Ratiometric biosensors based on dimerization-dependent fluorescent protein exchange

Abstract

We have developed a versatile new class of genetically encoded fluorescent biosensor based on reversible exchange of the heterodimeric partners of green and red dimerization-dependent fluorescent proteins. We demonstrate the use of this strategy to construct both intermolecular and intramolecular ratiometric biosensors for qualitative imaging of caspase activity, Ca2+ concentration dynamics and other second-messenger signaling activities.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Figure 1: The FPX strategy and application to imaging of protease activity.
Figure 2: Intermolecular FPX for imaging of second-messenger signaling.
Figure 3: Intramolecular FPX using tripartite single polypeptides.

Accession codes

Primary accessions

NCBI Reference Sequence

References

  1. 1

    Miyawaki, A. et al. Nature 388, 882–887 (1997).

    CAS  Article  Google Scholar 

  2. 2

    Xu, X. et al. Nucleic Acids Res. 26, 2034–2035 (1998).

    CAS  Article  Google Scholar 

  3. 3

    Hu, C.D., Chinenov, Y. & Kerppola, T.K. Mol. Cell 9, 789–798 (2002).

    CAS  Article  Google Scholar 

  4. 4

    Nagai, T., Sawano, A., Park, E.S. & Miyawaki, A. Proc. Natl. Acad. Sci. USA 98, 3197–3202 (2001).

    CAS  Article  Google Scholar 

  5. 5

    Nakai, J., Ohkura, M. & Imoto, K. Nat. Biotechnol. 19, 137–141 (2001).

    CAS  Article  Google Scholar 

  6. 6

    Alford, S.C., Abdelfattah, A.S., Ding, Y. & Campbell, R.E. Chem. Biol. 19, 353–360 (2012).

    CAS  Article  Google Scholar 

  7. 7

    Alford, S.C., Ding, Y., Simmen, T. & Campbell, R.E. ACS Synth. Biol. 1, 569–575 (2012).

    CAS  Article  Google Scholar 

  8. 8

    Wen, W., Meinkotht, J.L., Tsien, R.Y. & Taylor, S.S. Cell 82, 463–473 (1995).

    CAS  Article  Google Scholar 

  9. 9

    Kalderon, D., Roberts, B.L., Richardson, W.D. & Smith, A.E. Cell 39, 499–509 (1984).

    CAS  Article  Google Scholar 

  10. 10

    Simon, D.J. et al. J. Neurosci. 32, 17540–17553 (2012).

    CAS  Article  Google Scholar 

  11. 11

    Komatsu, N. et al. Mol. Biol. Cell 22, 4647–4656 (2011).

    CAS  Article  Google Scholar 

  12. 12

    Miyawaki, A. & Tsien, R.Y. Methods Enzymol. 327, 472–500 (2000).

    CAS  Article  Google Scholar 

  13. 13

    Bergsbaken, T., Fink, S.L. & Cookson, B.T. Nat. Rev. Microbiol. 7, 99–109 (2009).

    CAS  Article  Google Scholar 

  14. 14

    Zhang, Y., Tounekti, O., Akerman, B., Goodyer, C.G. & LeBlanc, A. J. Neurosci. 21, RC176 (2001).

    CAS  Article  Google Scholar 

  15. 15

    Botelho, R.J. et al. J. Cell Biol. 151, 1353–1368 (2000).

    CAS  Article  Google Scholar 

  16. 16

    Cui, J. et al. J. Neurosci. 31, 16227–16240 (2011).

    CAS  Article  Google Scholar 

  17. 17

    LeBlanc, A. J. Neurosci. 15, 7837–7846 (1995).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

We thank the University of Alberta Molecular Biology Service Unit, C.W. Cairo and R. Derda for technical assistance, and T. Meyer (Stanford) for the PLCδ-encoding gene. Funding support was provided by Canada Research Chairs (R.E.C.), the Alberta Glycomics Centre (R.E.C.), the Canadian Institutes of Health Research (NHG 99085 and MOP 123514 to R.E.C. and MOP 119425 to D.M.), the Natural Sciences and Engineering Research Council of Canada (Discovery grant to R.E.C. and a CGSD3 Scholarship to S.C.A.), Alberta Ingenuity PhD Scholarships (S.C.A. and Y.S.), a National Science Foundation of China Major Research Grant (91132718 to Y.Z.), the Beijing Natural Science Foundation (7142085 to Y.Z.), US National Institutes of Health DP1 CA174423 (to J.Z.) and 5R44NS082222 (to A.M.Q. and T.E.H.), and US National Science Foundation Small Business Innovation Research (SBIR) 1248138 and Montana SBIR matching funds #13-50 RCSBIR-003 (A.M.Q.).

