Molecular Movies for Calcium Ion Imaging
New Femtosecond Raman Technology Reveals Chemistry in Action
- Molecular Movies for Calcium Ion Imaging - New Femtosecond Raman Technology Reveals Chemistry in Action
- Fig. 1: (Left) The Ca2+ biosensor, GEM-GECO1, consists of cpGFP, green or blue shaded; CaM, shown in red sticks; and M13 peptide, shown in blue stick. The chromophore is highlighted by the orange box. (Right) The photoexcited chromophore navigates multidimensional PES of FPs, exhibiting characteristic structural motions to gate the fluorescence outcome.
- Fig. 2: (Left) Energy diagram in FSRS. Following photoexcitation (violet), the concurrent Raman pump (green) and probe (red) generates the vibrational coherence that relaxes in the excited state (Right) Pulse sequence and typical pulse parameters in FSRS.
- Fig. 3: New ways to generate a broadband tunable Raman probe for FSRS. Tunability is indicated by the asterisks.
Fluorescent proteins have revolutionized molecular and cellular biology by offering the genetically encodable and versatile biosensors to image life processes. However, the fluorescence mechanisms of these biosensors remains elusive. Using femtosecond stimulated Raman spectroscopy, we reveal the structural dynamics basis of a new calcium ion (Ca2+) biosensor by capturing images of one of the fastest chemical reactions in proteins, namely, excited state proton transfer, starting from time zero.
Imaging biological processes with ultrahigh spatial and temporal resolutions has been a dream for biophysicists, biochemists, and molecular biologists alike [1, 2]. Visualization of atomic choreography leads to a mechanistic understanding of biological functions, which can enable the rational design and control of cellular pathways to improve human health and fight diseases. This article presents an emerging structural dynamics tool, femtosecond stimulated Raman spectroscopy (FSRS) [3, 4], which allows us to capture a molecular "movie" of the chromophore inside a fluorescent protein biosensor . This ratiometric biosensor called GEM-GECO1  is part of an expanded palette evolved from genetically encoded Ca2+ indicators , which have become indispensible tools of cell biology, particularly for neural imaging . There remains room to improve these biosensors to have higher color contrast, larger dynamic range, or faster kinetic response.
Excited State Proton Transfer in Fluorescent Proteins
For bioimaging, fluorescent proteins (FPs) are mostly engineered via directed evolution from marine organisms including jellyfish and coral reefs  with a three-residue chromophore like Ser-Tyr-Gly in wtGFP (fig. 1). The dominant pathway leading to fluorescence of FPs has been proposed to involve excited state proton transfer (ESPT), where the proton at the phenolic hydroxyl end of the chromophore moves along an ESPT chain toward the other end. However, due to its small mass and fast speed, seeing transient proton movement has been extremely challenging. Moreover, imaging chemical reactions in a functional biomolecule in the electronic excited state is nontrivial .
Typical microscopy images cellular components using an objective lens and diode array or CCD camera  at video rate (~60 frames/second).
At the molecular level and timescale, to obtain structural evolution of the target system performing biofunctions, the reaction needs to be precisely initiated. To achieve high spatial resolution, the detection scheme needs to be sensitive to atomic motions. To achieve high temporal resolution, the detection setup needs to resolve molecular vibrations, typically on the femtosecond (fs, 10-15 s) timescale. Currently there is no detector with that speed for imaging.
FSRS provides a viable solution to this challenge, providing simultaneously high spatial and temporal resolutions to record sharp spectral features (fig. 2). Equilibrium structures in the ground state have characteristic vibrational features that can be exploited for high-sensitivity, label-free imaging in situ . The synchronization of the commercially available CCD camera with a table-top femtosecond laser amplifier at 1 kHz repetition rate, and a phase-stable optical chopper in the Raman pump beampath at 500 Hz, records one spectrum in 2 ms. As a result, thousands of data traces can be collected and averaged in seconds to increase the signal-to-noise ratio. In contrast to continuous-wave Raman, the stimulated Raman scattering photons in FSRS only emit in the probe direction, enabling measurements on highly fluorescent molecules such as FPs. The Raman probe pulse is an fs supercontinuum white light so a broad range of vibrational coherences can be observed at once.
