Three Things I Learned From a Decade Behind a Microscope

Near the start of my faculty position, I was sitting in front of my lab’s 2-photon microscope, watching fluorescent T cells on the screen as data was acquired. I realized that it was November 2021, and that I had done my first 2-photon experiment at the start of my postdoc in November 2011. That means I have been doing 2-photon experiments with live cells for 10 years! I can admit very frankly that the reason I have a faculty position is my expertise in 2-photon microscopy (also known as multi-photon microscopy).

Since its advent in the 90s, 2-photon microscopy has become a widely used biological microscopy technique. Its accessibility has increased as companies such as Nikon, Zeiss, Leica, Olympus Thorlabs, and Bruker have developed turnkey systems for biological laboratory use (the latter of which had built my own system here at Mayo Clinic). However, 2-photon is just complicated enough that it usually requires experience or dedicated staff in order to obtain the most meaningful results. This expertise is why though I am now faculty and the lab lead, I had been doing a lot of the 2-photon experiments towards our first publication. A major part of those experiments has been troubleshooting new imaging setups, as my own lab trainees are still developing their microscopy skills.

Our lab’s upright microscope, built by Bruker.

Though I did not perform 2-photon in my doctoral thesis, I did spend quite a bit of time puttering around with different fluorophores and learning about 2-photon excitation cross sections as a graduate student, my advisor being an expert in fluorescence calcium probes. However, my practical skills in microscopy were most cultivated by taking a wonderful class at the University of Texas at Austin from plant biologist Malcolm Brown, which involved 6 hours of microscope work and a full >10 page report to be written every week. As a postdoc, I joined a relatively new mentor that brought experience with the 2-photon imaging of T cells and from then on, I used 2-photon about 1-2 times per week for over 8 years. My skills were also complemented by a microscopy workshop at the Marine Biological Laboratory, an amazing scientific training institute that I had recently revisited to learn about stem cell techniques. So now that I have gotten my credentials established, let me share 3 things I have learned in over a decade of microscopy, both the practical and the abstract:

1. Visual data is a tremendously powerful thing. Like most people, scientists respond to visuals. It is one thing to read out cellular activity via a flow cytometry plot, a Western blot, or color change in an ELISA- it is another thing to see an image or video of cells in their native environment, interacting and moving. Cells and proteins that are invisible to the naked eye are suddenly more real. We can now almost picture them as entities with some agency of their own (though this isn’t true, of course, but it can make their behavior easier to understand). Microscopy can be dinged as too ‘qualitative’ or ‘descriptive’ as data, but used well it can help generate new ideas about phenomena, especially about how cells get from place to place, and how they interact with their environment towards some end.

2. Managing the specimen is the hardest part. Just one of many examples—For my postdoc, I studied the thymus, an organ that sits above the heart and is the seat of T cell development. Being next to the beating heart and inside the rib cage, it can be very difficult to image the organ by live microscopy. Our lab used a technique, derived from neuroscience, in which the thymus is quickly harvested onto ice and then prepared into thick tissue slices. Amazingly, the tissue can be maintained within cell culture conditions for several hours, and the cells within happily respond as if in a living organ. Unfortunately, you lose information regarding blood and lymphatic vessel dynamics, but you still retain a fairly intact microenvironment in which one can watch immune cells move and interact with one another.

For the most part, the optical part of the 2-photon microscope was the easiest to deal with—as long as I focused the objective onto fluorescent material, it would work. The only times this part would be mucked up is if there was some sort of software or hardware issue. The toughest part of the day-to-day experiments was maintaining the material the microscope would observe. The thymus slices had to be kept warmed to physiological temperature and immersed in an oxygenated medium supplemented with goodies to keep tissue and cells happy even when outside a living body. To keep the thymus slice viable then, I used a heated chamber that had an inlet and outlet for which to circulate the oxygenated medium. As a postdoc we used an IV regulator to gravity feed the medium, but the flow rate could be difficult to control as the salts in the medium started to gum up the regulator. Suddenly, the flow would stop, and my poor tissue slice would dry out on the stage. Or, the medium would overflow from the shallow chamber, and being so close to the resistors that heat the stage, would rapidly corrode the electronics.

Since the issue with fluidics was the bane of my postdoc existence, I planned my new microscope to mitigate this issue as much as possible. I found this microfluidic pump from Bioptech and it has really saved my life. The unexpected plus is that I also go through less medium, which is made from a costly powder formulation of phenol red-free RPMI1640. In conclusion, when it comes to imaging a biological specimen, it is usually the specimen part, not the imaging part, that is the most variable and requires the most optimization.

3. Have systems for experiment recording, data naming, and data storage. I am not a computer programmer, but in an alternate universe that would have been my career. However, as an older millennial (do y’all remember DOS? I sure do!) and also as someone who has used a bit of Matlab, I have become a little picky about naming files and practices like avoiding special characters. A regular imaging practice can generate a ton of data, and to be biologically meaningful you must be able to cross-reference it with an actual experiment. You don’t want to have a ton of data, all named “tumor 1.tiff”. Every experiment goes into a physical lab notebook with a date and a title, in chronological order. I still prefer paper ones, like these nifty ones I used as a postdoc and now buy for my lab. Here I record all the wet bench work associated with the specimen preparation. On the computer that runs the microscope and records the images, I have a “Jessica” folder, and each experiment day is a new subfolder with the experiment date and a general title. Lab notebooks remain the property of the lab, and all data should be backed up regularly to a server.

Were my tips surprising? Probably the last tip is that microscopy is akin to an art., even with all the automated imaging systems now popping up. Sometimes the biology is prohibitive, but in my opinion, the most impactful micrographs are also visually pleasing. There are few shortcuts for putting the time in and developing microscopy as a skill. I hope this continues to be an important part of biological science in the future.