Article Text
PDF
Abstract
The importance of monitoring post haematopoietic stem cell transplantation (hSCT) chimerism has been defined in numerous publications. Single-nucleotide polymorphisms (SNPs) are molecular markers that vary significantly among different populations. Allied to a very sensible technique, SNP assays seem to be very sensitive (0.001%) when post hSCT chimerism is measured. However, well known SNP frequencies are limited to certain populations, mainly in countries where there is a high level of diversity in its population, therefore restricting their use worldwide. Amplification by SYBR green based quantitative real time PCR of eight pairs of allele-specific SNPs (MLH-1, PECAM-1, ICAM-1, SUR-1, HA-1, rs715405, rs713503, rs2296600) was conducted in 88 patient/donor pairs, who underwent allogeneic myeloablative or non-myeloablative hSCT. One informative allele was detected in at least 42% (n=37) of the samples; 20% (n=18) had at least two informative alleles; 10% (n=9) had at least three informative alleles; 9% (n=8) had more than three informative alleles and 18% (n=16) showed no informative allele at all. Overall, the frequency of informative alleles for these SNPs in the Brazilian population was very low. Consequently, the amount of information attained reached 9% of those expected, being able to discriminate only eight pairs of donor/recipient samples with more than three informative alleles, making them useless for the quantification of chimerism in our routine.
- BMT
- Genetics
- Haemato-Oncology
- Molecular Biology
- PCR
Statistics from Altmetric.com
Introduction
Haematopoietic stem cell transplantation (hSCT) is an established medical procedure used to treat various malignant and non-malignant haematological diseases such as leukaemia, aplastic anaemia, multiple myeloma and lymphomas. The success of a bone marrow transplant—will rely on engraftment of the donor's transplant and restoration of the deficient haematopoiesis.1–3 The major causes of treatment failure are disease relapse, graft rejection and graft-versus-host diseases (GvHD).4–8 The key test used to predict relapses and graft rejections is monitoring the post-transplant chimerism.1–3
In Medicine, a chimaera is characterised as an individual in whom genetically different cell populations coexist. A chimaeric condition can arise naturally or it can be induced artificially. Common well-known classes of natural human chimaeras are cytomictical, whole body, fetal-maternal, germ cell and tumour chimaeras.9 Artificially generated chimerism is a consequence of hSCT. Following hSCT, the recipient becomes a chimaera because the newly acquired haematopoietic system consists mainly of allogeneic donor cells.3
Chimerism analysis involves methods used to detect the presence of donor cells in patients who have undergone hSCT. Surveillance of donor cell engraftment is crucial for the early detection of graft rejection, GvHD and disease relapse. This information is clinically relevant as a disease relapse can still respond positively to treatment at an early stage.10
The basic principle of chimerism detection is the utilisation of differences between the recipient and donor genomes. It seems challenging since the DNA sequence similarity between any two given individuals is thought to be approximately 99.9%.11 However, only a small fraction of the genome determines the genetic variability of individuals. In early studies, large-scale genomic differences were observed, such as quantity and structural variations of chromosomes. Technological advances and progress in molecular biology have allowed the detection of small-scale genome variations, and identified differences at a DNA level by adopting sequence polymorphisms. These sequence polymorphisms are single-nucleotide polymorphisms (SNPs), repetitive DNA elements (mini- and microsatellites), various size insertions, deletions, inversions, duplications and copy-number variations.12 ,13
SNPs are the most abundant form of genetic variability in the human genome. They are single-base substitutions of one nucleotide by another and both versions are observed in the general population at a frequency greater than 1%. Most SNPs are biallelic and display various frequencies within an experimental population. SNP genotyping has become more important after SNPs were discovered to contribute directly or indirectly to disease predisposition, drug response and disease progression, as well as genotype identification. Currently, a large number of SNP genotyping technologies are available addressing specific needs of research, including chimerism analysis.8 ,14–21
Several methods exist to predict the course of disease after hSCT is applied such as: genetic and/or immunophenotypic monitoring of minimal residual disease,2 ,5 ,6 ,22–25 as well as chimerism analysis.21 The methodology for assessing chimerism following transplantation must be informatively, sensitively and quantitatively accurate. Several techniques have been described in publications to establish haematopoietic chimerism.21 ,26–28
More recently, quantification of chimerism by real time reverse transcriptase PCR (qRT-PCR) methods have been developed showing high sensitivity, accuracy and faster performance than previous techniques. These techniques are based on TaqMan,14–16 ,20 ,21 ,29 ,30 hybridisation31 or fluorescent probes.3 ,8 A combination of a high polymorphic DNA marker—SNP—with a highly sensitive technique—qRT-PCR—and a fluorescent based platform—SYBR green—has allowed for a more rapid, accurate and low cost assay for quantifying chimerism, engraftment of host cells and disease relapse.3 ,8 ,21
In the present study, we are evaluating a panel of 16 allele-specific SNPs described in literature references to quantify the post hSCT chimerism by SYBR green-based real time PCR.
