Small Molecule Inducers of Autophagy in

The J. David Gladstone Institutes, University of , San INVENTORS Francisco Gladstone Center for Translational Research Steven Finkbeiner 1650 Owens Street, Andrey Tsvetkov , CA 94158 Jason Miller [email protected] Tel. (415) 734-2018; Fax. (415) 355-0930

Gladstone Docket No TBD UMB Docket No TBD

SUMMARY Autophagy is a critical intracellular protein turnover pathway that may have special relevance to human diseases characterized by the accumulation of misfolded proteins. The ubiquitin proteasome system, the other major protein turnover pathway, requires proteins to be unfolded monomers prior to degradation. By contrast, autophagy can, in principal, degrade aggregated misfolded proteins, organelles, and even parasites. We have developed novel methods to study autophagy in primary neurons and found that neurons appear to govern autophagy in ways that partly differ from non-neuronal cells. During these studies, we have discovered potent and novel small molecule inducers of autophagy in neurons with structure-activity relationships that suggest a novel pharmacophore.

APPLICATIONS ♦ Novel methods of treating neurological diseases and neurological manifestations of human CNS diseases, ranging from Alzheimer’s disease to Fragile X mental retardation.

ADVANTAGES ♦ The invention relates to small molecule inducers of autophagy. Small molecules have the advantage of the potential for high receptor activity, target enzyme selectivity, cell permeability, blood-brain barrier permeability, not eliciting immune responses, oral bioavailability, enhanced metabolic stability and the capacity for relatively cost effective large-scale manufacturing. ♦ The autophagy inducers of the invention induce the formation of autophagosomes in primary neurons, are relatively non-cytotoxic, and mitigate toxicity in primary models of polyglutamine-dependent neurodegenerative diseases. Several are known to be well tolerated after chronic oral administration and a few are FDA-approved drugs for other neuropsychiatric conditions, indicating that they are able to enter the brain.

BACKGROUND OF INVENTION In healthy cells, two major pathways are responsible for the degradation, disposal, or recycling of cellular proteins and other contents. Short-lived proteins are degraded by the ubiquitin- proteasome system (Pickart, 2001). Autophagy removes long-lived proteins and is the only pathway for degrading organelles and parasites (Levine, 2005).

Non Confidential Disclosure 17th July 2009 Three major types of autophagy occur in mammalian cells. In macroautophagy, cytoplasmic content or organelles are sequestered in autophagosomes, which fuse with lysosomes for degradation and reuse in essential cellular processes. In microautophagy, cytosolic contents are sequestered by invagination of the lysosomal membrane. In chaperone-mediated autophagy (CMA), proteins are recognized by a specific cytosolic chaperone and a receptor on the lysosomal membrane, which mediates protein translocation into the lysosomal lumen (Cuervo, 2004). Although the proteasome system and CMA require proteins to be unfolded monomers before degradation, macroautophagy (hereafter referred to as autophagy) can, in principle, engulf and degrade aggregated protein.

Autophagy has been characterized mostly in yeast cells and mammalian cell lines (Klionsky et al., 2007). However, Mizushima et al. (2004) suggested that mechanisms of autophagy differ in neuronal and other cells. They noticed that starvation, the best-known inducer of autophagy in most cells, does not induce autophagy in cortical neurons of starving mice even after 48 h. However, it was not clear whether this finding reflected protection of the brain from starvation by physiologic coping mechanisms or differences in the regulation of autophagy in neurons. Consistent with the latter possibility, the activity of mammalian target of rapamycin (mTOR), a kinase implicated in the majority of autophagic pathways in most cells (Ravikumar and Rubinsztein 2006), is not affected in most parts of the starving brain (Cota et al., 2006).

Recently, Komatsu et al. (2006) demonstrated that mice deficient in autophagy develop symptoms reminiscent of neurodegenerative diseases. A role for autophagy in is supported by findings in models of Parkinson’s disease (PD) and Huntington’s disease (HD). α-Synuclein, a pathogenic factor in some forms of PD, is degraded by the ubiquitin-proteasome system and CMA (Cuervo et al., 2004)—both of which are inhibited by aggregated α-synuclein, resulting in upregulation of macroautophagy, clearance of the aggregated protein, and better cell survival. Autophagosomes have also been observed in lymphoblasts from HD patients, and mutant huntingtin expression has been associated with an increase in endosomal-lysosomal activity and autophagy (Nagata et al., 2004; Kegel et al., 2000).

Despite its possible involvement in neurodegeneration, autophagy has been studied mainly in non-neuronal cells and cell lines, and it is not known whether the mechanisms that govern its induction and regulation in neurons are the same as in other cell types. Moreover, the regulation of autophagy in mammalian cells may be more complex than previously appreciated (Yamamoto et al., 2006).

STAGE OF DEVELOPMENT The program has successfully developed methods to monitor autophagy in primary neurons and has discovered important differences in the way that neurons respond to treatments that commonly induce autophagy in non-neuronal cells. These studies have led to the discovery of novel small molecules that reliably induce autophagosome formation in primary neurons.

