Culture Chambers for Live-Cell Imaging

Specimen chambers are an integral and critical branch in the history of live-cell imaging, and a wide spectrum of designs have been published over the years describing systems that offer excellent optical properties while allowing specimens to be maintained for varying amounts of time. Ranging in complexity from the simple preparation of a sealed coverslip on a microscope slide to sophisticated perfusion chambers that enable tight control of virtually all environmental variables, culture chambers are designed to to allow living specimens to be observed with minimal invasion at high resolution.

Regardless of their design, live-cell imaging chambers must fulfill a variety of requirements in order to be successfully employed in experiments. The chamber should be easily sterilized and totally isolated from the laboratory environment with a cover or seal during observation to minimize exposure to sources of contamination. On the other hand, the culture chamber should also offer uncomplicated access to the cells if the investigation involves microinjection, addition of reagents (such as drugs or metabolites), physical manipulation of the cells, or changes to the culture medium. If the live-cell experiment is followed by cloning or fixing and staining the culture with synthetic fluorophores or immunofluorescence after the imaging session, the chamber must be configured to allow removal of the coverslip.

The culture chamber dimensions, including overall size, surface area for cell culture, and the volume of medium (or buffer) bathing the cells, are important considerations. In addition to being readily accommodated by the microscope stage, the culture chamber must be large enough in its lateral dimensions to contain sufficient cells for a thorough sampling of the population. The depth of the surrounding medium should be minimized to ensure the highest possibility optical quality for transmitted light and accessibility to the cells, yet it must also be deep enough to provide a healthy environment. Plastic coverslips should be avoided due to their inherent autofluorescence and strain birefringence, which interfere with some imaging modes. Only the highest-quality glass coverslips should be employed, and these should be cleaned with a strong acid or base and thoroughly washed. Commercial culture chambers vary in reliability, versatility, and cost. These factors should be considered when designing experiments in order to determine whether a single complex (and often expensive) chamber is required, or if a larger number of simple culture dishes will suffice. In the final analysis, it should be recognized that there is no perfect culture chamber for all purposes and many compromises are often necessary to achieve success in live-cell imaging.

Two commercial high-performance incubation and perfusion chambers for live-cell imaging are illustrated in Figure 1. The open chamber system in Figure 1(a) is designed to be used with 35-millimeter Petri dishes having a coverslip fused to the base, and features a Peltier heat pump for controlling the operating temperature over a range of -5 to 50 degrees Celsius. Multiple perfusion system options enable continuous perfusion at a constant temperature, discontinuous perfusion, static incubation of culture medium (no perfusion), or the introduction of rapid temperature changes through perfusion. The chamber fluid height is adjusted using a screw mechanism coupled to a suction port that eliminates fluid flutter during media exchange. The chamber is also equipped with a single channel patch clamp coupled to a Teflon-coated well having a silver chloride electrode to form a salt bridge in electrophysiology experiments that require low noise. Gas flow over the chamber top provides improved temperature uniformity and pH control. The closed chamber system presented in Figure 1(b) contains a unique dual-role perfusion system in which laminar flow is produced by the controlled streaming of media over a channeled glass slide (termed a microaqueduct) that is also coated with an electrically conductive and transparent indium-tin oxide thin film for temperature control. This system can operate at flow rates ranging from static through rapid media exchange with low cell surface shear over a temperatures spanning from ambient to 50 degrees Celsius. Both bicarbonate and organic buffers are compatible with the chamber, which is interfaced to a stand-alone controller unit.

Historical Perspective

The first live-cell imaging chambers were designed and built shortly after mammalian cell culture techniques were developed in the early twentieth century. Most of these early chambers utilized a wick to control the flow of medium through glass vessels or cork housings with glass windows. Eventually, a number of perfusion-style growth chambers were constructed with the intention of being used for long-term imaging using the light microscopes available during the period. By the 1950s and 1960s, more sophisticated live-cell imaging chambers were being designed for high-resolution investigations that incorporated the emerging techniques of phase and differential interference contrast (DIC) microscopy. After affordable techniques and couplers to mate cinema film cameras with the microscope were perfected in the late 1960s and early 1970s, researchers were able to systematically investigate living cells in time-lapse experiments at high spatial and temporal resolution using increasingly more complex imaging chambers.

The imaging chambers developed during 1970s and 1980s were based on a standard design involving the sandwiching of two coverslips, separated by a pair of rubber O-rings or similar spacers, into a holder made of two metal plates. This common motif, which suffers a number of pitfalls, including the inability to exchange culture medium or to add measured aliquots of drugs or metabolites, prevailed for many years. More advanced perfusion chambers were ultimately constructed by modifying this basic design to enable the continuous addition of fresh culture medium and the intermittent addition of other reagents. The major drawback with many of these early chamber designs, however, was the lack of attention to laminar flow, a process in which the formation of chemical gradients is minimized through a smooth transition as the medium is replaced. Laminar flow ensures that two solutions do not become mixed by turbulence during the replacement process so as to avoid the generation of chemical gradients. Continued development of both simple and complex imaging chambers over the past few years has resulted in excellent designs that suit the requirements of most imaging scenarios.

Several live-cell imaging chambers that were originally designed in the late 20th century are illustrated in Figure 2. Variations of these chambers are commercially available and derivative configurations have been constructed in a number of laboratories. The chamber depicted in Figure 2(a) is a second-generation version from the original design of Robert Day Allen and Andrew Bajer, and is intended for use in perfusion experiments with an upright microscope. A large diameter exhaust port reduces the chances of cracking the coverslip, and the chamber can be easily modified for adaptation to an inverted tissue culture instrument. Although relatively inexpensive to build, this chamber does not produce ideal laminar flow characteristics. The Dvorak-Stotler chamber style presented in Figure 2(b), which features circular coverslips and gaskets, has periodically been made commercially available by several manufacturers, but is also relatively easy to construct. At a perfusion rate of 1 milliliter per hour, the Dvorak-Stotler chamber produces a clean, sweeping pattern of flow across the coverslip surface. Figure 2(c) represents a perspective drawing of a sophisticated perfusion chamber designed in the early 1950s that was originally milled from polished Plexiglas. Enabling the ability of view specimens at high resolution under continuous perfusion, the downside of this design is a lack of satisfactory laminar flow characteristics. The chambers presented in Figure 2 represent only a few of the many designs that have appeared in the literature over the past 50 years.

