CNWY40120 Advanced Biological Imaging Prof. Jeremy C. Simpson Lecture 7 Confocal microscopy
Today’s lecture ...
Introduction to confocal microscopy Wide‐field versus confocal microscopes Point scanning confocal microscopy Spinning disk confocal microscopy Software interfaces The 39 Steps ...
Types of fluorescence light microscopes Wide‐field microscopes ‐ light source is typically a lamp (eg: mercury, metal halide, xenon) ‐ appropriate fluorescence wavelengths are generated by passing the light through various filters ‐ images are captured on highly sensitive cooled CCD (charge‐coupled device) cameras ‐ images contain some ‘out‐of‐focus’ light ‐ relatively inexpensive (€20,000 ‐ €200,000)
Confocal microscopes ‐ light source is typically a laser (eg: helium, argon krypton gas, solid state) ‐ appropriate fluorescence wavelengths are generated by choice of laser and ‘tuning’ of their wavelengths ‐ images are captured on photomultiplier detectors (or occasionally on CCD cameras) ‐ images come from a single illumination plane generated by ‘pinholes’, and therefore contain no ‘out‐of‐focus’ light. True z‐axis imaging ‐ expensive (€200,000 ‐ €1,000,000)
Wide‐field versus point scanning confocal microscopy Confocal
Wide‐field
detector efficiency
3‐20% (PMT)
60‐80% (CCD)
optical sectioning
yes
no
signal level
low
high
frame rate
low
high
optimal usage
one plane in thick specimen
thin sparsely stained specimen
main limitation
acquisition speed
out‐of‐focus light
Principles of confocal microscopy
Principles of confocal microscopy
Wide‐field
Confocal
human medulla section
muscle fibres
pollen grain
Principles of confocal microscopy Point scanning confocal
Spinning (Nipkow) disk confocal
‐ laser light is from a point source ‐ galvanometer mirrors are used to direct the light ‐ laser is ‘scanned’ across the sample ‐ image is detected on photomultipliers
‐ laser light is from an ‘incoherent’ source ‐ light is passed through rotating disks containing lenses ‐ only small areas of the sample are illuminated ‐ image is detected on a CCD camera
‐ ideally suited to bleaching applications
‐ ideally suited to rapid live cell imaging
Point scanning confocal microscopy
6. Detectors / PMTs
1. Lasers 5. Objective & microscope 3. Excitation pinholes 4. Scanner
2. AOTFs
Principles of confocal microscopy ‐ AOTF
‐ the acousto‐optic tunable filter (AOTF) is an electro‐optical device that functions as an electronically tunable excitation filter to simultaneously modulate the intensity and wavelength of multiple laser lines from one or more source ‐ changes in the acoustic frequency alter the diffraction properties of the crystal (tellurium dioxide), enabling very rapid wavelength tuning, limited only by the acoustic transit time across the crystal ‐ AOTFs provide the possibility of precisely regulating the excitation wavelength, thereby replacing the need for unwieldy filter mechanisms
Principles of confocal microscopy ‐ scan head
‐ the scan head lies at the heart of the confocal system ‐ it is responsible for controlling (rasterising) the excitation scans, and collecting the emitted photons from the sample ‐ laser scanning is controlled by galvanometer motor‐based raster scanning mirrors (one x‐ and one y‐axis)
Principles of confocal microscopy ‐ photomultiplier detectors
‐ photomultiplier tubes (PMTs) are devices that respond when photons impinge on a photocathode and liberate electrons ‐ these electrons are accelerated towards an electron multiplier composed of a series of curved plates (dynodes) ‐ the output from the dynode chain is a current proportional to the number of photons striking the photocathode and to the voltage drops along the dynode channel ‐ because of the high level of electron multiplication by the dynode chain (gain), photomultiplier detectors provide extremely high sensitivity to light and have exceptionally low noise when compared to other photosensitive devices
Spinning disk confocal microscopy
Spinning disk confocal microscopy ‐ the point scanning laser confocal microscope is limited in image acquisition speed due to the need for extremely precise control of the galvanometer mirrors
‐ only a limited number of photons are emitted by the specimen during the pixel dwell time
‐ typically point scanning confocal microscopes scan at the rate of 1 microsecond per pixel, which translates to acquisition speeds ranging from one‐half to two seconds per image, depending upon the dimensions ‐ spinning disk confocal microscopes overcome this by imaging the specimen with multiple excitation beams operating in parallel
Spinning disk confocal microscopy
‐ whereas the point scanning confocal builds an image sequentially one pixel at a time, the spinning disk confocal ‘multiplexes’ by simultaneously illuminating the entire specimen field with parallel pinholes
‐ a partial rotation of the disk scans the specimen with approximately 1,000 individual light beams that can traverse the entire image plane in less than a millisecond
