Identification of the Minimal Essential RNA Sequences Responsible for Site-Specific Targeting of the Leishmania RNA Virus 1-4 Capsid Endoribonuclease

Abstract

The Leishmania RNA virus 1-4 capsid protein possesses an endoribonuclease activity responsible for single-site-specific cleavage within the 450-nucleotide 5′ untranslated region of its own viral RNA transcript. To characterize the minimal essential RNA determinants required for site-specific cleavage, mutated RNA transcripts were examined for susceptibility to cleavage by the virus capsid protein in an in vitro assay. Deletion analyses revealed that all determinants necessary for accurate cleavage are encoded in viral nucleotides 249 to 342. Nuclease mapping and site-specific mutagenesis of the minimal RNA sequence defined a stem-loop structure that is located 40 nucleotides upstream from the cleavage site (nucleotide 320) and that is essential for accurate RNA cleavage. Abrogation of cleavage by disruption of base pairing within the stem-loop was reversed through the introduction of complementary nucleotide substitutions that reestablished the structure. We also provide evidence that divalent cations, essential components of the cleavage reaction, stabilized the stem-loop structure in solution. That capsid-specific antiserum eliminated specific RNA cleavage provides further evidence that the virus capsid gene encodes the essential endoribonuclease activity.

The Leishmania RNA virus (LRV) genome comprises approximately 5.3 kb of double-stranded RNA that encodes two large open reading frames (ORF2 and ORF3) on the positive-sense strand in all isolates examined (31, 32, 35). When expressed by recombinant baculovirus in Spodoptera frugiperda 9 (Sf9) cells, the product of LRV1-4 ORF2 self-assembles into virus-like particles morphologically identical to native virions, demonstrating that ORF2 encodes the major capsid protein (5). Sequence similarities to the RNA-dependent RNA polymerases of other double-stranded and plus-strand RNA viruses further imply that ORF3 encodes the viral RNA-dependent RNA polymerases.

Since a short RNA transcript was first identified, both as a product of an in vitro polymerase assay and as a by-product of natural virus infection in Leishmania spp. (8), studies have focused on understanding the nature of this transcript and mapping the precise cleavage site on the full-length RNA substrate. The cleavage site in LRV1-4 RNA was mapped by primer extension (19) to nucleotide 320 of the virus 5′ untranslated region (UTR). Subsequent gene expression studies identified the LRV1-4 capsid protein as the endoribonuclease responsible for the cleavage event (see references 20 and 21 for a review). As with many other endoribonucleases (10, 11), divalent cations have been shown to be essential for RNA cleavage in LRV (19), although their precise role has not been identified. The original RNA substrate developed for use in the in vitro cleavage assay contains 447 nucleotides derived from the 5′ UTR of a full-length LRV1-4 transcript (19). Crude boundary mapping subsequently identified a 226-nucleotide RNA fragment that retains all determinants necessary to accurately target cleavage to the wild-type site (18).

Specific determinants within the 5′ UTR that contribute to cleavage specificity have not yet been elucidated, although it is clear that the presence of the consensus cleavage sequence alone is insufficient (18). In addition to conserved nucleotide sequences, the 5′ UTR of LRV1-4 transcripts is predicted to contain five conserved stem-loop structures (31). It has been hypothesized that the structures, not yet proven to exist, may be important in translation and/or targeting of viral transcripts for cleavage. Here, we delineate the minimal essential UTR sequence required for precise RNA cleavage and define by ribonuclease mapping and site-specific mutagenesis a critical stem-loop structure within that sequence. We also provide evidence of a role for Mg²⁺ in stabilizing the stem-loop to facilitate accurate RNA cleavage at the downstream target sequence.

MATERIALS AND METHODS

Parasite strains and cell culture.

Leishmania guyanensis stock MHOM/BR/75/M4147 (M4147) was grown at 23°C in M199 semidefined medium (GIBCO Laboratories) supplemented with 5% fresh, filter-sterilized human urine (1).

Virus purification.

