| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Virology |
Department of Medical Microbiology, Center of Infectious Diseases [M. C. W. F., R. B., F. M. d. S., L. S., E. v. d. M., J. t. S., W. J. M. S.] and Departments of Internal Medicine [R. G. J. W.] and Dermatology [J. N. B. B.], Leiden University Medical Center, 2300 RC Leiden, the Netherlands, and Department of Immunobiology, Netherlands Institute of Public Health, 3720 BA Bilthoven, the Netherlands [B. P. J. V.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recent studies indicated that EV-HPV types are not restricted to EV patients. SCCs and BCCs in the general population, together referred to as nonmelanoma skin cancer, also frequently contain EV-HPV DNA (5
, 6)
. In renal transplant recipients, who often develop nonmelanoma skin cancer in the years after transplantation, the percentage EV-HPV-positive SCCs increases up to 90% (7, 8, 9)
. EV-HPV is also frequently detected in actinic keratoses (
65%), the presumed SCC precursor (10
, 11)
. Increasing age, male sex, fair skin, sun exposure, and smoking are known risk factors for nonmelanoma skin cancer (12, 13, 14)
. EV-HPV infection may act as an additional risk factor, especially because EV-HPV interacts with UV-induced apoptosis (15)
.
In most HPV DNA prevalence studies in nonmelanoma skin cancer, no specific EV-HPV type stood out. Usually, a wide spectrum of EV-HPV types is detected, including unassigned new EV-HPV types (5 , 11 , 16, 17, 18) , and multiple EV-HPV types are frequently found in single biopsies (7 , 11) . The complexity of this situation, which hampers the search for oncogenic EV-HPV types, may be explained by the use of different PCR primer sets and the ubiquity of these viruses (18) . Alternative strategies to investigate the relation between EV-HPV infection and nonmelanoma skin cancer, such as serology, are therefore needed.
In individuals with a history of different types of skin cancer, we looked for serorecognition of six EV-HPV types (HPV5, 8, 15, 20, 24, and 38) frequently detected in nonmelanoma skin cancer biopsies and plucked hairs (7 , 11 , 16 , 19) and a common genital type associated with cervical cancer (HPV16; Ref. 1 ). By comparing EV-HPV seroreactivity and seropositivity among skin cancer cases and controls and by estimating skin cancer relative risks, we looked for associations between EV-HPV infection and skin cancer development.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
V, Fitzpatrick classification; Ref. 20
) were also excluded because they very rarely develop skin cancers. Ninety-three individuals (40 cases and 53 controls) were excluded from the current study because of lack of antigen to repeat the ELISA for sera that had unacceptable duplicate values (see below) or for sera that were located on plates with unacceptable positive control serum levels (see below). For the current study, 873 subjects were included: 333 controls without skin cancer; 456 cases with nonmelanoma skin cancer (160 SCC, 291 nBCC, and 141 sBCC); and 84 melanoma cases. Of the nonmelanoma skin cancer cases included, 134 (29.4%) had more than one type of skin cancer, including melanoma. Every nonmelanoma skin cancer case (SCC, nBCC, or sBCC) was analyzed accordingly, regardless of the presence of other skin cancer types, unless stated otherwise. The melanoma cases selected had no other type of skin cancer.
Data and Sample Collection.
The visit at the dermatology outpatient clinic was identical for cases and controls and consisted of a standardized interview and a physical examination. During physical examination, a dermatologist recorded skin type and the possible presence of skin cancer. Occasionally, controls appeared to have skin cancer, which was subsequently confirmed by histological examination. These persons were further considered as cases. In the same visit, blood was drawn to collect serum.
Generation of the HPV-L1 Recombinant Baculoviruses.
Recombinant baculovirus AcHPV5aL1 (21)
was a gift from Drs. G. Orth and M. Favre (Pasteur Institute, Paris, France). AcHPV8L1 (22)
and AcHPV15L1 (23)
were donated by Drs. H. Pfister and P. Fuchs (Virology Institute, University of Cologne, Cologne, Germany). AcHPV16L1 was provided by Dr. J. T. Schiller (NIH, Bethesda, MD; Ref. 24
). HPV20-L1, HPV24-L1, and HPV38-L1 expressing recombinant baculoviruses were made in the Bac-to-Bac expression system (Life Technologies, Inc.) according to the manufacturer’s instructions, with slight modifications. The pFastBacDual-GFP expression vector was used (25)
, which contains the GFP gene cloned in multiple cloning site II. As a result of this insertion, an additional HindIII site was introduced. This HindIII site was subsequently removed by the ligation of a mutated HindIII*-KpnI linker in the vector digested with HindIII and KpnI. The resulting vector is called pFBD-GFP*.
