Research-Super-Nonlinear Fluorescence Microscopy

It is extremely desirable to be able to probe biological activities deep inside living organisms. Harnessing a nonlinear excitation scheme, two-photon fluorescence is the most successful optical microscopy for this endeavor. Despite its popularity in broad biomedical applications, a fundamental imaging depth limit, accompanied by the inevitable background excitation, still exists for two-photon fluorescence microscopy when imaging scattering samples (illustrated in Fig. 1). Essentially, the conventional optical sectioning picture breaks down when approaching the fundamental depth limit, as the exponential increase of the incident laser power eventually outstrips the power-law fall-off of the excitation efficiency. Quantitatively, the depth where the in-focus signal and the out-of-focus background are equal to each other is defined as the fundamental imaging-depth limit.


Fig. 1. The fundamental imaging-depth limit of two-photon fluorescence microscopy. (a) Under a constant two-photon excitation power, volume images of a turbid gel sample homogeneously doped with fluorescent beads exhibit a decreasing fluorescence signal over depth. (b) Much deeper images can be acquired for the same sample when using compensating higher laser powers in deeper layers. However, increasingly reduced signal-to-background contrast was observed. (c) The reduced image contrast when imaging deep inside the scattering sample is due to the inevitable excitation of the fluorophores that are out of the focus.

We have conceived and realized an entire class of super-nonlinear fluorescence microscopy rendered by new spectroscopic transitions including photoactivation, photoswitching, stimulated emission, ground state depletion and frustrated FRET. Conceptually, unlike conventional multiphoton processes mediated by transient virtual states, our strategy constitutes a new class of fluorescence microscopy where high-order nonlinearity is mediated by real population transfer [23]. Application-wise, this class of super-nonlinear fluorescence microscopy has the prospect to image scattering samples with unprecedented contrast and depth penetration.

(1) Multiphoton activation and imaging (MPAI) employs photoactivatable probes which are initially in the dark state throughout the entire tissue volume. The subsequent two-photon activation will switch on a drastically higher percentage of photoactivatable probes at focal volume than those out of focus. As illustrated in Fig. 2 (c), the resulting spatial disparity of dark-to-bright transitions would lead to a significantly decreased background in the final two-photon imaging. Physically, the two-photon activation process and the subsequent two-photon imaging process accumulate together to make MPAI behave as an overall four-photon process.


Fig. 2. Principle of multiphoton activation and imaging (MPAI). (a) Imaging transparent samples. (b) Imaging deep into scattering samples. (c) When imaging with photoactivatable probes which are originally in the dark state, the multiphoton activation will switch on a higher percentage of probes at focus than those out-of-focus. Such a spatial disparity of dark-bright transitions would lead to a drastically reduced background fluorescence in the subsequent imaging. MPAI is physically a four-photon process.

We have recently demonstrated MPAI using genetically encodable photo-activatable green fluorescent protein (pa-GFP), which is a prototype of the molecular switch probes. As shown in Fig. 3 (a), pa-GFP can be turned on from its initial dark state by two-photon activation in the 750~850 nm range, and the resulting bright state can be readily imaged in the 900~950 nm range. We prepared tissue phantoms by embedding the E. coli cells expressing fluorescent protein into 3D agarose gel. The resulting sample is highly scattering due to the densely packed E. coli cells in 3D. As shown in Fig. 3 (b) for the control sample, when imaging at 100 µm, the out-of-focus background is completely overwhelming the focal signal for cells expressing regular GFP. In contrast, for the test sample in Fig. 3 (c), MPAI of cells expressing pa-GFP at the same 100 µm depth displays a much improved image contrast (S/B more than 50) after the initial activation at 830-nm and the subsequent imaging at 920-nm. Thus we have experimentally demonstrated the principle of MAPI in promoting the S/B contrast, which essentially breaks the imaging-depth limit of two-photon microscopy.


Fig. 3. Experimental demonstration of MPAI in tissue phantoms with pa-GFP. (a) pa-GFP could be activated by a pulsed laser at 830 nm to its bright state, which could be further excited by a 920 nm pulsed laser to emit fluorescence. (b, c) Deep imaging comparison of 3D turbid samples made of E. coli cells expressing free regular GFP (b) or pa-GFP (c) embedded in 2% agarose gel with the same cell densities. While out-of-focus background is overwhelming when imaging E. coli expressing regular GFP at a 100 µm depth inside the gel, MPAI with pa-GFP at the same depth offers a much improved image contrast.

