|
|
||||||||
Articles |
1
Servicio de Análisis Clínicos, Hospital San Agustín, Aviles 33400 Asturias, Spain.
2
Department of Biochemistry, University of Aviedo, 33006
Asturias, Spain.
3
Washington University School of Medicine, Department of
Pathology, Division of Laboratory Medicine, 660 S. Euclid Ave., St.
Louis, MO 63110.
4
Washington University School of Medicine, Department of
Pathology, 660 S. Euclid, St. Louis, MO 63110.
5
Comprehensive Health, Overland Park, Kansas City, KS.
a Address correspondence to this author at: Servicio de Análisis Clínicos, Hospital San Agustín, c/Camino de Heros, 4 Avilés, 33400 Asturias, Spain. Fax 34 985 123025; e-mail falvarez{at}arrakis.es
| Abstract |
|---|
|
|
|---|
Methods: To generate antibodies, we purified nucleated red blood cells (NRBCs) from fetal livers and used them as the immunogen to generate monoclonal antibodies (mAbs) directed against surface antigens.
Results: The four antibodies recognized at least two conformationally sensitive epitopes of the transferrin receptor. Isolation of NRBCs from 252 maternal blood samples using these antibodies in magnetic activated cell sorting after an initial density gradient centrifugation yielded 0419 NRBCs per 25 mL of maternal blood. One antibody, 2B7.4, not only isolated the highest number of NRBCs (>10 in 90% of the samples) but also isolated these NRBCs in 78 consecutive maternal samples.
Conclusion: Antibody 2B7.4 shows promise for the isolation of NRBCs from maternal blood and should allow studies concerning the source of these cells, fetal vs maternal, and the factors controlling their prevalence.
| Introduction |
|---|
|
|
|---|
,
, and
(5), providing useful markers of fetal cells. Finally,
erythrocytes are unlikely to persist from one pregnancy to the next
(6). However, controversy exists regarding the origin of the
NRBCs in maternal blood: some authors state that most
nucleated erythrocytes in first trimester maternal blood are of maternal
origin (7)(8). The rarity of fetal cells in maternal circulation necessitates several enrichment techniques. A single, double, or triple density gradient is commonly used before separation strategies such as fluorescence-activated cell sorting (FACS) (9)(10)(11)(12), magnetic activated cell sorting (13)(14)(15), immunomagnetic beads (16)(17), antibody-conjugated columns (18), and charge flow separation (19)(20).
The most common strategy to enrich NRBCs uses a monoclonal antibody (mAb) that binds a cell surface antigen present on fetal NRBCs. The most widely used mAbs for fetal NRBC isolation are those directed against the transferrin receptor (CD71) (21)(22), the thrombospondin receptor (CD36) (23), and glycophorin A, a red cell sialoglycoprotein (21)(24).
To improve the purity and yield of fetal NRBCs isolated from maternal blood and to investigate the biology of materno-fetal transfer of cells, we produced and characterized murine mAbs to human fetal liver NRBCs and used some of these antibodies to isolate NRBCs from maternal blood.
| Materials and Methods |
|---|
|
|
|---|
A negative procedure was used to remove leukocytes. Cells (2 x 109) in 2 mL of phosphate-buffered saline containing 5 mmol/L EDTA, 1 g/L human serum albumin, and 0.5 g/L NaN3 (PBS/HSA buffer) were incubated with 50 µL of microbeads coated with antibodies to CD45, CD14, CD15, or CD16 (Myltenyi Biotec) for 30 min at 4 °C. The cell-microbead mixture was washed and passed over a Myltenyi (Myltenyi Biotec) type CS+ magnetic separation column (2 x 108 bound cell capacity) according to the instructions of the manufacturer. The purity of the resulting NRBC preparation was assessed by visually examining cytospins (Shandon cytospin 3) stained with the nuclear stain giemsa and with 3,3'-dimethoxybenzidine, which stains hemoglobin.
development of hybridomas
Mice were immunized by intraperitoneal injections of 3 x
101
to 6 x 101
fetal
NRBCs every 2 weeks for 6 weeks. All mice received an additional
injection of NRBCs 72 h before fusion. Cell fusion was performed
by the method of Waldman (26) with the P3 x 63-Ag8.653
myeloma fusion partner by the Washington University Hybridoma Center
(St. Louis, MO). The hybridomas were screened for binding to cell
surface antigens on fetal NRBCs by FACS analysis. Fetal NRBCs (40 µL)
resuspended in PBS/HSA (1 x 101
cells) were
incubated in 96-well plates with 10 µL of hybridoma supernatant for
30 min at 4 °C with agitation. The positive and negative controls
were antibodies to CD71 and CD45, respectively (Becton Dickinson).
