A method is described enabling prenatal diagnosis of glycogenosis II (Pompe's disease) within 7—10 days after amniocentesis in the 14th—16th week of pregnancy. The procedure is based on amniotic fluid cell cultivation on thin plastic foil, freezing and freeze-drying of the cells and subsequent microdissection of pieces of plastic foil each containing 70-250 lyophilized cells. These are incubated in microliter volumes of substrate. The extinction or fluorescence values are measured in microcuvettes adapted to a normal spectrofluorometer or in microcapillaries using a microscope spectrofluoiometer design. The methods developed are compared with conventional biochemical analyses.

INTRODUCTION

Hers¹ delineated the basic defect in type II glycogenosis (Pompe's disease) as a deficient activity of the lysosomal acid a-i,4-glucosidase. In patients with Pompe's disease this enzyme deficiency has been demonstrated in various tissues and organs².³. Nitowsky and Gninfeld⁴ showed that the acid α-glucosidase deficiency is also ex­pressed in leucocytes and in cultivated skin fibroblasts. This not only facilitated the diagnosis of patients and heterozygous carriers but also formed the basis for prenatal detection of this autosomal recessive disease. The first examples of prenatal diag­nosis of Pompe's disease were based on a deficiency of α-glucosidase in amniotic fluid, in uncultivated and cultivated fluid cells⁵.⁶. Later, it became apparent that the use of amniotic fluid and uncultivated cells could result in erroneous interpretations (refs. 7, 8), because of the possibility of contamination with maternal components and because of the presence in amniotic fluid of an α-glucosidase with different properties than the lysosomal acid glucosidase which is deficient in Pompe's disease'.⁹.

The necessity to use cultivated amniotic fluid cells for a reliable prenatal diag­nosis implied a long waiting period. So far 4 6 weeks of in vitro cultivation (after amniocentesis in the 14th 18th week of pregnancy) was necessary to obtain sufficient cell material for α-glucosidase analysis⁵-⁶.¹⁰. Such long waiting periods not only mean

a psychological stress for the parents concerned but also in the case of an abnormal fetus interruption of a pregnancy at later stages may result in more obstetrical difficulties. The use of microchemical techniques⁸-^,1_15 enables a reduction of the number of cultured cells required for analysis and preliminary data on the application of such microtechniques in the prenatal analysis, of a number of lysosomal enzymes in cultured amniotic fluid cells have been presented"'¹⁶.

In this paper a micromethod is presented which allows prenatal diagnosis of Pompe's disease within 7-14 days after amniocentesis in the I4th-i6th week of pregnancy. The method developed has been compared with other assay procedures and some problems encountered in the practical use of the various methods will be discussed. The application of this micromethod in the prenatal monitoring of several pregnancies at risk for Pompe's disease will be described separately".

MATERIAL AND METHODS

In vitro cultivation and preparation of cell material

Control amniotic fluid was obtained by transabdominal amniocentesis between the 14th-!6th week of pregnancy after ultrasound localization of placenta and fetus. 5-15 ml amniotic fluid was centrifuged in siliconiscd glass tubes and the cell pellet was resuspended in Ham's Fio medium supplemented with 20% fetal calf serum. When microchemical studies were to be performed on dissected small clones of freeze-dried cells, in vitro cultivation was carried out on "mylar dishes" consisting of a 5 cm diameter glass ring and a bottom of thin plastic (Melinex polyester film type 0, I.C.I. Holland). For (micro)biochemical studies on celt homogenates cultivation was per­formed in 35 mm falcon dishes. Established fibroblast strains from control individuals, patients with Pompe's disease and heterozygous carriers were cultured under similar conditions, using Ham's Fio supplemented with 15% fetal calf serum.

Amniotic fluid cells were either directly used foi cell cultivation and subsequent analysis or were cultivated after storage of the unprocessed cells, using a new method (ref. 18). Cell homogenates were prepared by trypsinization with 0.25% trypsin solution and washing of the cells which were then suspended in 0.001% bovine albumin. After counting the cell number in a hemocytometer, homogenizatioii was carried out by sonic vibration (2 X 30 sec) or by repeated freezing and thawing. In most experiments a concentration of about io^e cells per ml was used (corresponding to a protein concentration of about 0.3 mg/ml). Analysis of cell homogenates was either performed directly or after storage of the cell sonicate at —70⁰. Depending on the type of experiments 1-10 μ\ aliquots of cell homogenate were used for α-i,4-glucosidase and protein analyses.

