Tel: 0049-6221-424480, Fax: 0049-6221-423755
Infectious Causes of Cancer
NF-κ B-dependent upregulation of ICAM-1 by HPV16-E6/E7 facilitates NK cell/target cell interaction
Article first published online: 6 MAY 2010
Copyright © 2010 UICC
International Journal of Cancer
Volume 128, Issue 5, pages 1104–1113, 1 March 2011
How to Cite
Textor, S., Accardi, R., Havlova, T., Hussain, I., Sylla, B. S., Gissmann, L. and Cerwenka, A. (2011), NF-κ B-dependent upregulation of ICAM-1 by HPV16-E6/E7 facilitates NK cell/target cell interaction. Int. J. Cancer, 128: 1104–1113. doi: 10.1002/ijc.25442
- Issue published online: 28 DEC 2010
- Article first published online: 6 MAY 2010
- Manuscript Accepted: 12 APR 2010
- Manuscript Received: 5 NOV 2009
- Human Papillomavirus;
- cervical cancer;
- NK cells;
NK cell recognition of tumor cells is mediated by a delicate balance of signals received by MHC class I-binding inhibitory NK cell receptors and activating NK cell receptors, which mainly bind to virus-, stress- or tumor-induced ligands. In addition, adhesion molecules such as the intercellular adhesion molecule-1 (ICAM-1) and its receptors, the lymphocyte function-associated antigen-1 (LFA-1) and Mac-1, are crucial for immune synapse formation and NK cell-mediated killing. In this study, we show that expression of the adhesion molecule ICAM-1 was rapidly induced by E6 and -E7 oncoproteins of HPV16, -18, -5 and -8, but not of HPV38 and -6 in primary human keratinocytes after retroviral transduction. ICAM-1 was upregulated in E6E7-expressing keratinocytes both at mRNA and protein levels. The observed ICAM-1 upregulation in HPV16-E6E7-expressing keratinocytes was partially dependent on activation of the NF-κB pathway. Importantly, the upregulated ICAM-1 expression in HPV16-E6E7-expressing keratinocytes led to enhanced conjugate formation with NK cells. We previously showed that HPV16-positive cervical carcinomas frequently express low levels of inhibitory NK cell ligands and high levels of activating NK cell ligands. Moreover, levels of the adhesion molecule ICAM-1 are enhanced by HPV16-E6/E7. Therefore, strategies that aim at harnessing NK cells might be beneficial for the treatment of cervical carcinoma.
Human papillomaviruses (HPV) are a family of small double-stranded DNA viruses including more than 100 different members that cause diseases in humans ranging from benign warts to malignant cancers. The two main HPV genera are the Alpha and Beta papillomaviruses.1 Alpha papillomaviruses include the so-called “high-risk” mucosal HPV types, e.g., HPV16 and HPV18 that are the causal agents of the development of malignant cancer of the cervix.2 Beta papillomaviruses, including HPV5, HPV8 and HPV38, are typically associated with unapparent cutaneous infections in humans. Nevertheless, in particular, in immunocompromised individuals, these viruses are associated with nonmelanoma skin cancer (NMSC).3 The two early viral proteins E6 and E7 of “high-risk” Alpha papillomaviruses, such as HPV16, are well-characterized oncogenes leading to a severe deregulation of the cell cycle by degradation of p534 or inhibition of pRb,5, 6 respectively. Little is known about the signaling pathways affected by E6 and E7 from Beta papillomaviruses.3
