Focused ultrasound with microbubble agents can open the blood brain barrier focally and transiently. This technique has been used to deliver a wide range of agents across the blood brain barrier. This article provides a detailed protocol for the localized delivery to the rodent brain with or without MRI guidance.
Stereotaxic surgery is the gold standard for localized drug and gene delivery to the rodent brain. This technique has many advantages over systemic delivery including precise localization to a target brain region and reduction of off target side effects. However, stereotaxic surgery is highly invasive which limits its translational efficacy, requires long recovery times, and provides challenges when targeting multiple brain regions. Focused ultrasound (FUS) can be used in combination with circulating microbubbles to transiently open the blood brain barrier (BBB) in millimeter sized regions. This allows intracranial localization of systemically delivered agents that cannot normally cross the BBB. This technique provides a noninvasive alternative to stereotaxic surgery. However, to date this technique has yet to be widely adopted in neuroscience laboratories due to the limited access to equipment and standardized methods. The overall goal of this protocol is to provide a benchtop approach to FUS BBB opening (BBBO) that is affordable and reproducible and can therefore be easily adopted by any laboratory.
Despite the many discoveries in basic neuroscience, the number of emerging treatments for neurodevelopmental and neurodegenerative disorders remains relatively limited1,2. A deeper understanding of the genes, molecules and cellular circuitry involved in neurological disorders has suggested promising treatments unrealizable in humans with current techniques3. Effective treatments are often limited by the need to be brain penetrable and site-specific4,5,6,7,8. However, existing methods of localized drug delivery to specific brain regions (e.g., delivery via injection or cannula) are invasive and require an opening to be made in the skull9. The invasiveness of this surgery prevents the routine use of localized delivery into the human brain. Additionally, tissue damage and the resulting inflammatory responses are ubiquitous confounds for basic and preclinical studies that rely on intracerebral injection10. The ability to noninvasively deliver agents across the blood brain barrier (BBB) and target them to specific brain regions could have a tremendous impact on treatments for neurological disorders, while simultaneously providing a powerful investigational tool for preclinical research.
One method of targeted transport across the BBB with minimal tissue damage is transcranial focused ultrasound (FUS) together with microbubbles to focally and transiently open the BBB11,12,13,14,15,16. FUS BBB opening has gained recent attention for the treatment of neurodegenerative disorders, stroke and glioma by localizing therapeutics to target brain regions such as neurotrophic factors17,18,19, gene therapies20,21,22, antibodies23, neurotransmitters24, and nanoparticles25,26,27,28,29. With its wide range of applications and its noninvasive nature30,31, FUS BBB opening is an ideal alternative to routine stereotaxic intracranial injections. Furthermore, due to its current use in humans30,32, preclinical investigations using this technique can be considered highly translational. However, FUS BBB opening has yet to be a widely established technique in basic science and preclinical research due to lack of accessibility. Therefore, we provide a detailed protocol for a benchtop approach to FUS BBB opening as a starting point for labs interested in establishing this technique.
These studies were conducted with either a high-power air backed FUS specific ultrasound transducer or a low power damped focused ultrasonic immersion transducer. The transducers were driven by an RF power amplifier designed for reactive loads and a standard benchtop function generator. Details for these items can be found in the Table of Materials.
All experimental procedures were done in accordance with UAB Institutional Animal Care and Use Committee (IACUC) guidelines.
1. Focused ultrasound driving equipment setup
2. Focused ultrasound benchtop setup
3. Intracranial targeting procedure
NOTE: Male Sprague Dawley rats weighing 250-350 g were used for these experiments. Animals had free access to water and rat chow, and were maintained on a 12:12 h light:dark cycle.
