Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

LOVTRAP: an optogenetic system for photoinduced protein dissociation


LOVTRAP is an optogenetic approach for reversible light-induced protein dissociation using protein A fragments that bind to the LOV domain only in the dark, with tunable kinetics and a >150-fold change in the dissociation constant (Kd). By reversibly sequestering proteins at mitochondria, we precisely modulated the proteins' access to the cell edge, demonstrating a naturally occurring 3-mHz cell-edge oscillation driven by interactions of Vav2, Rac1, and PI3K proteins.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Generation of Zdk and design of LOVTRAP.
Figure 2: Protein control by LOVTRAP.
Figure 3: Resonance in HeLa cell protrusion–retraction cycles upon acute release of Vav2.

Accession codes

Primary accessions


Protein Data Bank

Referenced accessions

Protein Data Bank


  1. 1

    Weitzman, M. & Hahn, K.M. Curr. Opin. Cell Biol. 30, 112–120 (2014).

    CAS  Article  Google Scholar 

  2. 2

    Nord, K. et al. Nat. Biotechnol. 15, 772–777 (1997).

    CAS  Article  Google Scholar 

  3. 3

    Kay, C.W. et al. J. Biol. Chem. 278, 10973–10982 (2003).

    CAS  Article  Google Scholar 

  4. 4

    Harper, S.M., Christie, J.M. & Gardner, K.H. Biochemistry 43, 16184–16192 (2004).

    CAS  Article  Google Scholar 

  5. 5

    Harper, S.M., Neil, L.C. & Gardner, K.H. Science 301, 1541–1544 (2003).

    CAS  Article  Google Scholar 

  6. 6

    Kanaji, S. et al. J. Cell Biol. 151, 277–288 (2000).

    CAS  Article  Google Scholar 

  7. 7

    Zoltowski, B.D., Vaccaro, B. & Crane, B.R. Nat. Chem. Biol. 5, 827–834 (2009).

    CAS  Article  Google Scholar 

  8. 8

    Kawano, F. et al. PLoS ONE 8, e82693 (2013).

    Article  Google Scholar 

  9. 9

    Levskaya, A. et al. Nature 461, 997–1001 (2009).

    CAS  Article  Google Scholar 

  10. 10

    Krause, M. & Gautreau, A. Nat. Rev. Mol. Cell Biol. 15, 577–590 (2014).

    CAS  Article  Google Scholar 

  11. 11

    Vicente-Manzanares, M., Webb, D.J. & Horwitz, A.R. J. Cell Sci. 118, 4917–4919 (2005).

    CAS  Article  Google Scholar 

  12. 12

    Machacek, M. & Danuser, G. Biophys. J. 90, 1439–1452 (2006).

    CAS  Article  Google Scholar 

  13. 13

    Abe, K. et al. J. Biol. Chem. 275, 10141–10149 (2000).

    CAS  Article  Google Scholar 

  14. 14

    Huang, N.E. & Shen, S.S. (eds.) Hilbert-Huang Transform and Its Applications 2nd edn. (World Scientific, 2014).

  15. 15

    Yang, H.W. et al. Mol. Cell 47, 281–290 (2012).

    CAS  Article  Google Scholar 

  16. 16

    Han, J. et al. Science 279, 558–560 (1998).

    CAS  Article  Google Scholar 

  17. 17

    Chen, H.C. & Guan, J.L. Proc. Natl. Acad. Sci. USA 91, 10148–10152 (1994).

    CAS  Article  Google Scholar 

  18. 18

    Weiner, O.D. et al. Nat. Cell Biol. 4, 509–513 (2002).

    CAS  Article  Google Scholar 

  19. 19

    Booden, M.A., Campbell, S.L. & Der, C.J. Mol. Cell. Biol. 22, 2487–2497 (2002).

    CAS  Article  Google Scholar 

  20. 20

    Cotton, S.W., Zou, J., Valencia, C.A. & Liu, R. Nat. Protoc. 6, 1163–1182 (2011).

    Article  Google Scholar 

  21. 21

    Kabsch, W. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132 (2010).

    CAS  Article  Google Scholar 

  22. 22

    McCoy, A.J. et al. J. Appl. Cryst. 40, 658–674 (2007).