Activation of NK cells is mediated by a delicate balance of signals received by inhibitory and activating receptors.7 Inhibitory receptors mostly recognize self-MHC class I molecules and activating receptors interact with virus-, stress- or transformation-induced ligands.8, 9 Thus, low MHC class I expression combined with high levels of activating NK cell receptor ligands make cells susceptible to NK cell-mediated killing. In addition to the “right” combination of activating and inhibitory NK cell receptor ligands on target cells, adhesion molecules on targets and NK cells are crucial for immune synapse formation and NK cell-mediated killing.10 Firm NK cell/target cell adhesion is mainly mediated by the two integrins, LFA-1 and Mac-1, on NK cells and ICAM-1 on target cells.11 The adhesion molecule ICAM-1 mediates not only cell–cell but also cell–matrix interactions.12 ICAM-1 is expressed on hematopoietic and nonhematopoietic cells13 and is rapidly upregulated in response to inflammatory stimuli, including virus infections or pro-inflammatory cytokines.12 Depending on the stimulus, ICAM-1 expression is regulated by several signal transduction pathways, including protein kinase C (PKC), mitogen-activated protein (MAP) kinases, Janus kinases (JAKs), signal transducers and activators of transcription (STAT) or NF-κB signaling pathway.14
In HPV-associated diseases, high ICAM-1 expression was observed in situ by immunohistochemistry.15, 16 Furthermore, increased expression of ICAM-1 in high-grade cervical intraepithelial neoplasia (CIN)15 and cervical carcinoma (CxCa) specimen16 was reported. Studies using immortalized or transformed in vitro-cultured HPV16-positive cell lines confirmed these results. ICAM-1 was reported to be upregulated in HPV16-immortalized human oral keratinocytes compared to normal human oral keratinocytes.17 In contrast, Coleman et al. described ICAM-1 expression only in fully transformed tumorigenic cells such as in the HPV16+ CxCa cell lines SiHa and CaSki, but not in HPV16-immortalized nontumorigenic human keratinocytes.15 In these studies, the mechanisms of ICAM-1 upregulation were not addressed.
Here, we describe for the first time that E6 and E7 of HPV16 rapidly induced ICAM-1 surface expression in their natural host cells, primary human keratinocytes. This ICAM-1 upregulation was partially mediated by the NF-κB pathway. Furthermore, our results reveal that significantly more conjugates were formed between NK cells and HPV16-E6E7-expressing cells as compared to control cells. Thus, expression of HPV16-E6/E7 in primary human keratinocytes might alert the immune system by upregulation of ICAM-1 that facilitates immune cell adhesion.
Material and Methods
Phoenix ampho cells were cultured in DMEM containing 4.5 g/ml glucose and 2 mM l-glutamine (Sigma-Aldrich, Munich, Germany) supplemented with 10% FBS (PAA, Pasching, Austria). Primary human epidermal keratinocytes from adult or neonatal foreskin were obtained from Invitrogen and cultured in EpiLife® medium supplemented with 1% human keratinocyte growth supplement, HKGS, containing 0.2 % bovine pituitary extract, 5 μg/ml bovine insulin, 0.18 μg/ml hydrocortisone, 5 μg/ml bovine transferrin and 0.2 ng/ml human epidermal growth factor (Invitrogen, Karlsruhe, Germany). The cervical cancer-derived cell line, CaSki, was maintained in DMEM containing 1 g/ml glucose and 2 mM L-glutamine supplemented with 10% FBS. The human NK cell line, NKL,18 was cultured in RPMI containing 1 g/ml glucose and 2 mM L-glutamine supplemented with 10% FBS and 100 U/ml human IL-2 (Chiron Corporation, Emeryville, CA). All cells were cultured in the presence of 100 U/ml penicillin G and 100 μg/ml streptomycin (Gibco/BRL, Karlsruhe, Germany).
