Open Access
Add to BibSonomyAbstract
We report a fiber optics-based intravital fluorescence imaging platform that includes epi-fluorescence microscopy and laser patterned-light stimulation system. The platform can perform real-time fluorescence imaging with a lateral resolution of ~4.9 μm while directly inserted into the intact mouse brain, optically stimulate vasoconstriction during real-time imaging, and avoid vessel damage in the penetration path of imaging probe. Using 473-nm patterned-light stimulation, we successfully modulated the vasoconstriction of a single targeted 37-μm-diameter blood vessel located more than 4.7 mm below the brain surface of a live SM22-ChR2 mouse. This platform may permit the hemodynamic studies associated with deeper brain neurovascular disorders.
© 2017 Optical Society of America
1. Introduction
In neuroscience, it is important to understand the delivery of blood to cranial tissue and links between functional units, e.g., the functional activities of the neurovascular or gliovascular units, in vivo [1–4]. Perivascular neurons, astrocytes, and vascular cells work together as functional units in the cerebrovascular system to maintain homeostasis, regulating body temperature, blood pressure, and pH. Functional changes in cell-level interactions can alter brain activity, hemodynamics, and metabolism in the brain, causing deviations in normal neurological function. Researchers have used brain imaging systems to identify brain structures and vascular distributions, perform specific functional assessments in specific brain areas, and study brain diseases related to various vascular diseases [1, 4–16]. Functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) scans are the most commonly used imaging techniques exploiting hemodynamic responses to map brain functions [1, 6, 7]. However, these imaging techniques have the disadvantage of difficulty in directly detecting intercellular responses, because they can only confirm the reaction to changes in blood flow caused by the simultaneous reaction of thousands of neurons [8]. High-resolution imaging optical systems such as conventional microscopy systems, including epi-fluorescence [9], confocal microscopy [10], and two-photon microscopy [11–14], have also been applied to cerebrovascular studies. These high-resolution imaging systems facilitate studies of the hemodynamics and blood cell–vessel wall interactions in the primary somatosensory cortex [11, 15], limb cortex, and parietal cortex [12, 14]. However, because of the limited penetration of light into brain tissue, many researchers experience difficulties in studying deep-brain vascular-related diseases, such as Parkinson’s disease [2, 16].
To overcome the limitations in spatiotemporal resolution and depth penetration of light into brain tissue, an intravital microscopy system has been designed that relies on microlenses and fiber optics to provide micrometer-scale resolution in deep-brain imaging by direct insertion into brain tissue [17–19]. This allowed researchers to study the deep neocortex [17], thalamus, and hippocampus [18, 19], which was impossible using conventional microscopy systems that obtain images through thinned cranial windows. These invasive techniques for deep brain imaging have already broadened the applications of the optical stimulation system and fluorescence imaging system for in-vivo brain research [8, 18, 20].
Recently, many brain imaging research groups have developed multimodal imaging systems combining two or more imaging systems to complement the individual limits of single systems [21]. These efforts emphasize the importance of developing a multimodal imaging system that enables the diagnosis and treatment of cerebrovascular disorders. Multimodal imaging techniques include combinations of photoacoustic microscopy (PAM) and fluorescence confocal microscopy and the integration of laser-scanning optical-resolution PAM with spectral-domain optical coherence tomography [22, 23]. However, a multimodal imaging system for deep-brain blood vessel studies incorporating intravital microscopy has not yet been developed.
Here, we introduce a fiber optics-based intravital fluorescence imaging platform that integrates epi-fluorescence microscopy and laser patterned-light stimulation systems. The platform can avoid blood vessels located along the insertion route by the insertion of flexible, position-free fiber bundles instead of microlenses [8, 24, 25]. The optical transmission properties of both sides of the fiber bundle enable easy and precise targeting of specific blood vessels located in the deep brain area, as well as immediate light stimulation. All depths of the brain are approachable, but in this experiment we have targeted a reticular thalamic nucleus at or above 3.5 mm in depth in the brains of living mice. To demonstrate that the integrated platform can perform therapeutic optical stimulation and diagnosis of vascular disorders at the same time in the deep brain, optical stimulation and monitoring vasoconstriction induced by optical stimulation were performed on a SM22-ChR2 transgenic mouse. In cerebrovascular disease studies, vasoconstriction is an especially important vascular response in studying vascular occlusion, hemostasis, and neurovascular circuits [4, 15, 26, 27]. Our proof-of-concept experimental results demonstrate that the platform possesses great potential for studying cerebrovascular diseases or hemodynamics accordingly in the functional unit connections in living animals [2–4, 15, 27].
2. Methods
2.1 Real-time fluorescent blood vessel imaging
To demonstrate the proposed integrated intravital imaging platform using a fiber bundle, we performed real-time fluorescent blood-vessel imaging in live wild-type male mice aged 8 weeks. The mice were placed under avertin (250 mg/kg) anesthesia during the tail-vein injection of a 200 μL bolus of fluorescein isothiocyanate (FITC)-dextran (2000 KDa, Sigma-Aldrich Co.) in phosphate-buffered saline (PBS, 7.5 mg/0.2 mL). The simultaneous insertion of the fiber bundle into the brain and real-time recording was initiated approximately 2 h after the injection. The fluorescence image was taken with green fluorescent protein (GFP) band-pass filters (FF02-482/18-25, FF03-525/50-25, Semrock), rotating the excitation and emission filter wheels to select a filter for GFP fluorescence imaging.
