A simple way to evaluate self-designed probes for tumor specific Multiplex Ligation-dependent Probe Amplification (MLPA)
- Kristina Pedersen†1, 2, 4,
- Emilia Wiechec†1, 2,
- Bo E Madsen3, 5,
- Jens Overgaard2 and
- Lise Lotte Hansen1Email author
© Hansen et al; licensee BioMed Central Ltd. 2010
Received: 5 February 2010
Accepted: 26 June 2010
Published: 26 June 2010
The Multiplex Ligation-dependent Probe Amplification (MLPA) is widely used for analysis of copy number variations (CNVs) in single or multiple loci. MLPA is a versatile methodology and important tool in cancer research; it provides precise information on increased or decreased copy number at specific loci as opposed to loss of heterozygosity (LOH) studies based upon microsatellite analysis. Pre-designed MLPA kits and software are commercially available to analyze multiple exons, genes, and genomic regions. However, an increasing demand for new gene specific assays makes it necessary to self-design new MLPA probes for which the available software may not be applicable. During evaluation of new self-designed reference probes, we encountered a number of problems, especially when applying the MLPA methodology to tumor samples.
DNA samples from 48 unaffected individuals and 145 breast cancer patients were used to evaluate 11 self-designed MLPA probes and determine the cut-off values for CNV, before applying the MLPA probes to normalize the target probes in a cohort of affected individuals. To test the calculation strategy, three probes were designed to cover regions in Regulator of G-protein Signaling 8 (RGS8), which we previously have identified as being affected by allelic imbalance by LOH analysis across RGS8 in the cohort comprising 145 breast tumors. Agreement between the LOH results and the results obtained by each of the three MLPA probes in RGS8 was found for 64%, 73%, and 91%, of the analyzed samples, respectively.
Here, we present a straightforward method, based upon the normalization pattern in both unaffected and affected individuals, to evaluate self-designed reference probes and to calculate CNV for the MLPA assay with specific focus on the difficulties when analyzing tumor DNA.
Copy number variations (CNV), including deletions or duplications of whole chromosomes and minor fragments down to single exons are implicated in the onset and risk of human congenital and somatic diseases . The Multiplex Ligation-dependent Probe Amplification (MLPA) is widely used both in research and clinical laboratories and is a promising methodology to assess CNV . MLPA assays provide highly detailed information on CNV, from chromosome aneuploidy down to single exons. The methodology is fast and straightforward to perform, standard laboratory equipment is used, and a limited amount of DNA (20 ng/analysis) is necessary to provide high quality and reproducible results. Ligation-dependent PCR amplification was first described by Hsuih et al.  further developed by Schouten et al.  and a growing number of predesigned gene specific MLPA assays is commercially available from MRC Holland http://www.mrc-holland.com. The MLPA methodology is versatile and can be applied to different analyses as assessment of genomic copy number variations (CNV), DNA methylation status of specific loci (MS-MLPA), evaluation of chromosomal break points, CGH results, and gene expression profiles (RT-MLPA) [5–8]. The advantage is that multiple loci can be analyzed simultaneously and software is available to perform the tedious calculations.
Current research involving MLPA is mainly based on the commercially available MLPA kits targeting specific genes (MRC-Holland, Amsterdam, The Netherlands), which significantly shorten the optimization time and facilitates data analysis, as specific software is available to analyze the results.
However, the increasing demand for new gene specific assays makes it necessary to self-design individual probes for the analysis . Establishment of CNV is based upon results obtained from reference probes, which are included in each individual assay and used to normalize the values obtained from the target genes. Specific care has to be taken when analyzing tumor genomes, as some of these reference genes may be located in regions affected by cancer specific CNV [10–12].
Here, we describe the evaluation of self-designed reference probes, using a cohort of unaffected and affected individuals, to determine the quality and the cut-off values determining CNV, before applying and using the reference probes to normalize self-designed target probes in tumor samples. To evaluate our calculation strategy we aimed to determine CNV in Regulator of G-protein Signaling 8 (RGS8), which previously has been identified as allelic imbalance by a combination of loss of heterozygosity (LOH) analysis and PCR mapping (unpublished results).
