Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Sep 1.
Published in final edited form as: Neuroimage. 2018 May 29;178:414–422. doi: 10.1016/j.neuroimage.2018.05.063

Focused Ultrasound Induced Opening of the Blood-Brain Barrier Disrupts Inter-Hemispheric Resting State Functional Connectivity in the Rat Brain

Nick Todd 1,3, Yongzhi Zhang 1, Michael Arcaro 2, Lino Becerra 3, David Borsook 3,4, Margaret Livingstone 2, Nathan McDannold 1
PMCID: PMC6046263  NIHMSID: NIHMS972861  PMID: 29852281

Abstract

Focused ultrasound (FUS) is a technology capable of delivering therapeutic levels of energy through the intact skull to a tightly localized brain region. Combining the FUS pressure wave with intravenously injected microbubbles creates forces on blood vessel walls that open the blood-brain barrier (BBB). This noninvasive and localized opening of the BBB allows for targeted delivery of pharmacological agents into the brain for use in therapeutic development. It is possible to use FUS power levels such that the BBB is opened without damaging local tissues. However, open questions remain related to the effects that FUS-induced BBB opening has on brain function including local physiology and vascular hemodynamics. We evaluated the effects that FUS-induced BBB opening has on resting state functional magnetic resonance imaging (rs-fMRI) metrics. Data from rs-fMRI was acquired in rats that underwent sham FUS BBB vs. FUS BBB opening targeted to the right primary somatosensory cortex hindlimb region (S1HL). FUS BBB opening reduced the functional connectivity between the right S1HL and other sensorimotor regions, including statistically significant reduction of connectivity to the homologous region in the left hemisphere (left S1HL). The effect was observed in all three metrics analyzed: functional connectivity between anatomically defined regions, whole brain voxel-wise correlation maps based on anatomical seeds, and spatial patterns from independent component analysis. Connectivity metrics for other regions where the BBB was not perturbed were not affected. While it is not clear whether the effect is vascular or neuronal in origin, these results suggest that even safe levels of FUS BBB opening have an effect on the physiological processes that drive the signals measured by BOLD fMRI. As such these effects must be accounted for when carrying out studies using fMRI to evaluate the effects of pharmacological agents delivered via FUS-induced BBB opening.

Introduction

Focused ultrasound (FUS) can be used to disrupt the blood-brain barrier (BBB) in a way that is non-invasive, spatially localized, and safe. This is achieved through the interaction of burst mode FUS with intravenously injected microbubbles (Hynynen et al., 2001). FUS applied through the intact skull can deliver energy focused to a few millimeters in extent anywhere from the cortex to deep lying structures. Circulating microbubbles that enter the FUS pressure field undergo rapid expansion and contraction, creating forces strong enough to disrupt the tight junctions of the cells in vessel walls and thereby open the BBB. Circumventing the BBB in this way overcomes a major hurdle in drug delivery to the central nervous system. FUS-induced BBB opening has been used to deliver a wide range of therapeutic agents into the brain (Aryal et al., 2014; Leinenga et al., 2016; Meairs, 2015) and is currently entering clinical trials for treatment of brain tumors, Alzheimer’s disease, and Amyotrophic Lateral Sclerosis.

Animal studies have shown FUS BBB opening to be safe from the standpoint of being able to achieve permeablization of the BBB without causing damage to tissue cells that is detectable on histology (Downs et al., 2015b; Kobus et al., 2016; McDannold et al., 2012). However, recent research has shown that FUS BBB opening may have other effects on local physiology. Investigators are looking into whether FUS BBB opening initiates an immune response, modulates neuronal activity, or changes vascular hemodynamics (Chu et al., 2015; Downs et al., 2015a; Kovacs et al., 2017a). Any effects that FUS BBB opening may have on neuronal function are critically important to understand. From a safety standpoint, potential changes to neuronal function must be assessed for severity and longevity to determine any risk of harm to patients. From an applications standpoint, studies delivering a pharmacological agent designed to affect brain function must account for the possible additional neuronal modulation caused by the FUS BBB opening itself. Additionally, if FUS BBB opening does effect neuronal function, it may be possible to harness that effect for intentional manipulation of brain activity.

Functional magnetic resonance imaging (fMRI) is a noninvasive technique that can map neuronal activity in the brain at a neurosystems level. It provides information over the entire brain with millimeter-scale spatial resolution on how the brain responds to an external stimuli, which brain regions are active during a particular task, and how different brain regions are functionally connected even in the absence of a task. This allows multiple and spatially distant brain regions to be evaluated simultaneously and can reveal if functional activity in one region is propagated to another region. Using rs-fMRI, it has been shown that functional connectivity patterns between spatially distinct regions exist in the rat brain which mirror those found in humans. This includes a default mode network (Lu et al., 2012), bilateral functional connectivity in the motor, somatosenory, visual, and auditory cortices, and several sub-cortical regions (Becerra et al., 2011; D’Souza et al., 2014; Hutchison et al., 2010).

In this study, we use rs-fMRI to investigate the effects that FUS BBB opening has on functional connectivity in the rat brain. Rats underwent either sham or actual FUS BBB opening targeted unilaterally to the right primary somatosensory cortex hindlimb region (S1HL). rs-fMRI data was collected and analyzed using functional connectivity between anatomically defined regions, whole-brain connectivity maps based on seeds in anatomically defined regions, and spatial connectivity patterns based on independent component analysis (ICA) that has no a priori knowledge of anatomical connectivity. These methods are sensitive to changes in both local and long-range functional connectivity and provide information about functional networks throughout the brain. We hypothesized that FUS BBB opening in the right S1HL region would affect local functional connectivity, functional connectivity to the contralateral hemisphere, but not the functional connectivity within other anatomically distinct brain networks.