Today, a wide spectrum of live-cell imaging chambers are commercially available, ranging from simple designs mounted on microscope slides to complex systems capable of controlling virtually all aspects of the environment. Although many investigators (especially those with access to a machine shop) continue to fabricate specialized chambers that meet the particular requirements of their experiments, a wide range of commercial chambers now feature advanced laminar flow, polymer construction, and high resolution capabilities in disposable units that can be purchased at reasonable cost. As emphasized, the most important aspect of any culture chamber design is the ability to guarantee and verify that cell cultures are healthy and undergoing growth on a normal timescale, as well as providing an adequate optical window for the microscope to ensure that imaging of the culture can be conducted at high enough numerical aperture to meet the resolution demands of the experiment.

Simple Microscope Slide and Petri Dish Imaging Chambers

Short term imaging experiments (20 to 30 minutes or less) can be conducted simply by attaching a coverslip containing adherent cells onto a microscope slide using spacers to keep the cells from being damaged (physical stress can induce autofluorescence in some cell lines). The coverslip can be secured with any one of a number of sealants, including molten agarose, rubber cement, vacuum grease, or a useful preparation known as VALAP (a 1:1:1 mixture of Vaseline, lanolin, and paraffin), to provide a watertight seal and eliminate evaporation of the culture medium (as illustrated in Figure 3(a)). Thin gaskets cut from silicone rubber (also commercially available) or broken pieces of coverslip can be used as spacers to keep the cells from coming in direct contact with the microscope slide (Figure 3(b)). When assembling these chambers, make certain that the coverslip surface containing the cells is placed face down on the spacer, and fill the void between the coverslip and slide with a physiological buffer (such as phosphate buffered saline; PBS) or nutrient tissue culture medium. Seal the edges around the coverslip using the reagent of choice and place the microscope slide on the stage for observation. In the absence of growth medium and temperature control, the cells will function normally for only a few minutes, but this is often enough time to obtain the necessary images. For short term storage without seriously compromising cell viability, the slide can be placed in a carbon dioxide incubator or on a small heating block next to the microscope between image gathering sessions. Alternatively, to avoid loss of buffer or culture medium during storage between experiments, the slide can be placed in a small heated humidity chamber.

Synthetic O-rings (Figure 3(c)) or similarly shaped circular gaskets (such as the reinforcement tabs used for three-ring binders) offer a convenient alternative to flat silicone gaskets (Figure 3(b)), and can be easily sealed with any of the mounting mixtures discussed above. In many cases, the circular gaskets are thick enough to increase the amount of media or buffer available to bathe the cells, and there is less risk of cell damage when mounting. Coverslips containing adherent cells, thin tissue sections, embryos, or small intact organisms can often be more readily imaged by substituting a thicker single-well concave slide for the standard slide (Figure 3(d)). These slides are approximately 3 millimeters thick (compared to a 1-millimeter traditional glass slide), enabling them to accommodate a relatively deep (2-millimeter) cavity. The well holds a greater volume of buffer or culture medium than can be sandwiched between a flat slide and coverslip and, similar to O-rings, the depression will be less likely to damage cells during the mounting procedure. On the downside, the thick design and curved walls are not compatible with high-resolution imaging in transmission contrast-enhancing (phase contrast, DIC, and Hoffman modulation contrast; HMC) modes. However, fluorescence imaging, which relies on episcopic illumination of the coverslip, is not affected by the well.

The simple microscope slide configurations described in the previous paragraphs are primarily intended for use with upright microscopes so that cells can be imaged through the coverslip rather than the thicker glass slide. In some cases, especially when using a solid adhesive, the mounted slide can be turned over and observed on an inverted microscope without deleterious effects. Alternatively, if a high-resolution dry objective containing a correction collar is available, the specimen can usually be imaged through the bottom of the slide if the objective aberration correction factor is able to compensate for the extra (1-millimeter) thickness. A more versatile configuration, which is suitable for both upright and inverted microscopes, can be produced by drilling a 1-centimeter hole in a glass or plastic microscope slide and mounting a coverslip to one side using nail polish, silicone sealant, or one of the newer advanced adhesives that are available at hardware distributors (as presented in Figure 3(e)). The chamber "well" can be filled with culture media or buffer and fitted on the upper side with a second coverslip containing adherent cells that is transiently mounted with VALAP. If the coverslip is carefully placed over a full well to avoid trapped air bubbles, an optically clear path is established between the two coverslips. Otherwise, a meniscus will be created by the bubbles, which can distort images captured in transmission illumination mode. This configuration is optimized for use with upright microscopes. An alternative loading technique, more appropriate for inverted microscopes, involves growing the cells on the permanently attached coverslip followed by covering the cells and growth medium with a second coverslip as described.

A wide variety of commercial microscope slide culture chamber designs are available to fulfill the requirements of many short-term live-cell investigations. These range in complexity from simple plastic culture bottles and multi-well enclosures fused to glass microscope slides to complex integrated systems that allow perfusion, manipulation, flow studies, and rapid assays of multiple isolated cultures, as illustrated in Figures 3(f) through 3(i). The so-called "lab-on-a-slide" system is depicted in Figure 3(f) with a multi-chamber design that is available in a number of formats, ranging from a single large chamber to several (up to 18) smaller chambers on a single glass slide (the chamber in Figure 3(f) has 4 rectangular wells). These versatile, self-contained chambers are fitted with a loose plastic cover that allows gas exchange, but can become a problem due to the meniscus and birefringent plastic when imaging with transmission modes. In addition, the gas exchange enabled by the chamber design limits the amount of time the cells can be outside of a carbon dioxide incubator without undergoing a rise in pH, unless the slide is imaged within an environmental control chamber. Similar chambers can be obtained that are fused either to 1-millimeter glass microscope slides or to 170-micrometer coverslips, both of which are intended for visualization of the growing cells with an inverted microscope. At the end of the live-cell observation session, the cells can be fixed and stained directly in the growth chamber, which is then removed for mounting of the slide or coverslip.

