Tobacco Documents Online

k I Frequency and Distribution of Papillomavirus Structural Antigens in Verrucae, Multiple Papillomas, and Condylomata of the Oral Cavity A. BENNETT JENSON, MD, WAYNE D. LANCASTER, PhD, DAN-PAUL HARTMANN, PhD, and EDWARD L. SHAFFER, Jr., DDS Sixty-seven proliferous squamous epithelial lesions of the oral cavity were examined for the presence of hu- man papillomavirus whole (structural) virion antigens by a peroxidase-antiperoxidase technique having im- munospecificity against genus-specific (common) an- tigens of the papillomaviruses. A positive reaction for papillomavirus genus specific antigens was found in 18 of 29 verrucae, 2 of 5 multiple papillomas, and 3 of 5 condylomata; common antigens were not detected in FOR YEARS, circumstantial evidence supported the idea that most cutaneous wartst' and some mucosal papillomas1J were caused by a single type of human papillomavirus (HPV). The human wart virus has never been well-characterized by the standard methods of virology, and even today its oncogenic potential remains largely unknown for several reasons: HPV cannot be grown and tested in tissue culture,• and, like other papillomaviruses, it is highly species-specifics.' (transmission studies cannot be car- ried out in animal models). Recent advances in molecular virology, however, circumvented some of the difficulties of working with the papillomaviruses, resulting in a number of reports'-9 indicating a re- markable plurality for HPV. The different types of HPV have little or no polynucleotide sequence ho- mology, and there is no cross-reaction between virion surface antigens.' Six.types of HPV (HPV-1 through HPV-5 and HPV-7) have been identified in a variety of proliferative lesions of the skin, and at least one type (HPV-6) was found in condylomata of the female genital tract. In fact, the type of papilloma- virus appears to determine, in part, the clinical and pathologic appearance, location, and natural fate of From the Departments of Pathology and Obstetrics and Gynecology, Georgetown University Medical Center, Washington, DC, and United States Army. Ft. Riley, Kansas 28 keratoacanthomas. The positive reaction was invari- ably intranuclear in cells having a focal or diffuse distribution in the superficial epithelium. This study shows that a variety of squamous epithelial lesions of the mucosa are associated with human papillomavi- ruses and suggests that these viruses may play an im- portant role in the etiology of some cases of squamous hyperplasia of the oral cavity. (Am J Pathol 1982, 107:212-218) cutaneous warts and perhaps some mucosal papil- lomas (Table 1). Many different clinical types of papillomas (focal epithelial hyperplasia, verruca vulgaris and plana, multiple and single papillomas, condyloma accumi- natum, laryngeal papillomas, and other papilloma- tous lesions) occur in the oral cavity. Although papo- vavirus-like particles have occasionally been ob- served by electron-microscopic examination in most of these lesions'-'-10-" papiltomavirus has only specifically been identified by immunologic and molecular virologic techniques in laryngeal papil- lomas.t'-t' In this study we examined selected papillomatous lesions of the oral cavity with the same immunologic techniques used to identify papillo- mavirus genus-specific (common) antigens in laryn- Supported by Grants CA24505-04 and CA28507-01 from the Public Health Service and a special project from the Council for Tobacco Research. 4ccepted for Publication December 29, 1981. Address reprint requests to Alfred B. Jenson, MD, Department of Pathology, Georgetown University, Schools of Medicine and Dentistry, Washington, DC 20007. 0002-9440/82/0510-0212$00.85 ® American Association of Pathologists 212 TI BU 33306

k I Frequency and Distribution of Papillomavirus Structural Antigens in Verrucae, Multiple Papillomas, and Condylomata of the Oral Cavity A. BENNETT JENSON, MD, WAYNE D. LANCASTER, PhD, DAN-PAUL HARTMANN, PhD, and EDWARD L. SHAFFER, Jr., DDS Sixty-seven proliferous squamous epithelial lesions of the oral cavity were examined for the presence of hu- man papillomavirus whole (structural) virion antigens by a peroxidase-antiperoxidase technique having im- munospecificity against genus-specific (common) an- tigens of the papillomaviruses. A positive reaction for papillomavirus genus specific antigens was found in 18 of 29 verrucae, 2 of 5 multiple papillomas, and 3 of 5 condylomata; common antigens were not detected in FOR YEARS, circumstantial evidence supported the idea that most