Characterization of Blood–Brain Barrier Opening Induced by Transcranial Histotripsy in Murine Brains

Abstract

Objective:

Transcranial histotripsy has shown promise as a non-invasive neurosurgical tool, as it has the ability to treat a wide range of locations in the brain without overheating the skull. One important effect of histotripsy in the brain is the blood–brain barrier (BBB) opening (BBBO) at the ablation site, but there is a knowledge gap concerning the extent of histotripsy-induced BBBO. Here we describe induction of BBBO by transcranial histotripsy and use of magnetic resonance imaging (MRI) and histology to quantify changes in BBBO at the periphery of the histotripsy ablation zone over time in the healthy mouse brain.

Methods:

An eight-element, 1 MHz histotripsy transducer with a focal distance of 32.5 mm was used to treat the brains of 23 healthy female BL6 mice. T1-gadolinium (T1-Gd) MR images were acquired immediately following histotripsy treatment and during each of the subsequent 4 wk to quantify the size and intensity of BBB leakage.

Results:

The T1-Gd MRI results revealed that the hyperintense BBBO volume increased over the first week and sub-sided gradually over the following 3 wk. Histology revealed complete loss of tight junction proteins and blood vessels in the center of the ablation region immediately after histotripsy, partial recovery in the periphery of the ablation zone 1 wk following histotripsy and near-complete recovery of tight junction complex after 4 wk.

Conclusion:

These results provide the first evidence of transcranial histotripsy-induced BBBO and repair at the periphery of the ablation zone.

Introduction

The blood–brain barrier (BBB) is a highly selective and restrictive barrier that exists at the interface between the blood vessel wall and neuronal tissue in the brain, it and serves as a critical defense to prevent harmful agents or hematologically mediated antigens from entering the brain parenchyma [1]. The unique property of the BBB is the existence of a well-developed tight junction complex between brain endothelial cells to eliminate any paracellular space and exchange between neighboring cells [2]. However, the BBB limits more than 98% of small-molecule drugs and nearly all large-molecule drugs from reaching the brain parenchyma in therapeutically relevant concentrations [3], which is the major hurdle for drug-based brain therapy. ZO-1 and claudin-5 are tight junctional proteins (TJPs) reported to play key roles in the restrictive permeability property of the BBB. ZO-1 is hypothesized to regulate the recruitment and anchoring of transmembrane TJPs (i.e., claudin-5), while claudin-5 plays a major role in limiting the paracellular space. Any dysfunction associated with the disorganization of the tight junction complex leads to BBB permeability, often seen in various neuropathological conditions such as neurodegenerative and neuroinflammatory disorders [2].

Low-amplitude ultrasound combined with microbubbles has been reported to temporarily open the BBB [4,5]. Using ultrasound applied from outside the skull, as well as extradurally implanted ultrasound transducers, microbubbles can be activated to oscillate in the blood vessels and apply shear stress to the vessel wall, resulting in damaged BBB TJPs. This mechanism can be used to open the BBB selectively and temporarily for localized drug delivery [6–8]. Substantial pre-clinical work has been done to investigate microbubble-mediated focused ultrasound (FUS) induced BBB opening (BBBO) for the delivery of various components such as antibodies [9,10] and therapeutic agents [11,12]. Microbubble-mediated FUS BBBO typically uses a center frequency of 200 kHz–5 MHz, peak pressure of 0.3–1.5 MPa, pulse duration of 10–100 ms and duty cycle of 1%–20% [13,14], although continuous wave [15] or microsecond-length pulses [16] have also been found to be effective for BBBO. Multiple clinical trials are ongoing using magnetic resonance–guided FUS (MRgFUS) and implantable FUS devices for drug delivery in human patients to treat brain tumors [17] and Alzheimer disease [18].

Histotripsy is a focused ultrasound ablation technique that generates cavitation to mechanically fractionate the target tissue using high-pressure, microsecond-length ultrasound pulses [19,20]. The rapid expansion and collapse of cavitation microbubbles exert high strain and stress on the targeted tissue and disrupt the cellular membrane and structures [21]. Transcranial histotripsy has been reported to successfully ablate various locations inside the brain without heating the skull in pre-clinical studies using in vivo porcine and mouse brains [22–27]. In contrast to the microbubble-mediated FUS that uses very low pressure (<1.5 MPa) and relatively long ultrasound pulses (millisecond length), histotripsy uses extremely high rarefaction pressure (>20 MPa) and microsecond-length pulses, and it does not require microbubble injection [28]. Microbubble-mediated FUS typically induces stable cavitation [29], while histotripsy produces inertial cavitation, resulting in significantly higher mechanical strain and stress [20].