Leishmaniavirus virions were purified as previously described (8). Briefly, Leishmania promastigotes (∼10¹⁰ cells) were harvested in early stationary phase, washed, and lysed in 1% Triton X-100. Cell lysates were fractionated on 10 to 40% sucrose gradients, and fractions containing the peak of viral double-stranded RNA were used in cleavage assays. Virus-like particles produced in a recombinant baculovirus system expressing the LRV1-4 capsid gene were purified as previously described (5). Native LRV1-4 virion and recombinant virus-like particles are both functional in the in vitro cleavage assay and yield identical cleavage products (20). The RNA cleavage assays presented here were done with the recombinant LRV capsid, unless otherwise indicated.

Deletion mutagenesis.

The parental plasmid pBSK-Full14 encodes a full-length cDNA copy of LRV1-4 under the control of a T7 transcriptional promoter (28) and was used as a template to generate several different LRV1-4 5′ deletion mutants by PCR. To construct a series of deletion mutants from the 5′ end of the LRV1-4 UTR region, pBSK-Full14 was amplified by PCR with Taq DNA polymerase (Boehringer Mannheim Biochemicals) and a pair of synthetic oligonucleotide primers (one of the 5′ M-series primers and the 3′ M-342 primer) (Table 1) according to the manufacturer's instructions. The desired PCR product was captured in transcription vector pCRII (Invitrogen), and the plasmid having a correctly sized insert was selected and digested with restriction enzymes XhoI and EcoRI (Boehringer). The small XhoI/EcoRI-digested fragments each having deleted 5′ ends of the LRV1-4 UTR region were gel purified and ligated into the large pBSK-Full14 fragment cut with XhoI and EcoRI by use of T4 DNA ligase (Boehringer). To show that the 5-bp stem structure of stem-loop IV is essential for accurate RNA cleavage by LRV1-4 capsid protein, three mutant RNAs of that region were generated by in vitro transcription from the plasmids constructed by PCR-directed mutagenesis with one of the SL4-M-series primers and the 3′ M-342 primer (Table 1). All constructs were verified by DNA sequencing.

TABLE 1.

Oligonucleotide primers used

Primera	Sequence (5′→3′)b
5′ M-38	ACTCGAGTAGCTGTCCGGATGG
5′ M-189	ACTCGAGACATCCTACATTTATG
5′ M-232	ACTCGAGCTGGTTGTATCCAGG
5′ M-285	ACTCGAGAATCCAATATGCTGACTAC
3′ M-342	GGAATTCAAGAAACTTGCTTACG
SL4-M1	ACTCGAGACTGCCGCGAGCGTAAGGGACACAGTTGGCAG
SL4-M2	ACTCGAGACTGCCGCGAGCGTAAGGGAGTGTTTTGGCATTGTGAATCCAATATG
SL4-M3	ACTCGAGACTGCCGCGAGCGTAAGGGACACAGTTGGCATTGTGAATCCAATATG

^aThe primer names are based on the nucleotide position from the LRV1-4 5′ end. 

^bViral sequences are underlined, and replaced nucleotides are in italic type. 

Synthesis of RNA cleavage substrates.

Plasmid DNA templates were linearized by digestion with EcoRI, and transcription was accomplished by use of an in vitro reaction with T7 RNA polymerase (Promega) according to the manufacturer's protocol. Transcription reaction mixtures were incubated at 37°C for 2 h, and template DNA was removed by treating the reaction mixtures with RQ1 DNase (Promega) for an additional 15 min at 37°C. The RNA product was treated with calf intestine phosphatase (New England Biolabs), extracted twice with phenol-chloroform, and precipitated in ethanol.

End-labeling reactions and purification of labeled RNA.

Dephosphorylated RNA was 5′ end labeled with T4 polynucleotide kinase (New England Biolabs) and [γ-³²P]ATP (New England Nuclear Corp.) as previously described (22). Kinase reaction mixtures contained 70 mM Tris-HCl (pH 7.6), 10 mM MgCl₂, 5 mM dithiothreitol, 0.6 μM [γ-³²P]ATP (6,000 Ci/mmol), and 1 to 15 μM RNA. After incubation at 37°C for 2 h, phenol-chloroform-extracted RNA was fractionated on a 0.4-mm-thick 5% polyacrylamide–8.3 M urea gel (National Diagnostics). RNAs located by UV shadowing were purified as previously described (22).

In vitro RNA cleavage assay.

