Figure 1.
Brd4 localizes to nuclear foci formed by HPV16 E1 and E2 proteins.
Human keratinocytes were transfected with pMEP4/9 (empty), pMEP9-HPV16 E1 and pMEP4-HPV16 E2 expression vectors, as indicated. E1, E2 and endogenous Brd4 proteins were detected by indirect immunofluorescence. Staining for E1 proteins is shown in green, E2 proteins in cyan, and Brd4 in red. High resolution 3D images reconstructed by deconvolution of stacks of confocal images are shown. At least five cells were digitally reconstructed from each condition and a representative image is shown. The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). A. Brd4 staining in a single optical slice through the center of the nucleus. B. An optical section through nuclei expressing E1, E2 or E1/E2. The crosshairs show the cut point through the x-, y-, and z-axes of the data set. These are the same cells shown in Figure 1A. C. An enlargement of a portion of the nuclei from the images shown in Figure 1B. The scale bar represents 1 µm.
Figure 2.
Salt extraction of cells expressing HPV16 E1 and E2 show that the E1 protein binds most tightly to host nuclei.
Human keratinocytes were cotransfected with expression vectors for HPV16 E1 (pMEP9-E1), HPV16 E2 (pMEP4-E2), or both proteins. Cells were either fixed directly in paraformaldehyde (PFA), or proteins were extracted from cells in buffers containing either 100 mM or 300 mM NaCl prior to fixation in PFA. A. E1, E2 and endogenous Brd4 proteins were detected by indirect immunofluorescence. Staining for the E1 protein is shown in green, the E2 protein in cyan, and Brd4 in red. The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). B. Level of E1, E2, and Brd4 were measured in individual cells extracted and fixed as described above using the Contour feature of Imaris software (Bitplane) to extract values from each channel in individual nuclei. Average values obtained from at least five reconstructed nuclei are shown and the error bars represent the standard deviation. Individual transfected cells could not be identified in cells transfected with the empty pMEP4/9 vectors or in cells transfected with HPV16 E2 vectors and extracted in 300 mM NaCl containing buffer. In these samples, random cells were selected for quantitation.
Figure 3.
Localization of Brd4 to E1–E2 replication foci changes in the presence of the replication origin.
Human keratinocytes were cotransfected with expression vectors for HPV16 E1 and E2 in the absence or presence of a plasmid containing the minimal HPV16 replication origin, p16ori. A. In the panels on the left, keratinocytes were transfected without the origin plasmid (instead the control plasmid pKS bluescript was used) and on the right in the presence of the origin. High resolution 3D images were reconstructed by deconvolution of stacks of confocal images and a representative optical slice is shown for a representative cell with and without the origin. Staining for E1 proteins is shown in green, E2 proteins in cyan, and the Brd4 protein in red. The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). B. To quantitate the amount of E1, E2 and Brd4 in nuclear foci in the absence or presence of the origin containing plasmid, at least five cells were digitally reconstructed from each condition by deconvolution of high resolution 3D stacks of confocal images. The replication foci were demarcated using the Surpass module of Imaris software (Bitplane) to define volumes expressing high levels of the E1 protein. These foci are shown in green on the image in the left panel. The proportion of E1, E2 and Brd4 in the E1 foci (levels in E1 foci volume/levels in nuclear volume) is represented in the graph in the middle. The graph on the right shows the enrichment of each protein in the E1 foci relative to the volume of the nucleus (levels of each protein multiplied by volume of E1 foci/volume of nucleus). Average values obtained from at least five reconstructed nuclei are shown and the error bars represent the standard deviation.
Figure 4.
In the presence of a replicating viral origin, the size of E1/E2 foci increases and Brd4 is displaced.
A. Human keratinocytes were cotransfected with expression vectors for HPV16 E1 and E2 in the presence of a plasmid containing the minimal HPV16 replication origin, p16ori (+ ori) or a control plasmid without the origin, pKS (− ori). The E1 and Brd4 proteins were detected by immunofluorescence. Representative foci from each condition are shown. The diameter of E1 foci was measured using Bitplane Imaris. Error bars represent the standard error of the mean. B. Human keratinocytes were cotransfected with expression vectors for HPV16 E1 and E2 in the presence of a plasmid containing the minimal HPV16 replication origin, p16ori (+ ori) or a control plasmid without the origin, pKS (− ori). E1 and E2 expression was induced four hours before fixation in the presence of 25 µg/ml AraC and E1, E2 and Brd4 proteins were detected by immunofluorescence. Representative cells from each condition are shown. The diameter of E1 foci was measured using Bitplane Imaris. Error bars represent the standard error of the mean.
