An easy-to-prepare, highly efficient antigen for cervical cancer serology There is ample evidence that human papillomaviruses (HPV) are involved in the etiology of cervical cancer (CC) (zur Hausen, 1991). These viruses do not grow in tissue culture. Therefore, sophisticated preparations such as genetically engineered viral proteins or synthetic peptides are used as antigens for detection of HPV antibodies in human sera. With some of these antigens, it has repeatedly been possible to demonstrate significantly higher prevalence of HPV antibodies in CC patients than in control subjects (for review, see Calloway, 1992). However, even the HPVl6 E7 antibody, which had been found to be associated with the disease most fiequentb, has not been detected in the majority of patients, including those in whom the presence of HPVI 6 DNA has been demonstrated (Dillner, 1990; Jha et al., 1993; Krchfiak, 1990;Mann et al., 1990;Onda et al., 1993). This has most probably been associated with the absence of conformational epitopes in the preparations used; antibodies formed in the course of tumour development may mainly be directed against such epitopes. We have observed that an extract from CaSki cells, a cell line derived from HPVl6 DNA-positive CC tissue and expressing HPVl6 E6 and E7genes (Baker et al., 1987) can serve as an antigen reactive with a great majority of CC sera but only rarely with sera jrom matched control women.CaSki cells were kindly provided by L. Gissmann, German Cancer Center, Heidelberg. They were cultivated like other cell lines kept in this laboratory (Vonka et al., 1979). Cell extracts were prepared employing the procedure we had successfully used for isolating EB virus nuclear antigen (EBNA-I) from EB virus-positive lymphoblastoid cells (Hirsch et al., 1978). HaCat cells, a spontaneously transformed human keratinocyte cell line (Boukamp et al., 1988), received through the courtesy of N.E. Fusenig, Heidelberg, served as the source of control antigen. IgG-specijic ELISA was essentially the same as in previous experiments (fichfihk et al., 1990). In the test, 4-8 units of antigen diluted in PBS (pH 7.3) were used, one antigenic unit being defined as the highest antigen dilution still giving a positive reaction in the presence of antibody excess.The cut-off value between positivity and negativity was the mean absorbance value (A) plus 3 SDs as determined in sera from healthy women; before the calculation was made, the outliers were eliminated. Control HaCat-derived antigen was used at the same dilution as the CaSki-derived antigen. Control sera known to be positive and negative were included in each experiment and tested on each plate. Sera from 34 CC patients (age 20-70 years) and 41 healthy women matched with the patients by age and living area were tested. All CC patients were incident cases. Sera were tested in a 1:25 dilution only.Results are shown in Figure 1. It can be seen that 31 (92.2%)patient sera and 3 (7.3%) controlsera gave a positive I I
1.mo600 0 controls cervical carcinoma patients a00 FIG...