Author information

Affiliations

Authors

Contributions

Y.D., J.Z., P.H.T., A.M.Q., T.E.H., D.M., S.C.A., Y.Z. and R.E.C. conceived of and designed experiments. Y.D. assembled all constructs except for the PIP2, PKA and ERK biosensors and performed imaging of all caspase-3, caspase-8 and caspase-9 biosensors and the three-polypeptide Ca2+ biosensor. J.L. performed imaging with the caspase-3 biosensor in neurons. J.R.E. determined heterodimer affinities in vitro. Y.S. performed imaging of the single-polypeptide Ca2+ biosensor. I.Z. performed imaging of the caspase-1 biosensor. P.H.T. assembled and performed imaging of the PIP2 and PKA biosensors. G.C.H.M. assembled and performed imaging of the ERK biosensor. All authors were involved in data analysis, and Y.D., J.Z., T.E.H., D.M., S.C.A., Y.Z. and R.E.C. wrote the manuscript.

Corresponding author

Correspondence to Robert E Campbell.

Ethics declarations

Competing interests

R.E.C., in partnership with the Alberta Glycomics Centre and the University of Alberta, has filed Canadian and US patent applications that describe the research in this manuscript. P.H.T., A.M.Q. and T.E.H. are employed by Montana Molecular, a for-profit company that develops genetically encoded fluorescent protein sensors. This does not alter the authors' adherence to all of the Nature Methods policies on sharing data and materials presented in this manuscript.

Integrated supplementary information

Supplementary Figure 1 B copies can rescue the fluorescence of the ‘wrong’ A copy partners.

(a) The B copy (B1) originally engineered to rescue the fluorescence of GA is able to rescue the fluorescence of RA. (b) B1 with the K153E mutation (B2) is also able to rescue the fluorescence of both A copies. (c) The B copy (B3) originally engineered to rescue the fluorescence of RA complements both RA and GA with a Kd of ~40 μM. For the FPX strategy, a lower Kd will favor one interaction over the other, assuming equal concentrations of the two A copies (i.e., RA favored over GA for binding to B2). Due to its closely matched, and relatively low, affinities for both RA and GA, B3 is our preferred B partner. The identity of the B partner used in each biosensor construct is provided in Supplementary Table 1. Source data

Supplementary Figure 2 Red-to-green FPX for detection of caspase-3 activity.

(a) Schematic representation of the biosensor. (b) Red and green whole cell intensities vs. time for a HeLa cell co-expressing RA-DEVD-B and GA and undergoing apoptosis. X-axis for all caspase activity imaging traces is time elapsed since 1 h after treatment with staurosporine. (c) Whole cell green-to-red intensity ratios vs. time for multiple cells treated and analyzed as in (b). Average green-to-red ratio change = 5.6 ± 1.6-fold (n = 7). Source data

Supplementary Figure 3 Representative whole-cell fluorescence intensity and ratio for the red-to-green protease biosensor based on translocation of the dark B copy, as shown in Fig. 1e–g and Supplementary Video 2.

(a) Whole cell intensities in the green and red channels. The average red intensity decrease = 4.4 ± 1.6-fold (n = 6) and the average green intensity increase = 4.7 ± 1.3-fold (n = 6). (b) Whole cell green-to-red fluorescence ratio. Source data

Supplementary Figure 4 Green-to-red FPX protease biosensor based on translocation of the B copy.

(a) Schematic representation of the biosensor. (b) Selected frames from imaging of HeLa cells co-expressing GANES-DEVD-BNLS and RANLS and undergoing apoptosis (Supplementary Video 3). Red fluorescence is represented as magenta. Scale bar represents 10 μm. (c) Left panel: Green cytoplasmic and red nuclear intensities vs. time for cell 1 using the ROIs represented in the left-most panels of the green and red channel images in (b). Time points of cytoplasmic and nuclear ROIs corresponding to the frames in (b) are represented as circles and squares, respectively. Middle panel: Whole cell intensities vs. time for the cell 1 ROI indicated in the left-most composite image of (b). Average green intensity decrease = 5.8 ± 1.7-fold (n = 3) and average red intensity increase = 4.9 ± 0.8-fold (n = 3). Right panel: Whole cell red-to-green ratio change using the data represented in the middle panel. (d) Data for cell 2, represented as in (c). Source data

Supplementary Figure 5 Imaging of caspase-3 activity during neuritic pruning.

