Re: Fwd: feasibility query

From: Howard Shapiro (hms@shapirolab.com)
Date: Sat Aug 21 1999 - 19:47:55 EST


Julie Auger passed on a query from Kevin Meisner about sorting phospholipid
vesicles containing single molecules of enzyme:At 04:14 PM 8/19/99 

>>My goal is to study the dynamics of single enzyme molecules. A recent
>>technological advance in phospholipid vessicle formation will enable me to
>>trap enzymes in vessicles less than 5 micrometers in diameter. Ideally,
>>each vessicle would contain a single enzyme molecule, and this could be
>>accomplished if vessicle formation were carried out at very low enzyme
>>concentration. Of course, the majority of the vessicles would be empty, so
>>I must find a way to isolate those containing the enzyme.
>>
>>My hope is that this could be done with a fluorescent cell sorter, but I
>>am woefully ignorant of this apparatus. Are single phospholipid
>>membranes too fragile to successfully pass through the machine? Can
>>vessicles of this size be accurately separated? Will tryptophan 
>>fluorescence (perhaps stimulated by a laser) from a single enzyme molecule
>>be brilliant enough to enable detection? Would another method of
>>detection be more robust?
>>

Flow cytometry has been used for studies of phospholipid vesicles, including
some too small to detect in scatter, which have been labeled with various
fluorescent dyes.  One question here is what the lower size limit for the
vesicles is; 5 micrometers is plenty big to detect in scatter, but, say, 0.2
micrometers is difficult for most stream-in-air sorters.

There is little likelihood of being able to use tryptophan fluorescence to
detect a single enzyme molecule; first, there will only be a few molecules
of tryptophan in the average protein, second, the quantum efficiency, as I
recall, isn't all that high, and third, you'd have to use a far UV laser
(e.g., a doubled argon or quadrupled YAG), with which few flow cytometers
are equipped.  The Jovins measured tryptophan fluorescence in cells in the
1970's with the flow cytometer they built at Goettingen, which did have a
far UV laser, but they were looking at a lot more tryptophan than you'd
expect to find in one protein molecule.

Now, I would presume that the reason for selecting vesicles containing a
single enzyme molecule is to watch the enzyme molecule work, which, at least
in the case of some enzymes, can be accomplished by watching the
accumulation over time of fluorescent product produced by enzyme action on a
fluorogenic substrate.  Thus, you'd want to look at the vesicles more than
once, which would be better done in a static or imaging system than in a
flow cytometers.  If you have a label for the vesicles, you can tell where
they are; a dye like "DiI" (DiICn(3), with n=14 or higher, and selected so
its fluorescence would not interfere with that of the enzyme product) can be
incorporated into the lipid bilayer when the vesicles are prepared.  If the
concentration of enzyme is low, you should expect a Poisson distribution of
vesicle occupancy; if you see enough empty vesicles (no product
accumulation), then there will be vesicles containing single molecules, and
the likelihood is that the kinetics of product accumulation will
differentiate the vesicles containing single molecules from those containing
more.  If this is the objective of the experiment, it would not require a
tremendously sensitive cytometer; a good fluorescence scope with CCD imaging
or a CompuCyte LSC should get the job done.  

I don't see that sorting would help much here for getting the 1-molecule
vesicles; you'd have to label the enzyme, which might change its behavior,
and, even if you labeled it with something bright and big, like
phycoerythrin, you'd need single-molecule PE detection to sort the vesicles,
which might be doable with Steve Quake's microfluidic molecule analyzer at
CalTech (P. N. A. S January 1999), but wouldn't with any commercial cell
sorter.  The more obvious label for the target enzyme would be another
enzyme, using a fluorogenic substrate for detection, and that would make it
hard to tell vesicles with one target enzyme from vesicles with more.

But again, the important, unanswered questions are - what's the target
enzyme, and what are you trying to do with it/learn from it?

-Howard



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:53:52 EST