7 mM by ultrafiltration, and D 2 O was added to 5%. Final volume was 200 l. NMR spectra were recorded on a Bruker DRX-500 spec-trometer at a probe temperature of 25°C. Solvent proton signal was suppressed by a jump-and-return pulse sequence (46) with a delay of 65 s, and 1000 -16,000 scans were accumulated. A 3-Hz exponential multiplication prior to the Fourier transformation and a polynomial base-line correction were applied.

RESULTS
To examine the RNA annealing activity of NCp7, we used our experimental system of a 39-mer RNA (DIS39) covering the whole DIS sequence. DIS39 forms the two types of dimers, the kissing-loop dimer and extended-duplex dimer (Fig. 1A), whose base pairing topologies were confirmed by NMR analysis (41). They can be readily discriminated by PAGE; both dimers retain dimeric states during flowing through polyacrylamide gel containing Mg 2ϩ , but in PAGE without Mg 2ϩ , the kissing-loop dimer separates into monomers, whereas the extended-duplex dimer retains the dimeric state because of their different Mg 2ϩdependent stability. As reported in the previous study (40), DIS39 mostly forms the kissing-loop dimer by itself at 37°C (Fig. 2, lane 1) and is converted into the extended-duplex dimer when incubated with an equimolar NCp7 at 37°C (Fig. 2, lane  2). Using this experimental system, we examined the relevance of the two basic regions of NCp7 to the RNA-annealing activity.
First, we prepared three peptides, NCBR1, NCBR2, and NCBR[1ϩ2], which correspond to the first basic region, the second basic region, and both connected by a linker of two glycine residues (Fig. 1B). Instead of intact NCp7, each of the peptides was mixed with DIS39 at 1 and 10 equivalent molar ratios of peptide to RNA. For DIS39 incubated with 10 equivalent molar amount of NCBR1, the upper band in Mg 2ϩ -free gel (Mg 2ϩ -independent dimer) became slightly intense (Fig. 2A,  compare lanes 1 and 4), indicating that shift to the extendedduplex dimer in a small population was induced by NCBR1. On the other hand, no difference was found in electrophoretic

NCp7 Basic Regions Convert Dimeric States of HIV-1 DIS
pattern between DIS39 incubated with NCBR2 and DIS39 without any peptide ( Fig. 2A, compare lane 1 and lanes 5 and  6), indicating that NCBR2 gave no effect to DIS39. In the case of NCBR[1ϩ2], although the annealing activity was weak when mixed at an equimolar ratio ( Fig. 2A, lane 7), 10 equivalent molar amount of NCBR[1ϩ2] clearly showed the activity; the lower band in Mg 2ϩ -free gel (monomer separated from Mg 2ϩrequired dimer) mostly disappeared, and the upper band (Mg 2ϩ -independent dimer) drastically increased ( Fig. 2A, compare lanes 1 and 8), indicating complete shift from the kissingloop dimer to the extended-duplex dimer. NCp7 and NCBR[1ϩ2] seemed not to show turnover because extended incubation did not change electrophoretic patterns (Fig. 2B).
There might be a possibility that the increased population of the upper band in Mg 2ϩ -free gel (Fig. 2A, lane 8) could not be the extended-duplex dimer but the kissing-loop dimer stabilized by NCBR[1ϩ2] that could be survived from the phenol/ chloroform treatment. This possibility of complex formation, if any, could be examined by using DIS mutants that only form kissing-loop dimers but never extended-duplex dimers. We previously designed a heterodimer of DIS39 mutants, DIS39K1 and DIS39K2, (termed DIS39K as the pair) as such an exclusive kissing-loop dimer (40) (Fig. 3A). As expected, DIS39K gave no upper band in Mg 2ϩ -free gel even if incubated with NCp7 (Fig. 3B, lane 4). Using this pair of mutants instead of the original DIS39, we carried out a similar experiment for the annealing activity assay of NCBR[1ϩ2]. Consequently, no upper band in Mg 2ϩ -free gel appeared for DIS39K incubated with 10 mol of NCBR[1ϩ2]/mol of total RNA molecules (Fig. 3B, lane  6). This indicates that the complex between the kissing-loop dimer and NCBR[1ϩ2] did not exist at least after the phenol/ chloroform treatment. Thus, it is highly probable that the major population in the upper band in lane 8 of Fig. 2B is the extended-duplex dimer of DIS39. This is confirmed by the NMR measurement as described below.
To investigate amounts of NCBR[1ϩ2] required for conformational conversion of a DIS39 dimer, the annealing assay was carried out at NCBR[1ϩ2] to DIS39 ratios between 1 and 10 (or ratios of NCBR[1ϩ2] to DIS39 dimer between 2 and 20). As shown in Fig. 4, although 2 molecules of NCp7 on average are enough to anneal one dimer of DIS39 (or 2 molecules of DIS39), 16 -20 molecules of NCBR[1ϩ2] on average are required for annealing of one dimer of DIS39.
To determine whether NCp7-and NCBR[1ϩ2]-converted DIS39 dimers take the same conformation and whether they are structurally the same one as the previously characterized extended-duplex dimer generated by heat-annealing (41), we measured and compared their imino-proton NMR spectra. As shown in Fig. 5, the NMR spectra of DIS39 samples treated by heat-annealing, NCp7 and NCBR[1ϩ2] (A-C, respectively) are all identical, whereas they are different from the spectrum of DIS39 samples incubated in the absence of peptides (D). This indicates that both NCp7 and NCBR[1ϩ2] treatment gave the same extended-duplex conformation that was previously analyzed in detail for DIS39 dimer prepared by heat-annealing (41). It should be noted that the sample for Fig. 5D is mainly in the kissing-loop dimer form but contained small amounts of the  7 and 8, respectively). K, kissing-loop dimer; E, extended-duplex dimer.  extended-duplex dimer, which gave the same signal with spectra A, B, and C. NCBR[1ϩ2] was active as described above. Which residue, then, is important for the activity? We examined the annealing activity of mutant peptides of NCBR[1ϩ2]. Basic residues Arg-3, Arg-7, Lys-11, and Lys-14 in NCBR[1ϩ2] were individually replaced with alanine (M3, M7, M11, and M14, respectively, as shown in Fig. 1B). Furthermore, the amino acid sequence of NCBR[1ϩ2] was randomized without alteration of amino acid composition; the N-terminal 14 residues and the whole amino acid residues were shuffled (RD1 and RD2, respectively, as shown in Fig. 1B). Basically, all these mutant peptides showed the annealing activity to an extent similar to that of the original NCBR[1ϩ2] (Fig. 6). Some point mutants showed a little reduced activity. This result indicates that an important factor for the activity is a large amount of basic residues rather than the amino acid sequence itself of NCBR[1ϩ2].

