Beetle luciferases are used as reporter enzymes for monitoring gene expression in living cells because the amount of luciferase present is correlated with emitted light intensity in the presence of excess intracellular luciferin. This method is widely used in chronobiological and pharmacological research (1,2,3). When the luciferase gene containing a target promoter region is transfected into target cells using a plasmid and the transfected cells are cultured in medium containing luciferin, the expression of luciferase can be estimated from bioluminescence for several days or even weeks using suitable equipment. Recently, Noguchi et al. developed a dual-color real-time reporter system in which the expression of two genes could be monitored simultaneously by splitting emissions from green and red-emitting beetle luciferases using an optical filtered luminometer (4). This system is very convenient for comparing the expression of target and control genes in living cell populations. However, although bioluminescence imaging of cells is a powerful tool (5,6,7), monitoring has been limited to single target genes. In this study, we constructed a dual-path luminescence imaging system to simultaneously monitor the expression of two genes in a single cell using a dual-color reporter system.

Our dual-path luminescence imaging system is based on a commercial single cell luminescence imaging apparatus, the Cellgraph AB-3000B (8) (ATTO Corp., Tokyo, Japan) (see Supplementary Materials, Method 1). Figure 1A shows a schematic diagram of our final imaging apparatus. Luminescence generated in a target cell is collimated by the objective lens and is divided by the dichroic mirror DM1 (DIM-50S-RED, Sigma Koki, Tokyo, Japan) into green and red light. Green light, with a wavelength shorter than the cutoff wavelength of DM1, passes through, whereas red light, with a longer wavelength, is reflected. Light is then reflected by the mirrors M1 and M2 to the dichroic mirror DM2, which has the same optical property as DM1, so that green light passes through and the red is reflected. The optical axes of the mirror are parallel but the distance between them is half the width of the charge-coupled device (CCD) sensor, which has dark current of 0.001 e⁻/pixel/s and a readout noise of <1 e⁻ at 1MHz with electron multiplication. The green and red luminescence emissions are focused on different areas of the CCD by the imaging lens. Therefore, the single CCD obtains two images of the cell simultaneously. Figure 1B shows the bioluminescence spectra of green- and red-emitting luciferases expressed by the pEluc-test and pSLR-SV40 plasmid vectors (Toyobo, Osaka, Japan), respectively, and the transmission spectrum of the dichroic mirror.

Figure 1. Dual-path luminescence imaging apparatus. (Click to enlarge)

For multicolor imaging, the green and red luminescence intensities (G and R) from the two luciferases in the living cell were calculated from the equation

where F₁ and F₂ are the relative light units (RLU) of light components that are either reflected by the dichroic mirror or pass through it, respectively; similarly, κRr and κRt are the RLU ratios of red luminescence that are reflected or pass through and κGr and κGt are the RLU ratios of green luminescence that are reflected or pass through. In this experiment, the respective κGr and κGt values were 0.235 and 0.765, and the κRr and κRt values were 0.97 and 0.03, respectively. Based on the above calculation, the green and red components could be reconstituted and visualized as a two-color image (Supplementary Materials, Method 2).

We visualized the expression of two circadian rhythm genes, mBmal1 and mPer2, using our system. The vectors, plasmid Bmal1-pSLR containing the mBmal1 0.9-kb promoter (9) and plasmid Per2-pEluc containing the mPer2 1-kb promoter (10), were transfected into NIH3T3 cells maintained in DMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 µg/mL streptomycin. Aliquots of 7.5 × 10⁵ cells were plated in 35-mm dishes, cultured for 24 h, and then transfected with 1.6 µg Bmal1-pSLRand 0.8 µg Per2-pEluc plasmids using Lipofectamine LTX (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Twenty-four hours after transfection, the cells were treated with 0.1 µM dexamethasone (Sigma-Aldrich, St. Louis, MO, USA) for 2 h to synchronize their circadian rhythms (4) and then the medium was replaced with 2 mL DMEM supplemented with 10% FBS, 10 mM HEPES buffer pH 7.2, 0.1 mM D-luciferin (Wako, Osaka, Japan), 100 U/mL penicillin, and 100 µg/mL streptomycin. The cellular bioluminescence signals were recorded continuously with the dual-path luminescence imaging apparatus for 4 days. Detection of luminescence required 29-min exposures every 30 min. The electron multiplier gain of the CCD camera was set to 900. The resulting images were analyzed and processed using PC software (Cellgraph Viewer, ATTO; see Supplementary Materials, Method 2).