The commercial imaging chamber illustrated in Figure 3(g) combines the convenience of cell culture in a Petri dish with the optical quality of a glass microscope slide for high resolution imaging using a variety of contrast modes. The observation channel connecting the two media reservoirs has a volume of 100 microliters and is 5 millimeters wide to enable observation of a large cell population. Flow through the channel can be initiated by filling one of the reservoirs. The bifurcated channel featured by the chamber in Figure 3(h) can be purchased either uncoated or treated with specialized coatings to grow endothelial cells for the simulation of blood vessels. Luer-style adapters at the channel termini allow adaptation to perfusion systems for precise regulation of flow rates and shear stress. This type of culture chamber is an excellent example of customized designs that target specific research arenas. A similar chamber, shown in Figure 3(i), is designed to accommodate very small volumes (10 microliters) in the performance of stem cell differentiation assays, molecular recognition and binding studies, as well as in the determination of binding constants based on fluorescence imaging. Five parallel channels are each equipped with a separate media reservoir to increase throughput. A wide variety of specialized chamber designs are commercially available, many targeted for economy of materials and the efficiency of execution in specific investigations.

A variety of Petri dish configurations specifically designed for high-resolution live-cell imaging are commercially available (see Figure 4). The simplest and perhaps most versatile design consists of a standard 35 or 50-millimeter disposable Petri dish featuring a circular opening (10, 14, or 20 millimeters) cut into the center of the attachment surface and covered by a high-quality borosilicate coverslip to enable observation of living cells at high resolution (Figure 4(a)). The chambers are available uncoated or coated with a thin layer of adhesion-promoting agent such as poly-D-lysine or collagen. Known as glass bottom dishes, these imaging chambers are available in 6, 12, and 24-well versions (Figure 4(c)) for multi-culture assays. A similar design (Figure 4(b)) features a customized Petri dish that contains a large circular coverslip spanning the entire base of the modified dish. These dishes are also available in a variety of sizes and can be purchased with tightly fitted covers to secure the atmosphere, unlike conventional Petri dishes that have loose fitting lids for gas exchange. Although standard polystyrene Petri dishes can be employed for routine low-resolution observations (directly through the plastic) with long working distance objectives having correction collars, the use of high numerical aperture (1.2 to 1.4) objectives coupled to oil or water immersion requires the modified glass bottom versions.

Petri dish culture chambers, in most cases, are filled with a substantially larger volume of growth medium or imaging buffer than can be accommodated on the microscope slides illustrated in Figure 3. Increased volumes of media minimize the danger of sudden changes in temperature or pH and provide a superior growth environment for long-term culture. Another benefit of these chambers is that the coverslip fused to the bottom of the Petri dish enables high-resolution imaging of cultured cells directly after transfer from the incubator rather than potentially inducing trauma by loading seeded coverslips or detached cells into a specialized imaging chamber. In practice, cultures containing transfected mammalian cells (expressing fluorescent proteins) can be repeatedly transferred back and forth between a humidified carbon dioxide incubator and the microscope stage for periodic observation or recording of images with little or no damage to the cells. Additionally, a growing number of commercial aftermarket heating stages and enclosed microscope incubators are designed to maintain the 35-millimeter glass bottom Petri dishes for long term time-lapse imaging sequences.

The simple live-cell imaging chambers illustrated in Figures 3 and 4 (in effect, microscope slides and Petri dishes) can often be washed, sterilized, and reused on multiple occasions. The first step is to thoroughly clean the chamber with detergent, followed by exhaustive rinsing with deionized or distilled water and drying in a dust-free container at room temperature for several days. In more difficult cases, especially for those chambers and slides that have been exposed to drugs, dried media, or cellular adhesion reagents, the units can be pre-cleaned by soaking with ethanol or dilute acid solutions, but this treatment should be avoided with chambers that use glue or another solvent-based adhesive to secure a fragile coverslip to glass, metal, or plastic. After cleaning and drying, the chambers are sterilized either by autoclaving (not recommended) or treatment for 15 minutes with an ultraviolet lamp in a dark cabinet. As an economical and quick alternative, many of the simple incubation chambers can be sterilized by washing with ethanol and drying in a sterile laminar flow hood.

Although cell cultures mounted on glass slides or grown in glass-bottom Petri dishes can be observed transiently on the microscope stage at room temperature for a few minutes, a variety of commercial devices have been manufactured to enable these simple imaging chambers to be employed for longer periods of time. Termed stage warmers or stage incubators, all of these units provide localized regulation of temperature (see Figure 5), while the more advanced designs also enable limited control of the atmosphere and contain capabilities for perfusion. The warming plates for standard dimension microscope slides (1 x 3 inches; see Figures 5(a) and 5(b)) are suitable for maintaining temperature control of the chambers illustrated in Figure 3, and are readily adapted to most microscopes, usually with the original stage clips. The slide warmer illustrated in Figure 5(a) features a metal top plate that fully encompasses the microscope slide within a heated compartment. Unfortunately, the specimen area on the coverslip under observation resides in the open aperture, which is not heated by the plate. The problem of heat transfer from the glass slide to the specimen is compounded by employing large apertures, although smaller viewing areas often create additional problems by obstructing the movement of physically large high numerical aperture objectives. The heated slide holder depicted in Figure 5(b) is designed to adapt to a standard inverted microscope and features a smaller aperture, providing more even temperature control of the specimen at the expense of available observation area.

Among the most advanced microscope slide warmers is the multi-well chamber slide incubator shown in Figure 5(c), which is designed to provide a highly controlled environment (for slides similar to the one illustrated in Figure 5(f)) that regulates pH, temperature, perfusion, and the localized atmosphere. The incubator takes advantage of Peltier technology to heat or cool the specimen through a temperature range of 5 to 50 degrees Celsius, and can accommodate up to four temperature-regulated perfusion lines as well as gas superfusion. Among the many unique features of this incubator system are a magnetic base for ease in microscope configuration and a specially designed aspirator to remove perfusate with minimal disturbance of fluid level. The slide warmers and incubator depicted in Figures 5(a) through 5(c) are each interfaced to a separate control unit that maintains temperature and the other necessary variables (such as perfusion and atmospheric regulation) for each device.

