Fourth Annual DNA Grantees' Workshop
Monday, June 23, 2003
MORNING SESSION
Completion of the New NIST Human Mitochondrial Sequencing Standard Reference Material 2392–I for Forensic and Medical Use
Barbara C. Levin
Biography
MR. FRANK: Our next speaker is Dr. Barbara Levin. Barbara is a principal investigator in mitochondrial research and genetic toxicology in the biotechnology division of the National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland. Dr. Levin received her B.A. from Brown University and her Ph.D. from Georgetown University. Her research areas include toxicity of combustion products, mitochondrial research, and genetic toxicology, and also development of an SRM (standard reference material) for sequencing mitochondrial DNA.
This morning she's going to speak to us about the completion of the new NIST human mitochondrial DNA SRM 2392 for forensic medical use.
DR. LEVIN: Good afternoon. You've all seen this a number of times. It seems to come up every time somebody tries to get theirs, and I hope that this will be the last time you see this slide. Levin: Slide 1
Lisa distinguished between projects that have been completed and projects that are new and exciting. Since our project has been completed, I hope that you'll still consider it exciting as well.
There are a number of NIST DNA standards that have been in process over a number of years. SRM 2390, which is called the DNA profiling standard, was first released in 1992 and was updated in 2001. This is the standard to do RFLP (restriction fragment length polymorphism). Levin: Slide 2
SRM 2391–B is the PCR-based DNA profiling standard, and this does your STRs (short tandem repeats). It was first issued in 1995 and revised in 1998 and 2002.
SRM 2392, which is the one that I have been involved in and is the human mitochondrial DNA sequencing standard, was first issued in 1999. With recent validation, its date of expiration has been extended to 2008. Our new work was on SRM 2391–I—that's Roman numeral I—human mitochondrial DNA sequencing standard, and I'll explain what the difference is between that and 2392. I believe you'll hear later this afternoon about the more recent SRM 2395, which is the human Y-chromosome DNA profiling standard.
What do we know about human mitochondrial DNA? Well, most of the people in the audience probably know all of this, but it's circular, double-stranded DNA, and it contains 16,569 base pairs. When I first started this work, the thought was we would just do hypervariable region I (HVI) and hypervariable region II (HVII). Then, I guess because I was kind of innocent and naive, I said, well if we could do that, we could do it all. And that turned out to be a bigger chore than just doing the HVI and HVII regions. Levin: Slide 3
The entire sequence was originally determined by Anderson and colleagues in 1981 [Anderson, S., et al., "Sequence and Organization of the Human Mitochondrial Genome," Nature 290 (5806) (1981): 457–65] and recently was corrected by Andrews and colleagues in 1999 [Andrews, R.M., et al., "Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA," Nature Genetics 23 (2) (1999): 47]. Andrews found the original placenta that had been used by Anderson in 1981 and resequenced it, and with today's state-of-the-art techniques, he was able to find a number of areas in that original placenta that were not correct or that had rare polymorphisms.
Mitochondrial DNA has both a noncoding region, which is the one primarily used by the forensic community, and a coding region. The coding region codes for 2 ribosome RNAs (ribosomal nucleic acids), 22 transfer RNAs, and 13 polypeptides, which are all involved in oxidative phosphorylation. The other thing that one needs to know is that the human mitochondrial DNA code differs from the nuclear DNA universal code.
Why are we interested in mitochondrial DNA? Well, there's a high copy number. You can have from 100 to 1,000 mitochondrial DNAs per cell. It's maternally inherited, and it has a mutation rate that is 10 to 20 times higher than nuclear DNA. And of course, it's highly polymorphic, especially in the HVI and HVII regions, which are considered the control regions. Levin: Slide 4
Human mitochondrial DNA is used by the forensic community for identification purposes and by the medical community for diagnosing diseases. It can be used for SNP (single nucleotide polymorphism) and mutation detection, and it's being used in evolutionary studies and aging research.
