Inhibition of HIV-1 integrase nuclear import and replication by a peptide bearing integrase putative nuclear localization signal

Background The integrase (IN) of human immunodeficiency virus type 1 (HIV-1) has been implicated in different steps during viral replication, including nuclear import of the viral pre-integration complex. The exact mechanisms underlying the nuclear import of IN and especially the question of whether it bears a functional nuclear localization signal (NLS) remain controversial. Results Here, we studied the nuclear import pathway of IN by using multiple in vivo and in vitro systems. Nuclear import was not observed in an importin α temperature-sensitive yeast mutant, indicating an importin α-mediated process. Direct interaction between the full-length IN and importin α was demonstrated in vivo using bimolecular fluorescence complementation assay (BiFC). Nuclear import studies in yeast cells, with permeabilized mammalian cells, or microinjected cultured mammalian cells strongly suggest that the IN bears a NLS domain located between residues 161 and 173. A peptide bearing this sequence -NLS-IN peptide- inhibited nuclear accumulation of IN in transfected cell-cycle arrested cells. Integration of viral cDNA as well as HIV-1 replication in viral cell-cycle arrested infected cells were blocked by the NLS-IN peptide. Conclusion Our present findings support the view that nuclear import of IN occurs via the importin α pathway and is promoted by a specific NLS domain. This import could be blocked by NLS-IN peptide, resulting in inhibition of viral infection, confirming the view that nuclear import of the viral pre-integration complex is mediated by viral IN.


Background
Active nuclear import begins in the cytoplasm with recognition of the transported cargo molecules by nuclear transport receptors designated as importins [1]. Proteins targeted to the nucleus contain a specific amino acid sequence, termed nuclear localization signal (NLS), which is recognized by either a member of the importin α family, or directly by importin β. The resultant complex then interacts with the nuclear pore complexes (NPCs), through which it is subsequently transported into the nucleus [2]. This nuclear translocation machinery is highly conserved among lower and higher eukaryotes [3].
Human immunodeficiency virus type 1 (HIV-1) belongs to the lentivirus family, which in contrast to other retroviruses can infect terminally differentiated cells [4,5]. The capability of HIV-1 to infect cell-cycle arrested cells has been ascribed to the ability of its pre-integration complex (PIC) [6,7] to translocate across the nuclear envelope via the NPC [1]. The karyophilic properties of the viral PIC have been attributed mainly to three viral proteins: matrix (MA), Vpr, and integrase (IN) [8][9][10]. The cellular Lens Epithelium-Derived Growth Factor p75 (LEDGF/p75) protein as well as the DNA flap structure of the viral cDNA have also been implicated in promoting the translocation of the PIC into nuclei of virally infected cells [11][12][13]. Yamashipa et al. have proposed that the HIV capsid protein plays a crucial role in controlling the nuclear import of the HIV genome [14]. However, despite these extensive studies and numerous reports, the nuclear import mechanism of the PIC and the involvement of viral or cellular factors driving such a process remain unclear and controversial [15].
The HIV-1 IN protein consists of 288 amino acids and three functional domains: the N-terminal domain (residues 1-50), which bears a zinc-binding motif [16,17]; the central core domain (residues 51-212), which includes the catalytic DDE motif [18][19][20]; and the C-terminal domain (residues 213-288), which has been shown to non-specifically bind the DNA [19][20][21]. To achieve integration of the viral DNA into the host chromosome, the IN must be translocated into the nuclei of infected cells [15].
Various studies have showed that IN is a karyophilic protein. Transfection of cultured mammalian cells with expression vectors bearing IN results in nuclear accumulation of the encoded protein [22]. Import of fluorescently labeled IN into the nuclei of digitonin-permeabilized mammalian cells was shown to be ATP-and temperaturedependent; and this import could be blocked by the addition of unlabeled IN, clearly indicating an active, receptormediated process [23,24]. Based on the ability of recombinant IN protein to bind in vitro to importin α and the ability of a peptide bearing the prototypic simian virus 40 T-antigen NLS (SV40-NLS) to block such binding, as well as nuclear import, nuclear transport of IN has been suggested to occur via the importin α pathway [8,23]. Moreover, interaction of IN with the importin α family has recently been reported [25].
The possibility of the IN protein being carried into the cell's nuclei by other cellular components has also been suggested [13,26,27]. The LEDGF/p75 was initially impli-cated in mediating the nuclear import of IN [13]. However, studies on the specific contributions of LEDGF/p75 demonstrated that it facilitates the interaction between IN and nuclear chromatin, but is not directly involved in the import process [28]. An interaction with importin 7, via a sequence located at the C terminus of IN [26], has been proposed. However conflicting results have been obtained regarding the necessity of this receptor [29,30]. Furthermore anti-importin 7 antibodies did not block nuclear import of IN [25]. More recently, the involvement of the transportin-SR2 (TNPO3) in the nuclear import of IN has been suggested [27]. This conclusion is based mainly on experiments showing that the knockdown of transportin-SR2 (TNPO3) resulted in the reduction of nuclear cDNA [27].
In the present work, we further confirm and emphasize the role that importin α plays in promoting nuclear import of the viral IN and thus in virus infection. Multiple approaches and various experimental systems such as transfected mammalian and yeast cells as well as virally infected cells have been used to answer the question of whether nuclear import of IN may be mediated by its own NLS via interaction with importin α. Our results clearly demonstrate that IN accumulates within wild-type yeast cell nuclei, but fails to do so in importin α-defective yeast mutants (srp1-31) [ cells. This appears to be due to the ability of the NLS-IN peptide to compete for the interaction between the viral IN and the cellular importin α. This peptide has also been found to significantly inhibit HIV-1 replication in TZM-bl cells and inhibit the integration of viral cDNA in infected cells. Thus, the present results support our [23] and others' previous results [33] claiming that the IN protein contains a specific functional NLS sequence, which is located between amino acids 161 and 173 and which confers to this protein the karyophilic property required to ensure productive viral infection.

