|Home | About | Journals | Submit | Contact Us | Français|
Retroviruses selectively package two copies of their RNA genomes by mechanisms that have yet to be fully deciphered. Recent studies with small fragments of the Moloney murine leukaemia virus (MoMuLV) genome suggested that selection may be mediated by an RNA switch mechanism, in which conserved UCUG elements that are sequestered by base pairing in the monomeric RNA become exposed upon dimerization to allow binding to the cognate nucleocapsid (NC) domains of the viral Gag proteins. Here we show that a large fragment of the MoMuLV 5'-untranslated region (5'-UTR) that contains all residues necessary for efficient RNA packaging (ΨWT, residues 147–623) also exhibits dimerization-dependent affinity for NC, with the native dimer ([ΨWT]2) binding 12 ± 2 NC molecules with high affinity (Kd = 17 ± 7 nM) and the monomer, stabilized by substitution of dimer-promoting loop residues by hairpin-stabilizing sequences (ΨM), binding 1–2 NC molecules. Identical dimer-inhibiting mutations in MoMuLV-based vectors significantly inhibit genome packaging in vivo (~100-fold decrease), whereas a large deletion of nearly 200 nucleotides just upstream of the gag start codon has minimal effects. Our findings support the proposed RNA switch mechanism, and further suggest that virus assembly may be initiated by a complex comprising as few as 12 Gag molecules bound to a dimeric packaging signal.
During the late phase of retroviral replication, two unspliced RNA genomes are selected for packaging into assembling virions from a cytoplasmic milieu containing a substantial excess of cellular and spliced viral mRNAs1. Both RNA molecules are utilized as templates during reverse transcription, promoting genetic diversity through recombination for heterozygous particles and allowing RNAs with otherwise deleterious break points to template intact proviruses2–4. Genome selection is mediated by the nucleocapsid (NC) domain of the viral Gag polyprotein, which specifically recognizes packaging elements within or near the 5’-untranslated region (5’-UTR) of the viral RNA1,5–7. Although well over a thousand copies of Gag assemble into each virion8–11, the number of Gag molecules that participate in genome recognition is unknown. In mature virions, genomes exist as non-covalently linked dimers, and sequences that promote dimerization in vitro also reside in the 5’-UTR and generally overlap with packaging elements12,13. Two genomes are packaged even under conditions in which only subsets of virions contain viral RNA14,15, and although small quantities of monomeric RNAs can be isolated from freshly budded particles under mildly denaturing conditions16, genetic studies provide strong evidence that the genomes form dimers (possibly weakly-linked dimers16) prior to packaging17–19, consistent with early proposals20,21.
The Moloney Murine Leukemia Virus (MoMuLV) is a prototypical retrovirus that has been extensively studied and widely used as a vector for gene delivery1. The virus is closely related to Xenotropic murine leukemia virus-related virus (XMRV), a recently discovered human pathogen22 linked to prostate cancer23 (although this link has recently been questioned24) and Chronic Fatigue Syndrome25. Previous nucleotide accessibility mapping experiments, phylogenetic analyses, in vivo mutagenesis, and free energy calculations indicate that the MoMuLV 5’-UTR consists of a series of stem loops connected by short linkers, and that the secondary structure changes upon dimerization26–29 (Fig. 1). This finding led to early suggestions that dimerization-dependent conformational changes might regulate genome packaging and other UTR-dependent functions26–29. Similar RNA structural switch mechanisms have also been proposed for other retroviruses, including HIV-130–34. Dimerization and packaging are promoted by four stem loops: DIS-1 and DIS-2, which form intermolecular duplexes26,35,36, and SL-C and SL-D, which can form intermolecular “kissing contacts” between their conserved GACG tetraloops37,38 (Fig. 1). In studies of small RNA fragments, DIS-1 and DIS-2 were shown to undergo register shifts in base pairing upon dimerization, exposing conserved UCUG elements that are capable of binding NC stoichiometrically and with high affinity39. These elements were sequestered by base pairing and unable to bind NC in the monomeric RNAs39. The 5’-UTRs of gammaretroviruses are enriched with UCUG and related Py-Py-Py-G (Py = pyrimidine) sequences39, all of which are potentially capable of binding NC with high affinity40. Early chemical accessibility experiments indicated that these elements are sequestered by base pairing in the monomeric MoMuLV 5’-UTR, but some exhibited increased reactivity and were predicted to become exposed upon dimerization26–29.
