Skip to main content

Thank you for visiting nature.com. 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.

  • Article
  • Published:

Imaging G protein–coupled receptors while quantifying their ligand-binding free-energy landscape

Abstract

Imaging native membrane receptors and testing how they interact with ligands is of fundamental interest in the life sciences but has proven remarkably difficult to accomplish. Here, we introduce an approach that uses force-distance curve–based atomic force microscopy to simultaneously image single native G protein–coupled receptors in membranes and quantify their dynamic binding strength to native and synthetic ligands. We measured kinetic and thermodynamic parameters for individual protease-activated receptor-1 (PAR1) molecules in the absence and presence of antagonists, and these measurements enabled us to describe PAR1's ligand-binding free-energy landscape with high accuracy. Our nanoscopic method opens an avenue to directly image and characterize ligand binding of native membrane receptors.

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

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Principle of FD-based AFM to detect ligand binding to PAR1.
Figure 2: Mapping ligand binding to human PAR1 using FD-based AFM.
Figure 3: Extracting energetic, thermodynamic and kinetic parameters from force curves describes the ligand-binding free-energy landscape.
Figure 4: Loading rate–dependent interaction forces of single ligand-receptor bonds quantitate the ligand-binding energy landscape of PAR1.
Figure 5: Free-energy landscape describing the thermodynamic (ΔGbu) and kinetic (xu) parameters of peptide-based ligands binding to PAR1.

Similar content being viewed by others

Accession codes

Accessions

Protein Data Bank

References

  1. Coughlin, S.R. Thrombin signalling and protease-activated receptors. Nature 407, 258–264 (2000).

    Article  CAS  Google Scholar 

  2. Macfarlane, S.R., Seatter, M.J., Kanke, T., Hunter, G.D. & Plevin, R. Proteinase-activated receptors. Pharmacol. Rev. 53, 245–282 (2001).

    CAS  PubMed  Google Scholar 

  3. Zhang, C. et al. High-resolution crystal structure of human protease-activated receptor 1. Nature 492, 387–392 (2012).

    Article  CAS  Google Scholar 

  4. Vu, T.K.H., Hung, D.T., Wheaton, V.I. & Coughlin, S.R. Molecular-cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 64, 1057–1068 (1991).

    Article  CAS  Google Scholar 

  5. Arora, P., Ricks, T.K. & Trejo, J. Protease-activated receptor signalling, endocytic sorting and dysregulation in cancer. J. Cell Sci. 120, 921–928 (2007).

    Article  CAS  Google Scholar 

  6. Kunishima, N. et al. Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor. Nature 407, 971–977 (2000).

    Article  CAS  Google Scholar 

  7. Siu, F.Y. et al. Structure of the human glucagon class B G-protein-coupled receptor. Nature 499, 444–449 (2013).

    Article  CAS  Google Scholar 

  8. Dufrêne, Y.F., Martínez-Martín, D., Medalsy, I., Alsteens, D. & Müller, D.J. Multiparametric imaging of biological systems by force-distance curve-based AFM. Nat. Methods 10, 847–854 (2013).

    Article  Google Scholar 

  9. Alsteens, D., Trabelsi, H., Soumillion, P. & Dufrêne, Y.F. Multiparametric atomic force microscopy imaging of single bacteriophages extruding from living bacteria. Nat. Commun. 4, 2926 (2013).

    Article  Google Scholar 

  10. Pfreundschuh, M., Martínez-Martín, D., Mulvihill, E., Wegmann, S. & Müller, D.J. Multiparametric high-resolution imaging of native proteins by force-distance curve–based AFM. Nat. Protoc. 9, 1113–1130 (2014).

    Article  CAS  Google Scholar 

  11. Butt, H.-J., Cappella, B. & Kappl, M. Force measurements with the atomic force microscope: technique, interpretation and applications. Surf. Sci. Rep. 59, 1–152 (2005).

    Article  CAS  Google Scholar 

  12. Zocher, M., Fung, J.J., Kobilka, B.K. & Müller, D.J. Ligand-specific interactions modulate kinetic, energetic, and mechanical properties of the human β2 adrenergic receptor. Structure 20, 1391–1402 (2012).