In this study, the temporal characteristics of transcranial histotripsy-induced BBBO were studied for the first time with T1-gadolinium (Gd) magnetic resonance imaging (MRI) and histology. Gd was used to visualize BBBO because it cannot normally pass through the BBB owing to its molecular weight (<400 Da). In the presence of a brain tumor or site of impaired BBB, Gd can enter the brain parenchyma and drastically shorten the T1 of surrounding molecules, resulting in a hyperintense region [30]. In this study, T1-Gd MRI, hematoxylin and eosin (H&E) histology and immunohistochemistry (IHC) with TJPs (claudin-5 and ZO-1) and blood vessel staining (collagen IV) were used to characterize BBBO within, at the periphery of and outside the histotripsy ablation zone over time. The knowledge gained will inform the development of transcranial histotripsy brain treatment.

Methods

Animal procedure

Transcranial histotripsy was delivered to a total of 23 healthy 8- to 10-wk-old B6 female mice (Taconic Farms, Rensselaer, NY, USA) with the stereotactic procedure previously described [23]. The stereotactic platform was used to ensure accurate targeting in the brain tissue. Of the 23 mice, 18 were used for MRI monitoring and H&E and TJP staining. From this cohort, 3 mice were euthanized at each time point to obtain the histology slides. The remaining 5 mice were used for collagen IV staining for blood vessel visualization at day 0 (n = 3) and day 7 (n = 2).

Each mouse was monitored via MRI with a T1-Gd sequence (fast spin-echo multislice T1, TR/TE (500/8 ms), slice thickness 1 mm, lateral resolution 0.5 mm, scan time 4 min) at the following time points: immediately after histotripsy (week 0, n = 18), week 1 (5–7 d after histotripsy, n = 15), week 2 (11–14 d after histotripsy, n = 5), week 3 (15–21 d after histotripsy, n = 6) and week 4 (22–28 d after histotripsy, n = 4). Two of the six mice that were used for week 3 MRI and three of the four mice that were used for week 4 MRI were not imaged at week 2 because of the unavailability of MRI. Gadolinium (Gadoteridol: 558.7 Da [31], ProHance, 279.3 mg/mL, Bracco, Monroe Township, NJ, USA) was injected via the intraperitoneal cavity. The T1-Gd image was acquired approximately 10 min after injection to allow peak circulation during data collection. For each time point, after MRI, mice were euthanized via CO₂ overdose, and their brains were extracted for H&E, claudin-5, ZO-1 and collagen IV staining (n = 3 for each time point).

On the day of histotripsy treatment and during MR image acquisition, mouse breathing rate and temperature were maintained with the protocol previously described [23]. Carprofen (Rimadyl, Pfizer, New York, NY, USA) analgesic (5 mg/kg) was administered subcutaneously on the day of histotripsy treatment and the following 2 d after the treatment. Before treatment, fur on the head was electrically shaved and chemically depilated (Nair, Church & Dwight Co., Ewing Township, NJ, USA) to ensure that no trapped air bubbles would block histotripsy pulses. After every histotripsy or MRI procedure, the animal was placed in a warm recovery chamber in isolation to fully wake up and rest before being returned to the original cage with other mice. The protocol described in this article and all mouse procedures were approved by the University of Michigan Institutional Animal Care Use Committee (IACUC).

