The prototype foamy virus protease is active independently of the integrase domain
© Spannaus et al.; licensee BioMed Central Ltd. 2012
Received: 14 February 2012
Accepted: 10 May 2012
Published: 10 May 2012
Skip to main content
© Spannaus et al.; licensee BioMed Central Ltd. 2012
Received: 14 February 2012
Accepted: 10 May 2012
Published: 10 May 2012
Recently, contradictory results on foamy virus protease activity were published. While our own results indicated that protease activity is regulated by the viral RNA, others suggested that the integrase is involved in the regulation of the protease.
To solve this discrepancy we performed additional experiments showing that the protease-reverse transcriptase (PR-RT) exhibits protease activity in vitro and in vivo, which is independent of the integrase domain. In contrast, Pol incorporation, and therefore PR activity in the viral context, is dependent on the integrase domain. To further analyse the regulation of the protease, we incorporated Pol in viruses by expressing a GagPol fusion protein, which supported near wild-type like infectivity. A GagPR-RT fusion, lacking the integrase domain, also resulted in wild-type like Gag processing, indicating that the integrase is dispensable for viral Gag maturation. Furthermore, we demonstrate with a trans-complementation assays that the PR in the context of the PR-RT protein supports in trans both, viral maturation and infectivity.
We provide evidence that the FV integrase is required for Pol encapsidation and that the FV PR activity is integrase independent. We show that an active PR can be encapsidated in trans as a GagPR-RT fusion protein.
Foamy viral assembly and maturation differs in many ways from the orthoretroviral counterpart. Foamy viruses (FVs) express the Pol protein from a specific transcript and not from a Gag-Pol precursor [1–6]. This leads to the expression of a separate Pol polyprotein consisting of a protease (PR), reverse transcriptase (RT) and integrase (IN) domain. Consequently, the assembly of FVs is complex since interaction of the viral RNA with Pol has been shown to be required for Pol uptake [7, 8]. Several PR cleavage sites in Pol have been determined in vitro , but the in vivo maturation of foamy viral proteins appears to be very limited as compared to orthoretroviruses. Single cleavage sites were identified in vivo in both, Gag and Pol. Pol cleavage results in the free IN domain and a PR-RT fusion protein, whereas a p3 peptide is cleaved off from the carboxyl terminal end of Gag . Despite the limited in vivo cleavage, the activity of the PR has to be tightly controlled to allow packaging of the full-length Gag and Pol precursor proteins into the virus. The PR activity is essential for viral infectivity . The separate PR domain was shown to exhibit only a weak tendency to form active dimers in vitro [11, 12]. PR dimer formation and activation are achieved, as we have shown recently, by the activation of the PR by a specific interaction of PR-RT with the viral PR activation RNA motif (PARM) [11, 13]. PARM is located in the IN region of the viral genome [7, 11]. Recently, Lee et al.  reported that PR activity requires the IN domain. Additionally, IN was suggested to be essential for dimerization of the Pol protein, which subsequently is needed for PR activity. Here, we wanted to re-analyse the Pol domains’ requirements for PR activity both in vitro and in vivo, using a novel approach to target Pol independently of RNA or Pol domains into FV particles.
Comparing both, Pol- and PR-RT-mediated Gag processing, a higher PR activity in Pol expressing cells was observed. This could be due to higher Pol expression levels (Figure 2). However, it is more likely that the somewhat lower PR activity with PR-RT was the result of a decrease of PR-RT incorporation into cell-associated or intracellular viruses, as IN is required for FV Pol encapsidation.
To confirm this result in the context of the PFV proviral clone (pHSRV13), the complete IN domain downstream of the RT/IN cleavage site was deleted. BHK cells were transfected with the proviral plasmids pHSRV13 or pHSRV13∆IN. Supernatants were collected after two days and viruses were pelleted through a sucrose cushion. The pelleted viruses were lysed in RIPA buffer. Gag and Pol proteins were analysed by Western blotting (Figure 2B). After deletion of the IN domain some Gag cleavage was observed, (Figure 2B), leading to the conclusion that the IN might enhance the PR activity, but is not strictly required for PR activity. The lower PR-RT incorporation can be observed in the recent publication by Lee at al ( Figure 1D) as well, underlining our results. In addition, our IN deletion mutants showed a strong decrease in cellular PR-RT amounts. Again, this is in line with the results shown previously .