For direct assays on freeze-dried cells (Fig. 4), the "mylar" dishes containing groups of cultured amniotic fluid cells or fibroblasts were frozen in CO₂-acetone (—70°) or in isopentane-liquid N₂ ( — 170°) after removing the medium by three times rinsing in isotonic salt solution. The dishes were then transferred to a freeze-drying apparature (W.K.F., type 05, Wetzlar Instr.) while kept at low temperature. Lyo-philization of the cultured cells was carried out in vacuo (io-^! mm Hg) at —45° during about 15 hours and the vacuum was released after the dishes were allowed to warm to 20°. Freeze-drying of the cells not only allows storage for long periods (at —70°) but it also enables isolation of cells without any loss of enzyme activity.

In a conditioned room (20⁰ ± 0.5 and relative humidity ^40%) pieces of plastic foil containing small numbers (tenths-hundreds) of freeze-dried cells were iso­lated using free hand microdissection under a stereomicroscope (zoomobjective, total magnification 60-120x). The number of cells on each dissected piece of plastic was counted using a raster specially designed for this purpose (Micropure, Arnhem, Holland) which is placed under the culture dish during microdissection. The dissected pieces of foil, containing a known number of cells (70-250) were used for further microchemical analyses.

(Micro)biochemical assays

Protein analysis. The method described by Lowry et al.¹' was adapted to a smaller final volume to decrease the amount of cell material required for analysis. To 5 μ\ of cell homogenate 50 ^1 copper sulphate is added and after 10 min 5 μ\ of Folin reagent is added. After 45 mm in the dark the extinction at 750 nm is read in ultra-microcuvcttes (volume of about 20 μ[ with an optical path of 10 mm) adapted to a Zeiss PMQ 2 spectrophotometer. Different concentrations of bovine albumin in dis­tilled water served as standards.

O.-1,4-glucosidase assays

(1) Maltose as a substrate. The procedure described by Nitowsky and Grunfeld⁴ was adapted for use in small final volumes to reduce the number of cells required for analysis. 10 μ\ cell homogenate was incubated with 10 μ\ substrate (5 mg maltose per ml in 0.05 M acetate buffer pH 4.0) during 30 min, 1 and 2 hours at 37⁰. After heating 2 min at ioo°, cooling in ice and 5 min centrifugation at 3000 rev./ min 10 μ\ of the incubation mixture was added to 50 μ\ glucose oxidase reagent and incubated at 37° during 1 hour. The extinction was read at 440 nm in Zeiss ultramicrocuvettes. Con­centrations up to 200 μg|ml glucose were used as standards.

(2) p-Nitrophenyl ot-D-glucopyranoside was also used as a substrate using various final volumes depending on the amount of cell material to be analyzed. 5 μ\ cell homogenate has been incubated with 25 μ\ 4 mM />-nitrophenyl a-D-glucopyranoside in 0.1 M citrate buffer, pH 4.0. Incubation times varied from 0.5-6 hours at 37° and the reaction was stopped with 30 μ\ 0.2 M NaOH in ice. Extinction at 405 nm was read in different ways depending on the final volume. Down to 60 μ\ extinctions were recorded in microcuvettes adapted to a Zeiss spectrophotometer. Extinction measure­ments in final volumes of 1 μ\ were carried out in glass microcapillaries (Drummond microcaps of 1 mm diameter) using a Leitz micrcspectrophotometer type MPV obj. 2.5 x PI. and an E.M.I., type: 9558 QA photomultiplier with a Knott power supply.

(3) 4-Methylumbeltiferyl tx-v-ghtcopyranoside was used as a substrate in a con­centration of 2.2 mM acetate buffer, pH 4.0. To obtain maximal sensitivity of the α-glucosidase analysis incubation was carried out in microliter volumes⁸. In most experiments incubation of cell homogenate (0.1-1 ^1) or of dissected pieces of plastic foil containing a small number of lyophilized cultured cells was carried out in 0.3-2^1 of substrate during 1-2 hours at 37⁰. To prevent evaporation incubation in such small volumes was performed under paraffin oil using teflon rack with small wells (oil well technique) according to principles described by Matchinsky²⁰ and Lowry¹² (Fig. 4). Cell samples and microdroplets of reagents were introduced into the "oil well" under microscopic control using quartz constriction pipettes which were kindly made for us

by Di. O. H. Lowry. After incubation the reaction volume was pipetted into 500 μ\ 0.5 M carbonate buffer (pH 10.7) and the fluorescence of methylumbelliferone was measured in a Perkin-Elmer fluorometer (extinction 365 nm, fluorescence 448 nm). The fluorescence of smaller volumes was measured in capillaries using the same micro-spectrofluorometer as described, but with a Ploem Opak illuminator²¹.