cutaneous wartst' and some mucosal papillomas1J were caused by a single type of human papillomavirus (HPV). The human wart virus has never been well-characterized by the standard methods of virology, and even today its oncogenic potential remains largely unknown for several reasons: HPV cannot be grown and tested in tissue culture,• and, like other papillomaviruses, it is highly species-specifics.' (transmission studies cannot be car- ried out in animal models). Recent advances in molecular virology, however, circumvented some of the difficulties of working with the papillomaviruses, resulting in a number of reports'-9 indicating a re- markable plurality for HPV. The different types of HPV have little or no polynucleotide sequence ho- mology, and there is no cross-reaction between virion surface antigens.' Six.types of HPV (HPV-1 through HPV-5 and HPV-7) have been identified in a variety of proliferative lesions of the skin, and at least one type (HPV-6) was found in condylomata of the female genital tract. In fact, the type of papilloma- virus appears to determine, in part, the clinical and pathologic appearance, location, and natural fate of From the Departments of Pathology and Obstetrics and Gynecology, Georgetown University Medical Center, Washington, DC, and United States Army. Ft. Riley, Kansas 28 keratoacanthomas. The positive reaction was invari- ably intranuclear in cells having a focal or diffuse distribution in the superficial epithelium. This study shows that a variety of squamous epithelial lesions of the mucosa are associated with human papillomavi- ruses and suggests that these viruses may play an im- portant role in the etiology of some cases of squamous hyperplasia of the oral cavity. (Am J Pathol 1982, 107:212-218) cutaneous warts and perhaps some mucosal papil- lomas (Table 1). Many different clinical types of papillomas (focal epithelial hyperplasia, verruca vulgaris and plana, multiple and single papillomas, condyloma accumi- natum, laryngeal papillomas, and other papilloma- tous lesions) occur in the oral cavity. Although papo- vavirus-like particles have occasionally been ob- served by electron-microscopic examination in most of these lesions'-'-10-" papiltomavirus has only specifically been identified by immunologic and molecular virologic techniques in laryngeal papil- lomas.t'-t' In this study we examined selected papillomatous lesions of the oral cavity with the same immunologic techniques used to identify papillo- mavirus genus-specific (common) antigens in laryn- Supported by Grants CA24505-04 and CA28507-01 from the Public Health Service and a special project from the Council for Tobacco Research. 4ccepted for Publication December 29, 1981. Address reprint requests to Alfred B. Jenson, MD, Department of Pathology, Georgetown University, Schools of Medicine and Dentistry, Washington, DC 20007. 0002-9440/82/0510-0212$00.85 ® American Association of Pathologists 212 TI BU 33306

Vol. 107 • No. 2 Table 1-Types of Human Papillomavirus° Virus Clinical association HPV-1 Plantar wart HPV-2 Common wart HPV-3 Flat warts HPV•4 Plantar and common warts HPV-5 Pityriasis•like lesions in epidermodysptasta verruciformis (EV) that may undergo malignant transformation HPV-6 Anogenital warts HPV-7 Common warts (butchers) Adapted from Howley.° geal papillomas"`° and genital tract condylomataZ° and now report on the prevalence of HPV in ver- rucae, condylomata, and multiple papillomas. Materials and Methods Papillomas Squamous papillomas of the oral cavity selected for this study were accessioned in the files of the Armed Forces Institute of Pathology (AFIP), Washington, DC. Specifically, we wanted to examine verrucae, multiple papillomas, and condylomata for the presence of papillomavirus antigens, since papil- lomavirus-like particles have been seen in these le- sions by electron microscopy; keratoacanthomas were examined because of their usual history of "wart-like" spontaneous regression. Verrucae and papillomas were retrieved from the files by location (oral cavity) and title (papillomas, squamous papillo- mas, verruca vulgaris, condyloma accuminatum, keratoacanthoma, florid papillomatosis, papilloma- tosis, verrucous hyperplasia). After hematoxylin and eosin (H&E) sections from each papillomatous lesion were reviewed, at least 5 consecutive 4-µ-thick sections from paraffin blocks containing formalin-fixed tissue were air-dried on glass and stored at room temperature until ready for use. Verrucae Twenty-nine patients (25 male, 4 female) with oral cavity verrucae were included