    CAS  Article  Google Scholar 

  23. 23

    Arnold, K. et al. Bioinformatics 22, 195–201 (2006).

    CAS  Article  Google Scholar 

  24. 24

    Emsley, P. et al. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010).

    CAS  Article  Google Scholar 

  25. 25

    Afonine, P.V. et al. Acta Crystallogr. D Biol. Crystallogr. 68, 352–367 (2012).

    CAS  Article  Google Scholar 

  26. 26

    Adams, P.D. et al. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221 (2010).

    CAS  Article  Google Scholar 

  27. 27

    Seidelin, J.B. et al. Am. J. Physiol. Gastrointest. Liver Physiol. 308, G92–G99 (2015).

    CAS  Article  Google Scholar 

  28. 28

    Lorenz, H., Hailey, D.W. & Lippincott-Schwartz, J. Nat. Methods 3, 205–210 (2006).

    CAS  Article  Google Scholar 

  29. 29

    Risse, P.A. et al. Am. J. Physiol. Lung Cell. Mol. Physiol. 300, L958–L966 (2011).

    CAS  Article  Google Scholar 

  30. 30

    Jia, L. et al. Mol. Biol. Cell 22, 4227–4235 (2011).

    CAS  Article  Google Scholar 

  31. 31

    Mukhopadhyay, P. et al. Nat. Protoc. 2, 2295–2301 (2007).

    CAS  Article  Google Scholar 

  32. 32

    Vilela, M. et al. Methods Enzymol. 519, 253–276 (2013).

    CAS  Article  Google Scholar 

  33. 33

    Zoubir, A.M. & Iskander, D.R. Bootstrap Techniques for Signal Processing (Cambridge University Press, 2004).

Download references


This work was funded by NIH grants R01-GM090317 (K.M.H. and G.D.), P01-GM103723 (K.M.H. and G.D.), R01-DA036877 (K.M.H. and B.K.), and R01-CA157738 (R.L.) and by the Max Planck Society, German Research Foundation DFG, FOR 1279 (I.S.). H.W. is a recipient of an Arthritis Foundation Postdoctoral Fellowship. We thank E. Hartmann for crystallizing Zdk–LOV complexes, R. Littlefield (UNC Chapel Hill Department of Biochemistry and Biophysics) for providing LOV2 peptides, and E. Trudeau for cloning help. Diffraction data were collected at the Swiss Light Source, beamline X10SA, of the Paul Scherrer Institute, Villigen, Switzerland. We thank the Heidelberg data collection team and the PXII staff for their support in setting up the beamline and C. Roome for expert support of the crystallographic software. The UNC Flow Cytometry Core Facility is supported in part by P30 CA016086 Cancer Center Core Support Grant to the UNC Lineberger Comprehensive Cancer Center.

Author information




H.W. performed the screening experiments, engineered proteins, and performed imaging studies. M.V. performed analysis of oscillating signaling behavior. A.W., M.T., and I.S. carried out crystallography. H.W., H.Y., and B.K. purified proteins for crystallography. R.L., G.D., and K.M.H. directed the work and carried out final edits for the paper, which was written using contributions from all authors.

Corresponding authors

Correspondence to Rihe Liu, Gaudenz Danuser or Klaus M Hahn.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Two strategies for application of LOVTRAP.

The LOV domain is never entirely open or closed, but rather is in equilibria that favor the open or closed forms in the light and dark, respectively. As shown by Yao et al.1, 1.6% of LOV2 molecules are in the open conformation in the dark and about 9% are in the closed conformation in the light. Based on this, the configuration with LOV2 on mitochondria (Strategy I) produces essentially no free protein in the dark, because LOV can be kept in excess over Zdk-POI. This offers an advantage for proteins that need to be tightly controlled, such as lethal proteins or proteins where a small amount of free material in the dark produces undesired biological effects. In Strategy I, the amount of protein released upon irradiation will be highly dependent on the ratio of LOV2:Zdk. With a ratio of [LOV2]/[Zdk1]=10:1, about 20% of the POI can be released with light. For Strategy II, about 90% of the protein will be released by light, regardless of the ratio Zdk:LOV2. However, even in the dark there will be a small amount of unbound LOV-POI in the cytosol. When this strategy is used with Zdk1, LOV2 can only be fused to the C-terminus of the POI, because the C-terminus of LOV2 needs to be kept free for Zdk1 binding.

