Department of Biological Sciences, The University of Memphis
This article is a part ofJoVE General. If you think this article would be useful for your research, please recommend JoVE to your institution's librarian.Recommend JoVE to Your Librarian
Current Access Through Your IP Address
Current Access Through Your Registered Email Address
Ramanathan, C., Khan, S. K., Kathale, N. D., Xu, H., Liu, A. C. Monitoring Cell-autonomous Circadian Clock Rhythms of Gene Expression Using Luciferase Bioluminescence Reporters. J. Vis. Exp. (67), e4234, doi:10.3791/4234 (2012).
In mammals, many aspects of behavior and physiology such as sleep-wake cycles and liver metabolism are regulated by endogenous circadian clocks (reviewed1,2). The circadian time-keeping system is a hierarchical multi-oscillator network, with the central clock located in the suprachiasmatic nucleus (SCN) synchronizing and coordinating extra-SCN and peripheral clocks elsewhere1,2. Individual cells are the functional units for generation and maintenance of circadian rhythms3,4, and these oscillators of different tissue types in the organism share a remarkably similar biochemical negative feedback mechanism. However, due to interactions at the neuronal network level in the SCN and through rhythmic, systemic cues at the organismal level, circadian rhythms at the organismal level are not necessarily cell-autonomous5-7. Compared to traditional studies of locomotor activity in vivo and SCN explants ex vivo, cell-based in vitro assays allow for discovery of cell-autonomous circadian defects5,8. Strategically, cell-based models are more experimentally tractable for phenotypic characterization and rapid discovery of basic clock mechanisms5,8-13.
Because circadian rhythms are dynamic, longitudinal measurements with high temporal resolution are needed to assess clock function. In recent years, real-time bioluminescence recording using firefly luciferase as a reporter has become a common technique for studying circadian rhythms in mammals14,15, as it allows for examination of the persistence and dynamics of molecular rhythms. To monitor cell-autonomous circadian rhythms of gene expression, luciferase reporters can be introduced into cells via transient transfection13,16,17 or stable transduction5,10,18,19. Here we describe a stable transduction protocol using lentivirus-mediated gene delivery. The lentiviral vector system is superior to traditional methods such as transient transfection and germline transmission because of its efficiency and versatility: it permits efficient delivery and stable integration into the host genome of both dividing and non-dividing cells20. Once a reporter cell line is established, the dynamics of clock function can be examined through bioluminescence recording. We first describe the generation of P(Per2)-dLuc reporter lines, and then present data from this and other circadian reporters. In these assays, 3T3 mouse fibroblasts and U2OS human osteosarcoma cells are used as cellular models. We also discuss various ways of using these clock models in circadian studies. Methods described here can be applied to a great variety of cell types to study the cellular and molecular basis of circadian clocks, and may prove useful in tackling problems in other biological systems.
1. Construction of Lentiviral Luciferase Reporters
A mammalian circadian reporter construct usually contains an expression cassette in which a circadian promoter is fused with the luciferase gene. Both ligation- and recombination-based strategies are commonly used for DNA cloning. As an example, here we describe a recombination-based Gateway cloning method for generating a P(Per2)-dLuc lentiviral reporter, in which the destabilized luciferase (dLuc) is under control of the mouse Per2 promoter.
2. Production of Lentiviral Particles
1. Seed 293T cells (day 1)
2. Transient transfection via CaPO4/DNA precipitation (day 2)
3. Harvest viral particles (days 3-4)
3. Infection of 3T3 Cells
1. Seed 3T3 cells (day 3)
Split and seed appropriate number (~12,000) of 3T3 cells on a 12-well plate to obtain 20-30% confluence by next day. Incubate at 37 °C overnight.
2. Infect 3T3 cells (day 4)
3. Select infected cells (day 5 and onward)
4. Bioluminescence Recording of Reporter Cells
1. Seed reporter cells
Propagate Blasticidin-resistant reporter cells and split onto 35-mm culture dishes. Incubate at 37 °C until confluent. We usually prepare ≥3 dishes for each reporter cell line under each condition for circadian phenotyping.