All experiments on cell homogenates were performed in triplo and in the micro-assays on freeze-dried cells at least 10 pieces of plastic containing a varying number of cells (70-250) were dissected. In the latter experiments empty pieces of plastic dissected adjacent to the lyophillized cells were incubated as blanks. When low cell concentrations are incubated addition of bovine albumin (0.2 mg/nil) is required.

RESULTS

(1) Microchemical analyses on cell homogenates

(a) Maltose as substrate. When the analysis of a-14-glucosidase activity is carried out according to the original procedure of Nitowsky and Grunfeld⁴ about io⁶ cultivated control fibroblasts are required to yield accurate extinction values. For a reliable diagnosis sufficient cell material must be available for triplo measurements, a protein assay and also the sensitivity of the assay should enable a clear distinction between low enzyme activities in cells from heterozygous carriers and deficient cells from patients with Pompe's disease. This means that at least 10' cultured amniotic fluid cells must be available for a prenatal diagnosis when conventional procedures are used.

In this assay, like in each spectrophotometric analysis, the amount of cell material required, decreases proportionally when the final volume of the assay is reduced. We have combined such a reduction ?f the incubation and measuring volume with a simplification of the procedure. 10 ^1 of several dilutions from a homogenate prepared from cultured amniotic fluid cells has been incubated for 2 hours at 37" in ro μ\ maltose substrate. After 2 min heating to ioo° 10 μ\ of the incubation mixture was pipetted into 50 μ\ glucose oxidase reagent and kept for 1 hour at 37°. The ex-

tinction at 440 11111 was read in 20 ji\ microcuvettes lilting to a specially designed cuvet housing for a Zeiss spectrophotometer. Several steps of the original procedure (addition of water, centrifugation, addition of strong acid in the glucose assay step) could be omitted thus enabling more easy analysis in microvolumes.

The results illustrated in Fig. iA show that under those conditions a minimum of about 10 000 normal amniotic fluid cells is required for one single measurement. Incubation of less cells results in unreliable extinction values. As mentioned earlier in prenatal diagnosis sufficient cell material must be available for triplo measurements and a Lowry protein assay which requires a few thousand cells even when performed with a 60 μl final volume. Furthermore low α-glucosidase activities as present in cells from heterozygotes should be detectable and results in Fig. iB illustrate that this will require more than twice the amount of cell material compared to normal controls. The α-glucosidase activity in normal cultured fibroblasts is in the order of 6 -io^⁹ mole-miii^-1mg^_1 protein or 1-2 io^¹³moleh^_1 per cell. As a result of these factors at least 70000 cultured amniotic fluid cells must be available for prenatal diagnosis of Pompe's disease when maltose is used as a substrate even when the assay is carried out in the smallest final volume that can be measured with microcuvcttcs adapted to commercially available spectrophotometer.

Theoretically, further reduction in the number of cells required for analysis could be accomplished by combining longer incubation periods with a further de­crease of the final volume using microscopic spectrophotometry to measure the ex­tinction in submicroliter volumes. As is shown in Fig. iB the enzyme reaction already shows some deviation from linearity during the first 2 hours of incubation and after longer incubation periods no reliable results were obtained. A further reduction in the incubation volume and final volume (down to 1 ,«1) implicate the necessity of in­cubation undci paraffin oil to prevent evaporation and of extinction measurements in microcapillaries using a microspectrophotometer. It was found that the various steps required in the α-glucosidase assay when maltose is used as a substrate, could not easily be combined with the manipulation of submicroliter volumes. For this purpose single step enzyme assays should be preferred.

