Why and How? Daniel Gitler Dept. of Physiology Ben-Gurion University of the Negev
Why use confocal microscopy? Principles of the laser scanning confocal microscope. Image resolution. Manipulating the data cube. Existing confocal systems. Operating a confocal. Additional features of a LSCM. Other optical sectioning methods.
Confocal Widefield http://www.olympusfluoview.com/theory/confocalintro.html
http://www.olympusfluoview.com/theory/confocalintro.html
To obtain clear optical sections of a sample. Necessary in thick or multilayered preparations. To acquire 3D morphological information. To achieve focus throughout a thin but not absolutely planar preparation. When point-scanning of the sample is of methodological importance.
Buying it and operating it is expensive. Usually necessitates special training. May photo-damage samples more than a conventional camera-based system. Usually slower. Fashion and prestige.
Laser Exciation Emission Detector Pinhole Dichroic mirror Laser beam is focused into a point at the focal plane. The point is scanned on the object by galvanometric mirrors to obtain a raster image in the computer. The detector is usually a PMT. Collimator Scanning mirrors Objective Sample Focal plane Spatial separation is achieved by the scanning procedure. Emitted light is descanned and passed through a stationary pinhole.
(laser scanning confocal microscope)
Detector Pinhole Conjugated planes Objective Sample Focal plane
In focal plane
Below focal plane
Above focal plane
Pinhole rejects most out of focus light Pinhole Aperture determines degree of out-of-focus light rejection. Smaller -> better. However: The smaller the pinhole, the less light is detected may reduce signal to noise levels. For each objective find out the minimal pinhole aperture that still contributes to sectioning capabilities. Anything smaller will just loose light.
Image Image Light from out-of-focus objects is rejected However, Out of focus objects are excited!
The PSF is the image of an infinitely small light source (in practice an object smaller than the resolution of the optical system). The confocal PSF is shorter in the Z direction, reflecting the removal of out-of-focus light. http://www.olympusfluoview.com/theory/resolutionintro.html
Resolution is determined mostly by the objective s Numerical Aperture (NA). Magnification is irrelevant when zoom is available, No advantage to 100x over 60x lens of same NA. The pinhole increases contrast by removing out-of-focus light. Scanning with a point increases resolution relative to conventional microscopy because both excitation and emission are diffraction-limited.
2D Airy Disk diffraction pattern Lateral resolution Airy Disk diameter: 0.61λ/NA (Rayleigh criterion). In a confocal: 0.4λ/NA. Lateral typical limit: 250nm Axial resolution; Section thickness 2ηλ/(NA) 2 Axial typical limit: 500-1000nm Deconvolution can increase resolution
Immersion medium should be matched to mounting medium. In live cell work water immersion objectives should be considered. When Stoke s Shift is large, focal planes of excitation and emission may differ!!! Detection may be hampered!
Which is above?
Z Y X Z Y
n P( x, y) = Max( S ( i i= 1 x, y))
Detectors Emission Filters and Monochromator Pinholes Dichroic Mirrors Excitation Scanning Mirrors Excitation Power Monitor Emission Zeiss LSM 510 Meta
Detectors Emission Scanning mirrors Pinhole Mostly manual Modular Dichroics Excitation User accessible 3 Channels Emission filters (not shown) AOBS attenuation Shared pinhole
Fully automated Spectral detector (in practice 8 channel output, 10nm wavelength resolution). AOTF attenuation of lasers. Not user serviceable!
Like Nikon C1, with 32 channel spectral detector (2.5, 5 or 10nm wavelength resolution) AOM or AOTF laser attenuation
Fully automated. 4 Channels, 2 spectral, one pinhole. Serial spectral detection, using monochromator and slit. Confocal is mounted on excitation light path and not camera port more efficient light path. User serviceable.
Spectral confocal based on prism. Slits return rejected part of spectrum for further selection. Little loss of emission light.
How to detect more photons - How to improve the signal to noise ratio
Let more laser light through Pro: + More photons will be emitted, reducing SNR. + Relatively linear. Con: - Increased Photobleaching - Increased Photo damage - Possible Dye saturation: If light flux is strong, excitation photons may arrive before fluorophore emits, resulting in non-linearity.
