This thesis deals with several methodological and practical aspects of prenatal diagnosis.

A genetic metabolic disease can only be detected in utero if the enzymatic defect or storage of a specific metabolite is expressed in cultured cells from an affected individual. This should be demonstrated for each separate disease. An^- example of such a study is presented in appendix paper I. The data in­dicate, that the enzymatic defect in the haemolytic anaemia caused by pyru­vate kinase deficiency (an autosomal recessive disease) is not detectable in fibroblasts and cultured amniotic fluid cells, since these only produce the M-type of this enzyme, whereas the anaemia is related to a deficient activi­ty of the L-type. As a consequence, prenatal diagnosis of this disease is not possible.

The prenatal diagnosis of enzymatic defects requires analysis of a sufficient number of amniotic fluid cell cultures cultivated under identical conditions and for a comparable period as the diagnostic sample. Sufficient amniotic fluid samples from normal pregnancies will generally not be available at the very moment these are required. To solve this problem a method was develop-ped to store uncultured amniotic fluid cells without loss of cell viability. This method is described in appendix paper II and has enabled us to collect suffi­cient control material for later use in prenatal diagnosis of enzymatic defects. The possible detrimental effects of transport periods of up to a few days on the viability of amniotic fluid cells was tested simultaneously and the results indicate that transport during periods of up to seven days did not interfere with succesfull cell cultivation. This means that prenatal diagnosis can be offered to people living at long distances from the center performing the ana­lysis.

The prenatal diagnosis of metabolic disorders usually required the biochemical a-nalysis of a few million cells and accordingly the time to obtain these cells during cultivation varied from 4-8 weeks after amniocentesis. Such a long waiting pe­riod is a serious psychological disadvantage and also interruption of the pregnancy should for medical and ethical reasons - when needed - be carried out as early as possible. This problem was experienced in a number of groups (Littlefield, 1971; Galjaard et al., 1972; Milunsky et al., 1972; Epstein et al., 1972; Golbus et al., 1974). To reduce this waiting period the sensitivity of the analytical techniques had to be increased to enable the investigation of small numbers of cultured amniotic fluid cells. Our group has developped such methods for the microchemical analysis of several enzymes involved in genetic diseases (Galjaard, 1972; Galjaard et al., 1974°; 1974V The methodology for one of these enzymes (acid α-l ,4-glucosidase, deficient in Pompe's disease) is described in appendix paper III. Micromethods for incubation of small numbers of cells and small volumes of reagents were tested as well as different substra­tes either for spectrophotometric or fluorometric assays. Significant reductions in the number of cultured cells required for a diagnosis could be achieved by using microvolumes of the substrate 4-Methylumbelliferyl a-D-glucopyranoside. Analysis could be performed on homogenates containing the equivalent of about 1000 cultured amniotic fluid cells for each incubation; the total cell number for a prenatal diagnosis, including a protein assay, could be estimated as the equivalent of 4000 homogenised cultured cells. A further reduction could be realised by adaptation of preparation techniques developed for tissue sections by Lowry(Lowry, 1953; Lowry & Passoneau, 1972). Small groups (about 100) of freeze dried amniotic fluid cells are dissected under microsco­pic control from a culture grown on a thin transparant plastic foil.

The increasing use of cultured cells for diagnostic and genetic studies on human metabolic diseases warrants the study of a number of factors influencing

the expression of enzymatic activities in in vitro cultivated cells. Appendix paper IV reports a number of such studies on the causes of variation of lyso­somal enzymatic activities. Wide variations in the activities of acid α-l,4-glucosidase in fibroblasts and cultured amniotic fluid cells from controls and heterozygotes for type II glycogenosis were observed. Part of these were cau­sed by differences in cultivation conditions. It became apparent, that for a number of lysosomal hydrolases a proportional relation exists between the du­ration of the culture phase (passage) and the enzymatic specific activity. Va­riations of α-l ,4-glucosidase residual activities observed in fibroblasts from different patients with Pompe's disease might be related to a genetic hetero­geneity, i.e. different mutant genes causeating identical diseases. Levels of acid α-l ,4-glucosidase in primary and later subcultures of control amniotic fluid cell cultures show considerable variations and this should be considered in comparative studies in actual prenatal diagnosis. Microassays of enzymatic activity in small cell numbers from the same culture show considerable vari­ation of activities among individual cells and these might be caused by locqf differences in in vitro cell proliferation or differences in the cell cycle. Studies on genetic heterogeneity are possible by fusing (appendix paper IV) cellstrains from patients with apparently identical metabolic diseases and stu­dying the occurrence of complementation, i.e. the restoration of enzymatic activity in hybrid cells. If the latter phenomenon occurs, this is an indication that the disease is caused by different gene mutations in the two patients. p-Galactosidase assays in individual binuclear hybrid cells were carried out to study the genetic background of different clinical variants of GM.-ganglio­sidosis.

In appendix paper V the practical application is described of microchemical enzymatic analysis in 6 pregnancies at risk for Pompe's disease. Another example of application in prenatal diagnosis is described in appendix paper

balanced fetal karyotypes were found.

Fetal chromosome studies in 43 pregnancies at risk for an X-linked disease indicated the presence of a male fetus in 21 cases.

47 Analyses of alpha-fetoprotein in amniotic fluid from pregnancies at risk for a neural tube defect indicated the presence of 3 affected fetuses in the se­cond half of pregnancy; these last fetuses were found to be anencephalic on ultra­sound. In 34 pregnancies prenatal monitoring was carried out for 11 different me­tabolic diseases and 3 affected fetuses were detected. This type of investigation requires specific experience in the microchemical analysis of cultured amnio­tic fluid cells. As most of these diseases are infrequent, cooperation between different European centers to this purpose is needed; in the present study 40% of the samples for metabolic diseases was referred by different European cen­ters.

The detection of a fetal abnormality or a male sex (in cases of X-linked dis­orders) was the indication for selective abortion in 36 of 350 pregnancies tes­ted. However, in 90% of the pregnancies the result of prenatal testing ex­cluded that the fetus was affected by the disease involved and as a result these pregnancies could be continued without further anxiety for the parents.