Histotripsy treatment

A 1 MHz, eight-element focused element rodent histotripsy transducer (F# = 0.6) with a focal distance of 32.5 mm was used to produce a peak rarefaction pressure of 36 MPa in the free field to generate histotripsy cavitation. This measurement was based on a fiberoptic probe hydrophone (HFO 690, Onda, Sunnyvale, CA, USA). A 3 to 5 mm³ volume in the brain of each mouse was treated with transcranial histotripsy. The treatment grid was formed by dividing the prescribed volume into 0.25 mm spaced grid points for a raster scan. The focus was scanned through the grid by mechanically moving the transducer with a three-axis motor positioner. Each grid point was delivered with fifteen 1.5-cycle pulses at 5 Hz pulse repetition frequency (PRF) (0.0008% duty cycle). The untreated contralateral side of the brain was used as untreated control tissue. This parameter set was tested and chosen to be safely handled by mice and produce homogenization of brain tissue.

Quantification of BBBO through T1-Gd MRI

T1-Gd MRI images were acquired with a 7T small animal scanner (Varian Inc., Palo Alto, CA, USA) to evaluate the histotripsy ablation zone and resulting BBBO. The hyperintense region surrounding the histotripsy ablation zone was used to characterize the BBBO and was defined as three times the standard deviation of the pixel intensity of the contralateral region (Fig. 1). To measure the pixel intensity of the contralateral region, a circular ROI with a fixed diameter was placed on the contralateral side of the brain in each treatment slice in MATLAB (Math-Works, Natick, MA, USA) (Fig. 1A). Pixel intensities within each contralateral ROI were averaged and the standard deviation was calculated to determine the threshold pixel intensity for determining the hyperintense region:

$$\text{Threshold}={\mu }_{\text{contralateral}}+3{\sigma }_{\text{contralateral}}$$

Figure 1.

(A, B) Representative magnetic resonance imaging slice of the brain revealing the treatment region (white arrow) and contralateral ROI (red circle) (A) and thresholded hyperintense ROI used for quantification of T1-Gd magnetic resonance images (B). A circular ROI (red circle) with a fixed diameter was placed on the contralateral side of the brain for each treatment slice, and the mean pixel intensity and standard deviation were calculated. Pixels with intensities greater than three times the standard deviation of the mean contralateral pixel intensity were used to determine the hyperintense ROI in (B). The hyperintense ROI was normalized by the mean contralateral pixel intensity for each mouse, and the volume and mean pixel intensity were recorded to determine the level of gadolinium leakage through the blood–brain barrier. Gd, gadolinium; ROI, region of interest.

Here, ${\mu }_{\text{contralateral}}$ is the mean pixel intensity of the contralateral ROI, and ${\sigma }_{\text{contralateral}}$ is the standard deviation of the pixel intensities of the contralateral ROI; $3{\sigma }_{\text{contralateral}}$ was selected to best mimic the region selected by the neuroradiologist who was asked to blindly select regions with BBBOs from T1-Gd images. Pixels with intensities above the threshold were included in the hyperintense ROI (Fig. 1B). Hyperintense ROIs were normalized with the mean pixel intensity of the contralateral ROI for each mouse. To obtain the hyperintense volumes, the ROIs were converted to millimeters by multiplying by the MRI x/y resolution (0.001 mm²/pixel) and then multiplying by the slice thickness (1 mm).

Normalized pixel intensities and volumes for each time point were compared across time with a mixed model for repeated measures test with Tukey–Kramer adjustments. A p value <0.05 was considered to indicate significance. Statistical analysis was done in GraphPad Prism (GraphPad, San Diego, CA, USA).

H&E histology

After euthanization, the mouse skull was partially cut open along the sagittal suture and the brain was fixed in 10% formalin for 24–48 h. After fixation, the brains were cut coronally in half, and the anterior half, including the olfactory bulbs and the histotripsy lesion, was processed and sectioned by the University histology core (iLab, University of Michigan, Ann Arbor, MI, USA) for H&E staining. The posterior half of the brain was discarded. All slides produced were 4 μm thick.

Immunohistochemistry

Paraffin-embedded brain tissue was dewaxed and rehydrated through xylene and a series of alcohols (100%, 95%, 70% and 50%). Antigen retrieval was performed by boiling slides in 10 mM sodium citrate buffer (pH 6.0) for 10 min, followed by cooling and washing in phosphate-buffered saline (PBS) (pH 7.2). For IHC staining, brain samples were pre-incubated in blocking solution containing 5% normal goat serum and 0.05% Triton 100X (Sigma Aldrich) in PBS. Samples were then incubated overnight at 4°C with the following primary antibodies: claudin-5-Alexa-Flour 488 conjugated, ZO-1-Alexa Flour 594 conjugated and collagen IV-Alexa Flour 561 conjugated. All samples were imaged on a confocal laser scanning microscope (Nikon A1, Tokyo, Japan).

