2.1.

Animal Preparation

Male C57BL/6 mice weighing 22 to 28 g were used for this study. In total, 90 mice were used, and 10 of those mice did not undergo craniotomy procedures, serving as the control group. All animal procedures were performed under guidelines set by the Institutional Animal Care and Use Committee of Sungkyunkwan University.

2.2.

Cranial Window Installation

Animals were anesthetized by inhalation with isoflurane (3% for the induction of anesthesia and 1.5% for maintenance during the surgical procedure; Hana Pharm, Republic of Korea). They were then fixed in a stereotaxic frame (David Kopf Instruments) for a full craniotomy with their body temperature kept at 37°C to 37.5°C using a heating pad-connected temperature controller (DC temperature controller, FHC). The skin on the mouse’s head was sanitized with an alcohol swab (isopropyl alcohol 70%, BD) and cut with surgical scissors. Then, the epidermal skin was removed, and cyanoacrylate glue (Loctite) was applied gently to hold the residual skin.

To monitor brain waves of the entire brain, a burr hole (1.2-mm diameter) was created over the left hemisphere olfactory bulb area with a dental drill (Ram Products, Inc.). A micrometer screw (Worcester Polytechnic Institute) was then inserted into the hole for electrocorticographic recording. For cranial window implantation, another hole was made in the right hemisphere, 4 mm in diameter centered at a point 3.3 mm from the midline, and 1.3 mm from the bregma. For the soft cranial window installation, PDMS (elastomer:curing agent = 10:1, over 24 h at 75°C, Sylgard 184, Dow Corning) film–corresponding to a conventional coverslip–was affixed all the way around the edge of the skull with cyanoacrylate glue. PDMS film of 320- to $340\text{-}\mu \mathrm{m}$ thickness was used in order to avoid artifacts from pulsation and breathing. During these procedures, extreme caution was taken not to touch the cortical tissue. The applied glue was allowed to dry for more than 10 min. Then, a chamber plate ($8\mathrm{mm}\times 10\mathrm{mm}\times 1\mathrm{mm}$, Narishige Co., Japan) was glued on the same plane with PDMS. The window edge was sealed with dental resin (OA2, Dentkist Inc., Republic of Korea), which was then hardened by UV exposure. The peripheral tissues were disinfected by povidone-iodine (Sung Kwang Pharm, Republic of Korea). After the cranial window implantation surgery, meloxicam ($1\mathrm{mg}/\mathrm{kg}$, Boehringer Ingelheim, Germany) and Baytril ($5\mathrm{mg}/\mathrm{kg}$, Bayer, Germany) were injected subcutaneously. Animals were supplied with diluted tylenol ($1\mathrm{ml}/10\mathrm{ml}$, Janssen) in water to relieve pain, and they were individually housed in their cages until the experimental sessions. The detailed schematic of the soft cranial window is illustrated in Fig. 1(a).

Fig. 1

Soft cranial window installation and the time course of reactive astrocyte and microglial expression postsoft cranial window installation. (a) Schematic of the soft cranial window installation with PDMS. (b) and (c) Quantification of GFAP(+) cells (reactive astrocyte) and Iba-1(+) cells of the cortex within the soft cranial window area at different postsurgery time points ($n=5$ in each group). (d) Immunohistochemistry of normal, 2 weeks postsurgery, and 6 weeks postsurgery mice. GFAP levels of astrocytes and microglia were highly activated in the brain of the animals at 2 weeks postsoft cranial window installation (*$p<0.05$, **$p<0.01$, ***$p<0.001$).

2.3.

Immunohistochemistry

Animals at 1 to 6 weeks postsoft cranial window installation were perfused intracardially with saline and fixed with 4% paraformaldehyde. Brains were then extracted after perfusion and conserved in 4% PFA at 4°C for postfixation. After 24 h in PFA, each brain was immersed in 30% sucrose and 0.1% sodium azide solution for 24 h, and the media was exchanged. The brains were frozen rapidly in liquid nitrogen and embedded with an optimal cutting temperature compound (Leica Microsystems, Germany). Each brain was coronally sectioned at $40\mu \mathrm{m}$ by cryostat (Leica CM1950, Leica Microsystems). Brain slices containing the barrel cortex area, 0.62 to 1.94 mm from the bregma, were selected for detailed analysis. Fluorescent image acquisition was performed by confocal laser scanning microscopy (TCS SP8, Leica Microsystems) with a white light laser (Leica white light laser, Leica Microsystems). The following primary antibodies were used: 4′,6-diamidino-2-phenylindole (DAPI) (Sigma Aldrich, 1:1000), Iba-1 (a marker of microglia, 1:300, rabbit, Wako, Japan), and GFAP (a marker of reactive astrocyte, 1:300, mouse, Millipore).

2.4.

Cell Counting and Quantification

The number of GFAP+ cells and Iba+ cells in the region of interest (RoI) was analyzed using IMARIS software (Oxford Instruments, United Kingdom). The cells that overlapped with DAPI signals were counted. In addition, GFAP+ cells that had an almost full shape were counted.

2.5.

Optical Imaging of Cerebral Blood Volume during Whisker Stimulation

Optical imaging of the intrinsic signal (ORIS) associated with changes in cerebral blood volume (CBV) was performed on animals at 2 and 6 weeks postcranial window surgery. For a single ORIS session, mice from each group were anesthetized by urethane ($1.3\mathrm{g}/\mathrm{kg}$, Sigma-Aldrich) through intraperitoneal injection. Animals were laid on a bed of paper towels and kept warm with a temperature controller (DC temperature controller, FHC). Each animal’s head was fixed onto a stereotaxic frame (Narishige Co., Japan). An LED lamp (CLS150LED, Leica Microsystems) was used for even illumination of the cranial window area. The reflected light signal from the surface was collected by an optical imager system (Imager 3001-Celox, Optical-Imaging, Inc., Jerusalem, Israel) with a CCD camera (PhotonFocus AG, Switzerland). To image CBV, a 570 nm bandpass filter was selected. This wavelength is an isosbestic wavelength sensitive to both oxygenated hemoglobin and deoxygenated hemoglobin; therefore, the measurement reflects total hemoglobin change (i.e., hemodynamic changes). The imaging session consisted of 20-s recordings (with 40 trials). The sampling rate was 10 Hz, and the pixel resolution was $17.7\times 17.7\mu \mathrm{m}$. For repeated ORIS imaging, animals were anesthetized with ketamine (Yuhanyanghang, Republic of Korea, $100\mathrm{mg}/\mathrm{kg}$) and xylazine (Rompun, Bayer, Germany, $10\mathrm{mg}/\mathrm{kg}$) through intraperitoneal injection.

For whisker stimulation, the C1 and C2 whiskers were stimulated with a piezo actuator (plate benders, Noliac, Denmark) and controlled by a pulse stimulator (Master 9, A.M.P.I., Israel). The stimulation was applied for 4 s. Whole whiskers were stimulated by air puffing from a picospritzer (Picospritzer III, Parker Hannifin) for 4 s and controlled by a pulse stimulator (Master 9, A.M.P.I., Israel).

2.6.

Data Analysis

Acquired optical images were analyzed with a MATLAB computer program (Mathworks, Inc.) written to calculate and display the hemodynamic changes. Baseline was defined as the average value obtained from the frames in the 5-s period prior to whisker stimulation. The values from frames obtained after the whisker stimulation were divided by the baseline ($\mathrm{\Delta }R/R$). Spatial mapping of $\mathrm{\Delta }R/R$ changes was displayed in a two-dimensional format, and hemodynamic changes were carefully examined. The time course of pixel changes from the selected RoIs ($6\times 6\text{pixels}$, 1 pixel: $17.7\mu \mathrm{m}\times 17.7\mu \mathrm{m}$) was plotted. The maximum $\mathrm{\Delta }R/R$ change and the time to the maximum $\mathrm{\Delta }R/R$ changes were quantified and averaged within each group. Spatial extent was also obtained by a MATLAB computer program. To quantify spatial extent, we selected the divided frame that the maximum $\mathrm{\Delta }R/R$ value of pixel existed in and counted the pixels that had values over 50% of the maximum value. The shape of the mouse barrel column was taken from Krook-Magnuson et al.¹⁴

2.7.

In Vivo Electrophysiology Local Field Potential Recording and Data Analysis

Following ORIS imaging, electrophysiological recordings were performed using tungsten electrodes (FHC) attached to a headstage (Plexon Instruments, Inc.). Three tungsten electrodes with 300 to $500\mathrm{k}\mathrm{\Omega }$ were affixed in a straight line. The central electrode was located on the peak hemodynamic activation site in the C2 column of the barrel cortex. The other electrodes were located at $625\mu \mathrm{m}$ rostral and caudal to the center location. Electrodes were stereotaxically inserted into cortex layer 2–3 (300 to $400\mu \mathrm{m}$ from the surface). Animals were given a 0.1 ms C2 single whisker deflection by a piezo actuator (plate benders, Noliac) controlled by a pulse stimulator (Master 9, A.M.P.I., Israel). LFPs were acquired using Plexon electrodes (Plexon Instruments, Inc.) and analyzed by software written in MATLAB (Mathworks, Inc.).

2.8.

Inflammatory Cytokine Assay

The cortex tissue under the cranial window area was homogenized with radioimmunoprecipitation assay buffer containing a 0.01% phosphatase inhibitor. After centrifuging the sample at 4°C and 12700 rpm, we performed protein extraction. The BCA protein assay kit (Pierce Rapid Gold BCA Protein Assay, Thermo Fisher Scientific) was used for measuring protein concentration. Equal amounts of protein (1500 mg) were used to estimate the concentration of inflammatory cytokines [interleukin-$1\beta$ (IL-$1\beta$), IL-6, and IL-10] by separate enzyme-linked immunosorbent assay (ELISA) kits (Mouse IL-1 beta/IL-1F2 Quantikine ELISA Kit, Mouse IL-6 Quantikine ELISA Kit, Mouse IL-10 Quantikine ELISA Kit, R&D Systems, respectively).

2.9.

MinoNAC Administration after Soft Cranial Window Installation

MinoNAC [minocycline (minocycline hydrochloride, Sigma-Aldrich) and $N$-acetyl-l-cysteine (NAC, Sigma-Aldrich)] was mixed together and prepared as $5\mathrm{mg}/\mathrm{ml}$ stock solution in phosphate-buffered saline and aliquoted. It was preserved frozen at $-80°\mathrm{C}$. After soft cranial window installation, mice were administered minoNAC intraperitoneally (minocycline $22.5\mathrm{mg}/\mathrm{kg}$, NAC $75\mathrm{mg}/\mathrm{kg}$) every day for 3 days postsoft cranial window surgery. ORIS imaging and LFP recording were performed at 2 weeks postsurgery under urethane anesthesia. C2 whiskers were stimulated for 4 s by a piezo actuator (plate benders, Noliac) and controlled by a pulse stimulator (Master 9, A.M.P.I., Israel). Experimental procedures for ORIS and LFP recording were the same as those described above.

2.10.

Statistical Analysis

The Wilcoxon signed rank test was used to compare $\mathrm{\Delta }R/R$ changes at 2 and 6 weeks within the same animals. The Mann–Whitney U test was performed for comparison between 2 and 6 weeks groups. Bonferroni correction by ranks was conducted for group comparison of cell quantification data and inflammatory cytokine levels. Independent $T$-test was conducted for comparison of LFP amplitudes between groups.

In the data figures, data are presented as the $\text{mean}\pm \text{standard error}$ of the mean. However, LFP amplitudes and cell quantification are presented as the $\text{mean}\pm \text{standard}\text{deviation}$. Statistical significance in the figures is indicated at one of three levels. *$p<0.05$, **$p<0.01$, ***$p<0.001$.

3.1.

Immunohistochemical Analysis of Cortical Tissue at Different Time Points Following Soft Cranial Window Installation

Each week, for up to 6 weeks after the soft cranial-window installation surgery, brains from a subset of mice that had undergone the procedure were perfused and prepared for immunohistochemical analysis. The samples were stained for GFAP, a reactive astrocyte marker,¹⁵ and Iba-1, a microglia marker related to the brain’s innate immune response,^16,17 as shown in Fig. 2(a).

Fig. 2

The expressions of reactive astrocytes and microglia. (a) Immunohistochemistry of reactive astrocytes and microglia at 1-week interval from 1 to 6 weeks in normal mice and soft window installation mice. (b) Immunohistochemistry of whole hemispheres at 2 and 6 weeks postsoft window installation mice.

The immunohistochemical results indicated that at 2 weeks after surgery, both GFAP and lba-1 showed the highest activity level [i.e., rapid increases in the number of activated astrocytes and microglia; Fig. 1(b)]. After this point, the number of activated cells gradually decreased to the level of normal control mice by 6 weeks after installation of the soft cranial window. All $p$ values are shown in Tables 1 and 2.

Table 1

The p value of GFAP (+) expression between groups of normal, at 1 to 6 weeks postsoft window installation.

Table 2

The p value of Iba-1(+) expression between groups of normal, at 1 to 6 weeks postsoft window installation.

In particular, the microglia in tissues obtained at 1 and 2 weeks postsurgery were highly clustered at the cortex surface, a similar pattern to that of microgliosis.¹⁸ This clustering was not observed at other postoperative time points or in normal mice. Differential morphological features of microglia were also noticed relative to the postoperative period. Microglia in mice 2 weeks postsurgery were in a highly activated form, known as the amoeboid form, with increased size of soma and disappearance of branches. At 6 weeks postinstallation, the microglia had returned to the same morphology as that found in control animals [Fig. 1(d)]. Based on this immunohistochemical analysis, we focused on the outcomes at two time points—2 and 6 weeks postoperatively—to examine the effect of soft cranial window installation on hemodynamic and neuronal activation over time.

Reactive astrocyte expression showed a pattern quite similar to the microglia. Morphology of reactive astrocyte has been shown to differ with physiological condition.¹⁹ In mice at 2 weeks postsurgery, the morphology of reactive astrocytes was much enlarged like the phagocytic form and showed features like astrogliosis. However, at 6 weeks after window installation, the expression level of reactive astrocytes was similar to that of normal animals [Fig. 1(d)]. The expression of reactive astrocyte and microglia in whole hemispheres at 2 and 6 weeks is shown Fig. 2(b).

3.2.

Altered Hemodynamic Changes during Whisker Stimulation at 2 and 6 Weeks Postsoft Cranial Window Surgery

Hemodynamic changes during whisker stimulation were compared between 2 and 6 weeks postsoft cranial window installation within the same animals when the whole whisker pad was stimulated with air-puffing [Fig. 3(a)]. $\mathrm{\Delta }R/R$ changes at 2 weeks postsurgery were smaller than those at 6 weeks [maximal $\mathrm{\Delta }R/R$; 2 weeks: $0.53\pm 0.05\%$, 6 weeks: $1.18\pm 0.12\%$, $p\text{value}=0.04$, $n=5$ in both cases; Figs. 3(b)–3(d)]. There is no significant difference between vessel diameters (at resting state) between at 2 and at 6 weeks postsurgery (see Sec. 5.1).

Fig. 3

Hemodynamic signal change ($\mathrm{\Delta }R/R$) within animals at 2 and 6 weeks postsoft cranial window installation. (a) Optical recording of intrinsic signal (ORIS) set up with whole whisker stimulation (4 s) by air puff. A 570-nm bandpass filter was used. (b) Spatial activation map of ORIS imaging for whole-whisker stimulation in an animal at 2 or 6 weeks postsoft cranial window installation. (c) The graph of $\mathrm{\Delta }R/R$ change of within animals at 2 and 6 weeks postsoft cranial window installation ($n=5$). (d) The peak value of $\mathrm{\Delta }R/R$ change changes within animals at the 2 and 6 weeks postsoft cranial window installation ($n=5$; $p\text{value}=0.0431$).

In addition, when a single whisker (C2) was stimulated for 4 s with a piezo actuator [Fig. 4(a)], $\mathrm{\Delta }R/R$ changes at 2 weeks postsurgery were smaller than those at 6 weeks [maximal $\mathrm{\Delta }R/R$: $2\text{weeks}=0.53\pm 0.14\%$, $6\text{weeks}=1.19\pm 0.14\%$, $p\text{value}<0.001$, $n=10$ in both cases; Figs. 4(b) and 4(c)]. After acute surgery of soft cranial window installation, hemodynamic change during whisker stimulation was lower than each of those between 2 and 6 weeks groups (see Sec. 5.5; peak value of $\mathrm{\Delta }R/R$ change: $\text{acute}=0.31\pm 0.02\%$, $2\text{\hspace{0.17em}\hspace{0.17em}weeks}=0.53\pm 0.14\%$, $6\text{\hspace{0.17em}\hspace{0.17em}weeks}=1.19\pm 0.14\%$, $p\text{value}<0.001$). These experiments showed that surgical process can affect hemodynamic response immediately and that at 2 weeks postsurgery, the hemodynamic responses were still significantly reduced compared with those at 6 weeks.

Fig. 4

Hemodynamic change upon piezo stimulus of mice at 2 or 6 weeks postsoft cranial window implantation. (a) Optical recording of intrinsic signal (ORIS) set up with single-whisker stimulation (C2 whisker, 4 s) by piezo actuator. A 570-nm bandpass filter was used. (b) Spatial activation map of ORIS imaging for single-whisker stimulation. (c) Graph of $\mathrm{\Delta }R/R$ change ($n=10$ in each group). (d) Peak value of $\mathrm{\Delta }R/R$ change ($n=10$ in each group, $p<0.001$). (e) Spatial extent map of $\mathrm{\Delta }R/R$ change at the activated site following piezo stimulus. The spatial extent map described the areas that have $\mathrm{\Delta }R/R$ change values within 50% of the peak response value of $\mathrm{\Delta }R/R$ in whole frames. (f) Pixel counts in the spatial extent maps ($p\text{value}=0.03546$). (g) The spatial extent map of $\mathrm{\Delta }R/R$ change of C1 and C2 whisker piezo stimulation of an animal 2 and 6 weeks postsoft cranial window installation (Top 50 = within 50% of the peak response value of $\mathrm{\Delta }R/R$, top 40 = within 40% of the peak response value of $\mathrm{\Delta }R/R$, top 30 = within 30% of the peak response value of $\mathrm{\Delta }R/R$).

In order to verify the spatial extent of hemodynamic responses upon single-whisker stimulation, the spatial extent map of $\mathrm{\Delta }R/R$ change was generated. As described in Sec. 2, the pixel values above the 50% of the peak responses were counted and reflected in a spatial extent map. The spatial extents at 2 weeks were broader than those obtained at 6 weeks ($2\text{weeks}=1163.4\pm 309.49$, $6\text{weeks}=393.4\pm 33.42$, $p\text{value}=0.04$), suggesting that soft cranial window installation may affect the focality of hemodynamic responses of the barrel cortex [Figs. 4(e) and 4(f)]. In addition, we calculated the mean responses of the $1.06\mathrm{mm}\times 2.5\mathrm{mm}$ area, which have a peak hemodynamic response. The hemodynamic responses were significantly different between the two groups, as expected, and max $\mathrm{\Delta }R/R$ change of the 6 weeks group was significantly higher than that of the 2 weeks group (see Sec. 5.2; 2 weeks: $0.34\pm 0.03$, 6 weeks: $0.56\pm 0.01$, $p\text{value}>0.001$; $n=10$ in each group). In case of change of cerebral blood flow (CBF), the 2 weeks group tended to have smaller $\mathrm{\Delta }\mathrm{CBF}$ than the 6 weeks groups. However, there is no significant difference for maximal $\mathrm{\Delta }\mathrm{CBF}$ between groups (see Sec. 5.3; 2 weeks: $5.53\pm 1.24\%$, 6 weeks: $8.95\pm 3.57\%$, $p\text{value}=0.132$; $n=6$ in each group).

To confirm this, single C1 and C2 whiskers were activated sequentially for 4 s, and the spatial extents of $\mathrm{\Delta }R/R$ were generated. In 2-week postoperative mice, the spatial extents of the C1 and C2 columns were not adequately separated. By contrast, in animals at 6 weeks postsurgery, the separation of each whisker column was clear, suggesting better focalization of evoked-hemodynamic signals [Fig. 4(g)].

3.3.

Differential Dendritic Potential Changes during Whisker Stimulation after Soft Cranial Window Installation

In order to investigate whether the decreased evoked-hemodynamic signal in mice at 2 weeks postsurgery was associated with decreased neuronal activity, dendritic membrane potentials from a single whisker column were measured electrophysiologically. Three LFP recording electrodes were inserted directly through the soft cranial window into the center of the C2 column and into areas rostral and caudal to the first. The electrodes were $600\mu \mathrm{m}$ apart from each other. The C2 column was identified using in vivo optical recording of intrinsic signals [Fig. 5(a)].

Fig. 5

LFP at 2 and 6 weeks postsoft cranial window installation. (a) In vivo LFP recording set up with C2 whisker single deflection. Three electrodes recorded LFPs simultaneously in the center of the C2 column and in both the rostral and caudal directions. (b) LFP following C2 whisker single deflection (trials = 100). (c) Peak amplitude of LFP following C2 whisker single deflection (trials = 500, all $p\text{values}<0.001$). (d) Correlation of LFP amplitude in center of column and the peak value of CBV change.

At 2 weeks postsurgery, whisker-evoked LFP traces at the center of C2 were not significantly different from the ones recorded at the neighboring electrodes [Fig. 5(b)]. All three LFPs showed similar levels of cortical activity. On the other hand, at 6 weeks postsurgery, whisker-evoked LFP traces at the center of C2 were significantly different from the other locations, that is, LFPs from the center of C2 column showed significantly higher amplitude. When the LFPs obtained at 2 and 6 weeks postsurgery are compared, the amplitude of center LFPs at 6 weeks is significantly higher than those at 2 weeks. LFPs from the rostral and caudal areas showed no significant difference between the two postoperative groups [trials $n=500$, R: $2\text{weeks}=-0.18\pm 0.04$, $6\text{weeks}=-0.16\pm 0.03$, M: $2\text{weeks}=-0.21\pm 0.06$, $6\text{weeks}=-0.55\pm 0.06$, C: $2\text{weeks}=-0.19\pm 0.04$, $6\text{weeks}=-0.23\pm 0.06$, all $p\text{values}<0.001$; Fig. 5(c)]. The LFP amplitude from acute surgery was similar with that of the 2 weeks and smaller than that of 6 weeks group ($\text{acute}=-0.20\pm 0.04$, $2\text{weeks}=-0.20\pm 0.05$, $6\text{weeks}=-0.54\pm 0.05$, $p$ value: acute versus $2\text{weeks}=0.003$, acute versus 6 weeks, 2 weeks versus $6\text{weeks}=p\text{value}>0.001$; trials = 500 in each group; see Sec. 5.5). Overall, the tuning of cortical activity for the C2 column was much stronger at 6 weeks after soft cranial surgery.

This less shapely tuned whisker column phenomenon also appeared in the spatial extents map of hemodynamic signals at 2 weeks postsurgery. Therefore, we plotted the evoked-hemodynamic changes against LFPs at the center of C2 column at both 2 and 6 weeks following PDMS surgery and calculated the correlation between the peak value of $\mathrm{\Delta }R/R$ changes and LFPs amplitude by measuring the magnitude between the trough of the first negative peak and the peak of the first positive deflection. The $R$-value for the correlation of the factors was significantly higher in the 6 weeks postsurgery group [2 weeks: $n=8$, 6 weeks: $n=8$; Fig. 5(d)]. Two animal groups underwent similar anesthetic levels (see Sec. 5.4).

3.4.

Increased Interleukin $1\beta$ Levels at 2 Weeks Postsoft Cranial Window Installation

To determine the proinflammatory cytokine expression level associated with soft cranial window surgery, we measured IL-$1\beta$ and IL-6 of the cortex in the soft cranial window area. We also measured IL-10 as anti-inflammatory cytokine. An ELISA was used for these measurements. The expression of IL-$1\beta$ was higher at 2 weeks postsurgery than at either 6 weeks postsurgery or in normal controls (normal: $6.18\pm 0.89$, $2\text{weeks}:15.5\pm 2.53$, $6\text{weeks}:5.89\pm 1.87$, $p$ value: normal versus $2\text{weeks}=0.002$, 2 weeks versus $6\text{weeks}=p>0.001$). IL-$1\beta$ expression at 6 weeks was not significantly different from that of the normal group [$n=5$ for each group; Fig. 6(a)]. IL-6 expression showed no significant difference among the groups [normal: $13.80\pm 1.23$, 2 weeks: $12.61\pm 0.85$, 6 weeks: $14.70\pm 1.34$; Fig. 6(b)]. IL-10 expression in the cortex of the soft cranial window area was similar in the 2 weeks postsurgery group and normal group [normal: $3.26\pm 0.88$, 2 weeks: $4.74\pm 0.74$, 6 weeks: $7.52\pm 0.52$, $p$ value: normal versus $6\text{weeks}=0.004$; $n=5$ in each group; Fig. 6(c)].

Fig. 6

Expression levels of inflammatory cytokines at 2 or 6 weeks postsoft cranial window installation. (a) Expression of IL-$1\beta$ ($p$ value: normal-2 weeks = 0.006, 2 to 6 weeks = 0.001). (b) Expression of IL-6. The expression of IL-6 was not significantly different among the groups. Expression levels IL-10 at 2 or 6 weeks postsoft cranial window installation ($p\text{value}=0.004$). (d) Expression level of IL-$1\beta$ ($n=5$; $p$ value in Table 3).

In the case of expression of IL-$1\beta$, the animals after acute surgery also expressed high levels of IL-$1\beta$, such as animals at 2 weeks postsurgery. The expression of IL-$1\beta$ of the animals injected only with urethane was not significantly different from the normal control and 6 weeks group [Fig. 5(d); normal: $6.76\pm 0.89$, urethane: $5.16\pm 0.99$, acute: $14.24\pm 2.25$, 2 weeks: $15\pm 2.53$, MinoNac.2w: $6.62\pm 0.91$, 6 weeks: $5.86\pm 1.87$, $p$ value in Table 3; $n=5$ in each group].

Table 3

The p value of expression level of IL-1β between groups of normal, urethane, acute, 2 weeks, MinoNAC.2w, 6 weeks postsoft window installation.

3.5.

Elevated Hemodynamic Response Following minoNAC Administration after Soft Cranial Window Installation Surgery

