Traitement en cours

Veuillez attendre...



Aller à Demande


Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]


All documents cited herein are incorporated by reference in their entirety.

This invention is in the field of non-invasive prenatal diagnosis, relying on the analysis of fetal cells which circulate in maternal blood.

Prenatal diagnosis of genetic disease traditionally requires invasive procedures, such as amniocentesis or chorionic villus sampling. These complex procedures require skilled personnel and are associated with a risk of inducing miscarriage. In addition, amniocentesis is typically performed after around 16 weeks of gestation, meaning that any termination following the test results will be in the middle trimester. Termination at this stage of pregnancy is associated with psychological stress and physical risk to the mother.

Methods for non-invasive prenatal diagnosis have been proposed, based on fetal material that escapes into maternal tissue. There are currently three sources of fetal genetic material:

(i) trophoblast cells exfoliated into the cervix have been isolated. Unfortunately, fetal cells and/or DNA can be isolated in <50% of pregnancies [Overton et al. (1996) Am. J. Obstet, Gynecol 175:382-7];
(ii) fetal cell-free DNA has been detected in maternal serum samples [Lo et al (1998) Am.J. Hum. Genet. 62:768-75]. Fetal genotyping is limited, however, to diagnosis of rare monogenic conditions for which the mother is not a carrier [e.g. Amicucci et al. (2000) Clin. Chem. 46:301-2];
(iii) fetal cells that escape into maternal blood circulation by feto-maternal haemorrhage have been extensively investigated for non-invasive prenatal diagnosis [e.g. international patent applications WO90/06509, WO91/07660 & WO95/26507; US patents 5275933, 5639669, 5716776 & 5861253; Bianchi (1999) Brit. J. Haematol. 105:574-583].

Approach (iii) is hindered by many factors, including the lack of truly fetal-specific markers. Even in aneuploidy, which predisposes to feto-maternal cell trafficking, clinical trials have revealed only 41% sensitivity [Bianchi et al. (1999) Prenat.Diagn. 19:994-95].

An alternative to locating fetal cells in a maternal blood sample is to expand the fetal cell population within the sample, exploiting differences in growth factor requirements and clonogenic assay kinetics. Fetal CD34+ve stem cells, for instance, are more sensitive to cytokines and possess higher proliferative potential than maternal cells [Migliaccio et al. (1996) J.Hematotherap. 5:161-70].

Studies to date, however, have shown only very modest expansions of fetal cells in maternal blood in liquid culture [e.g. Little et al. (1997) Blooά 89:2347-58]. Even where an optimal cytokine combination is used, there may be only a doubling of fetal cells amidst an expanded maternal cell population.

It is an object of the invention to provide methods for the selective expansion of fetal cells within a maternal blood sample. It is a further object of the invention to utilise these fetal cells in prenatal diagnosis.

The invention is based on the surprising discovery that certain CD45"ve fetal cells without a maternal equivalent can be purified and expanded from maternal blood in the first trimester, thereby providing abundant fetal material for genetic testing by a non-invasive route.

The invention provides a process for preparing cells, comprising the steps of (a) extracting a blood sample from a pregnant female, and (b) enriching nucleated CD45"ve fetal cells therein.

The blood sample
Any blood sample from a pregnant female (preferably a human) can be used according to the invention. To maximise the number of CD45"ve fetal cells, it is preferred that the blood sample should be taken during the first trimester (weeks 0-13) of pregnancy. More preferably the blood sample is taken during weeks 5-13, and most preferable weeks 7-13.

Typical samples will be venous blood, with a volume between 1 and 50ml (e.g. 20-30ml).

The CD45-ve fetal cells
The internationally standardised CD (cluster of differentiation) marker nomenclature is well known to the skilled person [International Worlshops on Human Leukocyte Differentiation Antigens]. Methods for the assay and determination of CD antigens are commonplace.