Figure 5.
HPV origin containing plasmids replicate to high levels in E1/E2 nuclear foci.
Human keratinocytes were cotransfected with expression vectors for HPV16 E1 and E2 in the presence of a plasmid containing the minimal HPV16 replication origin, p16ori (+ ori) or a control plasmid without the origin, pKS (− ori). The E1 and Brd4 proteins were detected by immunofluorescence, and following a brief fixation, origin containing DNA was detected by FISH. The images shown are optical slices from cells that were digitally reconstructed by deconvolution of high resolution 3D stacks of confocal images. E1 is shown in green, Brd4 in cyan, and origin DNA in red. The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown).
Figure 6.
Brd4 surrounds replication foci in differentiating cells harboring HPV31 genomes.
A and B Replication foci were detected by FISH for HPV31 DNA in uninfected keratinocytes (i), and undifferentiated (ii) and differentiated (iii and iv) CIN-612 9E cells. C and D Replication foci were detected by immunofluorescence for γH2AX (cyan) and were also stained for Brd4 (red) DNA in uninfected keratinocytes (i), and undifferentiated (ii) and differentiated (iii and iv) CIN-612 9E cells. A, B, C, D. 3D image stacks were deconvolved and digitally reconstructed using Huygens Essential software. Volume images are shown in A and C and high resolution surface renderings of all objects were generated by Bitplane Imaris software and are shown in B and D. E The γH2AX and Brd4 proteins were detected by immunofluorescence in differentiated CIN-612 9E cells., and following a brief fixation, viral DNA was detected by FISH. Volume images are shown in i and high resolution surface renderings of all objects were generated by Bitplane Imaris software and are shown in ii, iii and iv. γH2AX is shown in cyan, Brd4 in red, and HPV31 DNA in green.
Figure 7.
Homologous recombination marker, Rad51, colocalizes with HPV replication foci.
A. (i) Nuclear foci were generated by differentiation of CIN612 9E cells for five days. Cells were stained with antibodies against Brd4 (rabbit mcb2), Rad51 (mouse) and γH2AX (mouse). Isotype specific secondary antibodies were used to distinguish between the mouse antibodies. Cellular DNA is counterstained with DAPI (blue in merged panels). 3D image stacks were deconvolved using Huygens Essential software. (ii) High resolution surface renderings of all objects generated by Bitplane Imaris software. (iii) High resolution surface rendering of the replication foci shown in (ii). (iv) Cross-section through the replication foci shown in (i, ii, iii). Rad51 and γH2AX are localized within the foci and are surrounded by small speckles of Brd4. B. Nuclear foci were generated by transfection of HPV16 E1 and E2 expression vectors (with and without p16ori). Cells were stained with antibodies against E1, Brd4 (rabbit mcb2) and Rad51 (mouse). The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). C. The percentage of HPV16 E1/E2 foci containing Rad51, obtained from experiments as described in B, is represented by the blue bars. The percentage of γH2AX foci containing Rad51 in differentiated CIN612 9E cells (as described in A) is represented by the red bar. Error bars represent standard deviation.
Figure 8.
Comparison of the characteristics of foci formed by expression of E1 and E2 proteins and naturally occurring replication foci in differentiated HPV containing cells.
HFKs were transiently transfected with HPV16 E1 and E2 expression plasmids and the control plasmid pKS (left column), or with HPV16 E1 and E2 expression plasmids and the replication origin plasmid p16ori (middle column). HPV31 genome containing 9E cells (right column) were differentiated for five days as described in Methods. Cells were stained for E1, E2, Brd4, H3K4me1, H4K8ac and H3K56ac, Rad51, pATM and γH2AX, as indicated. The rabbit Brd4 mcb2 antiserum is used in most panels, except for H3K4me1, H4K8ac and H3K56ac staining in HPV31 9E cells (right column), where a mouse Brd4 antibody 8H2 was used. A goat anti-Rad51 antibody was also used in these panels. A mouse Rad51 antibody was used to stain HPV31 9E cells, along with Brd4 (rabbit mcb2) and γH2AX (mouse). Isotype specific secondary antibodies were used to distinguish between the mouse antibodies. Cellular DNA is counterstained with DAPI (blue in merged panels).
Table 1.
Properties of replication foci.
Figure 9.
