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.

In vivo single-cell labeling by confined primed conversion


Spatially confined green-to-red photoconversion of fluorescent proteins with high-power, pulsed laser illumination is negligible, thus precluding optical selection of single cells in vivo. We report primed conversion, in which low-power, dual-wavelength, continuous-wave illumination results in pronounced photoconversion. With a straightforward addition to a conventional confocal microscope, we show confined primed conversion in living zebrafish and reveal the complex anatomy of individual neurons packed between neighboring cells.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Dual-laser illumination of Dendra2 leads to photoconversion.
Figure 2: Spatially confined primed conversion enables individual neuron labeling in tightly bundled neural clusters in living zebrafish larvae.


  1. Pantazis, P. & Supatto, W. Nat. Rev. Mol. Cell Biol. 15, 327–339 (2014).

    CAS  Article  Google Scholar 

  2. Subach, O.M. et al. Nat. Methods 8, 771–777 (2011).

    CAS  Article  Google Scholar 

  3. Piatkevich, K.D., Subach, F.V. & Verkhusha, V.V. Nat. Commun. 4, 2153 (2013).

    Article  Google Scholar 

  4. Betzig, E. et al. Science 313, 1642–1645 (2006).

    CAS  Article  Google Scholar 

  5. Plachta, N., Bollenbach, T., Pease, S., Fraser, S.E. & Pantazis, P. Nat. Cell Biol. 13, 117–123 (2011).

    CAS  Article  Google Scholar 

  6. Müller, P. et al. Science 336, 721–724 (2012).

    Article  Google Scholar 

  7. Dempsey, W.P., Fraser, S.E. & Pantazis, P. PLoS ONE 7, e32888 (2012).

    CAS  Article  Google Scholar 

  8. Zhou, X.X., Chung, H.K., Lam, A.J. & Lin, M.Z. Science 338, 810–814 (2012).

    CAS  Article  Google Scholar 

  9. Brown, S.C. et al. Plant J. 63, 696–711 (2010).

    CAS  Article  Google Scholar 

  10. Gurskaya, N.G. et al. Nat. Biotechnol. 24, 461–465 (2006).

    CAS  Article  Google Scholar 

  11. Fu, D., Ye, T., Matthews, T.E., Yurtsever, G. & Warren, W.S. J. Biomed. Opt. 12, 054004 (2007).

    Article  Google Scholar 

  12. McKinney, S.A., Murphy, C.S., Hazelwood, K.L., Davidson, M.W. & Looger, L.L. Nat. Methods 6, 131–133 (2009).

    CAS  Article  Google Scholar 

  13. Durisic, N., Laparra-Cuervo, L., Sandoval-Álvarez, Á., Borbely, J.S. & Lakadamyali, M. Nat. Methods 11, 156–162 (2014).