The above findings suggested that MoMuLV genome packaging may be regulated by a structural RNA switch mechanism, in which an unknown number of NC binding sites are sequestered by base pairing in the monomeric RNA and become exposed upon dimerization to allow NC binding and promote packaging39. To test this hypothesis, we conducted quantitative NC binding studies in vitro, and RNA packaging experiments in vivo, with RNAs that contain both the native packaging signal and mutations that inhibit dimerization. Our findings support the proposed RNA switch mechanism, and suggest that virus assembly may be initiated by a complex of 12 Gag molecules bound to a dimeric packaging signal.
An RNA construct that spans from the primer binding site through the gag start codon and includes residues shown to function as an independent packaging signal41 (ΨWT, residues 147–623), was prepared by in vitro transcription for NC binding studies (Fig. 2a). ΨWT contains the four sequences known to promote dimerization (DIS-1, DIS-2, SL-C and SL-D), and incubation at 50 °C for 30 min led to stoichiometric conversion of the RNA from an initial monomeric species to a dimer (Fig. 2b).
NC is highly basic (+10 overall charge at neutral pH) and binds RNA weakly and in a salt-dependent manner if high affinity sites are unavailable. We therefore initially probed for high affinity NC binding using a filtration assay, in which solutions containing dimeric ΨWT were first titrated with a large excess (50×) of NC, and then extensively washed with salt solutions by centrifugal filtration to remove weakly-bound NC molecules. After five washes with buffers approximating physiological ionic strength (160 mM NaCl; 10-fold v/v dilutions per wash), a total of 14 ± 2 NC molecules was retained with the dimeric RNA (Fig. 2c,d).
The affinity and stoichiometry of NC binding to ΨWT was also measured by isothermal titration calorimetry (ITC). Titrations with NC gave rise to negative binding enthalpies consistent with results obtained for short RNAs that contain a single high affinity NC binding site40,42. Non-linear least squares fitting of the ITC data afforded 1:12 (dimer:NC) binding isotherms characterized by a single class of binding sites with dissociation constant (Kd) of 17 ± 7 nM. Essentially identical results were obtained using physiological-like buffer43 (140 mM KCl, 10 mM NaCl, 1 mM MgCl2; data not shown). Thus, the quantitative ITC data are consistent with the qualitative results obtained using the NC filtration assay, and indicate that [ΨWT]2 binds approximately 12 NC molecules with high affinity.
NC is an efficient RNA chaperone that readily promotes retroviral genome dimerization12. As such, addition of NC to the monomeric form of ΨWT, prepared by boiling and rapid cooling under dilute conditions, resulted in spontaneous dimerization of the RNA, and we were unable to directly measure the affinity of NC for ΨWT in its monomeric state. Previous studies showed that dimerization of DIS-1 and DIS-2 oligoribonucleotides can be inhibited by substituting their native, dimer-promoting loop nucleotides by GAGA (a member of the GNRA family44; N = any nucleotide and R = purine), and that kissing-mediated dimerization37 of SL-C and SL-D can be blocked by substituting their conserved GACG tetraloops by GUGA and GAGG, respectively39,42,45. These substitutions inhibit dimerization without altering the structures of the native, monomeric hairpins39,42,45. As expected, corresponding substitutions in Ψ (ΨM, Fig. 2a) blocked dimerization under conditions that favored dimerization of ΨWT (Fig. 2b).
