PerspectiveOPTOGENETICS

Optogenetics and Translational Medicine

See allHide authors and affiliations

Science Translational Medicine  20 Mar 2013:
Vol. 5, Issue 177, pp. 177ps5
DOI: 10.1126/scitranslmed.3003101

Abstract

Optogenetic tools enable light-mediated control of cellular excitability and signaling in vivo. By manipulating biological processes, scientists can determine the roles played by these processes in intact biological systems, such as the brain. Such cellular-level control has greatly affected basic science. Here, we discuss how optogenetic tools might be translated into clinical impact through identification of new molecular and circuit-level targets and provide temporally precise interventions for defined biochemical or cellular events.

TOOLS FOR TARGETED CELLULAR CONTROL

Optogenetic tools are genetically encoded molecular reagents that, when expressed in targeted cells within an intact organism, allow a specific biological process to be controlled by light in a temporally precise fashion. The optogenetic toolbox is functionally rich, rapidly expanding, and increasingly modular, having so far enabled optical control over electrical excitability (Fig. 1A) (14), cytoskeletal remodeling (Fig. 1B) (5), G protein–coupled receptor (GPCR) signaling (Fig. 1C) (68), protein-protein interactions (Fig. 1D) and downstream effects like gene expression (9, 10), and other physiological processes in genetically targeted cells in vivo in living model organisms and in human cells in vitro. Optogenetic tools are being used to determine how signaling pathways and cell types contribute to biological functions in complex systems such as the brain.

Fig. 1.

Optogenetic molecular tools. Yellow dot indicates light-recipient protein domain. (A) Microbial opsins are seven-transmembrane proteins that translocate positively or negatively charged ions into or out of cells upon illumination. (B) Enzymes can be engineered to be light-gated by appending a light-actuated module to a protein. For example, a fusion between Rac1 and a LOV domain [photoactivatable Rac1 (PA-Rac1)] enables light control of cytoskeletal conformation. (C) Light-gated signaling pathways can enable complex physiology to be controlled by light. For example, Gq-coupled melanopsin enables optical control of G protein–signaling pathways. (D) Light-driven protein dimerization can bring cofactors in close proximity to one another to drive signaling.

CREDIT: Y. HAMMOND/SCIENCE TRANSLATIONAL MEDICINE

Each optogenetic tool is a protein that contains a photoreceptive domain that is coupled to a biological function. These proteins are either naturally occurring or are engineered by fusing a selected effector domain to a photoreceptive domain. As an exemplar of the former case, one class of optogenetic tools, the naturally occurring microbial opsins used by many organisms for photosensory and photosynthetic purposes, enables depolarization or hyperpolarization of excitable cells expressing the opsin (Fig. 1A) (11). Channelrhodopsins, which are light-gated cation channels, can be expressed in targeted neurons so that they can be depolarized by light (1), whereas halorhodopsins, light-driven inward chloride pumps (2, 3), and archaerhodopsins, light-driven outward proton pumps (4), can mediate hyperpolarization of neurons by light. These molecules are widely used in neuroscience in species from Caenorhabditis elegans to primates, revealing how specific cells contribute to normal or pathological brain states and functions (12). Use of the technologies has spread in part because of commoditization of the tools—for example, through widespread dissemination of DNA and viruses through services like Addgene and creation of simple wireless light delivery devices (13)—as well as a diversity of easy-to-use transgenic animals. The surge of excitement about optogenetics in the neurosciences has largely been driven by basic-science needs for new ways to analyze brain circuits. However, many applications to disorders such as those of metabolism (6) are emerging and point toward clinical value in systems other than the brain.

In the case of fusion proteins combining an effector domain and a photoreceptive domain, many use the light-oxygen-voltage (LOV) domain as the photoreceptive domain, so that the LOV domain modulates the function of the effector domain upon illumination [for example, the pathway driving cytoskeleton remodeling mentioned earlier (Fig. 1B) (5)]. Such allosteric modulation of protein function suggests that optogenetics might enable biochemical control over virtually any signaling pathway, yielding collectively a toolbox of “synthetic physiology” reagents, analogous to the heavily genetic toolbox of synthetic biology (14). In general, for an optogenetic tool to be simple to use, chemical cofactors required for the photoreceptive domain to respond to light should be naturally produced by the body (all-trans-retinal for opsins, flavin for the LOV domain). These LOV domain–containing tools have begun to be tested in vivo—for example, enabling optical control of gene expression in living mice (15).