in this study. These verrucae had the following distribution: upper lip, 7; lower lip, 12; tongue, 4; alveolar mucosa, 2; hard palate, 2; buccal mucosa, 1; and floor of mouth, 1. The youngest patient was 8 and the oldest 61, but the majority were 20-30 years of age. PAPII.LOMAVIRUS 213 Condylomata Five patients (all male) of condylomata were ac- cepted for this study; 4 were on the tongue, and 1 was on the soft palate. Their ages were 23, 27, 31, 38, and 63. One patient had concurrent condyloma of the anal mucosa. Multiple Papillomas Five patients (4 male, 1 female) with multiple pap- illomas were included in this study. These lesions had the following distribution: 1) maxillary gingivae; 2) right and left commissures; 3) floor of mouth and tongue; 4) frenulum and mandibular gingivae; and 5) soft palate and left lateral tongue. The 4 males were 18, 24, 25, and 28 years of age; the female was 15. None of these patients stated in the record that they had other verrucous lesions. Keratoacanthomas Twenty-eight cases (27 male, I female) with kera- toacanthomas were accepted for this study: 27 were located on the upper lip, and I was located on the lower lip. The youngest patient was 19, and the oldest was 74; the remaining patients were evenly dis- tributed between these age limits. There was no men- tion of concomitant verruca at the time of removal of the keratoacanthomas. Peroxidase-Antiperoxidase (PAP) Staining of Papillomas Antiserums Purified bovine papillomavirus type I(BPV-1) virions, previously characterized,2t were extracted from a single naturally occurring bovine fibropapil- loma. BPV-1 protein was prepared as previously described,"-10•22 adjusted to 0.24 M 2-mercap- toethanol l% SDS, heated to 68 C for 2 minutes, diluted with 3 volumes of saline, and mixed to an equal volume of Freund's complete adjuvant. A vary- ing concentration of virus protein (480, 240, and 120 µg) was inoculated subcutaneously into a rabbit on Days 0, 15, and 28, respectively; and the animal bled on Day 38. This hyperimmune serum (BPV-1 [SDS]) is reactive by PAP at a 1:100 dilution with papilloma- virus positive (by electron microscopy) cutaneous warts and mucosal papillomas from all species tested (human, cattle, dogs, rabbits, deer and horses) but does not react with virus-negative papillomas (data not shown); it has the same specificity but a higher titer for the papillomavirus genus-specific antigen as TI BU 33307 k 0 I . r--- I

Vol. 107 • No. 2 Table 1-Types of Human Papillomavirus° Virus Clinical association HPV-1 Plantar wart HPV-2 Common wart HPV-3 Flat warts HPV•4 Plantar and common warts HPV-5 Pityriasis•like lesions in epidermodysptasta verruciformis (EV) that may undergo malignant transformation HPV-6 Anogenital warts HPV-7 Common warts (butchers) Adapted from Howley.° geal papillomas"`° and genital tract condylomataZ° and now report on the prevalence of HPV in ver- rucae, condylomata, and multiple papillomas. Materials and Methods Papillomas Squamous papillomas of the oral cavity selected for this study were accessioned in the files of the Armed Forces Institute of Pathology (AFIP), Washington, DC. Specifically, we wanted to examine verrucae, multiple papillomas, and condylomata for the presence of papillomavirus antigens, since papil- lomavirus-like particles have been seen in these le- sions by electron microscopy; keratoacanthomas were examined because of their usual history of "wart-like" spontaneous regression. Verrucae and papillomas were retrieved from the files by location (oral cavity) and title (papillomas, squamous papillo- mas, verruca vulgaris, condyloma accuminatum, keratoacanthoma, florid papillomatosis, papilloma- tosis, verrucous hyperplasia). After hematoxylin and eosin (H&E) sections from each papillomatous lesion were reviewed, at least 5 consecutive 4-µ-thick sections from paraffin blocks containing formalin-fixed tissue were air-dried on glass and stored at room temperature until ready for use. Verrucae Twenty-nine patients (25 male, 4 female) with oral cavity verrucae were included in this study. These verrucae had the following distribution: upper lip, 7; lower lip, 12; tongue, 4; alveolar mucosa, 2; hard palate, 2; buccal mucosa, 1; and floor of mouth, 1. The youngest patient was 8 and the oldest 61, but the majority were 20-30 years of age. PAPII.LOMAVIRUS 213 Condylomata Five patients (all male) of condylomata were ac- cepted for this study; 4 were on the tongue, and 1 was on the soft palate. Their ages were 23, 27, 31, 38, and 63. One patient had concurrent