    CAS  Article  Google Scholar 

  14. Adam, V., Nienhaus, K., Bourgeois, D. & Nienhaus, G.U. Biochemistry 48, 4905–4915 (2009).

    CAS  Article  Google Scholar 

  15. Nienhaus, K., Nienhaus, G.U., Wiedenmann, J. & Nar, H. Proc. Natl. Acad. Sci. USA 102, 9156–9159 (2005).

    CAS  Article  Google Scholar 

  16. Zhou, S. et al. Curr. Biol. 22, 668–675 (2012).

    CAS  Article  Google Scholar 

  17. Kanchanawong, P. et al. Nature 468, 580–584 (2010).

    CAS  Article  Google Scholar 

  18. Lichtman, J.W. & Conchello, J.-A. Nat. Methods 2, 910–919 (2005).

    CAS  Article  Google Scholar 

  19. Lelimousin, M., Adam, V., Nienhaus, G.U., Bourgeois, D. & Field, M.J. J. Am. Chem. Soc. 131, 16814–16823 (2009).

    CAS  Article  Google Scholar 

  20. Denk, W., Briggman, K.L. & Helmstaedter, M. Nat. Rev. Neurosci. 13, 351–358 (2012).

    CAS  Article  Google Scholar 

  21. Pantazis, P. & González-Gaitán, M. J. Biomed. Opt. 12, 044004 (2007).

    Article  Google Scholar 

  22. Truong, T.V., Supatto, W., Koos, D.S., Choi, J.M. & Fraser, S.E. Nat. Methods 8, 757–760 (2011).

    CAS  Article  Google Scholar 

  23. Wiedenmann, J. et al. Proc. Natl. Acad. Sci. USA 101, 15905–15910 (2004).

    CAS  Article  Google Scholar 

  24. Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H. & Miyawaki, A. Proc. Natl. Acad. Sci. USA 99, 12651–12656 (2002).

    CAS  Article  Google Scholar 

  25. Tsutsui, H., Karasawa, S., Shimizu, H., Nukina, N. & Miyawaki, A. EMBO Rep. 6, 233–238 (2005).

    CAS  Article  Google Scholar 

  26. Habuchi, S., Tsutsui, H., Kochaniak, A.B., Miyawaki, A. & van Oijen, A.M. PLoS ONE 3, e3944 (2008).

    Article  Google Scholar 

  27. Subach, F.V. et al. Nat. Methods 6, 153–159 (2009).

    CAS  Article  Google Scholar 

  28. White, R.M. et al. Cell Stem Cell 2, 183–189 (2008).

    CAS  Article  Google Scholar 

  29. Westerfield, M. The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish (Danio rerio) 5th edn. (University of Oregon Press, 2007).

  30. Song, L., Hennink, E.J., Young, I.T. & Tanke, H.J. Biophys. J. 68, 2588–2600 (1995).

    CAS  Article  Google Scholar 

  31. Rizzo, M.A., Springer, G.H., Granada, B. & Piston, D.W. Nat. Biotechnol. 22, 445–449 (2004).

    CAS  Article  Google Scholar 

  32. Dempsey, W.P., Qin, H. & Pantazis, P. Methods Mol. Biol. 1148, 217–228 (2014).

    CAS  Article  Google Scholar 

  33. Kim, C.H. et al. Neurosci. Lett. 216, 109–112 (1996).

    CAS  Article  Google Scholar 

  34. Sato, T., Takahoko, M. & Okamoto, H. Genesis 44, 136–142 (2006).

    CAS  Article  Google Scholar 

  35. Caneparo, L., Pantazis, P., Dempsey, W. & Fraser, S.E. PLoS ONE 6, e20230 (2011).

    CAS  Article  Google Scholar 

Download references


We thank all members of the Pantazis lab for discussion and advice and the Single Cell Facility (SCF) at the Department of Biosystems Science and Engineering (D-BSSE) for technical support, especially A. Ponti. We thank the Biophysics Facility of the Biozentrum, University of Basel, for providing access to their fluorimeter and help in collecting data. We would like to thank S. Fraser for discussions that initiated this work, and R. Paro, T. Bollenbach, W. Supatto, S. Tay and F. Rudolf for critical comments on the manuscript. This work was supported by the Swiss National Science Foundation (SNF grant no. 31003A_144048) and the European Union Seventh Framework Program (Marie Curie Career Integration Grant (CIG) no. 334552). W.P.D. was supported by the Swiss National Center of Competence in Research (NCCR) “Nanoscale Science.”

Author information

Authors and Affiliations



P.P., W.P.D. and T.V.T. conceived and P.P., W.P.D. and L.G. refined the idea. P.P., W.P.D. and L.G. designed and L.G., W.P.D. and A.Y.S. implemented the experiments. L.G., P.M.H., A.Y.S., W.P.D. and M.H. performed the experiments; W.P.D. set up primed conversion and L.G. set up confinement of primed conversion. L.G., W.P.D., P.M.H., A.Y.S., T.V.T. and P.P. analyzed the data. P.P., W.P.D. and L.G. prepared the manuscript with editing input from all authors.

