All-optical osteotomy to create windows for transcranial imaging in mice

Abstract

Surgical procedures as a prelude to optical imaging are a rate-limiting step in experimental neuroscience. Towards automation of these procedures, we describe the use of nonlinear optical techniques to create a thinned skull window for transcranial imaging. Metrology by second harmonic generation was used to map the surfaces of the skull and define a cutting path. Plasma-mediated laser ablation was utilized to cut bone. Mice prepared with these techniques were used to image subsurface cortical vasculature and blood flow. The viability of the brain tissue was confirmed via histological analysis and supports the utility of solely optical techniques for osteotomy and potentially other surgical procedures.

Fig. 1

Flow chart and schematic of the experimental set-up. (a) Procedures to produce a thinned skull window for transcranial imaging in mice. (b) Schematic of the set-up. An ultrashort pulsed Ti-sapphire oscillator (Mira HP pumped by a Verdi V18; Coherent, Inc.) is used for metrology of the skull based on the detection (H7422-40; Hamamatsu) of second harmonic light. A Ti-sapphire regeneratively amplified laser system (Libra, Coherent Inc.) is used as the source for the laser ablation. The two beams share the optical axis and are merged with a polarizing beam splitter placed before the water dipping objective (LUMPLANFL IR 40X 0.80 NA; Olympus). Translational stages (XYR-6060; Danaher) were used to scan incoming optical beam in X- and Y-directions. A third stage was used to move the preparation along the optical axis, i.e., Z-direction (LMB-600; Danaher). The intensity of the metrology and ablation laser beams were controlled with half-wave plates followed by a polarizing beam splitter; the plates were coupled to a stepper motor and rotated for the desired intensity. The entire procedure is under control of a computer algorithm that operated the shutters (LS3; Uniblitz) and the stage controller (DMC-4040; Galil) and synchronized the data acquisition (NI 6110; National Instruments).

Fig. 2

Height and thickness of the skull as measured via second harmonic generation. (a) Intact bone was scanned along the Z-direction with ultrashort pulsed laser light. The signal rapidly increased as the focus approached the surface of the skull. (b) Map of height versus medial-lateral position for a band across the skull. (c) Overlay of the curvature obtained in panel b with a wide-field fluorescent image of a cross-section of the skull. The section was obtained by mechanically cutting the skull along the previous scanned path.

Fig. 3

Creation of a thinned skull by plasma-mediated laser ablation. (a) The intrinsic curvature of the skull was mapped over a 2 mm by 2 mm area with SHG metrology. (b) A smoothed version of the data in panel a that was used to to calculate translational stage movements and shutter openings for thinning. (c) The decay in the SHG signal when the skull has been cut to within a thickness of 50 µm. Note the sharp fall-off in SHG signal at the interior surface of the skull compared with the signal from a thick skull (Fig. 2(a)). (d) The top surface of the thinned skull window measured with SHG metrology after completion of the ablation process. (e) Photomicrograph of the surface in panel d. (f) Histogram of the thickness obtained from measurements in panel d show the distribution of thicnesses; the points marked “Edges” are at the border of the window.

Fig. 4

In vivo two-photon laser scanning imaging demonstrates the utility of a laser ablated thinned skull window for functional imaging in mice. The blood plasma was stained with a high-molecular weight fluorescein-labeled dextran (2 MDa; Sigma). The window is above parietal cortex. (a) Maximum projection of a stack of images from a depth of 50 µm to 100um below the pia. The imaged area is 300 µm by 300 µm. (b) The same field as in panel a with a projection from 100 µm to 150 µm below the pia. (c) The stacks used for panels a and b were resliced to visualize the vessels along the optical axis. Note the penetrating vessels and fine microvessels. (d) A single microvessel that lies within the focal plane at a depth of 40 µm below the pia. The scan path lies along a straight section of the vessel that is used to track red blood cell flow. (e) Line-scan data from successive scans through the vessel in panel d. Individual red blood cells appear as dark streaks against the fluorescent blood plasma. The velocity of the flow is inversely proportional to the slope of the streaks. (f) Spectral power of the time series of the RBC speed in a vessel, located 80 µm below the pia, shows low frequency vasomotion broadly centred near 0.2 Hz and the heart rate at 8 Hz.

Fig. 5

Immunological assays of the viability of parietal cortex after laser ablation to create a transcranial window and functional imaging through the window. Alternate brain sections were stained with immunological markers to detect possible damage from the laser ablation and imaging. (a) Staining for reactive astrocytes with α-GFAP visualized with a fluorescent secondary antibody. The tissue under the thinned skull window had α-GFAP levels comparable to the contralateral control side. (b) Staining for the pan-neuronal marker α-NeuN, which lables essentially all neuronal nuclei. Both the window and control sides have comparable neuronal cell densities. (c) Staining for microtubules, as a test of neuronal integrity, with α-MAP. The processes of the neurons are preserved for both the window and control sides of the section.