The plasmids pLXSN, pLXSN-16E6, pLXSN-16E7, pLXSN-16E6E7, pLXSN-38E6E7, pLXSN-5E6E7, pLXSN-8E6E7, pLXSN-18E6E7, pLXSN-6E6E7 and pBabepuro-16E7 were kindly provided by M. Tommasino (IARC, Lyon, France) and the plasmids pBabe-puro and pBabe-puro-ΔN-I-κBα by B. Sylla (IARC, Lyon, France). The pBabe-puro-ΔN-I-κBα was generated by inserting the coding sequence of I-κBα lacking the first 32 amino acids into the pBabe-puro retroviral vector. An IRES-EGFP sequence was cloned 3′ of the multiple cloning site in pLXSN, pLXSN-16E6, pLXSN-16E7 and pLXSN-16E6E7 plasmids as a marker genes. Retroviral supernatants generated by transfection of the pLXSN and pBabe-puro constructs into Phoenix ampho cells were used to transduce primary human epidermal keratinocytes as previously described.19 Forty-eight hours after transduction, keratinoctyes were analyzed by flow cytometry for expression of the gene of interest gating on EGFP-positive cells and/or selection with 100 μg/ml G418 (Gibco/BRL, Karlsruhe, Germany) or 0.2 μg/ml puromycin (Sigma-Aldrich, Munich, Germany) was started.
RNA extraction, reverse transcription and PCR
Total RNA was extracted with the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Residual contaminating DNA was removed with the TURBO DNA-free™ Kit (Ambion, Huntingdon, UK). One microgram RNA was reversely transcribed using the SuperScript™ First-Strand Synthesis system for RT-PCR (Invitrogen, Karlsruhe, Germany) and Oligo (dT)12–18 primer. To exclude false positive results due to contaminating genomic DNA, samples without reverse transcriptase were included.
Nonquantitative PCR experiments for HPV-E6/E7 transcripts were performed on a Mastercycler (Eppendorf, Hamburg, Germany) with the following program: 95°C for 4 min; 30 cycles at 95°C for 1 min/55°C for 1 min/72°C for 1 min; 72°C for 10 min. Taq DNA polymerase (Invitrogen, Karlsruhe, Germany) with its respective buffer, 100 ng forward and reverse primers (Table 1) and a cDNA equivalent to 10 ng RNA was used.
All real-time PCR experiments were carried out on an ABI 7300 SDS system (Applied Biosystems, Darmstadt, Germany). Reactions were performed in duplicates using the RT2 Real-Time™ SYBR Green/ROX PCR Master Mix (Biomol, Hamburg, Germany) containing 300 nm forward and reverse primer (Table 2) and a cDNA equivalent to 10 ng RNA. The PCR program was the following: 50°C for 2 min; 95°C for 10 min; 40 cycles at 95°C for 15 sec/60°C for 60 sec. Threshold cycle (Ct) values were determined using Sequence Detection Software version 1.2 (Applied Biosystems, Darmstadt, Germany). Transcripts were considered present at a Ct ≤ 35. Relative ICAM-1 mRNA expression was calculated as: 2−ΔCt, where ΔCt = CtICAM-1 − CtGAPDH.
Whole cell lysates were prepared using 1× Cell Lysis Buffer from Cell Signaling Technology (Danvers, MA) according to the manufacturer's instructions. For the separation of nuclear and cytosolic proteins, the NucBuster™ Protein Extraction Kit from Novagen (Madison, WI) was used. Equal amounts of protein lysates (50 μg per lane) were separated by 12% SDS-polyacrylamide gel electrophoresis and transferred to Immobilon-P® PVDF membranes (Millipore, Bedford, MA). Subsequently, membranes were blocked in PBS/0.05% Tween 20/5% milk for 1 hr at RT. Rabbit Abs specific for phospho-I-κBα (Ser32), I-κBα, phospho-p65 (Ser536), p65 and p105/p50 (Cell Signaling Technology, Danvers, MA) were added at an 1:1,000 dilution in PBS/0.05% Tween 20/5% BSA for an overnight incubation at 4°C. Equal loading was controlled by reprobing the membranes with a rabbit Ab for the nuclear Poly (ADP-ribose) polymerase (PARP) (Cell Signaling Technology, Danvers, MA) at a final dilution of 1:1,000 and/or a mAb specific for actin (clone C4, MP Biomedicals, Solon, OH) at a final dilution of 1:10,000. Horseradish-peroxidase (HRP)-coupled anti-rabbit antibody at a final dilution of 1:2,000 (Cell Signaling Technology) or HRP-conjugated anti-mouse Ab at a final dilution of 1:5,000 (Dianova, Hamburg, Germany) were applied for 1 hr at RT. Signals were visualized by enhanced chemiluminescence (ECL Plus Western Blotting Detection Reagents, GE Healthcare, Chalfont St. Giles, UK). Western Blot signals were quantified using ImageJ (rsbweb.nih.gov/ij/).