Significantly, the ΨM RNA exhibited reduced affinity for NC compared to ΨWT, retaining only one NC molecule per RNA strand after treatment with excess NC and successive salt washes (Fig. 2c,d). It is important to note that the loop residues of the native hairpin RNAs did not exhibit detectable affinity NC in previous NMR and ITC-detected titration experiments39; therefore, the loop mutations of the current studies should not directly interfere with NC binding. Native PAGE experiments confirmed that ΨM remained monomeric in solution, even after the NC incubation and salt washes. The ITC NC titration profile observed for ΨM was also considerably different from that of ΨWT, exhibiting classical two-component behavior with an initial exothermic component and an overlapping endothermic component (Fig. 2e). Fitting of the ITC profiles afforded NC:ΨM stoichiometries of 1:1 to 2:1 (Kd = 29 ± 4 nM) for the exothermic component. The endothermic component is similar in appearance to ITC data obtained for NC titrations with RNAs that lack high affinity binding sites, including homopolymeric RNAs and tRNAs40,46. These findings are consistent with NC-RNA binding results obtained using fluorescence-detected assays47, and are indicative of weak, non-specific binding and NC-stimulated RNA unfolding46–48.
The effects of the above Ψ mutations on packaging were assessed initially by quantifying RNA packaging levels in virions produced by transient transfection. An MoMuLV-based vector construct, pGPP, containing the 5’ terminal two-thirds of the MoMuLV genome, including the 5’-UTR, gag and pol genes, and a puromycin N-acetyltransferase expression cassette in place of env 49 was used to generate both viral proteins and genomic RNAs (gRNAs) (Fig 1a). Derivatives of pGPP that contain the wild-type 5’-UTR and 5'-UTRs with loop mutations identical to those used in the in vitro NC binding studies were expressed in 293T cells, and gRNA levels in the cells and the resulting virions were determined by RNase protection assay via comparison with 7SL RNA, a host RNA that is packaged in HIV-1 and MoMuLV at levels proportionate to virion proteins50,51. Intracellular levels of wild type and mutant gRNAs differed by less than three-fold (Fig. 3a), and virus production (as determined by reverse transcriptase (RT) activity) was roughly proportionate to intracellular viral RNA levels, thus indicating that the 5’ UTR mutations did not significantly affect transcription, translation, or virion assembly. However, deletion of either the core encapsidation signal (Δ215–367, ΨΔCES) or the extended packaging region (Δ215–568, Ψ−) 41 reduced RNA packaging to roughly 1% of wild type levels, whereas deletion of a large fragment located immediately downstream of the core encapsidation signal (Δ375–568, ΨΔ193) decreased packaging by less than two-fold (Fig3b and 3c). In earlier studies we showed that 7SL RNA is encapsidated in MoMuLV in proportion to virion protein50, and as expected, gRNA packaging levels measured by quantifying gRNA per virion, normalized to either RT or to 7SL RNA packaging, yielded very similar results (Fig 3b and 3c). Both measures indicated that the ΨM is packaged at ~8% of wild type levels. Thus, the UTR variants that were dimerization defective in vitro were also packaging defective in virus. When normalized by RT levels, the puromycin-resistant colony forming unit (cfu) titer of ΨM mutant virion-containing media was about 10-fold less than either wild type orΨΔ193, and both Ψ− and ΨΔCES titers were reduced an additional 10- to 20-fold (not shown). Thus, reverse transcription was not altered by these 5’ UTR mutations
Although packaging levels of the ΨM mutants were reduced in the above experiments, the over-production of the mutant RNAs in transfected cells might have caused packaging levels to be artificially high, since retroviruses can efficiently package RNAs that lack Ψ sequences under conditions in which the native genomes are not present52. Therefore, we next measured packaging efficiencies using a competition assay, in which “test” genomic RNAs (gRNAs) containing either wild type or mutant 5’ UTRs were co-expressed with a control gRNA containing a wild type 5’ UTR, and ratios of test to control gRNAs were determined in both virus and cells (Fig. 3d). As shown in Fig. 3d, the presence of a wild type competitor gRNA depressed packaging of both the ΨM and ΨΔCES mutants to ~ 1% of wild type levels, suggesting that at least some Ψ mutant RNA packaging observed under the transient transfection conditions above was nonspecific.
