Genotyping Single Nucleotide Polymorphisms in the Mitochondrial Genome by Pyrosequencing

Summary

Pyrosequencing assays enable the robust and rapid genotyping of mitochondrial DNA single nucleotide polymorphisms in heteroplasmic cells or tissues.

Abstract

Mutations in the mitochondrial genome (mtDNA) have been associated with maternally inherited genetic diseases. However, interest in mtDNA polymorphisms has increased in recent years due to the recently developed ability to produce models by mtDNA mutagenesis and a new appreciation of the association between mitochondrial genetic aberrations and common age-related diseases such as cancer, diabetes, and dementia. Pyrosequencing is a sequencing-by-synthesis technique that is widely employed across the mitochondrial field for routine genotyping experiments. Its relative affordability when compared to massive parallel sequencing methods and ease of implementation make it an invaluable technique in the field of mitochondrial genetics, allowing for the rapid quantification of heteroplasmy with increased flexibility. Despite the practicality of this method, its implementation as a means of mtDNA genotyping requires the observation of certain guidelines, specifically to avoid certain biases of biological or technical origin. This protocol outlines the necessary steps and precautions in designing and implementing pyrosequencing assays for use in the context of heteroplasmy measurement.

Introduction

The mitochondrial genome exists in the form of small (16.5 kb) circular molecules (mtDNA) present in the innermost compartment of the mitochondria named the matrix and encodes 13 subunits of the mitochondrial respiratory chain, as well as the tRNAs and rRNAs necessary for their translation in situ by the mitochondrial ribosome¹. This genome represents approximately 1% of all the proteins necessary for mitochondrial function, the remainder of which are encoded by the nuclear DNA (nDNA). It is commonly assumed that mitochondria are derived from an endosymbiotic fusion event between an alpha-proteobacterial ancestor and an ancestral eukaryotic cell. Once this hypothetical symbiosis took place, the genetic information of the mitochondria was gradually transferred to the nucleus over eons, which explains the aforementioned compactness of the mtDNA when compared to the genomes of modern cyanobacteria². Such a transfer of genes is most strongly evidenced by the existence of long stretches of nDNA that are highly homologous to the sequences found in mtDNA. These nuclear mitochondrial sequences (NUMTs) are a common source of misinterpretation during genotyping, and certain precautions must be taken to avoid nuclear biases when genotyping mtDNA³ (Figure 1A).

Another distinctive feature of mtDNA is that its copy number varies depending on the cell type, numbering anywhere from tens to thousands of copies per cell⁴. Owing to this multi-copy nature, mtDNA can harbor a wide range of genotypes within a single cell, which can result in a more continuous distribution of alleles in contrast to the discrete alleles associated with nuclear genes when considering the zygosity of chromosomes. This heterogeneity of mitochondrial alleles is referred to as mitochondrial heteroplasmy, which is typically expressed in the percent prevalence of a given mutation as a proportion of the total mtDNA in a given cell. Heteroplasmy can be contrasted with homoplasmy, which refers to a unique species of mtDNA being present across a cell.

Measuring mitochondrial heteroplasmy is of particular interest when quantifying the proportion of mtDNA molecules harboring pathogenic variants. Such variants come in the form of single nucleotide polymorphisms (SNPs), small indels, or large-scale deletions⁵. Most humans are heteroplasmic for pathogenic variants; however, they do not exhibit any clinical phenotypes, which often only manifest at higher heteroplasmy levels of pathogenic mtDNA in a phenomenon referred to as the threshold effect⁶. While the values associated with pathogenicity are highly dependent on the nature of the pathogenic mutation and the tissue in which it occurs, they typically lie above 60% heteroplasmy⁷.