2.2 Stereotactic injection of viral vectors (Ad-CMV-mCherry) and imaging
Ad-CMV-mCherry (Vector BioLabs) was used for transfection of the cells in the deep brain. Male C57BL/6 (B6) mice aged 8 weeks were anesthetized with 3% isoflurane in 100% oxygen. Surgery was conducted after head fixation on a stereotaxic atlas (51730, Stoelting Co.). After a small craniotomy, the dura was removed for needle insertion. Using X and Y stereotaxic coordinates, a Hamilton syringe (7635-01, Hamilton Syringe, 10 μL, 33G) was positioned over the area to be injected. The coordinates of the craniotomy were as follows: in the reticular nucleus (prethalamus), the anterior–posterior (AP) was −2.06 mm, the medial–lateral (ML) was 1.6 mm from the bregma, and the dorsal–ventral (DV) was −3.5 mm relative to the skull surface. These stereotaxic coordinates were established by Franklin and Paxinos [28]. The needle was very slowly lowered to a depth of 3.5 mm, the injection parameters (0.5 μL/1 min) were entered into the injector controller (NanojectII Auto-Nanoliter Injector, Drummond Scientific Co.), and the virus injection was initiated (approximately 5–7 μL). After finishing the injection, the needle remained in position for 1 min. to prevent efflux of the virus during removal. After this period, the needle was very slowly removed from the brain. The imaging experiment was performed after approximately one week. The virally injected mice were anesthetized with urethane (1.5 g/kg, i.p.) after head fixation to the stereotaxic atlas. After a small craniotomy (approximately 1 mm), the dura was removed for the insertion of the fiber bundle. The fiber bundle was carefully inserted into the deep brain at an approximate rate of 200 μm/s. Fluorescence images were taken with mCherry band pass filters, rotating the excitation and emission filter wheels (FF01-562/40-25, FF01-641/75-25, Semrock) to select a filter for mCherry fluorescence imaging. After the imaging experiment, the brain was removed and sliced into thicknesses of 30 μm using a vibratome (CM1950, Leica Microsystems GmbH) to confirm the positions of the viral injection and fiber insertion. Coronal sections of the brain were viewed under a microscope (Eclipse Ti-U, Nikon Corp.), and images were captured with a Hamamatsu electron-multiplying charge-coupled device (EM-CCD) digital camera (C9100-23B, Hamamatsu Photonics K.K.) using HCImage Live image-processing software (Hamamatsu Photonics K.K.). Mosaic section image stacks were acquired with a fluorescence microscope (Eclipse Ti-U, Nikon Corp.) equipped with a 4 × objective lens (CFI Plan Fluor 4x, NA 0.13, Nikon Corp.) and XY scanning stages (MLS203, Thorlabs, Inc.). The image tiles were combined into a mosaic using Photoshop (Adobe Systems Inc.) to rotate, crop, and adjust the color balance of the images.
2.3 In vivo animal preparation and imaging for optical activation of cerebral vasoconstriction of SM22-ChR2 mice and C57BL/6 (B6) mice
We used male C57BL/6 (B6) mice aged 8 weeks as the control. To perform the real-time optical manipulation of cerebral vasoconstriction in vivo, we used channelrhodopsin-2 (ChR2) mice among the offspring of floxed optogene (RCL-ChR2 (H134R)/EYFP) mice (Jackson Lab.) and SM22α-cre KI mice (Jackson Lab.). Depolarization of smooth muscle cells expressing ChR2 by blue light induces vasoconstriction in brain of SM22-ChR2 mice. Recently, it has been reported that contraction of bladder smooth muscles and cardiovascular smooth muscles can be regulated by ChR2 [29, 30]. So, we used ChR2 mice to control cerebral vasoconstriction in vivo through optical light stimulation. The SM22-ChR2 mice were anesthetized with urethane (1.5 g/kg, i.p.) after head fixation to the stereotaxic atlas. After a small craniotomy (approximately 1 mm), the dura was removed for the insertion of the fiber bundle. The fiber bundle was carefully inserted into the deep brain at an approximate rate of 200 μm/s, depending on real-time imaging. The cerebrovascular system was imaged with the yellow fluorescent protein (YFP) excitation band-pass filter and emission band-pass filter (ET500/20X, ET535/30m, Chroma) to observe the enhanced YFP (EYFP)-labeled smooth muscle cells.
2.4 Animal surgeries
All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of the Korea Institute of Science and Technology (KIST, 2014-049-3) and followed the guidelines of the Institutional Animal Care and Use Committee of KIST. The body temperature was maintained at 37 °C using a feedback-controlled heating pad during both surgery and imaging.
2.5 Imaging post-processing
To improve image quality, the intensity of the cladding area was interpolated with that of a neighboring fiber core using an iterative segmentation-interpolation algorithm, as suggested by Ford [31]. This post-processing method is used to remove fiber bundle cores and heighten the contrast. The post-processing was implemented with MATLAB® (The Math Works, Inc.).
2.6 Data analysis
Image analysis was performed using Image-Pro Premier 9.1 (Media Cybernetics, Inc.) and ImageJ 1.48v (Wayne Rasband, US National Institutes of Health). In addition, Image-Pro Premier 9.1 was used to coat a pseudo-color and merge images. Adobe Photoshop CS6 was used to insert scale bars; animate the frame rate; combine image tiles into a mosaic; and rotate, crop, and adjust the color balance of the images. Statistical tests were performed with Microsoft Excel (Microsoft®) and OriginPro8 (OriginLab Corp.). All values are mean ± standard deviation (SD) unless otherwise noted.
3. Results
3.1 A fiber bundle-based integrated platform design and analysis
The fiber bundle-based integrated platform has three components, as shown in Fig. 1(A): (i) a multi-color fluorescent light illumination component; (ii) a wide-field fluorescence imaging component; and (iii) a patterned 473-nm light stimulation component. The proposed integrated optical platform transfers images through a polished bare fiber bundle.
Fig. 1 A fiber bundle-based integrated platform. A) Schematic of the fiber bundle-based integrated platform; includes multi-color fluorescent light illumination, fluorescence imaging, and patterned 473-nm light stimulation. The enlarged image inside the black box is a photograph of the fiber bundle. (B) Captured spatial light modulator (SLM) stimulation pattern of the Korea Institute of Science and Technology (KIST) logo at the output end of the fiber bundle (left); measurement of core-to-core distance at the output end of the fiber bundle (right). (C) Captured stimulation image (left) and line profile (right) of SLM stimulation pattern of rectangle pattern of 5 × 5 pixels at fiber bundle end-plane. This pattern matches the size of a single fiber core. Abbreviations: BS: beam splitter; PBS: polarizing beam splitter; ND: neutral density; L1&L2: achromatic doublet, f = 100 mm; L3: achromatic doublet, f = 150 mm; L4: achromatic doublet, f = 75 mm. Scale bars: (A) 1 mm and (B-C) 50 μm.
