A New Light In Biosensing: Engineering Photostable GFP Mutants for Fluorescence Resonance Energy Transfer (FRET)


  • Share
  • Share URL
  • menu
  • <
  • >
  • 1Short Project Description 
  • 2Summary 
  • 3About MeAbout Our Team 
  • 4Question / Proposal 
  • 5Research 
  • 6Method / Testing and Redesign 
  • 7Results 
  • 8Conclusion / Report 
  • 9Bibliography, References and Acknowledgements 

Biosensors employing fluorescence resonance energy transfer (FRET) between fluorescent proteins are powerful tools for non-invasively monitoring intracellular processes. Clover, the brightest existing fluorescent protein, suffers from low photostability, reducing its utility for time-lapse imaging. I developed a new fluorescent protein, Clover2S, which is now the brightest existing fluorescent protein. Clover2S, which differs from Clover by N149Y and G160S, shows increased photostability, quantum yield, and maturation, and confers increased dynamic range onto the calcium biosensor. Clover2S’s incorporation into a variety of biosensors is a promising avenue to elucidate the mechanisms behind cancer and neurological pathways.










Biosensors employing fluorescence resonance energy transfer (FRET) between fluorescent proteins (FPs) are powerful tools for non-invasively monitoring intracellular processes. Clover, the brightest existing FP, suffers from low photostability, reducing its utility for time-lapse imaging. The goal of this experimentation is to improve the photostability of Clover, while preserving its other optimized properties. 


Libraries of Clover mutants were screened for photostability, efficient FRET donation to a red FP acceptor, and performance in a FRET-based calcium sensor. Results show that Clover2S, a new mutant that differs from Clover by N149Y and G160S, shows increased photostability, quantum yield, and maturation, and confers increased dynamic range onto the calcium biosensor. The TN-XXL calcium biosensor was further improved through lengthening the linker sequence in between the FRET pair. Mapping the mutations onto the Clover structure implies that the mechanism of increased photostability may involve preventing oxygen from diffusing into Clover and reacting with the chromophore.


Clover2S is the now the brightest existing fluorescent protein to date. Clover2S’s incorporation into a variety of biosensors is a promising avenue to elucidate the mechanisms behind cancer and neurological pathways.

Photo of EmilyMy favorite scientist Louis Pasteur once observed, “Chance favors the prepared mind.”

Everyday after school, I leap down Pasteur Drive to Stanford University’s bioengineering lab, working to develop bright, photostable fluorescent proteins for biosensors. 

Working in a lab for the past year has taught me the diligence behind discoveries.

I learned that passion for research is persevering despite no significance in my findings, staying up past midnight reading PubMed papers, and transforming bacteria during the holidays.

At Henry M. Gunn High School, I am the President of Research Science & Invention Club (RSI), where students initiate their own creative projects. Moreover, I am leading a Lemelson-MIT InvenTeam, receiving a $10,000 grant to build a solar incubator for developing nations.

In last year's Google Science Fair, my brother Trevor and I were selected as Regional Finalists of the Americas; we investigated the anti-cancer effect of propolis.

As of 2013, the culmination of over 500 hours of lab work, my breakthrough project has already been recognized at the regional, state, national, and international levels, including:

  1. National Winner and International Finalist of BioGENEius Challenge
  2. Synopsys N+1 Breakthrough Award
  3. First Place at Synopsys Science & Engineering Championshionship
  4. First Place at California Bay Area BioGENEius Challenge
  5. Dupont Biosciences Award
  6. California State Science Fair Biochemistry Category Award
  7. UCLA Brain Research Institute Award

Winning the Google Science Fair would allow me to interact with other motivated scientists. Through my research, I discovered my ultimate goal to develop into a leader in the medical field.

While preserving its other desired characteristics, how can the photostability of the brightest fluorescent protein Clover be improved?

      Biosensors employing fluorescence resonance energy transfer (FRET) are powerful tools for investigating intracellular processes, with important applications towards cancer and neurological research. In 2012, Clover and mRuby2, a new fluorescent protein FRET pair was developed. Clover and mRuby2 shows the greatest dynamic range across a variety of biosensors, but the donor fluorescent protein Clover suffers from low photostability. Even though fluorescent proteins are more photostable than chemical dyes, one of the major problems limiting time-lapse imaging is the donor fluorescent protein’s lack in photostability. The goal of this experimentation is to improve the photostability of Clover, while preserving its other optimized properties, such as its brightness. 