There are several research areas in which mitochondrial genotyping is common. In the medical field, testing for or quantifying mtDNA mutations can serve as a diagnostic criterion for mitochondrial diseases, many of which have mtDNA aberrations as their origin⁵. In addition to the study of human pathogenic mutations, the prevalence of animal models harboring pathogenic SNPs in the mtDNA is likely to increase, given the recent advent of mitochondrial base editing enabled by mitochondrially targeted DddA-derived cytosine base editors (DdCBEs)⁸ and TALE-based deaminases (TALEDs) for adenine base editing⁹. This approach will be instrumental in understanding the interplay between aberrant mitochondrial genotypes and the resulting dysfunctions. There is also ongoing scientific research into remodeling the mitochondrial genome for ultimate use as a therapeutic strategy in human mitochondrial diseases via an approach known as heteroplasmy shifting. This field of research primarily involves directing mutation-specific nucleases to the mitochondrial matrix; this results in the preferential degradation of pathogenic mtDNA, leading to rescues in phenotype^10,11,12,13. Any experiments involving the remodeling of the mitochondrial genotype require a robust quantitative method to assess heteroplasmy shifts.

A wide variety of methods are used to genotype mtDNA, and these vary according to the nature of the mutation. Next-generation sequencing (NGS) methods are more precise when it comes to quantifying SNPs in mtDNA; however, these methods remain prohibitively expensive for the routine quantification of mitochondrial heteroplasmy, particularly if the number of samples is small. Sanger sequencing can also allow for the detection of SNPs; however, this approach is not quantitative and often fails to detect low levels of heteroplasmy or can be inaccurate when estimating high heteroplasmies. Pyrosequencing, as an assay that involves minimal preparation and enables the rapid quantification of heteroplasmy for any mtDNA sample, is proposed as an apt compromise between these two extremes. This method has been used routinely to quantify mitochondrial SNPs by numerous researchers in various contexts, including forensic analysis^14,15, clinical diagnosis¹⁶, or the genotyping of mtDNA from single cells¹⁷.

This assay involves a first PCR preamplification step of a region flanking the SNP in the mtDNA, which is followed by a sequencing-by-synthesis assay using one strand of the previously generated amplicon. One of the two primers used in the preamplification step must be biotinylated on the 5' end, which will enable the pyrosequencing apparatus to isolate the single strand of DNA to be used as template for the sequencing reaction. A third sequencing primer is then annealed onto the retained biotinylated strand, which allows for nascent DNA synthesis as deoxynucleotides to be dispensed in a predefined order into the reaction chamber. The pyrosequencer records the amount of each base incorporated based on a luminescent readout, allowing the relative quantification of mutant and wild-type mitochondrial alleles upon DNA synthesis (Figure 1B). The luminescence is generated by a luciferase enzyme, which emits light in the presence of ATP that an ATP sulfurylase synthesizes de novo at each incorporation event from the pyrophosphates released by each nucleotide. These two reactions can be summarized as follows:

1. PP_i (from nucleotide incorporation) + APS → ATP + sulphate (ATP sulfurylase)

2. ATP + luciferin + O₂ → AMP + PP_i + oxyluciferin + CO₂ + light (luciferase)

Detecting adenine bases by the pyrosequencer without ATP cross-reacting with the luciferase in the second reaction is a challenge. However, this is solved by using an adenine analog for DNA synthesis, namely dATPαS. Despite not being a perfect substrate for luciferase, it produces a stronger luminescence compared with the three other nucleotides, which is digitally adjusted by the pyrosequencer and set to a factor of 0.9. Due to this inherent variability, it is suggested to avoid sequencing adenine at the SNP position (see the discussion for further details).

The following protocol details the method of mtDNA heteroplasmy assessment by pyrosequencing and outlines the necessary precautions in designing the amplification primers to avoid biological or technical bias when genotyping SNPs in mtDNA. The latter involves digitally surveying and selecting the primer sets, optimizing the preamplification PCR, and finally, sequencing and refining the assay. Two applied example assays are demonstrated: first, the optimization of the most common human pathogenic variant m.3243A>G¹⁸, and second, the genotyping of mouse embryonic fibroblast (MEF) cells that have undergone heteroplasmy shifting using technologies developed at the Minczuk laboratory in Cambridge^{10,11,12,19,20,21,22}.

Protocol

Informed consent was provided for the use of the human 3243A>G cybrid cells and the immortalized m.5024C>T MEFs used in this study. Ethical approval was not required in this instance as the patient cells were not collected at the University of Cambridge. The use of human fibroblasts may, however, require ethical approval. It is highly recommended to follow best practices for PCR setup when preparing the sample DNA for pyrosequencing. Frequent amplification using identical primers can lead to amplicon contamination and introduce bias to the subsequent genotyping if strict separation between the pre-PCR and post-PCR areas is not observed. The pipeline presented here uses specific equipment from a sole manufacturer; the details can be found in the Table of Materials. The primer design for the PCR can be performed manually if so desired; however, it is recommended to use existing software for this purpose (see the Table of Materials).

