Fourth Annual DNA Grantees' Workshop

Tuesday, June 24, 2003

AFTERNOON SESSION

Charge Tags as Electronic Labels for DNA Microchip Testing
Victor W. Weedn
Biography

(Note: The PowerPoint presentation that supplements Dr. Weedn's transcript is not available.)

MR. COFFMAN: Our last speaker is Dr. Victor Weedn, Esquire, I guess. He's both a doctor and an attorney. Dr. Weedn is the director of Biotechnology and Health Initiatives and principal research scientist at Carnegie Mellon University . His talk today is on charge tags as electronic labels for DNA microchip testing.

DR. WEEDN: Well, I want to thank NIJ for the grant and for the opportunity to have me come before you and speak. They have clearly made a small investment in what could only be considered the development of high-risk technology. Admittedly, I am not as far along as some of our other speakers who've demonstrated quite tremendous technology.

In this particular case, I am looking into dielectric relaxation spectroscopy (DRS) as a means to detect a DNA species. It uses whole molecule resonance, and I believe that this is the only investigation of this type in any lab in the world right now.

In a previous publication, NIST has indicated that there are three major peaks for resonance. We're specifically looking in the microwave area. This is quite a different kind of technology for me. It's a real stretch. It's really not like typical electronic engineering at all. It's really quite difficult. It requires very specialized electronic technology. Even the connections with the wires to the devices have to be made specially.

Last year I showed you or demonstrated how the molecules would resonate, going back and forth between alternating polarity. It gets to a point at which it doesn't keep up. The transition phase, then, between when a molecule can and can't keep up is a particular frequency. That's what we try to measure here.

Now, the typical curve is hardly useful at all for diagnostic purposes, and that's why I think other people have really not looked at this area as a way to do DNA diagnostics. We're trying to engineer the target so that it in fact can be detected, by taking a positive and negative charge, as opposed to, say, fluorescent tags, as reporter molecules. When these bind together, all of a sudden out of nowhere we have a great big dipole moment.

This, of course, goes straight to the issue of fragment size. In this respect, we look directly at the STR species or the allele fragment size directly, as opposed to a secondary measurement, such as mobility through a gel.

So the bottom line is this: With just two electrodes and an electric field, we could interrogate a molecule for the DNA that we were looking for. And because of this, it's an automatic or an instantaneous result. We don't have to wait for the gel electrophoresis.

It's also incredibly inexpensive. Even the cartridge required for two electrodes would be incredibly inexpensive. It would be less expensive than the other three alternatives presented here this afternoon. Furthermore, the hybridization would be in solution and there are certainly advantages to that.

As we made our way into this technology, the first thing we did was just to see if we could get results. You know, reproduce what the literature had. We have the water and we have two different kinds of molecules, and we indeed were able to reproduce what the literature had. Then we started experimenting with different molecules at different sizes. With the different size molecules came different size signals and different frequencies. Next, we started to look at concentrations. Do we still see this effect? The answer was yes. Now, one molar is a very big number. It's certainly far beyond what we would normally look at. But this is what we're having to deal with right now. It's an insensitive technology. There's much to be said about that.

Along the X axis, we put the size of the molecule in terms of number of carbon units and plotted frequency of the response, as well as what's called "E," the dielectric loss figure. We then can relate the size of the molecule to those indicators. With that, we can begin to understand our system and predict where it is we need to do our frequency sweeps.

It led us to this equation by Einstein. He concluded that the relaxation frequency is related to the dipole rotational correlation time. Imagine starting with a whole bunch of molecules oriented in the same direction. In time, they will start going to randomness, right? The point at which they are about 60 percent lined up and 40 percent random—it takes a time to do that—is the rotational time. It's simply a matter of a charge along with the friction in the environment. Again, we looked at the literature, and indeed we were able to predict using this equation. We can predict how the size and diameter of the particles really relates to the DRS spectroscopy.

We have also looked at sensitivity levels, and these are in fact very concentrated levels. We also tried to interrogate the difference between a rigidized molecule and a nonrigidized molecule—one that would flop around. We were able to show that the rigid molecules actually have a slightly higher signal.

Ultimately, we've done a fair amount of characterization to begin to understand what it is that we're working with. It basically comes down to this. We have problems with precipitation (because we're dealing with such high concentrations of DNA), salt masking (because it takes a lot of salt to keep it nonprecipitated), and insensitivity.

So here's the issue: The signal is bound by a curve—water. There is a series of curves—salt. If we can get diluted salt, then we can open up a window to see our signal. Unfortunately, we're at a point right now where our signal is swamped by the salt concentration.

However, we're trying to do several things. We're trying to use fewer charges—for instance 6 charges versus 12 charges—that way we can better solubilize our probes and replace peptide nucleic acids with DNA to solubilize it. With fewer charges, we can go up to a higher concentration. We're talking about overhangs.

At any rate, the whole point is that we're trying to engineer the probes so we can better see our result. But this is the result that we look at.

When we put the two together, we should have a hybridization product that reveals a signal. But we think we're about one order of magnitude below the sensitivity to see the signal. You don't see the signal here, so that's a problem.

We're talking about decreasing the salt by replacing it with nonsalt counter ions. We can use spermine, urea, or some other reagents in place of the salt to open up that window, if you will.

Another thing we can do is to increase the signal, and we can do that, for instance, by increasing the field strength of the electric signal. One way of maximizing the signal is by decreasing the electrode gap. The smaller the gap, the more concentrated the forces. We use interdigitated electrodes (IDEs) to maximize the number of gaps.

Here is a simulation plot. Looking on end, this is one electrode, and here's the second electrode. Here's our scale. You can see there's a little red around this. That is a micro-gap that's about 5 microns across. But when we go to nanometers (e.g., 100 nanometers), now look how intense that electric signal is. Our first IDEs looked like this, and we found out there is parasitic capacitance—a problem with our micro-gap IDE. After nearly a year, we finally manufactured our nanogap IDE at the Cornell Nanofabrication Facility.

This is our first result, which looks like garbage. Because we just got it, we now need to figure out what it is that we're doing and try to get it right. This has really been very tedious research.

Another thing we can do to increase the signal is to concentrate the molecules at a local level by electrical bias on the electrodes or even going to surface binding and hopefully not get the precipitation

The bottom line is that this is a gamble and we're appreciative of NIJ trying to make it happen, because if we are able to succeed then I think we will have a truly inexpensive technology or the potential for a very inexpensive technology that goes to the heart of molecular sizing.

Lastly, here's the logo. Thanks to John Butler: No gels.

Thank you very much.

MR. COFFMAN: Well, thank you, Victor.

I know Lisa wants to say some things to us, but I wanted to see if there's any quick questions we have. I know we have gone over this afternoon, but anybody have any questions?

(No response.)

MR. COFFMAN: Okay. Well, Lisa, the floor is yours. Thank you.