Methods

Animal Preparation

All experiments were done in accordance with procedures approved by the Brigham and Women’s Hospital Institutional Animal Care and Use Committee. The animals were housed, fed, and watered according to the Office of Laboratory Animal Welfare and the Association for Assessment and Accreditation of Laboratory Care regulations. Male Sprague Dawley rats were used for all experiments (257 – 384 g). At the start of each experiment, the rats were anesthetized with one dose of ketamine and xylazine (K = 80 mg/kg, X = 10 mg/kg, intraperitoneal injection). The head of the rat was shaved and depilatory cream was applied to remove all fur for optimal ultrasound coupling. A tail vein catheter was placed for administration of microbubbles and MRI contrast. All imaging was performed in a 7 Tesla Bruker BioSpec small animal MRI scanner (Bruker Corp., Billerica, MA, USA). Experiments were carried out for N = 10 rats in both the BBB closed and BBB open cohorts.

FUS Blood-Brain Barrier Opening

FUS induced opening of the BBB was targeted to the right primary somatosensory cortex hindlimb region (S1HL). The center of the FUS focus was placed such that minimal subcortical regions were affected and the energy was confined to the right hemisphere only. The FUS power level was set based on previous experience and literature values (Chu et al., 2015; McDannold et al., 2008) so as to induce BBB opening but not cause observable damage to the local tissue.

The rat’s head was fixed in a custom made holder that slots into the bed of an in-house made MRI-compatible FUS system. The FUS system consists of a 690 kHz single element focused ultrasound transducer (4.0 cm diameter, 3.5 cm radius of curvature), a single element passive cavitation detector, a three-axis manual positioning system, and a transmit/receive MR RF coil. The rat was coupled to the ultrasound transducer by placing it upside down with its head in a bath of degassed and deionized water. The focused ultrasound sonications were applied using a function generator (33220A, Agilent), amplifier (240L, E&I) and custom Matlab user interface.

Targeting of FUS energy to the right S1HL was done in two steps. First, a ten second FUS sonication was applied to a silicone gel phantom, creating a hotspot detectable by MRI temperature-sensitive imaging which provided the location of the FUS focal point in the MRI coordinate system. Then, the rat was placed on the system and structural imaging was performed to allow visual identification of the right S1HL in MRI coordinates. The FUS transducer was moved such that the center of the focus would lie at the location of the right S1HL.

To achieve BBB opening, repeated bursts of FUS sonications were applied immediately after injection of microbubbles (200 µL/kg bolus injection of Optison, GE Healthcare). FUS sonication parameters were 10 ms bursts, 1 Hz repetition frequency, and 120 repeats. Five sets of sonications were applied over a 2 × 2 mm square (4 corners and one in the middle) to ensure coverage of the right S1HL region. Applied FUS power levels were fixed for each rat at either 0.32 or 0.34 MPa (peak negative pressure calibrated in water). These pressures were tested previously in mice and found no immediate significant tissue damage (McDannold et al., 2017). The passive cavitation detector was used to monitor the behavior of the microbubbles interacting with the ultrasound field, ensuring that stable cavitation was taking place but inertial cavitation was not (Arvanitis et al., 2012). A warming blanket was used throughout the BBB opening process to maintain body temperature and the rat’s breathing rate was continuously monitored. T1-weighted images obtained after administering an MRI contrast agent show the extent and location of BBB opening in each individual rat (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Extent of BBB opening for all rats. A) T1-weighted contrast images showing the BBB opening for each of the ten individual rats in the BBB Open cohort. The images have been converted to percent difference between pre- and post-contrast injection. B) Sum of binary T1-weighted contrast masks over all ten rats overlaid on a T2-weighted image. Color scale indicates number of rats with a contrast change of greater than 4%.

Functional MRI

After BBB opening, the rat was removed from the FUS system and set up in the conventional Bruker animal holder. The rat was laid prone and its nose was inserted into a nose cone for head immobilization and delivery of isoflurane and oxygen. A line was inserted into the rat’s abdomen for subcutaneous injection of Dexdomitor (Dexmedetomidine; Orion, Espoo, Finland) (Adamczak et al., 2010) and a 2 cm diameter surface coil was placed over the rat’s head. A warming blanket was used to maintain body temperature, and breathing rate was monitored throughout. Once the rat was set up, a bolus injection of Dexdomitor was given (0.025 mg/kg) and delivery of 0.25% isoflurane in 60% oxygen enriched air was started. This low level of isoflurane was kept constant throughout the remainder of the experiment, and periodic infusions of Dexdomitor were given as needed to keep the rat in a stable physiological state (Brynildsen et al., 2017; Paasonen et al., 2018). The resting state imaging run took place approximately 40 minutes after the end of BBB opening. This was observed to be enough time to allow physiological levels to stabilize but is still well within the ~4 hour time window that the BBB remains open after FUS sonication (Park et al., 2012).

