Detection of fetal mutations causing hemoglobinopathies by non-invasive prenatal diagnosis from maternal plasmaE D'Souza, PM Sawant, AH Nadkarni, A Gorakshakar, K Ghosh, RB Colah
Department of Haematogenetics, National Institute of Immunohaematology (ICMR), K.E.M. Hospital Campus, Parel, Mumbai, Maharashtra, India
Correspondence Address: Source of Support: Indian Council of Medical Research, Conflict of Interest: None DOI: 10.4103/0022-3859.109483
Source of Support: Indian Council of Medical Research, Conflict of Interest: None
Background: Prenatal diagnosis of hemoglobinopathies enables couples at risk to have a healthy child. Currently used fetal sampling procedures are invasive with some risk of miscarriage. A non-invasive approach to obtain fetal deoxyribonucleic acid (DNA) for diagnosis would eliminate this risk. Aim: To develop and evaluate a non-invasive prenatal diagnostic approach for hemoglobinopathies using cell-free fetal DNA circulating in the maternal plasma. Settings and Design: Couples referred to us for prenatal diagnosis of hemoglobinopathies where the maternal and paternal mutations were different were included in the study. Materials and Methods: Maternal peripheral blood was collected at different periods of gestation before the invasive fetal sampling procedure was done. The blood was centrifuged to isolate the plasma and prepare DNA. A size separation approach was used to isolate fetal DNA. Nested polymerase chain reaction (PCR)-based protocols were developed for detection of the presence or absence of the paternal mutation. Results and Conclusions: There were 30 couples where the parental mutations were different. Of these, in 14 cases the paternal mutation was absent and in 16 cases it was present in the fetus. Using cell-free fetal DNA from maternal plasma, the absence of the paternal mutation was accurately determined in 12 of the 14 cases and the presence of the paternal mutation was correctly identified in 12 of the 16 cases. Thus, this non-invasive approach gave comparable results to those obtained by the conventional invasive fetal sampling methods in 24 cases giving an accuracy of 80.0%. Although the nested PCR approach enabled amplification of small quantities of cell-free DNA from maternal plasma at different periods of gestation after size separation to eliminate the more abundant maternal DNA, an accurate diagnosis of the presence or absence of the paternal mutation in the fetus was not possible in all cases to make it clinically applicable.
Keywords: Cell free fetal deoxyribonucleic acid, hemoglobinopathies, non-invasive, prenatal diagnosis
Circulatory cell-free fetal deoxyribonucleic acid (DNA) in maternal plasma is now increasingly being used as the source of fetal DNA for prenatal diagnosis. From the time Lo et al. showed its application in the detection of Y chromosome sequences, maternal plasma has been widely used for non-invasive prenatal diagnosis. The presence of cell-free fetal DNA in maternal plasma avoids the need of any of the expensive enrichment or isolation strategies as is required for the isolation of fetal cells from the maternal circulation. The technique has been widely used for non-invasive prenatal diagnosis of fetal RhD status, fetal gender determination for sex-linked disorders like hemophilia, detection of achondroplasia, myotonic dystrophy, cystic fibrosis, Huntington disease, and congenital adrenal hyperplasia. 
However, in the background of abundant maternal DNA present in the maternal plasma most of the studies have concentrated on detection of fetal markers that were unique to the fetus itself or were inherited from the father so that they could be easily distinguished from the maternal DNA. To effectively use this cell-free fetal DNA, the focus has shifted to develop quantitative gender-independent means for non-invasive prenatal diagnosis for autosomal recessive disorders like hemoglobinopathies.
Chan et al. showed that fetal DNA molecules predominantly have an approximate size <300 bp, whereas most maternally derived DNA molecules are considerably larger. To assess if it was practical to use this principle for the non-invasive prenatal diagnosis of hemoglobinopathies, we used the combination of the size separation approach and semi nested polymerase chain reaction (PCR) for the detection of fetal mutations.
The study was conducted at the National Institute of Immunohaematology, Mumbai, where the couples were referred for prenatal diagnosis of hemoglobinopathies. Informed consent was taken and the study was cleared by the Ethics Committee of the Institute. 15 ml peripheral venous blood was collected from the mother in EDTA. 10 ml blood of the father and the previously affected child (if available) was also collected. After blood collection, the couples were referred for the invasive procedure for fetal tissue/blood sampling depending on the gestation age of the fetus.
The protocol devised had three main parts: Separation of plasma and extraction of the cell-free fetal DNA, size separation approach to isolate fetal DNA, and detection of fetal mutations causing hemoglobinopathies using Nested PCR.