The incorporation of an fs actinic pump pulse readily extends the FSRS detection to the excited state (fig. 2). Because the detected signal can only be generated upon the simultaneous arrival of the ps Raman pump and fs probe, the time resolution is determined by the time delay between the preceding actinic pump and the Raman probe, both fs pulses. This is crucial to resolve atomic motions, because a typical bond stretch has a ~20 fs period. As the photoexcited wavepacket navigates the multidimensional potential energy surface (PES) to relax energy and transfer proton (fig. 1), characteristic atomic motions will be activated to facilitate the reaction.
In contrast to preparing a protein crystal to solve for structure, FSRS captures structural snapshots following photoexcitation in physiologically relevant environments. In contrast to retrieving low-frequency modes that modulate the electronic response in transient absorption, FSRS monitors multiple vibrational modes and directly observes anharmonic couplings between them (with the impulsively excited coherent skeletal motions in the first 1-2 ps [4, 5]), dissecting PES along the reaction coordinate. Besides measuring the reactant and product, FSRS exposes nuclear motions in between, particularly prior to main photochemical events.
Imaging Key Structural Motions Inside a Biosensor
The newly developed ratiometric biosensor for Ca2+ imaging, GEM-GECO1 (fig. 1, left), consists of a chimera of circularly permutated (cp)GFP, calmodulin (CaM), and a peptide derived from myosin light chain kinase (M13). Upon UV excitation, it fluoresces green in the absence of Ca2+ but blue upon Ca2+ binding with a Kd of 340 nM . It was speculated that ESPT plays a role in this unusual color change because typical FP-based Ca2+ biosensors operate by intensity change . The power of FSRS is evident from the time-resolved excited-state Raman spectrum starting from time zero for photoexcitation : dramatic differences are revealed that provide a comprehensive picture of proton motions from the initial stage of searching phase space to reaching the transition state of either trapping (blue) or ESPT (green). The crucial time window spans from tens of fs to several ps, when key skeletal motions are coherently excited and strongly coupled to other degrees of freedom to determine the fluorescence outcome. The insights may establish the structure-function relationship to guide rational design of improved FPs from the bottom up.
FSRS is an emerging structural dynamics methodology that has a set of unique advantages to answer some of the most pressing questions in photochemistry and life processes. Besides its sensitivity to molecular structures at the chemical bond level, FSRS can capture transient conformational motions in the excited state from time zero. The collection of time-resolved structural snapshots of the photoexcited chromophore inside an FP biosensor for Ca2+ imaging is reminiscent of creating a molecular "movie" with a proton as the main "actor". With technical innovations including new ways to generate the Raman probe (fig. 3), we can extend the FSRS detection range to the low-frequency domain to directly monitor molecular skeletal motions evolving during a chemical reaction. This may provide FSRS more imaging power to further elucidate missing links of the dynamics in life processes.
Drs. Robert E. Campbell and Yongxin Zhao are thanked for materials, and Dr. Weimin Liu and Ms. Breland Oscar for instrumentation and discussion. Oregon State University Faculty Startup Fund provides the resources for FSRS development and implementation.
 Hochstrasser R. M.: Proc. Natl. Acad. Sci. USA 104, 14190-96 (2007)
 Miller R. J. D.: Science 343, 1108-16 (2014)
 McCamant D. W. et al.: Rev. Sci. Instrum. 75, 4971-80 (2004)
 Fang C. et al.: Nature 462, 200-04 (2009)
 Oscar B. G. et al.: Proc. Natl. Acad. Sci. USA 111, 10191-96 (2014)
 Zhao Y. et al.: Science 333, 1888-91 (2011)
 Miyawaki A. et al.: Nature 388, 882-87 (1997)
 Grienberger C. et al.: Neuron 73, 862-85 (2012)
 Ai H.-W. et al.: Nat. Protocols 9, 910-28 (2014)
 Freudiger C. W. et al.: Science 322, 1857-61 (2008)
Dr. Chong Fang (Corresponding author via e-mail request button below)
Department of Chemistry
Oregon State University