condyloma of the anal mucosa. Multiple Papillomas Five patients (4 male, 1 female) with multiple pap- illomas were included in this study. These lesions had the following distribution: 1) maxillary gingivae; 2) right and left commissures; 3) floor of mouth and tongue; 4) frenulum and mandibular gingivae; and 5) soft palate and left lateral tongue. The 4 males were 18, 24, 25, and 28 years of age; the female was 15. None of these patients stated in the record that they had other verrucous lesions. Keratoacanthomas Twenty-eight cases (27 male, I female) with kera- toacanthomas were accepted for this study: 27 were located on the upper lip, and I was located on the lower lip. The youngest patient was 19, and the oldest was 74; the remaining patients were evenly dis- tributed between these age limits. There was no men- tion of concomitant verruca at the time of removal of the keratoacanthomas. Peroxidase-Antiperoxidase (PAP) Staining of Papillomas Antiserums Purified bovine papillomavirus type I(BPV-1) virions, previously characterized,2t were extracted from a single naturally occurring bovine fibropapil- loma. BPV-1 protein was prepared as previously described,"-10•22 adjusted to 0.24 M 2-mercap- toethanol l% SDS, heated to 68 C for 2 minutes, diluted with 3 volumes of saline, and mixed to an equal volume of Freund's complete adjuvant. A vary- ing concentration of virus protein (480, 240, and 120 µg) was inoculated subcutaneously into a rabbit on Days 0, 15, and 28, respectively; and the animal bled on Day 38. This hyperimmune serum (BPV-1 [SDS]) is reactive by PAP at a 1:100 dilution with papilloma- virus positive (by electron microscopy) cutaneous warts and mucosal papillomas from all species tested (human, cattle, dogs, rabbits, deer and horses) but does not react with virus-negative papillomas (data not shown); it has the same specificity but a higher titer for the papillomavirus genus-specific antigen as TI BU 33307 k 0 I . r--- I

. 214 JENSON ET AL an antiserum (HPV [SDS]) prepared against deter- gent-disrupted, heat-aggregated pooled plantar wart virus as previously described.'= PAP Staining Sections of papillomas were deparaffinized, de- hydrated, and washed with phosphate-bufl'ered saline (PBS). After quenching of endogenous peroxidase activity with 0.507o hydrogen peroxide in methanol and a 30-minute incubation with 20°lo normal goat serum to reduce nonspecific staining, the sections were incubated with hyperimmune BPV-1 (SDS) serum at a dilution of 1:100 for 1 hour. The secon- dary antiserum was goat anti-rabbit immunoglobulin diluted 1:20 and left on the sections for 30 minutes. This was followed by a 30-minute incubation with a rabbit PAP complex diluted 1:50. The reaction was developed by the addition of 0.05°7o, 3,3'-diamino- benzidine tetrachloride and 0.0101o hydrogen perox- ide in 0.05 M-tris buffer (pH 7.6) for 5-8 minutes. Sections were then counterstained with hematoxylin, W- .; .. Ila f ;. , A ~r. e.ri. ~11 i ~ ;,`:. ..a~f, ,+:~~: 1it ; .. r :.. ~° ~ r 9 I; 3i ,1 ' ~ ~~i ! 1 ~~• s , ~~;`1~I ~' . s:,p ~~ 1 ,• . ;a . , tWr 0 , •. • • ••A1 . ~ O •• P 1e • i t . "' ; /// •:.: sr .b . r * .i_•. !4r'A `•f trr •• ~ ^.'.T t~•~,[I/f!!. 3r &A x1: ' ;: ; e » ~A.A . ~ ' : ~ ti ,, i'~ j . ts4. ~i . • • . . l ~ Figure 1- Plantar wart used as positive control for PAP reac• tions. HPV common antigens (brown precipitate. arrows) are seen in nuclei of prickle and granular cells before the latter undergo heratinizaUOn. (PAP and hematoxylm• x 197) TI BU 33308 I ~ AJP • M., l'M17 dehydrated, cleared, and mounted. The reactions and all washings performed between incubations were with PBS (pH 7.4). The specificity of the BPV-l (SDS) antiserum in the immunocytochemical reaction was verified by the control studies that were performed. A variety of normal mucosal tissues (cervix, foreskin, larynx, and gingivae) and brain tissue infected with herpes simplex virus Type 2 were reacted with the BPV-l (SDS) antiserum; these all failed to show a positive reaction with the antiserum (negative control). In contrast, sections of virus-positive (by electron microscopy) human plantar warts (Figure 1) and canine oral papillomas run simultaneously with sec- tions of oral cavity papillomas reacted positively with the BPV-1 (SDS) antiserum (positive controls). Negative controls performed simultaneously with the oral cavity papillomas under study consisted of serial sections incubated with normal rabbit serum and rab- bit type-specific antiserum prepared against intact but pooled BPV-1 and BPV-2 virions.