After two washes, the cells were incubated as above with 40 µL of a
1:50 dilution of goat anti-mouse IgG Fab conjugated to fluorescein
isothiocyanate (Sigma Diagnostics). The cells were washed again
and resuspended in 500 µL of PBS for analysis in a FACScan (Becton
Dickinson).
Adsorption of positive hybridomas to adult red blood cells (RBCs), subtypes A and B, eliminated mAbs to mature RBC antigens. We incubated 1 x 108 adult blood cells with 200 µL of hybridoma supernatant for 30 min at 4 °C with gentle agitation. After centrifugation, 150 µL of supernatant was incubated with 1 x 101 fetal NRBCs and screened as described above. The same supernatants without adsorption were analyzed concurrently. Only the hybridomas with unchanged positive fluorescence were considered specific for fetal NRBCs.
Anti-fetal NRBC mAbs were produced in ascites fluid and purified by protein A-agarose chromatography (Genex).
immunohistochemistry
The human T-cell lymphoblastoid cell line CEM, which expresses 5-
to 100-fold more CD71 than normal resting lymphocytes
(27), was purchased from American Type Culture Collection.
White blood cells (WBCs) were isolated on Histopaque gradients
(28) and were activated by incubation with 2 mg/L
Concanavalin A (Sigma Diagnostics) for 48 h at 37 °C and 5%
CO2. Cytospins of cells were fixed in methanol
for 5 min at room temperature. The Vectastain anti-mouse IgG ABC-AP kit
(Vector Laboratories) was used according to manufacturer's
instructions. Binding was visualized with Vector Red alkaline
phosphatase substrate and a hematoxylin counter stain (Sigma
Diagnostics). The slides were viewed on an Olympus B 40x microscope.
mAb COMPETITION STUDIES
mAbs were radioiodinated using Iodogen (Pierce Chemical) according
to manufacturer's recommendations, with a mean specific activity of
4 x 107 cpm/µg (280 Ci/mmol). For
competition studies, 1 x 105 freshly
purified fetal NRBCs were incubated with ~3 x
104 cpm 125I-labeled
mAb (2 x 10-3 to 4 x
10-3 µg) and 10 µg of unlabeled competing
mAb in a total volume of 100 µL PBS containing 10 g/L bovine
serum albumin (BSA) for 2 h at 4 °C. The cells were pelleted at
2000g for 10 min at 4 °C and washed twice by resuspending
in 200 µL of PBS/BSA/azide (PBS containing 10 g/L BSA and 0.5 g/L
NaN3). The final pellet was counted in a gamma
counter (Multiprias; Packard Instruments).
immunopurification of fetal nrbc antigens
Fetal NRBCs at 1 x 1010 to 4 x
1010 cells/L were lysed in 0.01 mol/L
Tris, pH 8.0, 0.14 mol/L NaCl, 2.5 g/L NaN3, 0.1
mmol/L Pefabloc® (Boehringer Mannheim), 5 nmol/L
iodoacetamide, 0.2 kIU/L aprotinin, and 5 mL/L Triton X-100 on ice for
1 h with occasional mixing.
Nuclei were removed by centrifugation at 300g for 10 min at 4 °C, and the resulting supernatant was made 2 g/L sodium deoxycholate. Unsolubilized membranes were removed by centrifugation at 48 000g for 1 h at 4 °C. Antigen was immunopurified from the supernatant as described by Springer (29) except that 0.1-mL columns of immunoabsorbant were used. The mean yield was 50 µg protein per 1 x 109 NRBCs, estimated by the Bradford (30) protein assay (Bio-Rad). Purified antigen was stored at -70 °C until use.
antigen identification
Antigen was digested with a CNBr mixture consisting of 239 µL of
880 mL/L formic acid (Aldrich), 46 µL of 0.35 g/L affinity-purified
antigen, and 15 µL of 5 mol/L CNBr in acetonitrile (Aldrich)
and reacted overnight. Samples were electrophoresed on 420% sodium
dodecyl sulfate-polyacrylamide electrophoresis (SDS-PAGE) minigels
(31) with subsequent electrotransfer to PSQ-grade
polyvinylidene difluoride membranes (Sigma), using a
three-buffer semidry blotting system (32)(ESA); the stained
bands were Edman sequenced (31). The unstained membrane was
analyzed by Western blotting, using Amersham Life Sciences AuroProbe BL
plus gold-labeled goat anti-mouse as the secondary mAb, and bands were
detected using the silver enhancement kit.
enrichment of NRBCs FROM MATERNAL BLOOD
After informed consent was obtained under a protocol approved by
San Agustín Hospital, 25-mL blood samples were collected from
252 pregnant women into Vacutainer Tubes containing EDTA anticoagulant.