Finally, gRNA packaging was assessed in virions produced by cells that contained single integrated proviruses. Pools of puromycin-resistant cells generated by infection with GPP derivatives at low (<0.01) multiplicity of infection were used to generate virions, and the virion and intracellular levels of the respective gRNAs were compared. As shown in Fig. 3e–3g, ΨM gRNA packaging levels were reduced 100-fold relative to wild-type levels under these conditions, which resemble those that occur during natural infection. These results are consistent with those obtained using the competition assay, and provide further evidence that much of the ΨM packaging observed under transient transfection conditions was due to non-specific packaging of the overproduced RNA.
The relatively efficient encapsidation of ΨM gRNAs under transient transfection conditions provided a way to generate ΨM gRNAs in sufficient quantities for examination on non-denaturing northern blots. This allowed us to test whether or not the encapsidated RNAs exist in the metastable dimer linkage characteristic of wild type MoMuLV gRNAs53. Vectors with ΨWT, ΨM, and ΨΔCES – containing 5’ UTRs were overexpressed by transient transfection, and virion RNAs were purified under non-denaturing conditions. Virion RNA samples normalized for gRNA content were subjected to heat treatments based on previously determined genome melting profiles53, and examined on non-denaturing northern blots (Fig. 4). The results showed that wild type gRNAs displayed properties as described previously15,20: running as a fairly discrete dimer when unheated, displaying a small amount of monomer but principally residual dimer at 58 °C, and denaturing fairly completely to the monomer form—albeit with a significant amount of the RNA degradation characteristic of retroviral gRNAs—when incubated at 65 °C. In contrast, unheated ΨM and ΨΔCES gRNAs appeared as diffuse, slowly-migrating bands that resolved mainly to monomers or a smear with reduced mobility at 58 °C, followed by nearly complete denaturation to the monomer form at 65 °C. These results are consistent with previous findings for a ΔDIS-1/ΔDIS-2 deletion mutant, which gave rise to a diffuse low mobility product rather than a discrete monomer at lower temperatures36. The thermal denaturation profile observed here is highly reminiscent of the so-called “tethering” interactions that have been described within and between co-packaged MoMuLV gRNAs in RNA regions outside the dimer linkage region53. Previous work has established that no significant RNA-RNA dimerization interactions form between MoMuLV gRNAs outside their 5’ ends15. Thus, these findings suggest that the slow migrating forms of both ΨM and ΨΔCES RNAs were retarded largely by non-specific interactions that occur during condensation of the RNAs in assembling virions rather than by specific, more stable interactions involving their 5’ UTRs.
Previous NMR studies showed that isolated DIS-1 and DIS-2 RNAs undergo register shifts in base pairing upon dimerization, exposing UCUG elements that are both strictly conserved at these positions and highly enriched (50-fold) within the 5’-UTR of the MoMuLV genome39. These and related Py-Py-Py-G elements are capable of binding the MoMuLV NC protein with high affinity when exposed in unstructured conformations, whereas sequences that lack the 3'-guanosine or contain purines at one or more of the upstream sites bind with substantially reduced (up to two orders of magnitude) and, in some cases, undetectably low affinities40. Thus, NC is not simply a single-stranded RNA binding protein - it binds preferentially to unstructured RNAs with exposed guanosines and is highly discriminative toward unstructured Py-Py-Py-G sequences40. Chemical accessibility mapping studies using traditional nucleobase protection assays indicated that nearly all of the Py-Py-Py-G elements are sequestered by base pairing in the monomeric genome, and that several of them become exposed upon dimerization29. These findings collectively suggested that diploid genome selection may be mediated by a dimerization-dependent RNA switch mechanism, in which the high-affinity NC binding sites are sequestered by base pairing in the monomeric Ψ-site and become exposed to promote packaging upon dimerization39. The present studies show that substitution of as few as 10 nucleotides in the dimer-promoting loops of the intact Ψ-site blocks dimerization in vitro and substantially attenuates NC binding, consistent with findings obtained for smaller fragments of the 5'-UTR39. The same substitutions in a MoMuLV-based vector significantly inhibit RNA packaging in vivo, whereas deletion of nearly 200 nucleotides between SL-D and the gag start codon (nearly 30 % of the 5'-UTR) has a minimal effect on packaging. Thus, the in vitro NC binding and in vivo packaging results obtained in the present studies provide strong support for the proposed dimerization-dependent RNA packaging mechanism39.
