Printer Friendly

[zeta]-, [epsilon], and [gamma]-globin mRNA in blood samples and [CD71.sup.+] cell fractions from fetuses and from pregnant and nonpregnant women, with special attention to identification of fetal erythroblasts.

The appearance, duration of presence, shifts, and amounts of embryonic and fetal hemoglobin forms during fetal life and in adult life have been studied intensively. Far less studied are the specific mRNAs for these different forms of hemoglobin. Specific mRNA transcripts for [gamma]-, [theta]-, [zeta]-, and to some degree [epsilon]-globin have been detected in adults (1-3), but the number of studies is few, and to our knowledge, investigations of the above forms and comparisons between pregnant and nonpregnant women have not been performed.

In the embryo, the earliest globin chains are the [zeta] and [epsilon] chains, and [[zeta].sub.2][[epsilon].sub.2] is the major hemoglobin until 5-6 weeks of gestation. The [[alpha].sub.2][[epsilon].sub.2] form has been found at 4 weeks but is absent after 13 weeks of gestation. As the liver replaces the yolk sac as the main site of erythropoiesis, synthesis of the [zeta] and [epsilon] chains decreases and that of the [alpha] and [gamma] chains increases. Although hemoglobin F ([[alpha].sub.2][[gamma].sub.2]) is present in very young embryos, it is not the major hemoglobin of fetal life until 10-12 weeks of gestation (4-7).

The possibility of isolating fetal nucleated red blood cells from maternal blood samples for use in antenatal diagnosis of fetal genetic or chromosomal abnormality has gained much attention in recent years (8, 9). Although several studies have been somewhat successful, further development of this approach of prenatal diagnosis into a broader clinical setting has been hampered by the lack of fetal-specific markers for use in cell isolation and identification procedures. Embryonic and fetal hemoglobin forms are obvious candidates for such markers, and monoclonal antibodies against some of these forms have already been used (10, 11). However, the use of specific globin mRNA molecules as targets for probes in cell identification procedures will require further studies to clarify the possibility of low transcription of these mRNAs in adults, which might be great enough to interfere with probe hybridization detection procedures.

Therefore, information about these embryonic and fetal hemoglobin mRNAs, especially in pregnant and nonpregnant women, is important from the perspective of using these molecules as markers of fetal erythroblasts. To investigate whether these embryonic and fetal globin transcripts are present in nonpregnant or pregnant women, we analyzed whole-blood samples and [CD71.sup.+] cell fractions with a sensitive reverse transcription-PCR (RT-PCR) method and with quantitative analysis of [gamma]-globin mRNA based on competitive RT-PCR. The CD71 receptor for transferrin is highly expressed on fetal erythroblasts, and monoclonal antibodies against CD71 have been used for cell sorting of these cells in several studies (9,12).

The amounts of the mRNA for a possible specific fetal marker sequence might be important when using probe hybridization technology (e.g., peptide nucleic acid probes) to identify fetal cells. If the number of specific mRNA copies in adult cells is less than a certain threshold limit, their presence may not lead to positive results in a detection assay.

Materials and Methods


Informed consent was obtained from the subjects, and the study was approved by the local Research and Ethics Committee. Whole-blood samples, anticoagulated with EDTA (adult samples) or heparin (fetal samples), were collected from fetuses between 17 and 20 weeks of gestation that had been aborted for genetic indications and from the peripheral blood of healthy, pregnant women (9-13 weeks of gestation) and from healthy women who, to their knowledge, had never been pregnant. After collection the samples were kept on ice until further processing ( [less than or equal to]3 h). The concentration of human choriogonadotropin measured in the samples from women who claimed never to have been pregnant was in all cases <5 IU/L and therefore negative.


[CD71.sup.+] cells were isolated from 1 mL of whole blood using Dynabeads M-450 CD71 (Dynal) according to the manufacturer's instructions. The [CD71.sup.+] cells were then lysed with 500 [micro]L of Lysis/Binding Buffer from the Dynabeads mRNA DIRECT reagent set. [Poly(A).sup.+] RNA from [CD71.sup.+] cells was extracted using the Dynabeads mRNA DIRECT reagent set according to the manufacturer's instructions; 50 [micro]L (250 [micro]g) of [Oligo(dT).sub.25] Dynabeads was used.


Whole blood (1 mL) was centrifuged in a benchtop centrifuge at maximum speed. The plasma phase was discarded, and the pellet was lysed with 500 [micro]L of Lysis/Binding Buffer. The [poly(A).sup.+] RNA was extracted as described above, except that 250 [micro]L (1250 [micro]g) of [Oligo(dT).sub.25] Dynabeads was used per sample. The concentration of the extracted [poly(A).sup.+] RNA was calculated from absorbance measurements at 260 and 280 nm.


