Working with a newer fluorescent molecule on the fluorimeter today, I wanted to verify the (old) solutions of it were all the same thing1. Just running a typical emission profile on the same dilution of each to make sure they were similar concentrations of the same compound. Averaging five repeat spectra for each, I noticed the emission intensity was dropping a bit after the first pass, so obviously there is some concern for photobleaching.

Considering some of my planned assays are looking at 5–10 minutes long for the kinetics to reach equilibrium, I wanted to see what I was working with in terms of photobleaching. Perfect time to take lunch! I setup a 1 hour scan—albeit probably overkill—to look at the time frame & extent of photobleaching. Coming back from lunch, I was a bit confused…

Signal Noise

Granted, this is all normalized against a stable section of the scan (35–45 minutes in), but I wasn’t expecting that obvious of a signal drift, considering the lamp had been warmed up for an hour before even the multi-pass emission scans before!

I’m slightly concerned because (a) these are within the scope/timeframe of the kinetics that I’m looking at, and (b) this is likely an issue with the excitation source, either the power supplying it, or the bulb itself.

Either way, I’m a little concerned about this. I’m not looking forward to having to examine the full length of time needed for signal stabilization (if it can be stabilized at all, presuming it’s not a power supply/line issue). What I really need is some sort of an internal control I can use, especially considering I have two detectors at my disposal. I was attempting to find out if I can use the RCQC signal to measure variance in this, but in some subsequent scans I saw no correlated signal variation in it with the signal drift. However, the drift I saw may have been poorly characterized kinetics for the proteins involved.

Soooooooooooo, yeahhhhhhhh…fun times. *sighs*

1These came from an older time with other students in the lab, so there are some “hazy” (if not completely absent) details about them.

Some days, you’re just better off not going back to the original experiment in its entirety.

I opted to return to the original buffer system for some tryptophan fluorescence in vitro previously done in the lab. Up until now, I was using a different buffer that I had optimized for doing fluorescence anisotropy back during the fall. Using the cross-optimized buffer, I could see 20-30% reductions in tryptophan fluorescence (after accounting for the 10% reduction in signal from ligand addition to induce the conformational change).

Now? Either I’ve gotten way too slow at returning the reaction to the fluorimeter for analysis (doubtful), my protein is crapping out (also unlikely as this is a pretty fresh stock), or this buffer just doesn’t play as nicely. I’m getting maybe a 5–10% change in signal upon ligand addition. And I’m like 3 hours into this 4.5ish hour set of samples.