It is noteworthy that the packaging defect induced by the ΨM mutations was about 10-fold less severe when assessed using the transient transfection assay, as compared to the competitive and single-copy integration assays. Transient transfection has been widely employed in previous studies of retroviral genome packaging, and can produce cells that individually contain 100 or more copies of the vector. Since retroviruses efficiently package cellular RNAs if their native packaging signals are not available54,55, it is likely that the higher packaging of ΨM RNA in virons produced from transiently transfected cells is at least partly due to a non-specific, mass-action packaging effect. Consistent with this hypothesis, the RNAs that are packaged at marginal levels under transient transfection conditions exhibit aberrant mobility and thermal denaturation properties similar to those observed previously for non-specifically packaged RNAs that lack Ψ-sequences53.
Recently, a secondary structure for the monomeric packaging signal of the closely related Moloney murine sarcoma virus was proposed, based on measurements of ribose reactivity (called SHAPE56), in which residues of DIS-2 form part of a large unstructured loop that contains nearly 50 unpaired nucleotides and three exposed Py-Py-Py-G segments57. This proposed structure, which differs significantly from structures determined for the intact 5'-UTR using traditional nucleobase protection assays26–29, would be expected to bind four NC molecules with high affinity and appears incompatible with our quantitative NC binding studies and the proposed dimerization-dependent packaging mechanism. Since the present studies involve a large portion of the 5'-UTR, these differences are unlikely due to non-native local folding that can sometimes occur for small, isolated RNA fragments. Previous NMR studies have shown that the ribose moieties of nucleotides in non-canonical base pairs of the monomeric DIS-2 hairpin adopt conformations with significant non-C3'-endo populations. As such, these ribose groups might be expected to be reactive to SHAPE-based reagents, and this could explain the disparities between structures predicted using nucleobase- and ribose-modifying reagents. NMR studies of larger 5'-UTR constructs, which should definitively address these discrepancies, are underway.
Although NC readily promotes dimerization of the native Ψ-site in vitro, it is not yet clear how dimerization is induced in vivo, and cellular chaperones could play a role58. In addition, subsets of the Gag molecules of Rous Sarcoma Virus59–61, HIV-162, and MLV63 have been shown to transiently access the nucleus in some studies, possibly to recruit the viral genome for packaging59. It is thus possible that the NC domains of these Gag molecules might initiate genome dimerization in the nucleus, thereby enabling high affinity binding and promoting the trafficking of the Gag:genome complex out of the nucleus to virus assembly sites on the plasma membrane64–67.
Very recently, Simon, Bieniasz and co-workers used a fluorescence microscopy approach to directly detect association of the HIV-1 genome and Gag proteins at the plasma membrane68. These studies revealed that the dimeric (and not the monomeric) HIV-1 genome is targeted to virus assembly sites by a ribonucleoprotein complex that contains a small number of Gag proteins (likely a dozen or fewer), and that large numbers of Gag molecules subsequently accumulate at these sites68. Our finding that the dimeric MoMuLV ΨWT RNA contains 12 high-affinity NC binding sites is consistent with this observation. The initial assembly of these evolutionarily distant retroviruses, which contain dissimilar 5’-UTRs and NC domains, may thus be initiated by a similar complex comprising approximately a dozen Gag molecules bound to a dimeric packaging signal.