All fMRI data were acquired with a 2D single-shot gradient echo-planar imaging (EPI) sequence. Sequence parameters were: 3.2 × 3.2 cm field of view; 64×64×18 imaging matrix; 0.5 × 0.5 × 1.0 mm resolution; 18 slices with 0.2 mm slice gap; TR = 1500 ms; TE = 18 ms; Flip Angle = 65°; four dummy scans prior to data acquisition; 300 image volumes acquired in 7 minutes and 30 seconds. Prior to the fMRI runs, main field homogeneity was optimized using the Bruker MAPSHIM protocol and an anatomical image was acquired with a T2-weighted RARE sequence (0.3 × 0.3 × 0.5 mm resolution; 60 axial slices with no slice gap; TR = 6500 ms; TE = 50 ms). For rats that underwent FUS BBB opening, contrast imaging was done with a T1-weighted RARE sequence (0.3 × 0.3 × 0.6 mm resolution; 18 coronal slices with 0.4 mm slice gap; TR = 609 ms; TE = 18 ms) before and after injection of Gd-DTPA (Magnevist, 0.25 mL/kg). The contrast images were converted to percent signal change, coregistered to the anatomical image and normalized into the template space (as described below). These images were acquired at the end of the experiment so that no contrast agent was given prior to the rs-fMRI runs.

Anatomical Atlas and Seed Regions

A set of anatomical brain regions were created in the normalized MRI space based on the rat brain atlas of Paxinos and Watson (Paxinos G., 2005). 34 total regions were chosen, which covered almost all of the cortex and some subcortical areas. There were 16 bilateral regions consisting of left and right: primary visual cortex (V1), secondary visual cortex lateral area (V2L), secondary visual cortex medial area (V2M), primary auditory cortex (Au1), fields CA1, CA2 and CA3 of the hippocampus (Hipp), the lateral and posterior areas of the hypothalamus (HyTh), the ventral posterolateral and ventral posteromedial thalamic nuclei (Thal), primary somatosensory cortex trunk region (S1Tr), primary somatosensory cortex barrel field (S1BF), primary somatosensory cortex hind limb region (S1HL), primary somatosensory cortex front limb region (S1FL), secondary somatosensory cortex (S2), primary somatosensory cortex jaw region (S1J), primary motor cortex (M1), the dysgranular and granular insular cortex (Ins), and the caudate putamen (CPu). There were two regions on the midline of the brain: retrosplenial cortex granular a and b (RSC), and the cingulate cortex areas 1 and 2 (Cing). All regions are shown in Figure 2, overlaid on a T2-weighted anatomical image.

Figure 2.

Figure 2

Open in a new tab

The 34 anatomical regions used in this study overlaid on a T2w anatomical image. 12 coronal slices are shown with their respective distances from the Bregma landmark.

Within each anatomically defined area, a smaller seed region of interest (ROI) was defined. The center of the seed ROI was manually located in the center of the anatomical area. Each seed ROI was a sphere of diameter 1.25 mm and comprised of 31 voxels in the normalized EPI image space.

Data Processing

All fMRI images were pre-processed using SPM12 (SPM12, 2014) and custom Matlab scripts (R2014b, Mathworks, Natick, MA, USA). The T2-weighted anatomical image was segmented in SPM12 using the template of Valdes et al. (Valdés-Hernández et al., 2011). Using SPM12, the EPI images were slice time corrected, realigned, coregistered to the anatomical image, normalized to the template space, and spatially smoothed with a Gaussian filter of 0.8 × 0.8 × 0.8 mm FWHM. Using custom Matlab scripts, the data were further temporally filtered (bandpass 6th-order Butterworth filter from 0.01 to 0.1 Hz), the motion correction time courses from realignment were regressed out, and a mean signal averaged over only voxels in white-matter was also regressed out. Finally, the mean signal over time from each individual voxel was calculated and subtracted out on a voxel-by-voxel basis.

Connectivity matrices were computed to summarize the resting state functional connectivity between every pair of the 34 seed ROIs. The BOLD signal time course was spatially averaged over each seed ROI. For each of the 10 rats in the BBB closed and BBB open cohorts, the time courses from each seed ROI were correlated with the time course of every other seed ROI to obtain a Pearson’s R correlation coefficient. These correlation values were summarized into a 34 × 34 correlation matrix for each rat.

Whole brain seed-based correlation maps were computed for each seed ROI for each rat in the BBB closed and BBB open cohorts. The spatially averaged BOLD signal time course from a seed ROI was correlated with the individual-voxel time course from every other voxel in the brain. The resulting correlation maps were averaged over the ten rats and thresholded at a correlation value of 0.2.

Finally, independent component analysis (ICA) was performed as a way of analyzing the data without any prior knowledge of anatomical regions. Group ICA was performed using the GIFT software package (Calhoun et al., 2001). First, data reduction was done in two steps doing principle component analysis (PCA) on the individual data sets first and then a second PCA reduction step on the concatenated group data. ICA was performed on the reduced data sets using the Infomax algorithm. Different runs were performed with the number of components fixed to 20, 40, 60, and 80. The main results shown are from using 40 components, and a comparison over the four different numbers of components used is also provided. The spatial ICA components were assigned to a particular anatomical region according to their average value in each of the 34 seed ROIs.

Statistical Analysis

Analysis was performed to determine if statistically significant differences existed in the functional connectivity measurements between the BBB Closed and BBB Open cohorts. The Fisher Z-Transformation was applied to the Pearson R correlation coefficients in order to obtain normally distributed data. An unpaired t-test was used to determine significant differences between the correlation values for the BBB Closed and BBB Open cases (18 degrees of freedom). P-values are presented after undergoing correction for multiple comparisons using the false discovery rate method (p<0.05) (Benjamini and Hochberg, 1995; Benjamini and Yekutieli, 2005). The mean of the correlation values over all rats was calculated for display in figures.