The invention utilises fetal cells from maternal circulation that are CD45"ve i.e. they do not express CD45, a marker which is characteristic of leukocytes and their precursors. The cells will typically also have one or more of the following immunophenotypes:

Preferred fetal cells for use with the invention are thus CD14"ve, CD34"ve, CD68've, vWF"ve, myoglobin ve, collagen-rve, cytokeratin ve, CD44+ve, CD105+ve, CD106+ve, fibronectin+ve, prolyl-4-hydroxylase+ve, α-smooth muscle actin+ve, and laminin+ve, as well as being CD45"ve.

Cells for use according to the invention preferably display the morphological characteristics of stromal cells, and/or can support the growth of CD34+ haematopoietic cells. More preferably, they have mutilineage potential and have the ability to differentiate into bone, fat, and cartilage. They may also have the ability to differentiate into neurons. Most preferably, they are multipotent mesenchymal stem cells (MSC).

Advantageously, the fetal cells are adherent, and present in fetal circulation.

Most preferably, the fetal cells are those which can be obtained from first trimester fetal blood samples by the following method: (i) extract 20-300μl fetal blood by ultrasound-guided cardiac aspiration between 7-13 weeks at clinically-indicated surgical termination of pregnancy; (ii) heparinise the sample immediately; (iii) plate the blood sample at a density of 105"6 cells/well; (iv) culture the cells in Dulbecco's Modified Eagle Medium (DMEM) at 37°C with 5% CO2 and 10% fetal bovine serum; (v) after 3 days, remove non-adherent cells and replace medium; (v) culture cells to confluence (around seven days); and (vi) trypsinise the adherent cell layers.

These cells are particularly suitable for non-invasive prenatal diagnosis because:
(a) the cell type is not present in adult blood [Lazarus et al. (1997) J.Hemotother. 6:447- 55], the only exceptions being after bone marrow transplant or chemotherapy [Fernandez et al. (1997) Bone Marrow Transpl. 20:265-67];
(b) their immunophenotype is not normally encountered in maternal blood cells, facilitating enrichment strategies;
(c) their expandability is surprisingly high;
(d) they are readily adherent, which is surprising for circulating cells;
(e) they can be readily identified amidst contaminating adherent maternal cells based on morphology, immunophenotype and growth pattern;
(f) they are replatable and clonal, being able to yield a large pure population of fetal cells;

(g) persistence from previous pregnancies is unlikely, because blood is not the natural home for MSC. Indeed, they do not persist in fetal circulation beyond 13 weeks.

Cell enrichment
Fetal cells for use according to the invention can be enriched from a maternal blood sample by various methods. These can generally be classified into two groups: (1) methods which separate fetal cells from maternal cells, typically based on differences in immunophenotype; and (2) methods which culture cells under conditions that favour the expansion of fetal cells compared with maternal cells, typically based on differences in functional properties. Either or both of these approaches can be used in the processes of the invention.

These methods permit even a single fetal MSC in a 20-30ml blood sample to be expanded into a pure source of fetal cells - 1 fetal stromal cell in 105 nucleated maternal cells, preferably 1 fetal stromal cell in 10 nucleated maternal cells, can be separated to yield a 100% pure population of fetal cells. This offers significant benefits compared with enrichment strategies previously used for fetal nucleated red blood cells in maternal blood, even though only a small proportion of fetal cells escaping into maternal circulation are CD45"ve mesenchymal cells.

The enrichment step in the process of the invention may initially include a step for separating mononuclear cells e.g. using density gradient centrifugation.

Advantageously, the fetal cells of the invention are adherent i.e. they can adhere to plastic in cell culture. This is surprising for circulating cells and facilitates their separation.

Methods which separate fetal cells from maternal cells
These methods typically employ the difference in immunophenotype between maternal cells and fetal MSCs. For instance, fetal MSCs are CD45"ve whereas the vast majority of adult adherent cells are CD45+ve. Other cell markers which may be used in addition or as alternatives for distinguishing maternal and fetal cells include, but are not limited to, CD 105 (endoglin), CD44 (H-CAM), CD29 (integrin βl), CD 106 (VCAM-1) and CD 120a (TNF receptor 1). SH2 antibodies (Osiris, Baltimore USA; see also WO92/22584) may also be used.

