Confocal Microscopy in Biomedical Research Paul J

COMMENT THIS ARTICLE SEE COMMENTS ON THIS ARTICLE CONTACT AUTHOR

March 1999 (Volume 40, Number 3) Confocal Microscopy in Biomedical Research Paul J. Rigby, Roy G. Goldie Biomedical Confocal Microscopy Research Centre, Department of Pharmacology, University of Western Australia, Nedlands, Western Australia, Australia

Confocal microscopy has allowed a major advance in biological imaging, since it represents a rapid, cost effective means of ecamining thick tissue specimens. In most cases, this involves fluorescence imaging and it is increasingly being used as a basic tool in biomedical research. Confocal microscopy allows the collection of thin optical sections, without the need for physical sectioning of the tissue. Additionally, confocal microscopes can usually produce images with greater sensitivity, contrast and resolution than those produced with normal light microscopes. We attempt to explain how this technology might be better used as a routine research tool. Since high quality, in-focus optical sections of thick tissue preparations can be generated quickly, confocal microscopy, in combination with immunofluorescence histochemistry, can now be used to examine complex three-dimensional distributions of distinct structures within tissues such as nerves within airways. Additionally, ultraviolet confocal microscopy allows the assessment of both dynamic and static phenomena in living cells and tissues. Thus, in addition to the imaging of fluorescence associated with structural elements, confocal microscopes can be used to quantitatively evaluate the distribution and fluxes of intracellular ions like calcium. Rapid, line-scanning confocal microscopes can be used in the assessment of dynamic events. For example, the in vivo imaging of microvascular permeability in airways becomes possible for the first time. By providing examples of some of our uses for confocal microscopy, we might encourage others to explore this relatively new and important texhnology for examining events and structures in single cells, tissue samples and in intact animals.

Key words: bronchi; calcium; immunofluorescence microscopy; fluorescence microscopy; microscopy, confocal; muscle, smooth; substance P; trachea

Light and electron microscopy have long been important research tools used in analyzing biological specimens and both techniques have provided valuable insights into cellular biology, structure, and function. Although electron microscopy offers excellent resolution, artifacts introduced in specimen preparation (fixation and sectioning) can severely limit some applications of this technique to biological and, in particular, to living specimens. Conventional wide field microscopy is routinely used to examine living tissue, but suffers from the physical resolution limits imposed by using visible light as the illumination source. Additionally, conventional microscopy of living or thick specimens suffers enormously from out-of-focus light degrading the image quality. Furthermore, in the absence of enormous operator effort, both of these techniques reveal little information about the three dimensional nature of biological specimens. In 1957, the first confocal microscope was patented by Marvin Minsky, and commercial systems for cell biologists only became available in 1987 (1). These developments heralded a quantum leap in the quality of light microscopic data that could be obtained from biological specimens. The impact of the enormous improvement in the efficiency of data collection and in the quality and power of the data provided to the researcher by such efficient image production and processing, cannot be over- estimated. It is perhaps little wonder that confocal microscopy has become so popular in less than ten years. Confocal Microscopy: Principles and Advantages Traditionally, in conventional microscopy, a condensor lens is used to uniformly and simultaneously illuminate a wide area and volume of the specimen (hence the term wide-field microscopy). In thick specimens, this results in out-of-focus blur originating from volumes above and below the plane of focus. This out-of-focus light can reduce contrast and resolution and can severely detract from the interpretation of fine microscopic detail. In an effort to reduce these problems, conventional microscopy has routinely utilized thin or ultra-thin sections of tissues to enable clear views of the tissue structure to be obtained. In confocal microscopes, the illumination is both sequential and focused on a small volume of the tissue, with regions away from the focal plane receiving less illumination, thus reducing the out-of- focus blur. Additionally, both the illumination and detection systems are focused on the same single volume element of the specimen (2). Thus, the illumination, specimen, and detector are all focused on the same volume and are therefore confocal. With the addition of a carefully aligned aperture at a focal point in the optical path, further reduction in out-of-focus information is achieved (Fig 1).