Separation of plasma and extraction of the cell-free fetal deoxyribonucleic acid
5 ml of the maternal blood sample was first spun at 1600 g for 10 minutes. The plasma was separated and respun at 16000 g for 10 minutes. Care was taken to avoid aspirating any cell pellet. The sample was then stored at −20°C until further use.
The mutations of the parents were first characterized using Covalent Reverse Dot Blot hybridization (CRDB) to screen for the six common Indian mutations causing β-thalassemia: (IVS 1 nt 5 (G→C), IVS 1 nt 1 (G→T), Cd 8/9 (+G), Cd 41/42 (-CTTT), Cd 15 (G→A) and Cd 30 (G→C)) along with Hb S and Hb E.  If the mutations were uncharacterized, a PCR based on Amplification Refractory Mutation System (ARMS) was then used to screen for other uncommon Indian β-thalassemia mutations.  The 619 base pair (bp) deletion was detected by PCR across the break points of the deletion and electrophoresis on a 2% agarose gel.
Based on the mutations of the parents, the strategy was designed. We first chose samples in which the parental mutations were different. There were 30 such couples. The plasma was first thawed at room temperature and then 800 μl of the plasma was used for extraction of DNA using the Qiagen Mini Blood Kit with minor modifications in the manufacturer's instructions. The DNA was eluted in 50 μl of the AE buffer provided.
Size separation approach to isolate fetal deoxyribonucleic acid
This DNA (50 μl) was loaded on a 1% agarose gel prepared in 0.5× Tris Borate EDTA buffer and was run for about 90 min at 80 V. Only one DNA sample was run per gel alongside a 100 bp ladder (Roche). The gel fragment corresponding to 300 bp and less was excised from the gel using a sterile scalpel and the DNA was eluted using the Gel extraction Kit (Qiagen). The sample was now eluted out in 50 μl EB buffer provided.
Detection of fetal mutations causing hemoglobinopathies using nested polymerase chain reaction
Since the parental mutations were different, the strategy that was used was to screen the fetal DNA sample for the presence or the absence of the paternal mutation. Primers were designed bearing in mind that the fragment size of the target DNA was 300-350 bp.
For the external PCR, four separate reactions were run based on the mutation of the father.
All the reactions were run using the same PCR protocol. The 25 μl reaction mixture contained 10× PCR buffer containing 15 mM MgCl 2 , 1.25 mM deoxynucleoside triphosphate (dNTP) mix, spermidine (1 M solution) (Sigma), and Taq polymerase (5 U/μl) along with 10 picomoles (pm) of the respective primers.
polymerase chain reaction
The nested PCR reaction was based on ARMS to screen for the presence or absence of the paternal mutation. A 1: 500 μl dilution of the external PCR product was then used as the template for the nested PCR based on ARMS. The 25 μl reaction mixture contained 10× PCR buffer containing 15 mM MgCl 2 , 1.25 mM deoxynucleoside triphosphate (dNTP) mix, spermidine (1M solution) (Sigma) and Taq polymerase (5U/μl). For the internal controls we used 10 pm of control primers and 10 pm of allele specific N and M primers [Table 1]. The PCR conditions were set at 94°C for 1 min, 59°C for 1 min and 72°C for 1 min 30 sec for 25 cycles.
The reactions that were set up for every mutation and the expected fragment size in shown in [Table 2]. All the samples were run on a 2% agarose gel along with marker VIII (Roche, IN, USA) and stained with ethidium bromide.
There were 30 couples where the parental mutations were different. To compare these results with findings of the invasive method, the results obtained after the semi nested PCR were divided into two groups: One group where the paternal mutation was present and the other group where the paternal mutation was absent. This division was based on the results obtained by the invasive method.
Of the 30, there were 16 cases where the paternal mutation was present. Among these, the presence of the paternal mutation was accurately detected in 12 cases. [Figure 1] shows the results of one family where the paternal IVS1-5 (G > C) mutation was present both by the invasive and non-invasive approaches. In the four cases where there was a misdiagnosis, there was no amplification of the mutant paternal allele and by the non-invasive method they appeared to be normal. The findings are summarized in [Table 3].
In the remaining 14 cases, the paternal mutation was absent. Among these, the absence of the paternal mutation was accurately detected in 12 cases. Here too there was a misdiagnosis in two cases. The fetal sample was diagnosed to be normal by the invasive method but in the semi nested PCR there was non-specific amplification of the mutant paternal allele. The findings are summarized in [Table 4] and the gel picture of one of the cases is shown in [Figure 2].
