2.1.

Subjects

Male mice (20 to 25 g, $n=89$) were used in all experiments. The animals were housed under standard laboratory conditions, with access to food and water, ad libitum. All procedures were performed in accordance with the “Guide for the Care and Use of Laboratory Animals.” The experimental protocol was approved by the Committee for the Care and Use of the Laboratory Animals at the Saratov State University (Protocol H-147, 17.04.2001).

2.2.

Assessment of the BBB Permeability

Four different techniques were used to evaluate the BBB disruption:

During the imaging, mice were kept under inhalation anesthesia with 2% isoflurane at $1\mathrm{L}/\min$ ${\mathrm{N}}_2\mathrm{O}/{\mathrm{O}}_2-70:30$. The body temperature was maintained at 37.5°C by a homoeothermic blanket system with rectal probe (Harvard Apparatus, Holliston, Massachusetts). The brain temperature was kept at 37°C using an objective heating system with a temperature probe (Bioptechs Inc., Butler, Pennsylvania). The fluorescein-labeled dextran (70 kDa, Sigma-Aldrich) in physiological saline (5% wt/vol) was injected through the tail vein ($\sim 100\mu \mathrm{l}$) at an estimated initial concentration in blood serum of 150 Mm.²¹

Imaging was done using Prairie View Ultima system with an Olympus BX51WI upright microscope and water-immersion LUMPlan FL/IR $20\times /0.50\mathrm{W}$ objective. The excitation was provided by a Prairie View Ultima multiphoton laser scan unit powered by a Millennia Prime 10 W diode laser source pumping a Tsunami Ti:sapphire laser (Spectra-Physics) tuned up to 810 nm center wavelength. Fluorescence was bandpass filtered at 510 to 550 nm. Real time images ($512\times 512\text{pixels}$, $0.15\mu \mathrm{m}/\text{pixel}$ in the $x$- and $y$-axes) were acquired using the time series mode of the Prairie View software (30-s intervals between images, 30 min sessions). Between the sessions (before and 1 to 4 to 24 h after sound exposure) mice were awake and were kept at the animal facility.

In offline analyses using ImageJ, microvascular BBB permeability was evaluated by measuring changes in perivascular tissue fluorescence in planar images of the cortex taken 50 and $150\mu \mathrm{m}$ depth 20 min after fluorescein dextran injection, as described previously.²⁰ On each image, fluorescence of 10 randomly chosen regions of interest over the vessel areas and 10 corresponding perivascular brain parenchyma areas were measured. The obtained measurements in the interstitial space were normalized to the maximal fluorescence intensity in the blood vessels and expressed as a percentage of fluorescence intensity (modified technique from Egawa et al.²²). The estimated quantity of fluorescein that leaked out of the vessel and left in the vessel 20 min after the injection was calculated from the initial vascular fluorescence equal to $\sim 150\mu \mathrm{M}$ of fluorescein. The time courses of fluorescein leakage for each group were also plotted.

To image the meningeal lymphatic vessels, 2PLSM was performed as described above with some modifications. Just before 2PLSM imaging, mice were preanesthetized in a box with 5% isoflurane in ${\mathrm{N}}_2\mathrm{O}/{\mathrm{O}}_2-50:50$, secured on stereotactic frame, and switched to inhalation anesthesia with 2% isoflurane at $1\mathrm{L}/\min$ ${\mathrm{N}}_2\mathrm{O}/{\mathrm{O}}_2-70:30$. To label the lymphatic vessels, mice were injected i.c.v. (into the cisterna magna) with 5 ml of Alexa 488-conjugated anti-Lyve-1 antibody (ALY7, eBioscience) as described in Ref. 1 and moved to the 2PLSM setup. After baseline imaging, mice underwent sound treatment, and imaging was repeated at 1 h after the end of sound treatment. In offline analysis, using a National Institutes of Health, diameters of the lymphatic vessels were quantified.

2.3.

Optical Coherence Tomography

The OCT imaging of meningeal lymphatic vessels was performed with the commercial OCT system (Thorlabs Ganymede) providing lateral and transverse resolution of this system is $6.2\mu \mathrm{m}$ (on air) and $8\mu \mathrm{m}$, respectively, with the corresponding 1.3 and $3\mu \mathrm{m}$ pixel size in the resulting image. The central wavelength of the OCT light source is 930 nm. Each B-scan consists of 768 A-lines. Contrast enchantment was done by equalizing the histograms of the images. Due to the fact that the lymph is optically transparent within a broad range of wavelengths, in the resulting B-mode OCT image, the lymphatic vessels will look like an “empty” tube with the inverted signal-to-noise ratio regarding background; this is a so-called “negative” contrast. The main feature of the meningeal lymphatic vessels is the lack of shadowing and decorrelation “tail” beneath them during the lack of high scattering turbid media (please see Video 1 for details). These two features allow us to easily identify optically resolvable vessels inside the tissue.