The majority of the women were average risk patients participating in
routine prenatal screening recruited from clinical sites throughout the
Principado de Asturias. In all cases, gestational age was confirmed by
sonogram. Because amniocentesis was performed in only a few cases, we
analyzed most of the samples without knowledge of fetal gender.
Mononuclear cells from peripheral maternal blood were isolated on a double histopaque gradient. The washed cells were resuspended in PBS/BSA/azide to a final concentration of 1 x 101 cells per 100 µL and incubated with 3 µg of one of our mAbs for 20 min at 4 °C. The cells were then incubated with magnetic microbeads conjugated to rat anti-mouse IgG antibody (20 µL per 107 cells; Miltenyi Biotech) for 20 min at 4 °C and separated on the Miltenyi miniMACS magnetic column type MS (1 x 107 bound cell capacity) as described by the manufacturer. The NRBCs were enumerated on cytospins.
pcr
Sorted cells were pelleted at 10 000g and snap frozen.
For PCR, the cells were resuspended in 35 µL of water and boiled for
10 min. The following primer sequences were used:
Primers XES14 and XES15 flanked a single copy of Y-specific sequence consisting of 600 bp; primers SRY4 and SRY5 flanked a 80-bp sequence nested within the 600-bp sequence (33).
PCR mixture containing 1.7 mmol/L MgCl2, 0.2 mmol/L each dNTP, 200 ng of each primer (Integrated DNA Technologies), and 1 U of Taq polymerase (Boehringer Mannheim) in 50 µL of the PCR buffer provided by the manufacturer. The first of two amplifications was with primers Xes14 and Xes15; the reaction was denatured at 94 °C for 10 min and then underwent 34 cycles of denaturation at 94 °C for 1 min, annealing at 58 °C for 45 s, and elongation at 72 °C for 1, followed by a final 10 min at 72 °C. The second amplification was performed with 10 µL of the reaction product in fresh reagents and using the primers SRY4 and SRY5; a 2-min denaturation at 94 °C was followed by 34 cycles of denaturation at 94 °C for 20 s, annealing at 58 °C for 45 s, and elongation at 72 °C for 1 min, and then by a final 10 min at 72 °C.
| Results |
|---|
|
|
|---|
|
|
To confirm the identity of the antigen as transferrin receptor, we used
our mAb immobilized on agarose as affinity columns to immunopurify
antigen from fetal liver NRBCs. Two Coomassie-stained bands of
approximately equal intensity were seen that migrate in the 85110 kDa
range under reducing conditions on 420% SDS-PAGE (Fig. 2
, lane 2). A Western blot of a duplicate lane indicated that
only the ~110-kDa band reacted with a commercial anti-human
transferrin receptor mAb H68.4 (Zymed) specific to residues 328 of
the cytoplasmic tail (Fig. 2
, lane 3). Immunoblots with our mAbs
yielded only a very weakly stained ~110-kDa band even after reaction
with gold-labeled secondary mAb and three 45-min rounds of silver
enhancement (data not shown).
|
In a separate experiment, the immunopurified proteins were transferred to polyvinylidene difluoride membranes and then Edman sequenced. No sequence was found for the ~110-kDa species, but the ~85-kDa band gave an 11-residue N-terminal sequence identical to human transferrin. The inability to sequence the 110-kDa antigen suggested a blocked NH2 terminus, thus requiring an internal amino acid sequence (35) for positive identification. Five major bands were discernible after CNBr digestion that corresponded to molecular masses of ~36, 29, 22.5, 17, and 14 kDa (not shown). Sequencing of 912 residues of these fragments allowed identification of the parent protein as the transferrin receptor, which is known to exist as a transmembrane, glycosylated, disulfide-linked homodimer (36). Under our experimental conditions, transferrin copurified with its receptor.
Data on the enrichment of NRBCs from maternal blood using our four mAbs
are presented in Fig. 3
. Blood was obtained from 252 pregnant women at 1518 weeks of
gestation and was processed between 2 and 24 h after collection.
The absolute number of NRBCs isolated from 25 mL of maternal blood
ranged from 0 to 419 with three of the mAbs and from 3 to 336 with mAb
2B7.4. This mAb isolated 310 NRBCs in 10% of the samples, 1150
NRBCs in 60% of samples, and 51336 NRBCs in 30% of the maternal
blood samples. We also found that the median and mean numbers of
isolated NRBC were higher in samples processed between 2 and 6 h
after maternal blood collection and that maternal WBC counts were
higher in older samples (data not shown). In four samples in which the
sex of the fetus was known, the enriched NRBC sample was split into
three aliquots: one for microscopic examination and two for
Y-chromosome PCR. Fig. 4
shows the PCR results for these four samples of enriched NRBCs.