(2) Multiphoton deactivation and imaging (MPDI) employs photoswitchale probes which are initially in the bright state throughout the tissue volume. After the subsequent two-photon-induced bright-to-dark switching, the in-focus probes will be switched off much more than those out of focus, creating a disparity of dark-bright states in space. The resulting differential image before and after the two-photon-induced deactivation would generate a pseudo background-free contrast which effectively behaves as an overall four-photon process.

We chose to use Dronpa-3 for demonstration, as it exhibits a high quantum yield for its bright-to-dark photoswitching. We demonstrate MPDI on tissue phantoms made of live bacterial or mammalian cells. Dronpa-3 expressing E. coli cells are embedded in 3D low melting-point agarose gel (2%). As demonstrated in Fig. 4 (a), the out-of-focus background is substantial for a quick pre-switching scan at a depth of 250 µm. The corresponding near-unity S/B indicates that this depth is close to the fundamental imaging-depth limit for regular fluorophores. After performing one deactivation scanning at 920nm, the post-deactivation image indeed became dimmer. The differential MPDI image offers a satisfactory image contrast with much reduced out-of-focus background. MPDI in mammalian cells was successfully shown in Fig. 4 (b). Thus we proved MPDI as a viable approach for high-contrast deep tissue imaging.

Fig. 4. Demonstration of multiphoton deactivation and imaging (MPDI) in tissue phantoms with photoswitchable Dronpa-3. (a) For Dronpa-3 expressing E. coli cells packed in 3D, the regular pre-switching image (at 920 nm) is overwhelming at a depth of 250 µm. After performing a deactivation scanning, the post-deactivation image is dimmer. The difference image (after auto-scaled) offers a much improved image contrast. (b) Similar contrast improvement is observed for HEK 293T cells (transfected by H2B-Dronpa-3 plasmids) placed on a dense layer of scattering E. coli cells expressing Dronpa-3.


(3) We then derived theoretically and showed experimentally the higher-order (more than quadratic) nonlinear intensity dependence of both the MPAI process (with PA-GFP) and the MPDI process (with Dronpa-3). As shown in Fig. 5, regarding MPAI, the experimental fluorescence signal from PA-GFP expressing E. coli cells shows a 3.4-order nonlinear power dependence on the geometric mean of the 830-nm laser power (for activation) and 920-nm laser power (for imaging). As for MPDI, the experimental signal from Dronpa-3 expressed E. coli cells shows a 3.8-order nonlinear power dependence on the applied power range of the 920-nm laser (which serves as both deactivation and imaging). These measured values of nonlinearity orders are close to the theoretical prediction of an overall nonlinearity order of 4.

Fig. 5. Super-nonlinear dependence of MPAI and MPDI signals on the incident laser intensity. (Left) For MPAI, the experimental fluorescence signal from pa-GFP expressing E. coli cells shows a 3.4-order nonlinear power dependence on the geometric mean of the 830-nm (for activation) and 920-nm (for imaging) laser powers. (Right)For MPDI, the experimental signal from Dronpa-3 expressed E. coli cells shows a 3.8-order nonlinear power dependence on the applied range of the 920-nm (for both deactivation and imaging) laser power.


(4) To generalize the concept of super-nonlinear microscopy, we set out to explore other spectroscopic mechanisms to produce more forms of high-order nonlinearity. Stimulated emission reduced fluorescence (SERF) microscopy was subsequently conceived. As shown in Fig. 6, an additional laser beam that is capable of inducing stimulated emission of the fluorophores from the excited states is introduced to collinearly overlap with the standard two-photon excitation beam. The two-photon fluorescence originated at the focus is then preferentially switched on and off by temporally modulating the stimulated emission beam. The resulting SERF image, constructed from the reduced fluorescence signal, will then exhibit an overall 3rd order nonlinear dependence on the incident laser intensities: quadratic from the excitation beam and linear with respect to the stimulated emission beam. We have published the principle of SERF microscopy, and are planning to implement it in our laboratory.

Fig. 6. Principle of stimulated emission reduced fluorescence (SERF) microscopy. (a) A simplified Jablonski diagram of a typical fluorophore under two-photon excitation and one-photon stimulated emission. (b) Setup. On a standard two-photon fluorescence microscope, a modulated stimulated emission beam is combined collinearly with the two-photon excitation beam. Reduced fluorescence at each pixel is measured by a lock-in amplifier to form the SERF contrast (which is effectively a 3-photon process).