Minocycline is a tetracycline antibiotic and known to inhibit the activation and proliferation of microglia.²⁰ $N$-acetylcysteine is an antioxidant and also used as an antiinflammatory drug.²¹ We mixed these two in order to synergically modulate microglia activation.^22,23 To test whether the elevation in active microglia of mouse with a PDMS soft cranial window is related with hemodynamic changes, we inhibited neuroinflammation by administering minoNAC to animals immediately following soft cranial window installation. Treatment was given every day for 3 days [Fig. 7(a)]. In animals that received minoNAC, whisker evoked-hemodynamic signals were not reduced at 2 weeks postsurgery when compared to the untreated mice 2 weeks [MinoNAC.2w: $0.80\pm 0.11$, 2 weeks: $0.54\pm 0.07$, $p\text{value}=0.005$; MinoNAC: $n=6$, 2 weeks: $n=10$; Figs. 7(b)–7(d)]. In addition, amplitude of LFP of received minoNAC animals was higher than the untreated 2 weeks group [MinoNAC.2w: $-0.34\pm 0.59$, 2 weeks: $0.20\pm 0.59$, $p\text{value}<0.001$; MinoNAC: $n=6$, 2 weeks: $n=10$; Figs. 7(e) and 7(f)]. They even maintained low levels of reactive astrocyte and deactivated microglia from IHC compared with the untreated 2 weeks group [Fig. 7(g)] and expressed similar levels of IL-$1\beta$ with normal animals [Fig. 6(d); $n=5$, $p$ value in Table 3]. These results suggest that neuroinflammation may have a direct effect on decreased hemodynamics and cortical activity at 2 weeks postsurgery.

Fig. 7

Hemodynamic and neural activity from animals with minocycline with NAC administration postsoft cranial window installation. (a) An experimental scheme of the mixture of minocycline with NAC administration for 3 days postsoft cranial window installation. (b) The spatial activation map of ORIS imaging for a single-whisker stimulation (4 s) in the animal with or without minocycline with NAC at 2 weeks postsoft cranial window installation. (c) A graph of $\mathrm{\Delta }R/R$ change at 2 weeks postsoft cranial window installation in mice with or without minocycline plus NAC (minoNAC: $n=6$, 2 weeks: $n=10$). (d) The peak value of $\mathrm{\Delta }R/R$ changes. (e) LFP following C2 whisker single deflection (trials =100). (f) The peak amplitude of LFP following C2 whisker single deflection (trials = 500, $p\text{value}>0.001$). (g) Immunohistochemistry of reactive astrocyte (green) and microglia (red) of an animal treated with minoNAC.

5.1.

Comparison of Blood Vessel Diameters of Resting State between 2 and 6 Weeks Postsoft Window Installation

In order to verify and compare the diameter of blood vessel in resting state, we measured the baseline diameter of an identical vessel at the 2 and 6 weeks images from our experiment (pixel resolution: 1 pixel: $17.7\mu \mathrm{m}$). We selected the middle cerebral artery (MCA) as the blood supply source of the barrel cortex for these comparisons. As shown in Fig. 3(a), three vessel segments were selected before and after the bifurcation based on the first bifurcation point of the MCA. The bold line, which has $\sim 100\mu \mathrm{m}$ width, was set perpendicularly to each selected vessel. We quantified five diameters with a parallel shift following the bold line using the “diameter” plugin function of ImageJ.⁴⁸ Thus, we measured 15 diameters from three blood vessels for each individual animal at two time points, i.e., at 2 and 6 weeks postsoft cranial window installation. For the comparison of the measured diameters, we first verified the normality of data through the Shapiro–Wilk test. Based on the normality results, all data were compared by paired $t$-test.

To sum, there is no significant difference between vessel diameters between 2 and 6 weeks postsurgery [Fig. 3(b)]. Thus, the resting state of the vessel diameter was not different between 2 and 6 weeks (Fig. 8) ($2\text{weeks}=83.5\pm 1.84\%$, $6\text{weeks}=84.39\pm 6.65\%$, $p\text{value}:0.730$; $n=5$ in each group).

Fig. 8

Blood vessel diameter comparison within a single animal at 2 and 6 weeks postsoft cranial window installation. (a) The images of vessels for the comparison from an animal at 2 weeks (left) and 6 weeks (right) postsoft cranial window implantation. (b) The averaged blood vessel diameters from an animal at 2 and 6 weeks postsoft cranial window implantation.

5.2.

Comparison of Hemodynamic Responses from the Large Field of View between 2 and 6 Weeks Postsoft-Window Installation

We calculated the mean responses of more RoIs, $1.06\mathrm{mm}\times 2.5\mathrm{mm}$ area, which have a peak hemodynamic response (Fig. 9).

Fig. 9

Hemodynamic responses in $1.06\mathrm{mm}\times 2.5\mathrm{mm}$ areas of animals at 2 and 6 weeks postsoft cranial window installation. (a) Spatiotemporal map of $\mathrm{\Delta }R/R$ change of animal 2 and 6 weeks postsoft window installation. (b) Graph of $\mathrm{\Delta }R/R$ change ($n=10$ in each group). (c) Peak value of $\mathrm{\Delta }R/R$ change ($n=10$ in each group).