Nuclear and cytosolic extracts were prepared using the NucBuster™ Protein Extraction Kit from Novagen (Madison, WI). Equal amounts of protein (30 μg per well) in nuclear and cytosolic extracts were loaded on a 96-well plate. Phospho-p65 or p65-specific signals were quantified with the PathScan® Phospho-NF-κB p65 (Ser536) Sandwich ELISA Kit or with the PathScan® Total NF-κB p65 Sandwich ELISA Kit, respectively (both Cell Signaling Technology, Danvers, MA) following the manufacturer's instructions.
Transduced keratinocytes and CaSki cells of less than 70% confluence were harvested by trypsinization and washed twice with FACS-buffer consisting of PBS containing 0.02% NaN3 (Applichem, Darmstadt, Germany) and 3% FBS (PAA). 1 × 105 cells were resuspended in 50 μl FACS buffer containing PE-labeled anti-CD54 mAb (clone HA58; BD Pharmingen, San Jose, CA) at a 1:50 dilution in 96-well plates. After washing, 7-AAD (Sigma-Aldrich, Munich, Germany; 2 μg/ml) was added and incubated for 20 min on ice. Flow cytometric analysis was performed using a FACS Calibur® and CellQuest Pro® software (BD Biosiences, San Jose, CA).
A 5 hr standard 51Cr release cytotoxicity assay was carried out as previously described.19 Blockage of LFA-1 was accomplished by adding anti-LFA-1 (clone L130; BD Pharmingen, San Jose, CA) or isotype control mIgG1 (clone 11711; R&D Systems, Wiesbaden, Germany) mAbs at 20 μg/ml final concentration to the effector cells 10 min before coculture with target cells.
Flow cytometry-based conjugation assay
The flow cytometry-based conjugation assay was performed as previously described.20 Briefly, effector cells (NKL) and target cells were harvested and washed once with HBSS containing CaCl2 and MgCl2 (Gibco/BRL, Karlsruhe, Germany) 2 × 106 target cells/ml were resuspended in 40 nM Calcein AM (Molecular Probes/Invitrogen, Karlsruhe, Germany) in HBSS and 1 × 106 effector cells/ml were resuspended in 0.2 μm-filtered 253 μM DHE (Molecular Probes/Invitrogen, Karlsruhe, Germany) in HBSS. Cells were incubated with the dyes for 30 min at 37°C with slow agitation every 10 min. Staining was stopped by washing twice with HBSS. Conjugate formation was achieved by mixing 2 × 105 Calcein AM-labeled target cells with 1 × 105 DHE-labeled effector cells. The cell mixture was slowly vortexed and spun down for 3 min at 300 rpm at 4°C followed by incubation for the indicated time periods at 37°C and 5% CO2. Conjugate formation was stopped by addition of ice-cold 1% paraformaldehyde. Conjugate formation was analyzed using a FACSCalibur™ and CellQuest™ software (BD Biosciences, San Jose, CA). Blockage of LFA-1 was performed as described for the cytotoxicity assay.
For the calculation of statistical significances of differences between experimental groups two-tailed Student's t tests for unpaired data were used. p values of less than 0.05 were considered as significant.