The separation step may not completely remove maternal cells. If this is the case, the process of the invention may include more than one step in which maternal and fetal cells are separated e.g. a step based on CD45 followed by a step using SH2 antibodies.
Preferred methods which utilise these immunophenotype differences to separate fetal and maternal cells are magnetically activated cell sorting (MACS) and fluorescent activated cell sorting (FACS). FACS advantageously allows multiparameter sorting and better sensitivity, whereas MACS is technically less complex and thus more suited to clinical situations.

After separation of fetal and maternal cells, the process of the invention may include a step of adherent cell culture. In particular, where cells are separated on the basis of CD45 expression, the process advantageously includes adherent cell culture followed by trypsinisation and re-plating. Adherent maternal cells are relatively resistant to trypsinisation, and the detached cells are therefore substantially of fetal origin.

Adherent cell culture may also be used itself to separate fetal and maternal cells. The separation step in the process of the invention may therefore begin with adherent cell culture.

Selective culture conditions
Maternal and fetal cells can behave differently in cell culture, and these differences can be exploited to expand fetal cells in a sample in preference to maternal cells (e.g. WO95/26507). The CD45've cells used in the present invention show an unexpectedly high ability to expand.

MSCs have different growth factor requirements compared with maternal blood cells. MSC expansion is favoured by growth factors such as bFGF, PDGF, EGF and IGF. These may be used singly or in combination in selective growth media.

The culture may be serum-containing, low-serum (e.g. 2-3% serum), or serum-free. A serum-free medium which favours MSC growth is disclosed in US patent 5908782, and comprises a minimal essential medium (e.g. MEM, DMEM, IMDM, IMEM etc.), serum albumin, an iron source (e.g. transferrin, FeSO4, ferritin), insulin or an IGF, glutamine and, optionally, a mitogen.

MSCs and maternal cells also display different abilities to differentiate, and thus media which induce mesengenic differentiation can be used to favour the expansion of fetal MSCs compared with maternal cells. Fetal MSCs can differentiate into adipocytes, osteocytes and chondrocytes. They may also differentiate into neurons. A suitable adipogenic medium that can be used according to the invention comprises hydrocortisone, isobutyl methylxanthine and indomethacin. A suitable osteogenic medium comprises dexamethasone, ascorbic acid and β glycerol phosphate. A suitable chondrogenic medium comprises transforming growth factors β3, dexamethasone, ascorbic acid, pyruvic acid, proline, insulin, transferrin, selenium, linoleic acid and BSA. A suitable neuronogenic medium comprises dimethylsulfoxide (DMSO) and butylated hydroxyanisole (BHA).

Testing for genetic disorders
The process of the invention results in cells which are ideally suited for genetic testing.

Thus the invention provides a process for prenatal testing of a fetus for a genetic disorder, comprising the steps of (a) extracting a blood sample from a pregnant female, (b) enriching nucleated CD45"ve fetal cells therein, and (c) testing the enriched nucleated CD45"ve fetal cells for a genetic disorder.

Any disorder which is known to have a genetic basis may be tested, including disorders with a chromosomal abnormality [e.g. Down's Syndrome (trisomy 21), Turner's Syndrome (XO chromosomes), Klinefelter's Syndrome (XXY), Edward's Syndrome (trisomy 18), Patau Syndrome (trisomy 13)] and single-gene disorders [e.g. cystic fibrosis, alpha and beta thalassaemia, haemophilia, muscular dystrophy, myotonic dystrophy, sickle cell disease, Huntington disease etc.].

The test may be at a chromosomal level, or may require more detailed genetic analysis (e.g. nucleic acid hybridisation, RFLP detection, PCR, or sequencing). The sex of the fetus can also conveniently be determined.