To achieve full-field multi-color fluorescent imaging, the multicolor fluorescent light illumination path starts from a broadband light source (120-W mercury lamp, X-cite® 120Q, Excelitas Tech. Corp.). The broadband light propagates through a 0.5-inch beam expander, excitation filters, and an achromatic doublet lens (L1, f = 100 mm). An excitation filter wheel (CFW6, Thorlabs, Inc.) is located after the lens to select the specific wavelength of the broadband light. The selected excitation light is reflected by the dichroic mirror (69011bs, Chroma Tech. Corp.) and propagates through a 40 × objective lens (PLN 40x, NA 0.65, Olympus Corp.). The dichroic mirror includes a triple-wavelength band filter that facilitates the functions of the other components, as described below. The end of the fiber bundle is placed at the focal plane of the objective lens, and the other end contacts the brain tissue directly. The dichroic mirror is fixed, but the excitation and emission filter wheels (CFW6, Thorlabs, Inc.) are changed to select a well-matched filter for multi-fluorescence imaging based on the target fluorescence color. In addition, a digital shutter (Uniblitz Shutter System, VCM-D1, VS25S2ZM0R3, Vincent Associates®) is activated during the interval when the CCD camera (Cool SNAP KINO, Photometrics) is exposed to light.
The patterned light illumination component for the optogenetic stimulation used a 473-nm diode-pumped solid-state (DPSS) laser light (MBL-H-473-300 mW, CNI Optoelectronics Tech. Co.) to activate ChR2 in the deep brain. The laser light is modulated by a function generator (DG4062, Rigol Tech. Inc.), and the intensity of the laser light is attenuated by a reflective neutral-density filter (NDK01, Thorlabs, Inc.). The laser light becomes spatially homogeneous and the speckle noise is removed by using a laser speckle-reducing diffuser (LSR-3005-12D-VIS, Optotune). The first polarizing beam splitter (PBS) (CCM1-PBS251/M, Thorlabs, Inc.) and a reflective spatial light modulator (SLM) (LC-R720, amplitude modulation, Holoeye) are used to create 473-nm laser light patterns for the real-time patterned optical stimulation. The laser patterns are projected onto the fiber bundle by the L3 lens (f = 150 mm) and the objective lens. The patterned laser light for optical stimulation and the illuminating light from the broadband light source for multicolor fluorescent imaging are combined by the second PBS and the dichroic mirror. First, the patterned 473-nm light is reflected by the second PBS, which changes the direction of the laser light by 90°. The patterned 473-nm light then passes through the dichroic mirror. The maximum intensity of the patterned laser light after the fiber bundle is measured as ~120 mW/mm2 using an optical power meter (S130C, PM100D, Thorlabs), controlled using neutral-density filters and the SLM. The 33.3 × magnification of the SLM projection system is designed to cover the surface of the fiber bundle without degrading the spatial resolution of the fiber bundle. In addition, the SLM is connected to a computer to project the desired shape easily into the brain tissue. Figures 1 (B) and 1 (C) show the patterned laser light images at the end of the fiber bundle that contacts the brain (IGN-037/10, Sumitomo Co.). The core size and core-to-core distance of the fiber bundle are measured as 1.6 μm and 2 μm, respectively, when expressed in full-width-half-maximum (FWHM). This is sufficient for patterned-light stimulation with brain structures of various sizes ranging from a few micrometers to hundreds of micrometers, such as blood vessels, neurons, and astrocytes.
For fluorescence imaging, the fiber bundle relays the emitted fluorescence light from the brain surface to the opposite side of the fiber bundle, and the emitted light passes through the dichroic mirror, achromatic doublet lens (L2, f = 100 mm), and emission filter to the CCD camera. The dichroic mirror passes the patterned laser light for optogenetic stimulation and the emitted fluorescence light (450–490 nm, 510–550 nm, and 580–650 nm), but reflects the broadband illumination light (~450 nm, 490–510 nm, and 550–580 nm) for multicolor fluorescent imaging. An emission filter wheel is used to select a specific wavelength of emitted light from the tissue surface for the fluorescence imaging. This fluorescence imaging system can be used with various wavelengths of fluorescent materials, such as DAPI, GFP, EYFP, mCherry, Fura-2, GCaMP, OGB-1, and X-Rhod-1, by using a broadband light source and multicolored fluorescence filter set. The magnification between the fiber bundle and the CCD camera is 22 × , which is designed to provide an exact match between the diameter of the fiber bundle (330-μm) and the sensing area of the CCD camera (8807 × 6628 μm).
3.2 Optical properties of the fiber bundle-based integrated platform
In order to determine the minimum resolvable feature size of the fluorescence imaging, a resolution test was preformed using a negative USAF 1951 resolution target (R3L3S1N, Thorlabs, Inc.) backed with a green fluorescence reference slide (#2273, Ted Pella, Inc.). The fluorescence image was taken with GFP band-pass filters (FF02-482/18-25, FF03-525/50-25, Semrock). The raw images are impaired by a honeycomb-like pattern that matches the individual cores and shared cladding of the fiber bundle, as shown in the left image of Fig. 2(A). After capturing the fluorescence image of the resolution target, image post-processing was performed to remove the fiber bundle pattern using MATLAB® (The Math Works, Inc.). This step is needed to improve visibility. The fiber bundle patterns are removed by implementing the iterative segmentation-interpolation algorithm suggested by Ford [31]. In Fig. 2(A), the left image shows a 1951 USAF resolution target before post-processing, while the right image shows the same target after the imaging post-processing. The red box represents the region of minimum resolvable vertical and horizontal patterns from which light intensity is obtained. Visibility is improved by reducing the pixelation of the optical fiber bundles seen in the raw image, even if the image is slightly blurred after post-processing. In Fig. 2(B), the line profiles through before and after processed image show contrast of elements 5 in group 7 (209 lp/mm) yielding a respective line width of 4.92 μm, spanning approximately two fiber cores (thereby satisfying the Nyquist sampling criterion). When tested on cultured dorsal root ganglion cells, the nucleus, cytoplasm, and axons of the cells can be distinguished (Figs. 2(C) – (D)). The measured diameter of the nucleus and cytoplasm located in the cell are 12 μm and 18 μm, respectively. The same iterative segmentation-interpolation algorithm was applied to all figures and videos.