Who does it affect? Forensic investigators have to know whether mitochondrial DNA evidence from the crime scene matches that of the suspect. In human identification, you need to know if the mitochondrial DNA from an unknown matches that of the living material relatives, maybe even generations apart. Levin: Slide 5
In patients with mitochondrial DNA diseases, the majority of these diseases are heteroplasmic and of low frequency; therefore, they're very difficult to detect. The question is does one mismatch equal an exclusion or is it a heteroplasmy?
Just to make sure everybody understands what I mean when I talk about heteroplasmy: If DNA in all of the mitochondria of two cells is the same, then it's considered homoplasmic. If, however, some of the DNA within the mitochondria within a cell has a mutation, then it would be considered heteroplasmic. Levin: Slide 6
Since people isolate tissues rather than individual cells, heteroplasmy could even be in the tissue. So you could have maybe one in a thousand—if you have a thousand mitochondria in a cell—that would have the mutation. The question is can you detect that low a level.
So the three SRMs that we're working on are:
- The human mitochondrial DNA standard number one. This contains two normal DNA templates; we call them CHR9947A. It has been available since 1999.
- SRM 2392–I, the new one, contains HL–60 DNA, and this DNA is from an individual that had promyelocytic leukemia. It has been available now since May 2003.
- The last one that we're working on is a heteroplasmic human mitochondrial DNA. Levin: Slide 7
Much of this work has resulted from a 1998 Federal Bureau of Investigation (FBI) directive (a standard numbered 9.5) that said, "The laboratory shall check its DNA procedures annually or whenever substantial changes are made to the protocols, against an appropriate and available NIST standard reference material or a standard traceable to a NIST standard." Levin: Slide 8
So the objective of SRM 2392 and SRM 2392–I was to develop mitochondrial DNA standard reference materials for quality control in amplification, sequencing, medical diagnostics, forensic identification, and mutation and SNP detection. Levin: Slide 9
As I mentioned before, the HL–60 cell culture, which is the culture that's in 2392–I, is a promyelocytic cell line from the peripheral blood leukocytes of a 36-year-old Caucasian female who had leukemia. We were rather interested in the fact that this was not just DNA from a normal individual, because the thought was perhaps we would find something unusual about this leukemic cell line that would lead us to know more about leukemia. However, we'd have to examine more cells in more different leukemia patients to determine this. Levin: Slide 10
Now, the FBI actually asked us to add HL–60 to standard reference material 2392 because they felt that it had well-spaced polymorphisms in the HVI region and HVII control region, whereas 9947A has only two polymorphisms in the HVI region and they're common differences. Levin: Slide 11
Having several polymorphisms allows for better differentiation between the positive control and the test samples. HL–60 also has no insertions in the HVII C-stretch region, which actually occurs between nucleotide pairs 303 and 310. It also does not have the C-stretch problem in the HVI region that we had in the CHR template.
Now, we knew that the CHR template had the C-stretch problem, but we used it primarily because of that. People had asked us to include that one because they wanted to be able to work through the problem.
The mitochondrial DNA standards 2392 and 2392–I include the extracted DNA; all the information you need for doing PCR (polymerase chain reaction), cycle sequencing, and data analysis; and the materials to assess the accuracy of the results. It also includes the information on the sequences of 58 sets of unique primers, which allow any area (small or large) or all of the mitochondrial DNA to be amplified and sequenced. There are no gaps using these primers. Levin: Slide 12
For the medical community, we wanted to provide the necessary quality control for DNA sequence data that would be used to determine the genetic disposition of certain diseases. What you don't want is a false negative, which would give you an incorrect sense of the system, or a false positive, which could lead to a loss of insurance and/or maybe unnecessary surgery.