. 214 JENSON ET AL an antiserum (HPV [SDS]) prepared against deter- gent-disrupted, heat-aggregated pooled plantar wart virus as previously described.'= PAP Staining Sections of papillomas were deparaffinized, de- hydrated, and washed with phosphate-bufl'ered saline (PBS). After quenching of endogenous peroxidase activity with 0.507o hydrogen peroxide in methanol and a 30-minute incubation with 20°lo normal goat serum to reduce nonspecific staining, the sections were incubated with hyperimmune BPV-1 (SDS) serum at a dilution of 1:100 for 1 hour. The secon- dary antiserum was goat anti-rabbit immunoglobulin diluted 1:20 and left on the sections for 30 minutes. This was followed by a 30-minute incubation with a rabbit PAP complex diluted 1:50. The reaction was developed by the addition of 0.05°7o, 3,3'-diamino- benzidine tetrachloride and 0.0101o hydrogen perox- ide in 0.05 M-tris buffer (pH 7.6) for 5-8 minutes. Sections were then counterstained with hematoxylin, W- .; .. Ila f ;. , A ~r. e.ri. ~11 i ~ ;,`:. ..a~f, ,+:~~: 1it ; .. r :.. ~° ~ r 9 I; 3i ,1 ' ~ ~~i ! 1 ~~• s , ~~;`1~I ~' . s:,p ~~ 1 ,• . ;a . , tWr 0 , •. • • ••A1 . ~ O •• P 1e • i t . "' ; /// •:.: sr .b . r * .i_•. !4r'A `•f trr •• ~ ^.'.T t~•~,[I/f!!. 3r &A x1: ' ;: ; e » ~A.A . ~ ' : ~ ti ,, i'~ j . ts4. ~i . • • . . l ~ Figure 1- Plantar wart used as positive control for PAP reac• tions. HPV common antigens (brown precipitate. arrows) are seen in nuclei of prickle and granular cells before the latter undergo heratinizaUOn. (PAP and hematoxylm• x 197) TI BU 33308 I ~ AJP • M., l'M17 dehydrated, cleared, and mounted. The reactions and all washings performed between incubations were with PBS (pH 7.4). The specificity of the BPV-l (SDS) antiserum in the immunocytochemical reaction was verified by the control studies that were performed. A variety of normal mucosal tissues (cervix, foreskin, larynx, and gingivae) and brain tissue infected with herpes simplex virus Type 2 were reacted with the BPV-l (SDS) antiserum; these all failed to show a positive reaction with the antiserum (negative control). In contrast, sections of virus-positive (by electron microscopy) human plantar warts (Figure 1) and canine oral papillomas run simultaneously with sec- tions of oral cavity papillomas reacted positively with the BPV-1 (SDS) antiserum (positive controls). Negative controls performed simultaneously with the oral cavity papillomas under study consisted of serial sections incubated with normal rabbit serum and rab- bit type-specific antiserum prepared against intact but pooled BPV-1 and BPV-2 virions.