The L1 genes of HPV20, 24, and 38, respectively, were inserted into multiple cloning site I of pFBD-GFP* by PCR-assisted cloning from pBR322-HPV20 (26) , pBR322-HPV24A (27) , and pBR322-HPV38 (27) , respectively. The primers used are as follows (their restriction site extensions are in italic): 5'-HPV20, nt 5889–5905 (CGAAGGCCTCCATCGATCCGCAGATGGCAGTTTGGC, StuI, ClaI); 3'-HPV20, nt 7462–7445 (GGCTCGAGCCACTAGTCCGTACCGAAAACGGTCGGG, XhoI, SpeI); 5'-HPV24, nt 5713–5739 (GCGCGGATCCAATATGTCGGTGTGGTTGCCAGCCAGTGGT, BamHI); 3'-HPV24, nt 7251–7231 (CGACAAGCTTATTTTACGTGCGTTTTCGTTTGGT, HindIII); 5'-HPV38, nt 5660–5677 (CGTCTCGAGCCGATGACACTTTGGCTTCC, XhoI); and 3'-HPV38, nt 7209–7193 (GCAAGCTTCCATCGATCCGGGCGACCGAAAACGGT, HindIII, ClaI).
The HPV20-L1 PCR product was first ligated into the StuI- and XhoI-digested pFastBac1 and subsequently cloned into pFBD-GFP* as a StuI-SpeI fragment. The HPV24-L1 PCR product was cloned into pFBD-GFP* as a BamHI-HindIII fragment. HPV38 was first isolated from pBR322-HPV38 by digestion with EcoRI and ligated as a circular fragment. Subsequently, L1 was PCR-amplified from the recirculized HPV38 viral genome and cloned into pFastBac1 after XhoI and HindIII digestion. Subsequently, it was cloned into pFBD-GFP* as a SalI-HindIII fragment.
pFBD20L1-GFP*, pFBD24L1-GFP*, and pFBD38L1-GFP*, respectively, were used to transform DH10Bac cells. Total DNA was harvested from gentamicin-resistant colonies and used to transfect Sf9 cells that were screened for fluorescence 4 days later. Supernatants from GFP-positive cultures were passaged to new Sf9 cells and checked for the presence of the HPV-L1 gene by PCR. GFP- and L1-positive passage 1 supernatants were plaque-purified and amplified to viral stocks of AcHPV20L1-GFP*, AcHPV24L1-GFP*, and AcHPV38L1-GFP*.
Generation of HPV VLPs.
To generate the HPV5, 8, 15, 16, 20, 24, and 38 VLPs, Sf9 cells were infected with AcHPV5aL1, AcHPV8L1, AcHPV15L1, AcHPV16L1, AcHPV20L1-GFP*, AcHPV24L1-GFP*, and AcHPV38L1-GFP*, respectively, at a multiplicity of infection of 5–10. Cells were harvested 3 days after infection, washed with PBS, and disrupted by sonification at 0°C during 2 x 45 s at 18 µm. Protease inhibitors (100 µg/ml pepstatin, 200 µg/ml aprotinin, 50 µg/ml leupeptin, and 10 mM benzamide) were added, and cell debris was spun down. The supernatant was purified on 40% and 70% sucrose cushions at 100,000 x g for 2 h at 4°C. The sucrose interface was resuspended in 27% CsCl in PBS (w/w) and centrifuged to equilibrium (200,000 x g, 24 h, 4°C). The VLP-containing band with a density of about 1.3 g/ml was isolated from the gradient with a syringe, followed by overnight dialysis against PBS. VLP batches were checked by SDS-PAGE (Fig. 1A)
and electron microscopy to confirm typical VLP morphology (Fig. 1B)
.
