Based on our experience in the field of live cell microscopy, we have compiled a primer on tips and tricks of imaging live cells that beginners may find helpful. Some adjustment may be required for individual needs, but this information may be a useful starting point for many experiments.

Media: We have found that most cells grow well in phenol red free L-15 + 10% FCS in ambient air conditions. This media is more complete than simple HBSS, allowing for longer imaging sessions (sometimes multiple days). For low density cell cultures, 2 ml of L-15 per 35 mm dish covered with a thin layer of light mineral oil works well. For slice cultures, perfusion may be necessary due to high oxygen demands, but we have seen good cell migration in brain slices cultured in non-perfused L-15/mineral oil. For situations in which bicarb-buffered media is required, try to obtain a phenol-red free version of the media, and use it in conjunction with a CO2 chamber.

Chamber: The most problematic aspect of these simple approaches is the thermal drift experienced when using high magnification lenses caused by temperature fluctuations in the room the microscope is located in. Often times, it is necessary to constantly re-adjust the focus during image acquisition due to this thermal-drift, which can introduce artifacts during quantitative microscopy due to the photobleaching that occurs during re-focusing. The optimal solution to this problem is to use a temperature controlled plexiglass chamber that encloses the entire imaging area (or at least the objectives and the stage). Using such a chamber will result in highly stable temperature control and minimal thermal drift.

Cameras: Higher quality cameras will allow for shorter exposure times, resulting in better resolution due to less motion-induced image smearing, reduced phototoxicity, and the ability to take more pictures before the onset of photobleaching/phototoxicity (thus resulting in higher frame rates or longer imaging periods).

Objectives: High quality objectives. Numerical Aperature, NOT magnification, is the most important aspect here, as long as you are using a camera with small pixels (e.g. 7 um pixels for a 60X 1.4 N.A. lens). The higher N.A. of the lens, the better the spatial resolution and the brighter the image. Upright microscopes are currently limited to 60X 0.9 N.A. water immersion lenses, although higher N.A. long working distance lenses are in development. The trick to using these objectives in live cell microscopy is to immerse the objective into the media, then add a layer of mineral oil to the media to prevent evaporation. As long as the oil is quickly cleaned off after imaging, no damage to the objective should occur. Inverted scopes are better in this respect, and lenses up to 1.65 N.A. are now available, although 1.4 N.A. lenses are most commonly used. For longer working distances, and imaging samples where the cells are not plated directly on the coverslips, water immersion coverslip corrected lenses are preferable (1.2 N.A. is currently the max for these lenses).

Software: Especially if complicated multi-color fluorescence microscopy is being performed, a more sophisticated software package than what will come with your camera will be required. We are most familiar with Metamorph, which has proven to be an excellent choice, albeit an expensive one. It is relatively user-friendly, and allows for sophisticated macros to be written (called “journals”) for even greater flexibility. Another good choice is Open Lab. Zeiss software is probably the most user-friendly, but experienced users may find it to be rather limited given the inability to create macros.

Fluorophores-GFP family proteins: The specific fluorophores employed during imaging make a big difference in terms of signal-to-noise. Although EGFP is widely employed to study protein dynamics, newer variants have proven to be superior for live cell microscopy. We currently use a new EYFP variant called “Venus” as our tag of choice, which is brighter than EGFP, and matures 20X faster than EGFP. In addition, since Venus is further red-shifted than EGFP, less scattering occurs in the sample being imaged, resulting in a greater signal-to-noise ratio. For multi-color imaging, ECFP-tagged proteins can easily be distinguished from Venus-tagged proteins in the same cell if the correct filter sets are used. The drawbacks to using ECFP are low brightness, increased scatter, and faster photobleaching compared to Venus. Alternatively, a new monomeric RFP variant has recently been developed by Roger Tsien, known as mRFP1, that should prove to be useful for imaging protein dynamics. The main benefits of using this protein are the ability to add this protein as a third color to CFP and YFP imaging, as well as the fact that red-fluorescent proteins can be imaged at deeper levels in living tissue due to less scattering. The major drawbacks for using this protein are its large size compared to GFP, as well as its relatively low brightness. Recently, a new generation of monomeric red fluorescent proteins have been developed by Roger Tsien that are much brighter and more photostable. These come in colors ranging from yellow-orange to near-infared.

Specific equipment required for multi-color imaging: Either a filter wheel based system (e.g. Lambda 10-2, Sutter) or a device that allows for simultaneous multi-color imaging (e.g. Dual-View, Optical Insights). The benefits of a filter wheel device are high transmission efficiency, and the flexibility to use a wide range of filter sets. The filter wheels sold directly by the main microscope manufacturers are relatively slow compared to the Sutter Lambda 10-2 system. The benefits of using a Dual-View device is the fact that there are no motion-related artifacts that can occur during the sequential imaging that takes place when using a filter wheel. This is especially important when doing quantitative, multi-color fluorescence experiments, such as FRET, where even single-pixel motion-related registration artifacts can cause major problems. The drawbacks are the limitation to only doing two color imaging, and the fact that the effective imaging area of the CCD chip is reduced by half. It is also important to consider the light source. Although mercury arc lamps are the most prevalent lamps used in epifluorescent microscopy, they do not provide uniform illumination across the field of view, and exhibit large fluctuations in light output over short periods of time. Both of these aspects become problematic when doing quantitative microscopy. To overcome the non-uniform illumination problem, one can use a liquid light guide to couple the lamphouse to the microscope. Using such a device scatters the light, and causes a uniform field of illumination. To overcome output fluctuations, either a feedback mechanism to control input power to the bulb can be used, or one can switch to a Xenon lamp, which has a much more constant output (although this is at the cost of brightness).

Additional Notes: Important things to keep in mind when conducting live-cell fluorescent imaging. Do not use phase-contrast objectives, if possible. The phase ring on the image will substantially reduce the amount of light collected.

Keep light exposure to a minimum. The use of neutral density filters is encouraged when you are trying to find a good cell for imaging. Keep the shutter closed whenever you aren’t looking at the cell being imaged, and minimize the time it takes to center and focus the cell being imaged.

If you aren’t doing sequential DIC microscopy and fluorescent microscopy, remove all prisms/analyzers from the optical path.

Make sure the gain settings on the camera you are using are optimized.

Consider using “binning” to increase the sensitivity of the CCD camera. A binning of 2 will increase the sensitivity 4X at the expense of a reduction in resolution by half.

If one experiences high background, check to see if it is caused by reflection off of the transmitted light optics; if so, remove these from the light path.