The cleavage assay was performed as previously described (19). Briefly, RNA cleavage activity was assayed by use of 20-μl reaction mixtures containing substrate RNA (100,000 cpm), sucrose-purified viral particles (approximately 9 μg of total protein), and 20 U of RNasin (Promega). In some studies, various cations were added to the cleavage reaction mixtures at various concentrations. In other experiments, equal amounts (10 μg of total protein) of an M4147 sucrose gradient fraction containing LRV1-4 viral particles were preincubated with LRV1-4 capsid protein-specific antiserum or preimmune serum (5) at room temperature for 10 min prior to the initiation of the RNA cleavage reaction. Incubation was done at 37°C for 40 min to 1 h unless otherwise indicated. Reactions were terminated by extraction with phenol-chloroform. Portions of the reaction mixtures were mixed with formamide loading dye and heat denatured at 90°C for 2 min. Reaction products were resolved on a denaturing 8% polyacrylamide–8.3 M urea gel and visualized by autoradiography.

Base-specific endoribonuclease mapping analysis.

Endoribonuclease mapping reactions (10-μl total volume) with RNase T₁ and V1 (Pharmacia) were performed as previously described (29) with minor modifications. Briefly, the reaction mixtures contained 200,000 cpm (Cerenkov) of 5′-end-labeled RNA and various quantities of diluted RNase. RNase T₁ reaction mixtures contained 40 mM Tris-HCl (pH 8.0), 1 mM EDTA, 0.25 mg of yeast tRNA (Sigma) per ml, and 0 to 10 mM MgCl₂. RNase V1 reaction mixtures contained 25 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 200 mM NaCl, and 0.25 mg of yeast tRNA per ml. Unless otherwise indicated, reactions were terminated after 14 min at 37°C by the addition of an equal volume of formamide loading dye. Reaction products were stored on dry ice until electrophoresis on 8 or 10% polyacrylamide–8.3 M urea gels. RNase V1 reactions were allowed to proceed for 2 min at 37°C prior to termination. Sequencing ladders were generated by partial digestion of the 5′-end-labeled RNAs with RNase T₁ according to the manufacturer's protocol. Alkaline hydrolysis was performed by heating the RNA at pH 9.0 for 5 min at 100°C (12).

RESULTS

The minimal essential sequences for precise RNA cleavage reside in viral nucleotides 249 to 342.

FOLD analysis (Genetics Computer Group) predicts five conserved stem-loop structures in the 5′ UTR of LRV1-1 and LRV1-4 transcripts (31), four of which reside upstream of the putative cleavage consensus sequence (Fig. 1A). To determine whether the structures constitute determinants for RNA cleavage at the downstream consensus site, in vitro transcripts generated from DNA constructs with successive stem-loop deletions were examined in the RNA cleavage assay. Transcripts which encode only stem-loop IV (RNA 5′232-342) showed cleavage at the wild-type site (Fig. 1B). Analysis of several additional 5′ deletion mutants further delineated the minimal RNA sequence required for optimal RNA cleavage activity to viral nucleotides 249 to 342 (see Fig. 3). A transcript containing viral nucleotides 251 to 342 showed reduced cleavage (data not shown), and another lacking stem-loop I through stem-loop IV (RNA 5′285-342) was not cleaved.

FIG. 1.

Conserved stem-loop structures predicted for the 5′ UTR of LRV1-4 transcripts and cleavage products obtained after deletion of the conserved sequences. (A) FOLD analysis predicts four conserved stem-loop structures (I to IV) in the 5′ terminus of LRV1-4 transcripts, upstream of the RNA cleavage site (asterisk). A putative consensus RNA cleavage sequence is shown (box). (B) The indicated 5′-end-labeled deletion mutant RNA transcripts were incubated in the absence (lane 1) or presence (lane 2) of sucrose-purified LRV1-4 viral particles for 40 min. Sequencing ladders generated by partial digestion of each mutant RNA with RNase T₁ are shown in lane 3. Molecular sizes are indicated (in nucleotides) on the right.

FIG. 3.