Material and methods
Patients and donors
Samples from 88 patient/donor pairs who underwent human leucocyte antigen (HLA) related or unrelated allogeneic myeloablative and non-myeloablative hSCT from two different institutions were selected. A written informed consent had been obtained prior to the transplant procedure from every patient and donor, and the ethics committees of both institutions had approved the study. Clinical data was collected manually or electronically from each patient file. Patient and donor characteristics are described in table 1.
Recipient and donor characteristics
Blood samples from both, donor/recipient, were collected pre-hSCT. In cases where the patient had already been transplanted, a mouth wash was collected from the recipient and a blood sample from his donor. There were three cases of non-related donors, consisting of two units of cord blood and a bone marrow. For these specific cases, a sample of blood was separated before cell infusion. DNA was extracted using a conventional commercial kit, (QIAmp—Qiagen) by following each of the manufacturers’ instructions. The concentration and purity of each DNA sample was determined using a NanoDrop 2000c Spectrophotometer (Thermo Scientific).
SNP selection
Eight SNPs were selected from different highly informative panels described in the references. Loci were selected based on their frequency and heterozygosity.3 ,15 ,32 ,33 Primers were then designed using Primer3 Design Software (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference gene. The selected SNPs, primer sequences and annealing temperatures are listed in table 2. In order to determine possible design mistakes and assure its specificity, all primers were sequenced with an ABI 3700 Applied Biosystems equipment (data not shown).
Selected single-nucleotide polymorphisms (SNPs) and primers
Quantitative and qualitative real time PCR
All reactions were performed at a 7500 Real-Time PCR System Applied Biosystems. The PCR conditions were as follows: 100 ng of genomic DNA, 6.25 µl of SYBR Green PCR Master Mix (Applied Biosystems), 0.75 pmol of each primer, and water to complete a final volume of 12 µl. Screening and quantification PCR reactions were performed with a 10-min initial denaturation at 95°C, followed by 45 amplification cycles at 95°C for 15 s, annealing (varying from 56 to 64°C) for 30 s and 72°C for 30 s.
Standard amplification curves
Standard Amplification Curves were done for each allele-specific SNP to evaluate the amplification efficiencies of each informative allele and its respective endogenous control, and also to evaluate the detection limit of the quantitative PCR assay. The curves were done with serial dilutions (100, 10, 1, 0.1 and 0.01%) of either each allele positive DNA template in allele negative DNA or water. The input of allele positive DNA and the total amount of chimeric DNA mixture within each dilution was 100 ng. All samples were in triplicate. Efficiency and slopes were determined for each primer.
Genotyping
All recipient and donor samples were genotyped in order to identify informative alleles. An SNP marker was considered informative when it presented a different nucleotide for a recipient end donor, so the selection criterion was the presence of an allele within a recipient/donor genome either in a homozygous or a heterozygous state. A specific cycle threshold (Ct) was determined for each PCR reaction where Ct>35 was considered as negative.3 ,8 In cases where an informative allele was identified, the chimerism was quantified as described above in post-transplant samples. The chimerism assay was performed in duplicate in a mixture of donor/recipient DNA samples post-hSCT with the same PCR conditions as previously described. The relative quantification ΔCt29 formula, modified from Gineikiene and colleagues,3 was used to determine the amount of relative proportion of informative allele in the post-hSCT sample whenever applicable.
