HIV-1 exploits importin 7 to maximize nuclear import of its DNA genome

Abstract

Background: Nuclear import of the HIV-1 reverse transcription complex (RTC) is critical for infection of non dividing cells, and importin 7 (imp7) has been implicated in this process. To further characterize the function of imp7 in HIV-1 replication we generated cell lines stably depleted for imp7 and used them in conjunction with infection, cellular fractionation and pull-down assays.

Results: Imp7 depletion impaired HIV-1 infection but did not significantly affect HIV-2, simian immunodeficiency virus (SIVmac), or equine infectious anemia virus (EIAV). The lentiviral dependence on imp7 closely correlated with binding of the respective integrase proteins to imp7. HIV-1 RTC associated with nuclei of infected cells with remarkable speed and knock down of imp7 reduced HIV-1 DNA nuclear accumulation, delaying infection. Using an HIV-1 mutant deficient for reverse transcription, we found that viral RNA accumulated within nuclei of infected cells, indicating that reverse transcription is not absolutely required for nuclear import. Depletion of imp7 impacted on HIV-1 DNA but not RNA nuclear import and also inhibited DNA transfection efficiency.

Conclusion: Although imp7 may not be essential for HIV-1 infection, our results suggest that imp7 facilitates nuclear trafficking of DNA and that HIV-1 exploits imp7 to maximize nuclear import of its DNA genome. Lentiviruses other than HIV-1 may have evolved to use alternative nuclear import receptors to the same end.

Figure 1

Generation of stable imp7 knock down (KD) cells. (A) Left panels: Western blot with an anti-imp7 antibody and an anti-Ran antibody on total cell extracts obtained from a low passage polyclonal population of DxR KD (lane 1) and imp7 KD HeLa cells (lane 2), a polyclonal population of DxR KD (lane 3), imp7KD (lane 4) and the same imp7 KD population expressing an imp7 cDNA with two silent mutations making it resistant to the shRNA (imp7+7R) (lane 5). Note that cells in lanes 3 to 5 were grown for 4 weeks to allow for selection of the imp7+7R line. Right panel: Western blot with an anti-imp7 antibody and an anti-Ran antibody on total cell extracts obtained from a polyclonal population of DxR KD cells (lane 1) and imp7 KD Jurkat cells (lane 2). (B) DxR KD (shDxR), imp7 KD (shimp7) and imp7 back complemented (shimp7+7R) HeLa cells were plated onto 24-well plates to have the same confluency by the next day and infected with five fold serial dilutions of a VSV-G pseudotyped HIV-1 vector (pHR') expressing GFP. Twenty-four hours after infection the percentage of GFP+ cells was measured by flow cytometry. (C) Cells were infected with three fold serial dilutions of a VSV-G pseudotyped SIVmac vector expressing GFP and analysed as in (B). Similar results were obtained in least three independent experiments using different virus stocks. (D) Western blot with an anti-imp7 antibody and an anti-Ran antibody on total cell extracts obtained from a polyclonal population of DxR KD HeLa cells (lane 1) and three different imp7 KD clones (lane 2, clone 2; lane 3, clone 4; lane 4, clone 8). The bands were scanned and the intensity of the imp7 band relative to Ran is shown below each sample. (E) DxR KD cells and imp7 KD clonal cell populations were infected with a VSV-G pseudotyped HIV-1 vector (pHR') or SIVmac vector expressing GFP at an MOI of 0.03 and the percentage of GFP+ cells counted 24 hours after infection by flow cytometry. Data are expressed as average percentage of infection relative the parent line control (shDxR) ± SD of three independent experiments. (F) DxR KD and imp7 KD Jurkat cells were infected with three fold serial dilutions of a VSV-G pseudotyped HIV-1 vector (pCSGW) expressing GFP and the percentage of GFP+ cells counted 24 hours after infection by flow cytometry. Similar results were obtained in two additional experiments using different virus stocks.

