Overview of Live‐Cell Imaging: Requirements and Methods Used

Live-cell imaging is an important analytical tool in laboratories studying biomedical research disciplines, such as cell biology, neurobiology, pharmacology, and developmental biology. Imaging of fixed cells and tissues (for which photobleaching is the major issue) usually requires a high illumination intensity and long exposure time; however, these must be avoided when imaging living cells. Live-cell microscopy usually involves a compromise between obtaining image quality and maintaining healthy cells. Therefore, to avoid a high illumination intensity and long exposure time, spatial and temporal resolutions are often limited in an experiment. Imaging live cells involves a wide range of contrast-enhanced imaging methods for optical microscopy. Most investigations use one of the many types of fluorescence microscopy, and this is often combined with transmitted light techniques, which will be discussed below. Continual advances in imaging techniques and design of fluorescent probes improve the power of this approach, ensuring that live-cell imaging will continue to be an important tool in biology. Successful live-cell imaging experiments can be a major technical challenge. An important caution is to ensure that cells are in good condition and function normally while on the microscope stage with illumination in the presence of synthetic fluorophores or fluorescent proteins. The conditions under which cells are maintained on the microscope stage, although widely variable, often dictate the success or failure of an experiment. Various cell culture media are available based on the particular biochemical requirements of cells. Culture media contain various constituents, including amino acids, vitamins, inorganic salts (minerals), trace elements, nucleic acid constituents (bases and nucleosides), sugars, tricarboxylic acid cycle intermediates, lipids, and co-enzymes. In tissue culture media, an important step is to control oxygen concentration, pH, buffering capacity, osmolarity, viscosity, and surface tension. Commercially available media formulations often include an indicator dye (e.g., phenol red) to visually determine the approximate pH value. A carbon dioxide and bicarbonate buffer system for regulating pH is needed for almost all cell lines. The cells need to be cultured in an atmosphere that contains a small amount of carbon dioxide (usually 5–7%) in incubators to control the dissolved gas concentration. For live-cell imaging, an appropriate atmosphere with carbon dioxide can be difficult to provide, and this usually requires specifically designed culture chambers for a regulated atmosphere. Oxygen requirements can vary among cell lines, but normal atmospheric oxygen tension levels are suitable for most cultures. With regard to osmolarity, most of the cell lines have a large tolerance for osmotic pressure, with good growth at osmolarities between 260 and 320 milliosmolar. When cells are grown in open-plate cultures or Petri dishes, hypotonic medium can be used to cope with evaporation. A constant cellular environment is important to maintain. Cells should be grown in culture medium in a carbon dioxide incubator. Mitotic index, viability, and morphological variability should be determined. Controlling temperature is a critical parameter for most cells in culture. Temperature control is often achieved by peripheral sources of heated air or infrared radiation. Temperature can also be controlled by a heating plate made of metal controlled by a thermistor coupled directly to the chamber, or by transparent thin coatings of electrically conductive metal oxides on the surface of the coverslip for providing efficient conductive heat transfer to the chamber. Temperature fluctuations can occur because the microscope stage, frame, and objectives can act as heat sinks and offset the heating system used. This problem can be solved using commercially available objective lens heaters. For permanent equipment, a box can be constructed around the microscope and heated with warm air. Most laboratories that image live cells with light or fluorescence microscopy have their own custom-designed viewing chamber (Khodjakov and Rieder, 2006). These chambers need to keep the specimens viable, while also providing optical properties that are optimal for imaging. The type of chamber that is used depends on various factors, such as the spatial and temporal resolutions necessary for the study, and the study duration. Short-term imaging experiments can be performed by sandwiching of two coverslips, which are separated from each other by a spacer between two plastic or metal plates. Long-term experiments can also be performed for imaging live cells on the microscope stage. The culture cells need to have optimum growth conditions for an extended period of time. In general, imaging chambers include a glass window, usually the thickness of a coverslip (approximately 