The OCT imaging was performed in the groups (1) before sound (the control, $n=5$) and 1 h after sound ($n=7$).

2.4.

Immunohistochemical Analysis

For the visualization of the meningeal lymphatic vessels, we used the protocol for the immunohistochemical (ICH) analysis with specific marker LYVE1—the lymphatic vessel endothelial hyaluronan receptor 1.^23–25 Tissue samples were fixed with formaldehyde and, after routine processing, were embedded into a paraffin block. Then samples were sectioned into 3- to $5\text{-}\mu \mathrm{m}$ sections; the sections were attached to poly-l-lysine-coated glass slides; they were dried at 37°C for 24 h; and they were sequentially incubated in xylene (three times, 3 min each), 96% ethanol (three times, 3 min each), 80% ethanol (two times, 3 min each), and distilled water (three times, 3 min each). The IHC reaction was visualized with a REVEAL—Biotin-Free Polyvalent diaminobenzidine kit (Spring Bioscience). Endogenous peroxidases were blocked by adding 0.3% hydrogen peroxide to the sections for 10 min followed by washing the sections in phosphate buffered saline (PBS). The antigen retrieval was conducted using a microwave oven in an ethylenediaminetetraacetic acid-buffer pH 9.0, and nonspecific background staining was blocked in PBS containing 0.5% bovine serum albumin and 0.5% casein for 10 min, after which the sections were washed in PBS for 5 min. Further, the sections were incubated in a humid chamber with diluted anti-LYVE1 antibody ($1:1000$) for 1 h at room temperature. After that, the sections were washed in PBS, incubated with secondary horseradish peroxidase-labeled goat antirabbit antibodies for 15 min, again washed in PBS, counterstained with hematoxylin for 1 min, washed in water, dehydrated in graded alcohols (three times, 3 min each) and then in xylene (three times, 3 min each), and finally embedded into Canadian balm.

The ICH analysis was performed after the OCT imaging in the same groups.

2.5.

Confocal Imaging of Lymphatic Drainage after the Opening of BBB

To induce opening of the BBB, we used an audible sound (110 dB, 370 Hz): 60 s sound, then 60 s pause, and this cycle was repeated for 2 h.²⁶ This procedure was performed using a sound transducer in a Plexiglas^® chamber (with volume of $2000{\mathrm{cm}}^3$) amplifying the effect of sound on the mice and not passing sound out.

To study the lymphatic drainage, we used the following experimental design: the mice underwent by sound during 2 h; then, 1 h after sound when the BBB opened, FITC-dextran 70 kDa was injected intravenously ($1\mathrm{mg}/25\mathrm{g}$ mouse, 0.5% solution in 0.9% physiological saline, Sigma-Aldrich) and allowed to circulate for 20 min. Afterward mice were anaesthetized by ketamine/xylazine injection i.p. and injected i.c.v. (into the cisterna magna) with 5 ml of Alexa488-conjugated anti-Lyve-1 antibody (ALY7, eBioscience). Ten minutes after this procedure, the mice were decapitated; their brains were quickly removed, fixed in 4% PFA for 24 h, cut into $50\text{-}\mu \mathrm{m}$ thick slices on vibratome (Leica VT 1000S Microsystem, Germany), and analyzed on a confocal microscope (Olympus FV10i-W, Olympus, Japan). Approximately 8 to 12 slices per animal were imaged. The confocal imaging was performed in the groups (1) before sound (control, $n=7$) and 1 h after sound ($n=7$).

2.6.

Statistical Analysis

The results are presented as a $\text{mean}\pm \text{standard error}$ of the mean (SEM). Differences from the initial level in the same group were evaluated by the Wilcoxon test. Intergroup differences were evaluated using the Mann–Whitney test and two-way analysis of variance (post hoc analysis with the Duncan’s rank test). Significance levels were set at $p<0.05$ for all analyses.

3.1.

Sound-Induced Opening of BBB

In the first step of our study, we analyzed the BBB permeability to EBd and FITC-dextran as a classical test for BBB integrity.

