Digital I/O card counts DNA fragments in nanobiotechnology system.
Not only is the physical sorting a challenge, the related electronics are, as well. It's not enough to run a digital-input counting subsystem at a high speed; the associated counters must be large enough so they don't overrun too quickly, and such a card must be able to move megabytes of data to a host machine without suffering from even the tiniest gap in the data. Those were some of the key requirements of Warren Zipfel, assistant director of the NIH-funded Developmental Resource for Biological Imaging and Optoelectronics (DRBIO) at Cornell University, who is involved in constructing the data-acquisition hardware of a new "nano-device" designed to quantify DNA fragments In collaboration with DRBIO researchers Jonas Korlach and group director Watt Webb, and with nanofabrication experts Mathieu Foquet and Harold Craighead of Cornell's Nanobiotechnology Center (NBTC), the group is using a combination of sophisticated silica substrates, electrophoresis, optics, and PC-based data acquisition to gather statistics on fragment sizing in DNA. When this technique is perfected, it could provide an inexpensive and fast alternative to conventional gel electrophoresis
One enabling technology has been the ability of the researchers to use high-end lithography techniques to create channels in silicon and glass substrates as small as a few hundred nanometers. Devices this small provide a means of achieving an effective optical resolution not possible with traditional optical techniques in the sense that a researcher can separate, confine, and observe a single molecule in a region much smaller than the optical spot size.
These "nano-channels" are so tiny that it's relatively easy to ensure that large molecules such as DNA pass through one at a time, and the researchers are putting this effect to good use. In their prototype system, the operator starts with a small sample of DNA that has been enzymatically cleaved into fragments of various lengths. Next, by binding fluorescent reagents onto each fragment the scientist can label each one in a way whereby the amount of fluorescent probe per fragment increases with fragment size. Fluorescence is used routinely in labs as a sensitive way to monitor biochemical reactions and to quantify important biomolecules such as proteins or DNA.
Once the sample is loaded into the device, an electric field in the range of 100V/cm induces the fragments to flow, one by one in single file, through a 250 x 500-nm wide channel in the device. A focused laser excites the fluorescently labeled fragments so they emit flashes of light while traveling through the illuminated region in the nano-channel. Each flash of light is a single photon, which the system then detects with a photomultiplier tube or avalanche photodiode. The photodetector sends out a TTL-level pulse for each detected photon, Then a digital counter card keeps track of the number of pulses that come in quick succession, and with this information the laboratory-written acquisition and analysis software can identify a "burst"--a collection of photons that correspond to a DNA fragment passing through the laser spot.
The software then analyzes the number of bursts as well as the number of photons in each one along with each burst's temporal width. It displays these results as a histogram and they correspond closely to a scan you'd get from an image resulting from gel electrophroesis--the traditional method of separating and analyzing DNA samples. Note that with this system you count every molecule in the sample, whereas with the gel method it's far more difficult to accurately quantify the amount of each DNA fragment. A further advantage is that this newer technique can work with a much smaller sample size compared to conventional electrophoresis methods.
To count the TIT, pulses from the optical-detection subsystem, Zipfel chose a PD2-D10-64CT card from United Electronic Industries. He uses the three counter/timers that come integrated in the DSP chip that controls the card's operation; two of the counters act as timing clocks to define bin size, while the third counts incoming pulses. The DSP reads the third counter on the fly and sends the increasing count value to the host PC. The 24-bit counters are sufficiently large to handle the thousands of photon pulses per bin that might come in while a large DNA fragment passes through the illuminated spot in the channel.
The DIO board is capable of continuously reading the counter output at a rate of 1.0 MHz if required, providing 1-[micro]sec resolution for determining photon burst width and time between bursts. At this datarate, gap-free acquisition for long periods can place an enormous workload on the PC, creating a problem if realtime data analysis is required. Fortunately, Zipfel knew about the special ACB (advanced circular buffer) on PowerDAQ cards, which allows them to stream large amounts of data to the host PC's hard disk and do so gap free. Several years ago, he designed a DOS-based system that had to digitize analog inputs for lengthy periods, and he chose a previous generation of streaming hardware and software from UEI.
When he needed the same capability in a PCI digital I/O card, however, he found that the current generation from UEI didn't offer that streaming capability for the DSP-based counters, Zipfel inquired because he felt like his request would get serious attention."UEI is small enough that when you call into that company, they are responsive. I had queried some larger companies about this project and even visited their booths at trade shows, but they never bothered to get back to me. But not only does UEI respond right away, I was speaking with the card's designer and the fellow who wrote the firmware. I explained what I needed, and they came back with an upgrade to the firmware command interpreter that runs on the DSP. The upgrade contained new calls into the UO function library that allowed me to do exactly what I needed with those timers. I don't believe I could expect this level of quick, personalized service from any other supplier." Because the PD2-DIO board's DSP handles the transfer of I/O data to RAM, the PC has plenty of resources available to analyze the incoming stream of photon counts in real time.
In fact, Zipfel adds that his group had spent considerable money on dedicated systems that couldn't do the job. "So here we had a $400 digital I/O card that enabled us to complete a project, and it worked better than a specialized card that sold for $15,000," he comments.
This system presently monitors one nanometer-sized channel on a device. With the proof of concept successfully behind them, the researchers are looking to expand the applications For instance, they envision a system with tens or hundreds of channels that could quantify DNA fragments by size, or even sort fragments into separate storage reservoirs for later retrieval. Zipfel has several other applications for the PD2-DIO board including an inexpensive autocorrelator for a biophysical technique know as Fluorescence Correlation Spectroscopy (FCS) and to replace more expensive "multichannel sealers" he has previously used for several other types of biophysical and photophysical measurements.
United Electronic Industries, Inc.
611 Neponset St.
Canton, MA 02021
Write In 117 or Reply Online