Dammit Jim, I’m a rocket scientist, not a cell biologist!
I’m selecting a project for my master’s in systems engineering, and I hoped that the Straight Dope could lend their cell biology expertise. I’m considering automating several steps of the optical grid method of cell counting. I don’t have to build the machine, but I do have to explain why it would be useful and define the machine well enough that someone else could design it from my specifications.
I understand – and correct me if I’m wrong – that optical counting is a prerequisite to plating, which is itself the first step in doing “blots” for RNA, DNA, and protein. Optical counting is also useful (again, as I understand it) for some tissue and blood studies. I keep coming across another cell-counting method called “flow cytometry”, however, and I can’t imagine why anyone would do optical counting when such an advanced tool exists for precisely measuring the cell content of a sample.
So, my question is, does optical-grid cell counting have any advantages over flow cytometry, or any specific applications where flow cytometry is unacceptable? Also, my Google-fu on this topic is surprisingly weak. Are there better / standard terms for the process I’m trying to automate?
I’ve never actually used FACS and have a limited knowledge of it’s implementation, but I think it’s a question of ease of use and the investment required in setting up a FACS versus the ultimate amount of cell-number statistics you’ll ultimately require.
That is, setting up a FACS to measure a certain cellular characteristic may not be an easy, fast, cheap, or off the shelf proposition. Searching out and/or designing the appropriate fluorescent labels can be a significant amount of work, particularly if you are working with a fairly novel field. Additionally, you might be looking for morphological changes in cells where there are little or no changes in surface epitopes, at least as far as you’ve been able to figure out so far. In other words, FACS might be a great way to acquire compelling statistics at the end of your project, but at the start when you’re trying to figure out what the hell two different cells are, maybe not.
Yes and no. The laser-scanning approach still relies on adding a fluorescent compound to the cell sample, and finding the right label, and including a laser. My intent is to preserve the simplicity and broad applicability of optical counts while increasing the speed to be competitive with other automated methods.
Have you used a laser-scanning cytometer? Do you find it to be superior to either optical counting or flow cytometry? Is the machine expensive to purchase or maintain?
Ok , I run a flow cytometry facility and I work with people who do cell counts all the time.
An easy way to explain Hemocytometer vs Coulter Counter vs Flow Cytometry would be this:
A Coulter Counter gives good cell counts and can give you a read out of particle size.
Hemocytometer is useful to count cells, true. But one could also add Trypan Blue to assess viability. You can do this and actually see what you have.
Seeing is believing. We sometimes get better counts by eye then with the coulter counter.
Flow cytometry can give you both counts, size, and viability. Also, add in the possibility of phenotype analysis using fluorescent conjugated antibodies. In flow, you can collect data from millions of events in a short period of time.
I would think that there are several methods out there that allow automated slide analysis. With rapid, high res cameras. Slides are now digitized and pixelated, so we can generate stats from these images.
There are companys like DAKO and Cellomics and Aperio.
Cellomics makes the ArrayScan , with this I could get counts from a slide.
Of course, a Hemocytometer is very cheap, a coulter counter is more expensive and a Flow cytometry setup very expensive. So you usually use the hemocytometer in the student’s lab, where they see that beside cells you may have clumps of cells and cell debries, or , combined with Trypan Blue staining, can assess what proportions of the cells survived whatever they did to them.
The coulter counter is routinely used in the cell culture lab if you need to seed cells at a defined density.
Flow cytometry is frequently used if you have to separate different cell populations and gather statistics, eg. on the appearance of differentiation-dependent cell surface markers, types and differentiation-status of white blood cells in leukemias
That’s exactly what I’m going for. There are a handful of innovations that would make such a system more useful to a university laboratory: basically moving all of the counting and statistics offboard to a desktop CPU, and incorporating recent improvements in digital photography (using COTS parts for semi-obsolete cameras). One or two other improvements are natural outgrowths of such a design but have definite value to the user community I’m working with.
The cost/complexity scale that MTRG and Anaglyph brought up is really good to know about – it helps me narrow my user community down. I assume Big Pharma labs are going to do Coulter or flow cytometry by default, and they’ve sunk capital into the biggest/best machines.
Would it be possible and/or useful for the optical cell recognition software to “guess” at the viability of the cells being examined?