Confocal microscopy revealed the accumulation of FITC-dextran outside of the brain capillaries at 1 h after sound exposure. Extravasation of FITC-dextran was determined as a clear fluorescence visualized outside of the vessel walls. Vascular leakage of FITC-dextran was defined as bright red clouds associated with a group of capillaries [Fig. 1(b)]. At 4 and 24 h after sound, the BBB rapidly recovered, which was accompanied by closing of the BBB to FITC-dextran. No extravasation of FITC-dextran was observed in the control group (no sound) [Fig. 1(a)].

Fig. 1

BBB permeability to FITC-dextran 70 kDa and EBd with albumin complex 68 kDa: (a) before sound exposure (no extravasation of FITC-dextran); (b) 1 h after sound exposure [strong extravasation of FITC-dextran (arrowed), which was defined as bright red clouds associated with a group of capillaries]; (c) the estimation of FITC-dextran concentration in brain parenchyma and microvessels 20 min after the injection. 2PLSM data are presented as $\text{mean}\pm \mathrm{SEM}$, $n=7$, **$p<0.01$; (d) EBd in $\mu \mathrm{g}/\mathrm{g}$ of mouse brain tissue before and after sound. Spectrofluorometric assay of EBd extravasation into the brain parenchyma shows an increase in the BBB permeability to EBd 1 h after the sound with rapid restoring of the BBB afterward.

2PLSM of FITC-dextran leakage, injected intravenously, confirmed the results obtained ex vivo using confocal microscopy imaging. The analysis of microvascular permeability to FITC-dextran using 2PLSM was performed on each of 7 mice before (no) and 1, 4, and 24 h after the sound exposure. Indeed, 2PLSM imaging showed an impaired BBB 1 h after the sound and the normalization of BBB permeability to FITC-dextran 4 and 24 h after the sound exposure [Fig. 1(c)].

Leakage of EBd also significantly increased 1 h after sound exposure [Fig. 1(d)]. Indeed, 1 h after the sound, a 19.7-fold increase in the leakage of EBd was observed ($7.3\pm 0.09$ versus $0.37\pm 0.02$, $\mu \mathrm{g}/\mathrm{g}$ of tissue, $p<0.05$). Four hours and twenty-four hours after the sound, the BBB permeability to EBd was completely normalized and the level of dye in the brain tissues was similar to the normal state ($0.54\pm 0.01$, $\mu \mathrm{g}/\mathrm{g}$ of tissue and $0.40\pm 0.03$, $\mu \mathrm{g}/\mathrm{g}$ of tissue, respectively).

Thus, in the first series of experiments, we clearly show that sound causes a temporal opening of the BBB with maximal effects 1 h after sound exposure.

3.2.

Visualization of the Meningeal Lymphatic Vessels before and after the Opening of BBB

In the second step, we analyzed the interaction between the meningeal lymphatic vessels and the increasing of BBB permeability. Following this task, we analyzed the changes in diameter of the meningeal lymphatic vessels in three independent groups using in vitro IHC analysis and in vivo OCT and 2PLSM.

OCT and 2PLSM data show similar depths of the meningeal lymphatic vessels, which is 30 to $50\mu \mathrm{m}$ from the surface of the brain, i.e., superficially in the meningeal layers of the brain [Figs. 2(b) and 3(c)].

Fig. 2

ICH (a and b) and in vivo two-photon laser scanning (c and d) visualization of the meningeal lymphatic vessels: (a) and (b) the meningeal lymphatic vessels stained with LYVE1 (specific marker of endothelial cells of lymphatic vessels, the brown color is positive staining); (a) before sound-induced opening of the BBB, the average diameter of lymphatic vessels is $3.00\pm 0.01\mu \mathrm{m}$; (b) after opening of the BBB (1 h after sound), the average diameter of some lymphatic vessels is $10.00\pm 0.07\mu \mathrm{m}$; (c) and (d) the meningeal lymphatic vessels before and after opening of the BBB, respectively. 2PLSM data show the increase of diameters of the lymphatic vessels after opening of the BBB: ($12.00\pm 0.03\mu \mathrm{m}$ versus $17.00\pm 0.08\mu \mathrm{m}$, $p<0.05$).

Fig. 3

The OCT images of the meningeal lymphatic vessels (black empty spaces) surrounding the sigmoid (a) and sagittal (b) sinuses. (c) The OCT analysis of diameter and depth of the meningeal lymphatic vessels. Scale bar is $100\mu \mathrm{m}$ (Video 1, MP4, 7.07 MB [URL: http://dx.doi.org/10.1117/1.JBO.22.12.121719.1]).