It was shown previously that FV Gag release is dependent on the viral Env and that Pol encapsidation requires the viral genome . To further investigate whether Pol uptake is genome dependent and GagPol particle release is Env-dependent, we analysed if Env or pMD9 is required for particle release. Therefore, HEK 293 T cells were transfected as described above with either the vector system or with pGagPol in the presence or absence of pMD9, harbouring the viral genome, or pcoPE expressing Env (Figure 3D). In order to increase particle formation, the cells were cotransfected with the gag expression vector pcoPG4 and the pGagPol plasmid. Viruses were partially purified by ultracentrifugation through a sucrose cushion. Cellular and viral Pol and Gag processing and amounts were analysed by Western blotting (Figure 3C). This analysis revealed that more Pol was incorporated into virus particles that were derived from expressing cells as compared to cells transfected with the vector system, indicating that FVs tolerate higher Pol incorporation than seen with the wt. In contrast to the wt virus, Pol incorporation was genome-independent. In addition, Pol was not cleaved off from Gag first and subsequently incorporated in a RNA dependent way, since Pol was encapsidated even in the absence of the viral genome. (Figure 3D, compare lanes 3 (separate Pol without viral genome)) and 7 (GagPol fusion protein without viral genome). Env was required for particle release of cells transfected either with vector system or pGagPol (Figure 3D, lanes 1 & 2 and lanes 7 & 8). These experiments demonstrate that a separate Pol is not required for FV assembly and that a Gag-Pol fusion protein results in infectious viruses.
Having established that a GagPol fusion protein leads to virus particles, a codon optimized GagPR-RT expression plasmid was created by PCR to investigate PR activity. This plasmid encodes the complete Gag, PR and RT ORFs followed by a stop codon. HEK 293 T cells were transfected with pcoPE, pMD9, pcoPG4, and either pGagPol or pGagPR-RT. Gag and Pol amounts as well as Gag processing of the harvested viruses was analysed by Western blotting (Figure 3D, lanes 9 and 10). The PR activity in the context of the codon optimized Pol constructs is not strictly dependent on the viral RNA, due to the high over-expression rate of pol. Viruses from both, pGagPol and pGagPR-RT transfected cells, showed similar Gag processing, indicating that the presence of the IN domain is not crucial for PR activity.
Therefore, in a second approach the plasmid pGagPolD/A was used as IN source in the context of the Pol protein. The pGagPolD/A plasmid encodes a GagPol fusion protein with a mutation in the PR-active site (Asp24 > Ala)  (Figure 4A). HEK 293 T cells were co-transfected with pGagPol alone or with pGagPolD/A and pGagPR-RT, as a PR source (Figure 4A and B). Western blotting analysis showed maturation of both cellular and viral Gag proteins. Furthermore, analysis of the viral infectivity revealed that both, recombinant viruses comprising an encapsidated GagPR-RT protein were infectious (Figure 4C) although viral titers (104/ml) were significantly lower than those of the wt (106/ml). These experiments prove clearly that the IN domain in cis is not necessary for PR activity, since we show here that the FV PR in trans is sufficient for viral maturation.
We provide evidence that the IN domain is required for efficient pol expression and for Pol encapsidation. Recapitulatory, all publications containing data on FV Gag and Pol cleavage in the presence or absence of the viral genome confirm the necessity of the viral RNA for PR activity. [11, 20–22]. In addition, we show that a Gag-Pol fusion protein can give rise to infectious viruses and that the expression of a separate Pol is not required for FV infecti-vity. Furthermore, with the GagPol fusion protein we can now separate Pol encapsidation from the genome and study genome incorporation separately.
In summary, we provide evidence that the FV IN domain is not required for PR activity: First, expression of PR-RT in cells is sufficient for Gag processing. Second, the PR activity in vitro is RNAPARM dependent but IN independent. Third, incorporation of PR-RT into viruses via a GagPR-RT fusion protein leads to Gag cleavage in purified viruses. Fourth, the PR activity can be encapsidated in trans as PR-RT molecule in the absence of the IN domain leading to mature and infectious viruses.
To determine proteolytic activity in vitro, PFV PR-RT and the GB1-GFP substrate containing the RT-IN cleavage site between GB1 (immunoglobulin binding domain B1 of the streptococcal protein G) and GFP (green fluorescence protein) were purified as described previously [23–25]. In a first assay, 10 μM of the GB1-GFP substrate was incubated for 2 h at 37°C in 50 mM Na2HPO4/NaH2PO4 pH 6.4 and 100 mM NaCl with 2.5 μM PFV PR-RT in the presence or absence of 0.5 μM RNAPARM in a total volume of 20 μl. Reaction products were separated by electrophoresis on 10% BisTris gels (Invitrogen, Karlsruhe, Germany) in 50 mM 2-(N-morpholino)ethanesulfonic acid buffer pH 7.3, 50 mM Tris base, 0.1% SDS and 1 mM EDTA.
In a second assay, the conversion of the GB1-GFP substrate was observed by measuring the change in fluorescence anisotropy of GFP upon cleavage of the substrate (excitation wavelength: 395 nm; emission wavelength: 517 nm) in an L-format Jobin-Yvon Horiba Fluoromax fluorimeter. Therefore, 10 μM GB1-GFP substrate was incubated at 37°C in 50 mM Na2HPO4/NaH2PO4 pH 6.4 and 100 mM NaCl. After 30 min 2.5 μM PFV PR-RT and after 60 min 0.5 μM of RNAPARM were added.
