Diary of a Failed Laboratory Method

My thesis is one chapter away from completion, and it’d be three chapters away if I hadn’t made a strategic decision to give up on a certain measurement I’d been hoping to make.  That measurement has been a pox on my career since I first proposed it.  Even as I acknowledge that actually getting it would have been an impressive boon for my thesis, I am glad to see it go.

Much of my Ph.D. research is about the homeoviscous response in fish.  Temperature acclimation in ectotherms involves many molecular and physiological changes, and the homeoviscous response is part of this process.  One of the effects of changes in temperature is that biological molecules become more rigid with cold and more fluid with warmth, and the lipids that comprise biological membranes are no exception.  Because the fluidity of membrane lipids affects protein activity and cell permeability, there is a response in place to keep membrane fluidity—the inverse of viscosity—constant: a homeoviscous response.  This is achieved by physically replacing membrane lipids with lipids better suited to the new temperature in several different ways, much as car owners once had to switch their motor oil between summer and winter versions.

Gas chromatograph of the sort used to analyze lipids.  By Mcbort [Public domain], via Wikimedia Commons
Gas chromatograph of the sort used to analyze lipids. By Mcbort [Public domain], via Wikimedia Commons
It is easy enough to measure lipid profiles in tissues, and one of my lab’s specialties is collecting lipid data.  Measuring membrane fluidity, and thus being able to determine how thorough a particular homeoviscous response actually is, is a much harder task requiring a completely different suite of specialized equipment.

A former student of the lab next door to mine had undertaken this measurement as part of her own thesis, so I concluded as part of my original proposal that such a measurement must be possible at this university.  My advisory committee tried to dissuade me, noting that the measurements were possible because of collaboration with the owner of the necessary equipment at another university, several hours southwest.  Trained to view such caution as a challenge and any response other than raring ambition as an admission of failure, I insisted I’d find a way, but would focus on the part of my proposal I knew in advance was, in fact, doable.

There really should be a directory of what laboratory procedures can be easily performed where.  So, so much of the least pleasant parts of my thesis-proposal process could have been avoided if I’d had a clear idea from the start what was and was not possible with readily accessible equipment.

I did searches for the kinds of plate readers and centrifuges that could handle the technique: fluorescence anisotropy/polarization.  I discovered other techniques preferred by more recent researchers, but even less available to me.  I looked into prices and accessories.  I sought every feeler I could manage on local labs that might have the machines or know how to use them.  Few of the people I consulted responded at all, in part because the generation that produced my professors is infested with the broken, anti-modern idea that Email is for fundamentally unserious inquiries and important business takes place only by telephone.  The ones who did usually claimed that I’d found them mistakenly.  Two avenues did bear fruit: I was redirected to speak to the co-author of the paper that started me on this foolhardy journey, and my supervisor learned that the needed equipment could indeed be found in two different locales at this university.  I established a baseline rapport with a grad student in one of those labs, learning as much as I could get out of her about the operation of her lab’s plate reader.  This included spending several hundred dollars of lab money on optical filters for their plate reader, so that it could perform measurements at the frequencies I required.  I also started an Email exchange with a student of the original collaborator, who got me started with a protocol.

I let the matter sit for a long time, busy with the work I’d already started and several other tasks.  Several ideas for thesis expansion withered on the vine until I felt no choice but to propose that I tackle this assay once more, for a new experiment.  My committee recognized the value of this inclusion, but again cautioned me.  My supervisor in particular, after watching how long it took me to become proficient at the laboratory tasks I’d been required to learn up to this point.

One more piece falling into place was the common molecular facility acquiring an ultracentrifuge, needed for the most difficult step and the last thing preventing me from actually attempting the protocol.

That summer was mostly devoted to getting phospholipid and cholesterol measurements from 315 and 539 tissue samples, respectively, and processing that data into results and figures, which was plenty of work on its own.  Once those tasks were done, I could focus on optimizing my received protocol for my actual situation.

I rapidly learned that the centrifuge tubes I’d picked out based on advice from other professors were in no way suited to this task.  They were too wide and also liable to shatter at the high-velocity spins I required.  It took several more tries and consultations with yet another professor unidentifiable from the outset but apparently possessed of vital knowledge to figure out this problem and get suitably-sized tubes, and a few more tries to start getting consistent membrane extracts.  By this time, my contact for the protocol had long since stopped responding to my Emails, so I had to rely on this new fellow to guide my trial-and-error.  He helped me figure out that part of the problem was that I needed to use more tissue than the protocol I was given assumed.  This, in turn, required me to pool samples between fish, reducing my sample size from 63 to 37 and making sure I would get exactly one go at extracting liver membranes.

But I got it done.  I extracted membranes from all of the liver samples for that experiment.  I set aside the muscle and gill samples, for a future moment when I had time to re-optimize for tissues with different textures.  All that was left was confirming how to use the plate reader, after years of no practice and never having made it all the way through the tutorial.

My contact in that lab graduated and passed my issue to her successor, who took longer than I’d have liked to agree to a meeting and was rather less invested in whether he even knew how to use it than he’d let on.  Then I had to buy another $150 optical filter because my previous contact told me the wrong number of filters.  After several more false starts, it seemed like we’d figured out how to get it to work…maybe.

I tried it.

I got a weird error message.

I tried it again on a Sunday, suspecting something odd I’d done with the plate.

I got the same weird error message.

The message pointed toward a burned-out internal deuterium lamp that would be expensive and time-consuming to replace, and the owner of that plate reader seemed disinterested at best in this apparently dire equipment malfunction.  He did, at least, point me at a potential alternative, a gigantic, robotic plate reader mostly used for much more involved chemistry work, but which could potentially solve my problem.  I had enough sample left, after those two failed tries, to attempt this measurement exactly once more before I simply would not get liver numbers, and looming thesis deadlines made it impossible to get data for other tissues regardless.

The fancy plate reader could indeed measure fluorescence polarization…but it didn’t currently have the right optical filters, and could not borrow the ones I’d already purchased for the other machine.

Out of time, out of patience, and out of options, I let the matter die.

Now, I have a shitload of prepared buffer, some leftover reagents, some extracted liver membranes that will, in all likelihood, never get used for anything, and years of badgering and rushing from my supervisor and declining interest from me, to show for my trouble.

The moral of today’s story is, don’t try to measure membrane fluidity.  It’s not worth the aggravation, and I’m getting the same thesis I would have gotten with it, without it.

Diary of a Failed Laboratory Method