2. Synchronization and change to recording medium
3. Bioluminescence recording of reporter cells
5. Data Analysis and Presentation
Reporter cells facilitate high-resolution quantitative luminescence recording, critical for determining phenotypic effects on circadian clock function. To obtain circadian parameters including phase, period length, rhythm amplitude, and damping rate, we use the LumiCycle Analysis program (Actimetrics) to analyze bioluminescence data5,14. Briefly, raw data are baseline fitted first, and baseline-subtracted data are fitted to a sine wave, from which the parameters are determined. For samples that show persistent rhythms, goodness-of-fit of >90% is usually achieved. Due to high transient bioluminescence upon medium change, we usually exclude the first cycle of data from analysis.
For data presentation, we usually plot raw data (bioluminescence, counts/sec) against time (days). When necessary, baseline-subtracted data can be plotted to compare amplitude and phase.
6. Representative Results
1. Phase-specific circadian reporters
The circadian clock is based on a biochemical negative feedback mechanism1. The core feedback loop consists of transcriptional activators BMAL1 and CLOCK, and repressors PERs and CRYs, which act on the circadian E/E'-box enhancer elements to produce rhythmic gene expression (with morning phase, e.g., Rev-erbα). The core loop regulates and integrates at least two other circadian cis-elements, the DBP/E4BP4 binding element (D-box; for day phase, e.g., Per3) and the ROR/REV-ERB binding element (RRE; for night phase, e.g., Bmal1)17. Combinatorial regulation by multiple circadian elements can generate novel intermediate phases. For example, Cry1 transcription is mediated by all three circadian elements (i.e., E/E'-box and D-box elements in the promoter and RREs in the first intron of the Cry1 gene), giving rise to the distinct Cry1 evening-time phase13.
Based on these mechanisms of gene regulation, we generated four different reporter constructs: P(Per2)-dLuc and P(Cry1)-dLuc reporters containing both E/E'-box and D-box elements in the regulatory region17,26,27; P(Cry1)-Intron-dLuc representing combinatorial regulation by all three elements (i.e., E/E'-box, D-box, and RRE)13,17; and P(Bmal1)-dLuc regulated exclusively by RRE9,17,19,21. We introduced these reporters into 3T3 cells to produce the anticipated distinct phases of reporter expression (Figure 2).
2. Gene knockdown via RNAi and pharmacologically active compounds
When transfection efficiency is high, synthetic siRNA can be transiently transfected into cells to knock down gene expression. When transfection is technically difficult, an shRNA expression vector can be stably transduced into cells via lentiviral infection, so that shRNA produced by the cell is processed to siRNA for gene knockdown (KD). Here we present KD effects of Cry1 and Cry2 genes using siRNA in U2OS cells (Figure 3A) and shRNA in 3T3 cells using a pLL3.7 Gateway expression vector9 (Figure 3B). In addition to 3T3 cells, the U2OS model has become another preeminent cellular clock model largely because it meets the key requirements for high-throughput screening of commercially available human siRNA libraries (e.g., human origin, capable of generating robust circadian rhythms, validated function of all known clock genes, and amenable to highly efficient transfection and quantitative luminescence recording). In both cell types, RNAi-mediated KD resulted in clock phenotypes consistent with previous mouse knockout (KO) and cellular KD studies5,10,11,28. For example, Cry1 KD shortens period length and reduces rhythm persistence, whereas Cry2 KD lengthens period. In addition, selected small molecules can be used to pharmacologically target and perturb protein function (Figure 3C).
Figure 1. The lentivirus-mediated gene delivery system. (A) Schematic diagram of two lentiviral P(Per2)-dLuc reporter vectors and a CMV-EGFP construct. Only the region for integration into the host cell genome is shown. In both reporter constructs, the transcription of dLuc is under direct control of the Per2 promoter. In the pLV7-Bsd-P(Per2)-dLuc vector (recombination-based cloning), a coexpressed Blasticidin resistance gene (Bsd) facilitates selection of infected cells. In the pLV156-P(Per2)-dLuc vector (ligation-based cloning), EGFP translation is mediated by an internal ribosome entry site (IRES) downstream of dLuc, allowing for visual observation and FACS sorting of infected cells. In addition, an SV40 promoter/terminator (P/T) is used as an insulator (see discussion 1.3). In the CMV-EGFP control vector, EGFP expression is under control of a strong CMV promoter. (B) Fluorescent images of transfected and infected GFP-expressing cells. Typically, we achieve high efficiency both in transient transfection of 293T cells and in lentiviral infection of cell lines of our interest, as indicated by GFP expression in these cells. Click here to view larger figure.