The NLS-IN peptide is functional in transfected and microinjected mammalian cells as well as in yeast cells
The results in Fig. 1  No toxic effect was exerted by this peptide during the time of the experiment, as estimated by MTT assay (data not shown), thus allowing for studies on its effect in cultured cells. As can be seen (Fig. 1), following the incubation of the transfected cells with the NLS-IN-Pen peptide, very little -if any -IN was intranuclear; most of it was located within the cytoplasm, clearly demonstrating the inhibition of nuclear import. Incubation with the Pen-peptide alone did not have any effect on nuclear import of IN (Fig. 1), strongly indicating a specific effect of the NLS-IN. The same results were obtained when the transfected cells were incubated with a cell-permeable SV40-NLS-Pen peptide ( Fig. 1), indicating an importin α-dependent nuclear import pathway [35].
Due to the high ambiguity surrounding the nuclear import pathway of IN and its NLS domain, we studied its translocation into nuclei in a non-mammalian cell environment as well, namely in yeast cells. W303 cells were transformed with expression vectors encoding the fulllength and truncated IN fused to the green fluorescence protein (GFP) (expressed proteins are schematized in Fig.  2A). As can be seen (Fig. 2B) in cells expressing GFP-IN, the fluorescence is packed into small intranuclear dots, as confirmed by DAPI staining of the cell's DNA, while both the cytosol and cell vacuoles appear relatively dark. The same was observed with the truncated GFP-180-IN, which includes the NLS-IN (Fig. 2B). Next, the ability of the NLS-IN sequence to promote nuclear import was studied. To create a molecule of high molecular weight, thereby avoiding passive nuclear import [2], the NLS-IN coding sequence was fused to the coding region of a double-GFP (GFP) 2 (Fig. 2B). Similar to GFP-IN and GFP-180-IN, the expressed GFP 2 -NLS-IN fusion protein also accumulated within the yeast cell nuclei (Fig. 2B). In contrast, no nuclear import was observed in yeast cells transformed with an expression vector encoding the truncated GFP-152-IN, which lacks the putative NLS-IN (Fig. 2B). INmediated nuclear import can be inferred from the results showing that in yeast cells transformed with vectors expressing GFP molecules alone, the fluorescence distributed within the intracellular space (Fig. 2B). Yeast nuclei in all described experiments were identified by DAPI staining: GFP fluorescence appeared mostly in the nuclei (Fig. 2B).  Fig. 3D). Moreover, no nuclear import was observed when only FITC-BSA molecules were microinjected (Fig. 3E). It should be mentioned that the various recombinant IN proteins were attached to BSA molecules in order to increase their solubility as well as their molecular size, thus avoiding passive diffusion via the NPC. Essentially similar results were obtained when the degree of nuclear import was quantitatively estimated using an ELISA-based system and biotinylated BSA (Bb) conjugates (Fig. 3F). A relatively high degree of nuclear import was