Vol. 107 • No. 2 Table 2-Detection of Papillomavirus Common Antigens in Proliferative Squamous Epithelial Lesions of the Oral Cavity Anti-BPV-1 (SDS) Type of lesion No.+ No.- Verrucae Upper hp 7 0 Lower lip 8 4 Other 3 7 Multiple Oral Papillomas 2 3 Condylomata 3 2 Keratoacanthomas 0 28 Results Anti-BPV-I (SDS) was reactive with 18 of 29 ver- rucae, 3 of 5 condylomata, and 2 of 5 multiple papillomas; it did not react with any of the 28 kera- toacanthomas (Table 2). The PAP stain (brown precipitate) was always localized to nuclei of squamous cells in the upper third of the epithelium (Figures 2-4); traces of nonspecific PAP staining was frequently seen in the cytoplasm but not in nuclei. In Figura 2-Verruca from hard palate. Papiilomavirus struc• tural antigens are identified by positive PAP reaction (arrows) in outer squamous epithelial cells of only t papillary frond in this photomicrograph. (PAP and hematoxylin, x 197) PAPILI,OMAVIRUS 215 verrucae and condylomata, the intranuclear PAP staining pattern resembled that of a "raisin" and was most often seen in vacuolated (koilocytotic) cells with prominent keratohyalin granules. In the multiple papillomas, the PAP stain was usually seen in small clusters of cigar-shaped nuclei as well as nuclei of koilocytotic cells in the outermost layer of squamous epithelial cells. Verrucae All verrucae (7) of the upper lip and 8 of 12 from the lower lip were positive by PAP staining. Three of 7 verrucae from other locations (buccal mucosa, hard palate, and maxillary gingivae) in the oral cavity were also positive (Figure 2) for genus-specific antigen(s). Of the 6 patients with concomitant cutaneous ver- rucae, 3 had oral cavity verrucae that were positive by PAP. Condylomata The 3 (of 5) PAP-positive condylomata (Figure 3) were located on the tongue. The I patient with a con- TI BU 33309 . r

Vol. 107 • No. 2 Table 2-Detection of Papillomavirus Common Antigens in Proliferative Squamous Epithelial Lesions of the Oral Cavity Anti-BPV-1 (SDS) Type of lesion No.+ No.- Verrucae Upper hp 7 0 Lower lip 8 4 Other 3 7 Multiple Oral Papillomas 2 3 Condylomata 3 2 Keratoacanthomas 0 28 Results Anti-BPV-I (SDS) was reactive with 18 of 29 ver- rucae, 3 of 5 condylomata, and 2 of 5 multiple papillomas; it did not react with any of the 28 kera- toacanthomas (Table 2). The PAP stain (brown precipitate) was always localized to nuclei of squamous cells in the upper third of the epithelium (Figures 2-4); traces of nonspecific PAP staining was frequently seen in the cytoplasm but not in nuclei. In Figura 2-Verruca from hard palate. Papiilomavirus struc• tural antigens are identified by positive PAP reaction (arrows) in outer squamous epithelial cells of only t papillary frond in this photomicrograph. (PAP and hematoxylin, x 197) PAPILI,OMAVIRUS 215 verrucae and condylomata, the intranuclear PAP staining pattern resembled that of a "raisin" and was most often seen in vacuolated (koilocytotic) cells with prominent keratohyalin granules. In the multiple papillomas, the PAP stain was usually seen in small clusters of cigar-shaped nuclei as well as nuclei of koilocytotic cells in the outermost layer of squamous epithelial cells. Verrucae All verrucae (7) of the upper lip and 8 of 12 from the lower lip were positive by PAP staining. Three of 7 verrucae from other locations (buccal mucosa, hard palate, and maxillary gingivae) in the oral cavity were also positive (Figure 2) for genus-specific antigen(s). Of the 6 patients with concomitant cutaneous ver- rucae, 3 had oral cavity verrucae that were positive by PAP. Condylomata The 3 (of 5) PAP-positive condylomata (Figure 3) were located on the tongue. The I patient with a con- TI BU 33309 . r