|
At the day of use, plates were thawed, washed three times with PBS, and blocked with a fresh solution of C-PBS for 1 h at 37°C. For 1 h at 37°C, 50 µl serum/well was added, diluted 1:100 in C-PBS. After washing three times in PBS, biotin-labeled antihuman total IgG (1 mg/ml), kindly donated by Dr. L. Aarden (CLB Sanquin, Amsterdam, the Netherlands), was added for 1 h at 37°C diluted 1:2000 in C-PBS, followed by three washes, and then streptavidin-poly-horseradish peroxidase (1 mg/ml, CLB Sanquin) diluted 1:10,000 in 1% casein in PBS was added for 30 min at room temperature. Finally, 50 µl of 0.1 mg/ml tetramethylbenzidine substrate in a 0.03% H2O2 0.11 M sodium acetate solution (pH 5.5) were added. The reaction was stopped after 5 min by adding 50 µl of 2 M H2SO4. The absorbance of the wells in each plate was measured at 450 nm in an automated plate reader.
Calculation of ELISA Values and ELISA Cutoff Values.
Sera for which the difference between duplicate wells exceeded 50% of the duplicate mean were retested. To correct for antigen-specific background, for each antigen, the blank value was subtracted from the test values. To correct for serum-specific background, for each serum, the uncoated value was subtracted from the test values. Subsequently, for each antigen, we calculated the mean of the positive control serum values of a total of 180 test plates. Serum samples from plates with positive control serum values that exceeded this mean value ± 2x SD were retested. All serum values were corrected according to the mean positive control serum values that we calculated for each antigen.
The ELISA cutoff values to determine the value above which a serum was called seropositive were based on a group of 100 sera from randomly selected anonymous LUMC hospital employees tested at appointment for serum antibodies against hepatitis B virus surface antigen. These sera were tested in exact the same manner as the test sera. For each antigen, we calculated the mean of the corrected absorbance values of this population + 3x SD, excluding outliers. The calculated cutoff values for HPV5, 8, 15, 16, 20, 24, and 38 were 0.620, 0.366, 0.405, 0.456, 0.348, 0.432, and 0.336, respectively. The percentages of seropositivity in this population, 0.0%, 0.0%, 1.0%, 0.0%, 2.9%, 3.7%, and 0.9%, respectively, were comparable with those measured in the study control population (Table 2)
.
|
2 test. Relative risks were estimated using exposure OR from cross tabulation. Multivariate logistic regression analysis was used to adjust for possible confounding factors. |
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
HPV Seroreactivity and Nonmelanoma Skin Cancer.
To investigate the association between EV-HPV infection and nonmelanoma skin cancer, we determined the reactivity in 873 sera against 6 EV-HPV types (HPV5, 8, 15, 20, 24, and 38) and a common genital type (HPV16). The absorbance values measured against each of the antigens are presented separately for cases and controls in Fig. 2
. For five of the six EV types tested (HPV5, 8, 15, 20, and 24), we observed a statistically significant increase in seroreactivity in the SCC cases compared with controls. HPV20 and 24 seroreactivity was also increased in the sBCC and nBCC cases, respectively, and HPV5 seroreactivity was increased in the melanoma cases. HPV16 seroreactivity was not increased among the skin cancer cases.
|
). Neither cumulative sun exposure nor smoking was associated with HPV seropositivity (data not shown).
|
). HPV20-seropositive individuals had a 3-fold increased risk to have had a nBCC or sBCC, respectively. In the calculation adjusted for age and sex, HPV38-seropositive individuals also had a 3-fold increased chance to have had SCC (Table 2)
.
To exclude confounding by the occurrence of different types of skin cancer within one individual, we repeated the relative risk calculations after exclusion of nonmelanoma skin cancer cases with more than one type of skin cancer. In addition, we confined these calculations to the seropositive persons to exclude a possible bias introduced by unknown differences in social behavior or immune responsiveness between HPV-seropositive and -negative individuals. Roughly, the same associations were observed (Table 3)
compared with those calculated in the complete dataset, although statistical significance was sometimes lost. This was mainly caused by the limited power of the smaller subgroups.