(a) Composite fluorescence images of a Hoechst (blue) stained neuron co-expressing GANES-DEVD-BNLS and RANLS during nerve growth factor (NGF) deprivation. Scale bar represents 20 μm. (b) Nuclear red fluorescence intensity vs. time for neurons cultured with and without NGF (p < 0.01). During the first few hours of NGF depletion, the morphology of neurites remained unchanged, but the activation of caspase-3 could be monitored by the accumulation of red fluorescence in the nucleus. The diminished rate of increase from ~6 h (relative to in vitro measured caspase-3 activity as shown in (c)) is attributed to depletion of free RANLS. Error bars represent mean ± SE. (c) In vitro measured caspase-3 activity vs. time for neurons cultured with and without NGF (p < 0.01). Error bars represent mean ± SE. (d) Images of neurons cultured with or without NGF for 24 h. Following 24 h of NGF deprivation, neurons exhibited very little nuclear fragmentation, substantially fewer neurites relative to control, and brighter red fluorescence in the nucleus relative to control. Scale bar represents 50 μm. Source data

Supplementary Figure 6 Neurons stimulated to undergo apoptosis exhibited increased nuclear red fluorescence and increased fragmentation compared to neurons deprived of nerve growth factor (NGF).

(a) Two color composite fluorescence images of a neuron during staurosporine-induced apoptosis. Scale bar represents 20 μm. Inset: Zoom-in showing nucleus fragmentation following caspase-3 activation during apoptosis. Scale bar represents 2 μm. (b) Images of a neuron before and after 24 h of NGF deprivation. Scale bar represents 10 μm. Inset: Zoom-in showing that the nucleus integrity was been preserved. Scale bar represents 2 μm. (c) Quantification of nucleus fragmentation during apoptosis vs. NGF deprivation1. Source data

Supplementary Figure 7 Caspase activity sensing using FPX and translocation of the fluorogenic A partner.

(a) Schematic illustration of the biosensor for caspase-3 activity with translocation of the fluorogenic GA partner. (b) Selected frames (see Supplementary Video 4) from imaging of HeLa cells co-expressing GANLS-DEVD-BNES and BNLS and undergoing apoptosis. Scale bar represents 10 μm. (c) Intensity vs. time for the cytoplasmic and nuclear ROIs indicated in (b) with markers to indicate the time points for the images. The average maximum cytoplasmic intensity decrease = 2.5 ± 1.0-fold (n = 6) and the average nucleus maximum intensity increase = 5.3 ± 1.2-fold (n = 6). Subsequent apoptosis-associated fragmentation of the nucleus led to a decrease in the green fluorescence intensity in the nucleus. (d) Selected frames (see Supplementary Video 5) from imaging of HeLa cells co-expressing RANLS-DEVD-BNES and BNLS undergoing staurosporine-induced apoptosis. Scale bar represents 10 μm. (e) Intensity vs. time for the cytoplasmic and nuclear ROIs indicated in (d). Time points for the frames in (d) are represented as circles (cytoplasm) and squares (nucleus), respectively. The average cytoplasmic intensity decrease = 3.4 ± 1.3-fold (n = 5) and the average nucleus intensity increase = 4.7 ± 2.9-fold (n = 5). Source data

Supplementary Figure 8 Two-color FPX with translocating GA (caspase-3) and RA (caspase-9) partners.

(a) Schematic illustration of strategy for monitoring both caspase-3 (green) and caspase-9 (red) activity with two different FPX constructs. (b) Selected frames from two-color imaging of tumor necrosis factor α (TNFα)-treated HeLa cells co-expressing RANLS-LEHD-BNES (Ref. 2), GANLS-DEVD-BNES, and BNLS. Scale bar represents 10 μm. See Supplementary Video 6. (c) Green and red fluorescence intensities in both ROIs (left hand panel) and green and red nucleus-to-cytoplasm intensity ratios (right hand panel) for cell 1 as labeled in (b). (d) Intensities and ratios for cell 2, represented as in (c). The average nucleus-to-cytoplasm ratio change for green and red fluorescence = 6.1 ± 1.5-fold (n = 3) and 2.8 ± 1.1-fold (n = 4), respectively. Source data

Supplementary Figure 9 Two-color FPX with translocating GA (caspase-3) and RA (caspase-8) partners.