Codon-optimized prototypic FV gag (pcoPG4) , pol (pcoPP) , env (pcoPE)  and the gfp encoding vector genome (pMD9)  plasmids were used. pGagPol: In the first step the Gag/P3 cleavage site was added by PCR using the primers PolHpaIs/PolHIa, Herculase II polymerase (Stratagene) and pcoPP as template. The product was digested with BamHI and cloned into the BamHI/HpaI digested pcoPP vector. The resulting vector was denominated as pcoCSPol. Then Gag was amplified with the primers GagHIs and GagHIa. The product was digested and cloned into the HpaI-site of the described pcoCSPol vector. The resulting plasmid was verified by nucleotide sequencing. pGagPR-RT: The Gag/P3 was added to pol and a stop codon was cloned at the RT/IN CS by PCR using the primers PolHpaIs and Pol∆INNota. The PCR product was cloned into the NotI/HpaI digested pcoPP vector. Then the Gag-coding sequence was added as described above. pPR-RT: An AsiSI-restriction site was introduced into pcoPP by site directed mutagenesis using the primers PolHIs and PolAsiSIa and the Pol100a and PolAsiSIs respectively. The PCR-product was digested with BamHI/XhoI and ligated into the BamHI/XhoI digested pcoPP vector. The pcoPPΔIN vector was digested with AsiSI and XhoI and the AsiSI/XhoI digested PCR product (primers: Pol100a and PolAsiSIs) was inserted. A stop codon was inserted 6 amino acid residues downstream of the PR RT-IN cleavage site (YVVN/CNTKKAI-Stop) by ligation of the INstop-oligo into the AsiSI site of pcoPP-AsiSI. pGagPolD/A (Asp24 > Ala): The catalytic centre of the PR was inactivated by amplifying the pcoPol vector with the primers pcoAflIIs and PolNarIa. The product was digested with AflII/NarI and ligated into the AflII/NarI digested pcoPol vector. The resulting pGagPolD/A vector was digested with HpaI and ligated with the Gag fragment amplified with the primers GagHIs and GagHpaw/ostop. pHSRV13∆IN: In order to delete the IN domain a AsiSI was introduced at position 5383–5389 of pHSRV13 by PCR mutagenesis using the primers HSRVAsiSIa and HSRVAsiSIs. In this site a stop codon was introduces by self-complementary primer pStop, resulting in a stop-codon at amino acid residue 562 of Pol.
All transfection reactions were performed using Turbofect (Fermentas) and HEK 293 T or BHK cells as described before .
PolHpaIs, AACACCGTGACCCAGATGAACCCCCTGCAGCTGCTGCAGCC; PolHIa, TCTTGTCGTAGTGGAACTGGATCCGGGG; Pol∆INNota, AATTGCGGCCGCTTAGTTCACCACGTAGGAGCCCTGGGTGG; GagHIs, AACGGTGGAGGGCAGTGTAGTCTGAGC; GagHIa, AACGGCTCTGTCCCGCTGGTCGCCGCCAGAGGC; GagHpaI∆CSa, ACTGAATTCGTCCCGCTGGTCGCCGCCAGAGGC; PolHIs, ATGACCTACCTGGAAGATCCCCGG; PolAsiSIa, ATATCGCTCGAGTCCAGCGATCGCCTT CTTGGTGTTGCAGT TCACCACG; PolAsiSIs, ATACGCGATCGCTAAGCCCAACCTGGACGCCGAACTGG; Pol100a, TATTAGGAAAGGACAGTGGGAGTGG; pcoAflIIs, ATGGAAGACTTAAGGCA-GCGGC; Pol21ANarIa, ATGGTGGCGCCGCTGGCCCAGTGGG; pStop, TTAACTAAGTAAGGATCCTTACTTAGTTAAAT; HSRVAsiSIa, TATTCCGGAATATGCGATCGCTTTTTTGGTATTACAATTAACTACATAACTTCCTTGGG; HSRVAsiSIs, ATAGCGAT CGCAT GTAAT ACCAAAAAACCAAACCTGGATGC.
To purify virus particles, cell culture supernatants were centrifuged at 474 x g to remove cells. The pre-cleared supernatant were loaded on a 20% sucrose cushion and centrifuged at 201,149 x g at 4°C for 2 h. To analyse the infectivity cell culture supernatants were titrated on BHK cells in triplicate assays. The number of GFP expressing cells was determined after two days using a fluorescence microscope. Viral titers were subsequently calculated.
Gag and Pol were detected using monoclonal antibodies against Gag and RT as described before . GAPDH amounts were visualized using an anti-GAPDH serum (Sigma-Aldrich).
streptococcal protein G
Green fluorescence protein
Protease activation RNA motif
Human immunodeficiency virus
Prototype foamy virus
This work was supported by the Deutsche Forschungsgemeinschaft (BO3006/2-1, IRTG1522, Wo630/7-3 and the University of Würzburg in the funding programme Open Access Publishing).