Results
Clinical data
Demographic and clinical data from all 88 donor/recipient pairs are listed in table 1. The average age, at the time of the transplant, was 39.2 years for patients and 33.9 years for donors. The percentage of the sex mismatch ratio M : F, 18 (20.4%) and F : M, 29 (33%) were higher than the sex match ratio, M : M, 25 (28.4%) and F : F 16 (18.2%). Acute myeloid leukaemia, 15 (17.04%); chronic myeloid leukaemia, 14 (15.90%); severe aplastic anaemia, 13 (14.77%) were most frequently found diseases. At the time of the hSCT, 53 (60.22%) were in complete remission (CR), 16 (18.18%) were in partial remission and 19 (21.60%) had disease activity. Most patients, 51 (57.95%) went through a myeloablative conditioning regimen, and 54 (61.36%) received peripheral blood as their source of haematopoetic stem cells. Only 3 (3.40%) samples were unrelated; 2 cord blood samples and 1 bone marrow sample. Forty-eight (54.54%) patients developed acute GvHD and 60 (68.18%) patients were found to have a chronic form. The overall survival rate was 61.75 months, whereas the disease-free survival rate was 14.9 months. Twelve (13.63%) had relapsed and 65 (73.86%) remained alive.
Genotyping and alleles frequency
Sixteen allele-specific SNPs, selected from previous studies 3 ,15 ,29 ,32 ,33 were designed in order to evaluate their level of informativity and applicability in the Brazilian population for quantification of post-hSCT chimerism. Informativity of each SNP allele was calculated as a percentage of alleles capable of discriminating the donor/recipient samples from the total amount of samples tested.
The slopes and efficiency ranged from −3.3985 to −2.705 and 1.985563 to 2.342533 respectively.
The overall informativity of the 16 alleles was 81% (72/88). Whereas 16 (18%) did not amplify any of the 16 alleles; 37 (42%) amplified only one of the 16 alleles; 18 (20%) amplified only two; 9 (10%) amplified only three and 8 (9%) amplified more than three alleles. In tables 3 and 4, the individual frequency and level of informativity of each allele of the researched population are listed respectively. HA1 (33.13%), PECAM1 (32.75%) and rs2296600 (29.23%) presented the highest levels of informativity whereas, on the other hand, MLH1 and ICAM1 were not informative within the sampled population, as they have showed no informative allele for any of the samples analysed.
Genotype and allele frequencies of the studied population
Level of informativity of each individual single nucleotide polymorphism (SNP) in our study compared with the literature references
Discussion
Chimerism quantification is an important tool in order to monitor the success of the hSCT, to evaluate the bone marrow reconstitution by the donor haematopoietic cells, and to predict any rejection and/or disease relapse. Among several well-described methodologies in the reference literature, the qRT-PCR has been described as the most sensitive.3 ,29 ,21 The possibility of using a fluorescent dye—SYBR green—makes the assays more accessible, less expensive and faster.8 SNPs are biallelic variants that differ at a single DNA base pair, making them exceptionally polymorphic among humans, with an estimated rate of occurrence of 1/1000 bp.34 Also, SNPs have been proven to be a particularly useful marker for monitoring chimerism after hSCT because they are stable and unique, and can be analysed by sensitive quantitative methods, such as qRT-PCR.15
The purpose of this study initiated with the necessity of implementing the quantification of chimerism by qRT-PCR using SYBR green in our routine, which had shown low costs and high sensitivity results.3 ,8 ,21 Since we have not made studies of SNP frequencies available in publications and SNP databases (dbSNP) for the Brazilian population, it was difficult to select an ideal panel based only on data from other populations. The Brazilian population is quite diverse as it is influenced by Europeans, Native Brazilians, African Negroes, Latinos and Japanese. Nowadays it is very hard to identify these isolated groups, as they had been miscegenated throughout the years.35
Being capable to discriminate donor/recipient cells, we have selected 16 allele-specific SNPs (table 2) previously referenced,3 ,15 ,29 ,32 ,33 in order to set up a panel that would be informative for our population. They were chosen based on their level of heterozygosity presented by Caucasians, Negroes and Latins, according to the dbSNP (http://www.ncbi.nlm.nih.gov). Gineikiene3 has already reported the importance of setting a panel of SNPs based on each specific population ethnicity. As a matter of fact, he and fellow contributors have also reported the uselessness of the Maars15 panel for their population.
