Inhibition of HIV early replication by the p53 and its downstream gene p21

Human colorectal cancer HCT116 p53^+/+ and HCT116 p53^−/− cell lines were generous gifts from Dr. B. Vogelstein. HEK-293 cells and TZM-bl cells were obtained from NIH AIDS Reagent Program (Germantown, MD, USA). HCT116 p53^+/+, HCT116 p53^−/−, HEK-293 and TZM-bl were propagated in Dulbecco’s modified Eagle’s medium (DMEM) (Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS), 2 mM L-Glutamine, 100 units/ml of penicillin and 100 μg/ml of streptomycin at 37 °C with 5% CO₂. A Countess II Automated Cell Counter (Thermos Fisher Scientific, Waltham, MA, USA) was used to count the number of cells in experiments as designed. Elutriated human monocytes were obtained from the University of Nebraska Medical Center. Donors were de-identified with consent procedures based on the anonymity. IRB approval has been obtained from the Committee on Research at Albany College of Pharmacy and Health Science. Monocytes were differentiated into macrophages (hMDMs) for 10–14 days in DMEM supplemented with 10% human AB serum (VWR, Radnor, PA, USA) by following procedures described previously [27]. Non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells were prepared by 24 h serum starvation. Both cycling and non-cycling cells were tested for cell viability by using trypan blue exclusion assay, and cell death was also measured by WST-1 assay (Roche, Indianapolis, IN, USA). The percentage of live cells was counted by the Countess II Automated Cell Counter (Thermos Fisher Scientific, Waltham, MA, USA) after cells were stained by the Trypan Blue solution (Thermos Fisher Scientific, Waltham, MA, USA). WST-1 assay (Roche) was performed by following kit instruction and O.D. was measured by Eppendorf Plate Reader AF2200 (Eppendorf, Hamburg Germany).

HIV VSV-G-pseudotyped viruses were produced by transient cotransfection of HEK293 cells with proviral HIV-1 or HIV-2 plasmids together with a vesicular stomatitis virus G protein (VSV-G) expression vector pVSV-G (Clontech Laboratories, Inc., Mountain View, CA, USA) by using X-tremeGENE 9 DNA Transfection Reagent (Roche, Indianapolis, IN, USA). HIV-1 plasmids pNL4–3 env(−)nef(−)gfp(+) was a gift from Dr. Vicente Planelles, pNL4–3 env(−)nef(−)luc(+) was a gift from Dr. Nathaniel Landau and HIV-2 luc(+) was a gift from Dr. Lee Ratner. Supernatants containing pseudotyped viruses were harvested 48 h after transfection, passed through 0.45-nm-pore-size filters, and stored at − 80 °C. Viral titers were determined by serial dilution on the TZM-bl indicator cell line as previously described [28]. 1 × 10⁵ cells/well were seeded in a 24 well plate for infection of HCT116 p53^+/+ and HCT116 p53^−/− cells. For non-cycling cells, the complete medium was replaced with DMEM medium without FBS after 24 h, and cells were infected after another 24 h. For cycling cells the medium was replaced with fresh complete medium after 24 h. At time of the infection, cell numbers of paired HCT116 p53^+/+ and HCT116 p53^−/− cells were counted by a Countess II Automated Cell Counter (Thermos Fisher Scientific, Waltham, MA, USA), the same MOI was used for infection in both cells. 0.5 × 10⁶ hMDMs cultured in 24 well plates were used for HIV infection and siRNA experiments. Azidothymidine (AZT) and Efavirenz (EFA) were obtained from NIH AIDS Reagent Program (Germantown, MD, USA) and were dissolved in dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St. Louis, MO, USA). 50 μg/ml AZT or EFA was used in infection experiments as controls. Inactivated virus control was made by heating virus at 65 °C for 1 h.

Luciferase Assay System (Promega, Madison, WI, USA) was used and luciferase assay was performed according to the manufacturer’s instructions. Cells infected with HIV-1 Luc⁺ virus were washed with PBS, and then lysed with lysis buffer. After centrifugation at 15,000×g for 1 min, 20 μl of sample supernatant was mixed with 100 μl of Luciferase Assay Reagent. Luciferase activity was measured in Relative Light Units (RLU) by using a GloMax®-Multi Jr Single Tube Multimode Reader (Promega, Madison, WI, USA).