IN VIVO SCREENS FOR THERAPEUTIC TARGETS

Optogenetic tools are beginning to be used to identify clinically relevant therapeutic targets, either for drugs or for electrical stimulation. The brain contains cells that differ in molecular composition, in connectivity pattern, and in how they change to cause disease states. Accordingly, there has been a long-standing desire to be able to activate and shut down different sets of neurons in order to see how they contribute to a given disease state or its remedy, revealing the mechanisms through which complex brain disorders arise. Using microbial opsins (Fig. 1A) and devices optimized for delivery of light into the brain, including wireless light delivery devices (Fig. 2A) (13) and multisite illumination devices (Fig. 2B) (16), investigators are probing how specific cell types, projection pathways, and regions contribute to Parkinson’s disease, epilepsy, and other human brain disorders (17, 18), most commonly in rodent models. By assessing whether the activation or silencing of certain neurons remedies a disease state in an animal model (assessed through biomarkers or behavior), these methods can reveal new cellular- or circuit-level targets for treatment of human brain disorders. Optogenetic activation or silencing of a given cellular or circuit target might also be useful to screen for side effects of modulating that target, measured behaviorally or in terms of neural activity patterns in cells or regions of the brain, perhaps measured by using functional magnetic resonance imaging (fMRI) (19).

Fig. 2.

Optogenetic hardware. (A) Wirelessly powered and controlled prototype device that activates a LED on the bottom of the device, with optical fibers or waveguides optionally attached to the LED. This type of device has been used to control mouse behavior (13). (B) Three-dimensional (3D) microfabricated multiwaveguide device capable of delivering light to a 3D distributed set of points in the brain (16).

Once a cellular or pathway target is identified in an in vivo screen, it can be therapeutically acted on in multiple ways. For example, one might develop a strategy for electrically stimulating the target via an implanted electrode. Another opportunity for optogenetics is to discover molecular targets that are preferentially associated with the identified cellular or circuit target. For instance, it is increasingly possible to identify the molecular contents of a cellular target by using novel technologies (20, 21); this could yield new drug targets that are specific to the cells that most contribute to a disease, and the resultant drugs may be more efficacious in remedying a brain disorder state.

HUMAN PHYSIOLOGY–TARGETED INTERFACES

Electrical and drug treatments may not be able to fix the subtle cellular- or systems-level changes that cause an intractable disorder. With optogenetics, one can genetically target a cell population to achieve a specific physiological effect with the temporal resolution of an electrical device. Because most drugs are specifically designed to have long pharmacokinetic properties, they effectively have limited temporal resolution of hours to months (for example, once or twice daily dosing to depot treatment). Active electrical devices can stimulate and inhibit circuits, but the ability to source current offers limited direct biochemical resolution over specific conductances or signaling events. It is tantalizing to envisage what new clinical opportunities exist by device-based modulation of kinase or GPCR activity, for example, because no such possibilities are in current clinical usage.

It is worth exploring what would make a disease a candidate for treatment with optogenetics. There must be a clear rationale or evidence for the importance of a specific cell type, circuit, or pathway to be activated or silenced in remedying a disorder state. In the absence of such a rationale, optogenetics might be considered unnecessarily complex relative to other treatment modalities. There also should not be a competing approach that can accomplish the same thing, without necessitating gene therapy, within a reasonable time scale. To these ends, it is possible that controlled release of hormones or peptides from cells delivered into the body, the treatment of blindness, and remedying spinal cord injury (22), among others, might be targets for direct optogenetic therapy at this time. Cell types that may be important for disorders such as epilepsy and Parkinson’s are also being probed with optogenetics (17, 18), which may suggest future targets for treating these disorders via optogenetic therapy.