NLS-IN mediates binding to importin in yeast cells and in vitro
Our results showing inhibition of IN nuclear import by the SV40-NLS peptide indicated the involvement of importin α in the translocation process. To verify this, we used the yeast srp1-31 temperature-sensitive mutant [31] in which importin α is inactivated following growth at the non-permissive temperature of 37°C. GFP-IN appeared as small fluorescent dots when expressed in the wild-type W303 or in the srp1-31 mutant yeast cells grown at 25°C (Fig. 4). These results clearly demonstrate accumulation within the cells nuclei, a localization which was verified by DAPI staining. Neither the cytosol nor the cell vacuoles were strongly fluorescent. The same fluorescently stained dots were observed after 4 h growth of the W303 cells at 37°C (Fig. 4), again indicating accumulation of GFP-IN within the nuclei under these conditions. On the other hand, the srp1-31 mutant yeast cells lost their nuclear import ability at the non-permissive temperature (37°C): most of the GFP-IN was distributed within the yeast cell's cytoplasm (Fig. 4). However, nuclear localization was restored in these mutant cells following re-incubation at the permissive temperature of 25°C (not shown). To confirm that at the non-permissive temperature only importin α-dependent nuclear import is blocked, we repeated previous experiments in which Pik1 protein [36] had been shown to be imported into nuclei of srp1-31 cells at 37°C (not shown and see [36]), and in which it was established that nuclear import of this protein is importin α-independent [36]. Thus, the blockage is specific to the importin α pathway in srp1-31 cells at the non-permissive temperature.
A specific IN-importin α interaction in vivo can be inferred also from the results obtained using the BiFC assay system in yeast cells ([32] and Fig. 5). As a control system, to confirm that restoration of fluorescence following the use of labeled IN is due to specific protein-protein interactions, the BiFC assay system was first employed to study the dimerization of the IN molecules themselves

NLS-IN inhibits IN and cDNA nuclear import as well as HIV-1 replication in cultured cells
In the light of the results showing the requirement of NLS-IN for nuclear import of IN in mammalian as well as in yeast cells, it became of relevance to study its effect in virally infected cells.
Co-immunoprecipitation (co-IP) experiments using a lysate obtained from HIV-1-infected cells revealed an interaction between the virus IN protein and the cellular importin α (Fig. 7A). Interestingly, when the virusinfected cells were treated with either the NLS-IN-Pen or the SV40-NLS-Pen peptides, no such interaction between the two proteins could be detected (Fig. 7A) Inhibition of nuclear import of the viral IN in HIV-1 infected cells as well is evident from the immunofluorescence microscopic study shown in Fig. 7B. From the immunostaining results, it appears that in infected cells, the IN is localized both within the cytosol and within the nuclei (Fig. 7B (no peptide)). However, no intranuclear fluorescence was observed in cells treated with the NLS-IN-Pen or the SV40-NLS-Pen peptides, indicating the inhibition of nuclear import (Fig. 7B). In contrast, some intranuclear fluorescently labeled IN could be observed when the infected cells were incubated in the presence of the SV40-mut-NLS-Pen or the Pen peptide itself (Fig. 7B). This is also evident from the fact that more cytosolic IN was present in such peptide-treated cells than in those incubated in the absence of any peptide or with the nonactive peptides (Fig. 7B).
Similar to their effect on IN nuclear import, both NLS-IN-Pen and SV40-NLS-Pen blocked nuclear import of the viral cDNA (Fig. 8A), as is particularly evident from the absence of 2LTR circles (Fig. 8B) in cells infected with wild-type HIV-1.
Inhibition of IN nuclear import is expected to result in the inhibition of virus replication, especially in cell-cycle arrested cells. Using TZM-bl cells [40] treated with aphidicolin to obtain cell-cycle arrested cells as an experimental system, a reduction in HIV-1 infection--as is reflected by the inhibition of reporter gene expression--was observed in the presence of NLS-IN-Pen or SV40-NLS-Pen (Fig. 9A). As expected, the inhibition was less pronounced when dividing cells were treated with the NLS-bearing peptides (Fig. 9B). The specific effect of NLS-IN and the requirement for cell permeability can be inferred from the results showing that no inhibition of HIV-1 infection was promoted by the Pen peptide alone or by the impermeable NLS-IN peptides. The results in Fig. 9C and 9D clearly demonstrate that the NLS-IN-Pen peptide due to its inhibitory effect on IN nuclear import inhibited the process of viral cDNA integration, reaching a higher degree of inhibition in non-dividing (cell-cycle arrested) than in dividing cells. Detailed kinetics studies (Fig. 9E) further support the view that the step which is blocked by the two NLS-Pen peptides (IN-NLS and SV40-NLS) is the nuclear import process. Evidently, nuclear import of the IN-DNA complex is required for the integration process to proceed. In addition, the time-dependent inhibitory pattern of the NLS-IN-Pen (Fig. 9E) is almost exactly half the way between that observed following the addition of AZT and that of the LEDGF 402-411 peptide, which has been demonstrated to directly block HIV-1 IN and thus the integration process [41]. Inhibition was not observed following the addition of the non-permeable NLS-IN peptide, a peptide bearing a SV40-mut-NLS-Pen or the Pen peptide, again indicating the specific effect of the NLS sequence (Fig. 9).