Figure 2. Phase-specific expression of bioluminescence reporters in 3T3 cells. The lentiviral reporter vectors used in this experiment are pLV7-Bsd-P(Per2)-dLuc, P(Cry1)-dLuc, P(Cry1)-Intron-dLuc, and P(Bmal1)-dLuc. Each reporter exhibits a distinct phase of oscillation, as indicated by the arrows. While the Per2 and Cry1 promoters drive peak bioluminescence at morning-day phases and the Bmal1 promoter at night phase, combinatorial regulation by the P(Cry1)-Intron harboring E-box, D-box, and RRE elements confers evening phase of peak bioluminescence. Click here to view larger figure.
Figure 3. Genetic and pharmacological perturbation of circadian bioluminescence rhythms in reporter cells. (A) Effects of Cry1 and Cry2 knockdown by siRNAs on cellular rhythms of U2OS reporter cells. The LumiCycle luminometer was used for bioluminescence recording of cells in 35 mm dishes. Figure is adapted from Reference #10, with permission from Elsevier (2009). (B) Effects of Cry2 knockdown by shRNAs on cellular rhythms of 3T3 reporter cells. A pLL3.7 Gateway vector containing a U6-shRNA cassette was used for Cry2 gene knockdown. shRNA2 has a better knockdown efficiency than shRNA1 as determined by Western blot analysis (data not shown). A Synergy luminometer was used for bioluminescence recording of cells in a 96-well plate. The settings for recording are as follows: incubator temperature, 33 °C; integration time, 15 sec; interval time, 30 min. (C) Effects of small molecule inhibitors on cellular rhythms of U2OS reporter cells. Chir99021 and Roscovitine are inhibitors directed against GSK-3 and CDK, respectively. The ViewLux system (Chir99021 assay) and a Tecan luminometer (Roscovitine assay) were used for bioluminescence recordings of cells in 384-well plates. Figure is adapted from Reference #19 (Copyright 2008 National Academy of Sciences, U.S.A.). Click here to view larger figure.
1. Modifications to Current Protocol
1.1 Recording devices and throughput considerations
Because of its commercial availability, the LumiCycle (Actimetrics) has become the most commonly used automated luminometer device for real-time recording4,5,9,19,29-31. The LumiCycle employs photomultiplier tubes (PMTs) as light detectors, which provide extremely high sensitivity and low noise14, and therefore is particularly suitable for data acquisition of extremely dim luciferase-based bioluminescence. Other similar PMT-based devices (e.g., Kronos, Atto Co.; and POLARstar Optima, BMG labtech Inc.) have also been used in many studies32,33.
The LumiCycle assay using 35-mm dishes can be adapted to 96-, 384- or even 1,152-well plate formats for high-throughput screening (HTS) experiments. HTS assay development mainly focuses on the following two aspects: 1) better reporter cells with brighter luciferase and/or higher reporter expression (see below), and 2) highly sensitive recording devices that accommodate multi-well plates. We and others have used several microplate readers such as Synergy SL2 (BioTek), Infinite M200 (Tecan)19, TopCount (PerkinElmer)28, and EnVision (Perkin Elmer)34. In our lab, both the LumiCycle and Synergy devices are placed in a temperature-controlled and light-tight incubator. Alternatively, if resources permit, the recording units can be placed in a temperature-controlled room. For HTS, critical settings such as integration/exposure time and interval between time points must be determined empirically for a given reporter and recording device.