(b) p-Nitrophenyl-cz-D-ghicopyranoside as substrate. The α-glucosidase assay using />-nitrophenyl-glucoside as a substrate has several advantages when the analysis

is to be carried out on little cell material. Linearity of the enzyme reaction with this artificial substrate was observed even during very long incubation periods (Fig. 2B). Again for this spectrophotometric assay a reduction of the incubation and final volume will enable extinction measurements with less cell material. In Fig. 2A results are illustrated of incubations of decreasing numbers of cultured control fibroblasts using a Go μ\ final volume and an incubation time of 6 hours. A minimum of about 4000 cells is required to yield extinction values of about 0.1. The activity in normal cultured fibroblasts was found to be about 3-4 x less than with maltose as a substrate. In order to be able to perform triplo measurements, a protein assay and to detect low enzyme activities in cells from heterozygous carriers at least 30000 cultured amniotic fluid cells should be available for a prenatal diagnosis of Pompe's disease when j!>-nitro-phenyl-glucoside is used as a substrate in a final volume of 60 p\.

Compared to the assay with maltose as a substrate the single step reaction with />-nitrophenyl-glucoside will more easily enable a further reduction of the final volume. When incubation is performed in microliter volumes the reaction should be performed under paraffin oil to prevent evaporation. The extinction measurements of the ^-nitrophcnol formed can be carried out with a microscope spectrophotometer design. To investigate the reliability of such measurements various dilutions of p-nitropnenol in 0.1 N NaOH were measured in a normal 3 ml cuvet (10 mm optical path) in a Zeiss spectrophotometer and the results were compared with those of 1 μ\ samples measured in rriicrocapillaries using a Leitz microspectrophotometer. From each dilution 10 samples were measured and the results are shown in Table I. A re­markable agreement is observed over the whole range of extinction values. The fact that the 1 μ\ samples show ten fold lower extinction values is the result of the ten fold smaller optical path in the microcapillaries. The results also show that in micro-spectrophotomctry much lower extinction values (about 10 x) can still be measured accurately compared with a normal spectrophotometer. The standard deviation of the measurements is quite small over the whole extinction range. These results indicate that the use of 1 μl final volume enables the analysis of 30-60 fold less cell material than is possible with the smallest microcuvettes adapted to normal spectrophotometers.

TABLE I

ACCURACY OF EXTINCTION MEASUREMENTS OF /J-NTTROPHENOL TN I μ\ VOLUMES USING MICRO-

spectrophotomf.tr v

For the prenatal diagnosis of Pompe's disease this means that a few hundred-thousand cultured cells would be sufficient for α-glucosidase analysis if a final volume of i μ\ is used. However, a protein determination in triplo microvolumes requires at least 3000 cells and hence a total of about 4000 amniotic fluid cells must be available. (c) Methylumbettiferyl a-D-glucopyranoside as substrate. Incubations with this substrate have been carried out at 37° and a substrate concentration of 2.2 mM which is about the maximal solubility. The K_m for cultured fibroblasts was found to be 3.6 mM. Mcthylumbelliferyl-glucopyranoside yields a fluorescent product which can be measured with greater sensitivity than a coloured substance. The minimum amount of methylumbelliferone which could be measured in a 500 μ\ final volume under ex­perimental conditions is about io~¹² mole. However, this does not necessarily mean that this amount of product can be accurately determined in an enzyme assay on cultured cells. Tike in any other fluorometric reaction, the sensitivity which can be attained for a given final volume and fluorometer design mainly depends on the blank value. This latter can be reduced by decreasing the incubation volume as this results in less nonspecific fluorescent material as well as less fluorescence from spontaneously hydrolysed substrate. The influence of a decrease in incubation volume and final volume on the minimum amount of methylumbelliferone that can be detected has been investigated in the following way. Different amounts of methylumbelliferone were dissolved in varying volumes of 2.2 mM iiiethylurnbelliferyl-a-D-glucopyranoside and incubated during 1 hour at 37°. Subsequently the samples were diluted with 0.5 M carbonate buffer to different final volumes. In 5 ml and 500 ^1 volumes the fluores­cence at 448 nm was measured with a conventional fluorometer and in volumes of 50,

10 and 5 μ\ by microfluorometry. The results in Table II indicate that for measure­ments with a conventional fluorometer the incubation volume must be reduced to 0.3-3^1 in order to be able to detect io~ⁿ-io~¹² mole of methylumbelliferone. Even then the difference between the measurement and the blank value is very small. Greater sensitivity can be obtained by also reducing the final volume and thus in­creasing the concentration of the fluorescent product. As shown in Table II the ratio between measurement and blank value improves when the reduction in final volume is combined with a further decrease in incubation volume. By combining in­cubation in 0.05 fi\ of substrate under paraffin oil and measurement of the fluorescent

product in a 5 ,ul final volume using microfluorometry less than io^-13 mole mcthyl-umbellifcrone can be detected.