(a) Schematic illustration of attempted strategy for monitoring both caspase-3 (green) and caspase-8 (red) activity with two different FPX constructs. (b) Green and red fluorescence intensities for the nucleus and the cytoplasm (left hand panel), and green and red nucleus-to-cytoplasm intensity ratios (right hand panel), for a representative staurosporine-treated HeLa cells co-expressing RANLS-IETD-BNES (Ref. 3), GANLS-DEVD-BNES, and BNLS. (c) Data for a second representative cell, presented as in (b). The average nucleus-to-cytoplasm ratio change for green and red fluorescence = 4.6 ± 1.8-fold (n = 6) and 1.5 ± 0.3-fold (n = 6), respectively. Consistent with a previous report4, we observe effectively simultaneous activation of both caspase-3 and caspase-8. Source data

Supplementary Figure 10 Attempted caspase-dependent FPX with translocation from the nucleus to the cytoplasm.

(a) Selected frames from imaging of HeLa cells expressing GANES-DEVD-BNLS and BNES during staurosporine-induced apoptosis. (b) Intensity vs. time for the ROIs shown in (a). (c) Selected frames from imaging of HeLa cells expressing RANES-DEVD-BNLS and BNES and undergoing apoptosis. (d) Intensity vs. time for ROIs shown in (c).We suggest two possible explanations for these results. The first possibility is that ANES (G or R) cannot be exported from the nucleus because the delayed activation of caspase-3 in the nucleus relative to the cytoplasm5 is causing a critical compromise in the integrity of the nuclear pore complex6,7. The second possibility is that the larger volume of the cytoplasm relative to the nucleus is diluting the A copy to a concentration well below the Kd of the heterodimer (see Supplementary Fig. 1). Source data

Supplementary Figure 11 Representative green and red intensity data for single cells expressing the PIP2 biosensor or the cAMP biosensor.

(a) Green and red intensity vs. time for a single cell expressing the PIP2 biosensor represented in Fig. 2d. (b) Green and red intensity for a single cell expressing the cAMP biosensor represented in Fig. 2f. Source data

Supplementary Figure 12 Ratiometric FPX for imaging of ERK activity using a single polypeptide RA–WW domain–ERK substrate–B protein and free GANES.

(a) Schematic of the ERK kinase activity reporter that rationalizes the observed ratio changes. In the absence of phosphorylation, either RA (intramolecular) or GA (intermolecular) can be bound to B. This equilibrium still exists in the phosphorylated form of the protein, but the association between RA and B is more favored due to the association of the WW domain and the phosphorylated substrate. (b) Ratiometric responses (left) and representative single channel intensity responses (right) of the cytosolic ERK kinase reporter co-expressed with GANES in HEK293 cells. Ratiometric intensity responses can clearly be seen approximately 5 min after stimulation with epidermal growth factor (huEGF, 100 ng/mL), consistent with results using a FRET-based biosensor8. The single channel intensity responses shown in the right panel correspond to the magenta trace on the left panel. Increases in the green and red channels after ~10 minutes are attributed to cell movement and morphological changes. These changes cancel out in the ratiometric plot, highlighting the advantage of ratiometric techniques such as FRET and FPX. Source data

Supplementary Figure 13 Intramolecular FPX for imaging of caspase-8 activity.

(a) Schematic representation of a single polypeptide FPX biosensor for caspase-8. (b) Whole cell green and red intensities from imaging of HeLa cells expressing RA-IETD-B-linker-GANES and undergoing staurosporine-induced apoptosis. (c) Green-to-red intensity ratio vs. time for multiple cells treated and analyzed as in (b). The average green-to-red ratio change = 5.1 ± 0.8-fold (n = 8). This intramolecular exchange approach is somewhat analogous to previously reported protease biosensors based on β-strand exchange within a single FP9. Source data

Supplementary Figure 14 Intramolecular FPX for imaging of caspase-1 activity.

(a) Schematic representation of a single polypeptide FPX biosensor for caspase-1. (b) Images of U251 glioblastoma cells expressing RA-linker-B-YVAD-GA and treated with inflammatory and non-inflammatory stimuli for 4 h. Scale bar represents 100 μm. Inset: Zoom-in showing red and green fluorescence. (c) Whole cell red-to-green intensity ratios from imaging U251 cells (n = 100) treated as in (b). The black line indicates the median. (d) Whole cell red-to-green intensity ratios for U251, MCF-7 breast adenocarcinoma, and HepG2 liver carcinoma cells treated with inflammatory and non-inflammatory stimuli for 4 h. An increase in the red/green ratio indicates an increase in caspase-1 activity (p < 0.01). Mean values are expressed relative to the control set to 1. Error bars represent mean ± SEM.Adenosine triphosphate (ATP) in combination with lipopolysaccharide (LPS), a prototypical pyrogenic stimulus10, was found to promote caspase-1 activity in all three cell lines. MCF-7 cells, which lack caspase-3, displayed strong caspase-1 activation resulting from curcumin and staurosporine, two potent cytotoxic agents11. In contrast, staurosporine did not stimulate caspase-1 activity in U251 cells, which instead undergo necrosis and caspase-3-mediated apoptosis12. Glioblastoma cells strongly responded to the pro-inflammatory cytokine tumor necrosis factor alpha (TNF-α)13. In HepG2 cells, caspase-1 activity related to metabolic stress was induced using palmitic acid, a saturated fatty acid implicated in the metabolic syndrome14. Although pro-inflammatory pathways in hepatocytes are also up-regulated in chronic alcoholic hepatitis, acute stimulation of HepG2 cells with ethanol did not significantly induce caspase-1 activity at 4 h15. Source data