Discussion
The question of how retroviruses and particularly HIV-1 cross the nuclear envelope in cell-cycle arrested cells is of specific scientific interest. After long and extensive research, it appears that no clear mechanism has yet emerged and the possibility that several pathways simultaneously mediate nuclear import of the viral PIC cannot be excluded. In the present work-following our previous experiments using in vitro systems [23]-we focused on the nuclear import of IN protein using yeast and mammalian An interaction between HIV-1 IN and importin α was first demonstrated by Gallay et al. [8]. Similarly an interaction between the IN and members of the importin family and that its nuclear transport appears to be dependent on the importin α/β heterodimer have also demonstrated by Hearps and Jans [25]. A functional NLS sequence was identified between amino acids 161 and 173 of the IN protein whose mutation disrupted IN translocation into nuclei [12,42]. However, later studies indicated that this sequence may be required for promoting viral DNA integration and not necessarily nuclear import [43]. However, our previous work clearly demonstrated that a peptide bearing IN 161-173 residues can mediate import of a conjugated protein into the nuclei of permeabilized cells, confirming the view that it can function as a NLS [23].
In the present work, we have further studied the involvement of IN amino acids 161-173 in promoting its nuclear import. Nuclear accumulation was observed in yeast transformed with an expression vector bearing only the NLS-IN sequence, which in order to avoid diffusion into the nuclei, was fused to two molecules of GFP, resulting in a molecule of about 58 kDa. It is assumed that molecules of up to about 30 to 40 kDa can passively diffuse via the NPCs into cell nuclei [2]. Therefore, the nuclear import observed here with the various GFP-IN conjugates, the molecular weights of which varied between 48 and 60 kDa, should be ascribed to a receptor-mediated, active process. Nuclear import was practically not observed when yeast cells were transformed with the 152-IN truncated protein, which lacks the putative NLS sequence. Results obtained in permeabilized or microinjected cells, as well as in IN-transfected intact mammalian cells, further supported these results. The failure of 152-IN to penetrate the cell nuclei suggests that if an additional NLS sequence, besides the one located between 161 and 173, were present, it must be located closer to the C terminus of the IN protein, a possibility that has been suggested previously [25,44]. However, the fact that the truncated 180-IN protein was translocated into the cell nuclei indicates that the identified NLS-IN is sufficient to provide the IN with karyophilic properties. Our results demonstrating inhibition of IN nuclear import in transfected cells by a peptide carrying the NLS-IN sequence further emphasize the view that this sequence gives IN its karyophilic properties.
SRP1 is the only known importin α protein in budding yeast, and previous studies have demonstrated that it is essential for proper maintenance of nucleocytoplasmic trafficking [45]. Indeed, a temperature-sensitive mutant in SRP1 has been isolated (srp1-31) and shown to be defective in nuclear import at the non-permissive temperature [31]. Thus, yeast, including SRP1-mutated strains, has been instrumental in studying various aspects of the nuclear import machinery [36,46] and in characterizing karyophilic properties. The availability of the temperature-sensitive srp1-31 mutant offers an advantage for studying nuclear transport of the IN protein in yeast cells. The involvement of importin α in the IN nuclear import pathway can be inferred from our experiments showing nuclear import at the permissive, but not at the non-permissive temperature where importin α is inactivated [31]. The amino acid residues of NLS-IN described in the present work have been implicated in additional viralrelated functions, such as specific binding of IN to the LEDGF/p75 protein and to the viral LTR region [38,49,50]. Due to its multifunctional activity, this region may be useful as a target for the development of inhibitors. As mentioned above our present as well as previous results describing the involvement of importin α in mediating nuclear import of IN do not exclude the possibility of an additional nuclear import pathway for IN in which TNPO3 is involved.

