Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocalTo add to an old thread:
We did some single molecule imaging using homebuilt sample-scanning
confocals equipped with APD detectors and this definitely works. The main
differences between the TIRF and the confocal are:
1) The confocal scans point by point while the TIRF with EMCCD camera
detects all pixel simultaneously. When you use 20-100 ms integration time in
the TIRF you need to match these on the confocal using the slowest scan
speed and some averaging. It takes quite long to image a 512 x 512 pixel
image (for 1 ms integration time/pixel 5 min.).
2) The TIRF has about twice the electric field at the surface (incoming and
reflected light), which gives four times higher intensity for the
fluorophore excitation but does also work for detection of the emitted light
(optical reciprocity).
3) The EMCCD has >90% quantum efficiency while the typical PMTs in the
confocals have less than 20%.
A good sample to start with is Alexa 647 labelled IgG (5 fluorophores /
molecule) excited with 633 nm, a drop 50 pM concentration on a coverslip,
add 10 mM CaCl to pin the IgG down and wash excess away.
You need to match the pixel size (100nm is good), laser power and
integration time of the confocal to the TIRF conditions. The lower
sensitivity of the detector can be compensated by increasing the laser power
at the expense of fluorophore lifetime. After bleaching of the majority of
fluorophores and background, there will be a few molecules left with longer
lifetime, which give a good indication of the signal of a single
fluorophore. I managed to image single Alexa 647 IgGs on a Fluoview 1000
using the PMT detector with bandpass filter but would not recommend this for
routine
imaging (too slow). However this would be an interesting benchmark to
compare S/N between microscopes.
Locked
