Since demonstration [1] of in vitro two-photon optogenetic stimulation (TPOS) of excitable cells and brain slices by using a point or scanning laser beam, there have been growing [2–5] applications of this method for better probing of neural circuitry. TPOS using near-infrared (NIR) laser beams provides deeper penetration as compared to conventional single-photon methods [6] due to low absorption and scattering coefficients of tissue in the NIR spectral region [7–9]. Further, high spatial precision (important for activating sub-cellular structures) is achieved by virtue of the nonlinear nature of ultrafast light interaction with the opsins. It is important to note that the two-photon cross section of opsins such as channelrhodopsin-2 (ChR2) has been estimated to be larger than that of most fluorophores [2], and therefore a two-photon beam has the potential for efficient stimulation of genetically targeted opsin-expressing cells in a dense tissue. It is important to note here that different scanning modes (spiral, raster) have been applied for the excitation of ChR2-expressing cells to optimize the efficiency of excitation [2]. While there have been recent advancements in TPOS technology by spatial sculpting [3] and/or temporal focusing [4] of the laser beam, to date, two-photon activation has only been demonstrated using bulky microscopic objectives.

In contrast to the use of a microscope objective and scanning two-photon laser beam, here we report on the development and implementation of a multimode beam for nonscanning fiber-optic TPOS (FO-TPOS) of ChR2-transfected excitable cells. This technology paves the way for precise in vivo probing of neural circuitry in a minimally invasive manner. Figure 1 shows the schematic comparison of various conventional objective-based scanning [raster and spiral, Fig. 1(a))] and fiber-optic nonscanning patterns [defocused single mode and multimode, Fig. 1(b)].

 

Fig. 1. (a) Schematic of conventional two-photon stimulation scanning pattern of targeted cell with femtosecond laser delivered by objective. (b) Schematic of fiber-optic two-photon activation. (c) Surface plot of the intensity pattern from a multimode fiber. Scale bar: 50 μm. (d) Composite image of two-photon fluorescence from polystyrene particles (green) and excitation intensity pattern (red).

Download Full Size | PDF

To confirm multimode fiber-optic two-photon excitation of fluorescent polystyrene microspheres (dia: 45 μm, Bangs Laboratories), an ultrafast laser beam (FSL, Maitai HP, Newport-SpectraPhysics Inc.) was expanded using a beam expander (BE) and coupled to a multimode optical fiber (core dia: 60 μm) using a fiber coupler (Newport Inc.), and the polystyrene particles were illuminated by the multimode fiber-optic beam. Figure 1(c) shows the surface plot of the intensity pattern from a multimode fiber imaged using $10\times$ microscope objective. The multimode fiber-optic beam (870 nm, 200 fs, 20 mW) was sufficient to induce two-photon fluorescence excitation in multiple particles simultaneously. In Fig. 1(d), we show a composite image of two-photon fluorescence (green) from the polystyrene particles overlaid on the multimode excitation intensity pattern (red).

For intensity-dependent characterization of multimode fiber-optically induced two-photon fluorescence excitation (or cellular stimulation), a circular neutral density filter (NDF) to control the intensity and a shutter (S) for controlling the exposure (macropulse) duration were used in the path of the tunable Ti:sapphire femtosecond laser beam (FSL) as shown in Fig. 2(a). The cleaved multimode fiber was mounted on a mechanical micromanipulator and positioned near the 45 μm polystyrene particles [Fig. 2(b)]. In the mode-lock-off condition, no fluorescence was observed by the electron-multiplying CCD (EMCCD, Photometrics Inc.) as shown in Fig. 2(b). Figure 2(c) shows a composite image of the bright-field and two-photon fluorescence excited by the multimode fiber-optic beam in the mode-lock-on condition. To characterize the two-photon excitation, the integrated fluorescence intensity was plotted as a function of incident laser beam power [Fig. 2(d)]. The nonlinear rise in fluorescence intensity confirmed the possibility of a multimode fiber-optic ultrafast laser beam enabling two-photon stimulation. Next, we implemented this method in FO-TPOS of cells.

 

Fig. 2. (a) Schematic of the fiber-optic two-photon irradiation and patch-clamp setup. (b) Bright-field image of fluorescent polystyrene particles and (c) two-photon fluorescence excited by fiber-optic multimode beam. Scale bar: 100 μm. (d) Variation of fluorescence intensity of arrow-marked bead with incident laser power.

Download Full Size | PDF

For FO-TPOS of cells, the HEK 293 cells were transfected with the ChR2-EYFP construct, and cloned into pcDNA3.1 neo (Invitrogen Inc.). EYFP was fused in-frame to the C terminus of ChR2 by PCR. Cells were maintained at 37°C, 5% ${\mathrm{CO}}_2$ in DMEM containing 10% fetal bovine serum. For generating light activation, cells were loaded with all-trans retinal (ATR, 1 μm) for at least 6 h and activated with a fiber-optic laser beam. Expression of CHR2 in HEK 293 cells was confirmed by fluorescence imaging of reporter fluorescent protein (YFP).

The cleaved multimode fiber tip was positioned (distance: 100 μm) near the ChR2-expressing cells (identified by YFP fluorescence). Light power at the fiber tip was measured using a standard light power meter (PM 100D, Thorlabs Inc.). Two NIR low-pass filters were placed in front of the EMCCD to reject the fiber-optic stimulation light. The macropulses were generated and controlled by an external shutter placed in the laser beam path, upstream from where the laser beam is coupled to the fiber.