Flow cytometry was used for both cell cycle analysis and quantification of infection. For cell cycle analysis by propidium iodide staining, cells were washed with PBS, fixed with ice-cold 70% ethanol, and stained with 0.1% (v/v) Triton X-100, 20 μg/ml propidium iodide (PI) (Sigma, St. Louis, MO, USA) and 100 μg/ml DNase-free RNase (Life Technologies, Grand Island, NY, USA). The Click-iT™ Plus EdU Flow Cytometry Assay Kit (Life Technologies, Grand Island, NY, USA) was also used to quantify S phase cells and the kit instruction was followed. For the infection assay, cells were disassociated by trypsin and washed with PBS. The infected GFP⁺ cells and uninfected cells were analyzed and quantified by a BD FACSVerse™ flow cytometer (BD Biosciences, San Jose, CA, USA). The FACSuite (BD Biosciences, San Jose, CA, USA) and the FlowJo (Ashland, OR, USA) software were used for data analysis.

For the quantification of late reverse transcription (RT) products, DNA was extracted from infected cells by using DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany, USA). The early, intermediate, late RT products and HIV-1 2-LTR cycle DNA were quantified by a TaqMan real time PCR, and the relative copy numbers were normalized to reference gene PBGD by using the ΔΔCt method [29]. The integrated HIV-1 provirus copy was also measured by a method described previously by our group. Sample DNA was amplified using the TaqMan Universal PCR Real time Reagent (Life Technologies, Grand Island, NY, USA) in a StepOne Plus real time PCR instrument (Life Technologies, Grand Island, NY, USA). StepOne software was used for quantitative analysis.

Proteins from cells were lysed with RIPA Lysis and Extraction Buffer (Thermos Fisher Scientific, Waltham, MA, USA). After being mixed with Laemmli buffer (BioRad, Hercules, CA, USA), protein samples were heated at 95 °C for 10 min. Protein samples were then separated by SDS-PAGE gel electrophoresis, and transferred onto PVDF membrane (Millipore, Billerica, MA, USA). After being probed with primary and secondary antibodies, protein bands in membranes were detected for chemiluminescence using SuperSignal™ West Pico Chemiluminescent Substrate (Thermo Fisher, Rockford, lL, USA). The primary antibodies used were: anti-p21^Cip1 (#2947) and anti-phospho-SAMHD1 (Thr592) (#89930) (Cell Signaling Technologies, Danvers, MA, USA); anti-SAMHD1 (#12586–1-AP), anti-GAPDH (#60004–1-Ig) and anti-RRM2 (#11661–1-AP) (Proteintech Group, Inc. Rosemont, IL, USA); anti-p53 (#sc-126, Santa Crus, Dallas, TX, USA); anti-actin (#A5441, Sigma-Aldrich, St. Louis, MO, USA) and anti-tubulin (# N-356, Amersham, GE Healthcare, Pittsburgh, PA, USA). Western blot images were detected by the ChemiDoc XRS+ system (BioRad, Hercules, CA, USA), and image analysis was performed by using the Image Lab™ software (BioRad, Hercules, CA, USA).

0.75 × 10⁵ HCT116 p53^+/+ and HCT116 p53^−/− cells were cultured a 24 well plate overnight, siRNA transfections were performed using Lipofectamine® RNAiMAX™ Transfection Reagent (Life Technologies, Grand Island, NY, USA) according to the manufacturer’s instructions. After transfected with siRNA for 2 days, cells were cultured in DMEM without FBS for another 24 h before infection. siRNA transfection of hMDMs was performed two times with a recovery period of two days between transfections to ensure knockdown of target mRNA. siRNA knockdown was confirmed by Western blot. The Silencer Select validated siRNA siRNA p21^Cip1 (#4390824) and negative control non-target siRNA (#4392420) were purchased from Life Technologies (Life Technologies, Grand Island, NY). siRNA p21 has a sequence of 5’-UAAAAUGUCUGACUCCUUGTT-3’. The FlexiTube siRNA p53 (# SI02655170) was purchased from Qiagen (Qiagen, Hilden, Germany, USA) and has a sequence of 5’- ACUCCACACGCAAAUUUCCTT-3’.