The LumiCycle assay does not provide spatial information at single cell resolution. Yamaguchi et al.35 and Welsh et al.4,5 pioneered real-time bioluminescence imaging by integrating a specially designed microscope with a highly sensitive, low-noise CCD camera15 to achieve single cell resolution. In recent years, more sensitive imaging systems have become commercially available, including LV200 (Olympus) and CellGraph (AB3000-B, Atto) with an ultrasensitive cooled and back-illuminated electron multiplying CCD and multicolor filters36,37. Imaging has also been used in more sophisticated HTS experiments, e.g., a ViewLux CCD imager (Perkin-Elmer) combined with the GNF automated robotic system (GNF Systems). This system enabled large-scale screens for novel genes and small molecules that affect clock function10,12,19.
1.2 Luciferase reporters
Luciferases were first used in real-time luminescence recording of circadian gene expression in the early 1990s in plants and cyanobacteria, and have been commonly used in the mammalian system since 200029,35,38-40 (also see reviews14,15). Luciferase reporters are generally less toxic and more sensitive than fluorescent reporters such as GFP, due to much lower background41. The most commonly used bioluminescent reporter is firefly (Photinus pyralis) luciferase (Luc+) constructed in the pGL3 vector series (Promega), in which the coding region of the native Luc was modified for optimized transcription and translation. The combination of firefly luciferase and its highly stable and cell-permeable substrate, D-luciferin, is ideal for long-term recording. dLuc is a modified version of Luc+ with a PEST sequence fused at its C-terminus to allow for rapid protein degradation. A further improved version, Luc2, is available in the pGL4 vector series (Promega) with higher and less anomalous expression. However, we did not observe a significant difference in performance between Luc+ and Luc2 in transiently transfected 3T3 and U2OS cells (data not shown). In a recent report, Brazilian click beetle luciferase (ELuc) was shown to exhibit a much brighter signal than Luc+, suitable for single cell imaging42.
Many recent studies have taken advantage of the mPER2::LUC fusion knock-in reporter mouse4,5,9,19,25,29-31,43. This reporter system allows circadian phenotyping of cells and tissue explants including the SCN. In our previous studies, we crossed this reporter mouse line with many of the behaviorally characterized clock gene knockouts and examined the dynamics of bioluminescence rhythms in SCN, liver and lung explants cultured ex vivo, and in dissociated SCN neurons and fibroblasts cultured in vitro4,5,9,31. These studies allowed us to gain important insights into the molecular details of clock operation at the cell and organismal levels.
1.3 Gene delivery methods
Vector-based clock reporters provide great versatility, as they can be introduced into various cell types via transient transfection11,13,16,17 or lentiviral transduction5,9,18. Although cumbersome and less reproducible, transiently transfected cells can also be grown in the continuous presence of antibiotics to generate stable cell lines. Lentiviral vectors are still preferred because of their stable integration into the cell's genome, as well as greater efficiency and versatility. Besides the pLV7 vector presented above in which infected cells can be selected with antibiotics, we have used other vectors. For example, the pLV156-P(Per2)-dLuc vector contains a P(Per2)-dLuc-IRES-EGFP expression cassette in which EGFP translation is mediated by an internal ribosome entry site (IRES) downstream of dLuc, allowing for visual observation and fluorescence-activated cell sorting (FACS) of cells with integrated vectors5 (Figure 1) (see below). To shield the reporter from integration site effects, a constitutive promoter and terminator (P/T) can be introduced as an insulator or decoy immediately upstream of the P(Per2)-dLuc reporter expression cassette (for example, bottom construct in Figure 1A), as reported by Brown et al.18 and Liu et al.5
Previous studies established the groundwork for exploration of tissue-specific genetic perturbation both in vivo and/or ex vivo44-48. For example, viral particles can be injected into the SCN region followed by SCN slice preparation. As an alternative to this in vivo followed by ex vivo assay, SCN slices can be dissected first and cultured ex vivo, followed by incubation with high-titer virus. However, due to low infection efficiency of relatively thick tissue slices, a highly consistent ex vivo protocol remains to be developed.