1. Pyrosequencing primer design and assay selection

1. Obtaining primer candidates with software
   1. Obtain an mtDNA sequence file for the species being genotyped, and identify the position of the SNP. Ensure that the reference sequence used employs the appropriate base numbering for the mutation studied so that the position of the SNP can be easily identified.
   2. Copy 1,000 base pairs upstream and downstream of the SNP site, and paste the truncated sequence into the software intended for the pyrosequencing primer design (see the Table of Materials).
   3. Set the analyzed SNP base as the target of the pyrosequencing assay in the software by highlighting the polymorphic base and right-clicking on and then selecting set target region.
   4. Press the play icon in the top right-hand corner of the interface to launch primer selection.
   5. Wait for the software to now automatically generate primer trios: two primers for preamplification of the template DNA and a third primer for sequencing by synthesis in the pyrosequencing machine. Retain all the primer sets by right-clicking on and selecting copy all primer sets.
2. Selecting the optimal primer sets from the generated list
   1. Retain the primer sets with the highest quality score based on the software output, which provides primer sets in descending order based on the quality score. If possible, omit primer sets with a score below 80.
   2. For the retained primer sets, identify and copy the amplicon produced by the amplification primers.
   3. Align the amplicons produced with the entire genome of the organism to be genotyped using NCBI Blast by using the online submission portal at https://blast.ncbi.nlm.nih.gov/Blast.cgi. From the dropdown menus during submission, select the following:
      1. Select Database | Genomic + Transcript databases.
      2. Select Human or Mouse in the dropdown menu, depending on the SNP analyzed.
      3. Select Optimize for | Somewhat similar sequences (Blastn).
   4. (Optional) If genotyping an organism other than a mouse or human, select the following during submission:
      1. Select Database | Standard database.
      2. Select RefSeq Representative genomes in the dropdown menu
      3. Select Optimize for | Somewhat similar sequences (Blastn)
   5. When possible, omit any amplicons that have perfect homology to the mtDNA amplicon, particularly if the primer binding regions are perfectly homologous (see the discussion).
      NOTE: The BLAST alignment tool will return any sequence with considerable homology to the pre-amplification amplicons. This allows for the detection of the NUMTs described in the introduction, which if not accounted for, can induce a bias as they are co-amplified with the mitochondrial DNA.
   6. Order the oligos obtained via this pipeline. Make sure that 5' biotin modification is added to the correct amplification primer during the synthesis order, such that the sequencing primer will be complementary to the biotinylated strand.
      ​NOTE: The choice of sequencing primer is more flexible than the amplification primers, and it is recommended to have at least one base separating the 3' end of the sequencing primer and the variable position.