Breast tumor DNA was isolated from 145 women diagnosed with unilateral primary breast cancer. The cohort was collected between August 1992 and January 1994 and described in . DNA isolated from peripheral blood was obtained from healthy Danish medical students in their first year of University. All DNA was isolated according to .
The respective local ethical committees have approved the use of these samples.
Probe design and MLPA assay
Probes for the MLPA assays were designed according to "Improved synthetic probe design protocol" available on http://www.mrc-holland.com, and the guidelines were followed carefully.
The set of synthetic oligonucleotide probes used in the MLPA assay.
Left probe hybridizing sequence
Right probe hybridizing sequence
The MLPA assay was performed using sets of self-designed probes and the EK5 SALSA MLPA reagents kit purchased from MRC-Holland (Amsterdam, The Netherlands). Briefly, 10 ng of genomic DNA in a volume of 5 μl was denatured for 5 min at 98°C, cooled and hybridized with MLPA probes (1.5 μl/sample) in the presence of SALSA MLPA buffer (1.5 μl/sample) for 16 h at 60°C. Ligation reaction was performed for 15 min at 54°C by adding 32 μl of ligation mixture (3 μl SALSA Ligase buffers A and B, respectively, 25 μl demineralized water, and 1 μl SALSA Ligase-65/sample) followed by heating to 95°C for 5 min. PCR amplification of the ligation product was performed according to the standard protocol. Briefly, 5 μl of the ligation product was added to the mixture consisting of (15.75 μl demineralized water, 2 μl SALSA-PCR buffer, 1 μl SALSA-enzyme dilution buffer, 1 μl SALSA FAM PCR primers and 0.25 μl SALSA polymerase/sample) while the samples were left at 4°C or on ice. The samples were pre-heated to 60°C for 20 s, followed by 35 cycles of 95°C for 30 s, 60°C for 40 s, 72°C for 1 min, and a final extension for 20 min at 72°C. From each PCR reaction, 1.0 μl product was mixed with 9 μl of deionized formamide (Amresco, Solon, Ohio) and 0.3 μl of Genescan-Rox 500 size standard (Applied Biosystems, Foster City, CA, USA) and all fragments were separated by capillary electrophoresis on the ABI 3130XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).
Evaluation of reference probes
Ten DNA samples obtained from unaffected individuals were analyzed in a pilot study to evaluate the general peak pattern of the reference probes in a multiplex set-up.
To secure that the self-designed reference probes were normally distributed, they were evaluated using DNA samples from 48 unaffected and 145 affected individuals. Following the MLPA analysis, all fragment-specific peaks, obtained from capillary electrophoresis, corresponding to specific probes were identified by GeneMapper™ (Applied Biosystems, Foster City, CA, USA) applying the Amplified Fragment Length Polymorphism (AFLP) analysis method. It should be taken into consideration that the AFLP algorithm automatically normalizes the peak heights and that this option should be turned off (recommended by Applied Biosystems). Data of peak height and fragment length were imported into a Microsoft Excel spreadsheet for further analysis.
Peak heights of the reference probes were analyzed in a Gaussian Q-Q plot and in a histogram, where all reference probe peaks have to show the same normalization pattern in both affected and unaffected individuals; otherwise, the probe is omitted from further studies 1. If the samples were overloaded, dilution of the MLPA product should be avoided as it may affect the final result. Instead, we suggest changing the injection time or voltage according to recommendations from MRC Holland. If the problem persists, the volume of probes in the probe mix can be decreased as well as the MLPA product before the capillary electrophoresis. The important issue is to treat all samples from unaffected and affected individuals in exactly the same way throughout the MLPA analysis.
Normalization of MLPA probes and CNV calculations
CNV was established in the tumor samples by normalization of the target probe to all reference probes present in the run. Next, standard deviation and mean value of the probe of interest were calculated from the cohort of unaffected individuals to establish the z score in the tumor samples. Normalized peak height ratios for each target probe in both tumor samples and unaffected individuals were calculated as:
, where a is the peak height of the target probe and C1-nis the peak height of each reference probe 1 - n. In this case the normalization of the probe height takes into account the specific peak profile and the ratio of the reference probe in the sample. When assessing the quantitative value of each probe, both the peak height and the peak area can be used. We experienced that the peak height provided the most consistent results (data not shown).