As the α-glucosidase activity in normal cultured fibroblasts is in the order of io~¹⁴ mole-hour^-1 per cell it can be predicted that only a few tenths-hundreds cul­tured cells are sufficient for a reliable assay provided that the incubation volume is kept small. In Fig. 3 the relation between the amount of methylumbelliferone formed, and the number of cultivated amniotic fluid cells is illustrated. In agreement with the expectation based on the model experiments from Table II the α-glucosidase activity could be detected in as little as 200-300 cells when 1 μi cell homogenate was incubated in 2 μl substrate. After incubation during I hour measurement of methylumbelliferone was performed after dilution to a final volume of 500 fil. The activity found for normal cultured amniotic fluid cells was about 10 fold less than with maltose as a substrate.

Again, for a reliable prenatal diagnosis more cell material will be required to enable triplo measurements and to make a proper distinction between an enzyme deficiency and a decreased α-glucosidase activity as present in cells from heterozygous carriers. As a consequence, using methylumbelliferyl substrate about 1000 amniotic fluid cells are required for a prenatal analysis using microliter volumes for incubation and 500 μ\ final volume. For a parallel protein assay another 3000 cells must be avail­able.

Further increase in the sensitivity of the methylumbelliferyl assay can be obtained by longer incubation periods as it was found that the reaction is linear up to over 8 hours. Another possibility is a further reduction of the incubation volume (down to 0.05 μ\) and measurement of the fluorescence in small final volumes (about 10-50 μ\) using microcapillarics and a microfluorometer design. According to the model experiments in Table II this procedure nearly allows α-glucosidase analysis down to the single cell level. However, in prenatal diagnosis based on studies of cell homogenates such extreme sensitivity of the enzyme assay is irrelevant, because of the necessity of a parallel protein determination which requires at least a few thousand cells. To avoid this disadvantage a method has been developed for a direct α-gluco-

sidasc analysis on isolated groups of freeze-dried cells thus enabling the expression of the enzyme activity per cell instead of per unit weight protein.

(2) Microchemical analysis on microdissected clones of freeze-dried cells

(a) Preparation and microchemical analysis of normal freeze-dried cultured cells. For those enzyme assays with a sufficient sensitivity to allow the analysis of a few tenths-hundreds cells the enzymatic acitvity can in principle be expressed per single cell. This requires the isolation of a counted number of cultivated cells without any detrimental effect on the activity of the enzyme to be analyzed. Methods originally described by Lowry¹¹ for the isolation and microanalysis of tissue structures have been adapted for the investigation of small groups of cultured cells. The procedures followed are described in the paragraph material and methods and Fig. 4 illustrates the main steps.

In vitro cultivation of established amniotic fluid-cell strains or fibroblasts used as controls in prenatal diagnosis should be done in such a way that the cell density is not too large to avoid erroneous cell counting. On the other hand the cell density should not be too small because the minimum number of cells (about 70-120 cells) required for a reliable a glucosidase assay must be present on an area of about 0.2-1 mm² of plastic foil on which the cells have grown. A piece of plastic foil with larger dimensions does not fit into the incubation volume of 0.3-1 μ\, which is necessary for a reliable fluorometric assay. In our hands seeding of about 100000 cells in 4 ml per "mylar

dish" and grown for 2-3 days usually gave the right cell density. In primary cultures of amniotic fluid cells as used in prenatal diagnosis the situation is more favorable as the first clones (after about 7-10 days cultivation) usual consist of closely packed epithelial like and/or fibroblast like cells, which are still sufficiently separated to allow accurate counting.

After freeze-drying the morphology of the cultured cells was found to be well preserved and under a stereomicroscope certain differences in structure, like different cell types within the same culture may well be detected. Free-hand microdissection of 10-15 groups of 100-200 cells can be carried out within an hour and during this period and even up to 2 days there are no changes in α-glucosidase activity in the lyophillized cultured cells.

