Noninvasive Delivery of Biologicals to the Brain

The interfaces of the blood vessels with parenchyma, and of blood vessels with cerebrospinal fluid, present relative barriers to the ingress of small molecules, oligonucleotides, particulates, and cellular elements into the brain (6). Small molecules may have specific carrier systems that allow for transendothelial transport (7). Phosphorothioate oligonucleotides, chemically modified to ensure nuclease resistance, are reportedly transported across the blood-brain barrier (BBB) through yet unidentified transporters, albeit with limited efficiency (8, 9). Transcytotic mechanisms also exist for larger materials, such as large proteins, exosomes, microbes, and immune cells (10). In the latter circumstances, there are specific ligands that allow luminal attachment and a complex sequence of events that eventuate in endo-vesicular delivery to the parenchymal side of the barrier.

The integrity of the tight intercellular junctions plays a role in essentially excluding bloodborne elements that are toxic to the brain, including glutamate, prothrombin, and plasminogen (6). The ionic channels active in endothelial membranes are also tasked with providing a relatively stable concentration of calcium and potassium in the brain interstitium, which is required for the normal electrophysiological function of neurons (11). By contrast, blood concentrations of these elements may fluctuate under differing physiological conditions, including fasting or aerobic exercise (12), and engagement during motor, affective, and cognitive tasks. Cerebral ischemic neuronal dysfunction can follow the incorrect balance of neural ions (13), and profound or prolonged disruption of these systems can lead to neurodegeneration (14). Therefore, successful delivery of therapeutics to the brain through the BBB must be accomplished without causing permanent damage or disruption to these systems.

One such mechanism for increasing permeability of the BBB for the delivery of therapeutics to the brain is transcranial focused ultrasound. Various forms of ultrasound equipment are available for human treatment and investigation, ranging from multiprobe devices used in a magnetic resonance imaging (MRI) environment, to single probe devices applied in the temple region. Transtemporal probes are targetable with integrated Doppler technology (15) or with optical tracking devices. Before the development of clot retrieval devices (16, 17), ultrasound delivered by a transtemporal window was used extensively for mechanical clot fragmentation and for improved blood flow, along with intravenous clot-dissolving agents (18). Safety has been established at frequencies of 2 MHz, but an increased risk of brain hemorrhage has characterized more penetrating, lower frequency treatment (19). Earlier work (20) on the opening of tight junctions after acute stroke or trauma has afforded delivery of stem cells to the brain through intravenous infusion. Mechanical effects of sonification have been used to pry open tight junctions after microbubble infusion (21). The latter technique presents significant safety issues, because the tight junction opening may lead to intracerebral hemorrhage (22, 23). Thus, increasing permeability of the BBB without using microbubbles has been shown to be effective. This experience with ultrasound treatment has suggested that such technology may have several effects on brain tissue, including temporary alteration of vascular properties (24, 25). Indeed, ultrasound has been shown to increase the adhesiveness of the luminal surface of capillaries (26, 27), fostering potential noninvasive delivery to targeted brain tissue (21, 28–32). Furthermore, in our experience of focused ultrasound in humans, blood flow can be increased up to eightfold, while also monitoring effects with quantitative arterial spin labeling (33). In the latter circumstance, poorly penetrating agents may be more efficiently delivered by increasing local capillary delivery. In other words, ultrasound may be used to facilitate focal delivery without risking injury to blood vessel integrity (34).

Stem cells may be given intravenously and then channeled to specific brain sites by using focused ultrasound. Much of the early research with stem cells in neurological conditions used fetal progenitors of unspecialized mesenchymal stem cells from autologous or allogeneic sources (35–37). Although transplanted stem cells in small numbers may survive to a certain extent, most will die while releasing packets of nucleic acids (e.g., microRNA, messenger RNA, genomic and mitochondrial DNA) and growth factors in the form of exosomes (38–40). There are several advantages to using stem cells, including their quality to migrate and home in on areas of inflammation and hypoxia in tumors (41–43). Furthermore, when they persist in tissue, they may provide a more extended period of exosome secretion, allowing for more effective delivery of exosome-encapsulated molecules. These qualities have been used to advantage in treating chronic stroke (44, 45), traumatic brain injury, and in earlier work in treating Parkinson’s disease (46, 47). The ability to migrate, home in on, and contain strategic cargo, such as lethal virions or prodrug-converting enzymes, has encouraged stem cell therapy to be utilized as a Trojan horse in the search and destruction of gliomas (42, 48–50).