Figure 2

The effect of imp7 KD on HIV-1 infection is independent of cell-cycle arrest. (A) HeLa cells analyzed by FACS after staining with propidium iodide. Top panel, dot plot of forward and side scatter; the gated population was used for cell cycle analysis, middle panel, no aphidicolin, bottom panel, cells treated with 1.5 micrograms/ml aphidicolin. M1, G1 phase; M2, S phase; M3, G2/M phase. (B) FACS analyses of HeLa DxR and imp7 KD cells infected at an MOI of 0.03 with a VSV-G pseudotyped HIV-1 vector (pCSGW), SIVmac vector and HIV-2 vector in the absence or presence (Aph) of aphidicolin. Cells were analyzed for GFP expression 20–24 hours post infection. Bars represent the mean value ± SD of three independent experiments. Infection levels with an MLV vector were 10 fold lower in aphidicolin-treated cells compared to control cells (not shown).

Figure 3

Lentiviral IN affinity for imp7 correlates with infection phenotype in imp7 KD cells. (A) GST-imp7 (top left), GST (middle), or GST-LEDGF^326–530 (bottom) immobilized on glutathione sepharose beads were incubated in the absence (lane 1), or presence (lanes 2–6) of untagged recombinant HIV-1, HIV-2, SIVmac, EIAV, or BIV IN in the pull-down buffer containing 130 mM NaCl. Proteins bound to glutathione sepharose beads were resolved in an SDS PAGE gel and detected by staining with Coomassie Blue. Input quantities of each soluble protein used are shown to the right. Migration positions of GST-Imp7, INs, GST, and GST-LEDGF^326–530, and the molecular weight markers are indicated. (B) Non-tagged Imp7 was incubated in the absence (lane 3) or presence (lanes 4–9) of C-terminally hexahistidine-tagged INs from HIV-1, HIV-2, SIVmac, EIAV, BIV and Ni-NTA agarose beads in a pull down buffer containing 150 mM NaCl. Proteins captured on the resin were separated in a tricine SDS PAGE gel and detected with Coomassie Blue. Lanes 1 and 2 show 100% and 20% Imp7 input, respectively. (C) GST (lanes 3, 6, 9, 12, 15), GST-LEDGF^326–530 (lanes 4, 7, 10, 13, 16), or GST-imp7 (lanes 5, 8, 11, 14, 17) were incubated without (lanes 3–5), or with non-tagged HIV-1 (lanes 6–8, 12–14) or HIV-2 (lanes 9–11, 15–16) INs. The pull-down buffer contained 150 mM (lanes 3–11) or 400 mM (lanes 12–17) NaCl. Lanes 1 and 2 contained input quantities of HIV-1 and HIV-2 INs, respectively. (D) DxR KD and imp7 KD cell populations were infected with VSV-G pseudotyped HIV-1, (pHR'), SIVmac, HIV-2 and EIAV vectors expressing GFP at an MOI of 0.03 and the percentage of GFP+ cells counted 24 hours after infection by flow cytometry. Data are expressed as average percentage of infection relative to control (shDxR) ± SD of two independent experiments performed in duplicate.

Figure 4

Rapid and efficient nuclear accumulation of HIV-1 DNA. Cells were plated in 10 cm plates, co-infected with a HIV-1 and a MLV vector at an MOI of approximately 0.1, incubated at 4°C for 2 hours and then at 37°C for an additional 24 hours. Cytoplasm and nuclei were fractionated and the distribution of MLV vector DNA (A) and HIV-1 vector DNA (B) was examined by Taqman PCR in the same HeLa DxR cells. DxR and imp7 KD cells were co-infected with a MLV and a HIV-1 vector after treatment with 1.5 micrograms/ml aphidicolin for 24 hours. The distribution of MLV vector DNA (C) and HIV-1 vector DNA (D) was examined by Taqman PCR. Bars represent the mean value of two independent experiments. To normalise across experiments, the total amount of viral DNA detected in the cytoplasm was given an arbitrary value of 100.

