Establishment of human papillomavirus infection requires cell cycle progression

Abstract

Human papillomaviruses (HPVs) are DNA viruses associated with major human cancers. As such there is a strong interest in developing new means, such as vaccines and microbicides, to prevent HPV infections. Developing the latter requires a better understanding of the infectious life cycle of HPVs. The HPV infectious life cycle is closely linked to the differentiation state of the stratified epithelium it infects, with progeny virus only made in the terminally differentiating suprabasal compartment. It has long been recognized that HPV must first establish its infection within the basal layer of stratified epithelium, but why this is the case has not been understood. In part this restriction might reflect specificity of expression of entry receptors. However, this hypothesis could not fully explain the differentiation restriction of HPV infection, since many cell types can be infected with HPVs in monolayer cell culture. Here, we used chemical biology approaches to reveal that cell cycle progression through mitosis is critical for HPV infection. Using infectious HPV16 particles containing the intact viral genome, G1-synchronized human keratinocytes as hosts, and early viral gene expression as a readout for infection, we learned that the recipient cell must enter M phase (mitosis) for HPV infection to take place. Late M phase inhibitors had no effect on infection, whereas G1, S, G2, and early M phase cell cycle inhibitors efficiently prevented infection. We conclude that host cells need to pass through early prophase for successful onset of transcription of the HPV encapsidated genes. These findings provide one reason why HPVs initially establish infections in the basal compartment of stratified epithelia. Only this compartment of the epithelium contains cells progressing through the cell cycle, and therefore it is only in these cells that HPVs can establish their infection. By defining a major condition for cell susceptibility to HPV infection, these results also have potentially important implications for HPV control.

Figure 1. Selected cell cycle inhibitors abrogate HPV infection.

The reporter RL activity (A) and cell viability (B) were measured by Renilla Luciferase Assay System (Promega) and CellTiter-Glo Luminescent cell viability assay (Promega), respectively. Human keratinocyte HaCaT cells were inoculated with HPV16 LCR-driven RL containing pseudovirion hpv16wpA-RL after 4 hr pre-treatment with 3 µM of each compound, and reporter expression/cell viability scored 48 hours later. The expression value is shown as % infectivity or % viability to untreated cells with hpv16wpA-RL.

Figure 2. Cell cycle arrest by serum starvation abrogates HPV infection.

(A) After 24 hr incubation in serum-free DMEM, HaCaT cells were inoculated with HPV pseudovirion hpv16wpA-RL, and RL activity was measured after 2 d. For serum conversion, serum-free DMEM was replaced with 10% FBS/DMEM before virus inoculation. The expression value is shown as % infectivity to untreated cells with hpv16wpA-RL inoculation. Columns, mean; bars, SD. (B) Flow cytometry was performed to confirm cell cycle arrest by serum starvation. The solid line indicates the HaCaT cells incubated with serum-free DMEM for 24 h, and the gray shaded area indicates control HaCaT cells in DMEM with 10% FBS.

Figure 3. Cell cycle progression through M phase is critical for HPV infection.

(A) HaCaT cells were synchronized with aphidicolin (sets 1, 2, 4, and 5) and etoposide (set 3) 24 hrs before hpv16wpA-RL inoculation. Cell cycle progression was released in set 1 and kept blocked in set 2 and 3 using aphidicolin and etoposide, respectively. In set 4, cell cycle progressed through S phase but arrested at G2 phase by switching the compounds in medium from aphidicolin to etoposide. In set 5, cell cycle was released for 20 hrs to progress through M phase and arrest at G2 phase by etoposide addition. Experimental protocols are shown as diagrams. Solid arrow and dotted arrow indicate aphidicolin and etoposide in cell culture medium, respectively. Histograms show the results of flow cytometry analysis following propidium iodide staining at 0, 24, and 48 hr time points using a cell cycle analysis program of FlowJo. (B) The reporter RL activity was measured by Renilla Luciferase Assay System (Promega).

Figure 4. Cell cycle progression through M phase is critical for HPV infection.

Cell cycle release after 24 hr G1 arrest was extended to 24 and 30 hrs before G2 arrest by etoposide. Expression of HPV16 early RNA transcripts was measured using E7 specific primers. Columns, mean; bars, SD. We could not detect any signal above the background in reverse transcriptase–negative controls, indicating that only mRNA expressed from infected virus could be detected (data not shown).

Figure 5. Late M phase arrest does not affect HPV infection.

(A) HaCaT cells were treated with different concentrations of monastrol, and hpv16wpA-RL infectivity was assessed as above. Dotted line is for mock-infected cells. (B) Cell cycle arrest at M phase by monastrol (100 µM) was confirmed using flow cytometry. Red and blue histograms indicate the cell cycle status of DMSO and monastrol-treated cells, respectively.

Figure 6. Early M phase arrest by CDK1 inhibition interferes with HPV infection.

(A) 293T cells were treated with a CDK1 inhibitor, purvalanol A (3, 6, and 12 µM), and inoculated with hpvSEAP. SEAP activity was assayed after 48 hrs incubation as described in Materials and Methods. p<0.05, significantly different from DMSO control. Columns, mean; bars, SD. (B) Cell cycle arrest at M phase by purvalanol A (12 µM) is shown. Red and blue histograms indicate the cell cycle status of DMSO and purvalanol A-treated cells, respectively.

Figure 7. Cell cycle arrest does not inhibit influenza virus.

(A) After 24 hr synchronization with aphidicolin and 4 hr treatment with etoposide or aphidicolin (3 µM each), 293T cells were inoculated in parallel with hpv16wpA-RL or an influenza virus vector in which the hemagglutinin and neuraminidase open reading frames in viral RNA were replaced with those of vesicular stomatitis virus glycoprotein and RL, respectively . (B) After 24 hr synchronization in serum-free DMEM, HaCaT cells were inoculated with hpv16wpA-RL and the RL-expressing influenza virus as (A) in the presence or absence of FBS. RL activity was measured after 48 hrs, as described in Materials and Methods, normalized to RL activity in equivalently inoculated cells maintained in 10% FBS, and expressed in the histograms as % infectivity. Columns, mean; bars, SD.