Figure 1. The principle of confocal microscopy. A light source, typically a laser, is reflected by a dichroic mirror or beam splitter and is brought to a point of focus by the objective lens at the level labeled ”Plane of Focus”. Fluorescence is emitted by the specimen from the point of focus (solid rays) and passes back through the objective lens, through the dichroic mirror and confocal aperture to the detector. However, fluorescence will also be emitted from planes above and below the plane of focus (dotted and dashed lines), but this light will be prevented from reaching the detector by the confocal aperture. To collect a two-dimensional (2D) image, the laser beam is usually scanned across the specimen in a ”rastor” pattern by a pair of mirrors (not drawn). To generate serial optical sections, the plane of focus is moved in the Z direction after each 2D image is collected.

The net result is that a confocal microscope can usually produce thin, optical sections with greater resolution and contrast and with greater sensitivity than conventional, wide-field microscopes. This means that confocal microscopy can allow the assessment of tissue structures within thick specimens using non-invasive optical sectioning techniques, such that fixation and sectioning of specimens is no longer mandatory. Given that tissue fixation is often not required for the purposes of confocal fluorescence imaging, it is now possible to image events within and around living cells in various preparations including thick tissue slices. In addition, images can be obtained up to 200 mm into the outer layer of whole organs using serial optical sections that can be rapidly processed by computer graphics programs into three dimensional (3D) reconstructions. Confocal microscopy can improve lateral (x,y) resolution to approximately 200 nm (compared to approximately 400 nm in wide field microscopy). Additionally, image resolution in the z (vertical) axis is also improved (approximately 500 nm). This improvement is at least partially obtained through rejection of out-of-focus light, particularly in thick tissue sections or in living specimens (1,3). Figure 1 is a schematic representation of the light paths involved in obtaining a confocal image. Out-of-focus light originating from the specimen is blocked by an aperture before it reaches the detector. Only light originating from the plane of focus is used to create the image, i.e., an optical section (Fig. 2a,b). Additional (serial) in-focus optical sections can be produced by changing the focus of the microscope in the z-axis to produce a z-series. This can then be written to a computer and manipulated to produce an extended focus image (projection) with great depth of field (Fig. 2c).

Figure 2. Confocal images taken from a ”whole mount” of rat tracheal tissue stained with an antibody to the neuronal protein marker PGP 9.5. The primary antibody has subsequently been localized with a second antibody labeled with fluorescein isothiocyanate (FITC). A 60 mm ”stack” or Z-series of images was collected at 2 mm intervals using a Bio-Rad MRC-1000/1024 UV confocal microscope. Single optical sections were taken (A) 18 mm and (B) 56 mm below the tissue surface. (C) A maximum intensity projection (extended focus image) of the entire data set. This projection method is most simply explained by considering each image of the original data set being mounted on individual transparent sheets. An image similar to the computer generated projection could be produced by shining a light through the transparent sheets and ”projecting” an image onto a wall. Scale bars = 50 µm.