It was found by our group that the infection of MLV vector based retrovirus was inhibited significantly in non-cycling HCT116 p53^+/+, while the inhibition was attenuated in non-cycling HCT116 p53^−/− cells [26]. To investigate whether p53 also inhibits HIV-1 infection depending on cell cycle status, non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells were prepared by serum starvation for 24 h, then the cell cycle statuses of both cycling and non-cycling cells were analyzed by flow cytometry after PI staining and EdU incorporation. The majority of HCT116 p53^+/+ and HCT116 p53^−/− cells were in S and G2 phases when cultured in complete medium with 10% FBS (Fig. 1a). After being cultured in DMEM without FBS for 24 h both HCT116 p53^+/+ cells and HCT116 p53^−/− cells became non-cycling cells, with very low percentages of cell populations in S phase and G2 phase (≤10%) (Fig. 1a). The cell viabilities of both cycling and non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells were analyzed by trypan blue exclusion test and by WST-1 assay (Fig. 1b). The results showed that there were no cell viability differences between cycling and non-cycling cells in HCT116 p53^+/+ and in HCT116 p53^−/− before infection.

Both cycling and non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells were infected by 2.0 MOI VSV-G pseudotyped HIV-1 GFP⁺ virus (Fig. 1c and d). In cycling cell status both HCT116 p53^+/+ and HCT116 p53^−/− were highly permeable to HIV-1 infection, and the infection in HCT116 p53^+/+ cells were inhibited by about 1.7 fold in comparison to HCT116 p53^−/− cells. However the fold of inhibition in HCT116 p53^+/+ increased to 4.6 times in non-cycling cells (Fig. 1c and d). To confirm this observation, both cycling and non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells were infected by 1.0 and 3.0 MOI of VSV-G pseudotyped HIV-1 Luc⁺ virus respectively. In 1.0 MOI HIV-1 infection, the inhibition changed from about 2.6 fold to 5.6 fold, and in 3.0 MOI infection, the observed inhibition in HCT116 p53^+/+ cells increased from 3.6 fold to 9 fold in comparison to HCT116 p53^−/− cells when cell cycle switched from cycling to non-cycling status (Fig. 1e). The amount of infection was dependent on virus dosage and no luciferase activities were detected in both uninfected cells and AZT treated cells. These results indicated the HIV-1 infection can be blocked by the presence of p53, and p53 dependent inhibition was increased with cell cycle change from cycling to non-cycling status.

In order to investigate whether HIV-2 infection is also inhibited by p53 dependent on cell cycle status, VSV-G pseudotyped HIV-2 Luc⁺ virus was used to infect both cycling and non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells (Fig. 1f). HIV-2 infection was significantly inhibited in HCT116 p53^+/+ in comparison to HCT116 p53^−/− cells. The inhibition in HCT116 p53^+/+ was about 9 folds in cycling cells when infected with 1 x and 5 x HIV-2 Luc⁺ viruses respectively, and inhibition in HCT116 p53^+/+ was about 20 and 15 folds in non-cycling cells when infected with 1 x and 5 x HIV-2 Luc⁺ viruses respectively (Fig. 1f). This finding suggested the inhibition to overall HIV-2 infection was increased compared to the inhibition to HIV-1 infection in HCT116 p53^+/+, and the inhibition to HIV-2 was also increased with cell cycle change from cycle to non-cycling status.

We found previously that p53 inhibited retrovirus reverse transcription [26]. To identify whether p53 dependent HIV-1 inhibition in non-cycling cells also happened at reverse transcription, a TaqMan real time PCR was used to quantify HIV-1 reverse transcription (RT) late product in infected cells. It was found there was no difference in amount of RT products between cycling HCT116 p53^+/+ and cycling HCT116 p53^−/− cells at 16 h post infection, however HIV-1 RT product was decreased about 2.1 times in HCT116 p53^+/+ cells compared to HCT116 p53^−/− cells in non-cycling cells (Fig. 2a). In order to identify whether the decrease of RT product in HCT116 p53^+/+ cells was either due to the inhibition in the process in viral cDNA synthesis by reverse transcriptase or because of possible degradation of virus cDNA by nucleases in infected cells, the amount of remaining late RT product was quantified at 1.5 h, 3.0 h and 4.5 h after cells were treated with the reverse transcriptase inhibitor EFA at 12 h post infection. No difference was observed in the percentages of virus cDNA degradation between non-cycling HCT116 p53^+/+ and HCT116 p53^−/− cells after blocking reverse transcriptase by EFA (Fig. 2b). This result indicated that the observed p53 dependent inhibition occurred in the synthesis of virus cDNA by HIV-1 reverse transcriptase. In a separate HIV-1 infection experiment, the amount of virus cDNA at different reverse transcription stages were quantified (Fig. 2c-e). HIV-1 early, intermediate and late reverse transcription products were found to be consistently decreased in non-cycling HCT116 p53^+/+ cells compared to non-cycling HCT116 p53^−/− cells over 8 h, 16 h and 24 h after infection (Fig. 2c-e). HIV-1 2-LTR cycle DNA and integration were also quantified (Fig. 2f and g). The amount of decrease of 2-LTR cycle DNA and integrated HIV-1 provirus copy in non-cycling HCT116 p53^+/+ cells in comparison to non-cycling HCT116 p53^−/− cells were about 2 times less, which are similar to the observed level of inhibitions in different RT stages. These results indicated that the block of HIV-1 infection in non-cycling HCT116 p53^+/+ cells occurred at the reverse transcription stage mostly by a mechanism involving in the processing of HIV-1 reverse transcription.

