MLPA® experimental setup recommendations
Slight differences between different experiments may affect the MLPA peak pattern. It is therefore necessary to include reference samples in each MLPA experiment. It is furthermore recommended to include positive and negative control samples and a no DNA control reaction, which aid in result interpretation and troubleshooting. This document contains recommendations for the use of reference samples, control samples and control reactions.
MLPA is based on the analysis of relative changes in probe signals. Thus, a given sample will not provide the information needed to estimate copy number changes without a reference sample to compare it to. Reference samples are samples in which the target sequences are assumed to have a ‘normal’ copy number. Usually, reference samples are DNA samples obtained from healthy individuals.
It is strongly recommended to use reference samples that have been purified by the same method and which are derived from the same type of tissue as the test samples to minimize non-biological differences between samples and reference runs and to minimize structural variation. Multiple reference samples are needed to estimate each MLPA probe’s reproducibility within each experiment.
It is recommended to aliquot (reference) samples and to store most of the DNA at -20°C. DNA is very stable when dissolved in TE, but storage at 4°C for several years is not recommended. Contamination with micro organisms and/or moulds can deteriorate your reference samples.
Use at least 3 reference samples in each MLPA run (some exceptions exist; see below). When using more than 21 samples, add 1 additional reference sample for every 7 samples. Reference samples should be spread randomly over the sample plate (see Figure 1) to avoid any bias and thus minimize variation.
Figure 1: Random distribution of reference samples enables optimal normalization and estimation of structural variation.
Although the use of reference samples in MLPA is recommended, there are some cases where their use is not essential. Reference samples might not be essential when a large number (>20) of independent samples (different families) are run at the same time AND the chance for each individual probe to be deleted or duplicated in a sample is low (<10%). For example, the SALSA MLPA P036/P070 Telomere probemixes, which contain one probe per telomere arm, can be used without reference samples when sufficient independent samples of individuals with mental retardation are tested. In this case, it is possible to normalize against the median of all samples. When in doubt, make sure to include reference samples.
Positive control samples
Positive control samples are samples with a known deletion, duplication, point-mutation or methylation change and can be used to check whether the entire MLPA procedure, including data analysis, is performed well. Inclusion of positive control samples in MLPA experiments can be very useful, but they are not essential. Similar to reference samples, positive control samples should preferably have been purified by the same method as the samples to be analyzed. It can be difficult to obtain positive control samples and MRC-Holland can not supply them. It is possible to use DNA from cell lines as positive control samples, but please take into account that cell lines may have acquired additional copy number changes, including gains of complete chromosomes.
If you have positive control samples available, it is recommended to aliquot these samples and to store most of the DNA at -20°C. DNA is very stable when dissolved in TE, but storage at 4°C for several years is not recommended because contamination with microorganisms and/or moulds can deteriorate your precious samples.
No DNA control reaction
Control reactions are performed to check whether the procedure was performed well and if the results are reliable. It is recommended to include a no DNA reaction in each experiment, as it will reveal the presence of a contamination in water, MLPA reagents, electrophoresis reagents or capillaries.
The no DNA control reaction (*), is an MLPA reaction on water or TE. In contrast to the MLPA probes, the Q-control fragments at 64-70-76-82 nt do not require ligation while bound to sample DNA for exponential amplification during the PCR reaction. Q-fragments are included in all SALSA MLPA probemixes and their amplification products will be very prominent in the no DNA control reaction.
Reference samples for MS-MLPA are assumed to have a ‘normal’ methylation profile. Reference samples are only used for normalisation of data for copy number determination–not for methylation quantification. However, reference samples are required for the result interpretation of methylation status.
Imprinting diseases (e.g. ME028/ME030/ME031)
For the detection of imprinting diseases, DNA samples can be used that are derived from the same type of tissue (usually blood), but from healthy individuals. Our product description will indicate the expected percentage of methylation for each probe in blood derived reference DNA samples.