Confocal fluorescence microscopy is ideal for the rapid assessment of the distribution and localization of extracellular and intracellular macromolecules, including proteins and nucleic acids, as well as intracellular ions such as calcium. The production of optical tissue sections in a few minutes, rather than the production of expensive and time-consuming slide-mounted tissue sections, means that the original tissue sample remains intact and can thus be repeatedly examined. Such sections can be rapidly combined to provide 3D tissue reconstructions of multiple cellular or extracellular entities. Extracellular tissue elements, surface, and intracellular profiles can be obtained following the use of multiple fluorescently tagged antibodies targeting specific antigens. For example, cell type-specific proteins such as neuronal protein gene product 9.5 (PGP 9.5) can be labeled with anti-PGP 9.5 antibodies. Similarly, peptides released as putative neurotransmitters, e.g., vasoactive intestinal peptide (VIP) and substance P (SP), can be simultaneously labeled and detected. This can also be achieved for standard fluorescence microscopy in slide-mounted sections, but confocal microscopy now allows the spatial relationships between these components to be appreciated ex vivo and undistorted by potential fixation artifacts. Combined with enormous time and financial efficiencies, this allows exploration of research areas previously considered unfeasible. Visualization of Confocal Images As previously described, confocal microscopes collect data from in-focus information. This can then be used to create images as optical sections in two dimensions (2D). However, to discern more than a fraction of the fluorescence associated with the antigen of interest, 3D images must be constructed. This in turn requires the collection of multiple serial optical sections through the thickness of the specimen which are then projected to form a representation of the whole 3D data set. The process of 3D reconstruction requires the use of specialized commercial software (e.g., AVS/Express, ImageSpace, and VoxelView) or public domain packages such as VolVis and Confocal Assistant. It is not our intention to discuss image processing and visualization systems in this article, and the reader is directed to several other reviews (4,5). Slit Scanning Confocal Microscopy Ideally, confocal microscopy should combine high resolution and sensitivity with high speed data collection and processing. However, at present, these properties cannot be found together in a single microscope design. It is most usual to require optimization of image resolution and sensitivity, properties that demand a compromise in image acquisition speed. Point scanning confocal microscopes collect information from a series of discrete points by scanning the laser beam across the specimen. At present, the maximum rate at which such point scanning can collect a full image field is approximately 1 Hz. Despite this, there remains a significant demand for higher speed data acquisition in cases where events are occurring very rapidly. Slit scanning confocal microscopes are ideal where fast image acquisition is required e.g., when monitoring the fast transit of fluorescing proteins into or out of cells or tissues. Since these microscopes scan the specimen using a line of laser light, data from consecutive full field images can be collected in real-time, i.e., at rates >30 Hz. Whereas this results in some decrement in image resolution, an added advantage of such systems is that events occurring within cells or tissues that are moving (e.g., contracting or pulsating) can be imaged without image distortion. Slit scanning and point scanning confocal microscopes are complementary rather than alternative competing technologies. Clearly, slit scanning confocal microscopy is not ideally suited to all applications, since poorer axial resolution and greater potential for photobleaching when viewing a specimen for extended periods (6) can be significant drawbacks. Slit scanning confocal microscopy is particularly useful where rapidly occurring events must be imaged in real-time. For this task, high sensitivity, cooled CCD cameras are required to deliver slit scanned confocal data to fast image capture computers. This allows later analysis and quantitation of data describing fast biological events that would otherwise be missed using slower point scanning technology. Slit scanning confocal microscopy can also offer the possibility of viewing confocal fluorescent images directly through standard microscope eyepieces as occurs with conventional fluorescence microscopes. This removes the need to view the image on a computer screen and increases the rate at which multiple specimens can be screened, e.g., by pathologists looking for particular structural or staining features. In addition and depending upon the microscope design, direct viewing in both the x and z directions within thick specimens (i.e., optical cross-sectioning) can possibly be achieved without the need for computer reconstruction. However, very fast and accurate focusing mechanisms need to be employed to realize the potential benefits of this feature. Immunofluorescence Histochemistry Immunofluorescence histochemistry requires association of fluorescently labeled antibodies with specific antigenic substances within a specimen. Fluorescence associated with an antigen-antibody complex is then detected as emitted light following excitation of the fluorophore with incident light of a particular wavelength. The optical sectioning technology and the high sensitivity of photomultipliers used in confocal microscopes makes them ideal for evaluations of the distribution of immunofluorescently labeled cell surface or intracellular macromolecules. Antigens in specimens are usually detected by first labeling specimens with a highly specific primary antibody, which is then itself tagged with a second antibody carrying a suitable fluorophore. This process can be repeated for multiple antigens, particularly where the primary antibodies are highly selective in their targeting. In our laboratory, we are particularly interested in neuronal function and distribution in the respiratory tract. Activation of mucosal and submucosal nerves of various types has marked effects on airway wall function, since sites including airway smooth muscle, microvessels of the bronchial circulation and mucous glands are stimulated. Perturbation of neuronal bronchial function as a result of the impact of inflammation, such as is evident in asthma, may be significant. Confocal microscopy can be used to assess asthma-associated changes in the distribution and density of neuronal pathways in the respiratory tract, since these nerves can be selectively imaged in 3D and their volume density evaluated (7). Extensive tracking of meandering neuronal pathways within a bronchial wall mass is a major undertaking unlikely to be feasible using conventional tissue labeling, embedding, and sectioning methodologies. In contrast, such a project is practicable using confocal microscopy, since serial optical sections can be produced within minutes. Confocal images of nerves stained for the neuronal cell marker PGP 9.5 can be rapidly obtained revealing the fine detail of the network of nerves and terminal varicocities (8). By using anti-PGP 9.5 in conjunction with a specific monoclonal antibody to smooth muscle a-actin, the relationships between nerves and airway smooth muscle in the airways can be assessed (Fig. 3). Routine histological methods, including serial sectioning may never provide an image of this quality despite the prosecution of a lengthy and costly study. However, confocal microscopy which employs non-destructive optical sectioning/techniques and the application of digital imaging technology, results in a high quality product in minutes.