The presence of [epsilon]-, [zeta]-, or [gamma]-transcripts in the [poly(A).sup.+] RNA preparations of the whole-blood samples was analyzed by three individual, specific RT-PCR reactions. The PCR step was performed in duplicate. Both steps included negative controls with no template in the reaction mixture. Reverse transcription was performed as follows (per reaction): 1.0 [micro]L of Oligo-dT primer (500 ng/[micro]L; Life Technologies), 200 ng of [poly(A).sup.+] RNA, 4.0 [micro]L of 5 X First Strand Buffer (Life Technologies), 2.0 [micro]L of 0.1 M dithiothreitol (Life Technologies), 2.0 [micro]L of 8.0 mM dNTPs (2.0 mM each dNTP; Amersham Pharmacia Biotech), 1.0 [micro]L of Superscript II RNase [H.sup.-] Reverse Transcriptase (200 U/[micro]L; Life Technologies), and nuclease-free water (Promega) in a total volume of 20 [micro]L. The mixture was incubated at 42[degrees]C for 55 min and then inactivated at 70[degrees]C for 15 min.

PCR was performed as follows (per reaction): 5.0 [micro]L of Mg[Cl.sub.2] (25 mM; MBI Fermentas), 5.0 [micro]L of 10 X PCR buffer (MBI Fermentas), 1.0 [micro]L of Taq polymerase (1 U/[micro]L; MBI Fermentas), 5.0 [micro]L of 8.0 mM dNTPs (2.0 mM each dNTP), 2.5 [micro]L of forward primer (10 [micro]M; TAGC), 2.5 [micro]L of reverse primer (10 [micro]M; TAGC), 1.0 [micro]L of cDNA from the reverse transcription reaction above, and nuclease-free water (Promega) in a volume of 50 [micro]L. Thermocycling conditions were as follows: 94[degrees]C for 3 min; 40 cycles of 94[degrees]C for 1 min, 65[degrees]C for 1 min, and 72[degrees]C for 1 min; and 72[degrees]C for 5 min, followed by holding at 4[degrees]C. PCR primers were as follows:

For [epsilon]-globin: forward, 5'-TTT TAC TGC TGA GGA GAA GGC TGC C-3'; reverse, 5'-CTT GCC AAA GTG AGT AGC CAG AAT AA-3'

For [gamma]-globin: forward, 5'-ACG CCA TGG GTC ATT TCA CAG A-3'; reverse, 5'-GAG CTC AGT GGT ATC TGG AGG A-3'

For [zeta]-globin: forward, 5'-CCC TGC CGC CAT GTC TCT GAC CAA G-3'; reverse, 5'-AGG CGG CGT GGG CCT CGG CC-3'

The RT-PCR products were analyzed by electrophoresis on a 2% agarose gel. Representative RT-PCR products were cloned and sequenced. Cloning was performed by using the TOPO TA Cloning reagent set with pCR-2.1 vector (Invitrogen) according to the manufacturer's instructions. Sequencing was performed with the Thermo Sequenase dye-primer 7-deaza cycle seq reagent set (Amersham Pharmacia Biotech) according to the manufacturer's instructions. The sequencing primers were as follows: [gamma], 5'-Cy5 ACT CAG CTG GGC AAA GGT GCC-3'; [epsilon], 5'-Cy5 ACT CAG CTT AGC AAA GGC GGG-3'; [zeta], 5'-Cy5 AGT TGC GCG CGC ACG GCT CC-3'.


The coamplification analysis was based on the simultaneous amplification of [gamma]- and [epsilon]-cDNA (Fig. 1). The same set of primers could be used for both templates because of the extensive sequence homology between [gamma]- and [epsilon]-globin. One primer was labeled with Cy5, which was incorporated into the PCR products during amplification. Products of similar length were produced by this amplification but were cut to different lengths with sequence-specific restriction enzymes and then analyzed on a denaturing polyacrylamide gel, where the ratios between the areas under the peaks corresponding to the restriction products were determined. To verify the validity of the method and to establish its detection window, we used different mixtures of [epsilon] and [gamma] sequences cloned into pCR2.1-TOPO vectors as templates for coamplification. The ratios in these mixtures were 1:1-1:100.


For one set of experiments, K562 cells (from the human chronic myelogenous leukemia cell line), stimulated with hemin chloride (Sigma-Aldrich) dissolved in dimethyl sulfoxide (25 [micro]mol/L hemin chloride was included in the culture medium for 48 h) to induce the expression of globin mRNAs, was used as starting material. [Poly(A).sup.+] RNA from K562 cells was extracted using the Dynabeads mRNA DIRECT reagent set according to the manufacturer's instructions. cDNA was synthesized as described above. The coamplification RT-PCR was performed using the reaction mixture described and thermocycling conditions of 94[degrees]C for 3 min; 30 cycles of 94[degrees]C for 1 min, 65[degrees]C for 1 min, and 72[degrees]C for 1 min; and 72[degrees]C for 5 min, followed by holding at 4[degrees]C. PCR primers for [gamma]- and [epsilon]-globin mRNA were as follows: forward, 5'-Cy5 TGG IGC AAG ITG AAT GTG GGA A-3' (where I = inosine); reverse, 5'-GCT TGA AGT TCT CAG GAT CCA CA-3'. This RT-PCR created two amplification products of similar lengths but different sequences. These products were cut to different lengths by restriction endonucleases specific for each RT-PCR product. The reaction mixture consisted of 1.6 [micro]L of 1O X NE2 buffer (New England Biolabs), 0.8 [micro]L of nuclease-free water (Promega), 1.6 [micro]L of 10X bovine serum albumin (New England Biolabs),1.1 [micro]L of MseI (4 U/[micro]L; New England Biolabs), 0.6 [micro]L of PvuII (5 U/[micro]L; New England Biolabs), and 10 [micro]L of the PCR products from the PCR reaction described above. After incubation at 37[degrees]C for 90 min, the restriction fragments were separated on a denaturing polyacrylamide gel (Amersham Pharmacia Biotech) using the ALF Express Sequencer (Amersham Pharmacia Biotech) with a run temperature of 40[degrees]C and a run time of 300 min. The results were analyzed by Allele Links software (Amersham Pharmacia Biotech).