4. Focused ultrasound procedure
5. MRI confirmation of BBB opening
6. Perfusion and tissue collection
Here, we demonstrate that focused ultrasound with microbubbles can induce localized BBB opening using the parameters specified above with both the low-power immersion transducer (Figure 3) and the FUS transducer (Figure 4). First, in early experiments, the low-power immersion transducer was targeted to one brain hemisphere either anterior (Figure 3b) or medial (Figure 3a). Animals were then sacrificed 2 hours later with perfusion (Figure 3a) or without perfusion (Figure 3b) and 10 µm frozen brain sections were collected. FUS BBB opening was evident by EBD autofluorescence (excitation: 470 and 540 nm, emission: 680 nm) in the target hemisphere (white arrows Figure 3a and 3b).
We have found it best to perfuse the animals for clear visualization of BBB opening with EBD autofluorescence. However, BBB opening can still be visualized without clearing the blood vessels (Figure 3b). Cellular uptake and clearance of EBD following BBB opening begins as soon as 30 minutes after BBB opening and increases over 24 hours37. For evaluation of BBB opening with EBD autofluorescence, it is best to sacrifice the animal between 15 minutes and 3 hours of BBB opening. Though ultimately, the time of sacrifice will depend on the agent that was delivered. For example, in an AAV study, 3 weeks post BBB opening and AAV delivery (Figure 5c) may be appropriate.
In later experiments, the FUS transducer was targeted to either the hippocampus (Figure 4a-c) or the anterior cingulate cortex (ACC) (Figure 4 d-f) and in addition to EBD, the MRI contrast agent gadobutrol (0.1 mL/kg) was IV injected to verify targeted opening of the BBB in vivo. Figure 4b,e show enhanced MRI contrast where the gadobutrol contrast has entered the tissue 1 hour after BBB opening and contrast agent injection. This contrast change is evident when comparing to the MRI prescans taken before the FUS procedure (Figure 4a,d). Animals were then sacrificed by perfusion 1.5 hours after BBB opening and 10 μm cryosections were collected. EBD autofluorescence is evident in FUS targeted regions further indicating location of BBB opening (Figure 4c,f). This figure highlights how MRI contrast can sometimes be difficult to see (as in the difference between Figure 4b and Figure 4e); therefore, it is helpful to confirm BBB opening with visualization of EBD autofluorescence as in the fluorescence micrograph in Figure 4f.
To assess whether this technique could be used for the targeted gene delivery AAV9-hsyn-GFP and gadobutrol contrast were injected IV (titer: 1.32 x 1014 GC/mL, 0.05 mL/kg) immediately after BBB opening in the hippocampus. The animal was then MR imaged 30 minutes after BBB opening and sacrificed 3 weeks later by perfusion. 10 μm cryosections were collected for fluorescent imaging of GFP expression. BBB opening was evident by gadobutrol contrast in the target hippocampus (Figure 5a,b). In addition, gene delivery was confirmed by GFP expression in the target hippocampus evident by green fluorescence (Figure 5c). Note that at this time point EBD has cleared out and is only evident in the ventricles (Figure 5c).
Figure 1: FUS benchtop setup. (a) FUS setup including the XYZ positioner, a 30 mm diameter PVC pipe for attachment of the transducer, the 3D-printed stereotaxic frame, and the infusion pump. (b) The end of the PVC pipe is capped, and a magnet is attached to it with epoxy. (c) Another matching magnet is attached to the top center of the high-power transducer with epoxy. (d) In addition, another matching magnet is attached to the pointer for nulling the positioner at the top and center of the MRI fiducial. (e) The pointer is eventually replaced with the high-power transducer and a water bath is coupled to the animal’s head with ultrasound gel. (f) The low-power immersion transducer can be attached to the PVC pipe with the 3D-printed transducer clip. (g) For positioning, the transducer is replaced with the 3D-printed pointer and the positioner is nulled at the top and center of the MRI fiducial. (h) The stationary 3D-printed frame holder allows for the animal to be returned to the same position after MRI if multiple FUS treatments are needed. Please click here to view a larger version of this figure.