Supplementary Figure 2 Crystal structures of LOV2:Zdk complexes.

Overall structures and close-up views of interactions between the C-terminal part of the Jα helix (in purple) and Zdk proteins. The residues involved in contacts are shown in ball-and-stick representation and labelled. Dashed black lines represent hydrogen bonds. Zdk1 binds to residues of the LOV globular domain and also requires an interaction with the C-terminus of the LOV2 Jα helix for tight binding. Zdk2 and Zdk3 bind similarly and have similar sequences in the variable region, but they do not require interaction with the LOV2 C-terminus. They bind at a site near that of Zdk1, but with a different orientation.

Supplementary Figure 3 Effects of C-terminal Jα modification on LOV2-Zdk1 binding.

Effects of LOV2 Jα helix mutations on binding. 1 = Binding tested by radiometric binding assay; 2 = Binding tested using LOVTRAP mitochondrial localization in living cells; 3 = Binding tested using Bio-Layer Interferometry assay. Modifications on the end of the Jα helix abolished the binding. Green: LOV2 core domain; Blue: Jα helix; Gray: Zdk1.

Supplementary Figure 4 Activation kinetics of LOVTRAP with different Zdk variants.

Zdk1, 2 or 3 was attached to the outer membrane of mitochondria via a TOM20 fragment, and mCherry was attached to LOV2 (for Zdk1, mCherry was attached to N-terminus of LOV2, while for Zdk2 and 3, mCherry was attached to both the N- and C- termini). When Zdk1 was used, substantial fluorescent protein was released after 100 ms of blue light irradiation, and equilibrium was reached within 500 ms irradiation. However, for the systems using Zdk2 or 3, release was slower and only partial release was achieved (b), likely due to the non-negligible affinity of these Zdk variants for the lit-state of LOV2. When using Zdk1, proteins need to be attached to N-terminus of LOV2, but for Zdk2 or 3, either terminus of LOV2 can be used. Scale bar: 10µm.

Supplementary Figure 5 Effectiveness of mutations that mimic the dark-state of LOV2.

Lov2 was labeled with mCherry fluorescent protein and Zdk1 was anchored at mitochondria. Fluorescence was monitored at a region of the cytosol away from mitochondria. Black: wild-type LOV2, from Figure 1; Red: LOV2 dark state mutant C450A; Blue: Because the widely used C450A dark state mutant shows some response to light, we combined the C450A mutation with other mutations that stabilize the closed form of LOV2 (C450A, L514K, G528A, L531E, andN538E). When this ‘superdark’ mutant was used in the mCherry-LOV2 construct, light no longer had an effect.

Supplementary Figure 6 Protein control by LOVTRAP.

(a) HEK293 cells with LOV2 attached to the plasma membrane and Zdk attached to mCherry (left). Fluorescence in a region in cytosol (circle) monitored before, during and after blue light irradiation (right). (b) Reversible cell protrusion produced by light-induced release of constitutively active Zdk-Vav2. (c) Cell area before, during and after light-induced release of constitutively active Vav2 (7 cells), Rac1 (7 cells), RhoA (4 cells) and Control (5 cells). Wilcoxon rank test and permutation test were performed to the medians before and during irradiation. For both tests: p < 1e-6 (Vav2, Rac1, RhoA) and p=0.9328 (Control). (d) Edge velocity (calculated from 2,017 edge locations sampled in 7 cells) during light-induced release of constitutively active Rac1. Dark blue line indicates median, band indicates 95% confidence interval. (e) Release of constitutively active RhoA led to irreversible contraction and reversible decrease in ruffling. (f) Quantitation of Vav2 effects on ruffling. White = percent cells in which ruffling was induced, blue = percent cells with no obvious effect on ruffling. Cells were transfected either with LOVTRAP (Vav2, n=14) or with LOV and Zdk1 only (Control, n=18). (g) Inhibition of ruffling by RhoA(Q63L), graphed as in (f). Control = expression of LOV2 and Zdk1 only (n= 9), RhoA = LOVTRAP (n= 18). All scale bars 10µm.

Supplementary Figure 7 Cell edge morphodynamic changes upon construct activation.