To investigate the p53 gene function change when cell cycle was changed from cycling to non-cycling status, and its response to HIV-1 infection, Western blot experiments were carried out to analyze the changes of p53, phosphorylated p-p53(S15) and the p53 downstream gene p21 (Fig. 3a and b). It was found that both the levels of p53 protein and p21 protein were significantly increased in non-cycling HCT116 p53^+/+ cells in comparison to cycling HCT116 p53^+/+ cells (Fig. 3a and b). p-p53 (S15) was slightly elevated at 1 h after HIV-1 infection in cycling HCT116 p53^+/+ cells, but no difference of p-p53 (S15) was found in non-cycling HCT116 p53^+/+ cells. Western blots were also performed on cycling and non-cycling HCT116 p53^−/− cells in the course of HIV-1 infection (Fig. 3a and b). The expression level of p21 was low in HCT116 p53^−/− cells and there were no significantly difference in p21 protein levels in HCT116 p53^−/− cells between cycling and non-cycling cells. This data suggests that the cell cycle status change resulting from serum starvation induced p53 expression, which in turn regulated the expression of p21 in HCT116 p53^+/+ cells. However the increase of p21 resulting from cell cycle change did not occur in HCT116 p53^−/− cells, which is in concordance with the lack of regulation by a functional p53.

It has been shown that p21 inhibits HIV reverse transcription by regulating cellular dNTPs level in differentiated macrophages. p21 was found to phosphorylate SAMHD1 at Thr592 [20], which inactivates its activity to degrade cellular dNTPs. p21 was also reported to down regulate expression of RNR2 [17], an enzyme responsible for cellular synthesis of dNTPs. Western blot experiments were carried out to compare p21, SAMHD1, pSAMDH1(T592) and RNR2 protein levels between non-cycling HCT116 p53^+/+ and non-cycling HCT116 p53^−/− cells (Fig. 4a). The levels of p21 protein were significantly higher over the course of HIV-1 infection in non-cycling HCT116 p53^+/+ in comparison to non-cycling HCT116 p53^−/− cells (Fig. 4a and b). There was no difference found in overall SAMHD1 levels (Fig. 4a and c), however both pSAMDH1 (T592) and RNR2 protein levels were elevated over multiple time points over HIV-1 infection in non-cycling HCT116 p53^+/+ in comparison to non-cycling HCT116 p53^−/− cells (Fig. 4a and d, e), even though the differences were not statistically significant.

