Barbara C. Levin,¹ Haiyan Cheng,² Mitchell Holland³ and Dennis J. Reeder^1
1 Biotechnology Division, National Institute of Standards and Technology, Gaithersburg, MD
² GEO-CENTERS, Inc., Newton Center, MA
³ Armed Forces DNA Identification Laboratory, Armed Forces Institute of Pathology, Rockville, MD

× Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø

Sequence analysis of mitochondrial DNA (mtDNA) is being used for human identification especially in those cases where nuclear DNA is highly degraded or non-existent. The distinction between individuals is primarily based on the considerable sequence variation found in the two hypervariable regions (HV1 and HV2) located in the non-coding displacement loop (D-loop). PCR amplification and sequencing of the PCR product of these regions have been generally successful and useful for human identification. In some individuals, however, either or both the HV1 and HV2 regions contain long homopolymer stretches of only cytosine (C) bases (C-Stretch region) which interfere with sequence analysis in the region following the C-Stretch. The objective of this research was to determine the reason for this interference. Effects of changing DNA polymerase, PCR additives and other parameters, such as denaturation and annealing temperatures, cycle number, enzyme concentration, dCTP and dNTP concentrations, magnesium concentrations, and adding formamide were examined. In all cases, the ability to sequence beyond the C-Stretch region was not improved. To determine whether the C-Stretch problem arises during the PCR amplification or cycle sequencing, we used capillary electrophoresis (Bio-Rad Laboratories, Inc., Richmond, CA) to analyze some of the amplification products to determine if they contained more than a single DNA product. The results showed that the control DNA (no C-Stretch problem) generated a symmetric peak indicating a single PCR product; whereas, each of the two DNA products with the C-Stretch problem induced a shoulder peak indicating more than one product. These results suggested that the C-Stretch problem may be due to a heteroplasmic mixture of mtDNAs, each containing C-Stretches of different lengths. The HV1 region was amplified by PCR and the PCR product was cloned into the M13mp18 vector which was used to transfect E. coli TG-1 cells. Using the cloned DNA, we were now able to obtain good sequence data from the HV1 region beyond the C-Stretch. The sequence of these clones showed different numbers of Cs. In 19 clones, four had 11 Cs, twelve had 12 Cs and three had 13 Cs. The questions that then arose were (1) Is the heteroplasmy due to errors made during the PCR process; or, (2) Does heteroplasmy exist in the mitochondrial DNA of the donor cells? To answer question (1), we reamplified the cloned PCR product DNA and found that the C-Stretch sequencing problem reappeared. To answer question (2), we needed a large quantity of cells in order to isolate and directly clone the mtDNA from the cells rather than the PCR product. White blood cells from the original donor were transformed with the Epstein-Barr virus and immortalized as a tissue culture cell line. MtDNA was isolated from 4 x 10⁸ tissue culture cells, cut with the restriction enzymes Sac1 and Kpn1, separated on a gel; the fragment containing the C-Stretch was cloned into the M13mp18 vector which was used to transfect E. coli TG-1 cells as described above. The DNA from 48 plaques was sequenced and 19 plaques were found to contain C-Stretches. Two had 10 Cs, six contained 11 Cs, nine contained 12 Cs and two had 13 Cs. We conclude that good sequence data following the C-Stretch can be obtained from cloned DNA. The inability to sequence through this region is due to a pool of heteroplasmic species containing different numbers of cytosine residues in the C-Stretch. We found that this heteroplasmy may be generated by PCR of the cloned DNA, but it is also present in tissue culture grown donor cells that have never had PCR.