RNA cleavage assay and RNase T₁ mapping of stem-loop IV mutant RNAs. (A) The indicated 5′-end-labeled mutant RNA transcripts were incubated for 60 min in the absence (lane 1) or presence (lane 2) of sucrose-purified LRV1-4 viral particles. Sequencing ladders (lane 3) were generated by partial digestion of each mutant RNA with RNase T₁. Molecular sizes are indicated (in nucleotides) on the right. (B) The identity of the relevant RNA substrate is indicated above the gel (see Fig. 2 for complete descriptions). Reactions with RNase T₁ were performed under native conditions and contained 0, 2, or 10 mM Mg²⁺ ions. The final concentration of RNase T₁ used in all reactions was 5 × 10⁻³ U/μl. The randomly cleaved RNA ladder was generated by alkaline hydrolysis (OH). Sequencing ladders (lane D) were generated by partial digestion of each RNA by RNase T₁ under denaturing conditions. The positions of important G residues and stem IV are indicated.

Stem-loop IV is an essential determinant for capsid-dependent RNA cleavage.

A panel of RNA transcripts in which putative stem-loop IV was eliminated through site-specific nucleotide substitutions and then reconstructed through complementary substitutions (Fig. 2) was prepared. RNA 5′SL4-M1 is a derivative of the parental transcript (RNA 5′249-342) in which sequences along the left side of the putative stem (5′ GUGUU 3′) were replaced by others identical to those along the right side of the stem (5′ GACAC 3′). In RNA 5′SL4-M2, sequences on the right side of the stem (5′ GACAC 3′) were replaced by sequences identical to those on the left. RNA 5′SL4-M3 is identical to the parental transcript, except that nucleotide sequences encoding the left and right sides of the stem are interchanged. Analyses of the RNA transcripts showed that mutants SL4-M1 and SL4-M2, in which the stem structure was eliminated by nucleotide replacements, were not substrates for capsid-dependent RNA cleavage, while the double mutant (SL4-M3) exhibited wild-type cleavage activity (Fig. 3A). Other derivatives with potential to form only a 3- or a 4-bp stem showed diminished activity in the RNA cleavage assay (data not shown).

FIG. 2.

Schematic of RNA constructs used to examine the role of stem-loop IV in RNA cleavage. The identities of the mutant RNAs are indicated above each diagram.

The results shown above support the notion that RNA 5′249-342 encode structural determinants that target RNA cleavage to the downstream consensus sequence. Nuclease mapping studies with RNase T₁ (cuts after unpaired guanosine (G) residues [13]) were performed to characterize structural elements in both cleavage-competent and -incompetent RNA transcripts. Because Mg²⁺ ions are essential to many endoribonucleases (11), including that of the LRV capsid (19), it was of interest to study the effects of Mg²⁺ on the structures detected by nuclease mapping. The results (Fig. 3B) showed that the relevant G residues in RNA 5′249- 342 (parental construct) (G-269, G-271, and G-280) and the double mutant SL4-M3 (G-282 and G-284) were RNase T₁ resistant, confirming the existence of stem-loop IV in the two cleavage-competent transcripts. The formation and/or stability of the stem-loop structure depended on the presence of Mg²⁺ ions in the assay mixture (compare Fig. 3B, T1, lanes 0, 2, and 10). Results obtained with RNase V1, which distinguishes residues present in a helical conformation (17), were also consistent with the presence of a double-stranded RNA structure (data not shown). In contrast to the findings obtained with cleavage-competent transcripts, the relevant G-residues in the cleavage-resistant RNAs 5′SL4-M1 (G-280) and 5′SL4-M2 (G-269, G-271, G-282, and G-284) were highly susceptible to RNase T₁, supporting the absence of base pairing at those residues. Taken together, the results show that stem-loop IV exists in cleavage-competent (but not cleavage-incompetent) transcripts, that the presence of Mg²⁺ stabilizes the structure, and that the presence of the structure imparts an element of specificity to the capsid-dependent RNA cleavage reaction.

Effects of divalent cations on capsid-dependent RNA cleavage.