I won't bore you with showing all 16,000 base pairs, but I will show you the mutations that we found at HL–60 that we didn't find in the previous cases. There are about 10 of them that are different. Most of these actually cause an amino acid change. There's only one here that's silent. Levin: Slide 13
What was interesting is that four of these mutations leads to the mitochondrial disease Leber's hereditary optic neuropathy (LHON), which causes people in their twenties to go blind. Now, since the linkage has been lost between the HL–60 cell line and the individual who provided it, we have no idea if this disease is actually in that individual.
This is a schematic of the human mitochondrial DNA showing all the differences from the Cambridge Reference Sequence. What you see here is the control region, which is where most of the forensic community is interested, and then the rest of the actual genes that are coded. Levin: Slide 14
Inside, I actually put the areas where the LHON mutations that have been associated with the disease occur, so that you can see which of these actually correlate with that disease.
What we have here is the CHR, which is in red, the 9947A, which is in green, and the HL–60, which is in black. We also did two other entire sequences multiple times. We did GM03798, which is another normal, and we did GM10742A, which is from an individual who was diagnosed with LHON.
So my conclusions are that we have two human mitochondrial DNA SRMs that sequence 16,569 base pairs. We use these 58 pairs of unique primers, using the same conditions. (You don't have to change conditions to use the different sets of primers.) We found many differences from the Cambridge Reference Sequence, and we found four differences in HL–60 that correspond to mutations associated with LHON. Levin: Slide 15
Tom Parsons, who worked with us on the interlaboratory evaluation, suggested that maybe these have nothing to do with LHON and maybe it just has to do with these particular mutations that cause this individual to belong to haplogroups J and J2. There are nine haplogroups that distinguish white Caucasians or Caucasian individuals of European descent, and HL–60 seems to belong to J and J2. Looking into this further, we came to the conclusion that those individuals in haplogroups J and J2 may actually be more prone to the LHON disease.
Three laboratories (Bodie Technology, ITT, and Locke Technologies) plus NIST participated in the interlaboratory evaluation of CHR in 1998. The FBI, Armed Forces DNA Identification Laboratory (AFDIL), and Georgia Bureau of Investigation (GBI) participated with us in an interlaboratory evaluation of HL–60.
We've had a number of publications result from this work. The first one is the publication on 2392. The second one was a review article that we wrote in Fresenius' Journal of Analytical Chemistry. The third one was another review article that we wrote in a book on standard reference materials. Levin: Slide 16
This is our most recent article, in which we actually name as coworkers all of the people that worked with us on this interlaboratory evaluation. So in addition to the people that worked with us at NIST, Mike Coble, Tom Parsons from AFDIL, Laura Kienker from the FBI, and Diana Williams from GBI, and Mary Pat Jones from NIH (National Institutes of Health) all worked with us on this. Levin: Slide 17
This article is now out. It's in Mitochondrion, and it has just recently come out.
In addition, we put together a special NIST publication, which is numbered 260–155. It's downloadable from the Internet (http://ts.nist.gov/ts/htdocs/230/232/sp_publications/documents/SP260-155.pdf) and allows people to see the differences between 2392 and 2392–I so that they can decide whether they want to buy both or just one.
I want to acknowledge Hayen Cheng, Lois Tully, Dennis Reeder, who worked on 2392 with me, and the funding that came from the Office of Standard Reference Material in the ATP (Advanced Technology Program) program. I want to acknowledge the people who worked on 2392–I—Diane Hancock, Korin Holland, and Kristi Richie—and the funding that came from the National Institute of Justice through NIST's Office of Law Enforcement Standards. We also want to acknowledge Gettysburg College for partially funding Korin Holland during her sabbatical year at NIST. Levin: Slide 18
These are some Web sites and e-mail addresses that you can use to find out more information about the standard reference materials program. The last one is our interactive Web page that allows you to put in any change you find in the mitochondrial DNA, and it'll tell you whether that change is a silent change that causes an amino acid change or whether it's been associated with a disease. Levin: Slide 19
And I thank you.
MR. FRANK: Thank you very much.