Brd4 is required for the formation of E1–E2 replication foci but viral genome replication obviates this need.
A. Effect of BET inhibition of Brd4 chromatin binding on E1–E2 foci. Keratinocytes were transfected with HPV16 E1 and E2 and treated with DMSO (solvent), 2 µM GSK525762A+ (a BET inhibitor) and 2 µM GSK525762A− (an inactive enantiomer of GSK525762A+), 0.5 µM JQ1 or a mixture of 10 mM KU-55933 and 2.5 mM caffeine. Replication foci were detected by immunofluorescence for E1 (green), Brd4 (red) and E2 (cyan). B. The percentage of E1/E2 expressing cells with nuclear E1/E2 containing foci with or without Brd4 after treatment with BET inhibitors. C. The percentage of E1/E2 expressing cells with nuclear E1/E2 containing foci with or without Brd4 in the presence of an origin containing replicon (p16ori) after treatment with BET inhibitors. D. HFKs were treated with either Brd4 siRNA or control siRNA for 24 hours before transfection with HPV16 E1 and E2 expression plasmids. Replication foci were detected by immunofluorescence for E1 (green), E2 (cyan) and Brd4 (red). The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). E. The percentage of E1/E2 expressing cells with nuclear E1/E2 containing foci with or without Brd4 after treatment with siRNA to Brd4 in the presence or absence of an origin containing replicon after treatment with siRNA to Brd4. F. Western blot showing reduction of Brd4 expression upon siBrd4 treatment.
Figure 10.
Analysis of the E1–E2 functions required for localization of Brd4 to nuclear E1–E2 foci in the presence and absence of viral origin containing DNA.
A. Representation of mutations that specifically inactivate E1 and E2 functions. E1: K230A, S231A (defective in specific DNA binding); K483A (ATPase deficient). E2: E39A (deficient in interaction with E1); R37A, I73A (deficient in transcriptional regulation and Brd4 binding); R302K, R304K (DNA binding defective). B. Wild-type or mutated HPV16 E1 and E2 expression plasmids were transiently transfected into keratinoctytes and cells were stained by immunofluorescence for E1 (green), E2 (cyan) and Brd4 (red). The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). The cells in the left panel were also transfected with a control plasmid (pKS) and those on the right with a minimal origin containing plasmid (p16ori). Fluorescent signals were collected at unsaturated levels, according to viral protein expression, and were later increased to enhance figure brightness. C. Cells expressing E1 and E2 in the presence or absence of origin containing DNA were counted and scored for the presence of foci that co-localized with Brd4.
Figure 11.
Replication and nuclear localization of HPV31 genomes containing mutations in E2 that disrupt Brd4 binding.
A. HFK cells were transfected with recircularized HPV31 genomes as described in Methods. The resulting drug-resistant colonies were pooled, cultured for eight passes and analyzed by Southern blot analysis. WT: wild type genome; RK: HPV31 E2 R37K; IL: HPV31 E2 I73L; pUC: control plasmid transfected cells. Total cellular DNA was either uncut (U) or digested with a restriction enzyme that cleaved the HPV31 genome once (C). The linear and supercoiled forms of the HPV31 genomes are indicated. B. An expression plasmid encoding HPV16 E2 with an I73L substitution mutation was transiently transfected into keratinoctytes with either control empty vectors, pMEP9 and pKS (top row); or HPV16 E2 I73L and wildtype E1 expression plasmids and control empty vector, pKS (middle row); or HPV16 E2 I73L, HPV16 E1 and p16ori plasmids (bottom row). Cells were stained by immunofluorescence for E1 (green), E2 (cyan) and Brd4 (red). The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). C. Immunofluorescence of differentiated normal HFKs and the HFKs described in A (immortalized with HPV31 wildtype and HPV31 E2 I73L genomes. Cells were differentiated with calcium for five days and stained for viral replication foci with γH2AX (red) and Brd4 (green). The dashed blue lines represent the perimeter of the nuclei (identified by DAPI staining, not shown). 3D image stacks were reconstructed by deconvolution using Huygens Essential software, and a single slice is shown. The scale bar represents 5 µm. D. Immunofluorescence of differentiated CIN-612 9E cells, W12 20863 cells, and B1 (a laboratory generated HPV18 cell line). Cells were differentiated with calcium for five days and stained for viral replication foci with γH2AX (red) and 2290, an antibody that recognizes all forms of Brd4 (green). Nuclei were identified by DAPI staining. The scale bar represents 10 µm.