The amount of a specific mRNA transcript in a sample can be determined by the use of competitive RT-PCR analysis. We used a modification of the method of Diviacco et al. (13) to design a [gamma]-globin DNA internal standard for the analysis. In addition to inserting an artificial linker sequence in the internal standard as devised by Diviacco et al., we incorporated 50 by of intron 1 in the genomic sequence of [gamma]-globin to increase the difference in length between the natural transcript and the internal standard. The insertion of an intron sequence into the DNA sequence of the internal standard was achieved by a combination of two separate PCR reactions, A and B. A third PCR joined the products of reactions A and B and generated the internal standard. The following reaction conditions were similar for reactions A and B: 5.0 [micro]L of Mg[Cl.sub.2] (25 mM; MBI Fermentas), 5.0 [micro]L of 10 X PCR buffer (MBI Fermentas), 1.0 [micro]L of Taq polymerase (1 U/[micro]L; MBI Fermentas), 5.0 [micro]L of 8.0 mM dNTPs (2.0 mM each dNTP), 7.5 [micro]L of forward primer (10 [micro]M; TAGC), and 2.5 [micro]L of reverse primer (10 [micro]M; TAGC). The template for reaction A was 2.0 [micro]L of human genomic DNA (100 ng/[micro]L), whereas that for reaction B was 1.0 [micro]L of K562 cDNA. For both reactions nuclease-free water (Promega) was added to give a volume of 50 [micro]L. Thermocycling conditions were as follows: 94[degrees]C for 3 min; 2 cycles of 94[degrees]C for 1 min, 50[degrees]C for 1 min, and 72[degrees]C for 1 min; 33 cycles of 94[degrees]C for 1 min, 65[degrees]C for 1 min, and 72[degrees]C for 1 min; and 72[degrees]C for 5 min, followed by holding at 4[degrees]C. The primers were as follows:

Forward primer A, 5'-ACG CCA TGG GTC ATT TCA CAG A-3'




The PCR products were analyzed on a 1% agarose gel. Separate pipette tips were dipped into the band for each product, A and B. The tips were incubated in 50 [micro]L of sterile water for 2 min at room temperature. Five micro-liters of this mixture was used as template with the following PCR reaction mixture: 10.0 [micro]L of Mg[Cl.sub.2] (25 mM; MBI Fermentas), 10.0 [micro]L of 10 X PCR buffer (MBI Fermentas), 2.0 [micro]L of Taq polymerase (1 U/[micro]L; MBI Fermentas), 10.0 [micro]L of 8.0 mM dNTPs (2.0 mM each dNTP), 5.0 [micro]L of forward primer (10 [micro]M; TAGC), 5.0 [micro]L of reverse primer (10 [micro]M; TAGC), and nuclease-free water (Promega) to give a volume of 100 [micro]L. Thermocycling conditions were as follows: 94[degrees]C for 1 min, 90[degrees]C for 1 min, 85[degrees]C for 1 min, 80[degrees]C for 1 min, 75[degrees]C for 1 min, 65[degrees]C for 1 min, 60[degrees]C for 1 min, 55[degrees]C for 1 min, 50[degrees]C for 2 min, and 72[degrees]C for 1 min, followed by 30 cycles of 94[degrees]C for 1 min, 60[degrees]C for 1 min, and 72[degrees]C for 1 min, and finally 72[degrees]C for 5 min. The PCR product was analyzed by gel electrophoresis on a 1% agarose gel. The PCR product was then cloned using the TOPO TA Cloning reagent set with pCR-2.1 vector (Invitrogen) according to the manufacturer's instructions. The plasmid carrying the internal standard sequence was used as a template (diluted 1:5000) in a PCR reaction with primers for the [gamma]-globin sequence, as described previously. Before quantification by absorbance measurements, the PCR product was cleaned up using the spin column technique with Sephacryl 5-400 HR (Amersham Pharmacia Biotech). The PCR product was henceforth used as a DNA internal standard.