Normalized velocities over time (for all three states). Each velocity profile was normalized by its respective median value for the dark state. Center dark colored line indicates median, band indicates 95% confidence interval about median, calculated from n edge locations sampled in m cells. Vav2 n=2,606, m=7; Rac1 n = 2,017, m=7; RhoA n=1,637, m=4; Control n=989, m=5.

Supplementary Figure 8 The effect of LOVTRAP on cell edge velocity.

The effects of Rac1 LOVTRAP (a) and Vav2 LOVTRAP (b) on cell edge velocity (normalized using the basal level before photo-activation). No dependence was observed between cell edge velocity and the expression level of the proteins.

Supplementary Figure 9 Effects of LOVTRAP on expression of endogenous Rac1.

Endogenous Rac1 expression is shown in the presense of LOVTRAP Vav2, RhoA, and Rac1. Control cells were transfected with LOV2 and Zdk1 only and non-transfected cells were not transfected with any DNA.

Supplementary Figure 10 Effect of LOVTRAP on protrusion speeds; Basal line analysis for LOVTRAP Vav2, Rac1 and RhoA.

Median cell edge velocity distributions for cells expressing LOVTRAP Vav2, Rac1, RhoA or Control (Zdk1/LOV2 only) calculated from n edge locations sampled in m cells. The blue box represents the 25th to 75th percentile and the center red line represents the median. The whiskers cover 99.3 of the data range and the red points represent outliers. Vav2 n=38,133, m=7; Rac1 n=21,157, m=4; RhoA n=27,213, m=7; Control n=29,148 m=7.

We performed the Kruskal-Wallis test to exam the null hypothesis that the all points come from the same distribution.

With a p-value of 0.2961, the test was unable to reject the null hypothesis at a significant level of 29.61%, showing that the cells were indistinguishable from control cells.

Supplementary Figure 11 Effects of LOVTRAP on mitochondrial superoxide generation.

Mitochondrial superoxide generation is shown in the presense of LOVTRAP Vav2, LOVTRAP RhoA, and LOVTRAP Rac1 in the dark. Control cells were transfected with LOV2 and Zdk1 only and non-transfected cells were not transfected with any DNA. Mitochondrial superoxide generation was measured using MitoSOX Red in flow cytometry2. In the dark, expression of LOVTRAP Vav2, Rac1 or RhoA had little effect on mitochondrial superoxide generation.

Supplementary Figure 12 Effects of LOVTRAP on mitochondrial membrane potential.

Mitochondrial membrane potential is shown in the presence of LOVTRAP Vav2, LOVTRAP RhoA, and LOVTRAP Rac1 in the dark. Control cells were transfected with LOV2 and Zdk1 only and non-transfected cells were not transfected with any DNA. Expression of LOVTRAP Vav2, Rac1 or RhoA had little effect on mitochondrial membrane potential. Membrane potential was measured using DilC1(5) Red using flow cytometry.3

Supplementary Figure 13 Time-frequency analysis with Hilbert-Huang decomposition algorithm.

Upper box: Hilbert-Huang Transform (HHT) algorithm. Left panel: Synthetic test signal generated as a mixture of noise and several sinusoidal waves with distinct frequencies at different times [1Hz for all times, 3Hz from 0-6 sec and 12-18 sec, 5Hz and 10Hz from 6-12 sec]. The 1Hz and 10Hz waves have twice the amplitude of the remaining ones. Middle panel: Decomposition of the test signal by the EMD algorithm into 8 intrinsic modes. Right panel: Spectrogram of the test signal after the numerical computation of the Hilbert transform of each mode and its respective analytical signal. Workflow for the time-frequency analysis of the cell edge motion. Cell edge is divided into windows, each tracked over time and velocities calculated. The HHT algorithm is then applied to the velocity profiles, each generating a spectrogram. An average spectrogram representative of the whole cell is calculated by bootstrapping the mean amplitude for each time-frequency pair from all individual spectrograms.

Supplementary Figure 14 Time-Frequency analysis of cell edge motion.

