The use of DNA from paraffin wax preserved tissue for predictive diagnosis in familial adenomatous polyposis.

Received 16 December 1991. Revised version accepted 7 February 1992. Familial adenomatous polyposis (FAP) is a dominantly inherited condition which predisposes family members to the development of colorectal carcinoma, usually in their thirties. Carcinoma is preceded by the development of widespread adenomatous polyps in the large bowel, and this feature has been used as a diagnostic marker for over 50 years. In 1987 the causative gene (the APC gene) was localised to a small area on the long arm ofchromosome 5 (5q21-22).' 2 Since that time it has been possible to identify gene carriers using polymorphic DNA probes closely linked to the APC gene on chromosome 5. With the development of flanking markers diagnostic accuracy can be greater than 99%.3 Unfortunately, such analysis needs complete family pedigrees and, as a consequence of early death from colorectal cancer, is not possible in many FAP families. In the West Midlands Region, of 47 families studied only 17 (36%) had suitable pedigrees for analysis. Similar findings have been reported from other centres.4 Fifteen of the families in our region (32%) were unsuitable because of the early death of key affected family members. For this reason we have investigated the analysis of DNA extracted from paraffin wax embedded tissue from dead family members in order to try to extend informativity in these families. Such tissue is widely available because most dead affected relatives have undergone surgery for colorectal cancer or its complications. In 1990 a known FAP family from Sheffield was referred for genetic counselling (figure). This was prompted by the death of the proband (III-6) from advanced colorectal carcinoma at the age of 32. She left two children who were at risk of developing the disease (IV1 and IV-2). Neither had evidence of colonic polyps at the time of bowel screening (15 and 18 years of age respectively), but at these ages such findings do not dramatically alter their future risk of developing the condition.5 Neither offspring had any evidence of congenital hypertrophy of the retinal pigment epithelium, but this was of limited prognostic significance because diagnostic lesions were not seen in their two affected cousins. It was not possible to establish phase for the two children primarily because the maternal DNA was unavailable for analysis. However, paraffin wax embedded tissue had been stored from her operation in 1987. Ideally a 04 g piece of histologically normal bowel mucosa is used for DNA extraction. In this case, however, only omental tissue (including tumour) was available. Tumour material is not ideal for analysis as deletion of the APC gene can also result in deletion of the DNA region containing the polymorphism. However, in this case there was no evidence of allele loss for Cllpll (figure) so the DNA results could be used. The sectioned tissue was treated with xylene to dissolve the wax, then washed in ethanol and water, and recovered by centrifugation. DNA extraction was performed by resuspending the tissue in 5 ml of lysis buffer (Tris 0-1 mol/l, EDTA 0 01 mol/l, pH 9,6 to which were added 0-5 ml 10% SDS and 250 gl proteinase K (20 mg/ml). The DNA was purified with phenol/chloroform and precipitated in two volumes of 100% ethanol overnight at 20°C. DNA was resuspended in 200 gl of TE buffer. Two primers for the 4 base pair deletion in C1 lpi 1 were used to amplify the genomic DNA fragment from each family member.7 The amplified products were run on a 12% polyacrylamide gel at 110V overnight. The alleles (66 and 70 bp) were identified by staining with ethidium bromide. The alleles associated with each tested family member are shown in the figure. Similarly, two PCR primers across a CA repeat region in ECB27 were used to amplify a second polymorphic region (fragment size 115 and 117 base fragments). These fragments were separated on an 8% denaturing polyacrylamide gel run at 40W for three to four hours. The bands were identified by inclusion of 32P radiolabelled dCTP in the PCR reaction mixture and subsequent exposure to x ray film. A discussion of the results is given in the legend to the figure.