Faster STORM using compressed sensing

Journal name:
Nature Methods
Volume:
9,
Pages:
721–723
Year published:
DOI:
doi:10.1038/nmeth.1978
Received
Accepted
Published online

In super-resolution microscopy methods based on single-molecule switching, the rate of accumulating single-molecule activation events often limits the time resolution. Here we developed a sparse-signal recovery technique using compressed sensing to analyze images with highly overlapping fluorescent spots. This method allows an activated fluorophore density an order of magnitude higher than what conventional single-molecule fitting methods can handle. Using this method, we demonstrated imaging microtubule dynamics in living cells with a time resolution of 3 s.

At a glance

Figures

  1. STORM image analysis using compressed sensing.
    Figure 1: STORM image analysis using compressed sensing.

    (a) Simulations that demonstrate the capability of compressed sensing to identify molecules efficiently at a high density. Scale bars, 300 nm. Also see Supplementary Figure 7 for a low-signal example. (b) Comparison of the efficiency of molecule identification using compressed sensing and single-molecule fitting. The simulation is for an average photon number of 3,000 per molecule and a background of 70 photons per pixel (see Online Methods). Error bars stand for s.d. from repeated simulations. Dashed line marks the case when the number of identified molecules equals the number of molecules in the simulation. (c) Comparison of localization precisions. The y axis is labeled in both FWHM and s.d. The dashed line marks the Cremer-Rao lower bound (CRLB) of single-molecule localization (8.8 nm FWHM). (d) Minimum number of frames to achieve a given overall image resolution for a continuous 2D sample. The line for the fitting method is calculated using a constant 0.58 μm−2 identified molecule density, whereas the curves for the compressed sensing are calculated using identified molecule densities that allow the corresponding localization precisions to match the desired image resolution.

  2. Experimental STORM images using compressed sensing.
    Figure 2: Experimental STORM images using compressed sensing.

    (a) STORM imaging of microtubules in Drosophila S2 cells immunostained with secondary antibody labeled with the Alexa Fluor 647–Cy3 dye pair. Left column, conventional fluorescence image and one raw image frame captured during STORM data acquisition, showing high density of activated fluorophores. Middle column, result of single-molecule fitting, reconstructed from 100 and 500 frames of camera images, respectively. Right column, result for compressed sensing using the same set of camera images. Scale bars, 300 nm. (b) STORM imaging of mEos2-tubulin in a living Drosophila S2 cell. The conventional fluorescence image in the leftmost panel is acquired before STORM imaging. Three snapshots from the STORM movie are displayed, each with 3-s integration time. The dynamics of the microtubules can be clearly observed. See Supplementary Video 1 online.

References

  1. Huang, B., Babcock, H. & Zhuang, X. Cell 143, 10471058 (2010).
  2. Hell, S.W. Nat. Methods 6, 2432 (2009).
  3. Rust, M.J., Bates, M. & Zhuang, X. Nat. Methods 3, 793795 (2006).
  4. Betzig, E. et al. Science 313, 16421645 (2006).
  5. Hess, S.T., Girirajan, T.P.K. & Mason, M.D. Biophys. J. 91, 42584272 (2006).
  6. Shroff, H., Galbraith, C.G., Galbraith, J.A. & Betzig, E. Nat. Methods 5, 417423 (2008).
  7. Jones, S.A., Shim, S.H., He, J. & Zhuang, X. Nat. Methods 8, 499505 (2011).
  8. Holden, S.J., Uphoff, S. & Kapanidis, A.N. Nat. Methods 8, 279280 (2011).
  9. Huang, F., Schwartz, S.L., Byars, J.M. & Lidke, K.A. Biomed. Opt. Express 2, 13771393 (2011).
  10. Quan, T. et al. Opt. Express 19, 1696316974 (2011).
  11. Cox, S. et al. Nat. Methods 9, 195200 (2012).
  12. Candès, E.J., Romberg, J. & Tao, T. IEEE Trans. Inf. Theory 52, 489509 (2004).
  13. Zhu, L. et al. Phys. Med. Biol. 53, 66536672 (2008).
  14. Bates, M., Huang, B., Dempsey, G.T. & Zhuang, X. Science 317, 17491753 (2007).
  15. Huang, B., Wang, W., Bates, M. & Zhuang, X. Science 319, 810813 (2008).
  16. Candès, E.J., Romberg, J. & Tao, T. Commun. Pure Appl. Math. 59, 12071223 (2005).
  17. Zhu, L. & Xing, L. Med. Phys. 36, 18951905 (2009).
  18. Mortensen, K.I., Churchman, L.S., Spudich, J.A. & Flyvbjerg, H. Nat. Methods 7, 377381 (2010).
  19. Grant, M. & Boyd, S. CVX: Matlab software for disciplined convex programming, version 1.21 <http://cvxr.com/cvx> (2011).
  20. Huang, B., Jones, S.A., Brandenburg, B. & Zhuang, X. Nat. Methods 5, 10471052 (2008).

Download references

Author information

Affiliations

  1. Nuclear and Radiological Engineering and Medical Physics Programs, The George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, Georgia, USA.

    • Lei Zhu
  2. Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, California, USA.

    • Wei Zhang &
    • Bo Huang
  3. Tetrad Graduate Program, University of California, San Francisco, San Francisco, California, USA.

    • Daniel Elnatan &
    • Bo Huang
  4. Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, California, USA.

    • Bo Huang

Contributions

L.Z. and B.H. conceived the project and developed the algorithms, W.Z. performed the experiments, D.E. and B.H. analyzed the data, and L.Z. and B.H. wrote the manuscript.

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Text and Figures (745K)

    Supplementary Figures 1–9 and Supplementary Note

Movies

  1. Supplementary Video 1 (2M)

    STORM 'movie' of microtubules in a living Drosophila S2 cell stably expressing mEos2-fused tubulin, with a time resolution of 3 seconds. The movie is reconstructed from 4,349 camera frames (77 seconds) and plays 11 times as fast as real time. Three snapshots from the movie are shown in Figure 2b. Scale bar, 1 μm.

Zip files

  1. Supplementary Software (229K)

    Matlab code for STORM data analysis using compressed sensing.

Additional data