Thus, using semi nested PCR, the ability to detect the presence or the absence of the paternal mutation in cell-free fetal DNA from maternal plasma was in 24 out of the 30 cases, giving an accuracy of 80%.
The use of cell-free fetal DNA has been very well established for sex-linked disorders but no large study has been carried out in the field of hemoglobinopathies. The main reason is that these disorders follow an autosomal recessive pattern of inheritance. The other reason is that it is difficult to discern between fetal and maternal DNA and it cannot be predicted whether the fetus is completely normal when the maternal and fetal mutations are the same.
Tungwiwat et al. have used real-time PCR for prediction of fetal mutations in α-thalassemia. Based on the differences in the threshold cycle values and the concentration of the amplified DNA, the fetal genotype was accurately predicted with the help of a semi-nested real time qPCR.  Fucharoen et al. in 2003  used nested PCR and restriction enzyme analysis to detect fetal mutations in five cases, when the mother was a carrier of thalassemia and the father was a Hb E carrier. Li et al. 2004,  combined the size separation approach with PNA clamping and screened 32 maternal plasma samples for the presence or the absence of the paternal mutation. Galbiati et al., 2008,  made use of PNA clamping along with microelectronic microchip analysis performing non-invasive prenatal diagnosis of thalassemia in 41 cases.
In all these studies, the main drawback has been that no definite protocol could be drawn that would allow for routine prenatal diagnosis. Also, the technique is restricted as the diagnosis is only possible when the parental mutations are different. Our study aimed to see if it was feasible to devise a simple approach to screen the maternal plasma sample for the paternal mutations. Nested PCR was the option of choice. The use of nested PCR has been successfully shown for the detection of fetal sex. In a study conducted by Smid et al., they showed that nested PCR on plasma DNA had the highest number of correct identifications of fetal sex. 
As we were unable to rule out the possibility of the presence of maternal DNA, the samples were chosen where the parental mutations were different. Thus, by this approach, we were only able to identify if the fetal DNA was a carrier of the paternal mutation or not. The approach would not find use when the parental mutations were the same. Although the sample number was only 30, the number of parental mutations being screened for were 11, 9 different β thalassemia mutations, and Hb S and Hb E. The β thalassemia mutations were Cd 41/42 (-CTTT), Cd 8/9 (+G), Cd 30 (G→C), Cd 15 (G→A), Cd 5 (-CT), Cd 16 (-C), IVS 1 nt 1 (G→T), IVS 1 nt 5 (G→C), and IVS II nt 837 (T→G). We were able to study the mutations spread all over the β globin gene.
We divided our results into two parts based on the results we obtained from the conventional invasive diagnosis approach. There were 16 cases where the paternal mutation was present and 14 cases where the paternal mutation was absent. On correlating our results with the results obtained by the invasive techniques, we found that accurate results could be obtained only in 24 cases restricting the accuracy of this method to just 80%. There were four cases where the misdiagnosis was due to allele drop out, a problem associated with the choice of the PCR technique.
Although simple and easy to perform without the need of any expensive isolation or enrichment techniques, this approach was found to have a limited application to parents having different mutations. Also the technique cannot be used by itself for diagnosis; when the paternal mutation is present it would have to be complimented with another test as the technique would not indicate if the fetus is going to be a heterozygote or compound heterozygote. However, by a size separation and nested PCR approach if the absence of the paternal mutation is shown then this will completely negate the need of an invasive procedure. The number of women who will have to undergo an invasive procedure will be drastically reduced. Today most of the methods described make use of Real-time PCR for the non-invasive procedures which may not be feasible to use due to the high cost of the machine and its reagents. Thus, we evaluated the simple nested PCR approach. For applying in a clinical setting, this approach would have to await developments that permit better separation of maternal and fetal DNA which is still challenging.
Recently, an innovative strategy based on co-amplification at a lower denaturation temperature PCR (COLD-PCR) which exploits differences in the melting temperature between variant and wild-type sequences has been developed and used for the detection of two paternally inherited mutations causing β thalassemia from maternal plasma. This may be a simple cost effective method for non-invasive fetal diagnosis but needs further validation.  Other emerging approaches are based on single molecule counting technologies such as digital PCR and massively parallel sequencing which are still very expensive and need large scale validation but they would play an important role in non-invasive prenatal diagnosis of monogenic disorders like the hemoglobinopathies from circulating fetal DNA in maternal plasma. 
We are grateful to all the obstetricians for the invasive fetal sampling procedures and the Indian Council of Medical Research for financial support for the study.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]