Corresponding author

Correspondence to Periklis Pantazis.

Ethics declarations

Competing interests

A patent application has been filed by ETH transfer, the technology transfer office of ETH Zurich, relating to aspects of the work described in this manuscript. Authors listed on the patent: P.P., W.P.D., T.V.T. and L.G.

Integrated supplementary information

Supplementary Figure 1 Two-photon (2P) photoconversion is negligible for many popular photoconvertible proteins.

a,b, After strong (50 mW average power) illumination, the red channel data (a) shows photoconversion for many different FPs (< 3 fold increase over background). The green channel (b) shows photobleaching. As a control, we used the same settings to try to photomodulate the dark-to-red photoactivatable FP PAmCherry (black data points). The dashed line in a indicates the fold change possible with primed conversion (see Fig. 1d). c, d, Comparison of long-term (4.5 min) 2P illumination (white bars) versus 405 nm photoconversion (gray bars) for all tested green-to-red photoconvertible proteins. As before, the red channel data (c) shows photoconversion, while the green channel data (d) shows photobleaching. In the green channel data, a value of 1 indicates that the green population is unaffected by pulsed illumination. For each pair in c, d, a t-test (two-tailed, unequal variance) was performed: c, *P ~ 0.005, **P < 0.005; d, *P < 0.005.

Supplementary Figure 2 In-depth characterization of Dendra2 confirms negligible photoconversion with 2P laser sources.

Green Dendra2 was irradiated with focused light from a tunable 2P laser source at different wavelengths and illuminating powers (25 mW, dark gray bars; 50 mW, gray bars; 75 mW, white bars) for either a, b, a short exposure time or c, d, an approximately 8-fold longer exposure time (see Online Methods). After laser illumination, the red channel data in a, c, show photoconversion is at most 5-fold above background, and b, d, green channel data show significant photobleaching at higher power illumination (fold change is less than 1). At each wavelength, a one-way ANOVA test was performed: a, P < 0.05 for only the 800 nm case; b, P < 0.01 for all wavelengths; c, P < 0.05 for 750, 850, and 900 nm; d, P < 0.05 for all wavelengths. * represents significance between a pair within a 95% confidence interval and ** represents significance between a pair within a 99% confidence interval after post-hoc analyses (see Online Methods).

Supplementary Figure 3 Rapid photobleaching occurs when performing simultaneous scanning of two distinct lasers in an intact, living sample.

With a commercial confocal microscope, we imaged individual nuclei within embryonic zebrafish at 1 day post fertilization (dpf) expressing nuclear-targeted Cerulean FP (see Online Methods). a, Nuclei were scanned continuously for 30 frames using a 458 nm laser alone (left column), a 561 nm laser alone (center column), or a simultaneous combination of 458 and 561 nm lasers (right column). Images were taken of the nuclei before (top row) and after (bottom row) the continuous time-lapse to monitor decreases in fluorescence intensity. No significant photobleaching was seen in the time-lapses with the 561 nm laser alone (< 10% decrease in fluorescence intensity, N = 5 nuclei). b, We quantified the rate of photobleaching within the imaged nuclei by fitting the time-lapse data from the left and right columns with a bi-exponential function (see Online Methods). The rate of Cerulean photobleaching increased nonlinearly when the 458 and 561 nm lasers illuminated the sample simultaneously (R2~1, N = 5 nuclei). Scale bar, 5 µm.

Supplementary Figure 4 mEos2 can be photoconverted with dual-wavelength illumination, suggesting that primed conversion is a general phenomenon.

These panels show single optical sections of the green and red channels after attempted photoconversion of mEos2 in a solidified gel. Results from photoconversion with 488 nm alone (left column), with 730 nm alone (middle column), and with primed conversion (right column) are shown. Scale bar, 10 µm.