MacBeth and Patterson (19) previously reported that EGTA-treated sucrose-purified particles lose the ability to generate the short transcript in polymerase assays. Other endoribonucleases are also known to require Mg²⁺ ions for activity (11). To test the effect of Mg²⁺ on capsid-dependent RNA cleavage, increasing amounts of Mg²⁺ ions were added to RNA cleavage reaction mixtures containing LRV1-4 viral particles. The results (Fig. 4A) showed that otherwise susceptible RNA sequences were poor substrates for cleavage in the absence of added Mg²⁺ ions and that cleavage activity was enhanced by the addition of Mg²⁺ to the reaction mixtures. Mg²⁺ at concentrations greater than 20 mM inhibited RNA cleavage. Other divalent cations were also tested for their ability to fulfill the role of Mg²⁺ in the cleavage reaction (Fig. 4B). The results showed that Ca²⁺ and, to a lesser extent, Mn²⁺ also enhanced RNA cleavage, while results obtained with Mn²⁺ or Zn²⁺ ions were inconclusive. Neither NH₄⁺ nor dithiothreitol (data not shown) could substitute for Mg²⁺ in these studies. EDTA completely abolished RNA cleavage, indicating that weak activity detected in the absence of added cations may result from residual Ca²⁺ in preparations of sucrose-purified viral capsids.

FIG. 4.

Effect of cations on RNA cleavage by LRV1-4 viral particles. (A) Transcripts of 5′-end-labeled RNA 5′249-342 were incubated with the indicated concentrations of Mg²⁺ ions and sucrose-purified LRV1-4 viral particles for 40 min as described in Materials and Methods. A sequencing ladder was generated by partial digestion of RNA 5′249-342 with RNase T₁ (lane T1). The expected location of the cleavage product is marked by an arrow. Numbers at left are nucleotides. (B) The cleavage activity of LRV1-4 viral particles on RNA 5′249-342 was examined in the presence of various cations or EDTA. The indicated reagent (10 mM final concentration) was added to the cleavage assay reaction mixture in the presence (+) or absence (−) of sucrose-purified LRV1-4 viral particles. Lane T1 contains a sequencing ladder generated by partial digestion of the RNA by RNase T₁. The RNA cleavage product is indicated by an arrow. Numbers at left are nucleotides.

Effects of capsid-specific antisera on RNA cleavage.

To provide supportive evidence that the LRV1-4 capsid encodes the endoribonuclease detected in these and previous studies, cleavage reactions with transcripts of RNA 5′249-342 were performed either as described above or in the presence of preimmune serum or capsid-specific antiserum. A time course experiment under standard reaction conditions (Fig. 5A) showed that the quantity of the 320-nucleotide cleavage product was increased with increasing reaction times. Some additional (smaller) cleavage products were detected after prolonged incubation (Fig. 5A, compare lanes 2 and 5); however, these nonspecific products were also formed from reactions supplemented with cell extracts prepared from wild-type baculovirus (AcMNPV)-infected Sf9 insect cells (data not shown). To demonstrate an active role for the LRV1-4 capsid in specific RNA cleavage, sucrose gradient-purified authentic LRV1-4 viral particles were incubated with LRV1-4 capsid-specific antiserum (5) or preimmune serum before the RNA cleavage assay was initiated (Fig. 5B). Incubation of native LRV1-4 with preimmune serum before initiation of the RNA cleavage assay had little effect on specific RNA cleavage (Fig. 5B, lane 2) relative to a buffer control (lane 1). In contrast, LRV1-4 particles preincubated with capsid-specific antiserum showed greatly reduced cleavage activity (Fig. 5B, lane 3). Increasing titers of antiserum abolished specific cleavage activity but had no effect on nonspecific nuclease activity (data not shown). These results are a further indication that the viral capsid, rather than a copurifying contaminant, encodes the site-specific endoribonuclease responsible for the RNA cleavage activity detected in these and other studies of LRV1-4.

FIG. 5.

Effects of virus capsid-specific antiserum on RNA cleavage. (A) Transcripts of 5′-end-labeled RNA 5′249-342 were incubated in the absence (lane 1) or presence (lanes 2 to 5) of sucrose-purified LRV1-4 viral particles. The reactions were stopped at the indicated times by phenol-chloroform extraction, and the products were analyzed as described in Materials and Methods. A sequencing ladder generated by partial digestion with RNase T₁ is shown (lane T1). The position of the expected 320-nucleotide product is indicated by an arrow. Numbers at left are nucleotides. (B) A 5-μl aliquot of a sucrose gradient fraction containing authentic LRV1-4 viral particles (10 μg of total protein) was preincubated with an equal amount of nuclease-free water (lane 1), preimmune serum (lane 2), or LRV1-4 capsid protein-specific antiserum (lane 3). After 10 min of incubation at room temperature, the RNA cleavage activity was assayed with transcripts of 5′-end-labeled RNA 5′249-342 as described in Materials and Methods. Sequencing ladders generated by partial digestion with RNase T₁ (lane T1) are shown, and the position of the expected 320-nucleotide product is indicated by an arrow. Numbers at left are nucleotides.