HPV16-E6E7 expression induces rapid upregulation of the adhesion molecule ICAM-1 in primary human keratinocytes
To investigate the impact of the two viral oncoproteins E6 and E7 on the expression of ICAM-1, we transduced primary human keratinocytes, the natural host cells of HPV, with a retroviral vector system containing E6/E7 of HPV type 16 and IRES-EGFP as a marker gene. Transduction efficiencies were ≥ 80%. After transduction, E6 and E7 expression were confirmed by RT-PCR (Fig. 1a). Full-length E7 and E6 transcripts were detectable as well as the two splice variants of E6, E6*I and E6*II (Fig. 1a). Importantly, HPV16-E6E7-expressing EGFP-positive keratinocytes expressed higher levels of surface ICAM-1 as compared to vector control (VC)-transduced keratinocytes 48 h after transduction (Fig. 1B, left and middle panel). ICAM-1 expression was significantly upregulated on protein level as determined by flow cytometry (Fig. 1b, middle panel) and at mRNA level as detected by Real-time quantitative PCR (Fig. 1b, right panel). Expression of other molecules, including the NKG2D ligand ULBP1, was unchanged (data not shown) indicating selective upregulation of ICAM-1 by HPV16-E6E7. Neither HPV16-E6 nor HPV16-E7 by itself was able to significantly upregulate ICAM-1 mRNA or protein in primary human keratinocytes (Fig. 1c). Taken together, our data indicate that E6 and E7 of HPV16 are potent inducers of ICAM-1 expression in primary human keratinocytes and that the two oncoproteins act synergistically.
ICAM-1 expression on target cells enhances adhesion and killing susceptibility to the NK cell line NKL
In a next step, we analyzed the influence of the enhanced ICAM-1 expression in transduced keratinocytes on the ability to form conjugates with the NK cell line, NKL that expresses high levels of LFA-1.21 Figure 2a shows that significantly more conjugates were formed between NKL cells and HPV16-E6E7-expressing keratinocytes as compared to VC-expressing keratinocytes. This enhanced conjugate formation between NKL cells and HPV16-E6E7-expressing keratinocytes was significantly reduced by addition of a blocking mAb directed against LFA-1 (Fig. 2b). In addition, conjugate formation of NKL cells with the HPV16-positive cervical carcinoma cell line, CaSki, which endogenously expresses high levels of ICAM-1 (Fig. 2c), was significantly reduced by LFA-1 blockage (Fig. 2d). Furthermore, killing of CaSki cells by NKL cells was markedly reduced upon addition of a blocking anti-LFA-1 mAb (Fig. 2e). Thus, enhanced ICAM-1 expression by HPV16-E6E7 leads to a better recognition of HPV16-E6E7-expressing cells by LFA-1-expressing NK cells.
The upregulation of ICAM-1 in HPV16-E6E7-expressing keratinocytes correlates with activated components of the NF-κB pathway
A variety of transcription factors, including NF-κB, AP-1, GAS and C/EBP, are implicated in the transcriptional regulation of ICAM-1.14 It was reported that induction of multiple NF-κB- and AP-1-responsive genes in HPV16-E6E7 expressing keratinocytes correlated with enhanced expression of components of the NF-κB signaling pathway.22 Thus, we investigated the involvement of NF-κB signaling in the induction of ICAM-1 expression by HPV16-E6E7. NF-κB activation is typically mediated through phosphorylation of its inhibitor Iκ-B by the Iκ-B kinase complex IKKα/IKKβ/IKKγ followed by the ubiquitination of I-κB and its degradation.23 During this process, NF-κB dimers, in most cells composed of p65 and/or p50, are also phosphorylated by the IKK complex and released from I-κB binding due to I-κB degradation. Phosphorylated NF-κB dimers shuttle into the nucleus and exert their function through binding to NF-κB binding sites upstream of NF-κB-responsive genes. As depicted in Figure 3a, we detected decreased levels of I-κBα and increased phosphorylation of p65 at Ser536 in keratinocytes expressing HPV16-E6E7 as compared to VC-expressing keratinocytes. Phosphorylation of additional phosphorylation sites at Ser276 and Ser468 of p65 was not observed (data not shown). Overall levels of p65 and p50 remained unchanged in whole cell lysates. After separation of the nuclear from the cytosolic cell fraction, more phosphorylated p65 and more total p65 and p50 were detected in nuclei of HPV16-E6E7-expressing keratinocytes than in nuclei of VC-expressing keratinocytes as assessed by Western blotting (Fig. 3b) and specific ELISAs (Fig. 3c). In the cytosolic cell fraction levels of phosphorylated and total p65 and p50 were similar in VC- and HPV16-E6E7-expressing cells (Figs. 3d and 3e). Hence, low levels of Iκ-Bα, phosphorylation of p65 at Ser536 and enhanced levels of p65/p50 in the cell nuclei indicate a constitutively activated NF-κB pathway in HPV16-E6E7-expressing cells.