2. Preamplification PCR optimization

1. Prepare the DNA samples by extracting the total genomic DNA using an appropriate method.
   NOTE: This will largely depend on the number of cells being genotyped and their origin. If isolating DNA from single cells using a lysis buffer containing proteinase K, the sample must be denatured at 95 °C for 10 min so as not to disrupt the polymerase during the subsequent PCR.
2. PCR setup and thermal block settings
   1. Set up the PCR with a high-fidelity polymerase of choice. Prepare a Master Mix for four reactions. As only 10 µL of PCR reaction are required for a pyrosequencing run, prepare 25 µL of PCR reactions (to allow for a technical repeat if necessary). Use 40 cycles of PCR with approximately 10 ng of genomic DNA as the starting material with 1 µM of both the forward and reverse primers.
   2. Set the extension time according to the manufacturer's specifications for the polymerase used, taking into consideration the length of the chosen preamplification primers.
   3. Use a thermal cycler with variable annealing temperature settings. As the annealing temperature can differ depending on the primer sequence, the salt content of the polymerase buffer, and other factors, program four different annealing temperatures into the thermal cycling program.
      NOTE: The worked example in the representative results used the following temperatures: 55 °C, 60 °C, 65 °C, and 70 °C.
   4. Split the Master Mix into four 200 µL PCR tubes, and set them to run at each of the four selected annealing temperatures in the thermal cycler for 40 cycles.
3. Preamplification band visualization
   1. During the PCR run, prepare a 2% (w/v) agarose gel in 1x TBE with SYBR Safe or ethidium bromide as a visualizing agent.
      NOTE: A higher percentage gel is recommended for visualization as the amplified fragments are usually short, typically ranging from 100 to 500 base pairs.
   2. After the PCR in step 2.2 is complete, mix 10 µL of the reaction with an appropriate amount of DNA loading buffer, and run on the 2% agarose gel at 7 V/cm for approximately 45 min.
   3. Visualize the resulting DNA fragments on a UV transilluminator.
      ​NOTE: The thermal cycling conditions selected for the preamplification step should produce a single clean band of the expected size. Lower annealing temperatures can occasionally result in off-target amplification, which can introduce biases to the subsequent pyrosequencing.
   4. Select the lowest annealing temperature that produces a clean band of the correct size for subsequent amplifications.
   5. (Optional) Excise a band of the expected size, purify using a gel extraction kit of choice, and analyze by Sanger sequencing to confirm the approximate heteroplasmy of the sample as a reference.

3. Instrument setup and run

NOTE: Once the PCR step in the previous section is optimized, the next step involves programming the pyrosequencer with the correct nucleotide sequence to analyze for the specific SNP. This involves entering 10 bases directly downstream of the 3' end of the sequencing primer. This is detailed in the following section.

1. Assay configuration
   1. Open the run software provided with the pyrosequencer, and select New Assay in the top-left corner of the interface.
   2. Select the Allele quantification assay template.
   3. Input the Sequence to be analyzed into the corresponding box by typing in the nucleotides to be incorporated directly downstream of the sequencing primer, which should be upstream of the SNP of interest. For the variable position, denote the two possible bases separated by a slash (e.g., A/T).
   4. Press Generate dispensation order, and let the software automatically determine a suitable order for the nucleotides to be dispensed in the sequencing-by-synthesis reaction.
   5. Save and provide a name for the assay.
2. Reagent preparation and storage
   1. Dilute the sequencing primer ordered to 4 µM in the Annealing Buffer provided in pyrosequencer reagent kit.
      NOTE: The sequencing primer can first be diluted in water to 100 µM as a stock solution and subsequently further diluted in Annealing Buffer to 4 µM as needed.
   2. Enzyme and substrate preparation and handling
      1. When first unboxing, redissolve the lyophilized enzyme and substrate as per the manufacturer's recommendations. Store at −20 °C when not in use.
      2. Upon subsequent use, thaw the enzyme and substrate vials from −20 °C.
         NOTE: All the other reagents in the provided kit can be kept refrigerated at 4 °C. Diluted sequencing primers can also be kept at 4 °C.
   3. Place the required amount of each reagent on ice.
   4. Keep the remaining components required, namely the streptavidin-coated magnetic beads, absorber strips, and pyrosequencing disks, at 4 °C.
3. Run execution
   NOTE: When an assay is first designed, it must be calibrated using PCR products of known heteroplasmy, which ensures that the assay can accurately distinguish heteroplasmies. Researchers can use mixtures of PCR products of known heteroplasmy as standards. Samples can also be verified by other methods mentioned in the introduction and the discussion, notably NGS.
   1. Dilute the DNA to be analyzed to 5 ng/µL. Particularly in the first run, be sure to include samples of known heteroplasmy as a reference and/or wild-type samples.
   2. Perform a presequencing PCR of 5 µL of diluted DNA in 25 µL reactions using the amplification primers and parameters identified in section 1 and section 2. Perform technical PCR replicates for each sample.
      NOTE: The preamplification PCR can be stored short term at 4 °C or long term at −20 °C before proceeding to sequencing.
   3. Run file setup
      1. Select New Run in the top-left corner of pyrosequencer software.
      2. The pyrosequencer can simultaneously sequence 48 separate preamplified reactions, with an empty square representing a single sequencing well on a pyrosequencing disk. Load the assays configured in section 3.1 by right-clicking on a square and selecting load assay. If required, sequence up to four separate assays with different sequencing primers.
      3. Set the primer dispensation mode to Automatic to automatically assign an injection chamber for each sequencing primer used for the run.
      4. Set Run mode to Standard unless running four different types of assays on one sequencing disk.
      5. Ensure the number of assays on the run template file matches the number of PCR reactions being amplified.
         NOTE: If running more than 48 samples, additional run files will need the be set up.
      6. Save the run files to a USB drive.
   4. Priming the pyrosequencer
      1. Press the Cleaning button on the main touch screen of the device, and clean all injectors with high-purity water following the instructions on the screen. When inserting the absorber strip into the machine, ensure that the ends meet in the "9 o'clock" position (directly left of the center).
      2. Plug in the USB stick with the runs set up in step 3.3.3, and load the run files defined in step 3.3.3.
         NOTE: The run files must be saved outside any directory or folder on the USB or they cannot be read by the machine.
      3. Follow the instructions on the device to load and prime the reagents as required into the corresponding injectors on the machine.
   5. Sample disk preparation and assay launch
      1. Allow the streptavidin-coated magnetic beads to equilibrate to room temperature.
         NOTE: These beads will enable the device to capture the biotinylated strand from the preamplification PCR step. They need to be purchased separately from the kit.
      2. Load 3 µL of beads into the wells defined in the run file from step 3.3.3.
      3. Load 10 µL of each PCR reaction to be sequenced into the corresponding sample, and pipette up and down to mix the sample with the magnetic beads. Avoid bubbles if possible.
      4. Look for the indication in the primed pyrosequencer that the disk can be loaded into the machine. Unscrew the plate holding nut before aligning the loaded sample plate with the metal pin in the disk compartment.
      5. Once the plate is firmly screwed into the plate compartment, launch the sequencing run by pressing the start button on the touchscreen interface.

