A modular single-construct tool for optical control of cell signaling

The study of cell signaling has revealed that cells can be particularly sensitive to spatiotemporal properties of the signaling dynamics. For instance, a cell can choose divergent fates (proliferation vs. differentiation, senescence vs. apoptosis, etc.) depending on the duration and strength of a protein signal. The development of fluorescent reporters has enabled us to observe these dynamics, but researchers still lack appropriate, general tools to perturb cells on the second- and minute-timescales on which signaling dynamics can occur.

Coupling light to protein activation may offer a powerful solution due to its fast, reversible, and highly tunable nature. In the March 2013 issue of Nature Methods, Bugaj, Kane, Schaffer and colleagues accomplish this by developing a method for light inducible homo-oligomerization of proteins. In this study, the authors report a genetically encodable, generalizable system that can activate multiple signaling proteins and networks in mammalian cells.

To do this, the authors leveraged the interesting ability of the Arabidopsis protein Cryptochrome 2 (Cry2) to form oligomers in response to blue light. To date, this clustering ability had only been shown in plant cells, but upon transfection into HEK 293Ts, the authors demonstrated that Cry2 can be clustered in mammalian cells as well (fig. 1). Since numerous signaling proteins are regulated through oligomerization, the authors hypothesized that Cry2 clustering could be used as an optogenetic method to regulate protein activity. They first demonstrated this by targeting the canonical Wnt/β-catenin pathway through clustering of the LRP6 receptor endodomain fused to Cry2-mCherry. Upon light induction, this fusion induced pathway activation in both 293Ts and neural stem cells through a large dynamic range, with a maximal response exceeding that achievable using the natural pathway agonist Wnt3a.

Figure 1. (a) Schematic depicting light-induced protein clustering. hν, light activation energy as given by the Planck relation. (b) Cry2-mCh cluster formation (top) in HEK 293T cells in response to 488-nm laser light, as well as dissociation (bottom) after light withdrawal. Scale bars (b–d), 20 μm. (c) Clustering does not occur with the light-insensitive mutant Cry2(D387A)-mCh. (d) Kymograph of mCherry fluorescence corresponding to the dashed line in b. (e) Measurement of cluster number over multiple light-dark cycles demonstrates repeatable, rapid clustering and declustering with consistent kinetics. A single exponential decay fit of cluster number allows measurement of cluster decay constant τ (mean ± 1 s.d., n = 3 decay cycles). (f) Representative plot of single-cell cluster formation over time. Similar behavior was observed in all cells for which cluster formation was measured. T1/2, time at which a half-maximal number of visible clusters was detected. (g) Representative plot of concentration-normalized T1/2 increase with decreased illuminating intensity (Online Methods). Plot shows mean ± 1 s.e.m., n = 4 cells for each condition.

Figure 1. (a) Schematic depicting light-induced protein clustering. hν, light activation energy as given by the Planck relation. (b) Cry2-mCh cluster formation (top) in HEK 293T cells in response to 488-nm laser light, as well as dissociation (bottom) after light withdrawal. Scale bars (b–d), 20 μm. (c) Clustering does not occur with the light-insensitive mutant Cry2(D387A)-mCh. (d) Kymograph of mCherry fluorescence corresponding to the dashed line in b. (e) Measurement of cluster number over multiple light-dark cycles demonstrates repeatable, rapid clustering and declustering with consistent kinetics. A single exponential decay fit of cluster number allows measurement of cluster decay constant τ (mean ± 1 s.d., n = 3 decay cycles). (f) Representative plot of single-cell cluster formation over time. Similar behavior was observed in all cells for which cluster formation was measured. T1/2, time at which a half-maximal number of visible clusters was detected. (g) Representative plot of concentration-normalized T1/2 increase with decreased illuminating intensity (Online Methods). Plot shows mean ± 1 s.e.m., n = 4 cells for each condition. From Bugaj et al., 2013 with permission from Macmillan Publishers Ltd (license 3102400399130).