Before subjecting the samples to competitive RT-PCR analysis, we had to establish an interval in which the amount of amplification product from the sample transcript roughly equaled the amount of amplification product of the added internal standard. This was accomplished by amplifying serial dilutions of the internal standard in a fixed amount of sample cDNA. The actual competitive RT-PCR analysis was subsequently carried out at the concentration at which the amounts of the internal standard and the sample products were equal and at the concentrations immediately bracketing this value. The reaction mixture for the competitive RT-PCR was similar to that described for qualitative RT-PCR except for addition of the standard template. The cycling conditions were also similar, except that the amplification steps were repeated for 35 cycles instead of 40. It was also necessary to verify that the amplification kinetics for the two templates were the same. To do so, we set up a 60-[micro]L PCR reaction using nearly equal amounts of sample transcript and internal standard as templates. Aliquots (10 [micro]L) were removed from the reaction mixture after 20, 25, 30, 35, and 40 cycles and analyzed to determine whether the original ratio between the two templates was conserved through the course of the reaction, as would be expected.


The number of samples positive for a specific globin mRNA in the groups of pregnant and nonpregnant women was compared by Fisher's exact test. Differences in the amount of [gamma]-globin mRNA in samples from pregnant and nonpregnant women were compared by the Mann-Whitney test.



Using the described method, we were able to construct a [gamma]-globin DNA internal standard (453 bp) for use in competitive RT-PCR vs the sample [gamma]-globin RT-PCR product (523 bp), obtained using the same set of primers and having considerable sequence similarity, which is a requirement for use in competitive RT-PCR. The sequence of the DNA internal standard fragment was verified by DNA sequencing and included both the intron sequence and the linker sequence. The sequence of the sample [gamma]-globin mRNA RT-PCR product was also verified by sequencing.


cDNA from the mRNA was coamplified with a nearly equal amount of the corresponding internal standard over a range of PCR amplification cycles (20-40). The concentration of (RT)-PCR products was determined by densitometry. In Fig. 2 the amount of (RT)-PCR product is plotted as a function of the cycle number, and the ratio (internal standard:sample) from the densitometry results is plotted against the cycle number. As shown, the internal standard and the sample sequence were amplified with the same efficiency for all numbers of amplification cycles. The CV of the ratios of the areas of two (RT)-PCR product bands in an agarose gel as measured by densitometry was 9%. The CV of the whole procedure from isolation of [CD71.sup.+] cells to quantification of [gamma]-globin mRNA by the competitive RT-PCR, determined with six replicates, was 17%. We consider this an acceptable CV, given that the procedure is divided into several parts, each containing many steps.


To verify the validity of the method and to establish its detection window, we performed coamplification analysis on different ratios of the artificially mixed templates of cloned [epsilon]- and [gamma]-globin sequences, and the electropherograms correctly reflected the ratios of the mixture down to a ratio of 1:25 ([epsilon]:[gamma] or [gamma]:[epsilon]); below this only the dominant template was detected. Use of RT-PCR coamplification analysis showed that the relative proportion of [gamma]- and [epsilon]-globin mRNA was 3:1 in the induced K562 cells. An example of an electropherogram of the PCR fragments is shown in Fig. 3. The quantitative [gamma]-globin mRNA RT-PCR procedure indicated that the amount of [gamma]-globin mRNA in induced K562 cells was ~1200 copies/cell (or 920 pg/[micro]g of total mRNA). An example of [gamma]-globin mRNA internal standard titrated in a fixed amount of cDNA derived from K562 cells followed by PCR amplification is shown in Fig. 4, together with a plot of the data. Given that the proportion of [epsilon]- to [gamma]-transcripts is 1:3, the amount of [epsilon]-globin mRNA in the induced K562 cells must be ~400 copies per cell.


[gamma]-, [zeta]-, AND [epsilon]-mRNA IN FETAL BLOOD SAMPLES AND FETAL [CD71.sup.+] CELL FRACTIONS

The qualitative RT-PCR results are listed in Table 1. In all fetal samples (17-20 weeks of gestation), [gamma]-, [epsilon]-, and [zeta]-mRNA could be detected in the [CD71.sup.+] cell fraction; in some whole-blood samples, however, [epsilon]- and [zeta]-mRNA could not be detected. The quantitative [gamma]-globin RT-PCR analysis showed that the median amount of [gamma]-globin mRNA in fetal [CD71.sup.+] cells (at 17-20 weeks of gestation) was 151 ng/[micro]g of total mRNA (range, 87-810 ng/[micro]g of total mRNA; n = 5). This corresponded to 1.6 X [10.sup.11] copies/[micro]g of total mRNA. Positive results for RT-PCR analyses or the quantities of [gamma]-globin mRNA showed no correlation with the number of weeks of gestation in the narrow gestational interval investigated. [gamma]-Globin mRNA should correspond to 15% of the total mRNA pool in the fetal [CD71.sup.+] cells. Furthermore, the coamplification RTPCR assay was not useful for coamplifing [gamma]- and [epsilon]-globin mRNA transcripts in the [CD71.sup.+] cells from the fetal blood samples because expression of [epsilon]-globin mRNA has ceased or decreased to a very low amount by 17 weeks of gestation.