1.4 Choice of cell lines and recording conditions
Much of what we know about the biochemistry and cell biology of the clock mechanism is based on three cellular models: 3T3 mouse fibroblasts16,17, U2OS human osteosarcoma cells10,11,19,28, and mouse fibroblasts derived from mice9,13,34. The lentiviral vector system permits efficient delivery and stable integration into the host genome of both dividing and non-dividing mammalian cells, and therefore is not limited to certain cell types as in transient transfection. Given the roles of circadian clocks in the regulation of a wide range of cellular and physiological responses, future studies will certainly extend to a wide variety of cell types, either commercially available cell lines or primary cells derived from mice. These cell-autonomous clock studies will help uncover tissue-ubiquitous, as well as tissue-specific, properties of circadian clocks.
It is worth noting that generating reporter cells may require empirical testing of different reporters (e.g., various vector and promoter types), and sometimes further optimization. Not all cell types have cell-autonomous rhythms. To improve cell health and persistence of cellular rhythms, optimization of recording medium is often necessary. For example, as an alternative to the 3T3 recording medium, the recording medium used for U2OS cells is as follows: 1x DMEM (Invitrogen), 2% B-27, 1 μM forskolin, 1 mM luciferin, 14.5 mM NaHCO3, 10 mM HEPES (pH 7.2). 1 mM luciferin is usually sufficient, but a final concentration (0.1-1 mM) may be determined for each cell/reporter type. Forskolin may not be necessary and its concentration may be empirically determined for each cell type.
1.5 Transfection efficiency in lentiviral prep
High transfection efficiency of 293T cells is critical for a good viral prep. Major considerations include health of 293T cells and choice of a transfection reagent. 293T cells can be very finicky and require careful handling. Thus, cell growth rate and morphology must be carefully monitored. For consistency, we recommend that serum be tested first and the same tested batch used thereafter for maintaining the cells. We always test new 2x BBS and CaCl2 stock solutions and optimize working concentration of CaCl2 around 0.25 M. Cells of high passage number and/or displaying slow growth rate should not be used. 293T cells are readily transfectable with a number of transfection reagents. In addition to CaPO4, we also achieved excellent transfection efficiency with Fugene 6 (Roche or Promega) or Lipofectamine-2000 (Invitrogen). These lipid-based transfection (lipofection) reagents are more expensive, but give better transfection consistency than CaPO4. For large-scale lentiviral preps, we recommend using CaPO4 for transfection because of lower cost.
1.6 Generation of clonal cell lines
Un-concentrated crude viral particles are of relatively low titer (106 viral particles/ml), and reporter expression in transduced cells is low. Although this is sufficient for most applications such as LumiCycle recording, brighter reporters are often needed, e.g. in high-throughput assays using a less sensitive recording device. Reporter expression can be increased by using higher titer viral preps. To do this, we recommend using larger culture vessels such as15 cm dishes for transfection, and collecting media twice at day 4 and day 5. For 10x concentration (107 viral particles/ml), perform filtration using centrifugal filter devices (Millipore). For 100x or more concentration (≥108 viral particles/ml), perform ultracentrifugation as previously described20. It is noted that, while viral titer may affect luminescence signal amplitude, it does not alter key circadian parameters such as period length5,18. If a homogenous cell population is desired, clonal cell lines may be obtained by FACS-based single cell sorting in 96-well plates. However, it is imperative that these clonal cells be carefully examined; cell morphology should be indistinguishable from parental cells, and clock properties such as period length and amplitude should be representative of the infected cell population.
2. Application of Cell-autonomous Reporter Clock Models
Unlike tissue or animal models, cell-based models are amenable to genetic and pharmacologic perturbations, and when necessary, also use in HTS formats. Perturbation of gene function can be achieved by over-expression or RNAi-mediated knockdown. Selective small molecules can be used to interfere with protein function. Here we briefly discuss some uses of cell-autonomous clock models in circadian studies.