Figure 1 shows: (A) adherent cells seen on early fetal blood film; (B) a monolayer of these cells formed in liquid culture; (C) Leishman staining after liquid culture; (D) fibronectin staining after liquid culture.

Figure 2 shows fetal MSCs grown in medium that favours (A) osteogenesis and (B) adipogenesis. In 2A, the cells are stained for calcium with silver nitrate after 7 days. In 2B, the cells are stained for lipid using Oil-Red-O after 3 days.

Figure 3 shows maternal and maternal adherent cells. 3A and 3C show fetal cells; 3B and 3D show maternal cells. 3 A and 3B show liquid culture; 3C and 3D show trypsinised cells. Figure 4 shows simultaneous immunophenotyping and XY FISH.

Isolation of stromal cells in fetal blood
Blood samples were taken from fetuses in first, second and third trimesters of pregnancy:

First trimester (median gestational age: 10+0, range 7+6 - 14+0 ; n=34) samples were collected under ultrasound guidance using a siliconised 20 GA 15 cm needle (COOK) from fetal heart under general anaesthesia before clinically-indicated termination of pregnancy. Both the syringe and the needle were heparinised before use (1000 U/ml). Fetal blood samples (average lOOμl) were diluted 1:9 with DMEM for a nucleated cell count using a haemocytometer chamber. Cells from blood samples <11 weeks gestation were cytospun onto cleaned slides and a differential cell count performed following Leishman' s staining. Blood films from samples >11 weeks were stained directly and differential cell counts performed.

Second trimester (median gestation age: 15+0, range 14+5 - 23+7; n=6) samples were collected either before embryo-reduction by cardiocentesis or during clinically-indicated termination. Cell count was performed using an automated cell count. A blood film was stained with Leishman' s stain and the differential cell count performed.

Third trimester (median gestation age: 39+1, range 38+0 - 40+2; n=5) samples were obtained from cord at delivery from uncomplicated pregnancies; all the newborns were healthy with birth weights appropriate for gestational age and had no evidence of birth asphyxia or congenital abnormalities.

In first trimester blood, the mean nucleated cell concentration was 73+7.3x109, of which 93.8±0.4% were erythroblasts, 3.7±0.4% were blasts and 2.3+0.3% were neutrophils. All first trimester samples from 7-13 weeks contained a small proportion (0.1 to 1.0%) of large mononuclear cells with the morphological appearance of stromal cells (figure 1), not present in second and third trimester samples.

These fetal cells may be frozen prior to use, as a freeze thaw cycle does not affect their growth or characteristics.

Culture of stromal cells
The stromal cells detected in the first trimester blood samples were cultured as follows:

Blood nucleated cells were diluted to 2.5x105 cells/ml with 10% FBS in DMEM with 50units/ml penicillin, 50μg/ml streptomycin and 2mM L-glutamine, plated in a well of a 6 well-plate and incubated at 37°C with 5% humidified CO2. After 72 hours, non-adherent cells were removed and the medium was replaced every 4 days until the cells grown to 80% confluence. They were harvested with 0.25% trypsin and ImM EDTA for 5 minutes at 37°C. The cells were re-plated in a 75 cm2 flask, grown to confluence and harvested. The cells were frozen in 10% DMSO and 30% FBS and 30% FBS and stored in liquid nitrogen. To expand the cells through successive passages, they were plated at 15000 cells per cm2, grown to near confluence and harvested with the same protocol. At the end of each passage, the cells were counted on a hemocytometer to calculate cell doublings.

When cultured in 10% FBS, the stromal cells rapidly (<5 days) formed a monolayer of confluent adherent fibroblast-like cells (figure 1). This culture can be maintained in culture over serial passages for at least six months.

