NDC for the Optical Control of Biological Function
A major challenge for biomedicine is to develop new ways of determining how proteins operate in complexes in cells. This requires molecularly focused methods for dynamic interrogation and manipulation. An attractive approach is to use light as both input and output to probe molecular machines in cells. While there has been significant progress in optical detection of protein function, little advance has been made in remote control of any kind, including optical methods. The NDC for the Optical Control of Biological Function is developing methods for rapidly switching on and off with light the function of select proteins in cells. The strategies are broadly applicable across protein classes.
This engineering challenge has already been solved in the course of the first two funding years for two very distinct protein targets, using distinct strategies that are variations on one of two core designs, that of a Photoswitched Tetehred Ligand or “PTL”. PTLs (and the other photoswitches) are tethered (attached) in a site directed manner to the protein of interest. The site of attachment is designed into the protein to be at a precise distance from a binding site for the ligand. The geometric precision has two important consequences. First, light of two different wavelengths (colors) is used to isomerize the linker (change its shape) in such a way that the ligand can only bind in one of the shapes, thus making it possible to toggle binding on and off with light. Second, native proteins are not affected by the PTL and remain insensitive to the light. This means that a specific protein in a cell, a tissue and even in an intact freely behaving organism, can have its biochemical signaling toggled on and off by remote optical control. The switching is very fast, taking place in ~1 millisecond, i.e. at the rate of the fastest nerve impulse.
The two variant PTL schemes that have already been demonstrated are a blocker of an inhibitory K+ channel pore and an agonist (activator) of an excitatory neurotransmitter receptor. These have been demonstrated to work in a variety of systems, including cultured mammalian brain neurons, leach neurons, mammalian cerebellar brain slices, mammalian retinal ganglion cells (RGCs), and behaving zebrafish sensory and motor systems. This success to date has laid the groundwork for expanding the toolkit of chemical photoswitches to other ligands, other photoisomerizable moeities, other attachment strategies, and from ligands to force actuators that are basically Photoswitched Cross-Linkers (PCLs) that can regulate the structure of peptides that interfere with the activity of specific proteins that are implicated in certain diseases, including inflammation, cardiac ischemic disease, and certain reactive skin diseases. Moreover, we are now in a position to expand our protein target list to other classes of channels and receptors, but also to enzymes. In addition, we are now in a position to expand our target tissues from ones that are directly accessible to light (isolated tissues, retina and skin) using optical fiber endoscopes that reach deep brain structures.
The success of the retinal work has made it realistic to now pursue our first disease model: blinding diseases that result from loss of photoreceptor cells. Such diseases often spare other cells in the retina for many years. Especially resilient are the output neurons of the retina, the RGCs. Our approach is to endow the surviving cells with an ability to sense light, using our engineered light-gated ion channels. We have successfully expressed light-gated channels selectively in a subset of ON RGCs using viral delivery and cell-specific promoters. We are now in pursuit of other promoter-virus combinations that would target OFF RGCs and other subtypes of ON retinal ganglion cells. We are confident that these will be identified within the next 2 years. While that effort is being pursued, we are already testing the ability of our initial ON RGC expression system to support vision (assayed electrophysiologically in the retina and primary visual cortex and by behavior) in mutant mice and rats that have lost their photoreceptor cells. Moreover, preliminary experiments have shown that the same promoter-virus system can be used in a primate model of human vision (Macaque), which opens the door for advanced psychophysical and electrophysiological experiments on higher visual processing.
Biological and disease targets: Three central signaling systems. We are focusing on ion channels, G-protein coupled receptors and kinases and phosphatases. These represent some of the largest protein families, together comprising about 7% of the human genome. The proteins are implicated as causes and drug targets in major human diseases. By enabling rapid, reversible and local switching these tools should make it possible to overcome two major problems in standard drug delivery: a) side effects due to systemic delivery, or systemic leak of locally delivered drugs, and b) long delays between delivery and action, and difficulty to terminate action.
Toolkits of light-gated signaling proteins and of light-gated peptides that selectively modify protein function. We are using structure-based design to chemically synthesize two classes of nanoscale photoswitches: 1) ligand photoswitches that attach to native and engineered target proteins, and remote control their function (with both activating and blocking ligands); and 2) light-gated peptides, which bind to target proteins (at interaction domains or substrates binding sites) in a light-dependent manner, interfering with function only in the bound state.
Addressing specific medical problems. A critical challenge is to transfer our methods from basic research in cell biology to therapeutics. We are focusing on animal models of human disease, including using light-gated channels to restore vision to the retina after loss of photoreceptor cells, treatment of neovascular retinal disease, prevention of cardiac damage due to ischemia, and certain models of inflammation and skin disease.
Progress in FY02. In FY02 we made progress in several areas: a) the fundamental properties of chemical photoswitches and target proteins, b) new classes of photoswitches, c) novel orthogonal methods of probe attachment to proteins, d) design of optical instrumentation for delivering light in vivo and in vitro, e) in vivo models of retinal disease and cardiac ischemia, f) delivery of light-gated channels to retinal ganglion cell subtypes in rodent and Macaque, g) use of light-gated channels to control neuronal activity and behavior in vivo in zebrafish, and h) pilot studies on photoswitches that can be triggered with 2-photon illumination.
Pathway to Medicine
Our most advanced disease model is retinal degeneration caused by gene defects in photoreceptor cells, which result in photoreceptor cell death by apoptosis. The heterogeneous nature of retinal disease makes the development of gene replacement therapies very challenging, as it currently appears that each genetic defect will require a specific therapy. Pharmacological and neuroprotective strategies are more broadly applicable, acting to postpone cell death in patients with surviving photoreceptors. Unfortunately, many individuals are not candidates for such treatments as their photoreceptor cells have died. In our NDC module, we proposed an alternative strategy for restoration of light perception in patients who have lost vision from photoreceptor cell death – this is to confer light sensitivity to surviving inner retinal neurons using light-activated ion channels and receptors developed in the NDC center.