    Article  CAS  Google Scholar 

  13. Venkatakrishnan, A.J. et al. Molecular signatures of G-protein-coupled receptors. Nature 494, 185–194 (2013).

    Article  CAS  Google Scholar 

  14. Zocher, M., Bippes, C.A., Zhang, C. & Müller, D.J. Single-molecule force spectroscopy of G-protein-coupled receptors. Chem. Soc. Rev. 42, 7801–7815 (2013).

    Article  CAS  Google Scholar 

  15. Coughlin, S.R. How the protease thrombin talks to cells. Proc. Natl. Acad. Sci. USA 96, 11023–11027 (1999).

    Article  CAS  Google Scholar 

  16. Kobilka, B.K. & Deupi, X. Conformational complexity of G-protein-coupled receptors. Trends Pharmacol. Sci. 28, 397–406 (2007).

    Article  CAS  Google Scholar 

  17. Bernatowicz, M.S. et al. Development of potent thrombin receptor antagonist peptides. J. Med. Chem. 39, 4879–4887 (1996).

    Article  CAS  Google Scholar 

  18. Pfreundschuh, M., Alsteens, D., Hilbert, M., Steinmetz, M.O. & Müller, D.J. Localizing chemical groups while imaging single native proteins by high-resolution atomic force microscopy. Nano Lett. 14, 2957–2964 (2014).

    Article  CAS  Google Scholar 

  19. Bell, G.I. Models for the specific adhesion cells to cells. Science 200, 618–627 (1978).

    Article  CAS  Google Scholar 

  20. Evans, E., Ritchie, K. & Merkel, R. Sensitive force technique to probe molecular adhesion and structural linkages at biological interfaces. Biophys. J. 68, 2580–2587 (1995).

    Article  CAS  Google Scholar 

  21. Evans, E. & Ritchie, K. Dynamic strength of molecular adhesion bonds. Biophys. J. 72, 1541–1555 (1997).

    Article  CAS  Google Scholar 

  22. Evans, E. Energy landscapes of biomolecular adhesion and receptor anchoring at interfaces explored with dynamic force spectroscopy. Faraday Discuss. 111, 1–16 (1998).

    Article  CAS  Google Scholar 

  23. Friddle, R.W., Noy, A. & De Yoreo, J.J. Interpreting the widespread nonlinear force spectra of intermolecular bonds. Proc. Natl. Acad. Sci. USA 109, 13573–13578 (2012).

    Article  CAS  Google Scholar 

  24. Moy, V.T., Florin, E.L. & Gaub, H.E. Intermolecular forces and energies between ligands and receptors. Science 266, 257–259 (1994).

    Article  CAS  Google Scholar 

  25. Dudko, O.K., Hummer, G. & Szabo, A. Theory, analysis, and interpretation of single-molecule force spectroscopy experiments. Proc. Natl. Acad. Sci. USA 105, 15755–15760 (2008).

    Article  CAS  Google Scholar 

  26. Sulchek, T., Friddle, R.W. & Noy, A. Strength of multiple parallel biological bonds. Biophys. J. 90, 4686–4691 (2006).

    Article  CAS  Google Scholar 

  27. Bustamante, C., Marko, J.F., Siggia, E.D. & Smith, S. Entropic elasticity of lambda-phage DNA. Science 265, 1599–1600 (1994).

    Article  CAS  Google Scholar 

  28. Evans, E. Probing the relation between force—lifetime—and chemistry in single molecular bonds. Annu. Rev. Biophys. Biomol. Struct. 30, 105–128 (2001).

    Article  CAS  Google Scholar 

  29. Seifert, U. Dynamic strength of adhesion molecules: role of rebinding and self-consistent rates. Europhys. Lett. 58, 792–798 (2002).

    Article  CAS  Google Scholar 

  30. Friddle, R.W., Podsiadlo, P., Artyukhin, A.B. & Noy, A. Near-equilibrium chemical force microscopy. J. Phys. Chem. C 112, 4986–4990 (2008).

    Article  CAS  Google Scholar 

  31. Li, F., Redick, S.D., Erickson, H.P. & Moy, V.T. Force measurements of the α5-β1 integrin-fibronectin interaction. Biophys. J. 84, 1252–1262 (2003).