DISCUSSION

Generally, viral capsid proteins serve a variety of functions that ensure the successful propagation of viral genomes and protect viral genomes from nucleases in the intracellular and extracellular environments. Identification of the LRV capsid protein as an endoribonuclease, however, is unprecedented among known viral capsid proteins. MacBeth (18) previously demonstrated that an RNA template encoding only 74 viral nucleotides, including the wild-type LRV1-4 consensus sequence, was not cleaved, indicating that the presence of the consensus site alone is insufficient to target endoribonucleolytic processing.

We have extended those previous observations through deletion mutagenesis studies in an effort to delineate the minimal essential RNA determinants necessary and sufficient for capsid-dependent cleavage. The studies presented here demonstrate that the minimal essential determinants for accurate RNA cleavage reside in nucleotides 249 to 342 of the LRV1-4 RNA genome. A hypothesis that conserved secondary structures predicted in the 5′ UTR of LRV1-4 (31) may constitute essential determinants for capsid-dependent RNA cleavage led us to perform structural mapping studies on the minimal RNA transcripts in solution. The studies presented here used nuclease mapping to document the existence of stem-loop IV and to show that disruption of the structure abolishes capsid-dependent RNA cleavage in LRV. The essential RNA structure comprises a 5-bp stem and a six-ribonucleotide hairpin loop (stem-loop IV) apparently conserved between LRV1-1 and LRV1-4 RNAs. Mutants lacking the structure were completely inactive in the cleavage assay, while those with a 3- or 4-nucleotide stem showed intermediate activity.

The structure of RNA 5′249-342 obtained by RNase mapping analyses in the absence of divalent cations (Mg²⁺ or Ca²⁺) was different from the structure predicted by FOLD analysis of the RNA sequences, confirming that RNA secondary structures can vary according to the experimental conditions. However, when Mg²⁺ ions were added to the reaction, RNase T₁ mapping yielded a structure nearly identical to that proposed by FOLD analysis, indicating that Mg²⁺ ions enhanced the formation of RNA double helices. Similarly, a potential pseudoknot structure generated by RNA oligonucleotides in vitro can be translated into two different hairpin structures, depending on the experimental conditions, such as ionic strength, ambient temperatures, metal ions, loop size, or loop sequences (37). Nuclear magnetic resonance studies have also shown that tRNA can exist in two different conformations, depending on the salt concentration (27).

The effects of divalent ions, especially Mg²⁺ ions, on endoribonucleases and RNA self-splicing enzymes (ribozymes) can be summarized according to their role(s) in catalytic activity and/or their interactions with RNA. The highly specific endoribonucleases that participate in RNA processing and turnover require divalent Mg²⁺ ions for catalysis (10). For example, an RNase which specifically degrades RNA in RNA-DNA hybrid structures requires Mg²⁺ ions and the presence of a sulfhydryl reagent (dithiothreitol) for maximal activity. The requirement for Mg²⁺ ions can be only partially replaced by Mn²⁺ (3). In RNase Q, the enzyme activity is stimulated by monovalent cations, such as K⁺ and NH₄⁺ at low concentrations. Mn²⁺ ions at a concentration of 0.1 mM are stimulatory but become inhibitory at higher concentrations (34). RNase E found in Escherichia coli and processed p5 rRNA needs Mg²⁺ or Mn²⁺ ions for activity, and this requirement cannot be fulfilled by Ca²⁺ or Zn²⁺ (24). Metal ions also play a crucial role in the catalytic activity of all characterized ribozymes. A number of approaches to monitoring the binding of metal ions to nucleic acids and to understanding models for ribozymatic cleavage have been described; these include X-ray crystallography (14, 33), biochemical techniques (9), and nuclear magnetic resonance spectroscopy (36). The presence of divalent metal ions (Mg²⁺ or Mn²⁺) is essential for pre-NanGIR1 (Naegleria andersoni group I ribozyme 1) activity. The failure of Ca²⁺, monovalent ions, or polyamines to substitute suggests that the cofactor is not simply a structural requirement and may instead participate directly in the hydrolysis reaction (15).