NF-κB signaling is required for the upregulation of ICAM-1 surface expression in HPV16-E6E7-expressing keratinocytes
Next, we investigated whether NF-κB signaling is involved in the increased ICAM-1 expression in HPV16-E6E7-expressing cells. For this purpose, we employed a NF-κB superrepressor (ΔN-I-κBα), a deletion mutant of I-κBα lacking the phosphorylation sites Ser32 and Ser36. Hence, VC- or HPV16-E6E7-expressing keratinocytes were transduced with a retroviral vector (pBabe) carrying ΔN-I-κBα (Fig. 4a). In HPV16-E6E7-expressing keratinocytes transduced with the superrepressor degradation of I-κBα was greatly impaired (data not shown). Three days after start of antibiotic selection, ICAM-1 expression was significantly reduced in HPV16-E6E7-expressing cells that were transduced with the NF-κB superrepressor. The expression of other cell surface molecules, like ULBP1, was unchanged (Fig. 4b). Moreover, the viability of target cells was not influenced by the NF-κB superrepressor at this time point (data not shown). In summary, these data indicate that the NF-κB pathway in HPV16-E6E7-expressing keratinocytes is, at least, partially responsible for the observed upregulation of ICAM-1 by HPV16-E6E7.
Mucosal and cutaneous HPV types differentially regulate ICAM-1 expression
As ICAM-1 expression was upregulated in HPV16-E6E7-expressing cells, the influence of additional related HPV types on ICAM-1 expression was investigated. Primary human keratinocytes from two different donors were transduced with retroviral vectors expressing HPV18-E6E7 (another mucosal high-risk Alpha PV type), HPV5-E6E7, HPV8-E6E7 and HPV38-E6E7 (cutaneous Beta PV types) and HPV6-E6E7 (a mucosal low-risk Alpha PV type). Expression of the respective HPV-E7 mRNA was confirmed by RT-PCR (Fig. 5a). Figure 5b shows that 5 days after transduction HPV16-E6E7 induced the strongest upregulation of ICAM-1 compared to E6 and E7 of other HPV types. E6 and E7 from HPV18 induced an intermediate upregulation of ICAM-1. In contrast to HPV38-E6E7, which did not upregulate ICAM-1 expression, E6 and E7 from the two other cutaneous Beta PV types, HPV5 and 8, significantly enhanced ICAM-1 expression. E6 and E7 of the low-risk mucosal Alpha PV type, HPV6, did not significantly change levels of ICAM-1. Analysis of components of the NF-κB pathway in whole cell lysates of these HPV types revealed that only in HPV16-E6E7 expressing cells I-κBα expression was diminished and p65 phosphorylated at Ser536 at this time point post-transduction (Fig. 5c). In HPV18-E6E7-expressing cells higher levels of phosphorylated p65 were detectable, whereas I-κBα expression was not altered. In HPV5- and HPV8-E6E7-expressing cells, the upregulated ICAM-1 expression did not correlate with activated components of the NF-κB pathway suggesting that in these HPV types additional signaling pathways are involved.