Figure 3 shows the difference between fetal (A&C) and maternal (B&D) adherent cells. In liquid culture (A&B) fetal MSCs (A) form a fibroblast-like monolayer of elongated cells, in contrast to maternal adherent cells (B) which only occasionally elongate and do not form stroma. After trypsinisation (C&D), the fetal cells (C) retain their morphology, with large irregular nuclei, vacuoles and pseudopodia, clearly distinct from the maternal cells (D).

Adherent cultured cells were tested using immunocytochemistry as follows:

Cytospin preparations (3x104 adherent cells following trypsinisation deposited onto a glass slide by centrifugation in a cytocentrifuge) or growing cells in double-chamber slides were fixed in equal volume of methanol and acetone for 1 minute at room temperature and then washed in TBST (Tris buffered saline containing Tween). The slides were then incubated with 3% H2O2 for 10 minutes followed by 10 minute incubation with 10%) normal goat or rabbit serum at room temperature. Washed slides were incubated with 50μl of the optimal concentration of diluted primary antibody (DAKO antibody diluent: 0.05M Tris-HCl containing 0.1% Tween and 15mM sodium azide) for 30 minutes at room temperature in a humidified chamber. The following antibodies were used: mouse anti-human CD45 mAb (clone T29/33; DAKO), mouse anti-human CD34 Class II mAb (clone QBEND10, DAKO), mouse anti-human CD68 mAb (clone PG-M1, DAKO), peroxidase-conjugated rabbit anti-human von WiUebrand factor polyclonal Ab (DAKO), mouse anti-human prolyl-4-hydroxylase mAb (clone 5B5, DAKO), mouse anti-skin-fibroblasts (Serotec), rabbit anti-human myoglobin mAb (clone A0324, DAKO), mouse anti -α-smooth muscle actin mAb (clone 1A4, SIGMA) and mouse anti-human CD 106 mAb (clone BB1G-V1, SIGMA). After 3 washes with TBST, all cytospins except those stained with peroxidase-conjugated rabbit anti -human vWF polyclonal Ab, were incubated with a 1 :100 dilution of the peroxidase-conjugated secondary antibody (anti-mouse or anti-rabbit IgG, SIGMA) for 30 minutes at room temperature in a humidified chamber. After 3 washes with TBST, 3, 3'-diaminobenzidine in chromogen solution (DAKO) was applied to the slides for 20 minutes at room temperature. The slides were counter-stained with 0.1% Mayer's haematoxylin solution and mounted with aqueous permanent mounting medium.

Extracellular matrix proteins were studied as follows: adherent cells growing on a double-chamber glass were fixed in situ with 1 :1 methanol: acetone, washed in TBST and labelled with each of the following antibodies: mouse anti-human fibronectin mAb (clone IST-4, SIGMA), mouse anti-collagen type I mAb (clone COL-1, SIGMA), mouse anti-collagen type II mAb (SIGMA), and mouse anti-laminin (SIGMA).

FACS analysis was performed as follows: first trimester adherent cells at the 2nd passage in culture were trypsinised and stained with anti-CD34-FITC, CD45-FITC, HLA-DR-PE, CD14-FITC, CD105-PE, CD44-FITC mAbs. Cell suspensions containing ~106 cells in 50 μl PBS with 1% BSA were first incubated with 50μl of 10% normal mouse serum and then incubated with 10-20μl of the corresponding antibodies for 30 minutes at 4°C. Non-specific isotype-matched antibodies were used to determine background fluorescence. Following two washings with PBS containing 1% bovine serum albumin and 0.1 % sodium azide, cells were analysed on a FACScan flow cytometer (Becton Dickinson) and data acquisition was performed with CELLquest software. Scatter gates were set to exclude debris, and 50,000 gated events were analysed.

The fetal cells are neither haematopoietic (CD34"ve, CD45"ve, CD14"ve, CD68"ve) nor endothelial (vWF"ve, CD34"ve), but are myofibroblastic (prolyl-4-hydroxylase+ve, fibronectin+ve, smooth muscle actinweak+ve) and mesenchymal (CD105+ve). Their immunophenotype and growth pattern are identical to stromal cells isolated from first trimester fetal liver and second trimester bone marrow.