|
and Tables 2
and 3
suggest that HPV type-specific serorecognition is often associated with more than one type of nonmelanoma skin cancer. We questioned the skin cancer type specificity of EV-HPV serorecognition by recalculating the relative risks of SCC, nBCC, and sBCC exclusively among the skin cancer cases. In these calculations, the melanoma cases were considered as the reference population. Most calculated ORs decreased, and statistical significance was lost in all cases (Table 4)
, partly due to considerable broadening of the CIs.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Serorecognition of HPV8 in particular posed a significant risk to have had nonmelanoma skin cancer. Because of the high ORs observed, it is tempting to believe that HPV8 is the cutaneous oncogenic EV-HPV type. However, the broad CI observed, due to the low HPV8 seroprevalence in our population, does not allow us to draw any definite conclusions yet regarding the oncogenic potential of this EV-HPV type. Because HPV8, 20, and 38 do not belong to the same EV-HPV subgroup (11) , we cannot relate nonmelanoma skin cancer to a specific EV-HPV subgroup.
The possible relationship between EV-HPV infection and skin cancer was also the subject of earlier, smaller serological studies. Favre et al. (28) could not detect any seropositivity in BCC and SCC patients against the HPV5a VLPs. This is in agreement with our finding that HPV5 seropositivity was extremely low in our study populations. Stark et al. (22) used similar HPV8 VLPs and also reported increased seroprevalence of HPV8 antibodies in patients with nonmelanoma skin cancer. The differences in HPV8 seroprevalence measured in our study compared with that measured by Stark et al. (22) are probably based on differences in ELISA methodology. We reported previously (29) that HPV8 VLP seropositivity was significantly associated with development of multiple actinic keratoses on the tropical island of Saba. In this population, SCC development was also associated with HPV8 seropositivity, although it did not reach significance, perhaps because of the low number of SCC cases. HPV15 VLP seropositivity measured by Wieland et al. (23) was comparable with ours and also showed no significant differences among BCC cases and controls.
As a measure of EV-HPV infection, we chose detection of serum antibodies. These are expected against HPVs that have infected the host, provided that this infection leads to immune response induction and not to immune neglect or tolerance. In the case of natural cutaneous EV-HPV infection, which may be persistent (11) , very little is known about the immune response induced. In particular, the onset and the longevity of these responses are unknown. In view of the relatively low number of seropositives, the ubiquity of the virus (18) , and the associations found between serorecognition and nonmelanoma skin cancer, the majority of EV-HPV-directed antibodies are probably generated later in infection, probably when tumor formation occurs. This belief is supported by our very recent findings that EV-HPV prevalence measured as the presence of type-specific EV-HPV DNA in plucked eyebrow hairs in the same population exceeds the EV-HPV seroprevalence about five times.4 Induction of EV-HPV-specific antibodies along with tumor formation may be a reflection of increased viral load or local inflammation. Because, in our population, the period between skin cancer diagnosis and serum collection was not associated with seropositivity (data not shown), we have no indication that the measured IgG responses are short-lived.
Throughout the study, particularly for HPV8, EV-HPV type-specific serorecognition was often associated with more than one skin cancer type. This prompted us to look into a possible association with skin cancer as such. In the calculations shown in Table 4
, where the melanoma cases served as a control for the individual SCC, nBCC, and sBCC cases, all associations found for the individual nonmelanoma skin cancer types were lost. This suggested to us that EV-HPV serorecognition is rather associated with skin cancer as such and not so much with a particular skin cancer type. If seroresponses to EV-HPV are induced upon skin cancer formation, they may be considered as a skin cancer epiphenomenon. As a consequence, HPV8 and 20, in particular, are expected to be relatively abundant in these cancers. Whether these types are involved in carcinogenesis remains to be solved. The concurrent appearance of EV-HPV seroresponses is probably not unique to skin cancer, because HPV5 antibodies were described as accompanying the epidermal repair process after extensive burns (28)
.
To monitor whether EV-HPV infection precedes nonmelanoma skin cancer development, the EV-HPV serological approach that we used is probably not the most appropriate. To this purpose, accurate studies that monitor EV-HPV type-specific infection in and outside skin tumors, e.g., by detection of EV-HPV type-specific DNA and RNA, remain indispensable. Ideally, these studies should be performed in a follow-up setting, because only then we will know if EV-HPV infection precedes nonmelanoma skin cancer development and in potential might be causally involved.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by European Community Grant QLK2-CT-2002-01179 and Dutch Cancer Foundation Grant LUMC 99-1913. 