Figure 5

Imp7 KD impacts on the efficiency of HIV-1 nuclear import. DxR and imp7 KD HeLa cells were infected with an HIV-1 vector (pCSGW) at an MOI of approximately 0.1 and analyzed at the indicated time points by (A) flow cytometry to measure GFP+ cells, (B) Taqman PCR to measure total viral DNA copy number, (C) Taqman PCR to measure 2LTR circular DNA copy number. Data are expressed as mean value ± SD of three independent experiments. Statistical significance was calculated by the Student t-test (n = 3), ** p < 0.005.

Figure 6

Imp7 KD inhibits viral DNA but not viral RNA accumulation in the nuclei. (A) HeLa cells were infected with a DNAse-treated and purified HIV-1 vector and incubated for 2 hours at 4°C and then 4–6 hours at 37°C. Infected cells were fractionated into nuclear and (a) cytoplasmic fractions and nucleic acids extracted, purified and divided into two aliquots. One aliquot was treated with RNAse A, re-purified and used for DNA quantification. The other aliquot was treated with RNAse-free DNAse and used for first-strand cDNA synthesis. Cyclophillin A cDNA was then amplified by PCR in each fraction to examine cross-contamination of nuclear fractions with cytoplasmic material and the overall efficiency of first-strand cDNA synthesis. Cyt, cytoplasmic fractions; Nu, nuclear fractions; W, wild-type virus; M, mutant virus; RT-, cytoplasmic fraction with no RT during first-strand cDNA synthesis; ctr-, primers only; Mw, GeneRuler 100 bp DNA molecular weigh marker. The band migrating at approximately 500 bp is cyclophilin A, lower molecular weigh bands are PCR artefacts. The experiments were performed using 10 fold serial dilutions of the cDNA mix. (B) HIV-1 RNA accumulates in the nuclei shortly after infection. HeLa cells were infected at an MOI of approximately 0.2 with a VSV-G pseudotyped HIV-1 vector (wild type) or with the same amount (p24 normalized) of vector with a mutation in RT and unable to reverse transcribe (RT-). Cells were incubated for 2 hours at 4°C and then 4 hours at 37°C following which samples were fractionated in nuclear and cytoplasmic fractions and treated as described in (A). Taqman PCR was used to measure the amount of viral DNA and RNA in each fraction. First-strand cDNA synthesis reactions carried out in the absence of RT gave undetectable signal. Values shown are average values ± SD of triplicate experiments. Similar results were obtained in two independent experiments. (C) Accumulation of HIV-1 DNA is reduced in imp7 KD cells. HeLa DxR or imp7 KD cells were infected with the same dose of a VSV-G pseudotyped HIV-1 vector, incubated 2 hours at 4°C and then 6 hours at 37°C, following which nuclear and cytoplasmic extracts were prepared and treated as described in (A). After first-strand cDNA synthesis, Taqman PCR was used to measure the amount of viral DNA and RNA in each fraction. First-strand cDNA synthesis reactions carried out in the absence of RT gave undetectable signal. Values shown are average copy number of viral RNA or DNA/μg total nucleic acids ± SD of triplicate experiments. Similar results were obtained in two independent experiments.

Figure 7

Imp7 KD inhibits plasmid DNA transfection. (A) HeLa DxR KD (shDxR), imp7 KD (shimp7) and imp7 back complemented cells (shimp7+7R) were plated onto 6-well plates and transfected the next day with different amounts of the linearized HIV-1 plasmid DNA (pHR') complexed with FuGene 6. Twenty-four hours after transfection, the percentage of GFP+ cells was measured by flow cytometry. (B) HeLa DxR KD cells (DxR) and two different imp7 KD clones (please refer to Figure 2D) were plated onto 24-well plates and transfected as described in panel A. GFP+ cells were counted twenty four hours after transfection by flow cytometry. Note that 1/4 of the DNA was used in experiments shown in panel B compared to panel A because cells were plated onto 24-well plates. Bars represent the average value ± SD of three independent experiments. (C) HeLa DxR and imp7 KD cells were plated into 6 well plates with or without 1.5 micrograms/ml aphidicolin for 24 hours. Cells were then transfected with the indicated amount of linearized plasmid DNA (pHR') as described in (A) and analyzed by flow cytometry 24 hours later. Data are expressed as fold inhibition relative to shDxR cells and are representative of two independent experiments.