Supplementary Figure 5 Photoconversion with either 405-nm illumination or primed conversion results in a photoconverted Dendra2 population with the same spectral properties and photostability.

a, The excitation and b, emission spectra (normalized to the peak value) of photoconverted red Dendra2 after primed conversion (black curve) and after near-UV illumination (gray curve). In the graphs, a reference spectrum (red dashed curve) for the photoconverted form of Dendra2 is also shown (see Online Methods). c, Dendra2 was continuously exposed to 561 nm light for ~290 seconds. Individual frames from this time-lapse were analyzed and fit to a bi-exponential function (see Online Methods). d, Samples of photoconverted Dendra2 that were kept in the dark between imaging experiments were imaged at several time points (see Online Methods) to characterize the chemical stability of the photoconverted form of the FP. The half life of the red form of Dendra2 in vitro is approximately 4 days. For each time point in (c) and (d), a t-test (two-tailed, unequal variance) was performed: differences between the two photoconverted Dendra2 populations were not significant (P > 0.1) for all time points.

Supplementary Figure 6 The efficiency of Dendra2 photoconversion decreases below pH 6.

Green-normalized fold change of photoconverted red Dendra2 over background (see Online Methods) at different pH values. A one-way ANOVA test was performed with these data: P < 0.005. Resulting from the post-hoc test (Tukey Range Test), * represents significance between a pair within a 95% confidence interval and ** represents significance between a pair within a 99% confidence interval.

Supplementary Figure 7 Efficient primed conversion is achieved using moderate imaging conditions.

Dendra2 was photoconverted either with a, b, varying priming beam power and a constant number of scan repetitions or c, d, constant priming beam power and a varying number of scan repetitions. In all cases, the converting beam was fixed at 5 mW. Photoconversion was achieved by scanning an entire frame (512 x 512 pixels) within the embedded protein gel (see Online Methods). a, c, Photoconversion fold change in the red channel for primed conversion (gray bars) and for 488 nm illumination (< 6-fold, white bars). b, d, Photobleaching of the unphotoconverted green Dendra2 population after laser illumination. For each pair (primed conversion versus 488 nm alone), a two-tailed, unequal variance t-test was performed: *, P < 0.05; ** P < 0.005.

Supplementary Figure 8 Line scanning reveals the peak primed conversion efficiency that can be achieved in vitro.

In this experiment, Dendra2 was photoconverted in a similar manner to Supplementary Fig. 7: the priming beam power was either a, b, varied while scanning repetitions were fixed, or c, d, fixed while scanning repetitions were varied. The major difference in the experiment was that photoconversion was achieved with a line scan (512 x 1 pixels). Photoconversion (maximum is ~ 60-fold) at a, a fixed number of iterations (2400 repetitions) and c, a fixed priming beam power (0.02 mW). b, d, Photobleaching of the green form of Dendra2 after laser illumination. For each pair (primed conversion versus 488 nm alone), a two-tailed, unequal variance t-test was performed: *, P < 0.05; **, P < 0.005.

Supplementary Figure 9 Schematic illustration of how the conversion filter plate confines primed conversion to the focus.

The conversion filter plate is placed in the infinity space of the beam path where the lasers (represented as individual ray beamlets) are collimated, within a filter cube beneath the objective. Each half of the conversion filter plate passes either the priming beam or the converting beam. Thus, the lasers only meet within the sample at the focus. An opaque strip aids in separating the beams, maintaining axial confinement by ensuring a sufficient time delay between both lasers as they scan areas above and below the focus.

Supplementary Figure 10 Confinement of primed conversion is most efficient when a region of interest is irradiated using a short pixel dwell time.