Ability to maintain haematopoietic cells
The ability of the fetal cells to support haematopoietic cells was studied as follows:

Following irradiation (8000 cGy), first trimester blood adherent cells were seeded at 2.0x105 per well of a 12-well-plate previously treated with 0.1% collagen. 24 hours later, 5xl04 cord blood CD34+ cells, isolated by staining with anti-CD34 antibodies (QBEND/10 mouse IgG, Miltenyi) conjugated with microbeads and elution through MiniMACS columns, were resuspended in long term culture medium (Myelocult, Stem Cell Technologies) supplemented with 10-6M hydrocortisone and seeded over the stromal cells in the absence of exogenous growth factors. The co-cultures were incubated at 37°C with 5% CO2 for 6 weeks. At weekly intervals, non-adherent cells in the collected half volume of the culture medium were counted, stained with Leishman' s stain and assayed for colony forming cells (CFCs) by colony assays in standard methylcellulose culture (Methocult, Stem Cell Technologies), and the medium was replaced by fresh medium. Control experiments were also performed by culturing cord blood CD34+ve cells either in the absence of a stromal layer and growth factors or over a murine stromal cell line (M210-B4).

The fetal cells can support the growth of haematopoietic cells, with co-culture leading to a <40-fold expansion of committed haematopoietic progenitors.

Differentiation of the fetal cells
The ability of the cells to differentiate into adipocytes, osteoclasts, cartilage and neurons was tested:

Adipogenic differentiation. Fetal adherent cells were grown to 90% confluence in double-chamber slides and then incubated with DMEM with 10% FBS supplemented with 0.5 μM hydrocortisone, 0.5 mM isobutylmehylxanthine and 60 μM indomethacin. The medium was replaced every 4 days. After 2-3 weeks cells were washed with PBS and fixed in 10%) formalin for 10 min. Cells were stained for 10 minutes with fresh Oil-red-O solution (SIGMA). The Oil-red-O solution was prepared by mixing 3 parts stock solution (0.5% in isopropanolol) with 2 parts water. Plates were washed three times with water. The percentage of adipocytes was assayed by counting 50-100 cells in multiple fields.

Osteogenic differentiation. Fetal adherent cells were grown to 90% confluence in double-chamber slides and then incubated with DMEM with 10%) FBS supplemented with 10-8 M dexamethasone, 0.2mM ascorbic acid and lOmM beta-glycerophosphate. The medium was replaced every 4 days. To assess mineralisation, cultures were washed with PBS and stained with silver nitrate and/or Alizarin red. Cultures were incubated for 1 hour with 2% silver nitrate and exposed to 100W bulb. They were then washed in water and fixed (30 seconds). Cultures were stained for 5 minutes with 2% Alizarin red (pH 4.2). Cultures were rinsed in acetone (30 seconds), treated with acetone-xylene 1 :1 (15 seconds). Mineralisation was assayed by examination of multiple fields for the area of mineralisation as a percent of the total area of the confluent cultures.