    Article  CAS  Google Scholar 

  32. Rosenbaum, D.M. et al. Structure and function of an irreversible agonist-β2 adrenoceptor complex. Nature 469, 236–240 (2011).

    Article  CAS  Google Scholar 

  33. Vassallo, R.R. Jr., Kieber-Emmons, T., Cichowski, K. & Brass, L.F. Structure-function-relationships in the activation of platelet thrombin receptors by receptor-derived peptides. J. Biol. Chem. 267, 6081–6085 (1992).

    CAS  PubMed  Google Scholar 

  34. Scarborough, R.M. et al. Tethered ligand agonist peptides. Structural requirements for thrombin receptor activation reveal mechanism of proteolytic unmasking of agonist function. J. Biol. Chem. 267, 13146–13149 (1992).

    CAS  PubMed  Google Scholar 

  35. Nanevicz, T. et al. Mechanisms of thrombin receptor agonist specificity. Chimeric receptors and complementary mutations identify an agonist recognition site. J. Biol. Chem. 270, 21619–21625 (1995).

    Article  CAS  Google Scholar 

  36. Feng, D.M. et al. Development of a potent thrombin receptor-ligand. J. Med. Chem. 38, 4125–4130 (1995).

    Article  CAS  Google Scholar 

  37. Morrow, D.A. et al. Vorapaxar in the secondary prevention of atherothrombotic events. N. Engl. J. Med. 366, 1404–1413 (2012).

    Article  CAS  Google Scholar 

  38. Frauenfelder, H., Sligar, S.G. & Wolynes, P.G. The energy landscapes and motions of proteins. Science 254, 1598–1603 (1991).

    Article  CAS  Google Scholar 

  39. Wolynes, P.G., Onuchic, J.N. & Thirumalai, D. Navigating the folding routes. Science 267, 1619–1620 (1995).

    Article  CAS  Google Scholar 

  40. Onuchic, J.N., Wolynes, P.G., Luthey-Schulten, Z. & Socci, N.D. Toward an outline of the topography of a realistic protein-folding funnel. Proc. Natl. Acad. Sci. USA 92, 3626–3630 (1995).

    Article  CAS  Google Scholar 

  41. Zocher, M., Zhang, C., Rasmussen, S.G.F., Kobilka, B.K. & Müller, D.J. Cholesterol increases kinetic, energetic, and mechanical stability of the human β2-adrenergic receptor. Proc. Natl. Acad. Sci. USA 109, E3463–E3472 (2012).

    Article  CAS  Google Scholar 

  42. Wildling, L. et al. Linking of sensor molecules with amino groups to amino-functionalized AFM tips. Bioconjug. Chem. 22, 1239–1248 (2011).

    Article  CAS  Google Scholar 

  43. Müller, D.J. & Engel, A. Atomic force microscopy and spectroscopy of native membrane proteins. Nat. Protoc. 2, 2191–2197 (2007).

    Article  Google Scholar 

Download references

Acknowledgements

We thank S. Weiser for SDS-PAGE; D. Martínez-Martín, R. Petrosyan, U. Hensen, S. Herzog and R. Newton for critically discussing the work; and A. Noy and R. Friddle for valuable discussions. The Swiss National Science Foundation (SNF; grant 200021_134521 to D.J.M.), ETH Zurich (grant ETH-03 14-1 to D.J.M.), the National Centre of Competence in Research (NCCR) Molecular Systems Engineering and the European Molecular Biology Organization (EMBO) (ALTF 265-2013 to D.A.) supported this work.

Author information

Authors and Affiliations

Authors

Contributions

D.A. and M.P. set up and performed the AFM experiments and developed strategies to chemically functionalize the AFM tip. D.J.M., D.A. and M.P. coanalyzed the experimental and performed calculations. C.Z. and B.K.K. provided some of the ligands and cloned, purified and reconstituted PAR1. P.M.S. performed SMFS. D.A., M.P., C.Z., S.R.C., B.K.K. and D.J.M. designed the experiments. All authors wrote the paper.

