Monday, November 29, 2004

Experimental Disasters

A discussion with my brother Rob this weekend inspired a post from me. Naturally we were talking shop about what's going on in our respective labs (his at Texas A&M, mine at Michigan State) and somehow we got on the subject of experiments getting screwed up.

It's a semi-regular occurence in research that someone forgets a step, or does it incompletely, or incorrectly. Little errors can have huge implications when it comes to the results. I thought I'd share some of the more spectacular ones I've heard of.

The first story I was first told when I was an undergraduate at Michigan State in a biochemistry class.
There are a lot of ways to examine DNA, but one is to run it through a porous gel so that it separates into bands of different weights. DNA is invisible though, so to visualize it one can use a dye called Ethidium Bromide to stain the DNA (it wedges itself in the spaces between bases on a strand of DNA). Ethidium Bromide glows brightly under UV light, so one can shine a light UV light on Ethidium Bromide stained DNA and visualize the band and determine the weight (and then approximate the size of the strand from that).

The problem with Ethidium Bromide is that it is a carcinogen, owing to the fact that it wedges itself in between DNA bases (this can cause mutations when the DNA is copied prior to mitosis). My biochemistry teacher was explaining this to us, and to emphasize the point told us of a new undergraduate he had hired in his lab. The student had stained a DNA gel and done his experiment, then walked off to the restroom after he finished. He did not remember (or didn't realize it was important) to remove his gloves and proceeded to put contaminate everything he touched with Ethidium Bromide (fortunately it can be cleaned off without a great deal of trouble). The professor told how they had to walk around the room with a UV light, following the Ethidium Bromide trail to the bathroom. The student didn't work for them after that, and I never heard at which point he remembered to take of his gloves.

Such stories are not uncommon to hear from chemical safety officers at universities, particularly when it comes to radioactive isotopes. Radioactive isotopes are useful in research because we can easily detect them, so attaching them to something that is normally undetectable can make such things easily detectable. The problem is they stick around for a long time. If one spills a radioactive material one may have to close off the area for use. An example that always sticks in my mind is the story about a person who had spilled a radioactive solution on the floor in the lab and accidently walked in it, getting the isotope on their shoes. The Office of Radiation, Chemical, and Biologicial Safety (ORCBS for short here at MSU) cleaned up the mess, and took the person's shoes! Apparently they were a good pair of shoes too, since the person had to call back every few months to see if their shoes had "cooled off" enough to be safe to wear. That's one case were a spill in the lab can put a real cramp on your style.

The next story was relayed to me when I was working in the Food Science and Human Nutrition department at Michigan State. bemoaning an error I made in mixing some buffers, using up most of a reagent that was about $100 per milliliter. The student I talked with said not to worry, she'd seen far worse mistakes. She told me the story of another student who had been working for a professor that did feeding studies (it was, after all, the nutrition department) in rats try to design better diets for people. For feeding studies one usually has to feed rats for a few months and track their weight gain. At the end of the study one does normal clinical chemistry on the blood of the animals to understand better how effective the diet was. Such a study gets to be very expensive because it costs money to buy, house, and feed rats (as well as graduate students).

At the end of the experiment the red blood cells need to be separated from the rest of the blood, so one uses a centrifuge to spin the blood and get the red blood cells at the bottom of the tube and the remainder of the blood (called plasma or serum depending on how it was treated) for clinical chemistry checks. When using a centrifuge it is important to make sure the weights on either side of of the spinning rotor are similar, otherwise the imbalance can damage the machine (consider an unbalanced washing machine on the high spin cycle can rock back and forth quite a but; now spin it 10,000 times faster). If things got really bad a centrifuge can fail in a spectacular manner.

The professor told the student to be sure to balance the rotor, and use water to even the weights of the blood samples. The student followed the instructions, adding water to each blood sample so that they were all the same weight, in the process ruining the blood samples (adding water to the blood will can the red blood cells to burst because there isn't enough electrolytes like sodium in the water). A case of months of work being partially undone in half an hour. Doh!

The final story I got was from Rob, who told me this story was told to him when he got to Texas A&M. I think it's the most impressive of the three.

Occasionally in research there are reagents left over. Being collegial, investigators often offer the leftovers to other investigators in case they have any use for them before sending the stuff for disposal. Apparently a researcher had finished with a tank of chlorine gas (a highly reactive, corrosive, poisonous, and useful reagent). It had been quite some time (years) since it had been used, so it was given to another group. The safety cover for the top of the gas cylinder had become rusted, so before a new regulator could be placed on the tank the old safety cap and regulator needed to be removed. A pair of graduate students were assigned the task of putting a new regulator on the tank so that it might be used. Despite their best efforts the safety cap couldn't be removed, so one took out a hammer and gently tapped on the cap to remove it. Success! Now the old regulator needed to be replaced. The succesful student stepped out of the room for a while to take care of something, leaving the other student to remove the regulator.

This is when things get stupid. The remaining student decided to try the hammer approach on the regulator. As Rob told me the story he described this stage like so:
A person in another room said all they heard was a "tink, tink, tink.... tink... Tink... TINK... OH SHIT! HISSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS"

The student knocked the entire top off the tank, producing a glorious stream of chlorine gas. Being that chlorine is a highly reactive, corrosive, poisonous, and useful reagent the student high tailed it out of the room and slammed the door behind him. Fortunately no one was injured in the faisco, other then the pipes in the ceiling of the room which the chlorine corroded quite badly. What's the moral? I'd like to phrase it like a professor put it to me when talking about how to plan an experiment: You've got to look at the tools you've got available to you and decide how to best approach a problem. If you need to cut a piece of 2x4 you can manage to do it with a hammer, albeit slowly and messily, or you could use a saw. That's where wisdom and experience come into play. So the moral is: a hammer is not the tool of choice for sawing wood or fixing the regulator on a tank of compressed gas.


Post a Comment

<< Home