Figure 3. Photomicrograph depicting the relationship between nerve pathways (green) and airway smooth muscle (red) within a ”whole mount” preparation of rat tracheal wall. Nerves were labeled with a primary antibody to PGP 9.5 and a fluorescein-labeled (FITC) secondary antibody. Airway smooth muscle actin was detected using an anti-a-actin antibody and a rhodamine-labeled (TRITC) secondary antibody. Scale bar = 50 µm.

Ultraviolet (UV) Confocal Microscopy Intracellular Ion Imaging Rapid shifts in intracellular Ca2+ concentrations are involved in the regulation of muscle contraction, cell motility, fertilization and cell division, and many other important cellular functions (9-11). However, intracellular Ca2+ concentrations in resting cells are very low (in the nanomolar range). Accordingly, accurate assessment of changes in these levels requires a very sensitive and accurate technique. With the introduction of selective ion chelating fluorescent dyes, it became possible to measure intracellular ions in living cells and tissues. Standard wide-field fluorescence microscopes were commonly used and, in an effort to overcome problems of dye redistribution and/or fading, ratiometric probes were developed. Fura-2 is perhaps one of the most commonly used wide-field microscopy ratiometric calcium probes (12). This dye has two excitation peaks in response to incident UV light, such that the ratio of the two fluorescent signals is directly related to intracellular Ca2+ concentration. This ratio is totally independent of the dye concentration within the cell. This approach minimizes variables, including cell thickness and photo-bleaching (fading) of the dye during the experiment, which complicate the assessment of intracellular Ca2+ concentration. However, the use of ratiometric probes in combination with wide field microscopes only partially improves the imaging of intracellular ions. Confocal microscopic images, particularly in thicker specimens, are not compromised by the blur that reduces the quality of data derived from conventional fluorescence microscopy. Therefore, single wavelength (non-ratiometric) dyes such as Fluo-3 were then developed to work with the wavelengths available from the lasers commonly used in standard confocal microscopy. These single wavelength probes are at best semiquantitative and their use necessitates that corrections be made for variations in intracellular dye concentration and photobleaching (12). Ultraviolet confocal fluorescence microscopy in combination with UV ratiometric fluorescent probes is ideally suited for the detection of intracellular ion fluxes, given its great sensitivity and capacity to finely resolve fluorescence within a clearly defined volume of the specimen. Indo-1 is a ratiometric intracellular Ca2+ probe that can be used with specialized UV confocal microscope systems (13). However, unlike Fura-2, Indo-1 is excited with a single UV wavelength of light (351 nm) that is readily available from an argon ion laser. Fluorescence emission intensity measured at 405 nm (violet) increases as intracellular Ca2+ rises, but decreases when measured at 490 nm (blue) (Fig. 4). Thus, a ratiometric approach can again be used to accurately quantify changes in intracellular Ca2+ concentrations and distributions within cells (Fig. 4). A further major advantage of confocal imaging in this field is that a visual appreciation of the relationships between different cellular and/or intracellular ion components can now be gained.