Petri dish incubators (see Figures 5(d) through 5(f)) are similar in design and execution to simple microscope slide stage warmers. The most elementary version, illustrated in Figure 5(d), consists of a heated platform that is readily adapted to any inverted microscope stage and can accommodate Petri dishes up to 10 centimeters in diameter. Adherent or suspended cells are observed through a central aperture in the warming plate. Even though the cells residing directly above the aperture are not warmed by the plate and may experience viability problems during longer observations, the dish can easily be rotated to move adjacent areas of the culture into the aperture for viewing. Alternatively, smaller Petri dishes (35 or 60 millimeters in diameter) can be more evenly heated and observed using this warmer. A more advanced Petri dish warmer design featuring a diagonally positioned slit instead of a circular aperture is depicted in Figure 5(e). This incubator can be employed with Petri dishes ranging in size from 60 to 100 millimeters, and a smaller version is available for 35-millimeter dishes. The diagonal slit spans the central to the peripheral regions of the Petri dish and can be used to examine virtually the entire culture by slowly rotating the dish during observation. Cells adjacent to the slit are maintained at the correct temperature, although those positioned directly over the opening are subject to fluctuations due to thermal gradients.

The 6-well plate stage incubator illustrated in Figure 5(f) is designed to be seated onto an inverted microscope, and provides full control of the environment to enable long-term imaging of several cultures. Temperature is regulated to a tenth of a degree Celsius (which is continuously monitored by comparing to a reference well) through a recirculating water system connecting an external thermostatically controlled bath to the chamber. A separate control unit mixes carbon dioxide with air, which is then continuously fed to the incubator chamber, while a water reservoir within the chamber serves to provide constant relative humidity. The chamber is supported by numerous accessories to accommodate both 35-millimeter Petri dishes and multi-chamber slides (Figure 3(f)), as well as interchangeable adapters to mount 12 through 96 multi-well plates and single-coverslip chambers. This advanced system is useful for a variety of applications ranging from pharmaceutical assays to fluorescence observations in widefield and confocal microscopy.

Isolated, unenclosed Petri dishes and microscope slides can also be maintained at or near physiological temperatures using hot air devices that resemble and operate in a manner similar to a hair dryer (as illustrated in Figure 6) or egg incubator heater. Although temperature regulation is limited to several tenths of a degree Celsius, even with the application of a proportional controller and a remote thermo-sensor monitor, the air blower focuses a stream of air directly onto the culture chamber and surrounding microscope parts (stage, objective, and condenser) to provide rapid response to temperature drift. Furthermore, access to the specimen for administering drugs and media changes is unimpeded. On the downside, open cultures grown in Petri dishes and heated with hot air blowers are subject to gradual drifts in pH and osmolarity as both carbon dioxide and water are slowly released from the chamber. This fact limits the use of air blowers with open chambers to short-term observations that can be conducted in a few hours. In contrast, closed system chambers (discussed below) can be maintained for much longer periods of time using air blowers. Another disadvantage of air blowers is that the temperature of the specimen constantly fluctuates (even if over a narrow range of temperatures), which can result in focus drifts whenever the air stream is switched on or off. Overall stability of culture conditions can be increased and the level of drift reduced by coupling a forced hot air blower to an environmental cabinet that surrounds the microscope. Infrared lamps have also been employed to warm cultures, but they suffer from problems similar to those encountered with air blowers.

A majority of the heating devices illustrated in Figure 5 have restricted areas for observation of living cells (often only a few millimeters), requiring careful attention to ensure that sufficient clearance is available for the objective to translate comfortably within the confines of the aperture. Modern high numerical aperture oil and water immersion objectives are quite large, have very short working distances, and may be so bulky that they collide with warmers having excessively thick bases or mounted too far above the microscope stage. Furthermore, the ability to translate an oiled objective across the glass surface of a culture chamber may be severely limited by the small apertures present on many stage warmers. When choosing a stage warmer for live-cell imaging, the investigator should compare the aperture and objective dimensions to ensure that a sufficiently large area of the culture chamber can be observed.

Several of the microscope slide and Petri dish warmers or incubator chambers discussed above (Figure 5) are optionally equipped with insulating pads to isolate the heated platform from the stage, a measure that retards heat loss through the thermally conducting metallic portions of the microscope frame. As an alternative, these heating plates can be mounted on a thin (1 to 2 millimeters) layer of insulating sheet made from plastic or composite to minimize heat loss. Note that the microscope stage and frame are excellent conductors of heat and can rapidly overcome the limited heating capacity of typical controllers used with heating plates. The usual result is a gradual drop in temperature (especially near the aperture) that will affect cell viability and render experimental results suspect. Another important point that must be considered relates to the actual performance of these devices. Although the Petri dish and slide warmers usually feature excellent temperature regulation (one-tenth of a degree Celsius) due to the accuracy of their proportional controllers, the actual temperature of the specimen region under observation often deviates from temperature setpoint due to the gradient created across the aperture. For this reason, coupling simple stage warmers to additional heating devices, such as objective heaters or enclosed environmental chambers, may be necessary to ensure cell viability during extended imaging sessions.

Virtually all of the simple imaging chambers discussed in this section are not in equilibrium with the environment and should be limited only to very short term investigations, usually ranging from a few minutes to an hour at most. Although the specialized glass bottom Petri dishes are useful for a number of experiments, chambers fabricated from polymers (in effect, standard Petri dishes and polystyrene tissue culture flasks) suffer from nonuniform optical surfaces and exhibit strain defects that prohibit use of advanced imaging techniques, such as polarized light and differential interference contrast. Furthermore, thick chambers (both plastic and glass) prevent the use of high numerical aperture immersion objectives. Perhaps the most detrimental factor in applications employing simple chambers, however, aside from the inefficient transfer of heat that occurs between stage-mounted warming devices and the imaging chamber, is the gradual increase in pH and osmolarity that accompanies slow evaporation of the medium. Collectively, the many disadvantages of simple culture chambers preclude their efficient application for serious imaging experiments and researchers are advised to consider one of the more sophisticated alternatives discussed below.