To investigate if the lysosomal enzyme α-glucosidase is sufficiently liberated into solution and accessible for the substrate, when a piece of foil containing freeze-dried cells is directly incubated without previous homogenization, the following

experiment was designed (Fig. 5). Control amniotic fluid cells were cultivated on plastic foil (for this purpose a culture dish was used, designed by Dr. P. Hc'sli) and after freezing and freeze-drying a strip of foil containing a known number of cells was transferred into 200 μ\ acetate buffer. After 30 min at 20⁰ a 10 μ\ aliquot of the buffer was used for an α-glucosidase assay (using maltose as a substrate) and a 5 /il sample for determination of the protein content. From the two values obtained the specific α-glucosidase activity in the buffer solution was found to be 3.1 X io~⁹ moles/min/mg protein (Fig. 5). This value corresponds remarkably well with the α-glucosidase activity determined in a cell homogenate prepared from the same cell strain at the same time (3.0 x io-'moles/min/mg protein). These data indicate that the interaction between α-glucosidase and substrate after direct incubation of a piece of plastic foil containing freeze-dried cells is comparable to that in a conventional assay on cell homogenate.

• Fluorescence values of 10 empty pieces of plastic ranged from 2-5 with a mean of 3% transmis­sion.

(b) a-Glucosidase activity in dissected freeze-dried cells and comparison with cell homogenates from controls, patients with Pompe's disease and heterozygote carriers. Fibroblasts from a normal control and a patient with Pompe's disease were cultivated on mylar dishes. The α-glucosidase activity was determined by mierochemical analyses of 10-12 pieces of plastic foil containing a counted number (65-170) of cells using methylumbelliferyl-glucoside as a substrate. The results in Table III show that a clear distinction between these two categories of cells is possible using the micro-

TABLE IV

techniques described. For the isolated groups of noimal fibroblasts consistently higher fluorescence values than the blanks (empty pieces of plastic dissected adjacent to the cells) were observed and a mean activity of 3Xio^_14mole-h"¹ per cell was found. In to out of 12 groups of cells from the patient with Pompe's disease no α-glucosidase activity could be detected which means an activity of less than 0.5 X io^_14mole-h^_1 per cell.

In Table IV the α-glucosidase activity in dissected groups of freeze-dried cells is compared with that in cell homogenates from corresponding cell strains. From fibroblasts of controls, patients with Pompe's disease and heterozygote carriers and from control amniotic fluid cells one strain was used for microanalysis of freeze-dried cells in three independent experiments. The results were compared with analytical data from cell homogenates obtained from 4-5 individuals in each category. The results in the first column of Table IV show that there is quite a variation between the experiments within one cell strain and also between the α-glucosidase activity in different groups of cells. It is, however, possible to distinguish the enzyme activities in control fibroblasts and deficient cells whereas fibroblasts from a heterozygote give intermediate values comparable to those in normal amniotic fluid cells. Similar differences are observed in the studies on cell homogenates, shown in the last column of Table IV. For each category of cells, except those from patients with Pompe's disease there is a wide range in the enzyme activity between different individuals.

The analytical data obtained by microchemical analysis of small groups of freeze-dried cells correspond very well with those obtained on cell homogenates when the latter are expressed per cell. This latter value is calculated by supposing that 1 mg protein corresponds to 3 io⁶ cells. This supposition is based on experiments where the cell number after trypsinization has been related to the protein content after homo-genization and also on direct microinterferometric measurements of the dry mass of single cells²². The level of detection in the microchemical analysis of a few hundred freeze-dried cells under the experimental conditions of the experiment in Table IV only allows the conclusion that in cells from patients with Pompe's disease the α-glu-cosidase activity is less than 5 â– io^-15mole-h^_1 per cell. In studies on cell homogenates the exact value of the deficient activity was found to be in the order of 4 io~¹⁶ mole • h^_1 per cell.

The experimental data obtained indicate that the microchemical analysis of α-glucosidase activity in small groups of freeze-dried cultured cells gives reliable results compared with conventional analytical procedures. The method developed thus offers the possibility of rapid prenatal diagnosis of Pompe's disease because the small number of amniotic fluid cells now required for analysis will be available within 7-10 days after amniocentesis.