4. Result acquisition

1. Once the run is complete, remove the USB drive from the machine, and plug it back into the computer running the pyrosequencer software. While proceeding with the following steps, proceed with cleaning the pyrosequencer following the prompts if it is the final run of the day.
2. Wait for the newly generated run file to appear on the USB drive, and double-click on the run result file. This automatically analyzes the luciferase output of each well upon nucleotide incorporation and quantifies the mitochondrial allele of interest defined during the assay configuration in section 3.1.
3. Look for the following color scores shown by the software based on the quality of the reads. Blue indicates an optimal run, yellow a run with a warning, and red a failed run. To save the results, select Reports in the top window | Full report to generate a .pdf file containing the pyrograms and results from each well of the run. Alternatively, download the results in other formats under the Reports tab.

Results

This section presents an example optimization of a pyrosequencing assay for a human pathogenic mtDNA mutation, as well as sequencing data from the genotyping of heteroplasmic (m.5024C>T) mouse embryonic fibroblasts (MEFs) treated with mitochondrial zinc finger nucleases (mtZFNs). Optimizing the assay for human cells and comparing two different assays demonstrates how to select the most accurate one, whereas genotyping genetically modified MEF cells in the second example serves as an applied example of detecting heter...

Discussion

A critical aspect for the success of the protocol is avoiding contaminations, particularly when using low amounts of starting material. It is recommended to use a UV hood and filtered pipette tips when preparing the samples wherever possible, as well as to keep the preamplification and post-amplification areas separate. Blank measurements and samples of known heteroplasmy (such as wild-type DNA) should always be included to be used as benchmarks to check for technical or biological bias.

A no...

Materials

Name	Company	Catalog Number	Comments
KOD Hot Start DNA Polymerase	Sigma-Aldrich	71086
PyroMark Assay Design 2.0	QIAGEN
Pyromark Q48 Absorber Strips	QIAGEN	974912
PyroMark Q48 Advanced CpG Reagents (4 x 48)	QIAGEN	974022
Pyromark Q48 Autoprep	QIAGEN	9002470
PyroMark Q48 Cartridge Set	QIAGEN	9024321
Pyromark Q48 Disks	QIAGEN	974901
Pyromark Q48 Magnetic beads	QIAGEN	974203
PyroMark Q48 Software License (1)	QIAGEN	9024325