Since the normalized peak heights were normally distributed, the z score of normalized peak heights (x) in tumor samples was found as , where μ and σ are the estimated mean and standard deviation of the normalized peak heights for the population of unaffected individuals. To determine CNV for each probe, the distribution of normalized peak heights can be determined from the group of unaffected individuals and the cut-off values determined as fractils from the normal distribution. When the z score was calculated, the cut-off values could be determined from the standard normal distribution, in which +/- 1.96 is equivalent to a 5% cut-off value.
Moreover, all probes in a region analyzed together can be evaluated in combination to provide an overall measure of gain or loss. If there are k probes in the region, and an individual has the probe z scores z1,...,z k , the combined z score is . Under the null hypothesis that z1,...,z k are standard normal distributed, the z' is standard normally distributed too and can be evaluated in the same way as the z scores for each probe.
The Microsoft Office Excell (version 2007) and JMP® Software were used for analyses.
Analyses of reference probes for MLPA assay
Validation of CNV by RGS8 specific MLPA assay
Our results strongly indicate that the MLPA probes should be designed using exon sequences as templates. The two intron-derived reference probes AFAP1L1a and ACTR1A could not be successfully analyzed due to high rate of variation between unaffected and affected individuals. In contrast, the exon specific AFAP1L1b reference probe presented normal distribution in both unaffected and affected samples enabling to validate MLPA data. The same issue might concern the set of synthetic probes designed to validate previously identified allelic imbalances within the RGS8 in breast tumor samples. The target-specific probes in the RGS8 except the fr.31 probe were designed in non-coding parts of the gene, due to specific unsuitable sequence composition in the target exons. The number of confirmed deletions was relatively lower for these intron - based probes when compared to fr.31. Different repeated sequences as SINE, LINE or Alu repeats may interfere with the probe hybridization lowering the specificity. Likewise, the presence of an yet uncharacterized single nucleotide polymorphism (SNP) may prevent hybridization or ligation of the probes . Another possible reason for discrepancy in validating CNV in breast cancer by MLPA might be due to non-microdissected tumor DNA.
In conclusion, our approach using a group of unaffected individuals to assess the quality of reference probes, and obtain cut-off values for CNV is to our knowledge new, and a solid supplement to existing analysis software. The aim has been to present simple calculations to facilitate the use of the versatile MLPA methodology especially in cancer research.
The Danish Cancer Society (K.P. and B.E.M.), Augustinus Fonden (E.W.) and the University of Aarhus Research Foundation (L.L.H.) supported this work. We thank Tina Kjeldsen for excellent technical assistance.