‐ in general, spinning disk instruments are capable of speeds that exceed those of point scanning confocals by 100 to 1,000 times
‐ spinning disk microscopes have no strict requirement for laser illumination and can be used with broadband‐ wavelength arc discharge lamps, such as mercury, xenon, or metal halide
‐ by coupling a CCD camera to the microscope instead of a photomultiplier, spinning disk microscopes generate a parallel area array image with less photobleaching due to the fact that the specimen can be adequately illuminated with lower excitation intensity and the emission is detected with higher quantum efficiency
Spinning disk confocal microscopy ‐ the most advanced spinning disk instruments are engineered by Yokogawa Electric Corporation of Japan
‐ these confocal scanning units (CSUs) are equipped with a unique architecture that consists of two coaxially aligned disks featuring a dichromatic mirror positioned between the disks. Each disk contains approximately 20,000 pinholes
‐ the upper disk is actually a glass plate containing microlenses on the top surface that direct and focus light onto perfectly aligned 50‐micrometer pinholes in the lower disk for transmission to the objective and specimen
‐ because significantly more input illumination is gathered by the microlens disk than a regular Nipkow disk, the light throughput of this system approaches 40‐60% as opposed to the 4‐6% normally observed
Spinning disk confocal microscopy
Spinning Disk Confocal Example: Andor Revolution
High resolution 4D imaging Four laser lines: 445, 488, 514, 561 nm Yokogawa CSU22 EM CCD camera Climate control system
Point Scanning Confocal Example: Olympus Fluoview FV1000
High resolution 4D imaging Six laser lines: 405, 458, 488, 515, 559, 635 nm Twin scanners Spectral detection Modules for FRAP/FRET etc Climate control system
Olympus Fluoview FV1000 Climate control systems
Stage and scanner controllers
Lasers, power supplies and combiner
Hg and transmission light illumination
Olympus Fluoview FV1000 ‐ Interface Microscope settings / acquisition control
Acquisition settings
Data manager
Image view
Olympus Fluoview FV1000 ‐ Interface Scanning direction Pixel dwell time (scanning speed) Image size (and shape)
Zoom and rotation
Laser lines and laser power
Spectral (lambda) scanning
Microscope control (objectives, z‐position)
Time scanning (time lapse series)
Olympus Fluoview FV1000 ‐ Interface Scan start / stop
SIM mode
Stimulus
Confocal aperture
Confocal / epi / transmission
Spectral settings
Transmission lamp
PMT settings Kalman (averaging) Sequential scanning
Olympus Fluoview FV1000 ‐ Confocal Aperture ‐ confocal aperture / pinhole size measured in Airy Units (AU) or μm ‐ sectioning and signal are optimum at 1 AU ‐ when AU 1 there is a rapid decrease in optical sectioning quality
C.A. = 80μm
C.A. = 120μm
C.A. = 800μm
Olympus Fluoview FV1000 ‐ Pixel Dwell Time (Scan Speed) ‐ time spent collecting information for each pixel ‐ values normally in the microsecond range ‐ can be defined directly, or indirectly from scan speed (Hz) and image size ‐ important characteristic for signal intensity and quality
Dwell time = 2μs Scan time = 6.5s
Dwell time = 8μs Scan time = 18s
Dwell time = 12.5μs Scan time = 27s
Dwell time = 40μs Scan time = 80s
Olympus Fluoview FV1000 ‐ Pixel Number (Image Size) ‐ the size of the output image, and therefore the quality, is determined by the number of pixels ‐ however, the acquisition of large images is very time consuming, especially if multiple z‐slices are also wanted
512 x 512 pixels
800 x 800 pixels
1024 x 1024 pixels
2048 x 2048 pixels
Olympus Fluoview FV1000 ‐ Detectors ‐ acquisition is controlled by a PMT for each fluorescent channel, specifically through ‘high voltage gain’, ‘photomultiplier gain’, and ‘offset’ ‐ REMEMBER, if your image is not bright, it is always better to increase the sensitivity of the detector BEFORE increasing the output from the laser
‐ PMT detector sensitivity in controlled by the voltage across them ‐ the most linear response of the PMTs varies, but is typically between 600V and 800V ‐ the offset control is used to adjust the background level to a position near zero volts (black) by adding a positive or negative voltage to the signal ‐ the photomultiplier gain adjustment is used to electronically stretch the input signal by multiplying with a constant factor prior to digitisation by the analog‐to‐digital converter
Olympus Fluoview FV1000 ‐ In Summary
To reduce image noise ...
‐ reduce scan speed (increase pixel dwell time) ‐ increase image size (more pixels) ‐ use line averaging / frame averaging ‐ optimise pinhole ‐ increase laser intensity
... but remember that all these changes will increase sample photobleaching !
The 39 Steps ... of Fluorescence Microscopy
Pawley J (2000). BioTechniques 28:884‐886. The 39 Steps: A Cautionary Tale of Quantitative 3‐D Fluorescence Microscopy.
Key take home point
In recent years advances in confocal microscopy have revolutionised cellular imaging, with specific systems designed and optimised for live cell imaging applications