The concept of a retinal prosthetic initially derives from histopathology observations from donor eyes and OCT images from patients with inherited retinal degenerations suggesting that retinal interneurons and ganglion cells survive for extended periods after photoreceptors are lost. We hypothesize that there is potential for restoration of light perception and some vision in patients blind from photoreceptor loss if the remaining retinal neurons were genetically modified to add a light receptive function, allowing these cells to transmit visual information to visual cortex. Adding a light receptive function to the inner retina would, in all likelihood, not intrinsically generate high-resolution shape and form vision. Development of an external optical interface to transmit signals encoded to mimic some of the lost intrinsic retinal processing, such as center – surround and edge detection will likely be required to stimulate the cells. In addition, the high amplification of the phototransduction cascade will have to be recreated to restore sensitivity.
Clearly, the ability to reproducible controlled action potential firing in RGC’s through optical stimulation will advance the state of the field in several ways. Viral mediated transduction methods have the potential to recruit the entire RGC population (estimated at >1 million cells) to this task. This is a significant advance from the current microfabricated electrode designs, which contain a few hundred electrodes. An additional advantage of transfection of a light-activated protein is the potential of removing the retinal contact electrode itself and the connecting cables from the eyeball and orbit. The current retinal ‘chip’ designs have a significant risk of damaging the delicate retinal cells through attachment of the electrode array to the inner limiting membrane of the retina. Replacing this with an optical interface would be a major step forward. Combining transfection methods of retinal gene therapy with DLP or other high-resolution spatial light modulation technology has the potential to restore vision.
Gene therapy is particularly well-suited for the treatment of retinal disorders, with several proof of concept studies underway in patients with Leber’s Congenital Amaurosis at U. Penn and Moorfields Eye Hospital, London. A potential therapy for late stage retinal diseases in which photoreceptors have been lost, is a gene therapy focused on imparting light-sensitivity to surviving neurons, in a manner designed to emulate normal retinal circuitry. In our experience, AAV vectors provide the best combination of high efficiency, low toxicity, and stable gene transfer to dividing and nondividing cells in retina. Gene therapy is a promising long-acting treatment for some retinal disorders. In fact, it may be more promising as a therapeutic in the retina than in many systemic applications partially because the eye is readily accessible for surgical injection of gene therapy vectors. One potential gene therapy for late stage retinal diseases is to impart light-sensitivity to surviving neurons, in a manner designed to emulate normal retinal circuitry. There are several systems available for retinal gene transfer. Most are based on viruses, as nonviral systems typically provide low efficiency and transient gene expression. Adenoviral, herpesviral, and lentiviral vectors have shown promise in animal models; however, the most promising vectors are arguably those based on adeno-associated virus, AAV, our delivery system of choice.
Significance and Uniqueness
An important pharmacology advance is the development of tethered ligands, in which the targeting of a small molecule to a protein of interest depends not only on the usual single condition of the quality of fit into the binding site, but also on a second condition, namely proximity of the binding site to a covalent attachment site. This two-key approach makes it possible to obtain high specificity even when binding is weak. By using a short linker between the ligand and the protein-attachment site, the effective concentration of the ligand can be made very high, permitting even low affinity ligands to have high occupancy. On its own this approach produces chronic liganding and thus begs for a modification that would enable controlled switching. An attractive solution is the use of photoisomerizable moieties. Photochemically induced changes in the shape or electronic character of functionally important amino acids have been used to control the function of proteins in response to light or to alter the backbone structure of peptides, thereby controlling their interaction with other biological macromolecules. We propose an alternative strategy in which photoisomerization of a tethered ligand or cross-linker is used to gate protein activity. We propose new applications that are generalizable to many classes of proteins, providing astonishingly rapid, reversible, and spatially constrained remote control of protein function. We will create a unique, powerful toolset of chemical photoswitches, and switched peptides and proteins, doe use in basic cell and tissue biology research. We expect that these tools will revolutionize the in vivo analysis of cell function, breaking new ground in challenging arenas such as the analysis of neural circuits and the relationship between the molecular events in specific cells and the behavior of the entire organism. The optical methods also open an avenue for cell engineering, synthetic cells, targeted “smart-cell” therapy and for high-throughput screening of small molecule and chaperone therapeutics.
Perhaps most enticing of all is our long-term goal of using light switches to control the function of individual proteins in specific cells as a novel strategy for therapy of human diseases. Optical control holds potential to circumvent problems of drug targeting by making the drug action a two – step process, which begins with the administration of the light sensitive drug. An inactive drug binds to the target proteins and once the necessary level of concentration is attained, the second step is to activate the drug with a specific dose of light of a particular wavelength, ensuring that the reagent acts only in a desired location that is pinpointed by light.
Evolution of Project
In the 2 years since we launched the NDC our goals have reoriented us ard in vivo applications and disease models, with a reduced emphasis on cell-based assays and synthetic signaling systems. We have also made advances in light delivery to superficial and deep tissue that now narrow our originally broad palette of approaches. These advances bring us to consider some alterations to the investigator group. Several promising new collaborations have been initiated with non-NDC investigators, who should become members of the NDC during next year’s scale-up. At the same time, 3 investigators have become more peripheral to the projects central goals and will either reorient their efforts or be dropped from the NDC (see Section 4).