Interpretation of MRI and histology

A blind reading of MR images was done to identify BBBO and BBB repair by a board-certified neuroradiologist (N.C., American Board of Radiology) and double-blinded reading of histology was performed by one board-certified neuropathologist (S.C.P., American Board of Pathology) and a professor of pathology (A.V.A.).

Results

MRI gadolinium leakage

T1-Gd MR images revealed hyperintense regions for all time points (Fig. 2). Immediately following histotripsy (week 0), the histotripsy ablation zone was depicted on MRI as having a hypo- to iso-intense core with a hyperintense rim at the periphery. At week 1, the ablation zone appeared as a region with high hyperintensity and minimal hypo-intensity. At weeks 2–3, the hyperintense rim was diminished and a larger hypo-intense core was seen. At week 4, the ablation zone decreased to a region with minimal hyperintensity and areas of hypo- to iso-intense cores. The volume of the hyperintense region significantly increased from 5.46 ± 7.26 to 11.28 ± 5.88 mm³ from week 0 to week 1, respectively (p = 0.029, n = 18/15 [week 0/1]) and significantly diminished from week 1 to week 2 (2.49 ± 0.97 mm³, p = 0.049, n = 18/5 [week 1/2]), from week 1 to week 3 (3.11 ± 2.48 mm³, p = 0.022, n = 18/6 [week 1/3]) and from week 1 to week 4 (1.87 ± 0.86 mm³, p = 0.030, n = 18/4 [week 1/4]), indicating reduced gadolinium uptake for these time points (Fig. 3A). The pixel intensity of the hyperintense region exhibited no significant changes over time but trended toward a maximum at week 1 (1.56 ± 0.18) and reduced intensity over the following weeks (week 2: 1.42 ± 0.14, week 3: 1.40 ± 0.08, week 4: 1.35 ± 0.62, p > 0.05 for all comparisons) (Fig. 3B).

Figure 2.

T1-Gd images of three histotripsy-treated mice over the 4 wk following treatment. Each row depicts the same mouse over time. Hyperintense regions (red arrowheads) indicate blood–brain barrier opening via Gd enhancement. At week 0, the histotripsy ablation zone appeared dark with a hyperintense rim. The hyperintensity reached a peak at week 1, followed by a gradual decrease over the next 3 wk. Gd, gadolinium.

Figure 3.

Quantification of blood–brain barrier opening from T1-Gd magnetic resonance images. (A) Histotripsy lesion hyperintense volume. (B) Hyperintense normalized pixel intensity. Both the volume and pixel intensities were measured from thresholded regions of interest determined by the mean contralateral pixel intensities for each mouse. The hyperintense volume significantly increased from week 0 to week 1, then significantly decreased from week 1 to weeks 2, 3 and 4. The hyperintense pixel intensity trended toward a similar pattern. *Statistically significant difference (p < 0.05) between two time points. Gd, gadolinium.

Post-treatment hemorrhage

Hematoxylin and eosin staining revealed acellular debris and bleeding at weeks 0 and 1 in the histotripsy ablation zone that correlated with the ablation region seen on MRI (Fig. 4). Transcranial histotripsy led to acute bleeding in the treatment region that resolved within the first week of treatment, as reflected by the decreased hemorrhage in the ablation zone for weeks 2–4. At week 0, T1-Gd MRI revealed a hyperintense rim and hypo-intense center. The hypo-intense center was correlated with blood products in H&E and limited gadolinium uptake. At week 1, the blood products were partially removed as seen on H&E staining, which was concurrent with the increased gadolinium leakage through the BBBO shown by the hyperintense regions on T1-Gd MRI. Macrophages (Fig. 4, green arrow; Fig. S1, online only) were observed in the periphery of the ablation zone at week 1, noted by the multinuclear cells surrounding the histotripsy ablation. At week 2, there was a continued decrease in the number of red blood cells and further reduction in histotripsy zone size on MRI. At weeks 3–4, the processed blood products remained, histotripsy lesion size decreased and hyperintensity induced by T1-Gd further decreased in intensity.

