prepared by
Markus Dürrenberger and Rosmarie Sütterlin
INTRODUCTION
Confocal laser scanning microscopy (CLSM) is a relatively new light microscopical imaging technique (introduced around 1980 by M. Petran and A. Boyde) which has found wide applications in the biological sciences [c.f. Pawley,1990; Boyde, 1994]. The primary value of the CLSM to the biologist is its ability to produce optical sections through a 3-dimensional (3-D) specimen - e.g., an entire cell or a piece of tissue - that, to a good approximation, contain information from only one focal plane. Therefore, as illustrated schematically in Fig. 1, by moving the focal plane of the instrument step by step through the depth of the specimen, a series of optical sections can be recorded [c. f. Lichtman, 1994]. This property of the CLSM is fundamental for solving 3-D biological problems where information from regions distant from the plane of focus can obscure the image (thick objects). With biological specimens, either the epi-fluorescence or the epi-reflection mode is generally employed. As a valuable by-product, the computer-controlled CLSM pro duces digital images which are amenable to image analysis and processing, and can also be used to compute surface- or volume-rendered 3-D reconstructions of the specimen.
PRINCIPLE
In Fig. 2, the confocal principle is illustrated schematically for the epi-fluorescence
imaging mode [c.f. Wilson & Sheppard, 1984; Lichtman, 1994]. To image the
specimen point by point, a collimated, polarized laser beam is deflected stepwise in the x- and y-direction by a scanning unit (not shown) before it is reflected by a dichroic mirror (beam splitter) so as to pass through the objective lens of the microscope,
and focused onto the specimen. The emitted, longer-wavelength fluorescent
light collected by the objective lens passes through the dichroic mirror
(transparent for the longer wavelength) and is focused into a small pinhole (i.e., the confocal aperture) to eliminate all the out-of-focus light, i.e., all light coming from regions of the specimen above or below the
plane of focus. Therefore, the CLSM does not only provide excellent resolu
tion within the plane of section (0.25 mm in x- and y-direction), but also
yields similarly good resolution between section planes (0.3 mm in z-direction).
The in-focus information of each speci men point is recorded by a light-sensitive detec tor (i.e., a photo-multiplier) positioned behind the confocal aperture, and
the analog output signal is digitized and fed into a computer. At the same time, the analog photo-multiplier signal can be used to generate a TV-like image on a video
monitor. The obvious advantage of having a stack of serial optical sections
through the specimen pixel by pixel in digital form is that either a composite projection image can be computed (Fig. 3), or a volume-rendered 3-D representation of the specimen can be generated on a graphics computer (see title page:
representing a shadow-projection image of dividing cells with the spindle
apparatus being labeled with a fluorescent antibody to tubulin and the condensed
chromatin with the DNA stain
TOTO-1).
Figure 1. Optical sectioning of a sphere by confocal planes.
Figure 2. The principle of a confocal laser scanning microscope.
The confocal part of a CLSM consists of an elaborate, highly folded optical
bench on which the laser, all the filters, an oscillating-mirror or acousto-optic scanning device , and the detector are mounted. When working in the epi-fluorescence mode, the laser beam is filtered to select the 488 nm, 568 nm or 647 nm
wavelength line from an Argon/Krypton laser, and a triple dichroic mirror
(to fit all three excitation wavelengths) is used to transmit - rather than
reflect - the longer-wavelength fluorescence signal to the detector. For
the epi-reflection mode no wavelength filters are needed. Instead, a semi-transparent mirror reflects 50% of the incident laser beam through the objective lens and
to reach the specimen, and it transmits 50% of the light reflected by the
specimen and collected by the objective lens to the detector. To suppress
light reflected by the various optical elements of the microscope, a 1/4-wavelength plate and a polarizing filter are put into the beam path.
Usually, objective lenses with a high numerical aperture (NA) are used to provide good resolution in x-, y- and z-direction (e.g., a 63x/1.4 NA Planapochromat for epi-fluorescence, or a 40x/1.0 NA for epi -reflection). Often, good specimen areas are sparse, so to find them the specimen is screened in the conventional transmission or epi-fluorescence mode. Once a good region has been located, the CLSM mode is activated, and serial optical sections are recorded at user-selectable depth increments which can be as small as 0.02 mm. In digital form, each image element yields an 8-bit intensity value in the range 0-255. Typically, the image frame size is 512x512 pixels, and in the best case images are recorded at video rate, i.e., at 25 frames per second in PAL or 30 frames per second in NTSC norm. To improve the signal-to-noise (S/N) ratio of individual image frames, several of them may be recorded in series and averaged. In the case of epi -fluorescence imaging, the total number of scans is generally adjusted to limit specimen bleaching to an acceptable level.
APPLICATIONS: From bone, cartilage and muscle to the cytoskeleton
In general, thick and opaque specimens that can barely be observed in a conventional light microscope are excellent specimens when it comes to demonstrate the power of a CLSM. For example, 20-25 mm thick sections of bone, cartilage or muscle are ideally suited for 3-D imaging in the CLSM. Independent of the thickness and surface quality of such tissue sections, individual confocal planes readily reveal a lateral resolution of 0.3 mm. By recording its auto-fluorescence, even a piece of wood can be optically sectioned to a depth of about 100 mm.
Cultured fibroblasts grown as monolayers can be multiple-labeled with fluorochrome-tagged antibodies against different cytoskeletal components: e.g., actin (labeled with rhodamine > red fluorescence) or tubulin (labeled with fluorescein > green fluorescence). In such cells, actin is primarily in the form of stress fibers (red fluorescence), actin filament bundles adhering to the plasma membrane attaching the cell to the coverslip on which it has been grown. In contrast, the microtubules (green fluorescence) usually radiate outward from the two perinuclear centrioles and reach all parts of the cell. During cell division the microtubules form the spindle apparatus that separates the condensed chromatin (i.e., the chromosomes).
The CLSM allows the 3-D distribution and relative spatial relationship of these two filament systems to be visualized directly.
Exploring the growth and differentiation of cultured cells in 3-D collagen matrices - a condition which more closely resembles the natural environment of cells - has been another application of the CLSM (Fig. 3). While difficult to assess by conventional fluorescence microscopy, the spatial relationship of the various cytoskeletal components relative to the different sub-cellular compartments in these matrix-embedded cells is readily determined by CLSM.
Figure 3. Seven representative optical sections selected from 81 confocal planes (corresponding to a depth of 50 mm) "cut" through a collagen matrix containing growing fibroblasts labeled with fluorescent antibodies to tubulin. Inset, composite shadow-projection image of all 81 confocal sections revealing the spindle apparatus of dividing cells and the regular microtubular network of interphase (i.e., non-dividing) cells.
REFERENCES
Boyde, A. (1994). Bibliography on confocal microscopy and its applications. Scanning 16, 33-56.
Lichtman, J.W. (1994). Confocal microscopy. Scientific American 271, 40-45.
Pawley, J.B. (1990). The Handbook of Biological Confocal Microscopy. Plenum, New York.
White, J.G., Amos, W.B. & Fordham, M. (1987). An evaluation of confocal versus conventional imaging of biological structures by fluorescence light microscopy. Journal of Cell Biology 105, 41-48 .
Wilson, T. & Sheppard, C. (1984). Theory and Practice of Scanning Optical
Microscopy. Academic Press, London.