Chondrogenic differentiation. Cells were resuspended in serum-free chondrogenic medium. This medium consisted of DMEM with 1 g/1 glucose supplemented with 10 ng/ml transforming growth factor-β (TGF-β3), lOOnM dexamethasone, 50μg/ml ascorbic acid, lOOμg/ml sodium pyruvate, 40μg/ml proline and ITS-plus (Collaborative Biomedical Products; final concentrations; 6.25μg/ml bovine insulin, 6.25μg/ml transferrin, 6.25μg/ml selenous acid, 5.33μg/ml linoleic acid and 1.25mg/ml bovine serum albumin). Aliquots of 250,000 cells suspended in 0.5ml were distributed to 15ml conical tubes. The cells were centrifuged for 5 min at 600g; tubes were placed at 37°C with 5% humidified CO2. The sedimented cells formed a spherical mass at the bottom of the tube within 24 hours. Medium was replaced three times per weeks. After 3 weeks cell pellets were harvested by rinsing in PBS followed by fixation for 1 hour in 4% formaldehyde. Samples were then transferred into 70% ethanol, dehydrated in ethanol and xylene series, and paraffin-embedded. Antibodies specific for extracellular matrix were used to assess chondrogenic differentiation. Sections were treated with 3% H2O2 in methanol (10 minutes), then digested for 30 minutes with 50μU/ml chondroitinase ABC (SIGMA) in lOOmM Tris-acetate pH 7.6, 0.1%) BSA, then incubated with mouse anti-collagen type I and type II mAbs (SIGMA) overnight at 37°C. Immunoreactivity was detected by incubating sections with goat anti-mouse IgG or IgM conjugated with horseradish peroxidase. Anionic sulphated proteoglycans were detected by Safranin O staining and toluidine blue metachromasia.

The fetal cells were shown to be able to differentiate into bone, fat and cartilage (figure 2). Neuronogenic differentiation. Fetal blood MSCs were isolated from a 9+5 week blood sample, based on their ability to expand in 10% FBS and adhere to plastic tissue culture dishes. At passage 2 in culture, 5xl04 MSCs were plated into a chamber of a 4-chamber slide and maintained in DMEM/10% FBS. To induce a neuronal phenotype, the cells were incubated with serum-containing preincubation media (DMEM, 10%o FBS, 10 ng/ml bFGF). To effect neuronal differentiation, the cells were washed with PBS and incubated with the neuronal induction serum-free media (DMEM, 2%» DMSO, 200 mM BHA) for 5 hours [cf. Woodbury et al. (2000) J Neuroscience Research 61:364-370]. Within 2 hours of exposure to DMEM/DMSO/BHA, morphology changes were apparent in ~30% of the MSCs. Responsive cells progressively assumed neuronal morphological characteristics over the following two hours. Initially, cytoplasm in the flat fetal MSC retracted towards the nucleus, forming a contracted multipolar cell body with peripheral extensions. Over the subsequent 3 hours, cell bodies became increasingly spherical and refractile and processes continued to form, displaying primary and secondary branches. To characterise neuronal differentiation further DMEM/DMSO/BHA treated cells were fixed after 5 hours and stained with a neuronal marker, the neurofilament-M (NF-M) antibody. Cells that exhibited contracted cell bodies and processes (-60%) stained positive for NF-M. Clusters of differentiated cells exhibited intense NF-M positivity and processes formed extensive networks.

Mesenchymal ability of cells
The fetal cells isolated in first trimester maternal blood display the characteristics of adult marrow-derived mesenchymal stem cells [Pittenger et al. (1999) Science 284:143-47]. The fetal cells are therefore defined herein as multipotent mesenchymal stem cells (MSC). They are believed to be produced in the fetal aorta-gonad-mesonephros, and migrate in the first trimester fetal blood to definitive sites of haematopoiesis and multi-organ mesenchymal tissue formation.

Separation of fetal cells from maternal cells
Fetal MSCs are CD45"ve, whereas the vast majority of adherent adult cells are CD45+ve. This distinction was used as the basis of cell separation. Starting with a mixture of fetal MSCs and gender-mismatched adult blood, mononuclear cells were separated by single density gradient (1077 g/ml) centrifugation. The separated mononuclear cells were resuspended in buffer and incubated for 15 minutes at 4°C with murine anti-human CD45 antibody directly conjugated to magnetic beads. This selectively labelled maternal cells, and fetal MSCs were negatively selected by a single MACS passage.

XY FISH can be used to distinguish by gender between maternal and fetal cells after 15 days. Other analyses include morphological evaluation after Leishman stain, immunophenotyping with CD45, and growth to confluence following replating. Figure 4 shows that fetal cells are CD45"ve XY (4A) whereas maternal cells are CD45+ve XX (4B).