Viruses
Wild-type HIV-1 was generated by transfection of HEK293T cells with pSVC21 plasmid containing the fulllength HIV-1 HXB2 viral DNA. Wild-type viruses were harvested from HEK293T cells 48 and 72 h post-transfection. The viruses were stored at -75°C.

Virus stock titration
Quantitative titration of HIV-1 was carried out using the MAGI assay, as described by Kimpton and Emerman [40]. Briefly, TZM-b1 cells were grown in 96-well plates at 1 × 10 4 cells per well. the cells were infected with 50 μl of serially diluted virus as described [40]. Two days post-infection (PI), cultured cells were fixed and β-galactosidase was estimated exactly as described previously [40]. Blue cells were counted under a light microscope at 200× magnification.

Synthesis of peptides
Peptides were synthesized on Rink amide resin using a model 433A Applied Biosystems peptide synthesizer with FastMoc chemistry, exactly as described previously [55]. The Pen peptide had the following sequence: RQIKIW-FQNRRMKWKK (Ant 43-58) [34].

Plasmid construction
All of the plasmids used in this study were constructed using PCR cloning techniques with the high-fidelity enzyme Platinum Pfx DNA polymerase (Invitrogen). Clones were subjected to automated DNA sequencing.

Construction of expression vectors for the BiFC assay
The yeast multicopy shuttle vectors pRS423 (with HIS3 as the selective marker) and pRS426 (with URA3 as the selective marker), both with the ADH1 promoter, were used as the cloning plasmids (a kind gift from D. Engelberg, Alexander Silberman Institute, Israel). The DNA coding region of the two GFP fragments [58], namely the N terminus (GN) and C terminus (GC), were cloned into pRS423 and pRS426 [59] to give pRS423-GN and pRS426-GC, respec-tively. A linker consisting of (GGS) 5  The transformed yeast cells were grown for 24 h at 25°C to reach the logarithmic phase, then divided into two cultures: one remained at 25°C (permissive temperature) and the other was transferred to 37°C (non-permissive temperature at which more than 95% of the SRP1 is inactivated [61]). After 4 h of growth, yeast cells were washed; GFP-IN expression was induced by the addition of galactose and the yeast cells were incubated for an additional 4 h at either 25 or 37°C. At the end of the growth period, yeast cells were harvested and then observed by confocal microscope. Yeast nuclei were identified by DNA staining in fixed yeast cells (4% v/v paraformaldehyde and 3.4% w/v sucrose) http://www.ciwemb.edu/labs/koshland/ Protocols/MICROSCOPY/gfpfix.html with 4',6-diamidino-2-phenylindole (DAPI). To confirm expression of all proteins, western blot with anti-GFP antibody was performed. Each experiment was repeated at least three times.

Analysis of protein-protein interaction by the BiFC assay
In this approach, a molecule of GFP is separated into two portions: the N-terminal part (GN) ending at amino acid residue 154 and the C-terminal part (GC) beginning with the methionine residue preceding residue 155 of GFP [62]. Neither of these two halves of GFP fluoresce when expressed alone; however, the fluorescence is restored when GN and GC are brought together as fusions with interacting proteins [32]. The two halves of the GFP were cloned separately with each of the indicated proteins as described above and the different plasmids were transformed into the yeast strain EGY48. After 48 h at 30°C, the plates were transferred to 23°C for 3 days and then yeast cells were visualized by confocal microscopy. Nuclei were stained with DAPI as described above. Each experiment was repeated at least three times.

Binding of transport substrates to importin
Binding to importin α was estimated using an ELISAbased system essentially as described in [23,63]. Briefly, Maxisorp plates (Nunc) were coated overnight at 4°C with a solution containing recombinant importin α (5 μg) in carbonate buffer (NaHCO 3 /Na 2 CO 3 buffer at pH 9.6). All subsequent steps were as described previously [23].

Nuclear import in permeabilized mammalian cells: quantitative estimation
Import of Bb-IN proteins or of Bb-NLS-IN conjugates into nuclei of digitonin-permeabilized Colo-205 cells was determined quantitatively by an ELISA-based system, exactly as described previously [23].