2 To whom requests for reprints should be addressed, at Department of Medical Microbiology, Leiden University Medical Center E4-P, P. O. Box 9600, 2300 RC, Leiden, the Netherlands.
3 The abbreviations used are: HPV, human papillomavirus; EV, epidermodysplasia verruciformis; EV-HPV, epidermodysplasia verruciformis-related HPV; OR, odds ratio; CI, confidence interval; SCC, squamous cell carcinoma; BCC, basal cell carcinoma; nBCC, nodular BCC; sBCC, superficial multifocal BCC; VLP, virus-like particle; C-PBS, 0.2% casein in PBS; LUMC, Leiden University Medical Center; GFP, green fluorescent protein; nt, nucleotide(s).
4 L. Struijk, J. N. Bouwes Bavinck, P. Wanningen, E. van der Meijden, R. G. J. Westendorp, J. ter Schegget, and M. C. W. Feltkamp. Presence of human papillomavirus DNA in plucked eyebrow hairs is associated with a history of cutaneous squamous cell carcinoma, manuscript submitted for publication.
Received 10/ 8/02. Accepted 3/13/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Z. Rizk, N. D. Christensen, K. M. Michael, M. Muller, P. Sehr, T. Waterboer, and M. Pawlita Reactivity pattern of 92 monoclonal antibodies with 15 human papillomavirus types J. Gen. Virol., January 1, 2008; 89(1): 117 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. McBride, R. Neale, N. Pandeya, and A. Green Sun-Related Factors, Betapapillomavirus, and Actinic Keratoses: A Prospective Study Arch Dermatol, July 1, 2007; 143(7): 862 - 868. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. N. C. de Koning, L. Struijk, J. N. B. Bavinck, B. Kleter, J. ter Schegget, W. G. V. Quint, and M. C. W. Feltkamp Betapapillomaviruses frequently persist in the skin of healthy individuals J. Gen. Virol., May 1, 2007; 88(5): 1489 - 1495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hall, L. Struijk, R. E. Neale, and M. C. W. Feltkamp Re: Human papillomavirus infection and incidence of squamous cell and basal cell carcinomas of the skin. J Natl Cancer Inst, October 4, 2006; 98(19): 1425 - 1426. [Full Text] [PDF]
|
|
|
|
|
|
M. R. Karagas, H. H. Nelson, P. Sehr, T. Waterboer, T. A. Stukel, A. Andrew, A. C. Green, J. N. Bouwes Bavinck, A. Perry, S. Spencer, et al. Human papillomavirus infection and incidence of squamous cell and basal cell carcinomas of the skin. J Natl Cancer Inst, March 15, 2006; 98(6): 389 - 395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Struijk, L. Hall, E. van der Meijden, P. Wanningen, J. N. B. Bavinck, R. Neale, A. C. Green, J. ter Schegget, and M. C.W. Feltkamp Markers of cutaneous human papillomavirus infection in individuals with tumor-free skin, actinic keratoses, and squamous cell carcinoma. Cancer Epidemiol. Biomarkers Prev., March 1, 2006; 15(3): 529 - 535. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Smola-Hess, J. Pahne, C. Mauch, P. Zigrino, H. Smola, and H. J. Pfister Expression of membrane type 1 matrix metalloproteinase in papillomavirus-positive cells: role of the human papillomavirus (HPV) 16 and HPV8 E7 gene products J. Gen. Virol., May 1, 2005; 86(5): 1291 - 1296. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. D. Schaper, G. P. Marcuzzi, S. J. Weissenborn, H. U. Kasper, V. Dries, N. Smyth, P. Fuchs, and H. Pfister Development of Skin Tumors in Mice Transgenic for Early Genes of Human Papillomavirus Type 8 Cancer Res., February 15, 2005; 65(4): 1394 - 1400. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Oldak, H. Smola, M. Aumailley, F. Rivero, H. Pfister, and S. Smola-Hess The Human Papillomavirus Type 8 E2 Protein Suppresses {beta}4-Integrin Expression in Primary Human Keratinocytes J. Virol., October 1, 2004; 78(19): 10738 - 10746. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. N. Bouwes Bavinck and M. C.W. Feltkamp Milk of Human Kindness? -- HAMLET, Human Papillomavirus, and Warts N. Engl. J. Med., June 24, 2004; 350(26): 2639 - 2642. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Cell Growth & Differentiation |