2.1 Study of core clock mechanisms
Cell-autonomous clock models have greatly facilitated mechanistic studies. Hogenesch and colleagues used the 3T3 model to show that feedback repression by CRYs is required for clock function16, and the U2OS model to probe the system-level properties of clock function11. We employed the Cry1-/-:Cry2-/- fibroblast model derived from mice to show that delayed feedback repression is necessary for clock function13, and the Bmal1-/- fibroblast model to show that Rev-erbα and β play redundant roles in regulating RRE-mediated rhythmic gene expression, and that Bmal1 rhythm is not critically required for core clock function9. Furthermore, chemical screening in reporter cells allowed identification and/or clarification of the functions of GSK-3β (period shortening when perturbed)19, CK1σ and ε (period lengthening when perturbed)49, and CK1α12. As discussed below, these models facilitate identification of additional clock components. Strategically, these cellular models are more tractable for rapid discovery of basic mechanisms, which then provide new entry points for in vivo validation and exploration.
2.2 Identification of clock factors
In addition to the core feedback loop, the clock mechanism also integrates diverse signaling and regulatory factors. For identification of additional clock components and modifiers, cell-autonomous clock models are advantageous over genetic screening in mice due to the inherent lethality, pleiotropic effects, and developmental genetic compensation associated with mutant animal models50. Cell-based screens, in combination with functional genomics approaches, have been effectively carried out using high-throughput recording systems to screen genome-wide siRNA and shRNA libraries for identification of novel clock factors10,28. Similarly, one can employ chemical biological approaches and screen for diverse small molecules to study their effects on clock function12,19,34,49. Although major clock genes and their functions have been identified, these screens have identified additional components or modifiers that are involved in regulation or modulation of the clock. Many of these modifiers represent particular modalities of integrating signal transduction of synchronous or asynchronous cues (clock inputs).
2.3 Study of clock outputs
Although virtually all cells in vivo have circadian clocks, cells cultured in vitro do not necessarily display overt circadian rhythms. In this context, the reporter approach provides a useful tool to test whether the clock mechanism exists in a particular cell type. Presence of cell-autonomous clock mechanisms warrant studies of clock outputs in these cells, as well as in tissues in vivo, including the transcriptome by microarray or RNA sequencing51,52, transcriptional regulatory networks, the proteome, and the metabolome. These studies examine genome-wide RNA and protein expression patterns and thus would complement cell-based luminescence reporter assays that do not necessarily reflect the endogenous expression.
3. Significance of Cell-autonomous Clock Models
Coordination of circadian function within the SCN and across different cell and tissue types is a critical aspect of normal physiology30. The central SCN and peripheral clocks are all essential parts of the overall circadian organization. The SCN receives light input directly from the retina to entrain to the light/dark cycle, and serves as a coordinator to synchronize peripheral clocks through both neuronal and hormonal connections. In addition to internal synchronization, the molecular clockwork within peripheral tissues and cells also orchestrates a myriad of transcriptional output networks, which ultimately govern overt circadian rhythms in physiology and behavior.
Thus, both systemic and local signals regulate circadian physiology, and there is a need to uncover circadian genes directly regulated by cell-autonomous clocks vs. those indirectly regulated by systemic signals emanating from the SCN. The role of peripheral clocks can be studied in cell-autonomous clock models, in which systemic influences are removed. In the case of the liver, tissue-specific regulatory networks generate rhythms in local physiology51,53. By comparing in vitro and in vivo circadian rhythms, we will be able to discriminate between cell-autonomous and systemic cue-driven clock functions. Indeed, recent studies have begun to reveal a significant role for the liver clock in hepatic gene expression6,51,52. Future research in the field warrants development of various tissue- and cell type-specific clock models representing different physiological outputs.
Previous studies of the biochemistry and cell biology of the clock mechanism have largely relied on a limited number of cell and tissue types, including particularly 3T3, U2OS, mouse fibroblasts, and liver. An implicit assumption in most circadian studies is that the clock works the same way in all cell and tissue types, and gene function determined in one cell or tissue type is generally considered to apply universally in all cells7. However, by establishing and characterizing more cell-autonomous clock models, future studies are expected to uncover tissue specific clock properties underlying local circadian biology.
No conflicts of interest declared.
This work was supported in part by the National Science Foundation (IOS-0920417) (ACL).
|DMEM||HyClone||SH30243FS||For regular cell growth|
|All other chemicals||Sigma|
|Tissue culture incubator||5% CO2 at 37°C|
|Tissue culture hood||BSL-2 certified|
|Light & fluorescent microscope||Phase contrast optional|