Antibody response to human papillomavirus (HPV) type 11 in children with juvenile-onset recurrent respiratory papillomatosis (RRP)

VIROLOGY

188, 384-387 (1992)

Antibody Response to Human Papillomavirus (HPV) Type 11 in Children with Juvenile-Onset Recurrent Respiratory Papillomatosis (RRP) WILLIAM BONNEZ,*~’ HASKINS K. KASHIMA, t BRIGID LEVENTHAL, t PHOEBE MOUNTS, t ROBERT C. ROSE,* RICHARD C. REICHMAN,* AND KEERTI V. SHAtit *Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York 14642; and tJohns Hopkins Medical Institutions, The Johns Hopkins University, Baltimore, Maryland 2 12 18 Received October 28, 199 1; accepted January 29, 1992

We previously established, using an ELISA, the presence of specific antibodies directed at human papillomavirus (HPV) type 11 virions in the sera of patients with condylomata acuminata, mostly a disease of young adults that, like recurrent respiratory papillomatosis (RRP), is caused by two closely related HPVs, types 8 and 11. The present study was done to investigate if children with RRP can make viral-specific antibodies to an infection that is acquired at birth. Using the same ELISA, we studied the sera of 32 children with biopsy-documented juvenile-onset RRP and compared them to the sera of 31 control children. The median (and interquartile range) of the OD values in the controls and the cases was 0.078 (0.003,O.lOl) and 0.230 (0.063,0.725), respectively, a statistically significant difference (P = 0.001). Among the cases, there was no difference in seroreactivity between children with HPV-11 -induced RRP and those with HPV-6-induced RRP (P = 0.31). Since HPV-11 viral particles do bind to the ELISA plate and remain intact and accessible to antibodies, we conclude that children with RRP, like adults with condylomata acuminata, develop antibodies directed at HPV-11 virions. o 1992 Academic press. hc.

from control children (1 1 boys and 20 girls). Most of the patients were participants in a recent trial of interferon for the treatment of RRP. HPV infection was confirmed by Southern hybridization of papilloma tissue DNA with HPV probes as described (2) and typing was available for 29 patients; 17 had HPV-1 1 and 12 had HPV-6 DNA in their tissues. The median age of the RRP patients was 5.5 years (range 1, 44), and 5 patients were older than 10 years of age. The controls were children from suburban Maryland who were bled during a previous rubella immunization study conducted in 1968 and 1969. The median age of the control children was 3 years (range 1, 7). The sera were coded and assayed by a blinded observer. Animal sera came from two rabbits. One animal was immunized with a /I%galactosidase fusion protein expressed in Escherichia co/i and derived from a pEX vector containing a 480-bp fragment of the Ll open reading frame of HPV-6b (pEX-480) as previously described (7). This protein contains the common papillomavirus antigen present in denatured papillomavirus particles (8). The other animal was immunized by a similar method against HPV-11 virions prepared as for the ELISA (9). From both animals pre- and postimmune serum samples were obtained, and seroconversion was demonstrated against the respective antigens. The ELISA technique has been previously described in detail for human sera (6). The same ELISA was also

Recurrent respiratory papillomatosis (RRP) is a rare disease in which benign papillomas occur in the respiratory tract, most commonly in the larynx (laryngeal papilloma) (1). It is caused by two closely related genital tract human papillomaviruses (HPVs) types 6 and 11, which are also responsible for exophytic genital warts (condylomata acuminata) (2, 3). Juvenile-onset RRP probably occurs as a result of transmission of HPV-6 or HPV-11, at birth, from mother to the newborn during its passage through an infected birth canal (4). While the HPV-6 and HPV-1 1 etiology of RRPis firmly established, it is not known if the child with RRP can make antibodies to an infection acquired at birth. Recently, using as antigen HPV-1 1 virions derived from the mouse xenograft system described by Kreider et al. (5), a type-specific serum antibody response to HPV infection was measured by ELISA in patients bearing HPV-6- or -11 -induced condylomata acuminata (6). Using this ELISA to determine if RRP patients develop an antibody response to their HPV infections, we tested the sera of 32 patients (16 males, 14 females, and 2 subjects for whom sex was not recorded) with a history of juvenile-onset disease and with a pathologically confirmed diagnosis of RRP as well as 31 sera ’ To whom reprint requests should be addressed at Infectious Diseases Unit-Box 689, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, New York 14642. 0042-6822/92

$3.00

CopyrIght 0 1992 by Academic Press, Inc. All rtghts of reproduction I” any form reserved.