Localization studies of transfected IN in cultured cells by immunostaining
HeLaP4/IN-expressing cells were grown on chamber slides (Nunc). After reaching 70-80% confluence, cells were arrested in the cell cycle by treatment with 5 μg/ml of aphidicolin, and then incubated with 150 μM of the indicated peptide for 6 h. Cells were fixed and immunostained essentially as described previously [66] with several modifications. Briefly, after fixation cells were blocked with 5% (w/v) BSA (IgG free) (Jackson) in PBS for 60 min. For detection of HIV-1 IN, cells were incubated with 1:100 rabbit α-IN (NIH AIDS Research & Reference Reagent Program, cat. no. 758) at room temperature for 60 min each. Cells were washed five times with PBS + 0.05% (v/v) Tween 20. Then the cells were incubated with a second antibody, Cy3-conjugated anti-rabbit antibody (Jackson) (1:200) at room temperature for 60 min followed by five washes with PBS + 0.05% Tween 20. For detection of DNA, cells were stained with DAPI according to the manufacturer's protocol. The slides were prepared with mounting media (Bio-Rad) and immunofluorescent cells were detected with a confocal microscope.

Localization studies of IN in HIV-1-infected cultured cells by immunostaining
HeLaP4 cells were grown on chamber slides. Cells were arrested in the cell cycle by treatment with 5 μg/ml aphidicolin, and then incubated with 150 μM of the indicated peptide for 2 h. After incubation with the peptides, cells were infected with wild-type HIV-1 at a multiplicity of infection (MOI) of 25. Cells were fixed and stained as described above with the following modifications: fixation was performed at 10 h PI, the first antibody was used at a dilution of 1:50, the second antibody at a dilution of 1:100. The remaining lysate or isolated fractions were incubated for 1 h at 4°C with either the anti-Impα or anti-IN antibodies. Following 3 h incubation with protein G-agarose beads (Santa Cruz) at 4°C, the samples were washed three times with PBS containing 1% (v/v) Nonidet P-40. SDS buffer was added to the samples and after boiling and running on an SDS polyacrylamide gel, the membranes were immunoblotted with either anti-Impα or anti-IN antibodies, and the complementary HRP-conjugated antibodies (Jackson) as second antibodies.

Study of in vivo protein-protein interactions using co-IP
When peptides were used, cells were incubated with 150 μM of the indicated peptide for 2 h prior to infection.

HIV-1 titration by multinuclear activation of a galactosidase indicator (MAGI) assay
Quantitative titration of HIV-1 was carried out using the MAGI assay, as described previously [40]. Briefly, TZM-b1 cells were grown in 96-well plates at 10 4 cells/well and incubated for 12 h at 37°C. Peptides were then added and after an additional 2 h of incubation, the cells were infected with 50 μl of serially diluted HIV-1. To obtain cell-cycle arrested cells, 5 μg/ml of aphidicolin was added 2 h before the experiment. Cultured cells were fixed 2 days PI and β-galactosidase was estimated [47]. Blue cells were counted under a light microscope at 200× magnification.

Quantitative analysis of integration
Real-time PCR experiments were performed to estimate integration as described previously [67].

Isolation of cytoplasm and nuclei from infected cells
The various fractions were obtained from virus-infected cells essentially as described previously [68,69] with several modifications. Briefly, cells were harvested and washed twice in buffer A (20 mM Hepes pH 7.3, 150 mM KCl, 5 mM MgCl 2 , 1 mM DTT and 0.1 mM PMSF). Cells were then suspended in 200 μl of buffer A with 0.025% (w/v) digitonin and incubated at room temperature for 10 min. Cells were centrifuged for 3 min at 1000 g at room temperature. The supernatant was then centrifuged at 8000 g and separated into supernatant (cytoplasm) and pellet (nuclei) and stored at -70°C.

Quantitation of total and nuclear viral DNA
Total viral DNA was estimated using SYBR green real-time quantitative PCR at 10 h PI from the total or nuclear-iso-lated fractions of the infected cells. DNA was isolated by phenol chloroform method. Briefly, DNA samples (1 μg of DNA) were added to 95 μl containing 1× Hot-Rescue Real Time PCR Kit-SG (Diatheva s.r.l, Fano, Italy), and 100 nM of each PBS (primer-binding site) primer: F5 (5' primer, 5'-TAGCAGTGGCGCCCGA-3') and R5 (3' primer, 5-TCTCTCTCCTTCTAGCCTCCGC-3'). All amplification reactions were carried out using an ABI Prism 7700 Sequence Detection System (Applied Biosystems): One cycle at 95°C for 10 min, followed by 45 cycles of 15 s at 95°C and 35 s at 68°C. In each PCR run, three replicates were performed. All other details are exactly as described in Casabianca et al. [70].

Quantitation of 2LTR circles
Quantification of 2LTR circles was estimated exactly as described in Butler et al. [71].