Figure 2: Stereotaxic frame with MRI fiducial for MRI-guided coordinates. (a) Animals are first positioned in a 3D-printed stereotaxic frame equipped with an MRI fiducial. (b) The frame is then placed inside the MRI bed and distance from the fiducial (dotted circle) to the target brain region is measured using both coronal images for the dorsal/ventral (D/V) measurements and axial images for the rostral/caudal (R/C) measurements, medial/lateral (M/L) measurement can be collected from both axes. Animals are kept in the frame and transferred to the FUS station where a pointer is used to null the XYZ positioner at the location of the fiducial. The pointer is then replaced with the transducer and which can then be moved based on the coordinates collected. Please click here to view a larger version of this figure.
Figure 3: Evans blue dye (EBD) confirms FUS BBB opening both with and without perfusion. (a) Micrograph of a 10 µm brain section 2 hours after FUS BBB opening with the low-power immersion transducer targeted to the medial left hemisphere. This is a representative image of an animal that was perfused with 4% buffered formalin prior to tissue collection. BBB opening is evident by EBD red autofluorescence (arrow). (b) Micrograph of a 10 µm brain section 2 hours after FUS BBB opening targeted to the anterior left hemisphere. This is a representative image of an animal who was not perfused prior to tissue collection therefore, EBD remains in the blood vessels. BBB opening is evident where EBD has leaked out of the blood vessels (arrow). Please click here to view a larger version of this figure.
Figure 4: FUS BBB opening confirmed with Gadobutrol MRI contrast and EBD expression. MR images before (a and d) and after (b and e) BBB opening. gadobutrol contrast enhancement confirms location of BBB opening in vivo (b and e, arrows). (c) 10 μm brain section showing further confirmation of BBB opening with EBD autofluorescence (red) in the hippocampus (blue DAPI nuclear stain). Scale bar; 500 µm. (f) Micrograph of a 10 µm brain section following BBB opening in the anterior cingulate cortex evident by EBD red autofluorescence (arrow). Please click here to view a larger version of this figure.
Figure 5: Localized delivery of AAV9-hsyn-GFP to the hippocampus via FUS BBB opening. MRI confirmation of BBB opening with MRI contrast agent, MRI contrast (arrows) in both coronal (a) and axial (b) T1-weighted images. (c) histological confirmation of GFP expression in the FUS targeted hippocampus (green) 3 weeks after FUS BBB opening and AAV9-hsyn-GFP IV injection. Blue indicates DAPI nuclear stain for overall cellular morphology. Please click here to view a larger version of this figure.
Here we described a benchtop approach to microbubble assisted FUS BBB opening with alternative approaches including, two different transducers and methods for intracranial targeting with and without MRI guidance. Currently, in order to establish MRI-guided FUS BBB opening in the lab, there is the option to purchase excellent ready-to use devices that provide highly standardized and reproducible results with user-friendly interfaces. However, many labs are not prepared for the cost of such instruments. Therefore, the main goal of this protocol is to provide a starting point that any lab could establish in order to build their expertise in the technique.
FUS BBB opening is now a widely used technique and it is often the case that different groups utilize a variety of agents and anesthetics, each of which can affect the degree of BBB opening and extravasation. Importantly, the particular anesthetic used can affect the magnitude of BBB opening so it is important to consider this when executing this protocol38. Here, the anesthetic isoflurane is used because animals can be maintained under isoflurane for the length of the protocol and levels of isoflurane gas can be easily adjusted based on the animal’s respiration rate and heart rate. In addition, we delivered isoflurane with oxygen because it was more accessible than medical air; however, medical air may allow more extensive BBB opening39. Some MRI contrast agents are more appropriate for this protocol than others. For example, in our hands, gadoteridol produced no contrast enhancement even when EBD leakage was clearly present in tissue postmortem. Microbubble formulation is also important. Here we use perflutren lipid microspheres. Other microbubble formulations such as perflutren protein-type A microbubbles are readily available, but the type of microbubble used will affect the results15.
