May 15, 2014 (Vol. 34, No. 10)

Striking the Right Balance between Signal Strength and Long-Term Cell Health

A key challenge with live-cell fluorescence imaging is the ability to visualize weak fluorescent signals over background without inducing photobleaching or phototoxicity from high excitation light intensities. Most cell culture media contain components that emit significant autofluorescence when excited, thus negatively impacting fluorophore signal-to-noise (S:N) ratios.

In turn, researchers are forced to image their cells in balanced salt solutions instead of growth medium, or to use suboptimal microscope settings, both of which lead to poor cell health. Here, we present FluoroBrite™ DMEM, a cell culture media designed specifically for live-cell imaging that, when excited at wavelengths commonly used for FITC/GFP (i.e., ~488 nm), has the optical properties of PBS while still possessing the nutrients required for maintaining long-term cell health during imaging experiments and routine cell culture.

Materials & Methods

Media, Cell Lines & Reagents
The following Thermo Fisher Scientific media and reagents were used: (1) Phenol red free (PRF)-DMEM, (2) FluoroBrite  DMEM, (3) GlutaMAX™, (4) U.S. Certified Fetal Bovine Serum, (5) Dextran Alexa Fluor® 488, and (6) Premo™ FUCCI Cell Cycle Sensor dye. The following ATCC cell lines were used: NIH 3T3, HEK 293, HepG2, and Sp2/0-Ag14.

Media Characterization
Spectral scans of media were performed to measure relative fluorescence units (RFU) emitted between 450 and 650 nm in response to a wide spectrum of excitation light (400–600 nm) using a SpectraMax M5 Microplate reader (Molecular Devices). Fluorophore S:N ratios were evaluated by diluting Dextran Alexa Fluor 488 into various media and measuring the fluorescent signal emitted at a single excitation (488 nm) and emission (509 nm) wavelength with an Enspire Plate Reader (PerkinElmer). Media autofluorescence in relation to light intensity was determined using an EVOS® FLoid Imaging Station (Thermo Fisher) and ImageJ (NIH).

Time-Lapse Microscopy
NIH 3T3 cells were labeled with Premo™ FUCCI Cell Cycle Sensor dyes according to the manufacturer’s protocol and imaged in either FluoroBrite DMEM or PRF-DMEM, both supplemented with 10% FBS and 4 mM GlutaMAX. Time-lapse experiments were performed at 37°C and 5% CO2 with an EVOS FL Auto Cell Imaging System. All wells were imaged once every 15 minutes for up to 10 hours under GFP (100 msec) and RFP (200 msec) imaging settings.

Long-Term Cell Culture Studies
HEK 293, HepG2, and Sp2/0-Ag14 cell lines were cultured in FluoroBrite DMEM or PRF-DMEM, both supplemented with 10% FBS and 4 mM GlutaMAX. Viable cell density and viability were determined by ViCell (BD) at the time of each subculturing for up to 24 passages.

Results & Discussion

Vitamins represent key components of all cell culture media; however, some negatively impact fluorescence microscopy by causing high levels of media autofluorescence. To evaluate the impact of vitamins on media-induced autofluorescence, spectral scans across a wide range of excitation (400–600 nm) and emission (450–650 nm) wavelengths were performed on standard PRF-DMEM and a PRF-DMEM prototype formulation with reduced vitamin content. The results demonstrated a direct correlation between vitamin content and media autofluorescence at wavelengths used to excite FITC/GFP (Figure 1A). Based on these findings, FluoroBrite DMEM, a phenol red-free cell culture media with background autofluorescence comparable to PBS, was developed to allow for both long-term live-cell imaging and routine cell culture.

The ability of FluoroBrite DMEM to enhance the S:N ratio of a green-emitting dye was evaluated using Dextran Alexa Fluor® 488 beads diluted into either FluoroBrite  DMEM or PRF-DMEM. When excited at 488 nm, Dextran Alexa Fluor 488 beads in FluoroBrite DMEM generated a S:N ratio that was ninefold greater than that observed for PRF-DMEM (Figure 1B).

The benefit of FluoroBrite DMEM was not as pronounced at excitation wavelengths outside of the FITC/GFP range; however, the elimination of autofluorescence in the green channel allowed for better resolution and visualization of other fluorophores in multi-color image overlays (Figure 1C). FluoroBrite DMEM also maintained consistent levels of background autofluorescence during long-term imaging compared to PRF-DMEM, allowing for easier and more accurate quantitation of fluorescence intensities over time without having to account for changing background levels (Figure 1C).