Supplementary Figure 15 Intramolecular FPX for imaging of caspase-3 and caspase-8 activity using a single polypeptide.

(a) Schematic representation of the single polypeptide FPX biosensor for both caspase-3 and caspase-8. (b) Selected frames from imaging of HeLa cells expressing RA-IETD-B-DEVD-GANES and undergoing staurosporine-induced apoptosis. Scale bar represents 10 μm. (c) Intensity vs. time of green and red fluorescence. Time points corresponding to the frames of green and red channel in (b) are represented as circles and squares, respectively. The average red-to-green ratio change = 7.9 ± 1.4-fold (n = 4).The mismatch in the amount of protease activity (i.e., caspase-3 cleavage (Fig. 3f) occurs more rapidly than caspase-8 (Supplementary Fig. 13c)) complicates the interpretation of these results and so we recommend the two-color translocation strategy shown in Supplementary Figs. 8 and 9 for imaging of two caspase activities. Source data

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–15 and Supplementary Tables 1 and 2 (PDF 8587 kb)

Intermolecular green-to-red FPX-based biosensor for caspase-3 activity.

Time-lapse fluorescence imaging of caspase-3 activity with GA-DEVD-B and RA. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 682 kb)

Translocating red-to-green FPX-based biosensor for caspase-3 activity.

Time-lapse fluorescence imaging of caspase-3 activity with NESRA-DEVD-BNLS and GANLS. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 628 kb)

Translocating green-to-red FPX-based biosensor for caspase-3 activity.

Time-lapse fluorescence imaging of caspase-3 activity with NESGA-DEVD-BNLS and RANLS. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 2677 kb)

FPX biosensor for caspase-3 activity based on exchange of GA.

Time-lapse fluorescence imaging of caspase-3 activity with GANLS-DEVD-BNES and BNLS. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 405 kb)

FPX biosensor for caspase-3 activity based on exchange of RA.

Time-lapse fluorescence imaging of caspase-3 activity with RANLS-DEVD-BNES and BNLS. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 3298 kb)

Two-color imaging of caspase-3 and caspase-9 activity by A copy exchange.

Time-lapse fluorescence imaging of caspase-3 and caspase-9 activity with GANLS-DEVD-BNES, RANLS-LEHD-BNES, and BNLS. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 487 kb)

Intermolecular FPX-based biosensor for Ca2+.

Time-lapse fluorescence imaging of histamine treated HeLa cells that are co-expressing RA-CaM, B-M13, and GA. The green and red fluorescence channels have been overlaid in this movie. (AVI 963 kb)

Intermolecular FPX-based biosensor for PIP2.

Time-lapse fluorescence imaging of PIP2 concentration in the membrane by co-expression of RA and B fused to PH domains, and free GA. HEK293 cells have been treated with carbachol. The change occurs within one imaging interval (7 s) and thus appears quite sudden in this movie. (AVI 123 kb)

Intramolecular FPX-based biosensor for Ca2+.

Time-lapse fluorescence imaging of histamine treated HeLa cells expressing the single polypeptide construct RA-CaM-B-M13-GA. The ratio of the green and red fluorescence intensities has been pseudocolored according to the look-up-table shown. (AVI 2531 kb)

Intramolecular FPX-based biosensor for caspase-3.

Time-lapse fluorescence imaging of caspase-3 activity in cells expressing the single polypeptide construct RA-linker-B-DEVD-GANES. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 642 kb)

Intramolecular FPX-based biosensor for caspase-8.

Time-lapse fluorescence imaging of caspase-8 activity in cells expressing the single polypeptide construct RA-IETD-B-linker-GANES. HeLa cells were treated with staurosporine to initiate apoptosis. (AVI 1617 kb)

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ding, Y., Li, J., Enterina, J. et al. Ratiometric biosensors based on dimerization-dependent fluorescent protein exchange. Nat Methods 12, 195–198 (2015). https://doi.org/10.1038/nmeth.3261

Download citation

Further reading

Search

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