The CD45"ve population was plated in duplicate onto a glass chamber in 10% FBS (105/chamber). Non-adherent cells are removed at 3 days, and the medium replaced every 3-4 days. The CD45+ve fraction was used as a control under the same conditions.

Whilst CD45-based MACS gave good separation, maternal CD45+ve cells often persisted in adherent culture (15-40%). In contrast to the fetal MSCs, however, these were relatively resistant to trypsinisation, so collection of detached cells after trypsinisation yields primarily fetal MSCs. Figures 4C and 4D shows that only CD45"ve XY fetal cells expand after trypsinisation and re-plating.

To increase separation specificity, and in order to eliminate any possible inhibition of fetal MSC growth by the maternal cells, alternative separation methods can be employed.

To improve MACS performance, alternative markers can be used. Endoglin (CD 105) is considered to be MSC-specific [Pittenger et al. (1999) Science 284:143-47], and endoglin-specific antibodies are widely available. FACS can be used as an alternative to MACS.

Culture conditions selective for fetal cell expansion
Maternal and fetal cells have different growth requirements in cell culture (e.g. different growth factor requirements). These can be exploited to expand fetal cells in a sample in preference to maternal cells. To favour expansion of fetal MSC over background maternal cells, optimal growth factor requirements for expanding undifferentiated fetal MSCs can be determined. Cytokines shown to favour adult MSC expansion (e.g. basic FGF, PDGF, EGF, IGF) can be added singly and in combination to fetal MSCs in low serum (2-3%) or serum free media, and expansion can be compared at 7 and 14 days. Optimal combination(s) can be evaluated in model mixtures. To establish the MSC-selectivity of these media, MSCs purified from fetal blood can be diluted down to 1 in 107 with maternal cells, and the ability to eliminate maternal cells can be evaluated. It has been routinely possible to amplify 1 fetal stromal cell in 105 nucleated maternal cells to yield a 100%) pure population of fetal cells (figure 4). Amplification of 1 fetal stromal cell in 106 nucleated maternal cells to yield a 100% pure population of fetal cells has been achieved using specific depletion columns ("LD" columns, Miltenyl Biotech GmbH, Germany, which were used in the same way a standard MACS columns in accordance with the manufacturer's instructions) in the magnetic activated cell sorting process, which allows retention of one order of magnitude greater numbers of positive cells within the system.

Culture conditions selective for fetal cell differentiation
MSCs and maternal cells also display different abilities to differentiate. Media which induce MSC differentiation can also be used to select MSCs. The expandability of fetal MSC cultured for 7-14 days in media known to induce their differentiation into adipocytes, osteocytes or chondrocytes can be evaluated.

Testing pregnant females
To assess the sensitivity of the process of the invention, venous samples (20ml) are collected from consenting women with viable first trimester singleton pregnancies undergoing clinically-scheduled vacuum curettage between 8-13 weeks. It has been reported that maternal blood collected shortly after suction termination of pregnancy contains approximately 300 times the number of fetal nucleated cells as from intact pregnancies [Bianchi et al. (1998) Am. J. Hum. Genet. 63(suppl.):A6], making this an ideal model system. Blood is drawn five minutes after the conclusion of the procedure, and processed without delay. Ideally, the fetus will be male, to facilitate the ability to distinguish fetal and maternal cells.

Maternal blood samples (20-30ml) are also taken from consenting women who are not undergoing termination. The enriched fetal cells can be compared, after birth, to definitive fetal genotype.

It will be understood that the invention has been described by way of example only and modifications may be made whilst remaining within the scope and spirit of the invention.

Note: blood collection for research purposes was approved by the Research Ethics Committee of Hammersmith and Queen's Hospital. The use of fetal tissues for research purposes complied with national guidelines (Polkinghome). All pregnant women gave signed consent for surgical procedure, fetal tissue sampling and percutaneous fetal blood sampling.