The FO-TPOS electrophysiology setup was developed on an Olympus upright microscope platform using an amplifier system (Axon Multiclamp 700B, Molecular Devices Inc.) as shown in Fig. 2(a). For a whole cell patch clamp, the micropipette was filled with a solution containing (in mM) 130 K-Gluoconate, 7 KCl, 2 NaCl, 1 ${\mathrm{MgCl}}_2$, 0.4 EGTA, 10 HEPES, 2 ATP-Mg, 0.3 GTP-Tris, and 20 sucrose. The electrode was mounted on an XYZ micromanipulator (Newport Inc.). The output from the amplifier was digitized using a National Instruments card (PCI 6221). For electrophysiological recording, the hardware was interfaced with patch-clamp software from the University of Strathclyde. The shutter (S) trigger was synchronized to the electrophysiology recording system. Electrical recordings were performed at a holding potential of $-60mV$. pClamp software was used for patch-clamp data analysis.

Figure 3(a) shows the bright-field image of a patch-clamped HEK cell transfected with ChR2. Here, we used a wavelength of 850 nm for fiber-optic stimulation in excitable cells [Fig. 3(a)]. Figure 3(b) shows raw recordings of inward current in response to five randomly selected macropulses (100 ms) of ultrafast ($\sim 300\mathrm{fs}$, measured after the fiber) laser pulses emanating from the multimode fiber tip positioned 100 μm away from the clamped cell. Taking into consideration the measured spot size of the fiber-optic beam at separation of 100 μm from the fiber tip and the laser power premeasured at the tip of the fiber, the average power density was calculated to be 0.012 $\mathrm{mW}/{\mathrm{\mu m}}^2$ in this configuration. The fact that a continuous-wave (control) laser beam (operated in mode-lock-off condition), having the same average power density (and exposure), did not induce inward current [Fig. 3(b)] confirms nonlinear interaction of the femtosecond NIR laser beam with the ChR2-expressing cells.

 

Fig. 3. FO-TPOS of ChR2-expressing cells. (a) Bright-field image of a patch-clamped HEK cell; scale bar: 10 μm. (b) Inward current in response to macropulses (100 ms) composed of FO-TPOS pulses. Also shown is the current response due to the control laser beam. (c) Inward current responses due to FO-TPOS (at 850 nm) using different average power densities (in $\mathrm{mW}/{\mathrm{\mu m}}^2$). (d) Dependence of inward current as a function of power densities (100 ms pulses) ($\mathrm{N}=4$).

Download Full Size | PDF

As compared to the single-photon activation spectrum [10] of ChR2, the peak of the two-photon activation spectrum was observed to be around 850 nm. This blue shift can be attributed to the fact that higher excited singlet states are reached with greater probability by two-photon excitation. Therefore, this wavelength was selected for studying the fiber-optic two-photon intensity-dependent current response of cells. Figure 3(c) shows representative inward current responses due to fiber-optic ultrafast NIR (850 nm) stimulation of a single cell at different average power densities (0.006–0.02 $\mathrm{mW}/{\mathrm{\mu m}}^2$, at cell membrane). The dependence of two-photon (at 850 nm, 100 ms pulses) induced inward current as a function of incident average power densities (near the cell membrane) ($\mathrm{N}=4$) is shown in Fig. 3(d). In contrast to expected nonlinear response of inward current, the FO-TPOS induced current in cells was found to be linearly dependent on the incident laser power density as it is a response to excitation rather than a direct measure of emitted photons, unlike two-photon fluorescence emission [Fig. 2(d)].

This is due to the fact that in the case of optogenetic activation, inward current is associated with conformational change of the molecule rather than a simple two-level transition as in the case of a dye molecule. It may be noted that the photophysics involved with ATR or ChR2 are not completely understood, and the electronic states involved in the process are not clearly identified. Therefore, it will be necessary to evaluate parameters such as the kinetics of the opening of the ChR2-channel and the two-photon absorption of ATR [11,12] in addition to intensity dependence (e.g., saturation at higher intensities). The 2-ph cross section at 850 nm was estimated to be $139\pm 123$ Goeppert–Mayer units for the fiber-optic multimode beam, using the inward current response to near saturation levels of two-photon activation for normalization. Accurate measurement of the TP cross section using a fiber-optic beam would require better modeling of ChR2 activation, and the use of a single-mode fiber-optic beam, or fast modulation of the multimode fiber by external piezo-actuation to smoothen the intensity profile.

Though we characterized multimode two-photon activation efficacy at different NIR laser intensities, further studies are required to optimize the multimode FO-TPOS strategy. The large two-photon cross section of ChR2 [2] should allow use of nanosecond or even microsecond compact NIR sources for FO-TPOS. Further, by tuning the wavelength of the two-photon light source for other opsins such as NpHR, the FO-TPOS method can be useful for combinatorial modulation (both excitation as well as inhibition) of the neural activity. Use of a microlens or axicon [13] on fiber tip should allow focus control of the FO-TPOS. The FO-TPOS technology will allow in-depth probing of neural circuitry in vivo since this technology permits minimally invasive and more precise anatomical delivery of stimulation. Further, due to the compactness and flexibility of the fiber, it would allow study of neuromodulation in freely behaving animals. Using the technology already developed for nonlinear endoscopy [14], both two-photon stimulation and optical imaging of neural activity can be achieved in vivo.