A disadvantage to the use of stem cells is the relative frailty of these therapeutic agents, which presents challenges to storage, transportation, reconstitution, and deployment (50). As an alternative, there has been increased interest in the use of exosomes (51–54). Exosomes are not motile, but they can diffuse through the interstitium and, with specific surface ligands, they stick to receptive cell surfaces (55, 56). The pioneering study by Alvarez-Erviti et al. (57) provided a blueprint for such targeted and exosome-dependent delivery of therapeutic short interfering RNA (siRNA) into neurons, microglia, and oligodendrocytes through intravenous administration. The exosomes were derived from dendritic cells engineered to secrete extracellular vesicles with neuron-specific rabies viral glycoprotein peptide on the surface and electroporated with therapeutic siRNAs before administration (57). Manufactured liposomes may also act in similar ways when coated with relevant ligands (58, 59). Lipid-modified ligands, such as DNA or RNA aptamers, provide a simpler strategy to modify exosome surface and thus target specificity by incorporating the ligand into the exosomal membrane (60–62). Because exosomes get into the brain by transcellular mechanisms from the bloodstream (63), no forceful opening of tight junctions is required (34). The packaging of therapeutic elements inside a lipid shell, such as an exosome, is likely to protect these elements from the enzymatic degradation that would likely occur with an unprotected bloodborne state.

Early efforts at stem cell and exosome therapy have often used regenerative elements from fetal or youthful donors. The initial hope was that these harvested elements would already complement progrowth and anti-inflammatory factors that would be useful in deployment. There is increasing interest in identifying and deploying natural or synthetic nucleic acid cargo that has specific actions on the recipient. The required effect will depend on the necessities of the condition being treated; desired outcomes may range from stimulating proliferation to suppressing inflammation (64, 65) to the promotion of intracellular processes, such as autophagy (66).

Targets that are based on clinical research experience with deep brain stimulation (DBS), such as the nucleus accumbens or the subgenual cingulate, may be considered for ultrasound-facilitated delivery of exosomes to treat refractory depression (67, 68). Degenerative conditions, such as Alzheimer’s disease or Parkinson’s disease, may be treated with exosome delivery to affected structures, such as the nucleus basalis of Meynert (69) and substantia nigra (70). Exosomes from youthful donors could be delivered to the hypothalamus to reverse aging-related frailty (5).

Other small molecules can be delivered to the brain by using focused ultrasound. Bosutinib, which has relatively poor intrinsic brain penetration, has been used along with ultrasound to treat Alzheimer’s disease and Parkinson’s dementia (71). Lipids generally fare better than hydrophilic substances in terms of brain penetration (72). Nevertheless, facilitated delivery to areas at risk for inflammation and degeneration may be targeted effectively, as in the delivery of plasmalogen precursors (73, 74) to treat Alzheimer’s disease and Parkinson’s disease. The anti-inflammatory effects of the docosahexaenoic acid (DHA) component of the plasmalogen precursor or the coadministration of curcumin extracts may be considered for targeting the nucleus accumbens of the subgenual cingulate in treating depression. Peptides, such as TREK-1 inhibitors, have been considered as treatments for mood disorders (75, 76). The short half-lives of these agents, due in part to renal clearance, limit their clinical applications. Facilitated delivery of peptides with focused ultrasound may be one strategy that may circumvent the limitations of short plasma half-lives (77–80).

Notably, the effects of ultrasound go beyond circumvention of the BBB. Ultrasound appears to produce mechanical effects on glial cells and neurons; mechanical transduction changes the biochemistry and physiology of nuclear, cytosolic, and membrane components, with potentially long-lasting impacts on targeted cells (19, 20, 81, 82). The potential use of ultrasound as a direct stand-alone therapy is the subject of other work.