    It was hypothesized that through random mutagenesis, mutations in amino acids near the chromophores would lead to greater photostability. In order to create a new photostable fluorescent protein, I will first conduct random mutagenesis to create a library of Clover mutants. Next, I will screen the libraries of mutants for photostability, efficient FRET donation to a red FP acceptor, and performance in a FRET-based calcium sensor. Then, I will repeat the screening processfor several rounds in order to develop a novel fluorescent protein and finally, incorporate the new fluorescent proteins into a FRET biosensor.

     In 2008, Dr. Shimomura, Dr. Chalfie, and Dr. Tsien received the Nobel Prize in Chemistry for discovering and developing the green fluorescent protein (GFP). Ever since, the green fluorescent protein has been mutated successively to produce a wide range of fluorescent proteins with various optimized properties. A myriad of non-invasive molecular probes have been developed for live-cell imaging using fluorescent proteins.

Emily conducting a PCRFluorescent proteins  ‚Äč                              

     Fluorescence resonance energy transfer, or FRET, is a distance-dependent transfer of energy from a donor fluorescent protein to an acceptor fluorescent protein. A donor fluorescent protein in its excited state emits a characteristic wavelength to excite the acceptor fluorescent protein through non-radiative dipole-dipole coupling. FRET efficiency is inversely proportional to the sixth power of the distance between the donor and acceptor fluorescent proteins, enabling high FRET sensitivity towards small fluorescent protein orientation and distance changes.

Diagram of FRET biosensor

     Moreover, FRET has been widely applied to develop fluorescent resonance energy transfer biosensors to visualize cellular activities at the molecular level with high spatiotemporal resolution. A biosensor is defined as "a detection system that relies on a biomolecule for molecular recognition and a transducer to produce an observable output."  These FRET biosensors typically comprise of a donor fluorescent protein, sensing domain, linkers, and acceptor fluorescent protein. When the sensing domain detects an environmental stimuli, there is a conformation change which causes the two fluorescent proteins to either become closer or farther apart, enabling or disabling FRET. Biosensors employing FRET are valuable and powerful tools for monitoring intracellular signals, such as pH changes, calcium wave induction, phosphorylation, reduction-oxidation reactions, and apoptosis.  

CloverClover-2Clover fluorescence

     Recently, researchers at the University of Illinois developed a new FRET biosensor with the ability to detect cancerous and drug-resistant cells. At the University of California, San Diego, researchers created a new FRET biosensor which monitors Src tyrosine activity to diagnose cervical cancer and screen for new anti-cancer drug candidates. The application of fluorescent proteins and FRET to cancer research has been an important breakthrough for elucidating the mechanism behind cancer pathways. 

     Clover, a new green fluorescent protein developed in 2012, shows increased dynamic range across a variety of biosensors. Moreover, Clover also improves upon the brightness of its parent fluorescent protein EGFP (enhanced green fluorescent protein). However, Clover suffers from low photostability, which is detrimental to time-lapse imaging.



Library of Mutants

Constructing Library of Mutants

            I ran a series of mutagenic polymerase chain reactions (PCR) using a low-fidelity DNA polymerase, Mutazyme II, in order to mutate the template Clover DNA. Simultaneously, I used restriction enzymes, BSLII and XHO1, to cut a vector. Then, I conducted gel electrophoresis and recovery to separate the DNA bands, which were ligated using In-Fusion, then transformed into bacteria after a series of cold and heat shocks.


     Microscope       I screened for photostability in three different contexts. First, I took a pre-fluorescence image using CDOps Lite, photobleached the plates with around 1,000 colonies using a LED array, and took a post-fluorescence image. Using the software Image J, I ratiometrically analyzed the post-pre ratios to identify photostable colonies.

          Next, I made patches to isolate photostable colonies and extracted the bacterial protein. I bleached the lysates on a 96-well plate using a LED array and measured fluorescence change using a plate reader.

            Finally, the lysates of the top mutants were purified using His-Pur beads, dispersed in mineral oil, and photobleached using a high-power microscope at 100% neutral density. Fluorescence was recorded through time-lapse imaging. An image analysis software, FIJI, was used to find values to graph photostability curves.

Characterizing and Sequencing Characterizing and sequencing flowchart

            I then conducted DNA minipreps, by lysing and washing the bacteria with buffers, extracting the mutants’ DNA for sequencing. Afterwards, the mutants were characterized to ensure their brightness was maintained with the increased photostability.