Controls of function generators can vary quite a bit, so refer to the manual for instructions on how to enter the settings listed in step 1. The appropriate command voltage (V peak to peak on the function generator) depends strongly on the transducer properties, RF amplifier gain, RF amplifier to transducer matching, the age and size of the animal, the microbubble type and concentration, and the desired treatment effect. The peak to peak V will need to be determined by trial and error. Start with the settings suggested in step 1 and determine the effect histologically. If there is tissue damage, lower the peak to peak V by 10% and try again. Likewise, if there is no BBBO then raise the peak to peak V by 10% and try again. Setting too high of a V can damage the low power immersion transducer. This will be apparent as a cracking or distortion of the transducer face. The manufacturing lead times on transducers can be long, so when starting, we suggest purchasing more than one transducer as a backup. Ultrasound transducers can also damage amplifiers if they are improperly matched. For simplicity and reliability, we suggest using a rugged power amplifier that can drive complex loads (such as the RF power amplifier in the materials table). Note that while the high-power transducer comes with a matching circuit, the low power immersion transducer does not. The suggested RF amplifier can handle the reflected power from the poorly matched transducer, but some amplifiers may be damaged in this configuration. Also, if step 2.7 proves problematic after installing the driver and software, double check that the proper serial port is selected in the software. Next try a different serial cable. If that fails, then seek local IT support.
From experience, it will take practice and multiple adjustments to achieve tight accuracy in brain region targeting. This can be seen in the differences of targeting between the early experiments (Figure 3) and the most recent experiments (Figure 5). We began the experiments using a classic stereotaxic frame and we include it as an option here if there is no access to a 3D printer or if there is no access to rodent MRI. However, MRI-guidance with MRI fiducials and the 3D printable frame provided (or of a custom design) is the ideal method. First, it accounts for individual differences between animals by gathering coordinates within the animal rather than relying on an averaged rat brain atlas. In addition, the use of MRI allows confirmation of FUS location targeting in vivo rather than relying on postmortem EBD expression. This is important when delivering agents that may require more than 24 hours to take effect such as AAVs (Figure 5). Lastly, the frame holder provided allows returning the animal back to the same position after MRI to fix any targeting errors or to repeat FUS after insufficient BBB opening without having to redo the coordinates. The portable 3D-printed stereotaxic frame can be used in any MRI with a clear bore of 200 mm or wider.
Depending on the blood vessel distribution in the target location, skull thickness40, the presence of ventricles, and other factors the degree of BBB opening can vary. For this reason, we provide a method for repeated targeting with the frame and frame holder. Accuracy of targeting depends critically on keeping the transducer focus at a consistent location with respect to the targeting pointer. The tip of this pointer should indicate the location in space of the center of the transducer focus when the transducer is attached to the XYZ positioner. The magnets allow the pointer and transducer to be easily swapped while maintaining this colocalization. The hole and protrusion in the magnets should match as precisely as possible. Any variability in this connection reduces the repeatability of the FUS focus targeting. However, there will be a spatial offset between the pointer tip and the ultrasound focus. Once it is confirmed that the offset is consistent, it can be corrected by using the MR images to calculate the difference in mm of the resulting BBB opening location (location of MRI contrast) and the intended target location. This difference can then be factored into the null location. Accuracy and repeatability also benefit greatly from using the same ultrasound frequency and using rats of a similar size and age in a given set of experiments. Ultrasound attenuation by the rat brain and rat skull varies with frequency and skull size and skull thickness varies with age. The skull is also a small cavity with respect to the ultrasound pulse and the incident ultrasound interacts with reflections inside the skull to produce a complex sound field that is dependent on the tissue, skull, frequency and the position of the transducer40.
