Mary Rennick and Dr. Brent Nielsen, Department of Microbiology and Molecular Biology

Beginnings

This research focused on the search for the probable mitochondrial DNA (mtDNA) mutation in one family affected with fibromyalgia and lactic acidosis. The hypothesis was that there is mixture of normal and mutated mtDNA, called heteroplasmy. I first obtained the four patients’ consent to participate in the study. The four patients were between the age of 19 and 25 years-­‐two males and two females. They donated urine samples to be used for DNA analysis. I isolated DNA using a Qiagen DNA extraction kit and obtained concentration measurements of 8-­‐ 10 nanograms per microliter using the nanodrop spectrophotometer in the RIC facility. I used the Taq polymerase reaction mixture for PCR amplification of the mtDNA using 17 sets of primers. These 17 segments of overlapped to provide coverage of the whole circular mitochondrial genome. I ran these reaction samples on DNA gel electrophoresis for about an hour to check that I had plenty of amplified DNA. I used the restriction enzyme, Surveyor Nuclease, to check for any mismatched mutant/nonmutant DNA that could have formed during the PCR reaction. I was so excited when I saw that one of the segments was cut by the restriction enzyme. I thought I had found the mutation location in the cytochrome b gene.

Sanger Sequencing

However, to verify that there was indeed heteroplasmy, I needed to sequence the DNA. Cleaning the DNA in preparation for sequencing was a bit of a challenge. Sometimes I ended up with a very low concentration of DNA after the PCR cleanup kit was used. I was able to buy a new kit and prepare some samples for Sanger sequencing in the Sequencing Center in the Widtsoe building. A couple of the samples yielded long reads, but only about one in three bases looked to be accurate. I could see that I would need many samples and a lot of money to sequence the entire mitochondrial genome of both the patient and the control subject.

454 Sequencing

Since I wanted to have very accurate and long PCR amplification in preparation for 454 sequencing, I decided to use a more accurate DNA polymerase called Phusion. Although this polymerase has a reputation for great accuracy, it is temperamental. I optimized the PCR reaction by using a gradient of temperatures and found the ideal temperature. However, I was still getting low yields of amplified DNA. I changed the buffer to be more suited for high GC (guanine-­cytosine) content DNA. It worked much better, but there were different lengths of DNA-­‐as I could see when checking the DNA on a gel. So, I investigated the possible cause of this, and did optimization using different magnesium and DMSO concentrations. This did the trick. There was ample DNA and only a few extra bands on the 17 segments of amplified DNA.

After cleanup of the PCR reaction samples, I submitted the DNA to the Sequencing Center to be run on the 454 instrument. After the 454 sequencing run, Dr. Ed Wilcox sent me a file with the 60,000 reads per sample. I was told that I could use the 454 software to check for variance (heteroplasmy). I worked on this with my computer programmer husband for a couple of months. We couldn’t get it to work. We then found out that this software would only work on a machine like the supercomputer-­‐with at least 40-­50 gig of RAM, and a lot more processing power than was available to me.

Assembly of 454 reads contradicts my hypothesis

Fortunately, Geneious software training was available at that time. I took advantage of it and learned how to map the reads from my four samples to the mitochondrial reference genome downloaded from the NIH website. However, finding the proper assembly parameters for detecting variance in the DNA took another several weeks of trial and many errors. Finally, I was able to see that extensive trimming of the reads before alignment reads reduced the errors in sequencing that often occur at the ends of the reads.

To make a long story short, I assembled four segments of the mitochondrial genome-­‐about 3000 base pairs in length over the same region of the cytochrome b gene. There was no heteroplasmy. No variance. I checked my results by having the principal subject and his control sister submit DNA to familytreedna.com to get the second opinion. Their results were the same as mine-­‐with the variants from the reference sequence listed. I queried this sequence in the human mitochondrial database (HmtDB) that has over 10,000 mitochondrial genomes organized into haplogroups of similar sequences. The results were interesting. There was not another haplotype like it and several of the base changes were never seen before.

Different hypotheses and further research

So, although I was not able to find heteroplasmy as I hypothesized, I did find unusual base changes that may be responsible for the disease phenotype. Perhaps these base changes affect the translation and shape of the protein. I will continue to investigate these possibilities.

Thank you for the ORCA grant and the opportunity to have gained so much experience.