Human papillomavirus type 16 E6 induces cell competition

Abstract

High-risk human papillomavirus (HPV) infections induce squamous epithelial tumors in which the virus replicates. Initially, the virus-infected cells are untransformed, but expand in both number and area at the expense of uninfected squamous epithelial cells. We have developed an in vitro assay in which colonies of post-confluent HPV16 expressing cells outcompete and displace confluent surrounding uninfected keratinocytes. The enhanced colony competition induced by the complete HPV16 genome is conferred by E6 expression alone, not by individual expression of E5 or E7, and requires E6 interaction with p53. E6-expressing keratinocytes undermine and displace adjacent normal keratinocytes from contact with the attachment substrate, thereby expanding the area of the E6-expressing colony at the expense of normal keratinocytes. These new results separate classic oncogenicity that is primarily conferred by HPV16 E7 from cell competition that we show is primarily conferred by E6 and provides a new biological role for E6 oncoproteins from high-risk human papillomaviruses.

Fig 1. An assay for enhanced competition by papillomavirus infected cells.

Keratinocytes were labeled by lentiviral transduction with either green (eGFP) or red (Fusion Red) proteins and then transduced with either vector or papillomavirus expressing retroviruses and cultured separately. On day 1 of the assay, 99.9% vector-expressing green cells and 0.1% oncogene expressing red cells are mixed and plated together at 10% confluency in a 10 cm dish. 24 hours later, those co-plated cells are trypsinized and each sample was plated onto 3 glass coverslips in a 6 well plate (2.1 10⁴ cells / cm²). Cells reach confluency by day 5–7 at which point one well is fixed and a second well is fed with F-media for another 7 days and then fixed, and a third well incubated for another 7 days before fixation and staining with DAPI. Pictures of fluorescent colonies are taken with a 4X objective from randomly selected fields and the relative size of the colonies ascertained using NIH ImageJ software.

Fig 2. Enhanced cell colony competition is induced by HPV16 and by HPV16 E6.

Vector-transduced green NIKS cells and oncogene-expressing red NIKS cells were seeded together as described in Fig 1 onto coverslips on day 2 and fixed at confluency on day 5 (Fig 2A, 2C, 2E, 2G, and 2I); a second coverslip was fixed on day 12 (Fig 2B, 2D, 2F, 2H, and 2J). The transduced, selected and stably expressed genes are indicated to the right of each pair. The quantified results of colony sizes (in arbitrary units) from 2 individual representative experiments (out of a total of 4 experiments) is shown (Fig 2K). Western blots for the stably transduced cell lines selected in Fusion Red expressing cells is shown (Fig 2L). MMC refers to treatment with mitomycin C that induces p53 expression, and MG132 refers to the presence of proteasome inhibitor MG132 which blocks the proteasomal degradation of p53. On Day 5 colony sizes were not statistically different between the samples and are not shown. HPV16 and E6 confer enhanced cell colony size at day 12 while E7 does not. Error bars in Fig 2K depict standard error of the mean for colony sizes.

Fig 3. 16E6 induces cell colony competition while 16E5 and 16E7 do not.

The individual HPV16 oncoproteins were transduced into Fusion-Red tagged NIKS cells and set into competition against EGFP-tagged NIKS cells as described in Fig 1; E5 (Fig 2E and 2F), E6 (Fig 2C and 2D), and E7 (Fig 3G and 3H). 16E6 with a stop codon at amino acid 12 (L12X, Fig 3A and 3B) was the negative vector control. Magenta colonies (indicated with a white arrowhead in Fig 3A) are patches of feeder cells and are not included in the colony analysis. Plates were stained at day 6 when confluent and day 17 when super-confluent. Pictures were taken with a 4X objective and relative colony sizes ascertained at day 17 and shown in arbitrary units (Fig 3I) from 4 separate experiments. Day 6 colony sizes were not statistically different between the samples and are not shown. Error bars represent standard error of the mean. **** is P<0.0001; n.s. is not significant. The pictures shown are representative from one of 4 separate assays. Expression of the papillomavirus oncoproteins in the same cell lines used in this experiment is shown in Fig 3J where the blot was sequentially probed with monoclonal antibodies to E7, E6, and the AU1 epitope on E5 and finally GAPDH.

Fig 4. Cell colony competition is induced by E6 in primary keratinocytes.

Primary foreskin keratinocytes maintained in the presence of the rho kinase inhibitor Y-27632 were transduced with the indicated fluorescent proteins and oncogenes and seeded onto glass coverslips as described in Fig 1. One set was maintained in media supplemented with Y-27632 (Fig 4A, 4C, 4E, 4G and 4I), but in a duplicate set Y-27632 was removed after seeding onto glass coverslips (Fig 4B, 4D, 4F, 4H, and 4I). Cells were fixed and stained with dapi on day 21 and red colony sizes ascertained by quantitation of photomicrographs in arbitrary units. Error bars represent the variation in colony sizes normalized to vector-transduced cells from 4 experiments and denote standard error of the mean. *** is P<0.001; ** is P<0.0.01; n.s. is not significant. Western blots for expression of p53, actin, 16E6, 16E7, and GAPDH are shown from cells growing in the presence or absence of Y-27632 in parts Fig 4K and 4L respectively.

Fig 5. Mutational analysis of 16E6-induced cell colony competition.

A. 16E6 mutations that interrupt interaction with p53 (F2V, E18K, F47R), NHERF1 (F69A/K72A), or PDZ-domain proteins (Δ150–151) were introduced into red NIKS keratinocytes and set into competition with green NKS keratinocytes as described in Fig 1. 16E6 with a stop codon at amino acid 12 served as the negative control and colony sizes are normalized to that sample. 16E6 with a stop codon at position 55 (E6*, R55-stop) produces 16E6* but cannot produce full length 16E6. The bar graph combines the results of 4 separate normalized experiments. *** signifies P<0.001, n.s. signifies non-significant differences. Fig 4B. Western blot analysis of 16E6 and mutant 16E6 protein expression. E6* is not shown because it is not recognized by the 6G6 monoclonal anti-16E6 antibody.

Fig 6. E6-expressing keratinocytes extend basally beneath normal keratinocytes, forcing the normal keratinocytes off the attachment substrate.

HPV16 E6 expressing keratinocytes tagged in red were seeded at 0.1% and vector transduced keratinocytes tagged in green were seeded at 99.9% and cultured for 21 days as described in Fig 1. Confocal Images were captured with a 10X objective; the bar at lower left indicates 50 um. Blue color is DAPI stained DNA signal. Fig 6A. Three color image. Fig 6B. DAPI-stained. Fig 6C.! 6E6-transduced cells tagged with Fusion Red. Fig 6D. Overexposure of Fusion Red image with superimposed red line indicating the extent of 16E6 expressing cells. Fig 6E. Red line superimposed onto 3-color image to show the extent of 16E6 expressing cells. Fig 6F. EGFP signal indicating vector-transduced cells. Fig 6G. Overexposure of Fig 6F, with line to indicate extent of vector transduced cells. Fig 6H. Green line from G superimposed onto 3-color image. Fig 6I. 3-color image with red and green lines indicating area of overlapping red-16E6 and green-vector expressing cells. Fig 6J. Three-color image with white boxes indicating the location of z-plane image construction; z-plane images to the right show that 16E6 red cells are basal to vector-green cells. Fig 6K. The same Z-stack image reconstructions shown in Fig 6J are enlarged for clarity; asterix shows the corresponding image and location.