Figure 4a A series of paired confocal images of human cultured lung epithelial cells, shown at 15 second intervals. The upper row of images shows the 405 nm (purple band) emitted fluorescence of the calcium indicator Indo-1. The lower row shows the emission at 490 nm (blue band). A rise in the concentration of calcium ions was induced by the application of a mediator and is apparent as an increase in the intensity of the 405 nm fluoresence with a corresponding decrease in the 490 nm fluorescence. Figure 4b. Mediator-induced change in calcium ion concentration in human cultured lung epithelial cells. The graph clearly shows an increase in the 405 nm fluorescence band and a corresponding decrease in the 490 nm fluorescence of the indicator Indo-1 following stimulation. The ratio of these two signals (405 nm/490 nm) is a function of calcium concentration, which can be quantified by means of a calibration procedure. These data were collected using the Bio-Rad Time Course Software Module (TCSM) for the confocal microscope. The arrows indicate the time points at which images in Fig. 4a were collected. (Reproduced with permission from Goldie et al., Pulm Pharmacol Ther 1997;10:175-88.)

Using suitable fluorescent probes, both UV and visible light confocal microscopy can be used to measure other intracellular ions including hydrogen ions (pH), magnesium and zinc. (12, 14-16). Additionally, the more common fluorescent histological stains AMCA, DAPI, Hoechst 33258 and 33342, and Fast Blue, which are excited by UV light can also be visualized. This increases the range of fluorophores that can be used in confocal microscopes, and is particularly valuable for labeling specimens with multiple probes. Studies Involving the Use of ”Caged” Compounds Fluorescence microscopy is the method of choice for the detection of compounds released intracellularly from a ”caged” source. Such compounds enter cells either by microinjection or as highly lipid soluble esters and are inactive. The active molecule is subsequently released from the ”cage” in response to an energizing stimulus such as UV light. Particular cells or cell clusters in a culture can be selectively targeted with UV light of a specified wavelength to activate the release of the compound. UV confocal fluorescence microscopy is ideally suited to the assessment of the effects of such mediators, since it incorporates the targetable UV light source and allows for the capture of in- focus fluorescence image data. Currently available photoactivatable probes include caged nucleotides (e.g., caged ATP), caged neurotransmitters (e.g., caged carbamoyl choline), caged nitric oxide donors, caged calcium, and caged calcium buffers. Caged compounds may be released in highly localized defined regions within a cell when used in conjunction with UV confocal microscopy. In Vivo Confocal Microscopy Assessment of Microvascular Permeability The permeability of tissue microvessels increases in inflammation as part of a natural process of recruitment of various plasma-derived molecules and inflammatory cells including macrophages (17). However, several disease states including asthma also involve chronic and often difficult to control inflammatory processes, with significantly elevated microvascular permeability leading to tissue oedema. In the case of asthma, bronchial wall edema seems to play a role in airway obstruction, with several neurotransmitter and other mediators perhaps playing a significant role (18). Accordingly, there has been considerable interest in studying this phenomenon with a view to devising effective therapeutic interventions. However, current methods are essentially static approaches, i.e., they do not provide continuous data in real-time and so the temporal dynamics of this process remains poorly understood. For example, although previous studies have measured microvascular permeability changes in response to sensory nerve stimulation and to applied agonists (19-22), the data have usually described the permeability changes at single time points for single doses of agonist. Such studies are extremely valuable but have not provided real-time information concerning the dynamics of response onset, maintenance, or resolution. We have used confocal microscopy in a novel way to simultaneously visualize the plasma leakage phenomenon in response to mediators, measure the sensitivity of airway microvessels to leakage- provoking stimuli, and evaluate the response peak and time course of these events as they occur in real time. The primary advantage of a confocal microscopic approach is that, by removing out of focus blur, it provides high resolution images of living microvascular tissue in situ. Plasma leakage from microvessels in a respiring, anaesthetized animal can be monitored as the accumulation of fluorescently tagged plasma albumin (Fig. 5a-c) (4,23) or fibrinogen. Most importantly, these measurements can be made continuously as the leakage response develops, peaks, and resolves (Fig. 5d). It is hoped that this approach will provide greater understanding of the characteristics of airway microvascular permeability changes and their role in airway disease.