Perfusion Chambers

The wide variety of culture chamber designs that are commercially available or can be readily constructed in-house generally fall into two basic functional categories. The first and most basic class, open chamber systems, are similar to the Petri dishes and custom microscope slides discussed above, which during use are slowly approaching equilibrium with the surrounding atmosphere. The more complex closed chambers are sealed to protect cells from evaporation of the culture medium and to ensure that environmental variables, such as pH, carbon dioxide concentration, and osmolarity, remain undisturbed. An open chamber cell culture system will allow easy access to the growing cells, thus readily permitting microinjection, patch clamping, addition of drugs and metabolites, changing of the culture medium, and other requisite manipulations to the cells. In contrast, closed chambers provide superior insulation from the external environment, but render access to the cells far more difficult. Most closed chamber designs include ports that permit the addition of fresh medium and drugs during the experiment without interrupting an imaging sequence. In these systems, perfusion is regulated by either a peristaltic pump, a motor-driven syringe, or through a gravity-controlled manifold. New solutions added to a closed chamber system should first be equilibrated to the same temperature and atmospheric conditions as the imaging chamber. Furthermore, many cell lines are sensitive to shear, so perfusion of adherent cells attached to a coverslip must be performed at very low flow rates. Several of the more advanced closed chamber systems described below are designed to provide tight control of shear forces.

The choice between using an open or closed culture chamber system is usually dictated by the characteristics of the investigation, including the necessary temporal and spatial resolutions, as well as the duration. High resolution brightfield imaging techniques that rely on DIC, HMC, or phase contrast require the numerical aperture of the condenser to be equal to or greater than that of the objective. Therefore, to achieve full resolution with a 60x objective of 1.4 numerical aperture, a matching condenser of equal numerical aperture must be employed. Because the maximum working distance of a high resolution condenser operating under Köhler illumination is only a couple of millimeters (depending upon the condenser working distance), this factor limits the physical dimensions of the observation chamber. Smaller closed chamber systems are capable of containing only limited volumes of culture medium (usually around several hundred microliters), which may become a problem with long term cultures if a perfusion system is not employed.

In contrast, fluorescence investigations that do not require accompanying high resolution transmitted light techniques can often be accomplished with open chambers, easing the task of changing medium or conducting other manipulations (such as electrophysiology measurements) to the culture. Closed chambers also require careful attention to the buffering capacity of the culture medium as pH cannot be controlled by the typical sodium bicarbonate/carbon dioxide system. As an alternative, media that are buffered without the gas can be utilized or synthetic buffer supplements (such as HEPES) should be considered. In any perfusion chamber, differences in the refractive indices between the perfusate, coverslip windows, and immersion medium can introduce significant error in depth measurements and result in the attenuation of signal. Many of these artifacts can be alleviated with water immersion objectives, but image quality is still superior in cases where the cells are attached to the coverslip.

Illustrated in Figure 7 are several commercial perfusion chambers designed for live-cell imaging experiments at varying optical resolution. The configuration shown in Figure 7(a) is a self-contained environmental chamber that houses a Petri dish along with capabilities for perfusion, temperature and humidity control, as well as regulation of atmospheric gas composition. Figure 7(b) depicts a diamond-shaped perfusion chamber (discussed below) in a heated stage adapter, while Figure 7(c) illustrates a similar design, but with a circular chamber. A variety of flow chamber stage inserts for the adapters in Figures 7(b) and 7(c) enable investigators to fine-tune the details of imaging experiments for both physiology and cell biology. Capable of being mounted in a rack, a Ludin-style perfusion chamber illustrated in Figure 7(d) is useful for imaging multiple samples with perfusion in a single experiment. Although the chamber itself is not heated, it can be housed on a heated stage plate or used in conjunction with an objective heater or fully enclosed environmental chamber. The Petri-style perfusion chamber in Figure 7(e) is molded to contain ports for introduction and exhaust of perfusion fluids, but contains no temperature control elements. Likewise, the chamber in Figure 7(f) contains ports and is designed to insert into standard 6-well plates that contain adherent cells on the plastic growth surface. The heated perfusion chamber presented in Figure 7(g) is designed to be mounted on a microscope stage with a circular aperture, and is provided with several ports around the perimeter of the central well. Note that literally hundreds of live-cell perfusion chambers are commercially available, so the investigator is encouraged to scan the availability of chambers for specific applications prior to committing to a single design.

Several of the important considerations in designing perfusion chambers are the mechanical attachment (if necessary) and stabilization of the specimen, fluid-mechanical flow characteristics and exchange time of the chamber, as well as the optical working distance and refractive indices of immersion oil, culture solutions, and glass plates through which observation will take place. Specimens for live-cell imaging in closed or open chambers can be immobilized either by their natural adherence to glass coverslips, by adhesion promoting agents (for example, poly-L-lysine and collagen), or by first embedding in a thin agar overlay. Requirements of the investigation will dictate the size of the glass surface area in a perfusion chamber that must be made available for observation. Random cell populations can often be used with relatively small coverslips, but much greater areas may be necessary for very large cells or networks, and in situations where many cells must be screened for particular events (such as mitosis).

Regardless of the coverslip surface area, the thin glass plates forming the specimen cavity (ideally, 170 micrometers in thickness) should be strain-free, optically flat, and positioned in a parallel orientation. Note that larger viewing areas are more susceptible to leaks and physical damage, but provide significantly greater access to high numerical aperture objectives that often have barrels ranging in diameter from 3 to 5 centimeters (and are even more bulky with a heater installed). Chambers should be constructed utilizing components that are not toxic to living cells. Modern commercially available culture chambers are constructed from glass, stainless steel, synthetic high strength polymers, Teflon, and silicone, often using adhesives during final assembly. Several of these materials may contain traces of toxic heavy metals or organic solvents that should be removed by a thorough cleaning prior to sterilization and use.

Perfusion chambers are usually necessary in order to ensure the viability of a cell culture when investigations require lengthy imaging sequences or if the serial addition of metabolites, tracers, and/or drugs must be accomplished without disturbing the viewfield. In addition, perfusion chambers enable the researcher to periodically sample the culture medium for production of cellular products and to conduct correlative studies where the specimen is monitored in the living state for a period of time followed by sudden fixation during observation to reconstruct structural or cytological events underlying a process of interest. The primary advantages of perfusion are the ability to maintain the pH and ionic concentration of the media during long-term experiments, as well as affording a mechanism to introduce a variety of reagents (infused into the chamber) without disrupting the cells. For example, perfusion enables the investigator to introduce a drug into the culture during a time-lapse sequence without interrupting image acquisition.