Figure 4.

T1-Gd images (first column), H&E staining at 2 × and 20 × (second and third columns) and TJP staining for claudin-5 and ZO-1 (40 ×, fourth column; 60 ×, fifth column). Each row represents the same mouse. Red arrows indicate the histotripsy ablation region, green arrow indicates an area of macrophages (Fig. S1, online only) and yellow arrows indicate regions of hemosiderin. For TJP staining, the treatment zone reveals little to no expression of claudin-5 and ZO-1, indicating disassembly of the TJ complex at week 0. At weeks 2–3, partial TJP recovery was seen, with ZO-1 recovering before claudin-5 at week 2. At week 4, there was a lack of coherent TJPs inside the lesion, but near-complete recovery of the TJPs at the periphery of the lesion, depicted by co-localization of claudin-5 and ZO-1. White dashed lines represent the border of the histotripsy lesion, and white arrowheads indicate zoomed-in blood vessels. Gd, gadolinium; H&E, hematoxylin and eosin; TJP, tight junctional protein.

BBB tight junction and vessel damage and repair

Blood–brain barrier tight junction damage was characterized by antibody staining for two TJPs: ZO-1 and claudin-5 (Fig. 5). On the contralateral untreated side, the BBB exhibited intensive and continuous staining for ZO-1 (red) and claudin-5 (green). Claudin-5 and ZO-1 (yellow = red + green) co-localization indicated a stable and functional TJ complex, as the interaction between these two proteins plays a critical role in regulating the paracellular space occlusion and BBB integrity. For all images, the treatment zone was defined as the area with a decreased density of intact nuclei (from DAPI staining) and increased non-specific TJP staining of acellular debris.

Figure 5.

Immunohistochemical staining for DAPI, TJPs (claudin-5, ZO-1) and collagen IV. (A) Intact blood vessels are seen in the periphery of the treatment zone at day 0 (white arrows, first row), illustrated by high collagen IV signal (red) and co-localization of claudin-5 and ZO-1 (white). In the treatment zone, there was discontinuous collagen IV staining (white arrows, second row) with absence of TJP staining, indicating damaged blood vessels. (B) At day 7, increased collagen IV signal and TJP signal inside the treatment zone and at the periphery suggest blood vessel repair. Staining for the contralateral untreated side of the brain is illustrated in the bottom row. Rows 2–4 are shown at 40 × magnification. DAPI, 4,6-diamidino-2-phenylindole; P, periphery; TJP, tight junction protein; TZ, treatment zone.

At week 0, blood vessels in the histotripsy ablated and surrounding tissue exhibited little to no expression of claudin-5 and ZO-1, indicating disassembly of the TJ complex. At weeks 1–2, partial TJP recovery was seen, as blood vessels exhibited both claudin-5 and ZO-1 staining at the periphery of the ablation zone but did not exhibit continuous staining. Hematoxylin and eosin staining and T1-Gd data for week 1 support this observation that the coagulated blood products in the center of the ablation have been cleaned and allowed gadolinium leakage, indicating the presence of a leaky BBB. At weeks 3–4, there was a lack of coherent TJPs in the center of the lesion. At the border of the lesion, most of the blood vessels exhibited intensive claudin-5 and ZO-1 staining and co-localization. The scattered TJPs in blood vessels in the center of the lesion and the close-to-completely recovered TJPs at the periphery agree with the T1-Gd images and H&E: in weeks 3–4 post-histotripsy, the gadolinium leakage around the lesion and the size of the lesion decreased substantially compared with weeks 0–2. The claudin-5 and ZO-1 staining indicate that BBBO induced by histotripsy is due to destabilization of TJ complex, which is reversible and recovered after 1–2 wk and within 4 wk post-histotripsy.