As stated elsewhere, this protocol is intended to provide a low investment alternative to excellent MRI compatible commercial solutions that already available for purchase. There are important limitations that result from keeping the cost low. There are also limitations inherent to the technique given physics and the current state of the art. Remarkably, despite these limitations, and as shown in the representative results, we can achieve consistent delivery of dyes, particles, and viruses to the hippocampus of rats with submillimeter accuracy. The most important limitations are that (1) the shape of the FUS focus depends on the intervening tissue and especially the shape and thickness of the skull{…}. The need to remove the animal from the MRI to perform the FUS treatment prevents real-time feedback on the localization and intensity of the FUS focus. Without this real-time feedback several experiments need to be done to confirm setting for each combination of targeting location. Once the settings are “dialed in”, we have found good repeatability. (2) The XYZ positioning and fiducial system as constructed, while precise, does not provide accuracy in the coordinate frame of the positioner from experiment to experiment. The relative locations of the XYZ home, the rodent’s skull, and the frame can move relative to each other from experiment to experiment. This is obviated by using an MRI image for targeting, a targeting pointer and fiducial with known location in space relative to the FUS focus, ensuring the MRI coordinate system is parallel to the XYZ positioner system, by doing test treatments prior to the set of real treatments and by doing the entire procedure within one session so that the animal does not need to be repositioned into the frame. Note that, because only one fiducial is used, frame rotations are not correctable, so it is critical to ensure the frame is level with respect to the MRI bed and bore. In summary, the pointer location does not indicate the true FUS focus, but we have found the offset is consistent for a given brain location provided the skull does not rotate relative to the ultrasound transducer. Also note that consistent timing of gadobutrol delivery relative to the BBB opening and the imaging is crucial for consistent results especially if MRI contrast change is used as a proxy for the amount of BBB opening (see 36).
We first began FUS BBB opening in the lab with the low-power immersion transducer described above. We found that it is an affordable option for getting started with this technique. Most importantly, adaptation of this protocol can provide a noninvasive alternative to intracranial stereotaxic surgery and preclinical research performed with this technique can be considered highly translational due to the current use of transcranial FUS in humans30,32,41. Once established in a lab, this technique can be used as a noninvasive alternative to stereotaxic surgery. Thus, transforming a strictly investigative tool into a highly translational tool. Our lab will use this technique for the MRI guided, localized delivery of viruses and nanoparticles to develop novel, non-invasive neuromodulation techniques that can be used in freely behaving awake rodents and non-human primates. The current work focuses on Designer Drugs Exclusively Designed for Designer Receptors (DREADDs) and the sensitization of neurons to low doses of high energy particles such as x-rays. The lab is also working on a new version of this protocol that can be performed in a human 3 T scanner to eliminate the need to move the animal during the treatment and to allow real-time targeting feedback.
The authors have nothing to disclose.
This research was supported in part by an NSF EPSCoR Research Infrastructure grant to Clemson University (1632881). In addition, this research was supported in part by the Civitan International Research Center, Birmingham, AL. The authors gratefully acknowledge the use of the services and facilities of the University of Alabama at Birmingham Small Animal Imaging Shared Facility Grant [NIH P30 CA013148]. The authors acknowledge Rajiv Chopra for his support and guidance.