All adult whole-blood samples were negative for [epsilon]- and [zeta]-globin mRNA (Table 1). Analyses of [CD71.sup.+] cell fractions showed that 19 of 20 nonpregnant and 10 of 14 pregnant women were positive for [gamma]-globin mRNA (Fisher's exact test, P = 0.13), and 3 of 20 nonpregnant and 5 of 14 pregnant women were positive for [zeta]-globin mRNA (Fisher's exact test, P = 0.23); any difference between the pregnant and the nonpregnant women was not significant. The amount of [gamma]-globin mRNA in [CD71.sup.+] cells from pregnant and nonpregnant women was in the same range: median, 85 pg/[micro]g of total mRNA (range, 26-1185 pg/[micro]g of total mRNA; n = 7) for pregnant women vs 161 pg/[micro]g of total mRNA (42-790 pg/[micro]g of total mRNA; n = 6; P = 0.37, Mann-Whitney test) for nonpregnant women. This corresponds to a median of 9.0 X [10.sup.7] copies/[micro]g of total mRNA for nonpregnant women and 1.7 X [10.sup.8] copies/[micro]g of total mRNA for pregnant women.


The occurrence of embryonic and fetal hemoglobin has been studied extensively in adult blood. In contrast, the expression of these hemoglobins has not been investigated to the same extent at the transcription level. Background transcription of the globin mRNAs may exist in cells that do not express the proteins. One study previously identified low amounts of [zeta]- and [epsilon]-globin mRNA in the reticulocytes of healthy adults and patients with sickle cell disease, chronic myelogenous leukemia, and polycythemia vera (1).

One approach to identifying fetal erythroblasts in the maternal blood for use in prenatal diagnosis could be based on a crude preselection of, e.g., [CD71.sup.+] cells in a maternal blood sample, followed by identification of the fetal cells with a hybridization probe directed against the mRNA of an embryonic or fetal globin sequence. A highly specific probe based on recognition of a nucleic acid sequence could be more specifically compared with antibodies directed against protein epitopes.

In this study, we investigated the presence of [epsilon]- and [zeta]-globin mRNAs and the amounts of [gamma]-globin mRNA in whole-blood samples and [CD71.sup.+] cell fractions from non-pregnant women, pregnant women (at 9-13 weeks of gestation), and fetal blood (at 17-20 weeks of gestation). We had no access to fetal samples from the first trimester of pregnancy.

We undertook the examination of a possible background transcription of fetal hemoglobin mRNAs in adult blood cells. The transferrin receptor CD71 was chosen as a marker for the preselection because it is highly expressed on nucleated red cells in early fetal blood (12). Having examined whole-blood samples and the [CD71.sup.+] fractions of these samples with RT-PCR, we found that a preselection step is necessary to increase the prevalence of erythroblasts in a sample to permit detection of the embryonic and fetal mRNAs. All adult whole-blood samples were negative for [epsilon]- and [zeta]-globin mRNA. Perhaps isolating the mRNA from 10-20 mL of whole blood and using more RT-PCR amplifications for each sample might increase the possibility of obtaining positive results. One adult (nonpregnant) whole-blood sample was positive for [gamma]-globin mRNA. All adult samples were negative for [zeta]-globin mRNA, whereas the fetal samples were positive. Albitar et al. (1) previously identified low amounts of [zeta]-and [epsilon]-globin mRNA in the reticulocytes of patients with sickle cell disease, chronic myelogenous leukemia, and polycythemia vera. They also detected [zeta]- and [epsilon]-globin mRNA by radioactively labeling RT-PCR products from reticulocyte RNA from four apparently healthy adults. Only trace amounts of [epsilon]-globin mRNA were detected in these samples after several days of exposure. This is consistent with our own findings, although we did not detect [zeta]- or [epsilon]-globin mRNA in all blood samples. Perhaps the content of these hemoglobin mRNAs in the heterogeneous cell fractions falls below the detection limit of the RT-PCR assay, or perhaps our method of detection (ethidium bromide-stained RT-PCR products on an agarose gel) is less sensitive than with RT-PCR radioactive labeling. Or perhaps no [zeta]- or [epsilon]-globin mRNA was present in the particular samples we examined.

In our search for a globin mRNA combination specific for fetal erythroblasts, we designed a coamplification RT-PCR assay for [gamma]- and [epsilon]-globin mRNA to determine whether the ratio of [gamma]- to [epsilon]-globin mRNA could be used as a positive identification of a fetal cell. The assay was initially tested on a model system of stimulated K562 cells or artificial DNA samples with different known ratios of [gamma]- to [epsilon]-globin and [epsilon]- to [gamma]-globin sequences. Analyses of the artificial DNA samples showed that the assay detected ratios between 1:1 and 1:25. Analysis of the stimulated K562 cells showed that in these cells the ratio of [gamma]- to [epsilon]-globin mRNA is 3:1. Subsequent analysis of both fetal whole blood and [CD71.sup.+] cell fractions did not detect any [epsilon]-globin mRNA. This indicates that the production of E hemoglobin at gestation week 17 has either ceased or decreased to a point where the ratio of [gamma]- to [epsilon]-globin mRNA exceeds 25:1. This is consistent with an immunocytochemical study of the expression of E and y hemoglobin in erythroblasts from fetal blood at 10 weeks of gestation (11). That study showed that 1-5% of the erythroblasts express only E hemoglobin at that stage, whereas 1-3% express both E and y hemoglobin. Furthermore, [epsilon]-positive blasts were seen in fetal liver tissue at 14 weeks of gestation at a frequency of 2-3%, and blasts positive for both phenotypes were seen in ~1%. This expression of [epsilon]-globin will have decreased even further by gestation week 17 and makes our results, which indicated a ratio of [gamma]- to [epsilon]-globin mRNA >25:1, very probable. The results from the immunocytochemical studies, together with our own results, also help explain why no [epsilon]-globin mRNA was detected in the samples from pregnant women at 9-13 weeks of gestation. Studies indicate that only 1-10 fetal cells can be detected in 1 mL of maternal blood (14,15). If some of these fetal cells contain only a little (low percentage) [epsilon]-mRNA/hemoglobin, then the probability of detecting [epsilon]-mRNA in such a sample would be very low.