It is well established that ICAM-1 expression in cells is often induced during viral infections by viral proteins, pro-inflammatory cytokines and interferons.12, 24 Thereby, ICAM-1 promotes interactions between infected cells and immune cells by allowing immune synapse formation.25 Here, we show that in primary human keratinocytes expression of HPV16-E6 and -E7 leads to a pronounced and rapid induction of ICAM-1 at mRNA- and protein levels already 48 hr after transduction (Fig. 1b). A previous study by Huang et al. described induction of ICAM-1 expression in HPV16-E6E7-immortalized human oral epithelial cells compared to normal oral epithelial cells without elucidating the mechanisms underlying the observed ICAM-1 upregulation in HPV16-E6E7-expressing cells.17 In addition, a study by Coleman et al. reported a significant induction of ICAM-1 in high-grade cervical intraepithelial neoplasia (CIN), in which the two HPV oncoproteins E6 and E7 are highly expressed.15 In this study, ICAM-1 expression was only detected in fully transformed tumorigenic cells, such as in HPV16+ CxCa cell lines, but not in HPV16-immortalized non-tumorigenic human keratinocytes. These data led to the assumption that the observed upregulation is not directly linked to E6 and E7 expression, but rather represents a late effect of transformation.15 However, our results demonstrate that ICAM-1 upregulation was already detectable 48 hr after transduction with HPV16-E6E7-expressing vectors suggesting that E6 and E7 expression in primary human keratinocytes directly led to increased ICAM-1 expression. Moreover, transfer of cell culture supernatant from HPV16-E6E7-expressing primary human keratinocytes to VC-expressing primary keratinocytes did not induce ICAM-1 expression (data not shown) indicating that ICAM-1 induction in HPV16-E6E7-expressing cells was not mediated indirectly by soluble factors.
Our data suggest a critical involvement of the NF-κB signaling pathway in the regulation of ICAM-1 in HPV16-E6E7-expressing cells. ICAM-1 upregulation in HPV16-E6E7-expressing cells correlated with decreased I-κBα expression, increased p65 phosphorylation at Ser536 and enhanced levels of p50 and p65 in cell nuclei (Figs. 3b and 3c). Most importantly, inhibition of the NF-κB pathway with a NF-κB superrepressor significantly reduced surface ICAM-1 expression (Fig. 4). Recently, Nees et al. also described activation of components of the NF-κB pathway in HPV16-E6E7-expressing keratinocytes compared to vector control-expressing cells.22 Furthermore, constitutive NF-κB activation in high-grade squamous intraepithelial lesions (SIL) and CxCa was detected in situ.26 In agreement with our study, these reports demonstrated that mainly p65/p50 NF-κB heterodimers translocated to cell nuclei. Of note, other studies reported an inhibition of NF-κB activity in HPV16-E6 or -E7-expressing cells.27–29 Harvard et al. described reduced activity of NF-κB associated with modulation of expression and subcellular localization of the NF-κB1/2 precursors p100 and p105 in long-term cultured HPV16-E6 or -E7 expressing keratinocytes compared to normal keratinocytes of different origin.27, 28 Furthermore, an inhibition of IKKα activity and TNF-α induced translocation of p65 by HPV16-E7 was reported.29 It is likely that the conflicting results regarding the modulation of NF-κB activity by HPV16-E6/E7 are due to different cell types and cell culture conditions used in these studies.