The PCR results confirmed the sex of the fetus established by
amniocentesis.
|
|
| Discussion |
|---|
|
|
|---|
The inconsistent presence of NRBCs in all pregnancies (7)(37)(39)(42) and the low yield reported suggest several questions that remain to be answered before the clinical utility of isolating fetal NRBCs for prenatal genetic diagnosis can be established. These questions involve whether fetal cells are present in all pregnancies and the extent to which the number of fetal cells are dependent on fetal status, gestational time, or maternal complications. To address these questions, we decided to develop reagents that might give better yields of fetal cells.
Our strategy was to generate high avidity, specific mAbs against fetal NRBCs by immunizing mice with fetal NRBCs purified from fetal livers. Our screening protocol selected mAbs that had high affinity for NRBCs but not for mature RBCs. The prominent surface antigen was identified as transferrin receptor (CD71), and competitive binding studies showed that mAbs 2E11.3 and 2B7.4 bind a similar epitope, whereas 4B2.1 and 2F6.3 bind to a different epitope on the transferrin receptor.
The weak reactivity of mAbs 4B2.1 and 2E11.3 in Western blots suggests that the mAbs recognize conformationally dependent epitopes on the receptor that are destabilized under denaturing conditions (43). Because 2E11.3 binds a different epitope than 4B2.1 and 2F6.3, both epitopes are conformationally sensitive.
The NRBC enrichment procedure, consisting of a double density gradient
followed by magnetic activated cell sorting with four of our mAbs, was
applied to 252 maternal blood samples. The mean number of NRBCs
obtained varied with the mAb (Fig. 3
). More NRBCs were isolated when
mAb 2B7.4 was used than the other three mAbs were used; 2E11.3 isolated
the least. The mean number of NRBCs obtained using our enrichment
procedure (0419 per 25 mL of blood) is similar to that (0220 per 30
mL of blood) reported by Oosterwijk et al. (44). However, we
isolated one or more NRBCs in 96% of the samples, whereas Oosterwijk
et al. detected NRBCs in only 66% of their cases. Our NRBC/leukocyte
(purity) ratio was improved compared with these results (44)
and those reported elsewhere
(20)(21)(39)(41). An
important factor in this ratio appeared to be the time between blood
collection and gradient separation, which was also observed by
Oosterwijk et al. (45). The lowest WBC contamination was
observed when the cells were processed within 6 h of collection.
Because our samples came from throughout Asturias, many of them could
not be processed that quickly; however, all were processed within
24 h. mAb 2B7.4 not only isolated the highest number of NRBCs
(>10 cells in 90% of the samples), but it isolated NRBCs in all 78
consecutive maternal blood samples studied, a success rate not
described in the literature previously. Thus, 2B7.4 isolates sufficient
NRBCs for fluorescence in situ hybridization and/or PCR in 90%
of the cases and in the remaining 10% should provide sufficient cells
for PCR analysis. It should be emphasized that the high NRBC yields
with our mAbs were obtained in blood samples from a healthy
population of pregnant women, not from a high-risk population in which
higher yields of NRBCs have been reported (38).
To determine whether any of the NRBCs isolated from maternal blood were
of fetal origin, PCR for the Y and X chromosomes was performed on some
of the samples. In two cases in which a male pregnancy was known, the
PCR correctly identified the Y chromosome, indicating a fetal source
for the NRBCs (Fig. 4
). Subsequent PCR analysis detected the Y
chromosome in 40% of 50 samples, and follow-up information will
provide the gender at delivery (F.V. Alvarez, personal communication).
Further optimization of the enrichment procedure may be possible. For
example, selective lysis of mature RBCs after the density gradient spin
to remove reticulocytes, which may still express some transferrin
receptor, may improve yields with mAbs 4B2.1 and 2F6.3. The
simultaneous use of mAbs to different epitopes on the transferrin
receptor may also increase yield. Work is in progress to identify the
isolated NRBCs as fetal or maternal in origin, using fluorescence in
situ hybridization for the Y chromosome and immunostaining for
or
fetal hemoglobin.
In conclusion, our mAbs, especially 2B7.4, successfully enriched NRBCs from maternal blood samples with low maternal WBC contamination. Thus, the mAbs we have developed against the transferrin receptor on fetal NRBCs possess high specificity for this type of cell and allow the separation of NRBCs from maternal blood, which may be useful in assessing fetal genetic disorders.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
A. Fernandez, B. Prieto, A. Escudero, J. H. Ladenson, and F. V. Alvarez A Monoclonal Antibody with Potential for Aiding Non-invasive Prenatal Diagnosis: Utility in Screening of Pregnant Women at Risk of Preeclampsia J. Histochem. Cytochem., March 1, 2005; 53(3): 345 - 350. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |