Arthur L Horwich MD
Sterling Professor of Genetics and Professor of Pediatrics; and Investigator HHMI
Chaperones in protein folding; ALS (Lou Gehrig's Disease)
SOD1-linked ALS mice and primary cultures derived from them (at E14). Projects are designed to address early toxic alterations produced by the mutant protein, using morphologic, electrophysiologic, and ultimately molecular biochemical techniques. A range of techniques including EM, mass spectrometry, fluorescent imaging procedures, and electrophysiology measurements (the latter in collaboration with the McCormick group), are being employed.
We are focused on misfolding caused by mutant forms of the cytosolic enzyme superoxide dismutase (SOD1), that produces an inherited form of ALS (Lou Gehrig’s Disease), with progressive, fatal motor neuron dysfunction. We are using mice overexpressing a mutant G85R SOD1-YFP fusion protein to study the mechanism of disease causation. Notably, other forms of ALS, including both inherited and non-inherited forms in humans, are indistinguishable at a clinical level. The mouse model studied affords one of the most powerful approaches to following the development of this non-treatable neurodegenerative condition. The hope is that basic understanding may lead to directed therapy.
Extensive Research Description
Studies of the past decade have shown that many diseases ofneurodegeneration are the result of protein misfolding, and we havebegun to seek an understanding of the mechanism of such degeneration. We have focused on misfolding caused by mutant forms of theanti-oxidant cytosolic enzyme SOD1 (superoxide dismutase), that producean inherited form of ALS (Lou Gehrig’s disease), with progressive,fatal motor neuron dysfunction. We are using mice expressing mutant G85R SOD1-YFP, containing a mutant version of SOD1 unable to reach the native state, studying the transgenic animals with a variety of approaches, to investigate how the mutant SOD1 produces motor neuron dysfunction.
We have identified that there is a progression of misbehavior of mutant SOD1 protein itself in the spinal cord of transgenic animals. Initially it is soluble and associated to a significant extent with the abundant cytosolic chaperone Hsc70 (by contrast the wild-type protein does not form such association, presumably because it occupies the native state). Subsequently the mutant SOD1 protein begins to form both soluble oligomers (observable by gel filtration chromatography) and insoluble aggregates and at this point an additional molecular chaperone, Hsp110 becomes associated with the soluble form. At this point the animal exhibits mild lower extremity symptoms (weakness or pulling in when held by the tail) which progress over the next two to three months to lethal paralysis, usually commencing in the lower extremities. We are currently seeking to understand ultrastructural correlates of this progression using EM, analyzing motor neuron cell bodies in the spinal cord, sciatic nerve axons, and neuromuscular junctions. Where do abnormalities first present? What is the progression? We have used laser capture microdissection of motor neurons and RNA-seq from animals at mid-course in their progression to inspect for changes in transcriptional expression, but find only a small number of differences with wild-type SOD1-YFP motor neurons. Notably, there is no heat shock response, no unfolded protein (ER) response, but there are effects on several calcium binding proteins, and the level of Hsp110 mRNA is increased. We surmise that most of the toxicity of mutant SOD1 is exerted post-transcriptionally, contrasting with pathogenesis by other proteins implicated in ALS, such as TDP43 and FUS. At the post-translational level, using affinity capture and mass spectrometry, we find a variety of associations of the mutant protein in spinal cord, the most prominent being with the molecular chaperone, Hsc70. It remains to be seen whether other lower affinity interactions, that have gone undetected by affinity capture and MS, are crucial to pathogenesis.
In the mouse system we have also recently been able, in collaboration with David McCormick's group to carry out electrophysiologic measurements of MNs by patch clamping them in spinal cord of wild-type and mutant SOD1 mice in slice preparations prepared by MSTP student Muhamed Hadzipasic in our lab. This has identified a vulnerable motor neuron that dies by the time animals are symptomatic. We are further characterizing this neuronal type as well as conducting additional circuit-related experiments to analyze how the motor sytem adapts to loss of a motor neuron firing type.
We are also studying a heterologous system that reports on mutant SOD1-YFP toxicity.
Our ability to produce and purify the mutant SOD1-YFP protein from E.coli (as well as the wild-type protein) has enabled testing in other systems. One such system involves the axoplasm from the giant axon of the squid Loligo pealei, a system in which the transport of vesicles can be examined using videomicroscopy. We observe that adding mutant SOD1-YFP (but not wild-type) produces marked slowing of anterograde but not retrograde vesicular traffic. This occurs when monomeric forms of the mutant protein are added, but also when oligomeric species, first crosslinked and then purified by gel filtration, are added. At the same time as the mutant protein produces slowing of anterograde fast axonal transport, we observe that the MAP kinase cascade is activated by the mutant protein, involving ASK1 (MAPKKK) and p38 (MAPK). Remarkably, addition of molecular chaperones, and most potently, Hsp110, restores anterograde transport to normal and abrogates activation of the kinase cascade. In affinity capture studies, we observe that Hsp110 physically associates with the mutant protein. Thus it appears that there is a pathway in which the misfolded protein, if not counteracted by molecular chaperones, is able to activate a kinase cascade that inhibits anterograde transport, most likely via the phosphorylation of kinesin. We are now seeking to understand how the misfolded protein “links” to the kinase cascade. For example, does it directly interact with ASK1, or with another component that does so?
In further tests we are studying neurotransmission in the squid giant synapse, an axonal-axonal connection, injecting mutant SOD1 protein presynaptically and testing effects on neurotransmission.