Here we demonstrated that divalent cations were an essential component of the capsid-dependent RNA cleavage reaction. While Mg²⁺ could be replaced by Ca²⁺ in this system, Mn²⁺ or Zn²⁺ ions were ineffective, indicating specificity in the requirement for a divalent ion. Divalent ions have been shown to affect the formation and stability of three-dimensional structures in many other types of RNA, including tRNA (4, 26), the Tetrahymena ribozyme RNA (6, 23, 38), and Bacillus subtilis P RNA (2, 25). Here, we showed by RNase mapping that Mg²⁺ ions stabilized specific regions (conserved stem-loop IV) of the RNA substrates for the LRV endoribonuclease. Electrophoretic mobility shift assay studies (data not shown) also showed that the presence of metal ions (Mg²⁺ or Ca²⁺) caused a mobility change in the RNA cleavage substrate (5′249-342). Nuclease mapping studies demonstrated that intact stem-loop IV was required for accurate RNA cleavage by the LRV1-4 capsid endoribonuclease. The changes in secondary or tertiary structure induced by point mutations within stem-loop IV correlated directly with a loss of cleavage specificity, suggesting that the conformational structure stabilized by divalent metal ions (Mg²⁺ or Ca²⁺) is essential for recognition and/or accurate RNA cleavage in this system. It remains unknown whether divalent cations might also be involved in RNA catalysis.

It is noteworthy that a specific structure (stem-loop IV) stabilized by metal ions (especially Mg²⁺ or Ca²⁺) is an essential component for capsid-dependent RNA cleavage. It remains unclear precisely how stem-loop IV imparts nucleotide specificity to the cleavage reaction. Chelladurai et al. (7) showed that in the absence of Mg²⁺, wild-type RNase III cannot engage R1.1 RNA in a stable gel-shifted complex, indicating that Mg²⁺ ions significantly enhance substrate R1.1 RNA binding to the E. coli RNA-processing enzyme, RNase III. Another gel shift assay, in which Ca²⁺ was substituted for Mg²⁺, provided a mobility similar to that of the Mg²⁺-stabilized complex, although Ca²⁺ was inactive as a catalytic cofactor (16). Our electrophoretic mobility shift assay (data not shown), however, revealed that there was no binding enhancement for the substrate RNA between Mg²⁺ ions and viral capsid protein.

We do not know yet whether fully assembled virus particles or unassembled capsid proteins exhibit endoribonuclease activity. Previously, MacBeth and Patterson (19) showed that CsCl₂-purified or EGTA-treated virus particles, which lack endoribonuclease activity, are partially disassembled. Also, a capsid mutant lacking 24 amino acid residues at the amino terminus is unable to self-assemble and is inactive in the RNA cleavage assay (Y.-T. Ro and J. L. Patterson, unpublished data). Disassembly of the viral capsid could induce a conformational change that alters cleavage activity or results in the dissociation of a cofactor that is required for cleavage. A comparison of RNA cleavage activity between disassembled and reassembled virus particles may provide further insight into the nature of the LRV capsid endoribonuclease.

The cleavage of viral transcripts by the capsid protein could be a mechanism by which the virus controls its gene expression. Removal of the 5′ end could presumably affect RNA stability, RNA packaging, replication, and/or translation. A recent study shows that Leishmania cytoplasmic proteins bind specifically only to the cleaved LRV1-4 RNA, not to the uncleaved RNA, in a gel mobility shift assay and a UV-cross-linking study (30). This interesting observation suggests that RNA cleavage alters the functionality of viral transcripts and that the cleavage of full-length transcripts unmasks a cryptic domain which is now accessible to bind host factors. This functional change in transcripts after RNA cleavage may affect the efficiency of the translation of the viral gene products, with either enhancement or inhibition. An understanding of the precise role of RNA cleavage in the life cycle of LRV awaits more direct evidence.