Novel Ubiquitin Specific Protease-13 Inhibitors Alleviate Neurodegenerative Pathology

Ubiquitin Specific Protease-13 (USP13) promotes protein de-ubiquitination and is poorly understood in neurodegeneration. USP13 is upregulated in Alzheimer’s disease (AD) and Parkinson’s disease (PD), and USP13 knockdown via shRNA reduces neurotoxic proteins and increases proteasome activity in models of neurodegeneration. We synthesized novel analogues of spautin-1 which is a non-specific USP13 inhibitor but unable to penetrate the brain. Our synthesized small molecule compounds are able to enter the brain, more potently inhibit USP13, and significantly reduce alpha-synuclein levels in vivo and in vitro. USP13 inhibition in transgenic mutant alpha-synuclein (A53T) mice increased the ubiquitination of alpha-synuclein and reduced its protein levels. The data suggest that novel USP13 inhibitors improve neurodegenerative pathology via antagonism of de-ubiquitination, thus alleviating neurotoxic protein burden in neurodegenerative diseases.

Spautin-1 is currently a known specific inhibitor to USP10 and USP13 with IC 50 s of 0.6-0.7 µM [23]. It inhibits autophagy by deregulating the formation of VPS34 complex acting over USP13 and USP10 that modifies Beclin-1 ubiquitination [23]. However, it has very poor penetration to brain. There are very limited studies about spautin-1 in neurodegeneration. It was mainly used as a small molecular tool to study autophagy in cancer research. For instance, spautin-1 enhances imatinib-induced apoptosis in chronic myeloid leukemia [24]. Spautin-1 ameliorates acute pancreatitis via inhibiting impaired autophagy and alleviating calcium overload [25].
To further elucidate the role of USP13 in neurodegeneration, we developed a small library of specific small molecule inhibitors of USP13. We demonstrate that brain-penetrant USP13 inhibitors regulate (de)-ubiquitination of neurotoxic proteins and may alleviate neurodegenerative pathology. Our previous findings and the current data suggest that USP13 inhibition prevents protein de-ubiquitination, thus providing a feasible neurotherapeutic target to facilitate misfolded protein clearance. The novel library of small molecules constitutes new tools for further investigation to assess the therapeutic potential of USP13 inhibition in neurodegeneration.

Chemical Synthesis and Cell Viability Assays
Spautin-1 does not cross the blood-brain barrier (BBB) and non-specifically inhibits USP13 and USP10 and affects proteasome and autophagy pathways [23]. Therefore, we developed analogues of spautin-1 that are more potent inhibitors of USP13 and are small molecules that can enter the brain. USP-13 inhibitors bearing 6-fluoroquinoline, thieno [3,2-b] pyridine and 3-nitrocoumarin backbones were synthesized as shown in Figure 1. Nucleophilic aromatic substitution reactions using aniline (2) with either 7-chlorothieno [3,2-b]pyridine (1) or 4-chloro-6-fluoroquinoline (3) in DMSO gave BK50118-A and BK50118-B in 98% and 95% yield, respectively. Similarly, BK50118-C was obtained from 4-chloro-6fluoroquinoline (3) and benzylalcohol (4) in the presence of NaH in 97% yield. Palladium catalyzed C-N bond formation produced CL3-499 from 4-chloro-6-fluoroquinoline (3) and benzylamine (5) in 81% yield. The amino acid derivatives CL3-512 and CL3-514 were synthesized from 4-chloro-3-nitrocoumarin (6) and the methyl esters of L-phenylalanine and L-tyrosine in 90% and 42% yield, respectively, in the presence of K 2 CO 3 at room temperature. The products were purified by column chromatography and characterized by 1 H, 13 C and 19 F NMR spectroscopy as well as high resolution mass spectrometry as described in more detail in Supplemental Figures S1 and S2. We thus generated a small library of six compounds ( Figure 2A).

Novel Small Molecules Are Potent USP13 Inhibitors
We next determined whether the novel compounds inhibit USP13 activities and clear toxic proteins. Stably transfected human neuroblastoma SH-SY5Y cells with human wildtype alpha-synuclein were treated with 1 mM, 100 µM, 10 µM, 1 µM, 0.1 µM, 0.01 µM and 0.001 µM of each of BK50118-A, BK50118-B, BK50118-C, CL3-499, CL3-512 and CL3-514 dissolved in an equivalent volume of DMSO for 5 h. USP13 activity via functional ELISA showed that all of the 6 compounds inhibited USP13 ( Figure 3A-F) with maximal inhibition coefficient (IC 50 ) ranging from 0.11 nM to 2.13 nM ( Figure 3G). Measurement of human alpha-synuclein via ELISA showed that BK50118-C ( Figure 3J) most potently decreased alpha-synuclein levels, reaching almost complete alpha-synuclein clearance at 1 mM and 100 µM. The other 5 compounds were also able to lower alpha-synuclein levels but with less potency ( Figure 3H,I,K-M).
1 Figure 3. The effects of the novel small molecule compounds on the USP13 activities and alpha-synuclein in vitro. The SH-SY5Y cells were transinfected with human alpha-synuclein plasmid. Cells were treated with a serial of concentrations of each individual compound for 5 h. ELISA was performed for (A-F) USP13, and (G) estimated IC50 to USP13 was calculated. ELISA for (H-M) alpha-synuclein. ** p < 0.01, **** p < 0.0001 to DMSO group, ## p < 0.01, #### p < 0.0001 to untreated group; ordinary one-way ANOVA. N = 3-9 per treatment group.