Optogenetic therapy in humans will require gene therapy and thus will require a vector to deliver in vivo the gene that encodes for the light-activated protein as well as an implantable or external light-delivery device. In a companion article in this issue of Science Translational Medicine, Williams and Denison discuss such practical barriers and the regulatory and engineering challenges associated with them (23). There are no U.S. Food and Drug Administration (FDA)–approved viral gene therapies, although several recent phase II and phase III clinical trials have used viruses such as adeno-associated viruses (AAVs) to successfully transduce cells in the central nervous system or other parts of the body directly [for example, for Leber congenital amaurosis and lipoprotein lipase deficiency (LLD), ClinicalTrials.gov identifiers NCT00999609 and NCT01109498, respectively]. Recently in a landmark decision, the European Medicines Agency approved the AAV-mediated gene therapy Glybera for LLD. Many of the aforementioned optogenetic tools are small enough to fit in AAV vectors, which hold ~4.5 kb of DNA. Newer serotypes of AAV might avoid immune responses seen for wild-type vectors and also improve specificity for targeting cell types.

Past gene therapy promoters have been strong, non–cell-specific promoters, whereas optogenetic promoters are commonly cell type–specific and tuned to yield the appropriate level of expression—in other words, appropriate to mimic a natural signaling pathway or to correct an aberrant signaling pathway. Few promoters have been published that are both highly cell type–specific and small enough to fit in an AAV. The use of lentiviruses may offer a route to delivering larger genetic payloads (~10 kb), albeit with fewer documented safety studies; however, lentiviruses have been used in recent clinical success stories in cell therapies for β-thalassemia (24) and chronic lymphocytic leukemia (ClinicalTrials.gov number NCT01029366). Alternatively, it may be useful to have drug-regulated promoters, such as antibiotic-driven promoters common in biology, that enable gene expression to be precisely tuned—or completely shut off, should undesirable effects occur in people. Importantly, although the use of new promoter-vector combinations delays clinical trial initiation owing to the toxicology studies required, they will be critical to the overall clinical progress and success of optogenetics.

Because many optogenetic tools come from species such as bacteria and algae, there are concerns that they may be detected as foreign by the human immune system or may be otherwise toxic to human cells, especially over the long time scales (many years) over which gene therapies are expected to operate. Adverse events that take a long time to develop may be difficult to assess in animal experiments. Optogenetic activation of cortical neurons in nonhuman primates has demonstrated the ability to alter awake-primate behavior (25). But no optogenetics experiment has looked ahead more than a few years in animals. Mammalian opsins, such as human opsins (melanopsins and rhodopsins), are slower in their light responses than are the microbial opsins but would avoid the need for trans-species gene therapy. Mammalian opsins have been tested in animal models of blindness and diabetes, achieving meaningful improvements, and in studying the vertebrate spinal cord (6, 8, 26). Furthermore, most optogenetic tools are fused to genetically encoded fluorophores for ease of visualization; various fluorophores can increase or decrease the level of microbial opsin expression in unpredictable ways, and their safety should not be taken for granted.

In addition to gene therapy, for many optogenetic therapies deep light delivery may also be required for use in humans. Wireless and multisite light delivery devices that contain light sources, such as light-emitting diodes (LEDs), and optical fibers (to convey light to deep targets) would minimize the need for connectors and wiring and increase the diversity of neural patterns that can be introduced into the brain ( Fig. 2 ). On the electronics side, it will be important to have medical devices that can deliver light to deep targets with a minimum of heating and side effects, with long life span and good efficacy (23).

CASE EXAMPLE: RETINITIS PIGMENTOSA

Retinitis pigmentosa (RP) fits several of the criteria outlined above as a candidate for optogenetics therapy. RP is an inherited retinal degeneration disorder that affects roughly 100,000 patients in the United States, and patients are currently underserved by available treatment options, such as large doses of vitamin A. As a rare genetic disorder, gene therapies are often considered, but because there are more than 100 mutations in RP, a traditional gene therapy via direct protein replacement will only help a small number of patients. What is needed is a therapy that treats multiple mutations in a broad-spectrum manner. Several optogenetic approaches have been reported recently for RP. By optically depolarizing so-called ON bipolar cells or retinal ganglion cells with an optogenetic reagent such as a channelrhodopsin (Fig. 1A), the retina can again capture light and transform it into signals the brain can comprehend (27). Another approach is to hyperpolarize cones that are dysfunctional but still alive by using a silencing optogenetic reagent (27) in order to enable the cones to be resensitized to light. It is important to note that these approaches to treating RP are complementary because they will target different patient populations, depending on the stage of degeneration.