Corresponding authors

Correspondence to David Alsteens or Daniel J Müller.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Principle of force-distance (FD) curve–based AFM.

(a) FD-based AFM contours the sample surface while oscillating the AFM tip with a sine wave at a frequency of 0.25 kHz. Pixel-by-pixel the AFM tip is approached (blue curve) and retracted (red curve) from the sample. The AFM cantilever deflection measures the force interacting between AFM tip and sample. During these approach and retraction cycles the force vs time (b) and force vs distance (c) is recorded. Thereby, the maximal force (imaging force) used to touch the sample Fi is kept constant using a feedback loop. (c) The mechanical deformation or distance of deformation DDef of a soft biological sample is described by the indentation of a much stiffer AFM tip. This indentation is detected at a certain repulsive force. (d) During retraction, adhesive force FAdh is recorded between the tip and the sample. Using a functionalized cantilever FAdh can detect the rupture of specific interactions between for example a functionalized tip and sample. (e) The parameters extracted from individual force curves can be displayed as maps such as the sample topography (height image) contoured at a given imaging force, the adhesion force or sample deformation1.

1. Medalsy, I., Hensen, U. & Muller, D.J. Imaging and quantifying chemical and physical properties of native proteins at molecular resolution by force-volume AFM. Angew Chem Int Ed Engl 50, 12103-12108 (2011).

Supplementary Figure 2 SDS-PAGE of PAR1 liposomes.

SDS-PAGE was performed on a 12% acrylamide gel with 150 V and 60 min in a loading buffer (50 mM Tris, pH 7, 6% SDS, 50 mM DTT, 8% glycerol, 0.1% bromphenol blue) and stained with Coomassie blue. Lane 1, molecular weight marker; Lane 2, 10 µL of empty lipid vesicles (liposomes made of 0.5 mg mL−1 DOPC and 0.05 mg mL−1 CHS); Lane 3, 10 µL of PAR1 proteoliposomes (10 µM PAR1 reconstituted in liposomes made of 0.5 mg mL−1 DOPC and 0.05 mg mL−1 CHS). The band at 40 kDa is in good agreement with the expected size of 43.9 kDa calculated from the PAR1 sequence.

Supplementary Figure 3 SMFS of human PAR1 and human β2AR reconstituted in liposomes.

Selection of FD curves of recorded upon unfolding of single PAR1 (a) and β2AR (b) embedded in lipid membranes composed of DOPC and CHS. Superimpositions of FD curves recorded of PAR1 (c) and β2AR (d)2,3. For each GPCR the FD curves recorded were superimposed following the procedure previously described for β2AR3,4. n gives the number of FD curves superimposed. For SMFS, PAR1 proteoliposomes were adsorbed for 1 h at room temperature to freshly cleaved mica in SMFS buffer solution (300 mM NaCl, 25 mM MgCl2, 25 mM Tris, pH 7.0) and β2AR proteoliposomes were adsorbed over night at 4°C to freshly cleaved mica in SMFS buffer solution. It has been shown that the stability of human and animal GPCRs embedded in lipid membranes and adsorbed onto mica in buffer solution does not alter within 24 h3-5. Thus, samples were newly prepared before becoming 24 h old. Within this 24 h time frame we could not observe any alteration of the FD curves, which would indicate denaturation of the GPCR. For each GPCR characterized by SMFS we had to prepare at least 20 different samples. Each time a new sample has been prepared a new AFM cantilever was taken. SMFS of both GPCRs was conducted as described3 and carried out using automated AFM-based SMFS (ForceRobot 300; JPK Instruments)2. SMFS data of both β2AR and PAR1 were recorded in SMFS buffer solution, at room temperature and at pulling velocities of 300-900 nm s−1.

2. Struckmeier, J. et al. Fully automated single-molecule force spectroscopy for screening applications. Nanotechnology 19, 384020 (2008).

3. Zocher, M., Fung, J.J., Kobilka, B.K. & Muller, D.J. Ligand-specific interactions modulate kinetic, energetic, and mechanical properties of the human beta2 adrenergic receptor. Structure 20, 1391-1402 (2012).