Dual-color bioluminescence imaging showed that the expression of mBmal1 and mPer2 oscillate in anti-phase, with a circadian period of ~24 h in individual cells over 4 days (Supplementary Movies S1 and S2). However, disorders such as ‘blinking’ and varying periodicities were observed several times at the single cell level although the bioluminescence monitoring data from the cell population shows robust and stable circadian rhythm during the same time (Supplementary Figure S1). Figure 2A shows the representative real-time imaging data at each time point, in which each spot is on a separate cell. We analyzed the bioluminescence intensity of mBmal1 and mPer2 at single-cell level inselected cells. As shown in Figure 2B and Supplementary Figure S2, the expression of mBmal1 and mPer2 shows a circadian rhythm but sometimes also an unstable oscillation pattern, which may be primarily due to toxicity resulting from DNA transfection, or may depend on cell conditions. Also, circadian rhythms in individual cells might not be synchronized perfectly even though the rhythmic pattern in a whole cell population is shown to be robust and stable.

Figure 2. (Click to enlarge)

This monitoring system revealed the dynamics of dual-gene expression in a single cell and illustrated the possibility of analyzing cell-cell interactions in the population. This system could be used for the detailed analysis of transcriptomes and promoteromes, and for screening new drugs or detecting harmful chemicals in a single cell.

2.) Buijs, R.M., and A. Kalsbeek. 2001. Hypothalamic integration of central and peripheral clocks. Nat. Rev. Neurosci. 2:521-526.

3.) Reppert, S.M., and D.R. Weaver. 2002. Coordination of circadian timing in mammals. Nature 418:935-941.

4.) Noguchi, T., M. Ikeda, Y. Ohmiya, and Y. Nakajima. 2008. Simultaneous monitoring of independent gene expression patterns in two types of cocultured fibroblasts with different color-emitting luciferases. BMC Biotechnol. 8:40.

5.) Inagaki, N., S. Honma, D. Ono, Y. Tanahashi, and K. Honma. 2007. Separate oscillating cell groups in mouse suprachiasmatic nucleus couple photoperiodically to the onset and end of daily activity. Proc. Natl. Acad. Sci. USA 104:7664-7669.

6.) Hiler, D.J., A. Bhattacherjee, S. Yamazaki, H. Tei, and M. E. Geusz. 2008. Circadian mPer1 gene expression in mesencephalic trigeminal nucleus cultures. Brain Res. 1214:84-93.

7.) Asai, S., K. Takamura, H. Suzuki, and M. Setou. 2008. Single-cell imaging of c-fos expression in rat primary hippocampal cells using a luminescence microscope. Neurosci. Lett. 434:289-292.

8.) Hoshino, H., Y. Nakajima, and Y. Ohmiya. 2007. Luciferase-YFP fusion tag with enhanced emission for single-cell luminescence imaging. Nat. Methods 4:637-639.

9.) Yu, W., M. Nomura, and M. Ikeda. 2002. Interactivating feedback loops within the mammalian clock: BMAL1 is negatively autoregulated and upregulated by CRY1, CRY2, and PER 2. Biochem. Biophys. Res. Commun. 290:933-941.

10.) Desmarais, D., M. Filion, L. Lapointe, and A. Royal. 1992. Cell-specific transcription of the peripherin gene in neuronal cell lines involves a cis-acting element surrounding the TATA box. EMBO J. 11:2971-2980.