Results

The functional connectivity matrices for the cases of BBB closed and BBB open are shown in Figure 3. These matrices summarize the functional connectivity strength between every pair of the 34 anatomical regions. The anatomical regions are organized into classes of sensorimotor, insula, auditory/visual, subcortical, and midline. Bilateral regions are positioned with the left and right region next to each other. In general, the sensorimotor, auditory/visual, and subcortical regions have the strongest connections to other regions within their same class, and homologous regions have strong connections between their left and right regions. The overall pattern of the connectivity matrices for the BBB closed and BBB open cases are similar, with only 6 cases of significantly different connectivity values surviving correction for multiple comparisons, including the connectivity between the right and left S1HL regions (p < 0.05, corrected for multiple comparisons).

Figure 3.

Figure 3

Open in a new tab

Connectivity Matrices. Correlation values are shown between all 34 pairs of defined anatomical regions (mean over all rats). Black dashed line highlights S1HL R correlations. On the BBB Open correlation matrix, significant differences between functional connectivity values for BBB Closed vs. BBB open are designated with * (p < 0.05, corrected for multiple comparisons).

Figure 4 highlights the major differences between the two connectivity matrices seen in the functional connectivity from the right S1HL region to other sensorimotor regions. The plots show correlation values between the right and left S1HL seed ROIs and all other seed ROIs (mean and standard error over all rats), directly comparing the BBB closed vs BBB open cases. When the BBB is open, the right S1HL region has reduced connectivity to several other sensorimotor and auditory/visual regions. This is not the case for the left S1HL region, which was not targeted for BBB opening. For the left S1HL region the only significant reduction is with the right S1HL ROI.

Figure 4.

Figure 4

Open in a new tab

Plots of correlations values between S1HL R or S1HL L and all other regions for cases of BBB closed and BBB open. Significant differences between correlation values for the BBB closed and BBB open cases are indicated by * (p < 0.05, corrected for multiple comparisons).

A second comparison of functional connectivity values for the right and left S1HL regions between BBB closed and BBB open cases is shown graphically in Figure 5. Only the connections to other sensorimotor regions are shown, with the strength of connection represented by the thickness of the line. The graphical connectivity highlights that the left S1HL connectivity is largely unchanged when the BBB is open, whereas the connectivity from the right S1HL to both ipsilateral and contralateral sensorimotor ROIs is markedly reduced.

Figure 5.

Figure 5

Open in a new tab

Connectivity graph for left and right S1HL regions to all other sensorimotor regions. The strength of correlation between two regions is represented by the thickness of the line connecting them. A general decrease in connectivity is seen for the S1HL R region when the BBB has been opened.

Figure 6 provides further evidence that FUS BBB opening is the cause of the reduction in functional connectivity between the left and right S1HL regions. The scatter plots shows the correlation values between the left and right S1HL ROIs. For the BBB closed case, they are plotted as mean and standard error over all rats; for the BBB open case, they are plotted individually as a function of the percent difference in the T1w Gadolinium images within the right S1HL ROI. The correlation coefficients are negatively correlated with the Gadolinium differences (R = −0.44), implying that stronger BBB opening leads to less functional connectivity.

Figure 6.

Figure 6

Open in a new tab

Functional connectivity as a function of the extent of BBB opening. The scatter plot shows the correlation values between the right and left S1HL regions as a function of Gadolinium change seen in the right S1HL region. For the BBB closed cases, the Gadolinium change was not measured and is assumed to be zero, and the correlation coefficients are plotted as mean and standard deviation over all ten rats. The correlation coefficients for the BBB open cases appear to be lower when the BBB is opened to a greater extent (correlation value is R = −0.44).

Seven different whole brain seed-based correlation maps are shown in Figure 7, comparing cases of BBB closed vs BBB open. The seed time courses were taken from the ROIs in the right M1, right S1HL, right S1BF, right Au1, right V1, right CPu, and cingulate regions. When the BBB is closed, the correlation maps for the first six ROIs show clear bilateral connectivity patterns. The last ROI, the cingulate, is known to be a part of the default mode network and shows a connectivity pattern similar to what others have shown for the rat default mode network (Lu et al., 2012). When the BBB is open, all of the correlation maps appear qualitatively similar to the BBB closed maps, with the exception of the right S1HL seeded map. For the right S1HL map, there is a clear reduction in the bilaterality of the connectivity pattern.

Figure 7.

Figure 7

Open in a new tab

Seed based correlation maps. Whole brain correlation maps are shown using seeds from seven different anatomical locations, comparing BBB closed and BBB open cases. Mean correlation over all rats is shown. Strong bilateral correlations can be seen in all cases except for S1HL R when the BBB is open. In this case, the correlation to the left hemisphere is much reduced compared to the BBB closed case.

Figure 8 shows the seven different spatial ICA components that most closely correspond to the seed-based correlation maps above. As with the seed-based correlation maps, these ICA components show clear bilateral patterns for each of the first six regions when the BBB is closed. For the cingulate, there is a clear pattern covering the cingulate region, but the spatial pattern does not extend to the auditory/visual regions that are a part of the default mode network and were seen in the seed-based correlation maps. When the BBB is open, the spatial patterns of connectivity are very similar for the last five regions shown but there are clear differences for the M1 and S1HL regions. The connectivity pattern for the M1 region is reduced in both extent and strength. For the S1HL region, the bilaterality of the pattern is substantially reduced. For the BBB open case, there was a separate ICA component identified for the left S1HL region that was distinct from the right S1HL component shown here. This was not the case for the BBB closed data.

