Re: Sorting by Chemiluminescence

Michael Corey wrote:

>It's very good of Howard to take the time to respond to my rather wacky 
>suggestion. I am about to concede defeat, on a related basis, which is 
>that a quick search indicated that the absolute kcat of bacterial 
>luciferases appears to be about 1/sec--probably just too slow. But by 
>nature, I still have to quibble--sorry. What if you had not just one 
>detector, but an array of detectors over a very long path length, so that 
>you could "age" the signal and thereby separate the contributions from 
>individual cells? I'm not sure how close together the detectors in a 
>phased array can be these days. Other obvious problems are the pressure 
>required for a high flow rate over a long path length, and the dreadful 
>software challenge of deconvoluting the time-dependent signal sequences. 
>And there might still not be enough photons. Well, talk is cheap.

As I said in my response to Michael:

"The only practical way to "sort" cells on the basis of a luminescence 
signal is to keep them at least more or less in place, and image them, all 
at once, over a period of seconds to minutes; then pick out the bright ones 
using a micromanipulator or laser capture. Or get rid of the dim ones with 
a pulsed laser beam. That's not fluidic sorting, and it isn't fast, but it 
can almost certainly be done."

A CCD *is* an array of detectors. If you keep the cells static, you can 
look at a very large number of them and see which are bright and which are 
dim, eliminating the need for the complex scheme that would be needed to 
separate (or attempt to separate) contributions from individual moving 
cells. In the above paragraph, I referred to relatively "brute force" 
methods of cell selection, but microfluidics allows production of chips 
that can keep thousands of cells fed and happy over a period of days, 
complete with the appropriate capabilities to extract individual cells for 
further use, i.e., sort them.

If you locked up the mirror of a Nikon D70 digital camera and slapped one 
of these chips onto the CCD, you'd be close to the first version of the 
LACS (good excuse to go out and buy the camera, right?). Actually, you'd 
probably need slightly higher resolution, but it's well within the range of 
many newer CCD-based array readers. And, with the long exposure time, you 
might want to cool the CCD if that wouldn't interfere too much with the 
luminescence signal. Getting the chip is nontrivial, but you could start 
off with cells on a coverslip. You could also emulate the "Bronx Box" 
designed for drug sensitivity testing of TB, and put the coverslip on a 
piece of Polaroid high-speed B&W film (remember film?), but I'd recommend 
the CCD route. Look what happened to Polaroid and Kodak.

If there is a demand for a LACS, it can be made, but it ain't gonna be a 
continuous flow system.

-Howard