Although we are aware of the possible limitations of the present study (RT-PCR amplification based on 1 mL of blood), the lack of detection of [epsilon]-globin mRNA in pregnant (9-13 weeks of gestation) and nonpregnant women leads us to speculate that [epsilon]-globin mRNA could be a candidate marker for fetal erythroblasts in a background of maternal blood cells at 6-8 weeks of gestation. Questions remain regarding whether fetal erythroblasts typically are present in maternal blood, and in what numbers they occur at this early stage of pregnancy.

The coamplification assay described here can be used in the analysis of [gamma]- and [epsilon]-globin mRNA ratios in very early fetal blood (weeks 6-8) with special attention to identifying fetal cells based on globin mRNA ratios or monitoring the switch from embryonic to fetal globin mRNAs. Furthermore, the assay might be used to identify fetal erythroblast colonies in in vitro cell cultures of cell populations isolated from maternal blood samples. These in vitro cell culture systems are a possible way to increase the number of fetal cells from maternal blood, as has been described by some investigators (16,17).

To investigate the possible differences between the amounts of [gamma]-globin mRNA in the blood cells of pregnant women and those of nonpregnant women, we designed a quantitative RT-PCR assay with a DNA internal standard. Essentially, we used the method described by Diviacco et al. (13) and evaluated the assay as described by Hayward et al. (18). We initially tested the assay with a definite number of stimulated K562 cells and found that the amount of [gamma]-globin mRNA in these was ~1200 copies/ cell. Accordingly, the number of [epsilon]-globin mRNA copies must be ~400. In subsequent tests of the [CD71.sup.+] cell fractions of whole blood from pregnant and nonpregnant women, we found that the [gamma]-globin mRNA was present in [CD71.sup.+] cells from nonpregnant women at a median of 85 pg/[micro]g of total mRNA, whereas the median in pregnant women was 161 pg/[micro]g of total mRNA. However, the median value in pregnant women, although higher, was not significantly different from that of nonpregnant women (P = 0.37). This result seems reasonable considering that the samples were from women only in the first trimester of pregnancy. We think that it is possible and rather likely that some of the [gamma]-globin mRNA found in the blood samples from the pregnant women originates from fetal cells. From the results in this study, however, whether any of the [zeta]-globin mRNA in the samples from the pregnant women at 9-13 weeks of gestation originates from fetal cells cannot be determined. Moreover, the [gamma]-globin mRNA in samples from pregnant and nonpregnant women might not be directly comparable because they represent the amount of [gamma]-globin mRNA in a [CD71.sup.+] cell fraction, not that in whole blood. The [CD71.sup.+] cell isolation procedure was used as a means of enriching for cells expressing the globin mRNAs of interest and an attempt to obtain a pure cell fraction. Pregnant women have a greater total white cell count (which in an activated state may express CD71) in the peripheral blood than do nonpregnant women [see, e.g., Ref. (19)]; therefore, the number of [CD71.sup.+] erythroblasts isolated from pregnant women might be relatively fewer than the number isolated from nonpregnant women because the erythroblasts compete with the activated lymphocytes for the antibodies to CD71. Thus, the possibility exists that the peripheral blood cells of pregnant women may contain embryonic and fetal globin mRNAs in greater numbers than stated above.

Because the [CD71.sup.+] cell fraction is a heterogeneous cell population, we chose to quantify the [gamma]-mRNA globin per total mRNA rather than per cell. Some cells might contain rather high amounts of [gamma]-mRNA and some very little. Accordingly, for determining the specific amount of [gamma]-mRNA in single [CD71.sup.+] cells or in cells positive for [gamma]-hemoglobin protein from nonpregnant and pregnant women, laser microdissection methods must be used. The methods described here could form the basis for such a study, which could be important in designing a detection assay for fetal erythroblasts based on [gamma]-globin mRNA contents in single cells as described below. Furthermore, RT-PCR-based quantification of [epsilon]-globin mRNA in single fetal erythroblasts at 6-8 weeks of gestation would be relevant for use of this marker with in situ hybridization techniques. Our study indicates that [epsilon]-globin mRNA is not present (or is present only in a very low amount) in maternal erythroblasts. Therefore, as stated previously, [epsilon]-globin mRNA might be a useful marker of fetal erythroblasts in early pregnancy.