(5) We have also exploited ground state depletion spectroscopy to achieve super-nonlinearity. We found that, in the region of weak ground state depletion (i.e., saturation), if the two-photon laser intensity is sinusoidally modulated at a fundamental frequency ω, the demodulated two-photon excited fluorescence at its third harmonic frequency 3ω will scale with the fourth power of excitation intensity. Such high-order nonlinearity is expected to offer a 2.3 times imaging depth extension over the regular two-photon microscopy. We have experimentally implemented this new microscopy on test samples. As shown by Fig. 7, the new method offers much higher signal-to-background contrasts for the fluorescent beads and GFP-expressing E. coli cells embedded in 3D scattering tissue phantom.


Fig. 7. Ground state depletion microscopy for high-contrast two-photon imaging. Comparison between the conventional two-photon imaging and our new method on fluorescent beads (a and b) or GFP-expressing E coli cells (c and d) embedded in 3D scattering tissue phantom. Much higher S/B contrasts are obtained with the new method.

(6) We harnessed the frustrated fluorescence resonance energy transfer (FRET) effect to create high-order nonlinearity for two-photon imaging. In brief, if the donor alone is excited, FRET will occur between the excited donor and the ground state acceptor, which quenches the donor fluorescence. In contrast, if the donor and the acceptor are both excited, energy transfer will be largely blocked and thus the donor fluorescence will be dequenched and enhanced (Fig. 8). Therefore, frustrated FRET is inherently a nonlinear process requiring excitation of both donor and acceptor. The concept of frustrated FRET was originally proposed to enhance the spatial resolution of confocal microscopy. Here we have exploited it for enhancing image contrast and penetration depth of two-photon microscopy in scattering samples.


Fig. 8. Principle of frustrated FRET. When only the donor is excited, the fluorescence resonance energy is transferred from the donor to the acceptor, quenching the donor fluorescence. b) When the donor and the acceptor are excited at the same time, FRET is inhibited, recovering the donor fluorescence.

(7) We could evaluate the imaging performance of our newly invented super-nonlinear microscopy by calculating the S/B ratio between the focal signal (S) and the out-of-focus background (B) as a function of the focal depth. We assume uniformly stained tissue samples, moderate anisotropy factors and a mean free path length of 200 m for brain tissues. Fig. 9 shows that the imaging-depth limit for the regular two-photon imaging is reached when zfocal=1023 m, which is very close to the experimentally measured 1 mm for mouse brain tissues labeled with GFP. For the new microscopy techniques with higher order dependence on laser intensity, the S/B ratios can be improved by orders of magnitude (i.e., much higher contrast). As a consequence, the new imaging-depth limit is expected to extend to zfocal= 1.8 mm for SERF microscopy, and zfocal= 2.4 mm for MPAI and MPDI microscopy (Fig. 9).

Fig. 9. Fundamental imaging-depth limits of nonlinear fluorescence microscopy of various orders.The depth limit is defined as the focal depth at which S/B reaches unity. The standard two-photon microscopy has a depth limit of zfocal=1023m. For microscopy with higher order nonlinearity, the corresponding S/B ratio and imaging-depth limit are significantly increased.


Relevant Papers:
1. F. Xu, L. Wei, Z. Chen and W. Min. "Frustrated FRET for high-contrast high-resolution two-photon imaging", Opt. Express 21, 14097 (2013).[PDF]
2. Y.-T. Kao, X. Zhu, F. Xu and W. Min. "Focal switching of photochromic fluorescent proteins enables multiphoton microscopy with superior image contrast", Biomed. Opt. Express 3, 1955 (2012).[PDF]
3. Z. Chen, L. Wei, X. Zhu and W. Min. "Extending the fundamental imaging-depth limit of multi-photon microscopy by imaging with photo-activatable fluorophores", Opt. Express 20, 18525 (2012).[PDF]
4. X. Zhu, Y.-T. Kao and W. Min. “Molecular-switch-mediated multiphoton fluorescence microscopy with high-order nonlinearity”, J. Phys. Chem. Lett. 3, 2082 (2012).[PDF]
5. L. Wei and W. Min. "What can stimulated emission do for bio-imaging?", Ann. N.Y. Acad. Sci. 1293, 1 (2013).[PDF]
6. L. Wei, Z. Chen and W. Min. "Stimulated emission reduced fluorescence microscopy: a concept for extending the fundamental depth limit of two-photon fluorescence imaging", Biomed. Opt. Express 3, 1465 (2012).