Figure 5. (A) Confocal image of tracheal microvessels imaged in situ in an anaesthetized rat. The airway microvessels were detected following intravenous injection of fluorescein-labeled plasma albumin. The same microvessels imaged (B) 30 seconds and (C) 60 seconds after the intravenous administration of substance P (SP; 1 mg/kg). Fluorescent albumin can now be clearly seen in the extravascular space, having leaked from the microvessels in response to SP. Figure 5d. The time-dependence of fluorescent plasma extravasation from rat tracheal microvessels can be quantified by integration of the fluorescence intensity signal in fields selected within and/or surrounding the vascular space. Reproduced with permission from Goldie et al., Pulm Pharmacol Ther 1997;10:175-88.

Summary Confocal microscopy is a relatively new technology that has the potential to vastly improve the quality of research output. Importantly, the enormous savings in labor time and costs associated with specimen preparation, expands the horizon for the researcher, since projects previously considered unfeasible can now be considered as practical and viable options. Other significant benefits include the retrieval of data in the form of high resolution fluorescence images that can be prepared as projections of 2D images with enormously increased depth of field (3D impression). Extracellular, cell surface and/or intracellular molecules can now be detected in both living and fixed tissue, without the need to destroy the tissue specimen by physical sectioning. Applications of confocal microscopy are only limited by the researcher's imagination, with the technology applicable at levels ranging from the single cell to the whole animal.

Acknowledgements The authors acknowledge the funding support to the Biomedical Confocal Microscopy Research Centre (BCMRC) from the Lotteries Commission of Western Australia, the Medical Research Foundation of Western Australia (MEDWA), the Arnold Yeldham and Mary Raine Medical Research Foundation, and from the National Health & Medical Research Council (Australia).