- Redon R, Ishikawa S, Fitch KR, Feuk L, Perry GH, Andrews TD, Fiegler H, Shapero MH, Carson AR, Chen W: Global variation in copy number in the human genome. Nature. 2006, 444 (7118): 444-454. 10.1038/nature05329.PubMed CentralPubMedView ArticleGoogle Scholar
- Stern RF, Roberts RG, Mann K, Yau SC, Berg J, Ogilvie CM: Multiplex ligation-dependent probe amplification using a completely synthetic probe set. Biotechniques. 2004, 37 (3): 399-405.PubMedGoogle Scholar
- Hsuih TC, Park YN, Zaretsky C, Wu F, Tyagi S, Kramer FR, Sperling R, Zhang DY: Novel, ligation-dependent PCR assay for detection of hepatitis C in serum. J Clin Microbiol. 1996, 34 (3): 501-507.PubMed CentralPubMedGoogle Scholar
- Schouten JP, McElgunn CJ, Waaijer R, Zwijnenburg D, Diepvens F, Pals G: Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res. 2002, 30 (12): e57-10.1093/nar/gnf056.PubMed CentralPubMedView ArticleGoogle Scholar
- Murthy SK, Nygren AO, El Shakankiry HM, Schouten JP, Al Khayat AI, Ridha A, Al Ali MT: Detection of a novel familial deletion of four genes between BP1 and BP2 of the Prader-Willi/Angelman syndrome critical region by oligo-array CGH in a child with neurological disorder and speech impairment. Cytogenet Genome Res. 2007, 116 (1-2): 135-140. 10.1159/000097433.PubMedView ArticleGoogle Scholar
- Buffart TE, Coffa J, Hermsen MA, Carvalho B, van der Sijp JR, Ylstra B, Pals G, Schouten JP, Meijer GA: DNA copy number changes at 8q11-24 in metastasized colorectal cancer. Cell Oncol. 2005, 27 (1): 57-65.PubMedGoogle Scholar
- Nygren AO, Ameziane N, Duarte HM, Vijzelaar RN, Waisfisz Q, Hess CJ, Schouten JP, Errami A: Methylation-specific MLPA (MS-MLPA): simultaneous detection of CpG methylation and copy number changes of up to 40 sequences. Nucleic Acids Res. 2005, 33 (14): e128-10.1093/nar/gni127.PubMed CentralPubMedView ArticleGoogle Scholar
- Eldering E, Spek CA, Aberson HL, Grummels A, Derks IA, de Vos AF, McElgunn CJ, Schouten JP: Expression profiling via novel multiplex assay allows rapid assessment of gene regulation in defined signalling pathways. Nucleic Acids Res. 2003, 31 (23): e153-10.1093/nar/gng153.PubMed CentralPubMedView ArticleGoogle Scholar
- Barbaro M, Cicognani A, Balsamo A, Lofgren A, Baldazzi L, Wedell A, Oscarson M: Gene dosage imbalances in patients with 46,XY gonadal DSD detected by an in-house-designed synthetic probe set for multiplex ligation-dependent probe amplification analysis. Clin Genet. 2008, 73 (5): 453-464. 10.1111/j.1399-0004.2008.00980.x.PubMedView ArticleGoogle Scholar
- Wu K, Li A, Rao M, Liu M, Dailey V, Yang Y, Di Vizio D, Wang C, Lisanti MP, Sauter G: DACH1 is a cell fate determination factor that inhibits cyclin D1 and breast tumor growth. Mol Cell Biol. 2006, 26 (19): 7116-7129. 10.1128/MCB.00268-06.PubMed CentralPubMedView ArticleGoogle Scholar
- Ythier D, Larrieu D, Brambilla C, Brambilla E, Pedeux R: The new tumor suppressor genes ING: genomic structure and status in cancer. Int J Cancer. 2008, 123 (7): 1483-1490. 10.1002/ijc.23790.PubMedView ArticleGoogle Scholar
- Croucher DR, Saunders DN, Stillfried GE, Ranson M: A structural basis for differential cell signalling by PAI-1 and PAI-2 in breast cancer cells. Biochem J. 2007, 408 (2): 203-210. 10.1042/BJ20070767.PubMed CentralPubMedView ArticleGoogle Scholar
- Hansen LL, Yilmaz M, Overgaard J, Andersen J, Kruse TA: Allelic loss of 16q23.2-24.2 is an independent marker of good prognosis in primary breast cancer. Cancer Res. 1998, 58 (10): 2166-2169.PubMedGoogle Scholar
- Hansen LL, Andersen J, Overgaard J, Kruse TA: Molecular genetic analysis of easily accessible breast tumour DNA, purified from tissue left over from hormone receptor measurement. Apmis. 1998, 106 (3): 371-377. 10.1111/j.1699-0463.1998.tb01359.x.PubMedView ArticleGoogle Scholar
- Wehner M, Mangold E, Sengteller M, Friedrichs N, Aretz S, Friedl W, Propping P, Pagenstecher C: Hereditary nonpolyposis colorectal cancer: pitfalls in deletion screening in MSH2 and MLH1 genes. Eur J Hum Genet. 2005, 13 (8): 983-986. 10.1038/sj.ejhg.5201421.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.