The IHC and 2PLSM analysis revealed that in the control group (before opening of the BBB) the average diameter of the meningeal lymphatic vessels is $3.00\pm 0.01\mu \mathrm{m}$ (IHC data) and $12.00\pm 0.03\mu \mathrm{m}$ (2PLSM data). When the BBB permeability increased (1 h after sound), some of the lymphatic vessels became larger with the average diameter—$10.00\pm 0.07\mu \mathrm{m}$ (IHC data) and $17.00\pm 0.08\mu \mathrm{m}$ (2PLSM data) [Figs. 2(a)–2(d)]. The differences in the size of the diameter of the meningeal lymphatic vessels can be explained by different regions of the brain taken for the analysis. Technically 2PLSM was performed in the area of superior sagittal sinus, an unpaired large venous channel between the endosteal and meningeal layers of dura mater that runs in a sagittal plane from the anterior aspect of the falx cerebri to its termination at the confluence of sinuses at the occipital protuberance. The IHC data were obtained from the area of sigmoid sinus, which begins beneath the temporal bone and follows a tortuous course to the jugular foramen, at which point the sinus becomes continuous with the internal jugular vein. The different sizes of the meningeal lymphatic vessels were also discussed by Aspelund et al.² Another possible reason is brain fixation, performed before brain harvesting, which could cause the lymphatic vessels to shrink.

The OCT can image the meningeal lymphatic vessels only after opening of the BBB but not in the normal state due to the very small size of the lymphatic vessels in the norm. Figures 3(a) and 3(c) show the OCT images of the meningeal lymphatic vessel (black empty spaces) surrounding the sagittal and sigmoid sinuses; the average diameter of the visible lymph vessels after the increase of BBB permeability is $-19.00\pm 0.11\mu \mathrm{m}$. Video 1 shows the decorrelation speckle beneath the blood vessels, which are close to the meningeal lymphatic vessels.

To be sure that we visualize the meningeal lymphatic vessels by the OCT, which look as a black empty space, we analyzed the size of the perivascular space (PVS), which also can potentially give a dark space on the OCT, especially after opening of the BBB when fluids move from the vessels into the PVS. Figure 4 shows that the PVS do not change significantly 1 h after the sound-induced opening of the BBB compared with the control. Due to this reason, we cannot detect the PVS in mice using 2PLSM in both the normal state and after opening of the BBB. The size of PVS is smaller than the black empty space on the OCT ($3.00\pm 0.03\mu \mathrm{m}$ versus $19.00\pm 0.11\mu \mathrm{m}$, respectively), which allows us to conclude that the black empty space is not the PVS or the PVS can be included only partly in the OCT imaging. Additionally, on the Video 1, we clearly show the differences between the OCT signals from the cerebral vein and the meningeal lymphatic vessels.

Fig. 4

The histological analysis of size of PVS (a) before and (b) 1 h after sound-induced opening of the BBB $3.8\mu \mathrm{m}$—in the normal state and $3.2\mu \mathrm{m}$ 1 h after the opening of the BBB.

3.3.

Lymphatic Drainage after the Opening of BBB

Since in our previous steps of study we observed increasing of diameter of the meningeal lymphatic vessels after opening of the BBB, we hypothesized that it can be associated with the drainage and clearing function of lymphatics. To test our hypothesis, we tried to answer the question, “If we open the BBB for dextran, is it possible to see this fluorescent tracer in the meningeal lymphatic vessels as one of the possible ways for its clearing from the brain?”

To answer this question, mice underwent sound during 2 h. Then 1 h after sound when the BBB opened, FITC-dextran was injected intravenously and circulated for 20 min. Then, the mice were anaesthetized and injected into the cisterna magna with Alexa488-conjugated anti-Lyve-1 antibody for the labeling of the meningeal lymphatic vessels. Ten minutes after this procedure, the mice were decapitated, and their brains were quickly removed for the confocal microscopy protocol. Thus, 30 min was the time for FITC-dextran extravasation through BBB, its accumulation in PVS and parenchyma, and then for clearance.

Figure 5 shows representative confocal image of the meningeal lymphatic vessels after opening of the BBB. Our results clearly demonstrate the presence of FITC-dextran directly in the meningeal lymphatic vessels after opening of the BBB.

Fig. 5

The confocal image of meningeal lymphatic vessels after the sound-induced opening of the BBB: the dextran (red color) is obseved in the meningeal lymphatic vessels (green color) 30 min after the opening of the BBB.