For many years, mannitol infusions have been used to shrink endothelial cells (83, 84), thereby opening spaces between them to allow for entry of small (85) and large (86, 87) therapeutic agents (88) into the brain. There is little control over targeting with this technique, unless by combining with real-time monitoring of injection via advanced MRI (89). Nasal insufflation has also been proposed (90–92). Nasal application presents an opportunity for rapid systemic uptake, with mucosal vascularity making this an ideal solution for drug delivery among children and those with poor intravenous access. Depending on the insufflation delivery device, some product may be delivered to the underside of the cribriform plate (93, 94) at the apex of the nasal cavity; numerous devices have been designed to allow targeted delivery to the olfactory mucosa (95). At this location, potential cerebrospinal fluid entry is available; however, with age and even more so in those with anosmia, the nose-to-brain pathways are potentially compromised. Uptake may be through olfactory nerves and trigeminal afferents; assuming that the latter pathways may be questionably patent among elderly individuals, product will likely be delivered off target to traverse the orbitofrontal cortex and brain stem, respectively. Pharmacokinetic studies have generally demonstrated very low penetration of substances through these paths. Producing carrier systems with improved penetrance is an area of active research (96).

In the meantime, there are other targetable energy sources that may play a role in the delivery of biologicals. Transcranial magnetic stimulation (TMS) is used extensively as a stand-alone therapy for depression (97–99) and obsessive-compulsive disorder (100, 101). After stimulation, blood flow to the targeted area is temporarily increased (102, 103), allowing for facilitated delivery of biologicals. A recent study (104) found successful delivery of tracers across the BBB with high amplitude repetitive transcranial magnetic stimulation (rTMS), in which large areas are potentially treated with rTMS; however, precision deployment and depth penetration beyond 3 cm are severely limited, making this energy source problematic for access to deep nuclear groups. Laser therapy at 1,024 nm has better potential penetration (down to more than 7 cm), and it is free of scalp stimulation that further limits rTMS deployment (105, 106). The use of multiple probes may enhance physiological changes and focal effects in deeper sites. Research relating to probe design and optimization of laser dosage is needed to understand the extent laser therapy may have in the future of delivery of biologicals to the brain (107). Microwave sources may also be considered for widespread and superficial delivery, but relevant research on this application is in its infancy (108, 109). There is potential in the combination of technologies, such as microwave and near infrared, for controlled drug delivery, which may be useful in getting medicine across the BBB (108). More research is needed to investigate the use of TMS, laser therapy, and microwave therapy in delivery of biologicals.

As previously stated, the success of therapeutics relies heavily on the accuracy of targeted delivery to the incipient region. By improving permeability of the BBB, ultrasound and other energy mechanisms offer the ability to improve the delivery of drugs and molecules. However, the accuracy of this delivery is directly proportional to the spatial resolution and precision of the method used to target the energy device. Until recently, TMS was conducted in the United Stated without any neuroimaging (97, 99, 100). Only recently has structural and functional neuronavigation been applied to TMS and other neuromodulation treatments (110–112), and the clinical outcomes have improved in direct accordance with the improved targeting of the brain structures (98, 105, 113). As this field continues to evolve, the use of multimodal neuroimaging to ensure accurate and successful targeting of specific brain regions will be of paramount importance to ensuring both safety and clinical outcomes. With the ability to target structural and functional regions with high specificity comes the ability to choose clinically relevant targets. Much of DBS requires a specific nucleus as the target; neuronavigated drug and molecule delivery will allow for small brain structures and nuclei to be identified as targets for therapeutic delivery. The targeting of these structures is the subject of another article within this issue.

The application of biological therapy to neuropsychiatric illness will likely result in an expansion of the therapeutic armamentarium. Several challenges to the use of biologicals include high cost, variable brain penetration, and uncertain target delivery. Focused ultrasound and other targetable energy sources promise to meet these challenges. More efficient delivery would reduce cost by enabling the potential use of lower systemic doses. Brain penetration may be improved by local enhancement of blood flow, facilitation of capillary adhesion, and stimulation of transcellular transport mechanisms. Specific target delivery is enhanced by aiming the energy source at network nodes that appear to be part of the relevant pathophysiology.