170 μm), through which the cells are viewed with an objective lens that has a high numerical aperture. Two types of imaging chambers are generally used: (1) open chambers that are exposed to the atmosphere and (2) closed chambers, which are sealed to avoid evaporation of the culture medium. Both types of chambers are available commercially. For some high-resolution studies, the culture medium needs to be replaced frequently, which has led to various chambers being designed where there is a continuous flow of fluid across the surface of cells. Cells are prone to photodamage, especially when fluorophores are present (Stephens and Allan, 2003). When fluorescent molecules are in an excited state, they react with oxygen to produce free radicals that can damage subcellular components and adversely affect the cells. Even when fluorophores are absent, mammalian cells are sensitive to ultraviolet light. Enhanced power of new imaging techniques can be problematic. For obtaining the maximum signal/noise ratio and resolution, cells need to be illuminated with very high light intensity (Khodjakov and Rieder, 2006). Many methods are used to reduce light-induced damage. An important protective step is to shut off the illuminating light when not required (Stephens and Allan, 2003). Shuttering of the light used in every type of illumination is the most important factor in imaging of living cells. Unwanted wavelengths of light also should be removed. Optimized emission filters should be selected for a maximal signal. Reducing oxygen levels can help limit photobleaching. Additionally, omitting phenol red (used as a pH indicator in media) and serum from the medium can help reduce background fluorescence. Contamination of the illuminating light with even trace amounts of ultraviolet or infrared wavelengths should be avoided. One of the most important aspects of live-cell imaging experiments is to establish criteria for determining the condition of the cells. These criteria will vary depending upon the type of experiment; however, one of the most important factors to determine is whether there is damage to the cells caused by the imaging process that might affect the results. One of the simplest ways for following an imaging experiment is to compare the morphology and condition of cells that were exposed to light during the study with neighboring cells in the same chamber that were not illuminated. Once the cells are mounted on the microscope stage, they need to be viewed to establish their morphology and condition. In the majority of experiments, particularly when cells have been transiently transfected with fluorescent proteins, a large amount of morphological variability of the cells can be present. Sometimes cells are not transfected or show poor localization of the probe. Some of the transfected cells might overexpress the fluorescent protein, which can lead to toxicity. Cells can show abnormal morphology, such as swollen mitochondria, blebbing, and detachment from the substrate. Cells that do not show normal morphology should not be used for imaging experiments. When choosing an optical microscopy system for live-cell imaging, the following three variables should be considered: sensitivity of the detector (signal-to-noise), specimen viability (discussed above), and the speed required for image acquisition. The system used should make maximum use of the light and use the fewest optical elements in the light path. To optimize signal-to-noise, the combination of filters selected for imaging of live-cells should closely match the spectral profiles of the fluorophores used for experiments. Advances in detector technology have enabled reduction in illumination levels. Photomultiplier tube cathodes for confocal microscopy have become increasingly sensitive. Sensitivity of the camera is also an important consideration. An intensified camera can be used for this purpose or a sensitive back-illuminated charged couple device (CCD) camera. With regard to speed of acquisition, particularly with simultaneous imaging of multiple fluorophores or radiometric analysis of a single probe, switching between filters or output from a monochromator will reduce the acquisition time. Currently, there is no all-purpose live-cell imaging system that is suitable for all possible investigations. Researchers must therefore compromise by determining the optimal parameters while minimizing cell damage or death. Researchers also need to understand the pros and cons of different microscopes. Most cellular processes occur in three dimensions over time. Therefore, cells need to be imaged in four dimensions to obtain a complete picture. Most epi-illumination microscopes and confocal systems acquire data series in four dimensions. Time-lapse imaging (the fourth dimension) is broadly applied to capture events that occur in live-cells over the period of a few seconds to several weeks or months. This technique enables repeated imaging