pMoMuLV-ΨWT: Polymerase chain reaction (PCR) was performed using plasmid pNCA42, which contains the proviral DNA of MoMuLV with oligonucleotide MoMuLV-147f (5'-CCGTCGAATTCTTAATACGACTCAC-TATAGGGGGCTCGTCCGGG-ATCGGGAGACCCCTG-3'), carrying an Eco RI site and a T7 RNA polymerase promoter, and oligonucleotide MoMuLV-MoMuLV-624r (5'-CCCTGAAGCTTCCCGGGCATATTCTCAGACAAATACAGAAACACAGTCAGACAGAGACAACACAG-3'), carrying a Sma I site and a Hind III site. The Sma I site was added to produce a blunt-end cut after the desired sequence with minimal incorporation of non-native sequences. The amplified product was inserted in the pUC19 plasmid after Eco RI and Hind III digestions (New England Biolabs). pMoMuLV-PsiAmBmCmDm: pMoMuLV-PsiAmBmCmDm was generated by PCR based site-directed mutagenesis using the plasmid pMoMuLV-PsiNative (from above) as a template for making the mutations in stem loops A, B, C and D. These mutations were designed to inhibit dimerization without affecting the known secondary structures of the loops36,42,45,69. Oligonucleotide MoMuLV-Amf (5'-GTAAGCTGGAGACAACTTATCTGTGTCTGTCCGATTGTCTAG-3') and oligonucleotide MoMuLV-Amr (5'-CTAGACAATCGGACAGACACAGATAAGTTGTC TCCAGCTTAC–3') were used to introduce the mutations in stem loops A. Oligonucleotide MoMuLV-BmCmDmf (5'-GTTGAGAAACTAGCTCTGTATCTGGCGGACCCGTGGTGGAACTGTGAAGTTCGGAACACCCGGCCGCAACCCTGGGAGAGGTCCC-3') and oligonucleotide MoMuLV-BmCmDmr (5'-GGGACCTCTCCCAGGGTTGCGGCCGGGTGTTCCGAACTTCACAGTTCCACCACGGGTCCGCCAGATACAGAGCTAGTTTCTCAAC-3') were used to introduce the mutations in stem loops B, C and D. PCR was performed with the oligonucleotide MoMuLV-147f and the oligonucleotide MoMuLV-Amr, and with the oligonucleotide MoMuLV-624r and the oligonucleotide MoMuLV-Amf, respectively. Each amplified product was separated on a 1.5% agarose gel, and purified using Gel extraction kits (Qiagen). These amplified products were mixed and were used as templates for another PCR with oligonucleotides MoMuLV-147f and MoMuLV-624r. Further PCR was performed on this amplified product, which contains mutations in SL-A, using the oligonucleotide MoMuLV-BmCmDmf and the oligonucleotide MoMuLV-BmCmDmr as above. This amplified product containing all mutations in stem loops A, B, C and D was digested by EcoR I and Hind III, and was inserted into pUC19.
Plasmids were linearized with Sma I (New England Biolabs), extracted twice with phenol-chloroform and precipitated with ethanol. The pellet was washed with 70 % (v/v) ethanol, and the DNA was dissolved in sterile distilled water.
RNA oligonucleotides were enzymatically synthesized by T7 RNA polymerase in 30 ml reactions containing 2.5 mg of template, 20 mM MgCl2, 2 mM spermidine, 80 mM Tris-HCl (pH 8.1), 4 mM of each NTP, 2 mM DTT, and 0.3 mg T7 polymerase. Reactions were incubated for three hours at 37 °C, and quenched with 25 mM EDTA. After denaturation at 96 °C for five minutes, the RNA was purified by electrophoresis on urea-containing polyacrylamide denaturing gels and isolated by electroelution (Elutrap, Whatman Inc.). The concentration of each sample was determined by measuring the optical absorbance at 260 nm.
MoMuLV NC was expressed and purified as described42, and exchanged into buffer containing Tris-HCl (10mM; pH 7.0) β-mercaptoethanol (0.1 mM), NaCl (10 mM) and ZnCl2 ( 0.1 mM).