Fig. 2 Optical properties of the fiber bundle-based integrated platform and imaging post-processing. A) Image of 1951 USAF resolution target. The red box represents the region of minimum resolvable vertical and horizontal patterns of elements 4 and 5 in group 7 (181 lp/mm, 209 lp/mm); before imaging post-processing (left) and after imaging post-processing (right). (B) Line profile of vertical and horizontal patterns of elements 5 in group 7 in the image of (A). (C) Immunocytochemistry (ICC) images of cultured dorsal root ganglion cells (DRGs) taken by the platform: the blue, green, and red images represent nucleus, tubulin (cytoplasm), and tau (axons) proteins, respectively, in DRGs. (D) Line profile of red dashed line in merged image of (C). Scale bars: (A, C) 50 μm.
3.3 Selection of the fiber bundle diameter to achieve favorable field of view for blood-vessel imaging in deep brain
A larger-diameter fiber bundle can provide a wider field of view (FOV) for the full-field image, but it is hard to avoid damage to the brain tissue when inserting larger fiber bundles. Damage to surrounding tissues can increase bleeding and permanent trauma. In order to study the in-vivo cerebrovascular system using intravital microscopy, it is especially important to minimize bleeding and reduce the impact on the surrounding tissue of the insertion of the imaging probe [19].
Before performing the imaging experiments, we tested two different fiber bundles of different diameters, selected based on the criteria of a wider FOV and higher density of image fiber elements per unit area, for imaging with better lateral resolution. The first fiber bundle contained 30,000 individual fiber elements, distributed in an imaging area of 620-μm in diameter, 0.4 numerical aperture (NA) (FIGH-30-650S, 1 m, Fujikura Europe Ltd.); the other fiber bundle contained 10,000 individual fiber elements, distributed in an imaging area of 330-μm in diameter, 0.35 NA (IGN-037/10, 1 m, Sumitomo Co.). When a fluorescence image was formed using the 620-μm-diameter fiber bundle, the entire fluorescent image magnification was changed from 22 × to 11 × by changing the objective lens positioned immediately before the fiber bundle from 40 × to 20 × (UPLFLN 20x, NA 0.5, Olympus Corp., Fig. 1A).
3.4 Angiography
Angiography was performed using a real-time fluorescent image of the cerebrovascular system during insertion. A diluted solution of FITC-dextran was injected into the blood vessel using tail-vein injection to visualize the cerebrovascular distribution at the surface and in the deeper structures. Injury or excessive contact of the blood vessel with the fiber bundle can cause bleeding, and the FITC mixture can be immediately visualized as it emerges from the injured vessel. The 620-μm-diameter fiber bundle is shown in Figs. 3(A) – (C), 3(F), and Visualization 1 and Visualization 3. This type achieves a wider FOV than the smaller fiber bundle. However, as shown in Fig. 3(A), the 620-μm-diameter fiber bundle causes frequently faster bleeding (Visualization 1) during deep-brain imaging. On the other hand, when using the smaller fiber bundle of 330-μm in diameter, as shown in Fig. 3(D), 3(E), and Visualization 2, damage and bleeding in the brain are minimized while maintaining continuous imaging during insertion. As shown in Fig. 3(E), the optical platform with a 330-μm fiber bundle achieves fine images even in the deep brain, as well as on the brain surface, compared to the larger one (Fig. 3B). Based on these results, we selected the fiber bundle of approximately 330-μm in diameter for use in all subsequent experiments to mouse brain imaging.
Fig. 3 Real-time fluorescent blood vessel imaging using two types of optical fiber bundles with different diameters. FITC-Dextran (200 kDa, 7.5 mg in 0.2 mL saline) was injected through the vein in the tail before the in vivo imaging. Frames A–C and F were obtained with the 620-μm fiber bundle, and frames D and E were obtained with the 330-μm fiber bundle. (A) Blood vessel image showing bleeding from excessive contact with the fiber bundle (see Visualization 1). (B) Image of uneven contact with the brain surface. (C) Blood-vessel imaging of the brain surface (see Visualization 3). (D) Blood vessel image at the superficial level, and (E) the image from the same frame at a depth of greater than 500 μm (see Visualization 2). (F) Cropped and magnified image sequence of the region in the white box in (C): the red, blue, green, and white arrowheads indicate individual red blood cells, and their progress can be tracked within the capillary; the relative times in seconds are displayed in the upper left corner of each image in the sequence. The mean blood flow rate measured for 2.6 s is 23 μm/s. The frame rate was 5 frames/s, and the exposure time was 200 ms. The illumination light output intensity was 1.2 mW/mm2. Scale bars: (A–C) 100 μm, (D, E) 50 μm, and (F) 20 μm.
In addition, we can observe the flow of red blood cells within capillaries by visualizing the dark features, because the unlabeled red blood cells appear dark against the brighter background of the FITC dye mixed with the blood plasma, as shown in Fig. 3(C), 3(F), and Visualization 3. The averaged blood flow rate for 2.6 s is measured as ~23 μm/s.
3.5 In-vivo real-time fluorescence images of Ad-CMV-mCherry expressing neurons in a mouse’s deep brain
To show that the proposed platform can accomplish the functions of accurate and wide-field cerebrovascular fluorescent imaging with near-cellular resolution, in-vivo real-time deep-brain imaging of a fluorescently labeled neuron cluster of a mouse was attempted. The location was found using a mouse brain atlas [28], and then the Ad-CMV-mCherry virus was injected into the deep brain region at a depth of 3.5 mm, as detailed in Methods and Fig. 4(A). Fluorescence imaging was performed one week after the injection (Fig. 4B). As shown in Fig. 4(E), fluorescence imaging of the live mouse brain is obtained continuously in real time from the surface to a depth of 5 mm during the insertion of the fiber bundle probe. The mCherry fluorescence images become brighter when the distal end of the fiber bundle probe approaches the neuron cluster that expresses mCherry at the viral injection site. When the fiber bundle probe moves away from the viral injection site, the fluorescence images become darker, as shown in Visualization 4. The depth of the brightest location, shown in Fig. 4(D), is approximately 3500 μm as measured with the proposed platform, which is well matched with the virus injection site in the brain slice fluorescence images, shown in Fig. 4(C), obtained by the fluorescence microscope. This result confirms the ability of the platform to target specific fluorescence areas using contextual image information from the brains of living animals through real-time monitoring.