Figure 8.

Figure 8

Open in a new tab

ICA based connectivity maps. Seven ICA components are shown, organized according to which anatomical location they most closely match. Both the M1 and S1HL components show differences between the BBB closed and BBB open cases. When the BBB is open, the component most closely corresponding to the M1 region is much weaker, and the component most closely corresponding to S1HL R is almost entirely unilateral.

A comparison of ICA results using 20, 40, 60 or 80 components is shown in Figure 9. The spatial ICA components were assigned to the most closely matched anatomical region. For ICA 20 and ICA 40, a single ICA component was picked out to cover both the left and right S1HL regions when the BBB was closed. When the BBB was open, ICA 20 and ICA 40 picked out two distinct components for the left and right S1HL regions. Using a larger number of components in the ICA 60 and ICA 80 analysis, separate components are picked out for the left and right S1HL regions for both the BBB closed and BBB open cases. However, the bilaterality of the components appears to be stronger for the BBB closed cases.

Figure 9.

Figure 9

Open in a new tab

Spatial components assigned to the right and left S1HL region when ICA was run using 20, 40, 60, or 80 as the number of fixed components.

One further experiment was performed to investigate the time duration of the reduced connectivity between the left and right S1HL regions after BBB opening. Figure 10 shows functional connectivity results from one rat acquired on the day of BBB opening, and 1 day, 3 days, and 7 days post-sonication. Figure 10A plots the correlation coefficient between the left and right S1HL region as a function of days from sonication. The full connectivity matrices are shown in Figure 10B, and seed-based correlation maps using the right S1HL as the seed are shown in Figure 10C. Together the results show reduced connectivity on the day of BBB opening as expected, followed by recovery towards the pre-sonication state that may be as fast as 24 hours.

Figure 10.

Figure 10

Open in a new tab

Duration of reduced connectivity following BBB opening (N=1). A) Correlation coefficients between the left and right S1HL regions on the day of BBB opening, and 1 day, 3 days, and 7 days post-sonication. Blue dashed lines indicate the mean +/− one standard deviation of correlation values seen for the N=10 BBB closed cases. B) Full connectivity matrices for the four days. C) Seed-based connectivity maps for the four days with the seed in the right S1HL region.

Discussion

We report that FUS-induced disruption of the BBB targeted to the right S1HL region of the somatosensory cortex in Sprague Dawley rats reduces the functional connectivity between that region and several other sensorimotor regions. Some of the regions experiencing reduced connectivity are in the ipsilateral hemisphere and physically adjacent to the right S1HL. However, the region experiencing the most pronounced reduction in connectivity is the physically distant left S1HL in the contralateral hemisphere. The connectivity reduction was seen by comparing the resting state BOLD signal time courses from seeds in right and left S1HL regions. Importantly, this disassociation of the two regions was also picked up by the ICA analysis, which does not use any anatomical information.

FUS BBB Opening and Local Effects on S1HL

Several pieces of evidence indicate that the FUS BBB opening did not cause any damage to tissue at the targeted site. We did not observe any abnormalities in T2-weighted MRI. Histology with hematoxylin and eosin staining was done on two of the rats and no indication of damage to neuronal cells was seen in either hemisphere of the cortex (Figure 11). In the scatter plot of Figure 6, these two rats were at points (5.9, 0.13) and (6.2, 0.16). These values correspond to the lower end of T1w contrast change, but have two of the lowest correlation values. The FUS peak negative pressure at the focus was approximately 0.3 MPa (0.4 mechanical index), which is in line with what others have used to open the BBB without seeing damage (Chu et al., 2015; McDannold et al., 2017). The data presented in Figure 10 suggests that the effect is resolved over a matter of days, which is not indicative of damage to tissue. In evaluating the BOLD signal time courses in the right S1HL ROI, the frequency power spectrum values appear somewhat reduced in the range of 0.01 to 0.03 Hz for the BBB open data compared to the BBB closed data. Additionally, comparison of the temporal standard deviation of the BOLD signal time course shows somewhat lower values in the right S1HL for the BBB open case. However, the differences rise to the level of statistical significance for only one point in the frequency power spectrum, and no significant differences in the temporal standard deviation were seen (Figure 12). Finally, the ICA analysis for the BBB open data still picks out as one of its 40 spatial components a region that corresponds to the right S1HL, indicating that there is still considerable spatial/temporal structure in those time courses. If the vasculature was damaged and the BOLD signal in the right S1HL was dominated by thermal noise, the ICA would not pick it out as a component. Nevertheless, further examination with additional histological techniques (Kovacs et al., 2017b; McMahon and Hynynen, 2017) at later times are needed to rule out tissue effects we may have missed.

Figure 11.

Figure 11

Open in a new tab

Histology after BBB opening. The two row shows the contrast difference image and hematoxylin and eosin staining from two different rats. Regions from the right and left somatosensory cortex are shown at 1.25×, 10×, and 40× magnification. No signs of damage to neuronal cells were seen in either rat.

Figure 12.

Figure 12

Open in a new tab

Plots of average power spectrum of the time courses for the BOLD signal in various seed regions, comparing cases of BBB closed vs BBB open. Inset bar plots show the standard deviation of the signal time course from the same region (arbitrary units, mean and standard deviation over all rats). Significant differences between power spectrum values for the BBB closed and BBB open cases at a particular discrete frequency are indicated by * (p<0.05, corrected for multiple comparisons). No significant differences were seen in the standard deviation comparisons.

Possible Factors Leading to Reduced Connectivity

This study is not able to determine the underlying reason why FUS-BBB opening leads to a temporary reduction in functional connectivity. The BOLD signal that rs-fMRI measures is driven by a complex combination of neuronal activity, neurovascular coupling signaling mechanisms, and vascular hemodynamics. Modulation of neuronal activity is possible. However, Chu et al. recorded somatosensory evoked potential measurements after performing FUS-BBB opening at similar power levels and did not see any change (Chu et al., 2015). It has been shown that FUS-BBB opening can induce vasoconstriction (Cho et al., 2011). In that study, the measured vessel diameter decreased after FUS-BBB opening, but then returned to its baseline full width at maximum value in less than a few hundred seconds, which is much faster than the approximately 40 minutes between FUS-BBB opening and rs-fMRI measurements done in this study. It is also known that FUS-BBB opening can induce an inflammatory response in the local brain tissue (Kovacs et al., 2017b), particularly when large microbubble doses are used (McMahon and Hynynen, 2017). In the Kovacs et al. study, most markers of inflammation persisted for at least 6 hours. It may be that the agitation from the FUS sonication and accompanying inflammatory response perturbs the neurovascular signaling mechanisms or hemodynamic response out of their ordinary state. This is speculation at this point and further investigation into the underlying cause is warranted.

BBB Disruption and Brain Networks

Focal changes in brain function can lead to non-local alteration of resting state networks and functional connectivity. For example, Parkinson’s disease is characterized by a loss of dopaminergic neurons in the substantia nigra which has been shown to lead to changes in cortico-striatal functional connectivity (Baudrexel et al., 2011) and also more distant alterations to the attentional network, the motor network and the default mode network (Sala et al., 2017; Wu et al., 2009). Similarly, resting state fMRI studies of localized stroke have shown that inter-hemispheric connectivity is reduced between the lesion site and the corresponding region in the contralateral hemisphere (Golestani et al., 2013; van Meer et al., 2010). The FUS BBB disruption performed for this study does not damage neurons as in the case of Parkinson’s disease or stroke. But it very possibly induces other changes that would affect neuronal function or the BOLD signal.

The disruption of functional connectivity or the reorganization of functional networks is most often associated with the negative outcome of a transformation in the brain from a normal state to a diseased state. In addition to Parkinson’s disease and stroke, changes in functional networks have also been seen in many other neurological disorders such as Alzheimer’s disease, schizophrenia, and depression (Greicius, 2008; Rombouts et al., 2005). Changes in functional connectivity due to FUS BBB opening therefore warrant caution and further investigation, particularly into the duration of the effect. However, alteration of functional networks can also be used as a therapeutic tool where deliberate interventions are designed to drive a diseased brain state back towards the normalized state. For example, resting state fMRI is one way to measure the efficacy of transcranial magnetic stimulation treatments of depression (Fox et al., 2012a, 2012b; Keeser et al., 2011). Due to its noninvasive and localized nature, FUS may be able to play a similar role.

Caveats

This study was limited in that it only acquired resting state fMRI data at one time point after FUS BBB opening. It is therefore not clear how long the effects of reduced functional connectivity last. After FUS-induced opening, the BBB re-closes over several hours (Marty et al., 2012; Park et al., 2012). It is possible that the functional connectivity effects follow this duration, but they may also persist for hours or days after the BBB has closed, depending on what underlying mechanism is responsible for the effect. In this study, half of the BBB closed sets (N=4) were acquired after a previous BBB open experiment had been performed on the same rat at least one week prior. The numbers are too small for reliable statistics, but no clear differences were seen between the BBB closed data sets that had previous opening versus those that did not. This puts an upper limit of one week on the duration of the functional connectivity reduction effects.

Conclusion

Focused ultrasound mediated opening of the blood-brain barrier is an exciting technology that enables non-invasive targeted drug delivery to the brain. Much work has been done to establish a safe window of parameters that produce BBB opening but do not cause tissue damage. However, it is clear that FUS BBB opening also produces several secondary effects that are not well understood including the initiation of an immune response, changes to vascular hemodynamics, and possibly suppression of neuronal activity. In this study we used resting state fMRI as a way of investigating these secondary effects. Our results indicate that a reduction in functional connectivity from the targeted area to other local and interhemispheric regions is correlated with the extent of BBB opening. Further investigation is needed to fully understand these secondary effects. This will help ensure patient safety as this technology moves into the clinic and possibly enable novel applications of FUS-mediated BBB opening.

Acknowledgments

The research was supported by NIH grants R25CA089017-13, K01EB023983 and P01CA17464501.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The authors declare no competing financial interests.

References

  1. Adamczak JM, Farr TD, Seehafer JU, Kalthoff D, Hoehn M. High field BOLD response to forepaw stimulation in the mouse. Neuroimage. 2010;51:704–712. doi: 10.1016/j.neuroimage.2010.02.083. [DOI] [PubMed] [Google Scholar]
  2. Arvanitis CD, Livingstone MS, Vykhodtseva N, McDannold N. Controlled ultrasound-induced blood-brain barrier disruption using passive acoustic emissions monitoring. PLoS One. 2012;7:e45783. doi: 10.1371/journal.pone.0045783. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aryal M, Arvanitis CD, Alexander PM, McDannold N. Ultrasound-mediated blood-brain barrier disruption for targeted drug delivery in the central nervous system. Adv. Drug Deliv. Rev. 2014;72:94–109. doi: 10.1016/j.addr.2014.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baudrexel S, Witte T, Seifried C, von Wegner F, Beissner F, Klein JC, Steinmetz H, Deichmann R, Roeper J, Hilker R. Resting state fMRI reveals increased subthalamic nucleus-motor cortex connectivity in Parkinson’s disease. Neuroimage. 2011;55:1728–1738. doi: 10.1016/j.neuroimage.2011.01.017. [DOI] [PubMed] [Google Scholar]
  5. Becerra L, Pendse G, Chang P-C, Bishop J, Borsook D. Robust Reproducible Resting State Networks in the Awake Rodent Brain. PLoS One. 2011;6:e25701. doi: 10.1371/journal.pone.0025701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brynildsen JK, Hsu LM, Ross TJ, Stein EA, Yang Y, Lu H. Physiological characterization of a robust survival rodent fMRI method. Magn. Reson. Imaging. 2017;35:54–60. doi: 10.1016/j.mri.2016.08.010. [DOI] [PubMed] [Google Scholar]
  7. Calhoun VD, Adali T, Pearlson GD, Pekar JJ. A method for making group inferences from functional MRI data using independent component analysis. Hum. Brain Mapp. 2001;14:140–151. doi: 10.1002/hbm.1048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cho EE, Drazic J, Ganguly M, Stefanovic B, Hynynen K. Two-photon fluorescence microscopy study of cerebrovascular dynamics in ultrasound-induced blood-brain barrier opening. J. Cereb. Blood Flow Metab. 2011;31:1852–1862. doi: 10.1038/jcbfm.2011.59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chu P-C, Liu H-L, Lai H-Y, Lin C-Y, Tsai H-C, Pei Y-C. Neuromodulation accompanying focused ultrasound-induced blood-brain barrier opening. Sci. Rep. 2015;5:15477. doi: 10.1038/srep15477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. D’Souza DV, Jonckers E, Bruns A, Künnecke B, Von Kienlin M, Van Der Linden A, Mueggler T, Verhoye M. Preserved modular network organization in the sedated rat brain. PLoS One. 2014;9 doi: 10.1371/journal.pone.0106156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Downs ME, Buch A, Karakatsani ME, Konofagou EE, Ferrera VP. Blood-Brain Barrier Opening in Behaving Non-Human Primates via Focused Ultrasound with Systemically Administered Microbubbles. Sci. Rep. 2015a;5:15076. doi: 10.1038/srep15076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Downs ME, Buch A, Sierra C, Karakatsani ME, Chen S, Konofagou EE, Ferrera VP. Long-term safety of repeated blood-brain barrier opening via focused ultrasound with microbubbles in non-human primates performing a cognitive task. PLoS One. 2015b;10:1–26. doi: 10.1371/journal.pone.0125911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fox MD, Buckner RL, White MP, Greicius MD, Pascual-Leone A. Efficacy of transcranial magnetic stimulation targets for depression is related to intrinsic functional connectivity with the subgenual cingulate. Biol. Psychiatry. 2012a;72:595–603. doi: 10.1016/j.biopsych.2012.04.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fox MD, Halko MA, Eldaief MC, Pascual-Leone A. Measuring and manipulating brain connectivity with resting state functional connectivity magnetic resonance imaging (fcMRI) and transcranial magnetic stimulation (TMS) Neuroimage. 2012b;62:2232–2243. doi: 10.1016/j.neuroimage.2012.03.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Golestani A-M, Tymchuk S, Demchuk A, Goodyear BG. Longitudinal Evaluation of Resting-State fMRI After Acute Stroke With Hemiparesis. Neurorehabil. Neural Repair. 2013;27:153–163. doi: 10.1177/1545968312457827. [DOI] [PubMed] [Google Scholar]
  16. Greicius M. Resting-state functional connectivity in neuropsychiatric disorders. Curr. Opin. Neurol. 2008;24:424–430. doi: 10.1097/WCO.0b013e328306f2c5. [DOI] [PubMed] [Google Scholar]
  17. Hutchison RM, Mirsattari SM, Jones CK, Gati JS, Leung LS. Functional Networks in the Anesthetized Rat Brain Revealed by Independent Component Analysis of Resting-State fMRI. J. Neurophysiol. 2010;103:3398–3406. doi: 10.1152/jn.00141.2010. [DOI] [PubMed] [Google Scholar]
  18. Hynynen K, McDannold N, Vykhodtseva N, Jolesz FA. Noninvasive MR imaging-guided focal opening of the blood-brain barrier in rabbits. Radiology. 2001;220:640–6. doi: 10.1148/radiol.2202001804. [DOI] [PubMed] [Google Scholar]