Bubble shaker | Lantheus Medical Imaging | VMIX | VIALMIX, actiation device used to activate Definity microbubbles |
Catheter plug/ Injection cap | SAI infusion technologies | Part Number: IC | Catheter plug/ Injection cap |
Evans blue dye | Sigma | E2129-10G | Evans blue dye |
Function generator | Tektronix | AFG3022B | Dual channel, 250MS/s, 25MHz |
FUS transducer, 1.1MHz | FUS Instruments | TX-110 | 1 MHz MRI-compatible spherically focused ultrasound transducer with a hydrophone |
Heating pad for Mice and Rats | Kent Scientific | PS-03 | Heating pad- PhysioSuite for Mice and Rats |
Infusion pump | KD Scientific | 780100 | KDS 100 Legacy Single Syringe Infusion Pump |
Kapton tape | Gizmo Dorks | https://www.amazon.com/dp/B01N1GGKRC/ ref=cm_sw_em_r_mt_dp_U_GbR7Db56HKD91 |
Gizmo Dorks Kapton Tape (Polyimide) for 3D Printers and Printing, 8 x 8 inches, 10 Sheets per Pack |
Low power immersion transducer, 1MHz | Olympus | V303-SU | Immersion Transducer, 1 MHz, 0.50 in. Element Diameter, Standard Case Style, Straight UHF Connector, F=0.80IN PTF |
Magnet sets | WINOMO | https://www.amazon.com/dp/B01DJZQJBG/ ref=cm_sw_em_r_mt_dp_U_JYQ7DbM32E5QC |
WINOMO 15mm Sew In Magnetic Bag Clasps for Sewing Scrapbooking – 10 Sets |
RF amplifier | E&I | A075 | 75W |
Tail vein catheter | BD | 382512/ Fisher Item: NC1228513 | 24g BD Insyte Autoguard shielded IV catheters (non-winged) |
Ultrasound contrast microbubbles | Lantheus Medical Imaging | DE4, DE16 | DEFINITY (Perflutren Lipid Microsphere) |
Ultrasound gel | Aquasonic | https://www.amazon.com/dp/B07FPQDM4F/ ref=cm_sw_em_r_mt_dp_U_D6Q7Db3J9QP7P |
Ultrasound Gel Aquasonic 100 Transmission 1 Liter Squeeze Bottle |
Winged infusion sets, 22ga. | Fisher Healthcare | 22-258087 | Terumo Surflo Winged Infusion Sets |
motor controller software | N/A | N/A | custom software written in LabView for controlling the Velmex motor controller |
runtime environment for the motor controller software | National Instruments | LabView runtime engine version 2017 or better | https://www.ni.com/en-us/support/downloads/software-products/download.labview.html |
3 axis Linear stage actuator (XYZ positioner) | Velmex | ||
bolts | Velmex | MB-1 | BiSlide Bolt 1/4-20×3/4" Socket cap screw (10 pack), Qty:3 |
motor controller | Velmex | VXM-3 | Control,3 axis programmable stepping motor control, Qty:1 |
mounting cleats | Velmex | MC-2 | Cleat, 2 hole BiSlide, Qty:6 |
mounting cleats | Velmex | MC-2 | Cleat, 2 hole BiSlide, Qty:2 |
usb to serial converter | Velmex | VXM-USB-RS232 | USB to RS232 Serial Communication Cable 10ft, Qty:1 |
x-axis linear stage | Velmex | MN10-0100-M02-21 | BiSlide, travel=10 inch, 2 mm/rev, limits, NEMA 23, Qty:1 |
x-axis stepper motor | Velmex | PK266-03A-P1 | Vexta Type 23T2, Single Shaft Stepper Motor, Qty:1 |
y-axis linear stage | Velmex | MN10-0100-M02-21 | BiSlide, travel=10 inch, 2 mm/rev, limits, NEMA 23, Qty:1 |
y-axis stepper motor | Velmex | PK266-03A-P1 | Vexta Type 23T2, Single Shaft Stepper Motor, Qty:1 |
z-axis damper | Velmex | D6CL-6.3F | D6CL Damper for Type 23 Double Shaft Stepper Motor, Qty:1 |
z-axis linear stage | Velmex | MN10-0100-M02-21 | BiSlide, travel=10 inch, 2 mm/rev, limits, NEMA 23, Qty:1 |
z-axis stepper motor | Velmex | PK266-03B-P2 | Vexta Type 23T2, Double Shaft Stepper Motor, Qty:1 |
3D printable files | |||
Immersion transducer mount and pointer | https://www.tinkercad.com/things/cRgTthGXSRq | ||
Stereotaxic frame | https://www.tinkercad.com/things/ilynoQcdqlH | ||
Stereotaxic frame holder | https://www.tinkercad.com/things/aZNgqhBOHAX | ||
9.4T small bore animal MRI | Bruker | Bruker BioSpec 94/20 | ParaVision version 5.1 |
AAV9-hsyn-GFP | Addgene | ||
Cream hair remover | Church & Dwight | Nair cream | |
gadobutrol MRI contrast agent | Bayer | Gadavist (Gadobutrol, 1mM/mL) | |
Stereotactic frame | Stoelting | #51500 | not MRI compatible |
turnkey FUS delivery device | FUS Instruments | RK-300 | ready to use MRI compatible FUS for rodents |