Fig. 4 In-vivo real-time fluorescence images of Ad-CMV-mCherry expressing neurons in a mouse’s deep brain. A) Brain atlas showing the injection region. (B) Photograph of the in-vivo deep-brain imaging experiment. (C) Images of the 100-μm-thick parasagittal slices, cut after the deep-brain imaging; the y-axis indicates the depth. (D) Intensity of the region of interest (ROI) according to the insertion depth; vertical gray lines indicate standard deviation ( ± SD, n = 308). (E) The deep-brain images according to various depths 6 days after the injection of Ad-CMV-mCherry (see Visualization 4); the depth is displayed in the upper left corner of each image. The frame rate is 10 frames/s, the exposure time is 100 ms, and the output intensity of the illumination light is 15 mW/mm2. Scale bars: (A) 1 mm and (E) 50 μm.
3.6 In-vivo modulation of vasoconstriction under optical stimulation and real-time monitoring in deep brain of living mouse
To demonstrate that therapeutic optical stimulation and diagnosis of vascular disorders is possible at the same time in the deep brain using the proposed platform, optical stimulation and the monitoring of vasoconstriction induced by optical stimulation were performed in a SM22-ChR2 transgenic mouse. In this research, the application of the optogenetic technique was used for guiding fluorescence and visualizing the change in responses according to the optical stimulation. The brain blood vessels of the SM22-ChR2 transgenic mouse express ChR2-EYFP (as described in Methods) because smooth muscle cells are located inside the blood vessel walls. Therefore, with the integrated optical platform, vasoconstriction can be visualized by EYFP fluorescence imaging of blood vessels while simultaneously applying a patterned 473-nm laser light to stimulate ChR2 in the smooth muscle cells inside blood vessels to modulate their contraction, as shown in Fig. 5 and Visualization 5. We used a C57BL/6 mouse as a control, injecting it with a diluted solution of FITC-dextran into the blood vessel via tail-vein injection to visualize the lumen of the cerebrovascular system, because the blood vessels are not seen alone. We targeted the blood vessels located 2.3 mm below the cortical surface of the brain of the control mouse. The diameter of the blood vessel lumen is not changed when the brain of the control mouse is stimulated with a 473-nm whole patterned light with an intensity of 1.4 mW/mm2 and an illumination light with an intensity of 100 μW/mm2 for FITC fluorescence imaging, as shown in Fig. 5(A). Meanwhile, Fig. 5(B) shows that, when we use the SM22-ChR2 transgenic mouse, some blood vessels located in the deep brain show dynamic vasoconstriction under illumination, even with an excitation light of 170 μW/mm2 for EYFP fluorescence imaging, because the ChR2 activation spectrum overlaps with the EYFP excitation spectrum (490 nm and 510 nm). The excitation light for EYFT imaging is then reduced until the blood vessels do not respond to the excitation light for imaging. The threshold intensity of the excitation light for imaging found to induce vasoconstriction is measured as exceeding 150 μW/mm2. As the intensity decreased to 50 μW/mm2, the change of vessel diameter was not observed in our imaging system compared to 150 μW/mm2. Following these process, unintended contraction by excessive illuminating light for EYFP imaging was excluded, enabling more accurate light stimulation and analysis.
Fig. 5 Control and monitoring of the single cerebral blood vessel using patterned light stimulation in the deep brain of ChR2 transgenic mice (SM22(CAG-ChR2-EYFP)) and B6 control mice. A) Images of vasoconstriction in a single cerebral blood vessel using whole patterned light stimulation at 2.3 mm below the cortical surface of the brain of the control mouse: i) No stimulation and ii) Stimulation using whole pattern. (B) Images of dynamic vasoconstriction in a single cerebral blood vessel illuminated with an excitation light of 170-μW/mm2 intensity for EYFP fluorescence imaging at 2.5 mm below the cortical surface of the brain of a ChR2 transgenic mouse: i) The first image frame, ii) After the seventh image frame, (arrowheads indicate the single blood vessel in which dynamic vasoconstriction is induced by illumination). (C) Images of vasoconstriction in a single cerebral blood vessel using patterned light stimulation at 4.7 mm below the cortical surface (see Visualization 5): (i, ii, v–viii) Fluorescence imaging of a blood vessel with EYFP in the deep brain of a ChR2 transgenic mouse (SM22(CAG-ChR2-EYFP)), where the arrows indicate the targeted single blood vessel that shrank in response to light stimulations, and the white dashed lines indicate the light patterns; i) No stimulation, enhanced by green pseudo-color, ii) Stimulation using whole pattern, enhanced by red pseudo-color, iii) Image formed by merging (i) and (ii), where yellow indicates the merged region, iv) Cropped and magnified image of the region in the red box in (iii), (v–viii) The distance in the upper left corner of the figures is the distance between the center of the blood vessel and the center of the 60 × 600 μm2 rectangular pattern. The diameter of blood vessel was measured at 2 seconds after optical stimulation. The optical stimulation parameters were controlled by a function generator remote controller, a SLM calibration software control, and the ND filters activated upon blood vessel contraction. The illumination light output intensities are A) 100 μW/mm2, B) 170 μW/mm2, and C) 50 μW/mm2; and the laser stimulation output intensities are A) 1.4 mW/mm2 and C) 200 μW/mm2. The duration of the stimulus pulse is fixed at 2 s; the frame rates are (A, C) 2 frames/s and (B) 3 frames/s; and the exposure times are (A, C) 500 ms and (B) 300 ms. Scale bar: (A, B, C (i–iii, v–viii)) 50 μm, (C (iv)) 10 μm.