We have developed five approaches for evaluating the status of the autophagic pathway in primary neurons. We have developed custom antibodies against microtubule-associated protein 1 light chain (LC3) that have enabled us to reliably distinguish in neuronal extracts lipidated (LC3- II) forms of LC3 from unlipidated forms (LC3-I). The ratio of LC3-II to LC3-I is commonly

Non Confidential Disclosure 17th July 2009 used as an assay of the accumulation of autophagosomes, which can indicate an induction of autophagy (Tsvetkov et al., 2009).

In addition, we have cultured primary neurons from mice that carry a transgene encoding GFP- LC3. GFP-LC3 is distributed diffusely throughout the cytoplasm of neurons at baseline. Upon induction of autophagy, GFP-LC3 forms punctae in the soma that can be quantified objectively from images with our automated microscopy and analysis techniques (Arrasate and Finkbeiner, 2005; Arrasate et al., 2004; Tsvetkov et al., 2007; Tsvetkov et al., 2009). As such, the assay and microscopy system forms a suitable platform for HTS for autophagy regulators in primary neurons. We have also confirmed our findings by performing electron microscopic assays for autophagy in primary neurons.

Finally, we have developed a new cell-based assay that harnesses and adapts novel features of new classes of fluorescent protein technology and combines them with the unique capabilities of our robotic microscopy platform (US 7,139,415) to create the first system that can measure flux through the autophagic pathway. Versions of the screening system have been created to specifically measure flux through the macroautophagy pathway and to specifically measure mitophagy. An assay focusing on flux through the autophagic pathway rather than snap shots of steady state intermediates is much less variable and better suited to HTS. As such, this new platform would be suitable to conduct an industry level small molecule discovery campaign for modulators of autophagy in neurons and other cell types.

With these assays in hand, we have assessed the ability of some commonly used treatments to induce autophagy in primary neurons and non-neuronal cells. To our surprise, a number of treatments and small molecules readily induced autophagy in non-neuronal cells but failed to do so in neurons (Tsvetkov et al., 2009). For example, rapamycin and everolemus, an orally available and more stable analog from Novartis, each demonstrably inhibited their intended target, mTOR, in neurons. However, both drugs failed to induce autophagy in neurons despite doing so in non-neuronal cells. Similarly, starvation, lithium chloride, and ceramide all potently induced autophagy in non-neuronal cells but were ineffective in neurons. We conclude that there are important differences in the pathways that mediate the ability of extracellular stimuli to govern autophagy in neurons.

During the course of these experiments and screening a series of small molecules, we discovered some that potently induce autophagy in primary neurons. The first, which we call AT1, is described as an Akt inhibitor. We found that nanomolar to low micromolar doses induce autophagy within minutes to hours after application. We have followed neurons treated with this small molecule for several days, and at doses that induce authophagy, we observe no overt toxicity. We have tested the drug in primary neuron models of two polyglutamine diseases— Huntington’s disease and spinomuscular atrophy—and we observe that the addition of drug significantly mitigates neurodegeneration.

We don’t think that this drug induces autophagy by inhibiting Akt. We have tested a series of structurally unrelated putative Akt inhibitors and, none induces autophagy despite demonstrable inhibition of Akt. We hypothesized that AT1 might be acting on a different cellular target to induce autophagy, so we tested scaffolds with related structures. In total, we have tested 14

Non Confidential Disclosure 17th July 2009 related structures and have found 11 (AT2-AT11) that induce autophagy in primary neurons with varying potency. Three of the 14, which lack key substituents, were inactive.

The structures of the 14 active and inactive compounds have been examined and similarities in ring structures, electron withdrawing substituents, tertiary amines, steric moieties, etc… have been noted. These structures have been correlated to their activity in silico and a preliminary pharmacophore has been developed. All have MW < 500. All are cell permeable and water soluble. A significant fraction are known to cross the blood brain barrier, and several are already FDA-approved drugs for other neuropsychiatric conditions.

PROPOSED R&D In the short term, we want to (1) perform additional off-the-shelf SAR to refine the pharmacophore and (2) obtain novel chemical entities based on these scaffolds. (2) We would also like to identify the molecular targets of these compounds by fluorescently tagging the most potent analogs and using biochemical approaches to identify the neuronal proteins that they bind. Longer term, we want to test the safety and efficacy of these compounds in cell and mouse models of AD, PD, ALS, and HD.

LICENSING POTENTIAL Gladstone seeks to develop and commercialize by an exclusive or non-exclusive license agreement and/or sponsored research with a company active in the area.

PATENT STATUS ♦ Invention disclosure filed.

Arrasate, M., and S. Finkbeiner. 2005. Automated microscope system for determining factors that predict neuronal fate. Proc. Natl. Acad. Sci. USA. 102:3840–3845. Arrasate, M., S. Mitra, E.S. Schweitzer, M.R. Segal, and S. Finkbeiner. 2004. Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature. 431:805–810. Tsvetkov, A., M. Arrasate, V.R. Rao, and S. Finkbeiner. 2007. Differences in autophagy in neuronal and non-neuronal cells: Relationship to a neuron model of Huntington’s disease. Soc. Neurosci. Abstr. 33:255.3. Tsvetkov, A.S., S. Mitra, and S. Finkbeiner. 2009. Protein turnover differences between neurons and other cell types. Autophagy:in press.

Non Confidential Disclosure 17th July 2009