Well, I was going to talk about some of these experiments during lab meeting, but why make a Powerpoint or Google CoLab link people won’t follow when I can write it as a blog post.
Regardless, we’ve recently been looking at how our various possible methods of spectrophotometry compare.
- “Amazon Spec” purchased for $235 back in November 2020.
- ThermoFisher Nanodrop in a departmental common room (I don’t actually know what kind as I’ve never used it)
- BioTek Syngergy plate reader, either… A) with 200uL of bacteria pipetted into a flat-bottom 96-well plate, or B) using their “BioCell”, which is a $290 cuvette that fits onto one of their adapter plates. I mistakenly label this one as “BioCube” in the plots, but they probably should have just named it that in the first place so I don’t feel too bad.
To test the methods, Olivia sampled bacterial optical densities while a batch of e.coli were growing out to make competent cells. Thus, the different densities in the subsequent data will correspond to different timepoints of the same culture growing out. Each time point was measured with all four methods.
Well, all of the methods correlated pretty well, so no method was intrinsically problematic. I’m not sure if the settings for any automated calculation of absorbance values, but the BioCell numbers were just off by an order of magnitude (The BioCell data also had a clear outlier). The Amazon spec and Nanodrop generally gave similar values, although the nanodrop gave slightly higher numbers, comparatively.
The plate reader option was also perfectly fine, although it required more back-end math to convert the absorbance values to actual optical density. This is also not the raw data, as the media only absorbance has to be collected and subtracted to yield the below graph.
Rather than try to figure out the path length and try to calculate the formula, I just used the above dataset to create a calibration for “nanodrop-esque optical density”. (Note: There was a second independently collected set of data I added for this analysis). Here, the goal was to actually use the raw values from the plate reader output, so people could do the conversion calculation on the fly.
Say you have a particular nanodrop-esque A600 of 0.5 in mind. The formula to convert to plate reader units is 0.524 * [nanodrop value] + 0.123, or in this case, 0.385. Checks out with the linear model line shown above.
Or, if you already have raw platereader values and want to convert to nanodrop-esque values, the formula here is 1.79 * [biotekp value] – 0.2 to get the converted value. Here, let’s pretend we have an absorbance value of 0.3, which calculates to a nanodrop-esque value of 0.338. So perhaps that’s a decent raw plate reader value to get with nearly grown bacterial cultures during the chemically competent cell generation process.
Lastly, it’s worth noting how surprisingly large dynamic range there seems to be for spec readings of bacterial cultures. It’s likely largely because we’re used to handling either mid-to-late log phase growth or saturated / stationary cultures, but we’re used to dealing with values in the 0.2 to 1.2 range, although the log-scale plots above suggest that we can be detecting cultures reasonably well within the 0.01 to 0.1 range as well.