Blood vessel damage at weeks 0 and 1 following histotripsy treatment was assessed with collagen IV staining, a marker for basement membrane (Figs. 6 and 7). At week 0, there were damaged blood vessels in the center of the treatment zone, as reflected by incomplete collagen IV staining and absence of TJP staining (Fig. 6). At the periphery (<200 μm from the lesion), some blood vessels appeared intact with destabilized tight junction complexes and intact basement membrane. At week 1, the size of the treatment zone decreased, and damaged blood vessels were partially recovered (Fig. 7). Although damaged blood vessels (defragmented collagen IV and absence of claudin-5 and ZO-1 staining) were still present in the treatment zone at week 1, segments of the BBB vessels appeared intact inside the lesion (bright and continuous collagen IV signal). At the periphery of the lesion, approximately 200 μm from the lesion core, most of the blood vessels were intact and similar to those on the contralateral side with continuous collagen IV staining and colocalization of claudin-5 and ZO-1. Quantification of IHC staining intensity for weeks 0 and 1 is illustrated in Figure S2 (online only) and confirms recovery of the blood vessels at the periphery of the treatment zone at week 1, but it indicates that peripheral TJPs are not yet fully recovered at week 1.

Figure 6.

Immunohistochemical staining for DAPI, TJPs (ZO-1 and claudin-5) and collagen IV for day 0 after histotripsy treatment at 20 × magnification of the TZ and P (first row) and 40 × magnification of the treatment zone (second row), periphery (third row) and contralateral region (fourth row) with zoomed-in images of specific blood vessels (arrowheads) in column 6. In the treatment zone, there was discontinuous collagen IV staining with the absence of TJP staining, indicating damaged blood vessels (arrowhead). In the periphery, there were intact blood vessels with high ZO-1, claudin-5 and collagen IV signals that resembled blood vessels in the contralateral region. Scale bars = 100 μm for 20 × images, 50 μm for 40 × images and 25 μm for zoomed-in images. DAPI, 4,6-diamidino-2-phenylindole; P, periphery; TJP, tight junction protein; TZ, treatment zone.

Figure 7.

Immunohistochemical staining for DAPI, TJPs (ZO-1 and claudin-5) and collagen IV on day 7 after histotripsy treatment at 20 × magnification of the TZ and P (first row) and 40 × magnification of the TZ (second row), P (third row) and contralateral region (fourth row) with zoomed-in images of specific blood vessels (arrowheads) in column 6. Some intact blood vessels are shown in the treatment zone with co-localized TJP staining and continuous collagen IV signal. Blood vessels in the periphery region exhibit a high TJP signal and collagen IV staining. Blood vessels in the TZ and P resemble blood vessels in the contralateral region. Scale bars = 100 μm for 20 × images, 50 μm for 40 × images and 25 μm for zoomed-in images. DAPI, 4,6-diamidino-2-phenylindole; P, periphery; TJP, tight junction protein; TZ, treatment zone.

Discussion

Blood–brain barrier opening resulting from histotripsy was investigated in healthy murine brains using gadolinium-enhanced T1-weighted MRI and histology with three markers for BBB integrity (TJPs claudin-5 and ZO-1) and blood vessel integrity (collagen IV). Histotripsy led to BBB disruption in the center of the ablation zone and BBBO in the periphery observed in T1-Gd MRI. Gd penetration was seen at the periphery of the histotripsy zone in week 0 and throughout the ablation zone in week 1, then decreased in volume and intensity from weeks 1 to 4. IHC staining for claudin-5 and ZO-1 revealed reversible TJP damage at the periphery of the histotripsy ablation zone, which indicated peak BBBO 1 wk after histotripsy and near-complete recovery after 3–4 wk. Collagen IV staining indicated damaged blood vessels at the center of the lesion at day 0, but intact and repaired blood vessels in the center and periphery of the lesion as soon as 7 d after histotripsy treatment.

At week 0, the histotripsy treatment zone consisted of a center hypo-intense region surrounded by a slightly enhanced hyperintense rim on T1-Gd MRI. The hypo-intense region within the histotripsy ablation zone was due to the coagulation of the ablated tissue and disrupted BBB. The hyperintense rim was explained by the BBBO and was supported with the TJP damage observed at the periphery zone by claudin-5 and ZO-1 absence of staining for weeks 0 and 1. At week 1, the hyperintensity increased on T1-Gd MRI and macrophages were observed along with a decrease in blood products in H&E staining. The combination of blood product decrease and immune activity of macrophages to clean cellular debris was thought to contribute to the increased permeability in the histotripsy zone, leading to increased gadolinium leakage into the brain parenchyma. Concurrently, partially intact BBB tight junctions appeared at the periphery of the histotripsy ablation zone. After week 1, the hyperintensity on T1-Gd decreased and a further decrease in blood products was observed. At week 2, ZO-1 expression was observed before claudin expression, which is in agreement with existing literature that suggests that ZO-1 is essential for anchoring claudin strands and occluding to the actin cytoskeleton [32]. By week 3, the immunofluorescent staining for the claudin-5 and ZO-1indicated BBB restoration in the periphery.