To further verify that the increase in p53 and p21 at protein level was responsible for the inhibition of HIV-1 reverse transcription in non-cycling cells, HCT116 p53^+/+ cells were transfected with either p53 siRNA or p21 siRNA. The non-targeting (NT) siRNA was used as negative control. After 48 h the medium were replaced with DMEM without FBS and cultured for another 24 h, and then siRNA transfected cells were infected with HIV-1 virus. As shown in Fig. 5a, both knockdown of p53 and knockdown of p21 by siRNA transfection were confirmed by Western blot experiment (Fig. 5a). Transfection with p21 siRNA resulted in knockdown of p21 to 13% (Fig. 5b). Transfection with p53 siRNA also led to knockdown of p21 to 18%. Transfection with p53 siRNA resulted in knockdown of p53 to 14% (Fig. 5c). To analyze the potential role of p53 and p21 in the regulation of genes responsible for maintaining host cell dNTPs pool size, SAMHD1, pSAMHD1(T592) and RNR2 protein levels were compared between HCT116 p53^+/+ cells transfected with p53 or p21 gene specific siRNA and non-targeting siRNA (Fig. 5a). It was found that the knockdown of p53 increased both RNR2 and pSAMHD1 (T-592), and the knockdown of p21 increased RNR2 and increased pSAMHD1 (T-592) slightly as well. Either the knockdown of p53 or the knockdown of p21 did not change SAMHD1 at protein level significantly (Fig. 5a). The cell cycle status after siRNA treatment was also tested by Click-iT™ Plus EdU Flow Cytometry Assay. There were a minor increase in S phase in siRNA p21 treated cells (15.5%) and siRNA p53 treated cells (13.5%) compared to siRNA NT treatment (12.9) (Fig. 5d). siRNA knockdown of p21 in non-cycling HCT116 p53^+/+ cells significantly increased infection of HIV-1 Luc⁺ virus, and siRNA knockdown of p53 in non-cycling HCT116 p53^+/+ cells also increased infection of HIV-1 Luc⁺ virus but the increase was not statistically significant (Fig. 5e). Meanwhile, either p53 siRNA or p21 siRNA treatment significantly increased HIV-1 reverse transcription in non-cycling HCT116 p53^+/+ cells (Fig. 5f). The above data indicated that it was the elevated p53 protein or p21 protein in non-cycling HCT116 p53^+/+ cells that caused the inhibition in HIV-1 reverse transcription.

To identify whether p53 and p21 impact HIV-1 infection in its natural host cells, VSV-G pseudotyped HIV-1 Luc⁺ was used to infect hMDMs. Protein levels of p53, p-p53(S-15) and p21 were analyzed by Western blot at multiple time points during the early stage of HIV-1 replication (Fig. 6a and b). In donor 1, p53 protein level was increased at 1 h after infection. The amount of p21 protein started to increase at 2 h post infection, and higher amounts of p21 proteins were found as early as 4 h and 8 h post infection. In donor 2, p21 protein was also increased at time points of 8 h and 16 h after infection, and there was minor increase in p53 proteins at time points of 4 h, 8 h and 16 h. No noticeable changes were observed in p-p53 (S-15) in infected hMDMs. Interestingly, p21 protein levels were increased in hMDMs in both donors at an early stage of HIV-1 replication, approximately when reverse transcription is in process. This result strongly suggests the role of p21 in the restriction to early stage of HIV-1 replication in hMDMs.

The function of p53 and p21 in HIV-1 infection was further examined by siRNA experiments. After the siRNA transfection, p21 protein level dropped about 28% in cultured hMDMs (Fig. 6c and d). The knockdown of p21 increased both RNR2 and pSAMHD1(T592), and HIV-1 infection increased significantly (Fig. 6c and e). After siRNA transfection, p53 protein level dropped about 40% in cultured hMDMs. The knockdown of p53 increased pSAMHD1(T592), and HIV-1 infection was increased but not statistically significant (Fig. 6c-e). There were no obvious changes in RNR2 protein levels after knockdown of p53, and knockdown of either p21 or p52 did not alter total SAMHD1 protein levels.

p21 was previously found to restrict HIV-2 infection [15, 16]. The Vpx protein from HIV-2 is able to degrade cellular restriction factor SAMHD1 [30], while p21 can inactivate SAMHD1 by phosphorylation [20]. To investigate the role of p21 in HIV-2 infection, hMDMs were transfected with p21 siRNA, then infected by HIV-2 Luc⁺ virus. In this experiment siRNA transfection was able to knockdown p21 to about 52% (Fig. 6f, g). In control hMDMs transfected with non-targeting siRNA, HIV-2 infection resulted in the decrease in SAMHD, and the levels of pSAMHD1 (T592) were almost undetectable before and after HIV-2 infection (Fig. 6f). p21 siRNA knockdown did not change SAMHD1 levels but greatly increased pSAMHD1 (T592) in both HIV-2 infected and uninfected hMDMs (Fig. 6f). Finally it was found that knockdown p21 in hMDMs significantly increased HIV-2 infection (Fig. 6h). This result also suggested that the degradation of SAMHD by HIV-2 Vpx was independent from phosphorylation of SAMHD1 resulted from p21 knockdown by siRNA in hMDMs.