Next, the authors aimed to show the modularity of this approach by targeting Rho GTPase signaling. By fusing Rac1 to Cry2-mCherry, light illumination induced rapid translocation of the Cry2-mCh-Rac1 protein to the membrane, indicating Rac1 activation and supporting previous in vitro data suggesting Rac1 sensitivity to oligomerization (fig. 2). Furthermore, the authors then showed that Rho GTPase family member RhoA is also sensitive to oligomerization and can be activated with light through clustering of a Cry2-mCh-RhoA fusion (fig. 2). This result was surprising, as clustering had not been recognized as a mode of RhoA activation.

Figure 2. (a) Cry2-Rac1 translocates to the cell membrane (left), and Cry2-RhoA translocates to the cell membrane and vesicles (right) upon illumination of HEK 293T cells. Scale bars, 25 μm. (b) Single-cell illumination of fibroblasts expressing Cry2-RhoA induces visible membrane retraction within minutes. Image shown at 8.5 min of activation; cell outline represents starting cell morphology, and circles depict illumination region. Scale bars, 20 μm. (c) ELISA analysis shows that cells expressing Cry2-RhoA have increased levels of GTP-bound (active) RhoA under blue light as compared to unilluminated controls (mean ± 1 s.d., n = 3 replicates, *P = 0.0495). OD, optical density. (d) Cry2-RhoA photoactivation induces enhanced stress-fiber formation in fibroblasts as visualized by the actin-binding Alexa Fluor 488–phalloidin conjugate. Scale bar, 50 μm. (e) Whole-field light activation induces contractility in fibroblasts expressing Cry2-RhoA. Inhibition with Rho-associated kinase inhibitor Y-27632 (ROCKi), but not myosin light chain kinase inhibitor ML-7 (MLCKi), attenuates this effect, thus demonstrating dependence of light-induced membrane retraction on RhoA pathway activation and independence from MLCK activity. Graph shows means ± 1 s.d., n = 6–10 fields, ***P < 0.001.

Figure 2. (a) Cry2-Rac1 translocates to the cell membrane (left), and Cry2-RhoA translocates to the cell membrane and vesicles (right) upon illumination of HEK 293T cells. Scale bars, 25 μm. (b) Single-cell illumination of fibroblasts expressing Cry2-RhoA induces visible membrane retraction within minutes. Image shown at 8.5 min of activation; cell outline represents starting cell morphology, and circles depict illumination region. Scale bars, 20 μm. (c) ELISA analysis shows that cells expressing Cry2-RhoA have increased levels of GTP-bound (active) RhoA under blue light as compared to unilluminated controls (mean ± 1 s.d., n = 3 replicates, *P = 0.0495). OD, optical density. (d) Cry2-RhoA photoactivation induces enhanced stress-fiber formation in fibroblasts as visualized by the actin-binding Alexa Fluor 488–phalloidin conjugate. Scale bar, 50 μm. (e) Whole-field light activation induces contractility in fibroblasts expressing Cry2-RhoA. Inhibition with Rho-associated kinase inhibitor Y-27632 (ROCKi), but not myosin light chain kinase inhibitor ML-7 (MLCKi), attenuates this effect, thus demonstrating dependence of light-induced membrane retraction on RhoA pathway activation and independence from MLCK activity. Graph shows means ± 1 s.d., n = 6–10 fields, ***P < 0.001. From Bugaj et al., 2013 with permission from Macmillan Publishers Ltd (license 3102400399130).

The Cry2 clustering system possesses several attractive features compared to previously published technologies to optically control protein signaling. Cry2 is entirely genetically encodable and does not require the addition of an exogenous cofactor, enabling its future use in live animal studies. Clustering on and off rates can be controlled on the order of seconds to minutes, which is sufficiently fast to regulate cell signaling processes. Furthermore, the entire photoactivatible module can be encoded on one peptide chain, eliminating stoichiometric considerations inherent with multi-component systems and likely minimizing the transcriptional load placed on the host cell. Most promisingly, clustering fusion proteins via Cry2 is highly modular with no additional protein engineering needed for each application, suggesting that this method could be broadly applicable to endow numerous other signaling pathways with the benefits of light control.

Link to original paper.

This entry was posted in Journal Club and tagged , . Bookmark the permalink.
Add Comment Register



Leave a Reply