            To measure the extinction coefficient, I measured the absorbance of the native FP compared the absorbance of the denatured FP. Using an online server, I also used the native and denatured FP absorbance values to calculate the maturation percent. For the quantum yield measurement, I took the absorbance and emission spectra of the FP, PBS, and a contro, fluoroscein.

Testing in FRET Pairs and Calcium Sensors

           I ran a PCR in order to attach the new photostable mutants to a red FP, mCrimson, to form a FRET pair. The new FRET pairs were isolated in patches and made into lysates, then tested for FRET efficiency using a plate reader to excite the Clover FP and record the emission of mCrimson. Then, I ran another PCR to incorporate the new Clover mutants into a TN-XXL calcium sensor. The sensor was tested in two conditions: with either calcium or EGTA (baseline), recording down emission in both states using a plate reader. The calcium sensor was then further optimized through a series of overlap PCRs to introduce point mutations to lengthen the linker sequence.

My experimental results can be highlighted through eight major findings:

Major Finding #1 Graph: All three initial mutants showed increased photostability over Clover.

Major Finding #2: Initial mutants also improve EC and QY.

Major Finding #3: Mutant Clover2 is 2-fold more photostable than Clover.

Major Finding #4: Mutant Clover2S is more photostable and bright. (Graph)

Monomeric mutation leaves Clover mutants photostable. (graph)

New FRET pairs have improved FRET efficiency. (graphs)

All sensors maintain or improve dynamic range. (table)

The new TN-XXL sensor with a lengthened linker shows an increase in dynamic range.


Clover2S, the mutant I engineered, is now the brightest existing fluorescent protein, as well as the most photostable Clover variant.


Clover2S , which differs from Clover by N149Y and G160S, shows increased photostability, quantum yield, maturation, and dynamic range in a biosensor. Clover2S's combination of improved aspects will greatly benefit time-lapse imaging. Moreover, this new fluorescent protein can improve the sensitivity of FRET biosensors, allowing for more accurate detection of subtle changes.

The development of Clover2S has a multitude of medical applications, including use for brain imaging, visualizing cancer cells, and elucidating the mechanism behind neurological disorders.

Further Research

Clover2S can be crystallized to better understand the mechanism behind its increased photostability and maturation rate. In addition to uncovering its structure, Clover2S will also be incorporated into a variety of biosensors to monitor kinase activity, protein-protein interactions, and protein conformation changes, as well as as to visualize activity in cancer-related pathways. Furthermore, Clover2S will be utilized to optimize time-lapse imaging.

Summary of findings.

This independent research project was carried out by Emily Wang under the guidance of Dr. Jun Chu at Stanford University from August 2012 – Present.

Many thanks to Professor Michael Lin and Dr. Jun Chu of Stanford University for their wholehearted support: from teaching me biotech techniques, to supervising me after school, on weekends, and holidays.

Thank you to Katherine Moser, AP Biology teacher, and Elana Zizmor, AP Chemistry teacher, for insight and encouragement.

I would like to further extend my thanks to Assay Depot and BioCurious for awarding me a grant for my proposal to fund my research. 


  1. Bacskai, B. J., Hochner, B., Mahaut-Smith, M., Adams, S. R., Kaang, B. K., Kandel,E. R. and Tsien, R. Y. Spatially resolved dynamics of cAMP and protein kinase A subunits in aplysia sensory neurons. Science 260: 222-226 (1993).

  2. Ballestrem, C., Erez, N., Kirchner, J., Kam, Z., Bershadsky, A. and Geiger, B. Molecular mapping of tyrosine-phosphorylated proteins in focal adhesions using fluorescence resonance energy transfer. Journal of Cell Science 119: 866-875 (2006).

  3. Bayle, V.Nussaume, L and Bhat, R. A. Combination of novel green fluorescent protein mutant T-Sapphire and DsRed variant mOrange to set up a versatile in planta FRET-FLIM assay. Plant Physiology 148: 51-60 (2008).

  4. Berney, C. and Danuser, G. FRET or no FRET: a quantitative comparison. Biophysical Journal  (2003).

  5. Chapagain, Prem. Fluorescent protein barrel fluctuations and oxygen diffusion pathways in mCherry. The Journal of Chemical Physics 13 (2011).