Quantification of [zeta]-globin mRNA in pregnant and nonpregnant women may seem imminently possible, but it was not performed in this study. Because [zeta]-mRNA can be detected in both nonpregnant and pregnant women, and probably not in high amounts in fetal erythroblasts after 10 weeks of gestation, [zeta]-mRNA would never be a completely specific fetal marker; indeed, it is probably present in substantially greater amounts in fetal erythroblasts than in maternal erythroblasts only during a few weeks in very early pregnancy.

Examination of the [gamma]-globin mRNA content of the [CD71.sup.+] fraction of fetal blood by the quantitative RT-PCR method showed a median of 151 ng/[micro]g of total mRNA. Again, this number may not be directly comparable to the numbers for the women because of a possible difference in the populations of isolated [CD71.sup.+] cells; nonetheless, the [gamma]-globin mRNA obviously is expressed to a considerably greater extent in the fetal cell fraction than in the adult fractions. Therefore, as the results in this work indicate, further studies on a homogeneous population of fetal or adult erythroblasts that produce [gamma]-globin mRNA may reveal whether the difference in the amount of [gamma]-globin mRNA produced by fetal and adult cells can be used as a marker for identifying or isolating a fetal cell. The amount of [gamma]-globin mRNA in adult cells might very well be below the threshold value for a positive signal in, e.g., an in situ probe hybridization detection system, such that only fetal nucleated red blood cells would show a positive signal.

In conclusion, using RT-PCR methods and 1-mL blood samples, we found that [gamma]- and [zeta]-globin, but not [epsilon]-globin mRNA, can be detected in the [CD71.sup.+] fractions of peripheral whole blood from pregnant (end of first trimester) or nonpregnant women. [CD71.sup.+] fractions of fetal blood (at gestation weeks 17-20) express all three mRNA forms. On the basis of our results, we speculate that [epsilon]-globin mRNA might be useful as a positive identification marker of fetal nucleated red blood cells in a background of maternal blood cells in early pregnancy.

This work was supported by The Danish Agency for Trade and Industry. T.V.F.H. was supported by grants from the Copenhagen Hospital Corporation, Danish Foundation for the Advancement of Medical Science, and the Danish Disability Foundation. We thank Lone Nielsen and Conny Andersen for excellent technical assistance.


(1.) Albitar M, Peschle C, Liebhaber SA. Theta, zeta, and epsilon globin messenger RNAs are expressed in adults. Blood 1989;74:62937.

(2.) Albitar M, Care A, Peschle C, Liebhaber SA. Developmental switching of messenger RNA expression from the human [alpha]-globin cluster:

fetal/adult pattern of [theta]-globin gene expression. Blood 1992;80:1586-91.

(3.) Leonova JY, Kazanetz EG, Smetanina NS, Adekile AD, Efremov GD, Huisman TH. Variability in the fetal hemoglobin level of the normal adult. Am J Hematol 1996;53:59-65.

(4.) Gale RE, Clegg JB, Huehns ER. Human embryonic haemoglobins Gower 1 and Gower 2. Nature 1979;280:162-4.

(5.) Kamuzora H, Lehmann H. Human embryonic haemoglobins including a comparison by homology of the human [zeta] and [alpha] chains. Nature 1975;256:511-3.

(6.) Hecht F, Motulsky AG, Lemire RJ, Shepard TE. Predominance of hemoglobin Gower 1 in early human embryonic development. Science 1966;152:91-2.

(7.) Huehns ER, Dance N, Beaven GH, Hecht F, Motulsky AG. Human embryonic hemoglobins. Cold Spring Harbor Symp Quant Biol 1964;29:327-31.

(8.) Elias S, Price J, Dockter M, Wachtel S, Tharapel A, Simpson JL, Klinger KW. First trimester prenatal diagnosis of trisomy 21 in fetal cells from maternal blood. Lancet 1992;340:1033.

(9.) Bianchi DW. Fetal cells in the maternal circulation: feasibility for prenatal diagnosis. Br J Haematol 1999;105:574-83.

(10.) Cheung MC, Goldberg JD, Kan YW. Prenatal diagnosis of sickle cell anaemia and thalassaemia by analysis of fetal cells in maternal blood. Nat Genet 1996;14:264-8.

(11.) Mesker WE, Ouwerkerk-van Velzen MC, Oosterwijk JC, Berini LF, Golbus MS, Kanhai HH, et al. Two-colour immunocytochemical staining of gamma ([gama]) and epsilon ([epsilon]) type haemoglobin in fetal red cells. Prenat Diagn 1998;18:1131-7.

(12.) Zheng YL, Zhen DK, DeMaria MA, Berry SM, Wapner RJ, Evans MI, et al. Search for the optimal fetal cell antibody: results of immunophenotyping studies using flow cytometry. Hum Genet 1997;100:35-42.

(13.) Diviacco S, Norio P, Zentilin L, Menzo S, Clementi M, Biamonti G, et al. A novel procedure for quantitative polymerase chain reaction by coamplification of competitive templates. Gene 1992;122: 313-20.