Figure 5(C) (i) shows an image of the blood vessels located 4.7 mm below the cortical surface with the illumination light intensity of 50 μW/mm2 for EYP fluorescence imaging, which is sufficiently lower than the threshold intensity. Therefore, vasoconstriction does not occur without 473-nm optical stimulation. This image is somewhat dark under the illumination intensity of 50 μW/mm2, but is nonetheless sufficient to visualize the shape of the blood vessels, as further illustrated in Visualization 5. Figure 5(C) (i) depicts blood vessels without optical stimulation, expressed in green pseudo-color. Figure 5(C) (ii) shows an image of blood vessels stimulated by whole-pattern 473-nm patterned laser light of 200 μW/mm2, expressed as red pseudo-color. The change in the diameter of the horizontal blood vessel is visualized by merging the two pseudo-color images in Figs. 5(C) (iii) and 5(C) (iv). The diameter of the lumen is decreased by 30% after 2 seconds of optical stimulation, from 37 μm to 26 μm. In addition, as shown in Fig. 5(C) (v–viii), when the brain is stimulated with a 60 × 600 μm2 rectangular-shaped light of 200 μW/mm2, vasoconstriction is induced even when the blood vessel deviates from the rectangular-shaped region, but occurs only at a distance of less than 170 μm (Fig. 5(C) (viii)). The 60 × 600 μm2 rectangular-shaped light stimulation pattern was selected to allow complete coverage of a single blood vessel of 37 μm diameter. The images in Fig. 5(C) were captured from Video 5 taken at a frame rate of 2 Hz.
4. Discussion
We have introduced an integrated intravital imaging platform including fiber optic-based epi-fluorescence microscopy and laser patterned-light stimulation systems. The platform successfully performed real-time fluorescence imaging during direct insertion into the intact brain of a live small animal. It allows the imaging probe to avoid damage to blood vessels in the penetration path. Additionally, the platform can optically stimulate the vasoconstriction of cerebral vessels during real-time fluorescence imaging simultaneously. We demonstrated that the proposed platform can accomplish the functions of accurate and wide-field cerebrovascular fluorescent imaging with near-cellular resolution by performing real-time deep-brain imaging of a fluorescently labeled neuron cluster in a living mouse. The vertical imaging function of the insertion optical system can be used to obtain accurate imaging data from deeper brain structures through contextual imaging, without using histological approaches that sacrifice unnecessary animals [8, 19]. It is also possible to reduce the inevitable damage to tissue along the penetration path, such as bleeding, tissue edema, and elevated intracranial pressure, by avoiding blood vessels at the site of penetration using real-time imaging during insertion. All of these advantages are useful for studying cerebral blood vessels and deep-brain disorders in vivo.
To demonstrate that the therapeutic optical stimulation and diagnosis of vascular disorders is possible at the same time in the deep brain, we optically stimulated and monitored, via real-time fluorescence imaging, vasoconstriction using 473-nm laser patterned optical stimulation in a living SM22-ChR2 transgenic mouse. We modulated the vasoconstriction of a single targeted blood vessel of 37 μm in diameter at a depth exceeding 4.7 mm through direct light stimulation of the smooth muscle. The induced vasoconstriction occurred even when blood vessels deviated from the 60 × 600 μm rectangular-shaped light, and this vasoconstriction was maintained only at distances less than 170 μm.
In the attached video (Visualization 5), it can be seen that the blood vessels contract repeatedly only when they are under patterned light stimulation. This means that the light stimulation does not damage the stimulated blood vessels, at least in the short term. Therefore, we can control vascular response using 473-nm laser patterned optical stimulation, not only to induce vasoconstriction with minimal damage, but also to control the appropriate parameters to selectively stimulate a single blood vessel.
We optically stimulated only smooth muscle cells to induce vasoconstriction using optogenetics, which has not been possible via electrical stimulation methods [26]. Vasoconstriction by electrical stimulation can be caused by various effects such as stimulation of the sympathetic nervous system connected to the blood vessels or the direct stimulation of smooth muscle [26, 32]. Thus, the mechanism of electrically induced vasoconstriction is difficult to distinguish clearly between these two aspects. With optogenetic technology and this platform, however, only the smooth muscle cells can be selectively stimulated, allowing clear interpretation of the results of photostimulated vasoconstriction, except for the effects of sympathetic nerves. Conversely, this experimental method can also be applied to studies of vasoconstriction induced by stimulating sympathetic nerves without stimulating smooth muscle cells.
As shown in Fig. 5, vascular contraction during imaging, even in the absence of patterned laser stimulation, was observed in real-time monitoring. This is because the fluorescence absorption band of EYFP overlaps slightly with that of ChR2, and thus ChR2 can be stimulated by the illumination light for fluorescence imaging of EYFP. Therefore, we reduced the illumination light intensity for fluorescence imaging to minimize the unintended activation of ChR2. This allowed us to observe the changes in the blood vessel that occurred during optical stimulation, which was difficult with optical stimulation and imaging systems for optogenetic applications using fluorescent indicators (e.g., GFPs) with absorption wavelength bands that overlap with that of ChR2 [33, 34]. However, we believe that the use of other fluorescence indicators that do not overlap with ChR2, such as RFP, is preferable.
In this study, the application of optogenetic techniques was used only for the control and imaging of vasoconstriction by optical stimulation. However, optogenetic technologies can be extended to new methods of studying neurovascular circuits that allow the observation and control of individual neurons, astrocytes, and vascular cells with excellent spatial and temporal resolution [9, 34–36]. If the developed platform is used as a general optical stimulus tool rather than an optogenetic stimulation, it can be adapted for focal photothrombosis using a photosensitive dye such as Rose Bengal solution [37], which could be applicable to studies of local ischemia and ischemic stroke [3, 26, 38].
Even with the fiber bundle of 330-μm in diameter to minimize brain damage during insertion, much care was necessary to prevent exerting pressure on the brain tissue and blood vessels in the image plane with the fiber bundle during insertion, as shown in Visualization 2. This problem may be mitigated by using an angled fiber bundle or a fiber bundle combined with a gradient index (GRIN) lens that can image at a working distance of a few micrometers in front of the distal end of the fiber bundle. However, additional optics such as the GRIN lens in front of the fiber bundle may cause problems such as a reduced FOV or chromatic aberration. A larger-diameter GRIN lens can be used to maintain the FOV, but may exacerbate tissue damage. To minimize chromatic aberration, an additional optical component should be added, which may complicate the optical probe.