Not only HPV16-E6E7, but also E6 and E7 proteins from other HPV types led to an upregulation of ICAM-1 in primary human keratinocytes (Fig. 5). E6/E7 from another high-risk mucosal HPV type, HPV18, and E6/E7 from the two cutaneous HPV types, HPV5 and HPV8, led to a significant increase in ICAM-1 expression (Fig. 5b). Noteworthy, E6/E7 from a low-risk mucosal HPV type, HPV6, and E6/E7 from another cutaneous HPV type, HPV38, did not significantly increase ICAM-1 expression (Fig. 5b). Although HPV38 belongs to the group of Beta papillomaviruses, like HPV5 and HPV8, it is located distantly from these types in the phylogenetic tree1 and, in contrast to HPV5 and -8, it is not regarded as a “high-risk” cutaneous HPV type.30 Thus, ICAM-1 upregulation in E6/E7-expressing cells appears to be associated with the oncogenicity of the respective HPV type. In this context, we observed that ICAM-1 expression in HPVE6E7-expressing keratinocytes increases with the number of cell doublings. Analysis at later time points after transduction (≥ 9 days) revealed that ICAM-1 upregulation in HPV5- and HPV8-E6E7 expressing cells even exceeded ICAM-1 expression in HPV16-E6E7-expressing cells (data not shown). Furthermore, it was reported that senescent cells upregulate ICAM-1 (31 and our unpublished observation). Keratinocytes transduced with the cutaneous HPV-E6E7-expressing vectors were not or less efficiently immortalized than HPV16-E6E7-expressing keratinocytes and usually became senescent soon after transduction. A recent study detected a most pronounced upregulation of ICAM-1 by HPV5-E6E7-expressing cells and to a lower extent by HPV38-E6E7- and HPV16-E6E7-expressing cells.32 In this study, the number of cell doublings or time after transduction of keratinocytes was not specified. Thus, it is possible that the experimental conditions and time of analysis after transduction were responsible for the different observations. Our results indicate that ICAM-1 upregulation in HPV5- and HPV8-E6E7-expressing cells did not correlate with decreased I-κBα expression and enhanced phosphorylation of p65 at Ser536 (Fig. 5c). These data suggest that additional mechanisms of ICAM-1 upregulation in cells expressing HPV-E6/E7 proteins exist.
The increased ICAM-1 expression in HPV16-E6E7-expressing cells led to enhanced conjugate formation with the NK cell line, NKL (Figs. 2a, and 2b). In addition, NKL cell-mediated killing of the HPV16+ CxCa cell line CaSki was significantly reduced by addition of a blocking LFA-1 mAb (Fig. 2e). In the study of Huang et al., increased ICAM-1 expression in HPV16-E6E7-immortalized human oral epithelial cells also correlated with enhanced adhesion to PBMCs and enhanced lymphokine-activated-killer cell cytotoxicity.17 Thus, ICAM-1 expression on tumor cells might contribute to lymphocyte retention and activation in cervical carcinoma. In this context, it was shown that ICAM-1 expression in breast cancer correlated with a better prognosis and survival rates of patients.33 In contrast, ICAM-1 expression in certain other cancer entities was associated with cancer progression. In melanoma and liver cancer patients, high ICAM-1 expression on tumor cells corresponded with an increased risk for metastasis.34–36 So far, there are no reports describing the impact of ICAM-1 expression on prognosis of cervical cancer or its precursors. Thus, it is unclear whether induction of ICAM-1 expression by high-risk HPV types is beneficial for tumor immunosurveillance or supports tumor progression.
Our previous study revealed that activating NK cell receptor ligands, such as CD155 and MICA, are often up-regulated and inhibitory NK cell receptor ligands, like MHC class I, are frequently down-modulated in CxCa in situ.19 In addition, here we demonstrate that ICAM-1 that is critical for conjugate formation of effector to target cells is induced by HPV16-E6E7 and high ICAM-1 expression was reported on CxCa in situ.16 These data suggest that CxCa should be well recognized by NK cells. However, in cancer patients NK cells are frequently inactive due to multiple suppressive forces exerted by other immune cells or the tumor microenvironment. Strategies that lead to the efficient activation of NK cells might be promising for successful treatment of CxCa.
We would like to thank Dr. Massimo Tommasino for providing valuable reagents and advice. We thank Dr. Carsten Watzl for helping to establish the FACS-based conjugate formation assay and Dr. Angelika Bierhaus for helpful discussions concerning NF-κB signaling.