DISCUSSION

Methods for prenatal diagnosis of chromosomal aberrations and about 30 inborn errors of metabolism have been developed (see for reviews²³'²⁵). The application of these methods gains increasing interest of centers involved in the genetic counseling of families with a high genetic risk. It has been established that the most suitable period for transabdominal amniocentesis is between the I4th-i6tk,week of pregnancy because at this stage sufficient amniotic fluid is available and the number of viable

..1 cells present, in most cases allow a successful in vitro cultivation²⁴-²⁸. The pre-:.il diagnosis of metabolic disorders requires biochemical analysis of the specific -i.zyme defect or the detection of specific accumulated metabolites. Using conven­tional analytical techniques io⁶-io⁸ cultured amniotic fluid cells corresponding to about 0.3-30 mg protein must be available. As a result the cultivation period and hence the waiting period for the parents involved has been 4-6 weeks for most pre­natal diagnoses and in some instances, like in the analysis of cystinosis and muco­polysaccharidoses even up to 8-10 weeks²⁷-²⁸. Prenatal diagnosis of glycogenosis type II (Pompe's disease) required about 5 weeks⁵-⁶. Although parents with a high risk for in affected child are usually strongly motivated to have a prenatal diagnosis, in our opinion it is advantageous to accomplish the analysis before the mother has felt child movements, i.e. before the 18th week of pregnancy.

By adapting prepaiation procedures and microchemical analytical techniques which have mainly been developed for histochemical investigation on tissue sections refs. 11, 12, 14, 20), a considerable reduction in the number of cultured amniotic fluid cells required for prenatal analysis can be obtained⁸-¹⁵-¹⁶.

The use of freeze-dried cultured cells for direct microchemical analysis was found to give results comparable to those from conventional studies on cell homo-genates (Fig. 5 and Table IV). As a consequence a protein determination can be avoided in the prenatal diagnosis of those enzyme defects which can be analyzed in a few hundreds of cultured cells. To obtain such sensitivity both in spectrophotometric and fluorometric reactions a reduction of the incubation volume is required. In spectro­photometry also the final volume has to be reduced and depending on the enzyme activity and the assay used, extinction measurements in volumes of less than 20-50/zl may be required. In such instances microcuvettes adapted to commercial spectro­photometers can no longer be used and a microscope spectrophotometer design has to be applied. Our analyses of />-nitrophenol using microspectrophotometry have shown that a 30-60 fold increase in sensitivity can be reached compared with a 6c μ\ volume using normal spectrophotometry when measurements are carried out in 1 μ\ (Fig. 2 and Table I). In this way a sensitivity could be reached which is comparable to that of a fluorometric assay using a methylumbelliferyl substrate.

Generally fluoiometric procedures will be preferred because in many instances the final measurement can be performed in a normal fluorometer. As far as the use of methylumbelliferyl substrates is concerned the minimum amount of fluorescent product detectable with a normal fluorometer is in the order of io^_11-io^-18 moles, provided that small incubation volumes (0.5-1 μ\) are used (see Table II). If the ac­tivity of the enzyme to be analyzed is in the order of io^-12 to io~¹⁴mole -h^_1 per cell a direct analysis of a few hundred isolated freeze-dried cells according to the procedure described can easily be performed. This was so far found to be the case for about 6 lysosomal storage diseases which can be analysed with methylumbelliferyl sub­strates⁸. For these diseases and for some defects that can only be detected by a (micro)spectrophotometric assay prenatal diagnosis can be accomplished within 7-20 days, depending on the enzyme defect.

However, one of the main problems in prenatal diagnosis of metabolic disorders with an autosomal recessive inheritance is the difficulty of a clear distinction between cultured amniotic fluid cells with an enzyme deficiency and those with a decreased activity from a heterozygous carrier. The few examples presented for Pompe's disease

(Table IV) show that clear distinction between these two categories of cells is possible when the mean activities are compared. But within the wide range of activities for each cell type, very low values may be encountered in cells from a heterozygous carrier. The sensitivity of the assay must therefore be sufficient to detect such low values to avoid erroneous interpretations. The results of the model experiments using microfluorometry (Table II) indicate a possibility to obtain further increase in sen­sitivity of the detection of methylumbelliferone.

By using these microfluorometric techniques, certain lysosomal enzyme activi­ties can even be measured quantitatively in one single cultured cell. This will provide new possibilities for more fundamental studies on metabolic cooperation and genetic complementation at the single cell level^29-31.