In conclusion, we have developed a fiber bundle-based integrated platform for vasoconstriction research and experimentally demonstrated that the platform can be utilized for cerebrovascular studies that require optical stimulation with temporal and spatial resolution and simultaneous real-time full-field cellular-resolution fluorescence imaging. The integrated platform can be applied to many types of studies in intact brains of living animals at cellular resolution from the surface level to the deep brain. The proposed platform can also be used to monitor neuronal activity using various fluorescence-based methods, such as fluorescent organic encoded activity indicators and genetically encoded calcium indicators [33, 34], as well as studying neurovascular coupling in deep-brain regions to gain insight into brain pathologies such as Parkinson’s disease or Alzheimer’s disease [2–4, 16], both of which are influenced by changes in local neurovascular coupling in deeper brain structures. Lastly, the proposed fiber bundle probe-based integrated optical platform is suitable for behavior research on live animals because the probe is minimally invasive, flexible, and lightweight [17, 24, 25, 39].
Funding
Institutional Grant, KIST (2E26890); Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science and Technology (2017R1A2B2008428).
References and links
1. R. B. Buxton, K. Uludağ, D. J. Dubowitz, and T. T. Liu, “Modeling the hemodynamic response to brain activation,” Neuroimage 23(Suppl 1), S220–S233 (2004). [CrossRef] [PubMed]
2. B. Rosengarten, V. Dannhardt, O. Burr, M. Pöhler, S. Rosengarten, M. Oechsner, and I. Reuter, “Neurovascular coupling in Parkinson’s disease patients: effects of dementia and acetylcholinesterase inhibitor treatment,” J. Alzheimers Dis. 22(2), 415–421 (2010). [CrossRef] [PubMed]
3. H. Girouard and C. Iadecola, “Neurovascular coupling in the normal brain and in hypertension, stroke, and Alzheimer disease,” J. Appl. Physiol. 100(1), 328–335 (2006). [CrossRef] [PubMed]
4. C. Iadecola, “Neurovascular regulation in the normal brain and in Alzheimer’s disease,” Nat. Rev. Neurosci. 5(5), 347–360 (2004). [CrossRef] [PubMed]
5. T. Misgeld and M. Kerschensteiner, “In vivo imaging of the diseased nervous system,” Nat. Rev. Neurosci. 7(6), 449–463 (2006). [CrossRef] [PubMed]
6. R. E. Jacobs and S. R. Cherry, “Complementary emerging techniques: high-resolution PET and MRI,” Curr. Opin. Neurobiol. 11(5), 621–629 (2001). [CrossRef] [PubMed]
7. R. M. Botnar, M. Stuber, K. V. Kissinger, W. Y. Kim, E. Spuentrup, and W. J. Manning, “Noninvasive coronary vessel wall and plaque imaging with magnetic resonance imaging,” Circulation 102(21), 2582–2587 (2000). [CrossRef] [PubMed]
8. M. J. Levene, D. A. Dombeck, K. A. Kasischke, R. P. Molloy, and W. W. Webb, “In vivo multiphoton microscopy of deep brain tissue,” J. Neurophysiol. 91(4), 1908–1912 (2004). [CrossRef] [PubMed]
9. T. J. Richner, R. Baumgartner, S. K. Brodnick, M. Azimipour, L. A. Krugner-Higby, K. W. Eliceiri, J. C. Williams, and R. Pashaie, “Patterned optogenetic modulation of neurovascular and metabolic signals,” J. Cereb. Blood Flow Metab. 35(1), 140–147 (2015). [CrossRef] [PubMed]
10. U. Dirnagl, A. Villringer, and K. M. Einhäupl, “In-vivo confocal scanning laser microscopy of the cerebral microcirculation,” J. Microsc. 165(1), 147–157 (1992). [CrossRef] [PubMed]
11. D. Kleinfeld, P. P. Mitra, F. Helmchen, and W. Denk, “Fluctuations and stimulus-induced changes in blood flow observed in individual capillaries in layers 2 through 4 of rat neocortex,” Proc. Natl. Acad. Sci. U.S.A. 95(26), 15741–15746 (1998). [CrossRef] [PubMed]
12. S. Zhang and T. H. Murphy, “Imaging the impact of cortical microcirculation on synaptic structure and sensory-evoked hemodynamic responses in vivo,” PLoS Biol. 5(5), e119 (2007). [CrossRef] [PubMed]
13. R. T. Megens, S. Reitsma, L. Prinzen, M. G. oude Egbrink, W. Engels, P. J. Leenders, E. J. Brunenberg, K. D. Reesink, B. J. Janssen, B. M. ter Haar Romeny, D. W. Slaaf, and M. A. van Zandvoort, “In vivo high-resolution structural imaging of large arteries in small rodents using two-photon laser scanning microscopy,” J. Biomed. Opt. 15(1), 011108 (2010). [CrossRef] [PubMed]
14. C. B. Schaffer, B. Friedman, N. Nishimura, L. F. Schroeder, P. S. Tsai, F. F. Ebner, P. D. Lyden, and D. Kleinfeld, “Two-photon imaging of cortical surface microvessels reveals a robust redistribution in blood flow after vascular occlusion,” PLoS Biol. 4(2), e22 (2006). [CrossRef] [PubMed]
15. S. Nakamura, D. W. Walker, and F. Y. Wong, “Cerebral haemodynamic response to somatosensory stimulation in near-term fetal sheep,” J. Physiol. 595(4), 1289–1303 (2017). [PubMed]
16. V. Gradinaru, M. Mogri, K. R. Thompson, J. M. Henderson, and K. Deisseroth, “Optical deconstruction of parkinsonian neural circuitry,” Science 324(5925), 354–359 (2009). [CrossRef] [PubMed]
17. M. Murayama and M. E. Larkum, “In vivo dendritic calcium imaging with a fiberoptic periscope system,” Nat. Protoc. 4(10), 1551–1559 (2009). [CrossRef] [PubMed]