of a cell culture at particular time points (Fig. 1). However, dynamic imaging data are complex and interpretation can be difficult. Therefore, specialized image processing methods for estimation of motion, object detection, visualization, and quantitation can be used (Gerlic et al., 2003). Time lapse images of biofilm formation by gfp-tagged Pseudomonas using confocal microscopy. The scale bar indicates 20 μm. The images were kindly provided by Thomas Moninger, Central Microscopy Research Facility of the University of Iowa and Pradeep Singh, University of Washington. Live-cell imaging is currently mainly performed by fluorescence microscopy. Some of the more common systems used for imaging live cells are discussed below; the more widely used systems are discussed first and then the less commonly used systems are described. A basic overview of these types of fluorescent microscopy has been published in a previous issue of The Anatomical Record. Many types of fluorescent proteins of different colors are available that are used as fluorescent tags for imaging live-cells. Currently, the most common fluorescent protein used in imaging of cells is green fluorescent protein (GFP; Shimomura et al., 1962; Crivat and Tarasaka, 2012). GFP-like proteins with emission colors ranging from cyan to red have been found in natural sources, including corals, sea anemones, hydrozoans, crustaceans, and even basic chordate animals (Wiedenmann et al., 2009). Table 1 shows some examples of the fluorescent proteins used by researchers, and use and advantages of these proteins. Widefield (conventional) systems provide a flexible system for live-cell imaging (Figs. 2 and 3). Fast acquisition can be achieved with flexible excitation and a low cost (Stephens and Allan, 2003). Many experiments using live-cells are best performed by widefield systems and subsequent deconvolution (method for reducing the amount of blurring in fluorescent imaging by reassigning each photon back its point of origin) of the data. However, deconvolution should be carefully and accurately applied to avoid generation of artifacts (Wallace et al., 2001). The process of phagocytosis of a ceramic fiber (RCF) by an LLC-MK2 monkey kidney epithelial cell in culture. The plasma membrane undergoes ruffling activity around the fiber as it becomes progressively internalized until it is completely contained within the cytoplasm of the cell. Labeling with DiI demonstrates that during phagocytosis the fiber becomes encased within a portion of the plasma membrane. The speed of incorporation is approximately 0.2 μm/min. Time (in hr:min:sec) is shown in the lower right corner. As an indication of magnification of the time-lapse, the diameter of the cell nucleus is approximately 20 μm. The images were kindly provided by Dr. Cynthia Jensen, the University of Auckland. Part of a time sequence of neonatal rat ventricular myocytes plated onto a dish and paced at 1 Hz. The white areas are free calcium ions. The scale is 0.19 μm per pixel. Immunofluorescence microscopy was used. The image was kindly provided by Gregory Bass, the University of Auckland. Imaging live-cells using transmitted light is often used with fluorescence microscopy to determine cell shape, motility, and position. This is important for investigating cells that undergo changes in shape, such as mitosis and apoptosis. Differential interference contrast and phase contrast are the most common types of microscopy used. Time-lapse imaging can be used in two spatial dimensions using widefield techniques. Time-lapse imaging of living cells is enhanced by the improved resolution in confocal microscopy. However, image collection during time-lapse is limited by the speed of the scanning mirrors. Imaging live tissues using confocal microscopy is more difficult than imaging fixed specimens, and is not always practical because the specimen may not tolerate the particular conditions. Choice of a cell type that is amenable to the conditions of imaging in confocal microscopy is important to avoid experimental problems. Live-cell confocal analysis is a good option because of its increased efficiency of photons and less phototoxic dyes for labeling. Images should be captured as rapidly as possible, using the minimal amount of laser power that enables imaging. Images taken with laser-scanning confocal microscopy are shown in Fig. 4. Laser-scanning confocal images of the Wistar rat lens. Two images are shown from the same experiment. (A) A cross-section of Wistar rat lens fiber cells is shown. Scale bar = 10 μm. (B) The summary of fluorescent intensity of 42 slides over a period of 84 sec is shown. The center cell of highest intensity is where the two-photon laser uncaged the fluorescein. Using another 488 nm laser, the fluorescein was excited and imaged. The image was kindly provided by Dr. Nancy Liu, the University of Auckland. Confocal microscopy has high irradiance, which may cause photobleaching of specimens (Conchello and Lichtman, 2005). Photobleaching from the laser beam is cumulative over multiple scans and, therefore, exposure to the beam should be reduced as much as possible for image acquisition. Another limitation of confocal microscopy is its inefficient use of light. In-focus light in a confocal microscope needs to pass through a small aperture, and this means that its transmission efficiency is lower than that in widefield systems. This loss can be reduced, but it will never be eliminated. Therefore, high intensity of excitation light is required for confocal microscopy, which increases the possibility of photobleaching to the specimen (Khodjakov and Rieder, 2006). Confocal microscopy provides control of the area of illumination and, therefore, it is useful for photobleaching techniques, such as fluorescence loss in photobleaching and fluorescence recovery after photobleaching (FRAP; see below). Spinning-disc confocal microscopy has some advantages for live-cell imaging compared with laser-scanning confocal microscopy (Stehbens et al., 2012). The frame acquisition rate is high for spinning-disc confocal microscopy because the specimen is scanned by thousands of points of light in parallel. Fast frame rates are advantageous because CCD cameras can be used for direct acquisition of confocal images rather than photomultiplier tubes, which are still the most commonly used light detectors in laser scanning confocal microscopy. Additionally, spinning-disc confocal microscopy is less affected by saturation of fluorescence or ground state depletion compared with laser scanning confocal microscopy. For live-cell imaging, a slight loss of confocality is associated with spinning disc confocal microscopy, but this is compensated for by a better signal-to-noise ratio and a reduced amount of photobleaching. Multiphoton microscopy has many advantages for imaging living cells. The use of longer-wave excitation greatly reduces photobleaching in contrast to single-photon confocal imaging, and this results in extended viability of living cells during long-term imaging. This type of microscopy allows three-dimensional (3D) fluorescence imaging of live cells that are hundreds of microns deep within thick, strongly scattering samples (Rubart, 2004). Examples of multiphoton fluorescence microscopy include imaging of intracellular calcium dynamics in various tissues, probing functional integration of donor cells after intracardiac transplantation, cellular redox state, and developmental studies, which require viability to be maintained for detecting sequential events in the same specimen over long time periods. The near-infrared laser based approach of multiphoton microscopy is very efficient; therefore, it has many potential uses for noninvasive, noncontact loading of fluorescent probes in various cell types, as well as in studies on cell to cell communication (Konig, 2000). In cell biology, TIRF microscopy is primarily used to observe cell–substrate interactions (Sako and Uyemura, 2002). TIRF microscopy is used for direct imaging of living cells that are within close proximity to the coverslip. TIRF can be used to follow biochemical reactions, structural changes, and movements of biomolecules at the level of single molecules in real time. This technique is useful for observing single living cells in aqueous solution near a glass surface, as well as single molecules on the plasma membrane of live-cells. Time-lapse imaging can be used to obtain information on the steady-state distribution of a protein over time but time-lapse imaging cannot determine the kinetic properties of a molecule. To achieve this, a selected pool of fluorescent proteins needs to be distinguished from other fluorescent proteins in cells. This can be accomplished by FRAP, which is a technique that is used to study the motion of cellular processes (Lippincott-Schwartz and Patterson, 2003). This method can be used to determine the mobility and dynamics of proteins that are fluorescently labeled in or associated with cells. In FRAP, fluorescently labeled molecules are intentionally photobleached in the area of interest of the cell, using high-intensity light. The bleached molecules diffuse out of this region and nonphotobleached fluorescently labeled molecules simultaneously move into the bleached region, which results in the fluorescence signal being recovered. The rate and the amount of recovery at the photobleached region is analyzed, revealing information on the dynamic properties of proteins or cytoskeletal structures (e.g., microtubules; Axelrod et al., 1976). FRAP is a time-course experiment in which data must be acquired quickly to successfully record recovery of the bleached area. However, this usually means that spatial resolution is traded for temporal resolution. Imaging chambers for FRAP experiments are constructed so that the cells remain healthy while optical resolution is maximized. Confocal microscopes are one of