Left column: axial (XZ) average intensity projection images of the photoconverted red Dendra2 channel show the confinement of Dendra2 photoconversion as a function of pixel dwell time (12.61 µsec in a, 2.55 µsec in b, and 0.64 µsec in c). Right column: These graphs show the normalized intensity profiles of the images in the left column. The areas outside of the approximated full width half maximum (FWHM) are grayed out. Confinement of photoconverted red Dendra2 signal in a, is ~ 4-5 µm, b, is ~ 6.5 µm, and c, is nearly 10 µm. In each case, the level of red signal decreases sharply to just above the background level (compare to Fig. 1g in the main text, without the conversion filter plate in place). Scale bar, 5 µm (lateral dimension) and 2 µm (axial dimension).

Supplementary Figure 11 Traditional photoconversion using 405-nm light cannot select out individual neurons in tightly packed neural ganglia in vivo.

A cell surrounded on all sides by labeled neighbors within the trigeminal ganglion of a zebrafish larvae expressing cytoplasmic Dendra2 under the neural-specific HuC promoter was targeted for photoconversion. a, b, As in Fig. 2 (main text), we attempted photoconversion of Dendra2 in a single neuron in the trigeminal ganglion using 405-nm laser light (maximum intensity projection, ~102 µm in depth). c, Higher magnification images of the boxed region in a. Many neighboring cells are seen in the green channel (green, top). In the red channel (magenta, middle), many nearby cells (arrows) were photoconverted along with the target cell, which can also be appreciated in the merged image (bottom). Arrowheads point out the target cell in the experiment. Lateral scale bars: 50 µm (a, b); axial scale bar, 10 µm (c; thickness of the bar represents 2 µm laterally).

Supplementary Figure 12 Regardless of the location in the zebrafish, cells can be individually labeled using spatially confined primed conversion.

These panels show the individual green (left column) and red (magenta pseudocolor, right column) channels for the images from Fig. 2 e, f, which were taken within 3 dpf zebrafish expressing neural specific Dendra2. a, c, Individual cell projections become indiscernible within clusters of a, synapsing neurons in the optic tectum and c, among parallel bundled neural tracts within the developing spinal territory, which can be seen in the green Dendra2 channel. b, d, We used primed conversion to select out a cell of interest in these regions of the brain (arrowheads), and the photoconverted Dendra2 quickly diffused out from the cell body, allowing us to uncover the neurites within these cells. Scale bars, 20 µm.

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–12, Supplementary Table 1, Supplementary Results and Supplementary Discussion (PDF 2915 kb)

Supplementary Data 1

Beta-Actin:H2B–Cerulean - SV40 polyA sequence Beta-Actin (GenBank: EF026002):H2B (Human cds GenBank: AF531293) –Cerulean - SV40 polyA FASTA sequence with H2B coding sequence emphasized first in bold font, and Cerulean coding sequence emphasized second in bold font. (DOCX 169 kb)

Supplementary Data 2

HuC:Dendra2-SV40 polyA sequence HuC (GenBank: AF173984):Dendra2-SV40 polyA FASTA sequence with Dendra2 coding sequence emphasized in bold font. (DOCX 170 kb)

Primed photoconversion with the conversion filter plate in place leads to spatially confined photoconversion of Dendra2.

In this cartoon movie, we depict the laser as a series of four beamlets that strike the conversion filter plate at different places. When the priming and converting lasers are scanned simultaneously through the conversion plate, they only meet in the sample at the focus, which enables us to photoconvert cells (red color) in a spatially confined manner. No off-target photoconversion above and below the focal plane should be visible when the conversion plate is in place. (MOV 896 kb)

Spatially confined primed conversion of a single cell within the tightly packed trigeminal ganglion can be visualized with high contrast in 3D.

This movie shows a rotation of a single snapshot in time for the 4 dpf zebrafish larva depicted in Figure 2d. Here, we focus mainly on the result of filament tracing and cell segmentation in the red Dendra2 channel (magenta pseudocolor) after photoconversion in context of the green Dendra2 cells in the surrounding environment. (MOV 3470 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Dempsey, W., Georgieva, L., Helbling, P. et al. In vivo single-cell labeling by confined primed conversion. Nat Methods 12, 645–648 (2015).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

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