  19. Keeser D, Meindl T, Bor J, Palm U, Pogarell O, Mulert C, Brunelin J, Moller H-J, Reiser M, Padberg F. Prefrontal Transcranial Direct Current Stimulation Changes Connectivity of Resting-State Networks during fMRI. J. Neurosci. 2011;31:15284–15293. doi: 10.1523/JNEUROSCI.0542-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kobus T, Vykhodtseva N, Pilatou M, Zhang Y, McDannold N. Safety Validation of Repeated Blood-Brain Barrier Disruption Using Focused Ultrasound. Ultrasound Med. Biol. 2016;42:481–492. doi: 10.1016/j.ultrasmedbio.2015.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kovacs ZI, Kim S, Jikaria N, Qureshi F, Milo B, Lewis BK, Bresler M, Burks SR, Frank JA. Disrupting the blood–brain barrier by focused ultrasound induces sterile inflammation. Proc. Natl. Acad. Sci. 2017a;114:E75–E84. doi: 10.1073/pnas.1614777114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kovacs ZI, Kim S, Jikaria N, Qureshi F, Milo B, Lewis BK, Bresler M, Burks SR, Frank JA. Disrupting the blood–brain barrier by focused ultrasound induces sterile inflammation. Proc. Natl. Acad. Sci. 2017b;114:E75–E84. doi: 10.1073/pnas.1614777114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Leinenga G, Langton C, Nisbet R, Götz J. Ultrasound treatment of neurological diseases — current and emerging applications. Nat. Rev. Neurol. 2016;12:161–174. doi: 10.1038/nrneurol.2016.13. [DOI] [PubMed] [Google Scholar]
  24. Lu H, Zou Q, Gu H, Raichle ME, Stein EA, Yang Y. Rat brains also have a default mode network. Proc. Natl. Acad. Sci. 2012;109:3979–3984. doi: 10.1073/pnas.1200506109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Marty B, Larrat B, Van Landeghem M, Robic C, Robert P, Port M, Le Bihan D, Pernot M, Tanter M, Lethimonnier F, Mériaux S. Dynamic study of blood-brain barrier closure after its disruption using ultrasound: a quantitative analysis. J. Cereb. Blood Flow Metab. 2012;32:1948–58. doi: 10.1038/jcbfm.2012.100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. McDannold N, Arvanitis CD, Vykhodtseva N, Livingstone MS. Temporary disruption of the blood-brain barrier by use of ultrasound and microbubbles: Safety and efficacy evaluation in rhesus macaques. Cancer Res. 2012;72:3652–3663. doi: 10.1158/0008-5472.CAN-12-0128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. McDannold N, Vykhodtseva N, Hynynen K. Blood-Brain Barrier Disruption Induced by Focused Ultrasound and Circulating Preformed Microbubbles Appears to Be Characterized by the Mechanical Index. Ultrasound Med. Biol. 2008;34:834–840. doi: 10.1016/j.ultrasmedbio.2007.10.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. McDannold N, Zhang Y, Vykhodtseva N. The Effects of Oxygen on Ultrasound- Induced Blood–Brain Barrier Disruption in Mice. Ultrasound Med. Biol. 2017;43:469–475. doi: 10.1016/j.ultrasmedbio.2016.09.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McMahon D, Hynynen K. Acute inflammatory response following increased blood-brain barrier permeability induced by focused ultrasound is dependent on microbubble dose. Theranostics. 2017;7:3989–4000. doi: 10.7150/thno.21630. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Meairs S. Facilitation of Drug Transport across the Blood-Brain Barrier with Ultrasound and Microbubbles. Pharmaceutics. 2015;7:275–93. doi: 10.3390/pharmaceutics7030275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Paasonen J, Stenroos P, Salo RA, Kiviniemi V, Gröhn O. Functional connectivity under six anesthesia protocols and the awake condition in rat brain. Neuroimage. 2018;172:9–20. doi: 10.1016/j.neuroimage.2018.01.014. [DOI] [PubMed] [Google Scholar]
  32. Park J, Zhang Y, Vykhodtseva N, Jolesz FA, McDannold NJ. The kinetics of blood brain barrier permeability and targeted doxorubicin delivery into brain induced by focused ultrasound. J. Control. Release. 2012;162:134–42. doi: 10.1016/j.jconrel.2012.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Paxinos G, W C. The Rat Brain in Stereotaxic Coordinates. Elsevier Academic Press; 2005. [Google Scholar]
  34. Rombouts SARB, Barkhof F, Goekoop R, Stam CJ, Scheltens P. Altered resting state networks in mild cognitive impairment and mild Alzheimer’s disease: An fMRI study. Hum. Brain Mapp. 2005;26:231–239. doi: 10.1002/hbm.20160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sala A, Caminiti SP, Presotto L, Premi E, Pilotto A, Turrone R, Cosseddu M, Alberici A, Paghera B, Borroni B, Padovani A, Perani D. Altered brain metabolic connectivity at multiscale level in early Parkinson’s disease. Sci. Rep. 2017;7:4256. doi: 10.1038/s41598-017-04102-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. SPM12. SPM12 Framework [WWW Document] 2014 [Google Scholar]
  37. Valdés-Hernández PA, Sumiyoshi A, Nonaka H, Haga R, Aubert-Vásquez E, Ogawa T, Iturria-Medina Y, Riera JJ, Kawashima R. An in vivo MRI Template Set for Morphometry, Tissue Segmentation, and fMRI Localization in Rats. Front. Neuroinform. 2011;5:26. doi: 10.3389/fninf.2011.00026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. van Meer MPA, van der Marel K, Wang K, Otte WM, el Bouazati S, Roeling TAP, Viergever MA, Berkelbach van der Sprenkel JW, Dijkhuizen RM. Recovery of Sensorimotor Function after Experimental Stroke Correlates with Restoration of Resting-State Interhemispheric Functional Connectivity. J. Neurosci. 2010;30:3964–3972. doi: 10.1523/JNEUROSCI.5709-09.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wu T, Wang L, Chen Y, Zhao C, Li K, Chan P. Changes of functional connectivity of the motor network in the resting state in Parkinson’s disease. Neurosci. Lett. 2009;460:6–10. doi: 10.1016/j.neulet.2009.05.046. [DOI] [PubMed] [Google Scholar]

RESOURCES