(a) Left, representative kymograph of cell edge velocity during an experiment with release of Vav2 between 1800 s and 3600 s in pulses of 50 s blue light alternating with 250 s of dark (equals 3.3 mHz blue light pulses). Above kymograph, distribution of edge velocity along cell periphery for each time point (blue line, median velocity; red band, 95% confidence about median). Right, spectrogram derived from velocities in left panel. Right, power density as a function of temporal frequency, collected during pulsatile release of Vav2. Center line indicates median density, band indicates 95% confidence interval about median, calculated from n = 1686 edge locations sampled in m = 5 cells. (b) Representative spectrogram of cells stimulated with 3.3mHz, 4mHz, 6.7mHz and 10mHz blue light pulses. Each panel contains spectral data from n = 2044 edge locations. (c) Representative spectrogram of cells with acutely stimulated activation of constitutively active Rac1Q61L (data from n = 357 edge locations); endogenous Rac1 via release of TIAM-1 DH/PH domain (data from n = 1546 edge locations); of constitutively active VAV2 in the presence of the PI3K inhibitor LY294002 (data from n = 2350 edge locations); and of VAV2 K401A mutant with impaired binding to PI3K-generated phospholipid products (data from n = 1439 edge locations).

Supplementary Figure 15 Hypothesis re mechano-chemical signaling pathways driving cell edge oscillations.

Vav2 is upstream of proteins controlling assembly and contraction that drive cell protrusion and retraction, and downstream of a mechanochemical feedback response modulating the local concentration of Vav2 signals at the cell edge.

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–15, Supplementary Tables 1–6 and Supplementary Note (PDF 3073 kb)

Supplementary Software

Spectral analysis code (ZIP 10 kb)

The crystal structure of LOV2 bound to Zdk1.

Crystal structure of the LOV2-Zdk1 complex. Zdk1: magenta; LOV2 core domain: green; LOV2 Jα helix: blue; FMN cofactor: cyan. (MOV 15041 kb)

Release of mCherry from mitochondria using blue light.

HeLa cells were imaged without blue light for 5 min, then irradiated with blue light for 5 min (a blue dot appears), then again imaged without blue light for 6.5 min. Before irradiation, mCherry is localized on mitochondria. Blue light causes release of mCherry, leading to diffuse distribution in the cytosol. When the blue light is turned off, mCherry returns to the mitochondria. (MOV 1870 kb)

Dark mutant of LOV2 abolishes the release of mCherry from mitochondria.

HeLa cells were imaged for 5 min, then irradiated with blue light for 5 min (a blue dot appears) and then imaged without blue light for an additional 6.5 min. The dark mutant of LOV2 (LOVSD) showed only mitochondrial localization even after prolonged irradiation. (MOV 2143 kb)

Release of mCherry from plasma membrane using blue light.

HEK-293 cells were irradiated with 5 second pulses of blue light (a blue dot appears) every 100 seconds. mCherry was released from the plasma membrane during irradiation and returned to the plasma membrane when the blue light was turned off. (MOV 755 kb)

Release of Vav2 from mitochondria induced reversible increases in the velocity of cell edge oscillations.

HeLa cells were imaged for 30 min, then irradiated with blue light for 30 min (a green dot appears), and then imaged again after blue light irradiation for 30 min. Images are scaled to optimize visualization of the edge. (MOV 3023 kb)

Release of Rac1 Q61L from mitochondria induced reversible increases in the extent and velocity of cell edge ruffling.

HeLa cells were imaged without blue light illumination for 30 min, then irradiated with blue light for 30 min (a blue dot appears), and then imaged for 30 min without blue light. (MOV 5159 kb)

Release of RhoA Q63L from mitochondria induced contraction.

HeLa cells were imaged without blue light for 30 min, then irradiated with blue light for 30 min (a blue dot appears) and then imaged without irradiation for 30 min. (MOV 9032 kb)

Tracking cell edges.

Upper left, a HeLa cell with the cell edge highlighted before, during and after optogenetically stimulated VAV2 release. Upper right, traces of the cell edge overlaid at 10 sec intervals before optogenetic stimulation; Lower left, traces overlaid during VAV2 release; Lower right, traces overlaid after blue light irradiation had been halted; Warmer colors indicate later time points. Scale bar: 10 μm. (MOV 18713 kb)

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wang, H., Vilela, M., Winkler, A. et al. LOVTRAP: an optogenetic system for photoinduced protein dissociation. Nat Methods 13, 755–758 (2016).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing