Optical sections are produced in the laser-scanning confocal microscope by scanning the specimen point by point with a laser beam focussed on the specimen, and using a spatial filter, usually a pinhole or a slit, to remove unwanted fluorescence from above and below the focal plane of interest (Figure 1). The power of this confocal approach lies in the ability to image structures at discrete levels within an intact biological specimen.
Fig1. Information fl ow in a generic laser-scanning confocal microscope. Light from the laser passes through a neutral density filter and an excitation filter on its way to the scanning unit. The scanning unit produces a scanned beam at the back focal plane of the objective lens, which focusses the light at the specimen. The specimen is scanned in the x - and the y -direction in a raster pattern, and in the z -direction by fi ne focussing. Any fluorescence from the specimen passes back through the objective lens and the scanning unit and is directed via dichromatic mirrors to three pinholes. The pinholes act as spatial filters to block any light from above or below the plane of focus in the specimen. The point of light in the specimen is confocal with the pinhole aperture. This means that only distinct regions of the specimen are sampled. Light that passes through the pinholes strikes the photomultiplier tube (PMT) detectors and the signal from the PMTs is built into an image in the computer. The image is displayed on the computer screen often as three monochrome images (K1 – K3) together with a merged colour image of the three monochrome images (K4) and. The computer synchronises the scanning mirrors with the build-up of the image in the computer frame store. The computer also controls a variety of peripheral devices; for example, it correlates movement of a stepper motor connected to the fi ne focus of the microscope with image acquisition in order to produce a Z-stack . Furthermore, the computer controls the area of the specimen to be scanned by the scanning unit so that zooming is easily achieved by scanning a smaller region of the specimen. In this way, a range of magnifications is imparted to a single objective lens so that the specimen does not have to be moved when changing magnification. Images are written to a hard disk and exported to various devices for viewing, hard-copy production or archiving.
There are two major advantages of using laser-scanning confocal microscopy in preference to conventional epifluorescence light microscopy. First, glare from out- of-focus structures in the specimen is reduced and resolution is increased, both laterally in the x - and the y -directions (0.14 μm) and axially in the z -direction (0.23 μm). Second, the image quality of relatively thick specimens such as fluorescently labelled multi-cellular embryos is substantially improved. Note that for thinner specimens, though, such as for example chromosome spreads and the leading lamellipodium of cells growing in tissue culture (<0.2 μm thick), there is no significant improvement of image quality when using laser-scanning confocal microscopy.
For successful confocal imaging, a minimum number of photons should be used to efficiently excite each fluorescent probe in the specimen, and as many of the emitted photons from the fluorophores as possible should make it through the light path of the instrument to the detector.
Laser-scanning confocal microscopy has found many different applications in bio medical imaging. Some of these applications have been made possible by the ability of the instrument to produce a series of optical sections at discrete steps through the specimen (Figure2). This Z-series (Z-stack) of optical sections collected with a confocal microscope are all in register with each other, and can be merged together to form a single projection of the image (Z-projection) or a three-dimensional representation of the image (3D reconstruction).
Fig2. Computer 3D reconstruction of confocal images. (a) Sixteen serial optical sections collected at 0.3 μm intervals through a mitotic spindle of a PtK1 cell stained with antitubulin and a second rhodamine labelled antibody. Using the Z-series macro program, a preset number of frames can be summed, and the images transferred into a fi le on the hard disk. The stepper motor moves the fi ne focus control of the microscope by a preset increment. (b) Three-dimensional reconstruction of the dataset produced using computer 3D reconstruction software. Such software can be used to view the dataset from any specified angle or to produce movies of the structure rotating in three dimensions.
Multiple label images can be collected from a specimen labelled with more than one fluorescent probe using multiple laser light sources for excitation (Figure 3). Since all of the images collected at different excitation wavelengths are in register, it is relatively easy to combine them into a single multi-coloured image. Here, any overlap of staining is viewed as an additive colour change. Most confocal microscopes are able to routinely image three or four different wavelengths simultaneously.
Fig3. Optical sections produced using laser-scanning confocal microscopy. Comparison of alkaline phosphatase (a) and tyramide-amplified detection of mRNAs (b,c). Staining patterns obtained using DIG labelled antisense probes directed against the CG14217 mRNAs, through conventional AP-based detection (a) or tyramide signal amplification (b), using tyramide-Alexa Fluor 488 (green fluorescence). Close-up images of tyramide-amplifi ed samples are also shown (c). In (b) and (c), nuclei were labelled in red with propidium iodide. Triple-labelled Drosophila embryo at the cellular blastoderm stage (d,e,f,g) . The images were produced using an air-cooled 25 mW krypton argon laser, which has three major excitation wavelengths at 488 nm (blue), 568 nm (yellow) and 647 nm (red). The three fluorophores used were fluorescein (λexc = 496 nm; λem = 518 nm), lissamine rhodamine (λexc = 572 nm; λem = 590 nm) and Cyanine 5 (λexc = 649 nm; λem = 666 nm). The images were collected simultaneously as single optical sections into the red, the green and the blue channels, respectively, and merged as a three colour (red/green/blue) image (Figure 1). The image shows the expression of three genes; hairy (in red), Krüppel (in green) and giant (in blue). Regions of overlap of gene expression appear as an additive colour in the image, for example, the two yellow stripes of hairy expression in the Krüppel domain (f,g). (Images a,b and c were kindly provided by Henry Krause, University of Toronto, Canada.)
The scanning speed of most laser-scanning systems is around one full frame per second. This is designed for collecting images from fixed and brightly labelled fluorescent specimens. Such scan speeds are not optimal for living specimens, and laser-scanning instruments are available that scan at faster rates for more optimal live-cell imaging. In addition to point scanning, swept fi eld scanning rapidly moves a micrometre-thin beam of light horizontally and vertically through the specimen.