384

385

SHORT COMMUNICATIONS

s 8 25

500

-

0

-

8

A

B

FIG. 1. Reactivity of the sera of (A) controls, and (B) cases with RRP. For each group of values, a boxplot is superimposed on a dotplot. Each box includes the mid-50% of the values, and the horizontal bar within represents the median absorbance measurement. The short horizontal lines at the end of the vertical lines extending below and above the box are the inner fences (18). Values outside the inner fences are indicated by an asterisk or a full circle.

carried out using the four rabbit serum samples (one per plate) as primary antibody, instead of the human sera. The sera were tested in triplicate at two dilutions (1:lOOO and 1:2000), and against three dilutions of the antigen (1 :l, as with the human sera, 1:5, and 1:25). The secondary antibody was a goat anti-rabbit IgG (y + light chains) conjugated to alkaline phosphatase (Tago), used at a 1: 1000 dilution. Statistical analysis was done with Mann-Whitney’s U test for scalar variables and Fisher’s exact test for categorical variables. Logistic regression was used to adjust the analysis of the absorbance values for age of the subject. Two-sided P values are reported, and a P < 0.05 was considered statistically significant. The absorbance values of the case and control sera in ELISA with HPV-1 1 virions are shown in Fig. 1. The median (and inter-quartile range) of the absorbance values for juvenile-onset RRP patients was 0.230 (0.063, 0.725) as compared to the corresponding values in controls of 0.078 (0.003, 0.101); this difference was highly significant (P = 0.001). This difference remained significant (P = 0.04) after adjusting for age of the subjects, a variable for which case and control groups were different (P < 0.001). Nine of the 32 cases (28%) had absorbance values greater than the highest control value of 0.649. We chose the value of the upper inner fence of the boxplot of the controls (i.e., 0.249) (18) as our cut-off absorbance value to classify individual serum specimens as antibody-positive or antibodynegative. By this criterion, 15 of 32 cases (47%) and 3 of 31 controls (100/o) were positive (P < 0.005). The

three positive controls were 2, 6, and 6 years old. HPV6 and HPV-1 1 are closely related but distinguishable viruses. We therefore compared the serum reactivity of 17 patients with HPV-1 l-induced RRP with that of 12 patients with HPV-6-induced RRP (Fig. 2). The median OD (and inter-quartile range) of the HPV-1 1 group, 0.383 (0.100, 0.794) was not significantly different from that of the HPV-6 group 0.132 (0.027, 0.480) (P = 0.31). A complementary part of our study was to show that HPV-1 1 virions bound intact to the ELISA plate and were recognizable as antigens. The rabbit antibody (anti-pEX-480) directed against the common papillomavirus antigen, although able to react with denatured HPV-1 1 virions by Western blot (data not shown), did not bind to the antigen used in the ELISA (Fig. 3A). In contrast, the antibody directed against HPV-1 1 viral particles reacted in a dose-dependent manner to the antigen (Fig. 3B). We therefore conclude that the HPV-11 ELISA was disease specific and that RRP patients developantibodies to HPV-1 1 capsid antigens. Using the same assay, we have previously reported remarkably similar results in condylomata acuminata patients (6). In particular, in both studies about 309/o of sera from patients had reactivities higher than the highest reactivity of the respective controls. Almost all cases of RRP and a large majority of condylomata acuminata are caused by HPV types 6 and 11. Therefore, a similar serological profile for the two diseases is not surprising. In addition, the results strengthen the notion of a common pathogenesis for condylomata acuminata and RRP and provide additional evidence for the specificity of the virion-based ELISA. As with condylo-

1500

/I

1000 s 8 25

500

8

A

B

FIG. 2. Reactivity of the sera from cases with (A) HPV-64nduced RRP, and (B) HPV-1 l-induced RRP. The values are otherwise displayed as in Fig. 1.

386

SHORT COMMUNICATIONS

B

Anti-pEX-480 Antibody

Anti-HPV-11 Virion Antibody

OD 2.000

1.500

1.000

0.500

0

1:l

1:5

1:25

Antigen Dilution

I

I

1:5

1:25

Antigen Dilution

FIG. 3. Reactivity of rabbit antibodies directed against (A) the pEX-480 protein product, bearing the common papillomavirus HPV-1 1 virions. For each point the vertical bar, wherever large enough to be visible, represents the standard deviation.