(14.) Hamada H, Arinami T, Kubo T, Hamaguchi H, Iwasaki H. Fetal nucleated cells in maternal blood: frequency and relationship to gestational age. Hum Genet 1993;91:427-32.

(15.) Bianchi DW, Williams JM, Sullivan LM, Hanson FW, Klinger KW, Shuber AP. PCR quantitation of fetal cells in maternal blood in normal and aneuploid pregnancies. Am J Hum Genet 1997;61: 822-9.

(16.) Bohmer RM, Zhen D, Bianchi DW. Differential development of fetal and adult haemoglobin profiles in colony culture: isolation of fetal nucleated red cells by two-colour fluorescence labelling. Br J Haematol 1998;103:351-60.

(17.) Lo YM, Morey AL, Wainscoat JS, Fleming KA. Culture of fetal erythroid cells from maternal peripheral blood. Lancet 1994;344: 264-5.

(18.) Hayward AL, Oefner PJ, Kainer DB, Hinojos CA, Doris PA. Kinetics of competitive reverse transcriptase-PCR. In: Innis MA, Gelfand DH, Sninsky JJ, eds. PCR applications. Protocols for functional genomics. San Diego: Academic Press, 1999:231-61.

(19.) Kuhnert M, Srohmeier R, Stegmuller M, Halberstadt E. Changes in lymphocyte subsets during normal pregnancy. Eur J Obstet Gynecol Reprod Biol 1998;76:147-51.


Departments of [1] Clinical Biochemistry 339, and [4] Gynecology and Obstetrics, Copenhagen University Hospital, H:S Hvidovre Hospital, 30 Kettegaard Alle, DK-2650 Hvidovre, Denmark. [2] The Chromosome Laboratory, Prenatal Research Unit, and s Department of Obstetrics and Gynecology, Juliane Marie Center, H:S Rigshospitalet, University of Copenhagen, 9 Blegdamsvej, DK-2100 Copenhagen O, Denmark. [3] Dako A/S, 42 Produktionsvej, DK-2600 Glostrup, Denmark.

* The contributions by T.V.F Hviid and A.M. Hogh are equal and the order of authorship is arbitrary.

[dagger] Author for correspondence. Fax 45-3675-0977; e-mail or

Received September 11, 2000; accepted January 12, 2001.
Table 1. Qualitative results of RT-PCR amplification of [gamma] -,
[epsilon]-, and [zeta] -globin transcripts.

 [gamma] -Globin [epsilon] -Globin

 Whole CD71+ Whole CD71+
 blood fraction blood fraction

Nonpregnant women, n

Positive 1 19 0 0
Negative 19 1 (a) 20 20
Total 20 20 20 20

Pregnant women
 (9-13 weeks of
 gestation), n

Positive 0 10 0 0
Negative 14 4 (a) 14 14
Total 14 14 14 14

Fetal blood samples
 (17-20 weeks of
 gestation), n

Positive 8 8 7 8
Negative 0 0 1 0
Total 8 8 8 8

 [zeta] -Globin

 Whole CD71+
 blood fraction

Nonpregnant women, n

Positive 0 3
Negative 20 17 (b)
Total 20 20

Pregnant women
 (9-13 weeks of
 gestation), n

Positive 0 5
Negative 14 9 (b)
Total 14 14

Fetal blood samples
 (17-20 weeks of
 gestation), n

Positive 4 8
Negative 4 0
Total 8 8

(a, b) Fisher's exact test: (a) P = 0.13; (b) P = 0.23.
COPYRIGHT 2001 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

 Reader Opinion




Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Molecular Diagnostics and Genetics
Author:Hogh, Anne Mette; Hviid, Thomas Vauvert F.; Christensen, Britta; Sorensen, Steen; Larsen, Rasmus D.;
Publication:Clinical Chemistry
Date:Apr 1, 2001
Previous Article:DNA melting analysis for detection of single nucleotide polymorphisms.
Next Article:Specific reverse transcription-PCR quantification of vascular endothelial growth factor (VEGF) splice variants by LightCycler technology.

Related Articles
Placental mRNA in maternal plasma and its clinical application to the evaluation of placental status in a pregnant woman with placenta...
Fetal rhesus D mRNA is not detectable in maternal plasma.
In-cell PCR method for specific genotyping of genomic DNA from one individual in a mixture of cells from two individuals: a model study with specific...
Quantitative analysis of the bidirectional fetomaternal transfer of nucleated cells and plasma DNA.
Large amounts of cell-free fetal DNA are present in amniotic fluid.
Plasma [gamma]-globin gene expression suggests that fetal hematopoietic cells contribute to the pool of circulating cell-free fetal nucleic acids...
The concentration of circulating corticotropin-releasing hormone mRNA in maternal plasma is increased in preeclampsia.
Increased fetal DNA concentrations in the plasma of pregnant women carrying fetuses with trisomy 21.
Fetal cell-free plasma DNA concentrations in maternal blood are stable 24 hours after collection: analysis of first- and third-trimester samples.
Development, characterization, and use of monoclonal antibodies made to antigens expressed on the surface of fetal nucleated red blood cells.

Terms of use | Copyright © 2015 Farlex, Inc. | Feedback | For webmasters