Figure 1. Impact of media autofluorescence on fluorophore S:N ratios and multicolor image overlays. (A) Spectral scan of media autofluorescence for PRF-DMEM, PRF-DMEM with reduced vitamin content and PBS excited at 480 nm. (B) S:N ratio of Dextran Alexa Fluor® 488 beads in FluoroBrite DMEM, PRF-DMEM and PBS. (C) Time-lapse microscopy of NIH 3T3 cells labeled with Premo™ FUCCI Cell Cycle Sensor dyes imaged in FluoroBrite DMEM and PRF-DMEM for eight hours. Green, red, and yellow cells are in G2/M, G1, and S phases of the cell cycle, respectively.

FluoroBrite DMEM was next evaluated for its ability to maintain low autofluorescence when excited by increasing intensities of 482 nm light. In contrast to PRF-DMEM, FluoroBrite DMEM autofluorescence levels remained low even when excited with extremely high excitation light intensities (Figure 2A and B), a benefit in instances where intense excitation light is necessary to resolve weak fluorophore signals over background.

In addition to low autofluorescence, FluoroBrite DMEM was specifically formulated for use in routine cell culture maintenance to eliminate the need for a media exchange into PBS or similar buffer at the time of imaging. Multipassage cell culture experiments were carried out in FluoroBrite DMEM, PRF-DMEM, and a competitor’s low autofluorescence culture media for up to 24 passages using adherent HEK 293 cells, vitamin-sensitive HepG2 cells, as well as the Sp2/0-Ag14 cell line, which is well known for its sensitivity to nutrient depletion. All cell lines showed equivalent growth and viability between FluoroBrite DMEM and control PRF-DMEM for all passages tested; however, when compared to the competitor’s imaging media, FluoroBrite DMEM achieved significantly better growth for all cell lines, as well as superior viability for the Sp2/0-Ag14 cell line (Figure 3). Indeed Sp2/0-Ag14 cell density and viability decreased rapidly in the competitor’s media and cultures could not be maintained past three passages of growth.


Figure 2. FluoroBrite DMEM maintains low levels of media autofluorescence even when excited with high-intensity incident light. Equal volumes of FluoroBrite DMEM and PRF-DMEM were excited with an EVOS® FLoid Imaging Station for the same time duration with increasing intensities of 482 nm light. The amount of media autofluorescence emitted at 532 nm was (A) imaged and (B) quantified for integrated light density for each media.

Conclusions

FluoroBrite DMEM represents the next generation of cell culture media that allows for both enhanced visualization of fluorescent probes in the absence of media-induced autofluorescence and routine cell culture, thus eliminating the need to exchange media prior to imaging. With optical properties similar to PBS, FluoroBrite DMEM delivers superior S:N ratios for fluorophores excited at 488 nm, making it ideal not only for microscopy, but also for high-throughput/high-content screening and multi-well cell-based assays.

Additionally, FluoroBrite DMEM provides an indirect benefit for the visualization of other fluorophores by eliminating the unwanted haze of green autofluorescence in multicolor image overlays. The extremely low autofluorescence of FluoroBrite DMEM allows for the use of lower-incident light intensities, which in turn reduces cellular phototoxicity and helps sustain long-term cell health during live-cell imaging experiments.


Figure 3. FluoroBrite DMEM maintains consistent cell health during long-term cell culture. Adherent HEK 293, vitamin-sensitive HepG2, and nutrient-sensitive Sp2/0-Ag14 cell lines were cultured in FluoroBrite DMEM, PRF-DMEM, and a competitor’s low-autofluorescence media and monitored for viable cell density and viability for up to 24 passages. Viable cell density (left Y-axis) is plotted as a solid line representing cells per cm2 for HEK 293 and HepG2 cell lines and cells/mL for the Sp2/0-Ag14 cell line. Percent viability (right Y-axis) is plotted as a dotted line for all cell lines.

Virginia A. Spencer ([email protected]) is staff scientist, Shyam Kumar is scientist III, Brian Paszkiet is staff scientist, Jeffrey Fein is scientist III, and Jonathan F. Zmuda is R&D leader at Thermo Fisher Scientific.

Previous articleU.K. Research Organizations Commit to Openness on Animal Testing
Next articlePatentability of 3D-Printed Organs