References 1 Inoué S. Foundations of confocal scanned imaging in light microscopy. In: Pawley J, editor. Handbook of biological confocal microscopy. 2nd ed. New York: Plenum Press; 1995. p. 1-17. 2 Wright SJ, Centonze VE, Stricker SA, DeVries PJ, Paddock SW, Schatten G. Introduction to confocal microscopy and three-dimensional reconstruction. In: Matsumoto B, editor. Cell biological applications of confocal microscopy. Meth Cell Biol 1993;38:1-45. 3 Sheppard CJR. Confocal microscopy: Principles, practice and options. In: Mason WT, editor. Fluorescent and luminescent probes for biological activity. A practical guide to technology for quantitative real time analysis. London: Academic Press; 1993. p. 229-36. 4 Goldie RG, Rigby PJ, Pudney C, Cody SH. Confocal microscopy and the respiratory tract. Pulm Pharm Ther 1997;10:175-88. 5 White NS. Visualization systems for multidimensional CLSM images. In: Pawley J, editor. Handbook of biological confocal microscopy. 2nd ed. New York: Plenum Press; 1995. p. 211-54. 6 Amos WB, White JG. Direct view confocal imaging systems using a slit aperture. In: Pawley J, editor. Handbook of biological confocal microscopy. 2nd ed. New York: Plenum Press; 1995. p. 403- 15. 7 Goldie RG, Spalding LJ, Pudney CJ, O' Connor BJ, Page CP. Confocal microscopy and the quantitative assessment of neuron densities in human asthmatic and non-asthmatic bronchus. Eur Respir J 1998;12 (suppl 28):453S, Abstract p. 2943. 8 Weichselbaum M, Everett AW, Sparrow MP. Mapping the innervation of the bronchial tree in fetal and postnatal pig lung using antibodies to PGP 9.5 and SV2. Am J Respir Cell Mol Biol 1996;15:703- 10. 9 Williams DA, Cody SH, Dubbins PN. Introducing and calibrating fluorescent probes in cells and organelles. In: Mason WT, editor. Fluorescent and luminescent probes for biological activity. A practical guide to technology for quantitative real time analysis. London: Academic Press; 1993. p. 320-34. 10 Nuccitelli R, editor. A practical guide to the study of calcium in living cells. Meth Cell Biol 40. San Diego: Academic Press; 1994. 11 Schild D. Laser scanning microscopy and calcium imaging. Cell Calcium 1996;19:281-96. 12 Tsien RY, Waggoner A. Fluorophores for confocal microscopy. Photophysics and photochemistry. In: Pawley J, editor. Handbook of biological confocal microscopy. 2nd ed. New York: Plenum Press; 1995. p. 267-79. 13 Girard S and Clapham DE. Simultaneous near ultraviolet and visible excitation confocal microscopy of calcium transients in Xenopus oocytes. In: Nuccitelli R, editor. A practical guide to the study of calcium in living cells. Meth Cell Biol 1994;40:263-84. 14 Haugland R. Intracellular ion indicators. In: Mason WT, editor. Fluorescent and luminescent probes for biological activity. A practical guide to technology for quantitative real time analysis. London: Academic Press; 1993. p. 34-43. 15 Kurtz I, Emmons C. Measurement of intracellular pH with a laser scanning confocal microscope. In: Matsumoto B. editor. Cell biological applications of confocal microscopy. Meth Cell Biol 1993;38:183-93. 16 Haugland RP. Handbook of fluorescent probes and research chemicals. 6th ed. Eugene (OR): Molecular Probes Inc; 1996. 17 Goldie RG, Pedersen KE. Mechanisms of increased airway microvascular permeability: role in airway inflammation and obstruction. Clin Exp Pharmacol Physiol 1995;22:387-96. 18 Barnes PJ, Chung KF, Page CP. Inflammatory mediators and asthma. Pharmacol Rev 1988;40:49-84. 19 Lundberg JM, Saria A. Capsaicin-sensitive vagal neurons involved in control of vascular permeability in rat trachea. Acta Physiol Scand 1982;115:521-23. 20 Rogers DF, Belvisi MG, Aursudkij B, Evans TW, Barnes PJ. Effects and interactions of sensory neuropeptides on airway microvascular leakage in guinea-pigs. Br J Pharmacol 1988;95:1109-16. 21 Pedersen KE, Rigby PJ, Self GJ, Goldie RG. Quantitative assessment of increased airway microvascular permeability to [125I]-labelled plasma fibrinogen induced by platelet activating factor and bradykinin. Br J Pharmacol 1991;104:128-32. 22 Erjefalt I, Grieff L, Alkner U, Persson CA. Allergen- induced biphasic exudation responses in guinea-pig large airways. Eur J Pharmacol 1993;48:695-701. 23 Rigby PJ, Paterson JW and Goldie RG. Evaluation of the dynamics of airway microvascular leakage using in vivo confocal microscopy. Am J Resp Crit Care Med 1998;157:A533.

Received: April 19, 1999 Accepted: May 13, 1999

Correspondence to: Paul J. Rigby Department of Pharmacology University of Western Australia Nedlands, WA 6907 Australia [email protected]