ΨWT and ΨM RNAs prepared by in vitro transcription were adjusted to 10 µM in 10 mM Tris-HCl (PH 7.0) and various concentrations (40, 80 and 160mM) of NaCl. To dimerize, 10 µl of Psi-native was incubated at 90 °C for one min and 50 °C for 60 min. 10 µl of RNA was mixed with 10 µl of 320 nM of NC in the same buffer as the RNA dimerization buffer. RNA and NC were incubated at 37 °C for 30 min. This 20 µl mixture of RNA-NC was placed on a filter (microcon-50, Millipore), which allows the passage of particles less than 50kDa, and was washed five times with 400µl of the dimerization buffer. Only free NC can pass the membrane of microcon-50 and NC that binds to RNA remains on the membrane.
The RNA-NC complexes that remained on the membrane of filter were adjusted to an RNA concentration of 500nM, and 10µl of this mixture was run on a 15% SDS-PAGE. Retained NC was determined by measuring intensities of stained bands (Brilliant Blue R-250, Fischer BioReagents) after denaturation and electrophoresis under denaturing conditions, and comparison data obtained for control NC samples prepared under identical conditions. Gels were photographed using a Kodak digital camera and band intensities of NC were analyzed with ImageJ software (www.rsbweb.nih.gov/ij).
Stoichiometry and dissociation equilibrium constants for MoMuLV NC binding to both native-Psi and Psi-M were determined by standard ITC methods using a VP-isothermal titration microcalorimeter (VP-ITC) (MicroCal Corp., Northampton, MA). NC concentration was adjusted at 100 µM, and RNAs concentrations were adjusted at 1 µM. NC and RNAs were dialyzed for 16 hours in 10 mM Tris (pH 7.0), 150 mM NaCl, 1 mM MgCl2. Exothermic and endothermic heats of reaction were measured at 30 °C for 28 injections of NC into 1.4 mL of RNA, and heats of dilution were measured by titrating NC into a buffer under identical conditions. Baseline correction was performed by subtracting heats of dilution (obtained by titrating NC into buffer), and binding curves were analyzed and dissociation constants determined by nonlinear least-squares fitting of the baseline-corrected data (Origin Version 5.21, MicroCal Inc., Northampton, MA).
MoMuLV RNA packaging was assessed using derivatives of the MoMuLV-based gag-pol-puro plasmid, pGPP (indicated in the figures here as “wt”), which contains an intact provirus modified by the replacement of the env open reading frame by a puromycin N-acetyltransferase gene (puroR) driven by a simian virus 40 (SV40) promoter 49. The “classic Ψ” deletion of Mann and Baltimore (Δ215–568) 41, sub-deletions Δ215–367 and Δ375–568, and the ΨM mutant were generated by overlap extension PCR, sequenced, and used to replace the corresponding portions of pGPP. The helper plasmid pNGVL-3’-gag-pol, which expresses MoMuLV Gag and Gag-Pol from a 5’ leader-deleted transcript driven by the CMV promoter, has been described previously 70. The competitor plasmid pBAG encodes a retroviral vector with a wild type MoMuLV 5’ UTR and lacZ 71.
Riboprobe templates were constructed as follows. The insert in pEG467-10, which was generated by PCR and subcloned into the EcoRV site of pBSII SK(+) (Stratagene), included complementarity to portions of both the MoMuLV 5’ untranslated region (nts. 55–214) and 100nt of 7SL RNA 50. pSRK38-7 is an HIV-1 pol-7SL chimeric riboprobe template described previously and used here as a probe for 7SL RNA 51. pSRK1216-71 was generated by sequentially subcloning a Cla I-Ssp I fragment of lacZ and a Bgl II-Xba I MoMuLV pol fragment from pNCA72 into pBSII SK(+) digested from Cla I-EcoRV and BamH1-Xba I, respectively.