the most common types of microscopes used for FRAP experiments (Houtsmuller, 2005). A widefield system with a laser for photobleaching may be a solution to improve temporal resolution. Fluorescence correlation spectroscopy (FCS) is based on a confocal optical setup and is similar to a laser-scanning microscope. FCS measures dynamics in the time range of 1 μs to 100 ms (Bacia and Schwille, 2003). Autocorrelation measurements evaluate a wide variety of dynamic behavior in vivo. Because FCS has the advantages of high temporal resolution and short analysis time, FCS is useful for studying conformation dynamics, chemical kinetics, and molecular diffusion in solution as well as on membranes (Tian et al., 2011). The FCS focal volume is positioned in a cell's plasma membrane, in the cytoplasm, the nucleus, or outside of the cell. When analyzing dynamic processes on a very fast scale, special precautions are necessary to avoid artifacts. Because FCS temporal resolution is under 100 ns, limitations of FCS for in vivo applications are more likely to be in the slow time range. When slowly moving molecules are analyzed (e.g., nuclear DNA, molecules interacting with the cytoskeleton, or membranes), acquisition times need to be long enough to capture the process. In addition, molecules under examination should not become photobleached during their transit time through the focal volume. The resulting photobleaching decay in the fluorescence trace causes decay in the correlation curve, and this often prevents accurate curve using of fluorescence transfer increases the spatial resolution of the fluorescence microscope to less than 10 This in resolution as a technique for studying events in particularly because can be performed in living cells. can be used to interactions live-cells et al., 2012). However, in live-cell studies, a possibility is that measurements will be by recovery of the fluorophores when using the method of photobleaching to FRAP Therefore, use of photobleaching in live-cell experiments may be Fluorescence imaging is applied for measurements of events single living cells. acquisition is enough of to to measurements in live cells and The fluorescence an The steady-state fluorescence image also enables the resolution of parameters of the state of proteins cellular is an method for of on living cells because of of and of fluorescence on concentration. are inorganic fluorescent that many limitations of the more commonly used fluorophores in live-cell imaging et al., 2004). can be imaged with type of fluorescence microscope. are useful for live-cell imaging because of their enhanced and their to and et al., et al., 2011). are useful in studies of long-term and imaging of cellular and molecular have been used for imaging of cells and cell membrane including on the cell membrane. can cells and at the subcellular and are used to to the cytoplasm when there are a large of cells. is used for single cell have advantageous optical inorganic are to which can limit their use in live-cell imaging. is an area of However, many experiments have that have limited of surface coatings has et al., 2009). Live-cell imaging using fluorescent proteins and fluorescent probes is a tool for cell biology, but a need for techniques to study tissue and cell dynamics chemical or scattering microscopy is used by cell to live-cell dynamics with chemical in a noninvasive, et al., 2012). is a multiphoton process that provides The image contrast is achieved by light that is by of chemical An of the process is that it in the focal volume where high of photons are spatial resolution. A detector is not Most of the research microscopy for cell has dynamics by et al., because has a signal from In fluorescence microscopy, obtaining is difficult because the process function and Therefore, in live a is to achieve time lapse and long time experiments on and and accurate is difficult. microscopy has to be an method for live-cell dynamics with chemical microscopy is a of scanning microscopy that was for the microscopy light microscopy because microscopy provides information of the cell not by light microscopy. Various imaging can be used in microscopy, but a few (e.g., signal or or and and are used in live-cell imaging 2003). type of imaging information on living cells. For imaging provides structural and The signal is for imaging very such as by cells moving on a substrate. microscopy enables a flexible analytical approach that can provide intracellular in live which has advantages over techniques, such as scanning microscopy et al., A limitation of microscopy is among cultured cells with regard to their to to ensure microscopy of all cell imaging of live cells using microscopy still has many to no other available method provides dynamic imaging and spectroscopy at high resolution. The Thomas Moninger, Pradeep Singh, Gregory Bass, Cynthia and Nancy for providing and for with processing University of on confocal microscopy is greatly