In this study, we observed a lack of TJP expression in the center of the histotripsy ablation and reversible TJP damage at the periphery of ablation, indicating temporary BBBO at the periphery. At the center of the ablation zone, histotripsy generated high mechanical strain and stress that disrupted the target brain tissue, including the small blood vessels. It is known that the mechanical stress induced by histotripsy-generated cavitation decreases with increasing distance from the cavitation site [21,33]. The literature on BBBO induced by microbubbles and low-pressure ultrasound pulses indicates that the shear stress induced by stable cavitation in the vessel wall causes disassembly of the TJ complex, leading to temporary BBB “opening” [6,7]. Therefore, the mechanical stress at the periphery of the ablation zone is low enough to conserve tissue viability, but high enough to damage the TJ complex. This would explain the enhanced rim surrounding the histotripsy zone on T1-Gd MRI immediately after histotripsy. This was observed in one mouse at week 0, which showed a large hyperintense rim immediately after treatment. This was explained by incomplete shaving of the mouse head and trapped air bubbles in the fur, which may have led to decreased effective pressure in the target tissue and/or defocusing of the ultrasound leading to a wider region of ultrasound exposure. Because of variation in fur length still present on the head after shaving, as well as small differences in skull thickness, the effective pressure could have varied across mice, leading to inconsistent degrees of BBBO. Further investigation is needed to understand the exact mechanisms underlying histotripsy-induced BBBO and the impact of parameters (such as pressure) on BBBO.

Recovery of claudin-5 and ZO-1 at the treatment boundary indicated that BBBO is reversible in this area of the histotripsy lesion. The TJ complex began to recover 1 wk post-histotripsy and appeared similar to healthy BBB TJ complex by week 4. Additionally, collagen IV staining revealed intact blood vessels at the periphery of the lesion at week 0 and an increased number of intact blood vessels in the treatment zone at week 1. This indicates that damaged blood vessels mostly exist in the center of the treatment region, with a tendency to recover structurally within the first week after injury. The periphery of the lesion mostly revealed diminished brain endothelial barrier integrity, while structurally, shown by the intact basal membrane, blood vessels were not modified. Leveraging this BBBO effect, histotripsy may be combined with drug delivery for brain tumor treatment, where histotripsy can be used to debulk the tumor core, and the BBBO induced by histotripsy can be used to enhance drug delivery (e.g., chemotherapy or immunotherapy drugs) at the periphery of the ablation zone to treat any residual tumor cells. The understanding of the BBB recovery timeline post-histotripsy can be beneficial to guide the drug delivery protocol.

Gliosis is a central nervous system wound healing process in response to trauma [34] and could play a role in transcranial histotripsy ablation response. Our results indicate that the pixel intensity of the hyperintense region at week 4 was approximately 30% higher than that of the contralateral side and the hyperintense volume was non-zero. This enhancement may be attributed to the absence of BBB in the center of the ablation region but could also be due in part to gliosis in this area. Further investigation is needed with respect to the extent to which gliosis affects the BBB permeability of the center of the histotripsy region.

Conclusion

This study provides early evidence of temporary BBBO at the periphery of the histotripsy ablation zone in the brain. Further studies need to be conducted to characterize the size of particles that can penetrate through the BBBO [35], the spatial extent of the BBBO and the time point at which the tight junction recovery starts at the periphery of the histotripsy zone. These future studies will also inform the optimal timing for efficient drug delivery to the brain tissue along with the extent and duration of the BBBO.

Data availability statement

Partial magnetic resonance imaging and histological data are available in the supplemental data repository. Only partial data are available because 10 GB is the maximum allowed storage space in Mendeley Data. Please email the corresponding authors for access to more magnetic resonance imaging and histological data.