+ Pro: + Does not affect acquisition time or sample exposure time. Con: - Not linear, should be taken into account in quantitative experiments! - Increases non-poisson noise
Pro: + More photons are collected + Signal to noise ratio is improved Con: - Increases Photobleaching - Increases Photo damage - Increases scan time
256X256 128X128 (zoomed up) Pro: Increases SNR without elongating scan time Does not increase sample exposure Con: Image resolution is lowered => Should be considered, since over-sampling is common
Pro: + Increase photon detection without elongating exposure or scan time. Con: - Loose confocal effect. - Non-linear, sample dependent. Ask yourself: How thin does the section really need to be?
Expanding the acceptable range of emission wavelengths will increase number of detected photons. Pro: Con: + No effect on scan time or length of exposure. - Unintended photons, emitted from other fluorophores or from auto-fluorescence, may contaminate the image.
Argon ion (UV (water cooled), 457, 476, 488, 514nm) Krypton- Argon (488, 568, 647nm) HeNe (543, 594, 633nm, IR) (not all at once) Laser diodes (630-680nm) Blue laser diode (405nm, 454nm, near UV) DPSS (1064nm, 532nm, 354nm, 488nm, others) HeCd (325, 442nm)
Short wavelength dyes should be detected with sharp bandpass filters to remove signal from cross excited long wavelength dyes. Long wavelength dyes should be excited with long wavelength source to prevent cross excitation of short wavelength dyes.
Emission of CFP and YFP overlap to a large degree.
457nm HQ480/20 457nm cross excites YFP HQ480/20 eliminates cross talk, but misses most of CFP
405nm HQ470/40 405nm stronger, more separated, yet less efficient for excitation HQ470/40 eliminates cross talk, and passes more CFP
440nm HQ470/40 440nm even better HQ470/40 eliminates cross talk, and passes more CFP
514nm HQ540/30 514nm does not excite CFP
30 Fluorescence (AU) 20 10 CFP Sum YFP 0 450 500 550 600 Wavelength (nm) Useful for separating acquired images into images of component dyes (unmixing) Useful for FRET studies
Nikon C1Si brochure
Zooming by controlling the scanning mirrors Flexible region of interest Acousto- optic control: rapid excitation switching, multiscanning, retrace blanking 12bit output: high dynamic range Spectral detectors: spectral unmixing, FRET. External sync: dynamic experiments, FRAP. Transmitted light detector.
Fast intensity attenuation (within a pixel) allows controlled bleaching or activation within a region of interest of the image. Useful for: FRET Fluorescence Resonance Energy Transfer FRAP Fluorescence Recovery After Photobleaching IFRAP Inverse FRAP Photo-Activation; PA-GFP, Kaede etc Uncaging; caged Ca 2+, amino acids, peptides And other photophysical protocols.
For quantitative research, a confocal should be calibrated. -> Resolution: Use subresolution beads to measure PSF. -> Chromatic matching: Use multiply labeled beads to verify colors match in X,Y and Z. -> Size: Against a calibrated target slide. -> Intensity: Against a fluorescence standard, such as known-fluorescence beads.
Z-stack of a sub-resolution PS-speck bead XZ image of PSF obtained by imaging a PS-speck bead Molecular probes fluorescent microsphere standards brochure
Using multi-colored FocalCheck beads to correct channel alignment Molecular probes fluorescent microsphere standards brochure
FocalCheck beads Molecular probes fluorescent microsphere standards brochure
Well calibrated fluorescence intensity standards (InSpeck) allows acquisition of quantitative data in a (confocal) imaging system Molecular probes fluorescent microsphere standards brochure
Multi Photon Microscopy Spinning disk confocal Structured light systems (Optigrid/Apotome) Deconvolution
Image Image Image
Fast (360Hz) Uses sensitive CCD as detector Saturation not an issue Directly viewable Possibly gentler on live cells Less flexible: Fixed size pinholes No zooming, ROIs etc.
Uses 50% of the available light Works well with conventional arc lamp illuminators Not real time, but faster than deconvolution Requires high dynamic range imagers Inexpensive add-on to conventional microscopes
Actual Signal IRF Signal convolved with IRF Obtained by measuring response to spike Deconvolution peels away the IRF from the measured signal Problem very sensitive to noise
Acquires wide field images at various heights, and uses a mathematical model to calculate the 3D distribution of light from the object. Blind deconvolution estimates the instrument parameters. Non-blind deconvolution requires measurement of the PSF for the system (or a reasonable guess thereof)
Requires highly accurate z-movement Is not real time (but computing is getting faster all the time) Makes maximal use of sample exposure (good for living cells). Relatively inexpensive. Not great for largely diffuse samples.