Among the most important features of a perfusion chamber is that the overall design must incorporate laminar flow characteristics so that solutions passing through can be exchanged rapidly and efficiently with relatively little mixing. For some investigations, laminar flow may not be of critical importance, but more rigorous studies often demand highly controlled media exchange, especially when other agents are being added. The fluid-mechanical properties of a perfusion chamber are closely related to the geometry (as illustrated in Figure 8). Laminar flow occurs when the chamber cross section is very similar, or even identical, to that of the inlet tube. The most unfavorable flow patterns occur when the chamber is large compared to the inlet tube. Figure 8 compares flow patterns in several popular perfusion chambers. The circular chamber illustrated in Figure 8(a) has poor exchange characteristics due to the sudden expansion of media from the inlet tube opening, which results in secondary flow and non-uniform streamlines. Gradual expansions, as depicted in Figures 8(b) and 8(c), result in more uniform streamlines and rapid exchange. The tapered regions at the entrance and exit tubes of the chamber design shown in Figure 8(b) improve the flow profile, while the diamond-shaped geometry of the chamber in figure 8(c) results in more even flow through the vessel. The microaqueduct chamber presented in Figure 8(d) has superior laminar flow characteristics due to the etched channel that directs media gently across the specimen.

As discussed above, it is often necessary to develop a reliable method of perfusing cells directly in their imaging chamber without impeding the experiment in terms of the acquisition of electrical, chemical, or optical data. Fluid flow through a perfusion chamber occurs by the exchange of new fluid passed into the inlet port, which in turn forces the spent fluid out through an exhaust port. Because most live-cell imaging chambers contain relatively minute volumes of tissue culture medium, the sensitivity of cells to shear stresses during perfusion, as well as the potential for creating vibrations of the coverslip at the aperture (often referred to as the diaphragm effect) must be considered when selecting a pumping system. In many cases, this selection process, as with most other factors in the imaging of living cells, involves one or more compromises. The ideal perfusion pump would provide a closely controlled, purely analog flow of media for an indefinite period, but this type of system has yet to be developed.

The standard techniques for delivery of fluids through perfusion are gravity flow, automated or manual injection using a syringe (see Figure 9(b)), and mechanical pumps. Gravity flow is simple to install and relatively inexpensive, but difficult to control at the flow rates typically necessary for microscopy. In the most basic configuration, a gravity feed system requires a large media reservoir placed in a location high above the microscope with a section of sterile tubing connected to the imaging chamber. The flow rate is adjusted by a needle valve controlling an exhaust tube attached to the opposite side of the chamber that drains into a catch basin. Alternatively, the rate of perfusion in a gravity system can be controlled by varying the vertical distance between the two reservoirs or by a stopcock that controls the flow from the upper reservoir. Laminar flow characteristics of gravity feed systems are often less than adequate, but will suffice in situations where the primary requirement is the ability to quickly exchange the culture medium.

For experiments that rely on the precise addition of drugs or other chemicals, as well as for superior control of sheer forces on the surface of the coverslip (where the cells are attached), a more sophisticated perfusion apparatus is necessary. Fluid exchange rates, which can be extremely critical in specialized experiments, have increased with improvements in chamber geometry. Half-times for fluid exchange of fewer than 50 milliseconds have been achieved in perfusion chambers without disturbing the specimen lateral position or axial focal plane. A step change in concentration may be achieved in fewer than 100 milliseconds with a relatively low linear flow velocity where the junction of the two perfusion solutions is positioned close to the chamber, provided that the chamber has a small internal volume and is designed to maintain laminar flow profiles. Many of the advanced perfusion chambers that are commercially available meet these criteria.

The periodic addition of growth factors, inhibitors, drugs, and other metabolites during perfusion can be readily accomplished using manual syringes. Mechanical pumps, available in peristaltic (Figures 9(a) and 9(b)) or motorized syringe (Figure 9(e)) designs, are the most reliable devices for delivering perfusate. The syringe pump is limited in volume for long-term experiments and subject to variations in delivery on a microliter flow scale (due to temporary sticking of the plunger). Pumps employing stepper motors can be controlled to very slow flow rates, but the instantaneous rotor movement results in hydrodynamic pulses that can produce coverslip flex or dislodge the cells. Advanced perfusion pumps feature regulated direct current motors coupled to a step-down transmission driving a roller spindle equipped with speed control. The result is a flow profile that is free of the sudden pulsations typical of cheaper peristaltic pumps. In choosing a perfusion delivery system, the investigator should consider the desired flow rate, uniformity of flow over various time scales, and total volume delivered, as well as the quality of flow.

In addition to ensuring suitable perfusion chamber fluid flow characteristics, the feeder solutions may also require attention with respect to the temperature and gas composition. If a carbon dioxide and bicarbonate buffer system is employed, all perfusion solutions should remain equilibrated with the appropriate partial pressure of carbon dioxide (dependent on the bicarbonate concentration) and maintained at the same temperature as the culture chamber. Thus, the perfusion solutions must be gassed and their temperature adjusted prior to their introduction into the perfusion chamber. Note that the solubility of carbon dioxide diminishes as temperature is increased, so the solutions must be gassed at the chamber temperature setpoint or the gas mixture must be adjusted for the temperature difference. A variety of commercial equipment is available for maintaining perfusion solution temperature and gas composition. Alternatively, suitable devices can be readily constructed in-house using common laboratory components. Simple humidity chambers can be fabricated with a media bottle, tubing, and a fish tank air stone (Figure 9(c)), while temperature can be controlled with a circulating water bath or desktop incubator. More sophisticated units (illustrated in Figure 9(d)) are designed to regulate both pH and humidity simultaneously. Because static solutions of media have a tendency to cool and lose carbon dioxide (subsequently producing an increase in pH), all solutions should be connected to the perfusion chamber using gas-impermeable tubing that passes through the heated water bath or incubator.

The choice of tubing couplers for live-cell perfusion experiments should be limited to those products (preferably bearing a USP Class VI certification) that exhibit a maximum degree of biocompatibility, thus ensuring the lack of tubing-derived toxicity during imaging sessions. Currently, there are three varieties of commercial tubing that meet this requirement: PharMed (available through numerous distributors), Tygon 2275, and C-Flex. Even though it is well suited for cell culture, the primary drawback of PharMed tubing is low opacity, which reduces the visibility of perfusate traveling through the tubing. As a result, bubbles that may form as the perfusate travels through the tubing are difficult to detect. In contrast, both C-Flex and Tygon tubing are nearly transparent and not prone to leaching plasticizers into the perfusate. The latter products are ideal for a majority of perfusion experiments.