DISCUSSION
There are controversial reports on dispensability or requirement of the zinc fingers for the RNA annealing activities of NC protein. Regarding one of the annealing activities, the maturation reaction of the dimeric viral RNA, we previously established a simple in vitro assay system using a short RNA fragment (DIS39) corresponding to the DIS of the HIV-1 (40). The immature dimer (loose dimer) and NCp7-converted mature dimer (tight dimer) of DIS39 can be easily distinguished by their different Mg 2ϩ -dependent stabilities when they are subjected to gel electrophoreses with and without Mg 2ϩ . This electrophoretic method was originally developed by Laughrea and Jette (33). We have also confirmed by NMR measurements that the two types of DIS39 dimer take the secondary structures proposed in the kissing-loop model (38,39), the kissing-loop dimer and extended-duplex dimer, respectively (Fig. 1A) (41). Using this assay system, we demonstrated previously that NCp7 did not require Zn 2ϩ to generate mature DIS39 dimer (40). In the present study, we further investigated dispensability of the zinc fingers themselves and found that the two basic regions flanking the N-terminal zinc finger, which were connected by a linker of two glycine residues instead of the zinc finger (NCBR[1ϩ2]), were sufficient for the activity although about 10 times the amount of peptide was needed in comparison with intact NCp7 to convert the same amount of DIS39 dimer (Fig. 2). The possibility that complexes between the kissing-loop dimer and the basic peptide behave like the extended-duplex dimer in gel electrophoreses was excluded by using a pair of DIS39 mutants that forms only a kissing-loop dimer (DIS39K, Fig. 3) and, more directly, by NMR measurements of DIS39 dimers converted by the basic peptide NCBR[1ϩ2] and intact NCp7.
Remy et al. (28) reported on requirement of the zinc fingers for the tRNA-PBS annealing activity of NC protein. They eliminated the phenol/chloroform treatment; thus, there is still be a possibility that they just observed the ternary complex of tRNA, PBScontaining viral RNA, and NCp7 without annealing. Feng et al. (29) have investigated the NCp7-induced maturation of dimeric RNA of Harvey sarcoma virus. They measured the heat stability of RNA dimer and discriminated the two types of RNA dimer. Their results showed the requirement of the basic residues flanking the N-terminal zinc finger for the activity of NCp7. However, they also showed that the mutation of the aromatic residue within the N-terminal zinc finger as well as the removal of the N-terminal zinc finger lost the maturation activity, although the cysteine residues in the zinc fingers could be mutated without the loss of the activity. We note the possibility that an increasing amount of their NCp7 mutants devoid of the zinc finger could lead to the detection of the maturation activity, like in our case of NCBR[1ϩ2].
Approximately 10 molecules of NCBR[1ϩ2]/DIS39 molecule on average were required for the maturation activity. In the case of an equimolar mixture between NCBR[1ϩ2] and DIS39, a prolonged incubation did not change the fraction of the mature dimer of DIS39, as shown in Fig. 2B. However, an increasing amount of NCBR[1ϩ2] raised the fraction of the mature dimer linearly in an incubation time of 2.5 h (Fig. 4). The binding affinity of NCBR[1ϩ2] may be decreased in comparison with intact NCp7, and the fraction of a possible specific binding that leads to the maturation reaction may be decreased by the competition with the nonspecific, and non-active, binding. The zinc finger domains of NCp7 may take a role to bind DIS39 specifically and locate the basic regions in the best position on DIS39 for the maturation reaction.
tain basic regions, their amino acid sequences themselves are not conserved (2). Our study demonstrated that peptides with randomized sequences of NCBR[1ϩ2] had the same level of the annealing activity as that of the wild type sequence of NCBR[1ϩ2]. On the hand, little and no activities were seen in the shorter peptides NCBR1 and NCBR2, respectively. These indicate that some number of basic residues is required for the annealing activity. If NCBR[1ϩ2] is assumed to form a helical structure, a constructed three-dimensional model of NCBR [1ϩ2] shows that positive electrostatic potential on the molecular surface is localized in one side of the helix. It should be noted that, unexpectedly, the two randomized peptides also show a similar localization of positive electrostatic potential in a helical structure model (data not shown). This possible common structural feature may be an important factor to anneal nucleotide duplexes. Further mutational analysis of the basic peptide NCBR[1ϩ2] and structural analysis of a complex between NCBR[1ϩ2] and DIS39 are required to understand detailed atomic interactions and molecular mechanism for the maturation reaction of dimeric retroviral RNA, one of important processes of viral life cycle.