Fourth Annual DNA Grantees' Workshop
Tuesday, June 24, 2003
MORNING SESSION
Quantitation of DNA by qPCR for Forensic DNA Typing: Single and Multiplex Assays for Nuclear and Mitochondrial Genomes
Mark Timken
Biography
MR. TAMBASCO: Our next speaker, Dr. Mark Timken, is with the California Department of Justice. Mark earned his Ph.D. in chemistry from the University of Illinois. Today his topic will be quantitation of DNA by qPCR (quantitative polymerase chain reaction) for forensic DNA typing among single and multiplex assays for nuclear and mitochondrial genomes.
DR. TIMKEN: Thank you.
First of all, I'm really going to focus on singleplex assays today, partly because of time and partly because we've really just started to multiplex and we have more singleplex results. Timken: Slide 1
I also want to begin by acknowledging my colleagues: Christian Orrego and Martin Buoncristiani at Jan Bashinski DNA Laboratory in California. They're the ones who actually wrote the proposal, and I'm the one who has been doing the lab work. They were responsible for getting the money.
Janice Nicklas covered this in a lot of detail yesterday, so I'm not going to go through the background very much. But let me just point out that, again, we're monitoring qPCR (quantitative polymerase chain reaction). It's just PCR, except you're monitoring the amount of product that forms as the PCR progresses in real time. Timken: Slide 2
This is potentially a great method. You can have great dynamic range, potential quantification down to 10 copy numbers or so. It's target specific, so you can target the human mitochondrial, the Y chromosome, or whatever. Of course, what everyone likes in our lab is the opportunity to get away from QuantiBlot, which is time consuming. Here you just do PCR setup, walk away, and come back and get your quants. Timken: Slide 3
We are using two methods to detect the fluorescence of our amplicons. That's typically the way you monitor this. We develop our assays with SYBR Green, which is a nonspecific detection method. All the assays that I'm going to talk about have been developed with SYBR Green, but we've designed them so that we can use TaqMan detection chemistry, which was developed by ABI (Applied Biosystems). Timken: Slide 4
As spelled out in our proposal, we want to develop, identify, and validate qPCR assays for three applications. First is just a simple singleplex nuclear quantitation assay, and Janice described a nice one yesterday with the ALUs. We would use this for databank and reference samples. We're also interested in developing two multiplex assays. One assay would be for the missing persons program, where we would duplex nuclear and mitochondrial quantitation. The idea is that an analyst in our missing persons program could actually extract DNA from some samples and in a single tube, stick a little bit of their extractate in the qPCR machine, come back in 2 hours, and have two numbers—a clear quant and a mito quant—and decide to go with either the autosomal STRs (short tandem repeats), the mitochondrial sequence here, or linear strip arrays. The other assay would be used with sexual assault samples, with nuclear and Y. Timken: Slide 5
I want to talk about what we've done in the way of nuclear targets for potential multiplexes—especially the mito multiplex—and what we've thought about during design. We've thought about and designed single copy targets, that is, one target per haploid genome. We've also thought about multiple copy targets, things like D17Z1 and the ALUs, which Janice and Eric Buel were kind enough to provide us a preprint, so we've worked with these a little bit. But since we wanted to multiplex with the mito, which is multicopy, we decided that it would be simpler to design a multiplex with a single copy target, so we really focused on single copy targets in our design. Timken: Slide 6
We've looked at and considered a few. One is a retinoblastoma susceptibility gene. This was actually used by Marie Allen's group in Sweden, and they published this about the time that we were writing the proposal. We're a little concerned because by sequence, it looks to be pretty conservative with maybe too much sequence homology to nonprimates, but we wanted to consider it.
Because we had some of it around, we looked at beta actin. It's part of an ABI's nonforensic TaqMan kit. Perhaps you know that ABI is coming out with a forensic kit. This is a nonforensic assay.
As we thought about what we wanted in a nuclear target, that is, primate specificity and well-characterized primers, we decided that we didn't want to design a target that had primer binding site mutations, so we thought, well, maybe we should just look at the CODIS STR loci. Maybe those are good enough to design a Taqman assay around. If we could design a good assay, that's what we're interested in. We want to design a quantitation assay for STR input.
We've looked at designing Taqman assays around the STR loci. Here is kind of a recipe for what we've done. The idea is that we use known kit primers for the STR loci. Actually, Promega Corporation has published them but ABI hasn't. Timken: Slide 7
We've done this for all 13 CODIS loci for different kits, and what you find is that the THO locus is a pretty good locus for a number of reasons. It's short, which should make the PCR efficient, it's not very polymorphic, and there's plenty of room for design. We've actually designed a number of Taqman assays based on Promega- and THO-type primers. We've also designed assays where we have used one of the primers and stuck another primer on the side. Some of them are big enough where you can design kind of a partial assay based on half an STR. Timken: Slide 8
We've looked at both full and the partial assays, and we basically ran them all under identical conditions. For now at least, we've found that we're going to stick with these putative ABI-type primers. We've modified them a little bit. Timken: Slide 9
We work with SYBR Green first to see if it looks good. If it looks good, then we work with Taqman. Timken: Slides 10–13
We've done specificity studies comparing, say, THO1, beta actin, and retinoblastoma, and we've actually looked at ALU (the Nicklas-Buel primer set) a little bit. We've looked at this menagerie of DNA with 25 nanograms separately of calf, mouse, cat, dog, rockfish, rat, rhesus monkey, orangutan, gorilla, chimpanzee, and human DNAs. Not surprisingly, THO1 is very primate specific, as is ALU. They both have very nice specificity. Timken: Slide 14
RB1 retinoblastoma is actually pretty specific for most DNAs, except for cow, cat, and dog. Cow DNA actually amps up with retinoblastoma just as well as human DNA, so we don't think that RB1 is a great forensic target. Beta actin, on the other hand, is not very specific, as one might imagine.
So everything works pretty well and we're pretty happy with this. We're working with this in the Taqman approach now and optimizing that for our nuclear target. However, I want to mention one surprise here, and we have an explanation for this surprise, but someone out there might have a better explanation, so you can share that with me later, or sooner if you want. Timken: Slide 15
One of the things that you can do with the SYBR assay is you can run a melt curve. That is, after you've gone through 40 cycles on your assay on your instrument, you tell the machine at the end of the last extension to ramp up to 95°C. As you ramp up to 95°C, you'll go through a melt, and when the amplicon melts, you should get a single peak if there's a single amplicon.
This usually happens, but for this assay we always get a shoulder. We get a single peak and a shoulder. This was kind of unexpected for an assay that seems to work so well. We don't see anything in the negative control, which is this green line, and we've done gel studies and we don't see any evidence for primer dimers.
The results are very reproducible. There's always a major peak, and the melting temperature, approximately 81°C, of that major peak never changes. Timken: Slides 16 and 17
Well, anyway, right now we're working on optimizing Taqman and once we get a Taqman assay that we're really happy with, we're going to work on multiplexing. Timken: Slide 18
I'm not going to spend much time on the mito assays, but we're working on three of them. One is M. Allen's, and we've got two that we've developed. They all work seemingly very well, so we're going to pick the one we like best and characterize and optimize it. Timken: Slide 19
I want to end by thanking Janice Nicklas and Eric Buel for their preprint that allowed us to do some work with the ALUs and especially the National Institute of Justice, obviously, for funding, because if we didn't have their money, we might not be doing this project. Thank you for your attention. Timken: Slide 20