mata acuminata patients, no difference in serum reactivity was discernible between patients with HPV-6- or HPV-1 l-positive biopsies, possibly a reflection of the fact that HPV-6 and HPV-1 1 DNAs have 829’0 nucleotide sequence homology (10). All of the RRP patients had a history of significant clinical disease, and most of them had clinical disease when they were bled. In many instances, the disease was of long duration. Yet, only 47% of the case sera had a reactivity above the cut-off value. This modest sensitivity of ELISA reflects, at least in part, technical limitations, such as low virion concentration (W. Bonnez, unpublished data). The marked difference between the reactivities of RRP case and control sera that was seen with HPV-11 virion-based ELISA in this study was not demonstrable in earlier studies when we tested sera from the same two groups of cases and controls by ELISA with HPV-6 Ll, L2, and E2fusion proteins (11; R. R. Arthur, C. C. Li, and K. V. Shah, unpublished data). Reactivities of both groups of sera with all three fusion proteins were very low. Less than 5% of the sera in either group was scored as positive, and there was no difference in reactivity between cases and controls. This suggests that conformational epitopes, dependent on the tertiary and quaternary structure of viral proteins, are responsible for a specific humoral immune response. Studies with HPV-1, the causative agent of plantar warts, also indicate that antigenic specificity is located on the virion. For instance, sera of patients with plantar warts react better with HPV-1 virions than sera of patients

antigen; and (B)

with other papillomavirus-related diseases (12- r5), including RRP ( 15- 17). Despite high cost and low sensitivity, the HPV-1 1 virion-based ELISA could serve as a reference for the development of less costly, more sensitive methodologies.

ACKNOWLEDGMENTS We are grateful to John W. Kreider for the use of the HPV-1 lHarshey strain. We are also indebted to Came DaRin for excellent technical assistance. We thank Debbie Reiss and the staff of the xenograft facility at the University of Rochester for their support in the care of the animals. David S. Strike provided us with some of the HPV-1 l-infected human xenografts. Reid Mattison and Elizabeth Woodward graciously gave us access to human neonatal foreskins. This study has been partially supported by Contract NOl-AI-82509 from the National Institute of Allergy and Infectious Diseases (R. C. Reichman), and Grant POl-Al-16959 from the U.S. Public Health Service (K. V. Shah).

REFERENCES 1. STRONG, M., VAUGHAN, C., and HEALY, G.. In “Laryngo-Tracheal Problems in the Pediatric Patient” (G. B. Healy. and T. 1. I. McGill, Eds.), pp. 88-98. Springfield, IL, 1979. 2. MOUNTS, P., SHAH, K. V., and KASHIMA. H., Proc. /Vat/. Aced. Sci. USA 79, 5425-5429 (1982). 3. GISSMANN, L., DIEHL, V.. SCHULTZ-LOULON,H. J., and ZUR HAUSEN, H., 1. tiroi. 44, 393-400 (1982). 4. SHAH, K., KASHIMA, H., POLK, B. F., SHAH, F., AEBEY, H., and ABRAMSON, A., Obsfet. Gynecol. 68, 795-799 (1986).

SHORT COMMUNICATIONS 5. KREIDER,J. W., HOWET, M. K., WOLFE, S. A., BARTLETT, G. L., ZAINO, R. J., SEDLACEK, T. V., and MORTEL, R., Nature 317, 639-641 (1985). 6. BONNEZ, W., DARIN. C.. ROSE, R. C., and REICHMAN, R. C., J. Gen. Viral. 72, 1343-1347 (1991). 7. STRIKE, D. G., BONNEZ, W., ROSE, R. C.. and REICHMAN, R. C., /. Gen. Viral. 70, 543-555 (1989). 8. JENSON,A. B., ROSENTHAL,I. D., OLSON, C., PASS, F., LANCASTER, W. D., and SHAH, K., /. Nat/. Cancerhst. 64,495-500 (1980). 9. BONNEZ,W., ROSE, R. C., and REICHMAN, R. C.J. Infect. Dis. 165, 376-380 (1992). 70. DARTMANN, K., SCHWARTZ,E., GISSMANN, L., and ZUR HAUSEN, H., Virology 151, 124-130 (1986). 11. LI, C. H.. SHAH, K. V., SETH, A., and GILDEN, R. V., /. Viral. 61, 2684-2690 (1987).

12. KIENZLER,J. L., LEMOINE, M. T., ORTH, G., JIBARD, N., BLANC, D., LAURENT, R., and AGACHE, P., Br. J. Dermatol. 108, 665-672 (1983). 13. ANISIMOVA, E., BARTAK, P., VLCEK, D., HIRSCH, I., BRICHACEK,B., and VONKA, V., J. Gen. viral. 71, 419-422 (1990). 14. STEELE, J. C., and GALLIMORE, P. H., Virology 174, 388-398 (1990). 75. VIAC, J., CHOMEL, J.-J., CHARDONNET,Y., and AYMARD, M., /. Med. Vii-o/. 32, 18-2 1 (1990). 16. PFISTER. H., and ZUR HAUSEN, H., ht. J. Cancer 21, 161-165 (1978). 77. PFISTER,H., HUCHTHAUSEN,B., GROSS,G., and ZUR HAUSEN, H., /. Nat/. Cancer Inst. 62, 1423-l 425 (1979). 78. TUKEY, 1. W. In “Exploratory Data Analysis.” Addison Wesley Publishing Co., Reading, MA 1977.