The case example of RP highlights many translational practicalities for optogenetic approaches. The light delivery could be noninvasive, if optoelectronic technologies, such as goggles that deliver light to the eye, were implemented. This is an important consideration because current optogenetic tools do not have anywhere near the dynamic range of the natural human eye. Also, there are efficient viral vectors to deliver the transgene for RP, which are known to be safe in the retina (27). Of course, the opportunities in ocular gene therapy are not limited to RP, with dry age-related macular degeneration and Stargardt Disease being a possibility. However, key questions remain, beyond the obvious ones regarding safety and efficacy. Are the current opsins, promoters, and optics sufficient, and if not, what clinically driven molecular engineering efforts are required? What will be the clinical end points? How will such a treatment be regulated? Will the therapeutic effects of a one-time transformative therapy be reduced by retinal remodeling accompanying photoreceptor degeneration? With several teams having voiced their interest in taking RP optogenetics therapies into humans, these questions will need to be answered sooner rather than later.

REMOTE-CONTROLLED BIOLOGICS

A therapeutic approach with a potentially less daunting safety hurdle than RP is the prospect of dynamically controlled cell therapy, in which a cell is made sensitive to light ex vivo and then transplanted into the body. For example, Ye et al. engineered a cell containing the light-sensitive opsin melanopsin and a melanopsin-signaling–dependent gene expression cassette encoding for glucagon-like peptide 1, a therapeutic peptide for type 2 diabetes treatment. Upon cell implantation in mice and transdermal illumination (6), mice experienced decreased glucose levels. Whereas no somatic gene therapy has yet to be approved by the FDA, there are currently eight FDA-approved cell therapies (www.fda.gov/BiologicsBloodVaccines/CellularGeneTherapyProducts/ApprovedProducts/default.htm). Such transplanted cells could be engineered to contain specific promoters (not constrained by the size of viruses) and to be equipped to respond to (or be destroyed by) a small-molecule drug.

One could imagine an easier regulatory path of delivering purified optogenetic proteins into the body (more than 130 FDA-approved protein therapies exist) and then illuminating them at a given point in time, at a defined site in vivo. This would be a dynamically titratable protein therapy—a biologic version of photodynamic therapy, if you will. This could enable interfacing to specific signaling pathways, albeit without genetic cell targetability.

INSPIRING NEW TECHNOLOGIES

The road to the clinic is slow. Innovation moves fast. Thus, another potential clinical impact will result from optogenetic tools inspiring new ways of coupling energy into intact biological systems for on-demand modulation of biochemical events and signaling pathways. One recent study presented a genetically encoded mechanism for cellular activation using a radio frequency (RF)–sensitive molecule, ferritin (28). RF activation of ferritin caused local heating, which activated a temperature-sensitive ion channel, thus allowing calcium influx that could drive downstream signaling. Thus, instead of light, RF energy can be noninvasively delivered to a genetically targeted cell, causing it to respond. Compared with visual spectrum irradiation, RF or other forms of energy, such as ultrasound, may enable tissue to be addressed noninvasively in deeper and larger volumes while still focal and temporally precise to some degree (millimeter to centimeter voxels, second to minute time precision), which might be appropriate for diseases that do not require ultraprecise spatial or temporal control (such as modulating metabolic disorders).

The clinical impact of optogenetics may thus be indirect as well as direct: It may inspire new technology platforms that beget biochemically specific and spatiotemporally precise therapeutic interventions, should we imagine them. By catalyzing an open discussion between optogenetic engineers, disease biologists, clinicians, industry leaders, and regulators, we are optimistic that closed-loop innovation will help reveal translational paths for this exciting new technology.

References and Notes

View Abstract

Navigate This Article