I have always been fascinated by technology and its ability to drive science in new directions. One of the reasons I enjoy working at a US National Laboratory is because of our abundance of new technologies, and my appetite for the newest, smartest designs is one reason I was happy to serve as a judge of this year's top 10 innovations (see Top Innovations of 2008).
I have realized, however, that using fancy devices is not always the best way to solve a problem. I learned this lesson the hard way when I was a young assistant professor 25 years ago. I had just started my laboratory and had extra start-up money, so decided to find a better way of measuring the rate of endocytosis. The typical approach to studying this process was to add radiolabeled ligands to cells and then measure the relative amount internalized over time using a "strip" solution that selectively removed surface-associated ligands. It was a tedious assay that gave lousy results. Compared to typical biochemical assays, the technique was unsatisfyingly crude. I felt it was time for a technological upgrade.
I decided to build a device that was capable of measuring endocytosis continuously with <1 sec time resolution. My idea was to grow cells on a pair of microscope slides, which were then suspended vertically in a chamber. Medium containing radiolabeled ligand was pumped up the slides at a constant rate so that it took, for example, 30 seconds to go from bottom to top. At the end of the experiment, each spot along the slides would correspond to a different time of ligand exposure, from 0 to 30 seconds. You then rinsed the slides, treated one with a strip solution to remove surface ligand, and then quantified the amount of radioactivity on both slides by autoradiography. The stripped slide yielded the pattern of internalized ligand and the unstripped slide the total amount. You used the difference to calculate endocytosis rates.
It was a very clever contraption, but very complex, with pumps, vacuum valves, electronics and a temperature-control system. Because this was the early days of computers and scanners, we had to build our own video densitometer from scratch and calibrate it using an extensive set of standards. It took over a year to get the device working, but finally, everything clicked. It worked exactly as planned and everyone was impressed by the continuous binding curves that the instrument generated. I was even invited to give a platform session on the device at the annual American Society of Cell Biology meeting. It seemed like a great start to my technogeek career.
There was a problem, however. My results showed me that that the process I was investigating, namely endocytosis, did not happen very fast. Instead of having a "characteristic time" of seconds, which is what my device could measure, it was clearly happening in minutes. The noisy results of the manual technique obscured this fact. Slowing the newer device down did not work because the cells would not survive the longer treatment. In addition, the complexity of the device meant that it frequently broke down. I spent more time tending to the care of the instrument than doing experiments.
All the time I spent working on the instrument did give me some good ideas of how to fix the original problems with the manual technique, which turned out to be mostly the inherent inconsistencies in rinsing plates of cells by hand. To solve this, I took a foot-operated, repeating dispensing pump and strapped on a vacuum hose with a piece of duct tape. This simple apparatus allowed me to do manual endocytosis experiments with extreme reliability and with a 15 second time resolution. This turned out to be plenty of accuracy for my needs.
The fancy instrument with all of the electronic controls ended up in a corner of my lab, where it collected dust for the next decade. I never could find another use for it. The simple pump-vacuum hose apparatus remained in the center of my lab where it was used routinely for almost 20 years. I was glad that I learned this valuable lesson early in my career: Technology can be great, but only if the data it provides are worth the cost and trouble. I also learned that the simplest solution to a problem is, very frequently, the best.
Steven Wiley is a Pacific Northwest National Laboratory Fellow and director of PNNL's Biomolecular Systems Initiative.