4. Zocher, M., Zhang, C., Rasmussen, S.G., Kobilka, B.K. & Muller, D.J. Cholesterol increases kinetic, energetic, and mechanical stability of the human beta2-adrenergic receptor. Proc Natl Acad Sci U S A 109, E3463-3472 (2012).

5. Sapra, K.T. et al. Detecting molecular interactions that stabilize bovine rhodopsin. J Mol Biol 358, 255-269 (2006).

Supplementary Figure 4 Overview AFM topograph of PAR1 proteoliposome.

(a) Topography (height image, 2.5 x 2.5 µm2) showing membrane patches on mica. Proteoliposomes were adsorbed to freshly cleaved mica in buffer solution. To remove weakly attached membrane patches, the sample was rinsed several times with the buffer (see Online Methods ). After adsorption to mica the proteoliposomes break open so that they showed single-layered membrane patches. (b) Cross-section (white dashed line in (a)) showing a lipid membrane protruding 4.5 ± 0.7 nm (average ± S.D., n=10) from the supporting mica. The sparsely distributed single protrusions originating from single or clustered PAR1. The FD-based AFM topograph was recorded in imaging buffer (300 mM NaCl, 20 mM Hepes, 25 mM MgCl2, pH 7.0) at room temperature.

Supplementary Figure 5 Structural analysis of PAR1 reconstituted in proteoliposomes.

(a) Topography of PAR1s sparsely distributed in lipid membranes made of 0.5 mg mL−1 DOPC and 0.05 mg mL−1 CHS. Histogram of diameter (b) and height (c) of PAR1 particles imaged in (a). (b) The diameter distribution showed two peaks centered at 8.1 ± 1.3 nm (average ± S.D.) and 14.7 ± 0.6 nm. Diameters were measured at full-width half maximum of particle heights. (c) The height distribution showed two peaks centered at 1.2 ± 0.2 nm (average ± S.D.) and 2.0 ± 0.3 nm, which could correspond to the height of the extracellular or intracellular surface emerging from the DOPS/CHS membrane, respectively (d). The FD-based AFM topograph was recorded in imaging buffer (300 mM NaCl, 20 mM Hepes, 25 mM MgCl2, pH 7.0) at room temperature. n gives the number of PAR1 particles analyzed.

Supplementary Figure 6 AFM topograph and multiparametric maps of PAR1 reconstituted in liposomes.

(a) Topograph showing single and clustered PAR1 molecules protruding from the lipid bilayer. (b) Applied force error map showing low errors of <20 pN. (c) Adhesion map showing the SFLLRN functionalized AFM tip interacting sparsely with the lipid bilayer and mainly with PAR1 (see (a)). (d) Deformation map showing enhanced deformation values of PAR1 molecules. The FD-based AFM data was recorded as described (Online Methods ).

Supplementary Figure 7 Mapping the binding of the SFLLRN ligand to human PAR1 proteoliposomes using FD-based AFM.

Topographs (left column) of human PAR1 reconstituted in proteoliposomes taken with the SFLLRN-ligand functionalized AFM tip oscillated at 0.25 kHz and amplitude of 50 nm. As described, the SFLLRN has been attached via the PEG-polypeptide linker to the AFM tip (Fig. 1c). Corresponding adhesion maps (right column). To increase their visibility adhesive pixel were enlarged by a factor 2. The FD-based AFM data was recorded as described (Online Methods ).

Supplementary Figure 8 Validating that SFLLRN ligand–functionalized AFM tips detect specific interactions with PAR1.

Height images (a,c,e,g) and corresponding adhesion images (b,d,e,h) recorded with either (a,b) a bare AFM tip, (c,d) a hexa-glycine, (e,f) a scrambled peptide (FLLNSR) or (g,h) a SFLLRN functionalized tip in the presence of peptide mimetic antagonist (1 µM BMS). Every functional group (hexa-glycine, FLLNSR, or SFLLRN) was attached via the same PEG-polypeptide linker to the AFM tip (Fig. 1c). This supplementary figure correlates to Fig. 2, which maps receptor-ligand interactions on PAR1 using SFLLRN functionalized AFM tips. The FD-based AFM data was recorded as described (Online Methods ).