Individually analysing each marker with the highest levels of informativity, PECAM1 presented an indifferent frequency from the groups described by Maars,15 Nichols and Behar (1996); HA1. Our results for heterozygotes—CATG—matched with that of the reference literature; however, it showed that in our population the CA allele was more frequent than the TG allele (TG), differing from Maars,15 (CA-23%; TG-35%) and Wilke36 and Tseng,37 (CA-17%; TG-34%). SUR1 has proved to be a powerful marker for us and Maars;15 however, our frequency differs from theirs (CC-29%; CT-48%; TT-23%); MLH1 was the marker with the lowest level of informativity described by Maars.15 This study did not show any allelic differences between the donor and recipient, as with ICAM1, which for Maars,15 was the one with the highest level of informativity. This difference assembles the SNP's characteristic genetic variation. A biallelic locus has only three possible genotypes, as the frequency of each possible combination is crucial for a successful high informative panel. Virtually any biallelic SNP locus with allele frequency of 0.4–0.6 could be used for genotyping.3 However, the main issue arises when there exists a very racially mixed population like the Brazilian population, and the numbers of studies describing the SNP frequency in a mixed population are inexpressive. There is almost no data available in the international SNP database (dbSNP), thus making the effort to find the right SNPs difficult for our study. Therefore, we do not know if by using a larger number of non-randomised SNPs, we would be able to demonstrate their efficacy for chimerism assays.
Considering the combination of the alleles, the power of discrimination for our panel of 16 markers for the population studied, showed to be the lowest when compared with other reports from publications listed on table 5. This may entail that even by being a powerful genetic marker, SNPs are hypervariable among populations where screening for a right informative SNP would be time consuming due to a large quantity of loci to be researched. As previously mentioned, Maars’15 panel did not seem to be informative enough for Gineikiene's population or ours, thereby reflecting the genetic diversity seen among distinct populations.
Level of informativity of the combined panel of 16 single-nucleotide polymorphisms to the reference studies.
When analysed, the combined numbers of informative alleles, capable of discriminating donor/recipient cells for the same pair (donor/recipient), had stratified the level of informativity, showing that this panel became useless—as only eight pairs (9%) proved to have more than three informative alleles and 9 (10%), 3—this suggests that the alleles alone may be frequent in a certain population, but by the time they are combined, different alleles with different frequencies end up diluted, unless chosen to work with SNPs with a least similar frequency.
In order to find the best SNP and consequently a very informative panel, SNPs with an allele frequency of 50% or higher would be powerful enough to provide a 99% probability of identifying at least a single informative SNP. As a result, a panel with at least six SNPs within this frequency would be sufficient to provide 99.6% of probability in identifying at least one informative locus in related and in unrelated pairs,,16 if you apply the right SNPs with the right population.
In summary, a combination of a high number of SNPs from different international panels may not be enough to have a successful panel with a high level of informativity, proposing that SNPs highly vary among the population's ethnicity. Another important point to mention is that our results demonstrated an important aspect of these specific alleles in the studied population, contributing to the knowledge and distribution of them in our population.
Take home messages
-
Chimerism quantification is very important to monitor the success of the post hSCT;
-
With a good panel of informative markers, VNTR may be a possible option for those who do not have access to other techniques such as STRs and or SNP/qRT-PCR;
-
STR still the GOLD standard method for quantifying chimerism;
-
A SNP/SYBR Green qRT-PCR can be a good, cheap, fast and reliable option for chimerism analysis when you have informative primers.
Acknowledgments
The authors would like to thank both São Paulo University Hospital and Santa Marcelina Hospital (CSSM) for allowing them to perform this study. The authors are also very thankful to CAPES and CNPQ for the financial support and to UNIFESP and UFAL.
References
Footnotes
-
Contributors JSRdO and MSF were the mentors and instructors for this project. JSRdO was also responsible for selection of the patients. JLD and MMA-S helped with the primer selection and DNA preparation. CACA also helped with patient selection, processing samples including DNA extraction, he also choose and designed the primers, and have done all the molecular work (qRT-PCR) and data analysis.
-
Funding None.
-
Competing interests None.
-
Ethics approval UNIFESP and CSSM Ethical Committees and The National Information System for Ethics and Research with Human Beings—SISNEP approved this study.
-
Provenance and peer review Not commissioned; externally peer reviewed.