I recently had the chance to visit the core Flow Cytometry facility at the University of Alberta’s Faculty of Medicine and Dentistry, and to see a real FACS (Fluorescence Activated Cell Sorting) machine in action. Saturated as we are in technology that would have seemed miraculous even a decade ago, it’s sometimes hard for me to appreciate the sheer ingenuity of modern scientific apparatus. But somehow, stepping out of my familiar stomping grounds of smartphone programming, and seeing what we can now do with things in the real world instead of just ones and zeros, snapped me out of my complacency.
The facility supervisor – Dorothy Kratochwil-Otto, who kindly demonstrated the inner workings of one of the FACS machines as she spoke – explained to me in relative layman’s terms how the FACS technology works. Here’s what I understood: Cells, presumably in suspension, are sucked up into a fast-flowing, very narrow stream of liquid (some sort of buffer solution). This stream is then fed through a vibrating element that breaks the continuous stream into a series of very precisely formed droplets of uniform size and distribution. To the naked eye, these droplets still seem to make up a thin strand, they’re so dense and traveling so fast. The instrument is calibrated so that, on average, one in ten of these droplets contains a single cell from the original sample.
Now the magic happens.The droplets, some containing cells, fall past a series of lasers (anything containing lasers in any capacity other than as presentation pointers is automatically cool). Some of the light from the lasers passes through the droplets, and the resulting pattern of light that falls on the detectors can tell the instrument whether the droplet contains anything in it, and if so, how big it is. Thus, the instrument can instantly tell, fairly confidently, if any particular droplet contains nothing at all, bits of debris, a single cell, or multiple cells glommed together. All but the droplets containing a single cell are discarded.
For the next part to make sense, I first need a little background. Modern molecular biology makes extensive use of fluorescent tags: These are molecules that, when illuminated with a particular frequency of light (called the excitation frequency), emit light at a different, lower frequency (the emission frequency). The excitation and emission frequencies are characteristic to each kind of fluorescent tag, and can be used to distinguish between them. These fluorescent tags can be attached to other molecules in the cell, such as particular proteins, that are of interest to the researcher. Cells containing these proteins have the tags in them, others don’t. It just so happens that the lasers in a FACS machine are tuned to the specific excitation frequencies of most commonly available tags.
In other words, when cells rush past the lasers, they fluoresce at all the emission frequencies of all the tags that are expressed in them. Some of that emitted light is then gathered up and fed to a bunch of detectors, that can instantly tell which tags were present in the cell that just rushed by. So, by this point the machine can successfully distinguish between the different kinds of cells it’s seeing. Some of the flow cytometry machines stop here, but the coolest ones do something about all the individual cell data they obtain, by sorting them: Each droplet, once it clears the lasers, can optionally be electrically charged by the machine, and can then get redirected into one of several openings at the bottom by traveling through an electrical field maintained between two parallel plates. The amount of charge each droplet gets, and thus its eventual destination, is determined by how it fluoresced when illuminated by the lasers (i.e., by what fluorescent tags it may have contained).
Phew, so there you have it: a machine that can, nearly instantaneously, distinguish between individual cells based on the tags that they contain, and that can then sort these cells just as fast based on what it saw. I didn’t ask how many cells a second this machine processed, but I’m guessing quite a lot: The stream of droplets looked to the naked eye like an uninterrupted, impossibly thin thread. And when the sorting mechanism kicked in, it looked like this strand had split into several, even thinner but just as uninterrupted threads, each going towards its own designated container at the bottom of the machine.
The operating concept, as well as seeing the lasers hitting the thin streams of droplets were impressive. I’m a camera geek in my spare time, so I also greatly enjoyed learning about the particulars of the optical system, the beam splitters and detectors that make up much of the bulk of the machine. But what really amazed me and stuck in my mind is how this technology is beginning to mimic the scale of the molecular processes that it is employed to study: Not only can it operate on a microscopic scale, but it does so at lightning speed, repeatedly and predictably. In that way, it’s not unlike molecular processes in the cell: Pathways composed of dozens of different proteins get activated, pass along information and then shut down, often dozens of times a second. This combination of speed and scale is what makes cellular life as we know it possible. And now we’re making machines that are beginning to catch up.