A potential problem in using perfusion chambers is that the specimen often bounces out of focus as the coverslip flexes to a limited degree when the process of forcing fluid through the chamber is initiated or terminated, as discussed previously (diaphragm effect). This artifact is evident in chambers where the perfusion system is poorly controlled. Coverslip flexing may not present significant problems in cases where the image capture sequence can be interlaced between perfusion sessions, but it can seriously affect those studies that require relatively high temporal resolution (2 seconds or less). The severity of coverslip flex can be minimized by reducing the diameter of the input port (and tubing couplers) compared to that of the exhaust port so that hydrostatic pressure does not accumulate during fluid flow. In addition, substituting a gravity flow system for a peristaltic pump can reduce or eliminate the diaphragm effect.

A major concern in the use of all perfusion chambers is the formation of air bubbles (due to leaks, surface tension forces, and degassing), which can interfere with the laminar flow of culture medium as well as reduce contrast and sharpness during imaging. A properly configured perfusion chamber should form a closed system, and hidden air or potential leaks in the chamber or at tube junctions should be eliminated after the specimen is mounted on the microscope, but prior to initiating the experiment. In gravity systems, the simplest method to clear the bubbles is to fill the reservoirs and chamber, stop flow through the system, and then slowly purge the stopcock while examining all components. Chambers using a pump system are usually more difficult to correct, and careful attention should be paid to minute bubbles that may hide in the chamber housing between the coverslip and the perfusion ports. In all cases, the entire system should be completely closed and loaded with fluid prior to the first perfusion in order to minimize coverslip flex as perfusions are initiated and terminated. Ensure that the exhaust tube is immersed in the catch basin to avoid bubbles creeping back into the chamber.

For many investigations, the positioning of the input and exhaust ports of a perfusion chamber must not interfere with changing the microscope objectives. Most sandwich-style chambers are designed so the ports are positioned on the same or opposing chamber edges and can often be used with both upright and inverted microscopes. However, chambers having ports installed on the upper portion of the housing may obstruct motion of the objective turret on upright microscopes. Many of the modern commercial perfusion chambers (see Figures 1 and 7) are designed to enable full rotation of the microscope turret without obstruction, but this factor should first be tested before deciding on a specific chamber. Furthermore, plastic perfusion and imaging chambers that are embedded in large metallic stage adapters on inverted microscopes may not provide sufficient clearance for objective rotation without first lowering the objectives away from the chamber.

Although a vast majority of the investigations in live-cell imaging require maintenance of the culture at temperatures elevated above the ambient, some studies involve observations of organisms that must be grown at very low temperatures while others study the effects of sudden cooling on a specimen. For long-term experiments, the microscope should ideally be located in a cold room to avoid the problem of water condensation that forms on cold surfaces at room temperature. In the absence of a cold room, perfusion chambers designed for cooling applications are commercially available. These devices often rely on the flow of chilled alcohol solutions through pre-coiled lines or tubing to reduce temperature. Specialized objective and condenser coolers may be necessary to avoid heating of the specimen with these microscope elements, and frosting can be prevented by coating external glass surfaces with a chemical similar to Photo-Flo (Kodak). Regardless of the temperature requirements of the investigation, however, it should be noted that sustaining an exact temperature value is often not as important as consistently achieving and steadily maintaining a constant temperature.

Advanced Closed and Open System Imaging Chambers

In general, inverted (tissue culture) microscopes are the most common instruments employed for live-cell imaging, but upright microscopes can also be adapted for this purpose, with several important concessions. The investigator must exercise caution in designing an experiment (regardless of the microscope frame configuration) by determining the necessary magnification range, working distance restrictions, and optimum numerical aperture, all of which are related to the depth of field. In most cases, fluid coupled objectives (oil and water immersion) will benefit from being temperature controlled when utilized with mammalian cell cultures. In transmitted light applications requiring a high numerical aperture condenser, the chamber must be able to accommodate the working distance, physical size, and proximity of the condenser front lens element, especially when oiled to the upper chamber glass. The microscope stage geometry should also be matched with the dimensions of the culture chamber to ensure adequate clearance. Generally, inverted microscopes are superior for live-cell investigations due to the longer working distance condensers and the ability of these instruments to use high numerical aperture objectives directly on the optical surface supporting the cells.

A number of variables should be considered when employing advanced open chamber culture systems. In particular, the volume, aperture size, construction material, coverslip thickness, chamber geometry, and biocompatibility must be addressed. Other factors that could impact experimental success include evaporation (and condensation), ambient light, the view angle, and laboratory environmental conditions. For sophisticated closed system chambers, many of the same considerations apply; however, additional characteristics are also important. Among these are fixed or variable chamber volume requirements, the separation distance between optical surfaces, fluid exchange rate, laminarity, shear stress, perfusion flow gradients, and flow channel geometry. In addition, temperature control (and uniformity) and the thermal effect of oil or water immersion objectives are important factors for both system designs.

Closed system live-cell imaging chambers are required in situations where the culture must be completely isolated from the external environment, or in applications using advanced contrast-enhancing techniques for transmission optical microscopy (such as high resolution DIC). In these cases, the cells are placed in a temperature-controlled, perfusable optical cavity or grown directly on a coverslip that forms the lower optical window for the chamber. A wide variety of traditional and advanced-design aftermarket configurations for closed system imaging chambers are commercially available. The vast majority of these designs utilize the similar characteristics of providing two optical surfaces separated by a perfusion ring sealed with gaskets to form a sandwich that is clamped together in a metal or polymeric housing. Among the shortcomings found in many closed chamber systems are excessive separation between optical surfaces, fixed optical cavity volume, lengthy fluid exchange rates, turbulent perfusion, difficulty in loading, and leakage. Several of the more advanced open and closed system imaging chambers designed to address common problems and provide excellent imaging characteristics are described below.