18. J. C. Jung, A. D. Mehta, E. Aksay, R. Stepnoski, and M. J. Schnitzer, “In vivo mammalian brain imaging using one- and two-photon fluorescence microendoscopy,” J. Neurophysiol. 92(5), 3121–3133 (2004). [CrossRef] [PubMed]
19. R. P. Barretto, T. H. Ko, J. C. Jung, T. J. Wang, G. Capps, A. C. Waters, Y. Ziv, A. Attardo, L. Recht, and M. J. Schnitzer, “Time-lapse imaging of disease progression in deep brain areas using fluorescence microendoscopy,” Nat. Med. 17(2), 223–228 (2011). [CrossRef] [PubMed]
20. J. C. Jung and M. J. Schnitzer, “Multiphoton endoscopy,” Opt. Lett. 28(11), 902–904 (2003). [CrossRef] [PubMed]
21. L. D. Liao, V. Tsytsarev, I. Delgado-Martínez, M. L. Li, R. Erzurumlu, A. Vipin, J. Orellana, Y. R. Lin, H. Y. Lai, Y. Y. Chen, and N. V. Thakor, “Neurovascular coupling: in vivo optical techniques for functional brain imaging,” Biomed. Eng. Online 12(1), 38 (2013). [CrossRef] [PubMed]
22. Y. Wang, S. Hu, K. Maslov, Y. Zhang, Y. Xia, and L. V. Wang, “In vivo integrated photoacoustic and confocal microscopy of hemoglobin oxygen saturation and oxygen partial pressure,” Opt. Lett. 36(7), 1029–1031 (2011). [CrossRef] [PubMed]
23. T. Liu, Q. Wei, J. Wang, S. Jiao, and H. F. Zhang, “Combined photoacoustic microscopy and optical coherence tomography can measure metabolic rate of oxygen,” Biomed. Opt. Express 2(5), 1359–1365 (2011). [CrossRef] [PubMed]
24. V. Szabo, C. Ventalon, V. De Sars, J. Bradley, and V. Emiliani, “Spatially selective holographic photoactivation and functional fluorescence imaging in freely behaving mice with a fiberscope,” Neuron 84(6), 1157–1169 (2014). [CrossRef] [PubMed]
25. B. A. Flusberg, E. D. Cocker, W. Piyawattanametha, J. C. Jung, E. L. Cheung, and M. J. Schnitzer, “Fiber-optic fluorescence imaging,” Nat. Methods 2(12), 941–950 (2005). [CrossRef] [PubMed]
26. Y. Mandel, R. Manivanh, R. Dalal, P. Huie, J. Wang, M. Brinton, and D. Palanker, “Vasoconstriction by electrical stimulation: new approach to control of non-compressible hemorrhage,” Sci. Rep. 3(1), 2111 (2013). [CrossRef] [PubMed]
27. L. Krizanac-Bengez, M. R. Mayberg, and D. Janigro, “The cerebral vasculature as a therapeutic target for neurological disorders and the role of shear stress in vascular homeostatis and pathophysiology,” Neurol. Res. 26(8), 846–853 (2004). [CrossRef] [PubMed]
28. K. B. J. Franklin and G. Paxinos, The Mouse Brainin Stereotaxic Coordinates (Academic Press, 1997).
29. J. H. Park, J. K. Hong, J. Y. Jang, J. An, K. S. Lee, T. M. Kang, H. J. Shin, and J. F. Suh, “Optogenetic modulation of urinary bladder contraction for lower urinary tract dysfunction,” Sci. Rep. 7, 40872 (2017). [CrossRef] [PubMed]
30. Y. Wu, S. S. Li, X. Jin, N. Cui, S. Zhang, and C. Jiang, “Optogenetic approach for functional assays of the cardiovascular system by light activation of the vascular smooth muscle,” Vascul. Pharmacol. 71, 192–200 (2015). [CrossRef] [PubMed]
31. T. N. Ford, D. Lim, and J. Mertz, “Fast optically sectioned fluorescence HiLo endomicroscopy,” J. Biomed. Opt. 17(2), 021105 (2012). [CrossRef] [PubMed]
32. W. R. Ferrell and A. Khoshbaten, “Responses of blood vessels in the rabbit knee to electrical stimulation of the joint capsule,” J. Physiol. 423(1), 569–578 (1990). [CrossRef] [PubMed]
33. J. Akerboom, N. Carreras Calderón, L. Tian, S. Wabnig, M. Prigge, J. Tolö, A. Gordus, M. B. Orger, K. E. Severi, J. J. Macklin, R. Patel, S. R. Pulver, T. J. Wardill, E. Fischer, C. Schüler, T. W. Chen, K. S. Sarkisyan, J. S. Marvin, C. I. Bargmann, D. S. Kim, S. Kügler, L. Lagnado, P. Hegemann, A. Gottschalk, E. R. Schreiter, and L. L. Looger, “Genetically encoded calcium indicators for multi-color neural activity imaging and combination with optogenetics,” Front. Mol. Neurosci. 6, 2 (2013). [CrossRef] [PubMed]
34. O. Yizhar, L. E. Fenno, T. J. Davidson, M. Mogri, and K. Deisseroth, “Optogenetics in neural systems,” Neuron 71(1), 9–34 (2011). [CrossRef] [PubMed]
35. A. Urban, A. Rancillac, L. Martinez, and J. Rossier, “Deciphering the neuronal circuitry controlling local blood flow in the cerebral cortex with optogenetics in PV:Cre transgenic mice,” Front. Pharmacol. 3, 105 (2012). [CrossRef] [PubMed]
36. P. Bregestovski and Y. Zilberter, “Optogenetics to help exploring the cerebral blood flow regulation,” Front. Pharmacol. 5, 107 (2014). [CrossRef] [PubMed]
37. L. Talley Watts, W. Zheng, R. J. Garling, V. C. Frohlich, and J. D. Lechleiter, “Rose Bengal photothrombosis by confocal optical imaging in vivo: a model of single vessel stroke,” J. Vis. Exp. (100), e52794 (2015).
38. V. Labat-gest and S. Tomasi, “Photothrombotic ischemia: a minimally invasive and reproducible photochemical cortical lesion model for mouse stroke studies,” J. Vis. Exp. (76): (2013).
39. A. M. Aravanis, L. P. Wang, F. Zhang, L. A. Meltzer, M. Z. Mogri, M. B. Schneider, and K. Deisseroth, “An optical neural interface: in vivo control of rodent motor cortex with integrated fiberoptic and optogenetic technology,” J. Neural Eng. 4(3), S143–S156 (2007). [CrossRef] [PubMed]
