Safe and stable noninvasive focal gene delivery to the mammalian brain following focused ultrasound

Restricted access


Surgical infusion of gene therapy vectors has provided opportunities for biological manipulation of specific brain circuits in both animal models and human patients. Transient focal opening of the blood-brain barrier (BBB) by MR-guided focused ultrasound (MRgFUS) raises the possibility of noninvasive CNS gene therapy to target precise brain regions. However, variable efficiency and short follow-up of studies to date, along with recent suggestions of the potential for immune reactions following MRgFUS BBB disruption, all raise questions regarding the viability of this approach for clinical translation. The objective of the current study was to evaluate the efficiency, safety, and long-term stability of MRgFUS-mediated noninvasive gene therapy in the mammalian brain.


Focused ultrasound under the control of MRI, in combination with microbubbles consisting of albumin-coated gas microspheres, was applied to rat striatum, followed by intravenous infusion of an adeno-associated virus serotype 1/2 (AAV1/2) vector expressing green fluorescent protein (GFP) as a marker. Following recovery, animals were followed from several hours up to 15 months. Immunostaining for GFP quantified transduction efficiency and stability of expression. Quantification of neuronal markers was used to determine histological safety over time, while inflammatory markers were examined for evidence of immune responses.


Transitory disruption of the BBB by MRgFUS resulted in efficient delivery of the AAV1/2 vector to the targeted rodent striatum, with 50%–75% of striatal neurons transduced on average. GFP transgene expression appeared to be stable over extended periods of time, from 2 weeks to 6 months, with evidence of ongoing stable expression as long as 16 months in a smaller cohort of animals. No evidence of substantial toxicity, tissue injury, or neuronal loss was observed. While transient inflammation from BBB disruption alone was noted for the first few days, consistent with prior observations, no evidence of brain inflammation was observed from 2 weeks to 6 months following MRgFUS BBB opening, despite delivery of a virus and expression of a foreign protein in target neurons.


This study demonstrates that transitory BBB disruption using MRgFUS can be a safe and efficient method for site-specific delivery of viral vectors to the brain, raising the potential for noninvasive focal human gene therapy for neurological disorders.

ABBREVIATIONS AAV = adeno-associated virus; AAV1/2 = AAV serotype 1/2; BBB = blood-brain barrier; BSA = bovine serum albumin; DAB = 3,3′-diaminobenzidine; DAPI = 4′,6-diamino-2-phenylindole; Gd-DTPA = gadopentetate dimeglumine; GFP = green fluorescent protein; MRgFUS = MR-guided focused ultrasound; PBS = phosphate-buffered saline; rAAV = recombinant AAV; RF = radiofrequency; TBST = Tris-buffered saline with Triton.

Article Information

Correspondence Michael G. Kaplitt: Weill Cornell Medical College, New York, NY.

INCLUDE WHEN CITING Published online April 27, 2018; DOI: 10.3171/2017.8.JNS17790.

Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

© AANS, except where prohibited by US copyright law.



  • View in gallery

    MRgFUS facilitates AAV-mediated gene delivery to the brain. A: Gd-DTPA–enhanced T1-weighted images collected after sonication showed disruption of the BBB and Gd-DTPA extravasation in brain parenchyma (dashed line). B: The brain tissue was harvested 3 weeks after sonication and the colocalization of GFP and NeuN was visualized using immunostaining. Bar = 50 μm. C: DAB visualization of GFP transduction from serial sections centered on the targeted area (serial sections through the center of the targeted points).

  • View in gallery

    MRgFUS facilitates stable, long-term GFP transduction. A: High-magnification immunostaining for GFP and NeuN reveals a dominantly neuronal population of GFP-transduced cells in the striatum. Bar = 50 μm. B: Bar graph showing that quantification of striatal GFP transduction is stable over time. GFP-positive neurons are expressed as a percentage of the total number of striatal neurons per 20 hpf per animal (see Methods). C: Bar graph demonstrating that the MRgFUS-mediated GFP transduction is restricted mainly to neurons.

  • View in gallery

    Detection of striatal MRgFUS-facilitated AAV-mediated GFP transduction 16 months after sonication. Gd-DTPA–enhanced T1-weighed images collected after sonication showed disruption of the BBB and Gd-DTPA extravasation in brain parenchyma (dashed line). Histological analysis of the brain harvested 16 months after sonication showed GFP transduction mostly in neurons. Bar = 50 μm.

  • View in gallery

    MRgFUS-facilitated AAV-mediated gene delivery in peripheral organs is present in the short term but not the long term. Analysis of high-power immunofluorescent images of tissue collected from animals with MRgFUS with AAV1/2.GFP, and animals with AAV1/2.GFP stereotactically administered in the striatum shows no long-term GFP transgene expression in the liver, heart, and lungs. Bar = 100 μm. FUS = MRgFUS.

  • View in gallery

    No evidence of long-term inflammatory response induced by unilateral striatal MRgFUS-mediated gene transfer to the brain. Both Iba1 (microglia marker) and GFAP (astrocytic marker) returned to baseline levels by week 2, suggesting that local inflammatory response is transitory. Bar = 50 μm.





All Time Past Year Past 30 Days
Abstract Views 150 150 60
Full Text Views 77 77 36
PDF Downloads 69 69 40
EPUB Downloads 0 0 0


Google Scholar