293T cells (human embryonic kidney cells expressing SV40 T antigen) were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum and 1% penicillin-streptomycin. Virus was produced by transient transfection of the pGPP or pNGVL-3’-gag-pol plasmids into 293T cells by calcium phosphate precipitation70. Virus-containing media were harvested at 24, 36, and 48 hours post transfection, pooled, filtered through a 0.2 µm MCE filter (Fisher Scientific), and stored at −70°C prior to use. Some of this transiently produced virus was used to infect ET cells, which are a 293T derivative that constitutively express ecotropic envelope 49. These infections were used for measuring titers and creating the pools of stably integrated proviruses used in later experiments. Virus-containing media harvested from the ET cell pools were harvested and analyzed for RT content and packaged RNA content using the assays described below.
Virus was harvested, filtered through 0.2 µ m MCE filters, and stored at −70 °C. Quantitative RT assays were performed using modifications of a standard protocol 73 as described previously 74. Products were quantified by PhosphorImager analysis.
Cellular RNA was extracted using TRIzol® reagent (Invitrogen) according to the manufacturer’s protocol. To generate riboprobes for measuring cellular gRNA/7SL levels, two riboprobes were used. To detect 7SL, pSRK38-7 was linearized with SalI and transcribed using T7 RNA polymerase (Promega) and [α-32P] rCTP to create a 344 nt transcript which protected 101 nt of 7SL RNA. To detect MoMuLV gRNA, pSRK1216-71 was linearized with XhoI and transcribed using T3 RNA polymerase (Promega) to generate a 693 nt transcript which protected 206 nt of MoMuLV pol RNA (nts. 5120–5325). In packaging competition experiments, pSRK1216-71 693 nt transcript was used to detect both Test RNAs (by protecting a 206 nt pol fragment) and the Control pBag vector (by protecting a 408 nt lacZ fragment).
For measuring viral gRNA/7SL ratios, RNA was isolated from pelleted virions using a previously described proteinase K-based extraction protocol 17 and quantified by RPA using a 365 nt chimeric MoMuLV-7SL riboprobe templated by linearized pEG467-10, which protected 160 nt of ΨM or 163 nt of wild type and Ψ deletion mutant gRNAs, plus 100nt of 7SL RNA. Previously described RPA approaches 50 were modified by extending hybridization times to 16h and digesting using only RNase T1. Bands were quantified by PhosphorImager and adjusted for the number of radiolabelled Cs incorporated. RNA packaging was quantified by normalizing the amount of MoMuLV RNA in each lane to the amount of co-packaged 7SL.
All RPAs were performed in probe excess as determined by quantification of the amount of undigested probe relative to protected fragments. The amount of probe in each undigested probe lane represents 1% the amount of probe used in each hybridization reaction in a given experiment.
MoMuLV viral RNA was isolated from pelleted virus by the proteinase K-based extraction protocol above and resuspended in 1× TENS buffer (10 mM Tris pH [8.0], 1mM EDTA, 1% (w/v) SDS, and 100 mM NaCL). Samples were incubated for 10 min at the indicated temperatures then placed on ice until loaded on a non-denaturing 0.7% (w/v) agarose gel. RNA was electrophoretically transferred to a Zeta-Probe® GT nylon membrane (Bio-Rad). Prehybridization was performed at 45°C for 2h in 6× SSC (1× SSC is 0.15 M NaCL plus 0.015 M sodium citrate) – 5× Denhardt’s solution–0.5% sodium dodecyl sulfate (SDS) – 0.025 M sodium phosphate – 625 µg/ml of denatured salmon sperm DNA. The oligonucleotide probe was an anti-MoMuLV R probe: 5’-ACTGCAAGAGGGTTTATTGGATACACGGGTACC-3’ that was 5’-end labeled using [γ-32P] ATP (Perkin-Elmer) and T4 polynucleotide kinase (NEB). Hybridization was performed at 45°C for 16 h. The blot was washed 2× with 2× SSC – 0.1% (w/v) SDS at 50°C for 15 min followed by 2 washes with 0.33× SSC – 0.1% (w/v) SDS at 50°C for 15 min. Washed blots were exposed to PhosphorImager screens for subsequent analysis.
Support from the NIH (GM42561 to MFS and CA069300 to AT) is gratefully acknowledged. P.S. and S.S. were supported by an HHMI undergraduate education grant (to UMBC).
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.