Bioptechs Delta T Open System Chamber

The Bioptechs Delta T culture chamber system features a number of innovative design motifs that can be utilized to a surprising advantage for a wide spectrum of static and perfusion-based live-cell imaging experiments. Modeled after traditional open chamber Petri dishes, the Delta T culture chamber consists of a 35-millimeter dish fused to a 170-micrometer coverslip that spans the entire bottom and employs a first-surface thermal transfer technique to maintain temperature control (illustrated in Figure 10). The thermal control strategy is based upon a thin-film coating of indium-tin oxide (ITO), a transparent, electrically conductive layer applied to the lower surface of the coverslip. Conductive contacts deposited at opposite edges of the coverslip engage a series of electrodes when the dish is properly seated in the specialized stage adapter. Temperature is regulated through a thermistor feedback loop (housed in an external controller unit), which applies an electrical current to the coated underside of the coverslip, thereby heating the entire dish to the desired temperature. The thermal response using this technique is about 0.1 degree Celsius per second.

The fast response time of the Delta T system enables the controller to switch current flow on and off within seconds, making it possible to compensate for temperature changes that occur in the dish due to surface evaporation, entropy, or perfusion. Coupled to the rapid response time is a high-speed safety circuit designed to protect the cells in the event of a controller error. The system provides high resolution imaging capabilities through a strain-free uniform glass surface and is available with a variety of cover glass thicknesses, including the number 1.5 coverslip designed for high numerical aperture applications. The dish environment is compatible with the popular contrast-enhancing modes of brightfield, darkfield, polarized light, phase contrast, DIC, HMC, fluorescence, multiphoton, and confocal microscopy.

Introduced in 1993, the first-surface thermal transfer technique using a thin-film coating on the coverslip provides the basis for a wide spectrum of adaptations and variations of the basic principle. For example, the Delta T and similar systems are useful for the control of temperature during investigations of tissue slice specimens that can be suspended in an optical reference plane at the appropriate focal distance from the objective. In addition, adapters can be readily installed to maintain low-volume perfusion for extended periods of time. A tightly fitted lid is also available for the Delta T dishes for use in high numerical aperture experiments to define an optical surface above the specimen that is completely filled with media or imaging buffer. This device eliminates the interface between air and the surface of the medium, thus preventing the shifts in image contrast that result from minute fluctuations in the media volume above the objective. In addition, an electrically heated, optically transparent cover glass lid can be used for low magnification transmission or high magnification fluorescence experiments. Thermal control of the lid cover glass prevents condensation from forming on the lower surface. Furthermore, addition of a gas port enables the heated lid to be used for maintaining the carbon dioxide concentration in the culture atmosphere over extended periods of time. Cooling of specimens with the Delta T system can be accomplished by submerging into the dish a coiled tube containing recirculated refrigerant fluid.

Bioptechs FCS-2 Closed System Chamber

Introducing a unique microaqueduct-controlled perfusion technique, the Bioptechs closed system FCS-2 culture chamber is an advanced solution to many of the difficulties commonly encountered with shear stress and flow geometry in perfusion chambers. In this system, the microaqueduct observation chamber is fabricated by incorporating perfusion grooves into one of the surfaces that define the optical cavity (see Figures 1(b), 8(d), and 11(b)). This design eliminates the perfusion ring common to many other chambers by creating an optical cavity with only a single gasket separating the perfusion slide from the coverslip. The physical configuration of the microaqueduct grooves produces an excellent laminar flow region in the optical cavity, while the single-gasket design enables the investigator to define the size, volume, thickness, and shape of the chamber. In addition, the microaqueduct perfusion chamber provides large aperture flow inputs and outputs, thus minimizing the problem of volume exchange rates. In order to further enhance the performance of the FCS-2 chamber system, the first-surface thermal transfer technique (described above) is applied to the microaqueduct glass, thus adding thermal uniformity to the chamber even during periods without flow. Thus the cell culture is maintained safely in a temperature-controlled optical environment that is compatible with all modes of optical microscopy.

Tokai Hit Stage Incubator

Marketed under a variety of trade names and configurations by several of the major microscope manufacturers, the Tokai Hit stage top incubator (model INU) is a combination system that can be utilized either as a perfusion device or as a static chamber to simulate the conditions found in a humidified carbon dioxide incubator (illustrated in Figure 11(a)). High humidity levels are maintained through a circular, heated water tank equipped with carbon dioxide injectors fed by a remote gas-mixing unit. The condenser aperture cover glass is heated by a first-surface coating to prevent the formation of condensation while enabling the application of transmitted light differential interference contrast and polarized light imaging techniques at low numerical aperture. Access ports in the cover glass allow the injection of metabolites and drugs during the image acquisition session. The chamber is designed with perfusion accessories and an anti-drift plate positioned above the heated stage insert to isolate the culture dish from the heating elements, thus minimizing focus drift. An optional objective heater can be used with immersion objectives to maintain specimen temperature. A variety of chamber designs and gas mixers are available for the most popular commercial microscopes.

The most comprehensive commercial package of imaging chambers is produced by Warner Instruments, specialists in tools for electrophysiology and cell biology research. The company offers a staggering array of imaging and recording chambers, stage adapters (see Figure 7(c)), temperature controllers, perfusion systems, objective warmers, spill sensors, and in-line perfusion heaters designed for a wide spectrum of applications. In addition, as live-cell imaging becomes a more popular technique in cell biology laboratories, a growing number of innovative culture chamber architectures, intended both for perfusion and static culture, will continually be introduced by aftermarket manufacturers. Before committing to a specific commercial live-cell imaging culture chamber system, the investigator is advised to carefully peruse the abundant on-line and published scientific literature.


BACK TO LIVE-CELL IMAGING

Contributing Authors

Michael E. Dailey - Department of Biological Sciences and Neuroscience Program, 369 Biology Building, University of Iowa, Iowa City, Iowa, 52242.

Daniel C. Focht - Bioptechs Inc., 3560 Beck Road, Butler, Pennsylvania, 16002.

Alexey Khodjakov and Conly L. Rieder - Wadsworth Center, New York State Department of Health, Albany, New York, 12201, and Marine Biological Laboratory, Woods Hole, Massachusetts, 02543.

Kenneth R. Spring - Scientific Consultant, Lusby, Maryland, 20657.

Nathan S. Claxton, Scott G. Olenych, John D. Griffin, and Michael W. Davidson - National High Magnetic Field Laboratory, 1800 East Paul Dirac Dr., The Florida State University, Tallahassee, Florida, 32310.