5.3.

Comparison of Cerebral Blood Flow (CBF) between 2 and 6 Weeks Postsoft-Window Installation

For measurement of the local microcirculatory blood perfusion in the capillaries of animal’s cortex, a small fiber probe (0.25 mm, 780-nm wavelength laser) connected to a laser Doppler instrument (Periflux system, Perimed, Sweden) was placed at the C2 barrel column, which was confirmed by ORIS imaging. Piezo stimulation (10 Hz) delivered to C2 whiskers was given during 4 s at 5 s after onset recording of LDF simultaneously. Laser Doppler instruments were connected with an AC convertor (Plexon Instruments, Inc.) and signals were digitized (1000 Hz) and recorded. Acquired signals from laser Doppler instrument were analyzed by MATLAB. 1000 Hz signals were downsampled to 10 Hz and smoothed by a moving average filter with window span of 10. The baseline was determined by signals from 1 to 5 s, and all signals were normalized with the baseline.

The results were shown that the 2 weeks group tend to have smaller $\mathrm{\Delta }\mathrm{CBF}$ than the 6 weeks groups [Fig. 10(a)]. However, there is no significant difference for maximal $\mathrm{\Delta }\mathrm{CBF}$ between groups [Fig. 10(b); 2 weeks: $5.53\pm 1.24$, 6 weeks : $8.95\pm 3.57$, $p\text{value}=0.132$; $n=6$ in each group]. All $\mathrm{\Delta }\mathrm{CBF}$ changes of two groups were higher than 1%.

Fig. 10

CBF changes at 2 or 6 weeks postsoft window installation. (a) $\mathrm{\Delta }\mathrm{CBF}$ change of animals at 2 and 6 weeks postsoft window installation. (b) peak value of $\mathrm{\Delta }\mathrm{CBF}$ change of animals at 2 and 6 weeks postsoft window installation ($n=6$ in each group).

5.4.

Anesthetic Levels of Animals at 2 and 6 Weeks Postsoft Window Installation

The distribution of power spectral of brain oscillation may reflect the depth of anesthesia.^49–51 In order to check the level of anesthetic depth during our experiment, we applied fast Fourier transformation (FFT) for spontaneous neuronal activity with no stimulation for 20 s. FFT results showed that there is no difference between two groups in power spectrum [Figs. 11(a) and 11(b); $n=8$ in each group].

Fig. 11

The power spectrums of spontaneous neuronal activity. (a) Power spectrum of spontaneous neuronal activity of 2 weeks group ($n=8$). (b) Power spectrum of spontaneous neuronal activity of 6 weeks group ($n=8$).

5.5.

Acute Assessment of Hemodynamic Responses and Neuronal Activity Immediately Following Soft-Window Installation

We performed the acute surgery and conducted hemodynamic and neural activity measurements. We followed same experimental protocol as other condition. Although this was a group comparison, hemodynamic change during whisker stimulation was lower than those between 2 and 6 weeks group, respectively [Figs. 12(a), 12(b), and 12(d); peak value of $\mathrm{\Delta }R/R$ change : $\text{acute}=0.31\pm 0.02\%$, $2\text{weeks}=0.53\pm 0.14\%$, $6\text{weeks}=1.19\pm 0.14\%$, $p\text{value}<0.001$; acute: $n=7$, 2 and 6 weeks: $n=10$]. The LFP amplitude from acute surgery was similar with LFP amplitude from 2 weeks and smaller than that of the 6 weeks group [$\text{acute}=-0.20\pm 0.04$, $2\text{weeks}=-0.20\pm 0.05$, $6\text{weeks}=-0.54\pm 0.05$, $p$ value: acute versus $2\text{weeks}=0.003$, acute versus 6 weeks, 2 weeks versus $6\text{weeks}=p\text{value}>0.001$; trials = 500 in each group; Figs. 12(c) and 12(e)]. This demonstrated that surgical process can affect hemodynamic response and neuronal activity immediately.

Fig. 12

Hemodynamic change and LFP postsoft window installation. (a) Spatial activation map of ORIS imaging during single whisker stimulation. (b) Graph of $\mathrm{\Delta }R/R$ change ($n=7$). (c) LFP following C2 whisker single deflection (trials = 100). (d) Peak value of $\mathrm{\Delta }R/R$ change (acute: $n=7$, 2 weeks, 6 weeks: $n=10$, $p$ value: acute versus 2 weeks = 0.003, acute versus 6 weeks, 2 weeks versus $6\text{weeks}=p\text{value}>0.001$). (e) Peak amplitude of LFP following C2 whisker single deflection (trials = 500, all $p\text{values}<0.001$).