Aberrant methylation in tumours (e.g. ME001/ME002)
For the detection of aberrant methylation in tumour samples, it is recommended to use reference samples derived from similar, but healthy tissue. Methylation is cell type specific and can thus vary between different tissues. For most probemixes that are intended for application on tumour derived DNA, MS-MLPA probes are selected that detect sequences which are unmethylated in DNA derived from blood of healthy individuals. It is however possible that one or more CpG sites are methylated, and that some of the MS-MLPA probes show a signal after HhaI digestion, on DNA purified from certain healthy tissues.
In case it is not possible to obtain a DNA sample from similar but healthy tissue, it is possible to use blood derived DNA samples as references. Please note that the sequences detected by the MS-MLPA probes can be methylated in other healthy tissue.
The choice of reference samples for RT-MLPA depends on the application. Sometimes, it is possible to use RNA samples obtained from the same tissue, but from healthy individuals. Alternatively, RNA reference sample(s) can be obtained from the same individual, or cell line, before a certain treatment.
In RT-MLPA control reactions the four small MLPA Q-fragments (64, 70, 76, 82 nt) should be very prominent.
Control reaction lacking reverse transcriptase enzyme.
Use an RNA sample and replace the reverse transcriptase enzyme by a 50% glycerol solution.
Control reaction on DNA.
The majority of the RT-MLPA probes will not generate a signal on contaminating genomic DNA, as the probes target sequences that span an intron boundary, and will thus only generate a signal on cDNA. Of course this does not apply for single exon genes.
Sample type / DNA isolation method / DNA concentration
The most accurate results will be obtained by using sample and reference DNA with a common origin, similar DNA concentrations and the same extraction method. Although MLPA allows a DNA concentration range of 50-250 ng and results are independent of the concentration within this range for high quality DNA samples, it is recommended to keep the sample DNA concentrations more or less equal to minimize any possible variation. Laboratories using a standard 1 µl sample for each MLPA experiment can experience problems when using old, partly evaporated and/or viscous DNA samples. For DNA samples that are less pure, the use of smaller amounts of sample DNA (around 50 ng) is often beneficial.
Control fragments: Up to 10 different control fragments are present in SALSA MLPA probemixes, resulting in amplification products between 64 and 118 nt in length. Control fragments are important as they can indicate that insufficient sample DNA was present (Q-fragments at 64-70-76-82 nt), that sample denaturation was incomplete (D-fragments at 88-96 nt) or that samples switching errors have occurred (chromosome X and Y-specific fragments at 100-105-118 nt).
Control reactions: Various MLPA reactions that can be used to check for reliability of MLPA results, e.g. MLPA reactions without sample DNA.
Digestion control probes: MLPA control probes, only present in MS-MLPA probemixes, which check for a complete digestion of the sample-probe hybrids with the methylation sensitive HhaI restriction endonuclease.
Positive control sample: A sample with a known copy number or methylation alteration or a known (point) mutation.
Mutation-specific probe: Probe which has been designed to detect a certain point-mutation.
Reference probes: MLPA probes detecting sequences that are assumed to have the same copy number in both patient and reference samples. These are present in most SALSA MLPA probemixes and detect sequences in genes/chromosomal regions that are not (known to be) involved in the disease studied.
Reference samples: Samples in which the sequences detected by the probes are assumed to have a normal copy number. In case of MS-MLPA, reference samples are also assumed to have a “normal” methylation profile. Usually, reference samples are DNA or RNA samples from healthy individuals.
(*) Please note that the no DNA control of several of our probemixes is not completely blank. In addition to the Q-fragments, some longer aspecific amplification products are sometimes present. With increasing amounts of sample DNA, both the size of the small Q-control fragments as well as the size of these aspecific fragments will decrease. When 50 ng or more human DNA is used for analysis, the four Q-control peaks at 64-70-76-82 nt will be small or almost absent and any aspecific amplification products will have decreased peak heights that will not influence the results obtained. The cause of these aspecific amplification products is not contamination, but some sequence similarity between two of the (up to 120 different) oligonucleotides in a MLPA probemix. We try to eliminate these aspecific amplification products in all our products but this is often very difficult.
Last updated 31-07-2015