Supplementary Figure 9 Determining the effective spring constant keff of the cantilever-PEG-polypeptide linker system.

(a) The effective spring constant keff is the combination of the cantilever stiffness kc and the stiffness kL of the PEG-polypeptide linker. As described (Supplementary Note), kL was estimated using the combination of the elasticity of the PEG linker (kPEG) and the elasticity of the polypeptide linker (kpolypeptide). The linker consists of the 27-unit long PEG spacer and the 28 aa long polypeptide, which mimics the thrombin cleaved N-terminal end of PAR1 (comp. Fig. 1). (b) The extension of the PEG-polypeptide linker applied to a stretching force F in water is described by the combination of both the PEG elasticity model6 and the WLC model7. (c) Calculated stiffness kL of the PEG-polypeptide linker (blue curve) vs extension and effective stiffness keff of cantilever and linker (red curve). (d) The force vs keff relationship was used to determine the effective spring constant of cantilever-PEG-polypeptide linker system exposed to a given force Feq and to subsequently calculate ΔGbu.

6. Sulchek, T., Friddle, R.W. & Noy, A. Strength of multiple parallel biological bonds. Biophys. J. 90, 4686-4691 (2006).

7. Bustamante, C., Marko, J.F., Siggia, E.D. & Smith, S. Entropic elasticity of lambda-phage DNA. Science 265, 1599-1600 (1994).

Supplementary Figure 10 Validating the elastic potential of the PEG-polypeptide linker attaching the ligand to the AFM tip.

Shown are individual FD curves, each one detecting the rupture of a specific SFLLRN-PAR1 bond. Each of these representative FD curves was fitted using the effective spring constant keff of the cantilever-PEG-polypeptide linker system (Supplementary Fig. 9). The individual fits were shifted along the distance axis to minimize the sum of squared residuals between the fit and the adhesion force peak. The ∆d values given correspond to the shifted distance. FD curves were recorded as described (Fig. 2) using a SFLLRN functionalized AFM cantilever and imaging a PAR1 proteoliposome by FD-based AFM.

Supplementary Figure 11 The position of the PEG-polypeptide linker bound to the tip affects the distance of the force peak stretching the linker.

The stretching of the PEG-polypeptide linker under force is described combining the model describing the PEG elasticity and the WLC model describing the polypeptide elasticity (see Supplementary Note). Independent of the position at which the PEG-polypeptide linker is chemically anchored to the AFM tip the force-distance characteristics describing the stretching of the linker is always similar since in all cases the linker is bound via one terminal end to the AFM tip and with the other terminal end adheres to the receptor. However, due to the vertical position at which the linker is anchored to the tip, the first part of the force-distance characteristics describing the stretching of the PEG-polypeptide linker can be hidden by relative distance "∆d" value (dashed line).

Supplementary Figure 12 Influence of the AFM tip drive frequency on ligand-receptor unbinding force.

DFS plot showing that the tip drive frequency (0.25 kHz, red or 0.5 kHz, blue) has no influence on the force required to separate SFLLRN-ligand and PAR1. Although the higher oscillation frequency reduced the contact time between tip and sample and, thus, lowered the frequency of bonds formed between ligand and PAR1, we could record a sufficient amount of specific unbinding events. Data were recorded in imaging buffer (300 mM NaCl, 20 mM Hepes, 25 mM MgCl2, pH 7.0) at room temperature and fitted using the Friddle-Noy-de Yoreo model (thin red line)8. Each circle represents one measurement. Darker red shaded areas represent 99% confidence intervals and lighter red shaded areas represent 99% of prediction intervals.

8. Friddle, R.W., Noy, A. & De Yoreo, J.J. Interpreting the widespread nonlinear force spectra of intermolecular bonds. Proc Natl Acad Sci U S A 109, 13573-13578 (2012).

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–12, Supplementary Table 1 and Supplementary Note (PDF 7786 kb)

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Alsteens, D., Pfreundschuh, M., Zhang, C. et al. Imaging G protein–coupled receptors while quantifying their ligand-binding free-energy landscape. Nat Methods 12, 845–851 (2015). https://doi.org/10.1038/nmeth.3479

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nmeth.3479

This article is cited by

Search

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