I 216 JENSON ET AL .~P ror ~ .. f `~ . ... , ~ • ~ ,~a~,o ^. .~ *~ '~ , `~ `1tf t~ 7 it ' !v'~YF•'! 1!r" ;4 q j t* ~ ~ft Y AJP • May 1982 Figure 3-Multiple papilloma from maxillary gingiva. Tangen• tial section shows several clus- ters of superfic al squamous cells with nuclei that are PAP• positive (arrows) for papoloma• virus. (PAP and hematoxylin, x 197) comitant anal-genital condyloma had a PAP-positive oral cavity condyloma. Multiple papillomas The 2 (of 5) PAP-positive multiple papillomas (Figure 4) were from the soft palate and maxillary gingivae and lateral tongue, respectively. Keratoacanthomas All (28) keratoacanthomas were PAP-negative. Discussion The papillomavirus genus, together with the poly- omavirus genus, constitutes the family Papova- viridae." A previous study showed that all viruses of the same genus possess common whole (structural) virus antigens that can be detected in nuclei of in- fected cells by antiserum prepared against disrupted but not intact virions of any member of the genus.l= The antiserum used by us and others to identify papillomavirus genus-specific antigens in formalin- fixed paraffin-embedded verrucae, condylomata, and multiple papillomas (Table 2) and laryngeal papil- lomasl' of the oral cavity and cervical dysplasias and vulvar condylomata of the female genital tract'° was prepared from detergent-disrupted bovine papillo- mavirus Type 1(BPV-1 [SDSJ) obtained from cutaneous fibropapillomas of cattle. Besides being a readily available source of large quantities of papillo- mavirus, the use of purified BPV-1 (SDS) as antigen provides an antiserum that would be less likely to give false reactions with endogenous intranuclear proteins and DNA in human tissue than an antiserum derived from HPV. Since papillomaviruses do not appear to exist in nature as passenger viruses, identification of HPV in a cutaneous wart or mucosal papilloma in all likeli- hood establishes the etiology of the lesion. Failure to detect HPV, however, does not necessarily exclude it as the etiologic agent. In this study papillomavirus common antigens were identified in 62010 (18 of 29) of TI BU 33310

I 216 JENSON ET AL .~P ror ~ .. f `~ . ... , ~ • ~ ,~a~,o ^. .~ *~ '~ , `~ `1tf t~ 7 it ' !v'~YF•'! 1!r" ;4 q j t* ~ ~ft Y AJP • May 1982 Figure 3-Multiple papilloma from maxillary gingiva. Tangen• tial section shows several clus- ters of superfic al squamous cells with nuclei that are PAP• positive (arrows) for papoloma• virus. (PAP and hematoxylin, x 197) comitant anal-genital condyloma had a PAP-positive oral cavity condyloma. Multiple papillomas The 2 (of 5) PAP-positive multiple papillomas (Figure 4) were from the soft palate and maxillary gingivae and lateral tongue, respectively. Keratoacanthomas All (28) keratoacanthomas were PAP-negative. Discussion The papillomavirus genus, together with the poly- omavirus genus, constitutes the family Papova- viridae." A previous study showed that all viruses of the same genus possess common whole (structural) virus antigens that can be detected in nuclei of in- fected cells by antiserum prepared against disrupted but not intact virions of any member of the genus.l= The antiserum used by us and others to identify papillomavirus genus-specific antigens in formalin- fixed paraffin-embedded verrucae, condylomata, and multiple papillomas (Table 2) and laryngeal papil- lomasl' of the oral cavity and cervical dysplasias and vulvar condylomata of the female genital tract'° was prepared from detergent-disrupted bovine papillo- mavirus Type 1(BPV-1 [SDSJ) obtained from cutaneous fibropapillomas of cattle. Besides being a readily available source of large quantities of papillo- mavirus, the use of purified BPV-1 (SDS) as antigen provides an antiserum that would be less likely to give false reactions with endogenous intranuclear proteins and DNA in human tissue than an antiserum derived from HPV. Since papillomaviruses do not appear to exist in nature as passenger viruses, identification of HPV in a cutaneous wart or mucosal papilloma in all likeli- hood establishes the etiology of the lesion. Failure to detect HPV, however, does not necessarily exclude it as the etiologic agent. In this study papillomavirus common antigens were identified in 62010 (18 of 29) of TI BU 33310