  6. Chan, F. K. M., Siegel, R. M., Zacharias, D., Swofford, R., Holmes, K. L., Tsien, R. Y. and Lenardo, M. J. Fluorescence Resonance Energy Transfer analysis of cell surface receptor interactions and signaling using spectral variants of the green fluorescent protein.Cytometry A44: 361-368 (2001).

  7. Chen, Y. and Mills, J. D. Protein localization in living cells and tissues using FRET and FLIM. Differentiation 71: 528-541 (2003).

  8. Chiu, Y. L., Cao, H. and Rana, T. M. Quantitative analysis of RNA-mediated protein-protein interactions in living cells by FRET. Chemical Biology and Drug Design 69: 233-239 (2007).

  9. Dale, R. E., Eisinger, J. and Blumberg, W. E. The orientational freedom of molecular probes The orientation factor in intramolecular energy transfer. Biophysical Journal 26: 161-194 (1979).

  10. Day, R., N. and Piston, D. W. Spying on the hidden lives of proteins. Nature Biotechnology 17: 425-426 (1999).

  11. Day, R. N. and Booker, C. F. and Periasamy, A. Characterization of an improved donor fluorescent protein for Förster resonance energy transfer microscopy. Journal of Biomedical Optics 13: 0311203 (2008).

  12. Gordon, G. W., Berry, G., Liang, X. H., Levine, B. and Herman, B. Quantitative fluorescence resonance energy transfer measurements using fluorescence microscopy. Biophysical Journal  (1998).

  13. Hazelwood, K. Fluorescent protein FRET pairs for ratiometric imaging of dual biosensors. Nature Methods 5 (2008).

  14. Kenworthy, A. K. Imaging protein-protein interactions using fluorescence resonance energy transfer microscopy. Methods (2001).

  15. Lam, A. et al. Improving dynamic range with bright green and red fluorescent proteins. Nature Methods (2012).

  16. Lu, Shaoying and Wang, Yingixiao. FRET Biosensors for Cancer Detection and Evaluation of Drug Efficacy. Clinical Cancer Research  (2010).

  17. Nikon. “Fundamental Principles of FRET Microscopy with Fluorescent Proteins.” MicroscopyU: The Source for Education. Web. 17 Jan. 2013. <http://www.microscopyu.com/articles/fluorescence/fret/fretintro.html>.

  18. Ormo, M. et al. Crystal structure of Aequorea victoria green fluorescent protein. Science  273, 1392-1395 (1996).

  19. Ouyang M., Sun J., Chien S., Wang Y. Determination of hierarchical relationship of Src and Rac at subcellular locations with FRET biosensors. Proceedings of the National Academy of Sciences. (2008). 

  20. Shaner, N.C. et al. Improving the photostability of bright monomeric orange and red fluorescent proteins. Nat. Methods 5, 545-551 (2008).

  21. Tsien, Roger. "The green fluorescent protein.” Annu Rev Biochem (1998). Print.

  22. Tsutsui, H. et al. Improving membrane voltage measurements using FRET with new fluorescent proteins. Nat. Methods 5, 683-685 (2008).

  23. Wu, X., Simone, J., Hewgill, D., Siegel, R. Lipsky, P. E. and He, L. Measurement of two caspase activities simultaneously in living cells by a novel dual FRET fluorescent indicator probe.Cytometry 69A: 477-486 (2006).

  24. Xia, Z. and Liu, Y. Reliable and global measurement of fluorescence resonance energy transfer using fluorescence microscopes. Biophysical Journal 81: 2395-2402 (2001).

  25. Xu, X., Soutto, M., Xie, Q., Servick, S., Subramanian, C., von Arnim, A. G. and Johnson C. H.Imaging protein interactions with bioluminescence resonance energy transfer (BRET) in plant and mammalian cells and tissues. Proceedings of the National Academy of Sciences (USA) 104: 10264-10269 (2007).

  26. Yang, X., Xu, P., Xu, T. A new pair for inter- and intra-molecular FRET measurement.Biochemical and Biophysical Research Communications 330: 914-920 (2005).

  27. Yasuda, R. Imaging spatiotemporal dynamics of neuronal signaling using fluorescence resonance energy transfer and fluorescence lifetime imaging microscopy. Current Opinion in Neurobiology 16: 551-561 (2006).

  28. You, X., Nguyen, A. W., Jabaiah, A., Sheff, M. A., Thorn, K. S. and Daugherty, P. S. Intracellular protein interaction mapping with FRET hybrids. Proceedings of the National Academy of Sciences (USA) 103: 18458-18463 (2006).

Back to Dashboard