Absorption, Distribution, Metabolism and Excretion (ADME) Studies
We focused on BK50118-C as the most potent USP13 inhibitor to determine the pharmacokinetics (PK) of this molecule. Wild type C57BL6 mice were injected with 10 mg/kg BK50118-C versus DMSO, and the brain and serum were isolated at 0, 1, 2, 4, 6, 8 and 12 h; extracted in water and examined by mass spectrometry. The concentration of BK50118-C peaked at 1 h (T max ) both in the serum and brain (Table 1), and a maximal concentration (C max ) of 81.49 nM and 354.63 nM were reached in the brain and serum, respectively. The bioavailabilities of the drug (AUC, area under the curve) were 164.3 nM×hand 599.4 nM×h in the brain and serum, respectively. Elimination (T 1/2 ) was 2.32 h for brain and 1.84 h in serum. The ratio of serum: brain reached 28%, indicating that this drug abundantly enters the brain.

BK50118-C Reduces Alpha-Synuclein, Increases Alpha-Synuclein Ubiquitination and Improves Neuronal Survival in Mice
Transgenic A53T mice harbor the arginine to threonine (A53T) mutation of human alpha-synuclein under the control of prion promoter and have abundant alpha-synuclein in the striatum as early as 3 months of age [26].
Male and female TgA53T mice (15 months old) were treated daily with intraperitoneal injection of DMSO versus 10 mg/kg or 40 mg/kg BK50118-C for 7 days. WB of midbrain lysates showed that human alpha-synuclein was significantly reduced at the above dosages ( Figure 4A,B 1st blot). Immunoprecipitation of alpha-synuclein protein from midbrain extracts followed by ubiquitin WB ( Figure 4B, 2nd blot) showed increased ubiquitination in 40 mg/kg BK50118-C compared to DMSO and 10 mg/kg BK50118-C. Immunoprecipitation of ubiquitin from midbrain extracts followed by alpha-synuclein WB ( Figure 4B, 3nd blot) showed no monomeric alpha-synuclein but increased ubiquitination of alpha-synuclein in 40 mg/kg BK50118-C compared to DMSO and 10 mg/kg BK50118-C. ELISA measurement of alpha-synuclein in the same extracts ( Figure 4C) confirmed that human alpha-synuclein was reduced by about 30% (10 mg/kg) and 56% (40 mg/kg), respectively ( Figure 4C).
TgA53T mice also demonstrated an elevated state of tauopathy in striata, suggesting that tauopathy is a common feature of synucleinopathies [27]. Therefore, we also measured tau levels. There was no effect on murine tau levels ( Figure 4D).
Immunostaining of 20µm thick brain sections showed human alpha-synuclein staining in both cortex ( Figure 4E,G,I) and striatum ( Figure 4J,L,N) in DMSO treated mice. BK50118-C treatment (40 mg/Kg) significantly decreased human alpha-synuclein staining in both the cortex ( Figure 4F,H,I) and striatum ( Figure 4K,M,I) compared to DMSO. Quantification showed that BK50118-C (40 mg/Kg) significantly reduced the number of human alphasynuclein positive neurons by 42% in cortex and 40% in striatum ( Figure 4I,N).
Nissl staining showed that BK50118-C significantly increased neuron counts in cortex (Supplementary Figure Figure S3I). tification showed that BK50118-C (40 mg/Kg) significantly reduced the number of human alpha-synuclein positive neurons by 42% in cortex and 40% in striatum ( Figure 4I,N).
Nissl staining showed that BK50118-C significantly increased neuron counts in cortex (Supplementary Figure

BK50118-C Increases Alpha-Synuclein Ubiquitination and Has Minimal Effects on Tyrosine Hydroxylase Levels in Striatum of TgA53T Mice
To ascertain the mechanistic effects of BK50118-C on alpha-synuclein clearance, we performed immunostaining in striatum of TgA53T mice. Human alpha-synuclein staining showed the level of human alpha-synuclein in DMSO ( Figure 5A) compared to BK50118-C (40 mg/Kg) treated mice ( Figure 5B). Ubiquitin staining in DMSO treated striatum ( Figure 5C) showed the level of ubiquitin compared to mice treated with 40 mg/Kg of BK50118-C ( Figure 5D). Merged alpha-synuclein and ubiquitin staining in DMSO ( Figure 5E), compared to BK50118-C ( Figure 5F) showed that alpha-synuclein co-localized with ubiquitin in mice treated with BK50118-C. Optic density of alpha-synuclein-ubiquitin staining showed a significant increase (130%) in co-localization in the striatum when mice were injected with BK50118-C compared to DMSO ( Figure 5G). We also performed the staining of tyrosine hydroxylase (TH) in the striatum of TgA53T mice. TH is the enzyme responsible for catalyzing the rate limiting step in synthesis of L-3,4-dihydroxyphenylalanine(L-DOPA) which is a precursor for dopamine (DA). Staining of TH will help to evaluate the status of DA producing neurons in the SN and their terminals in the striatum. Staining of tyrosine hydroxylase (TH) in the striatum of TgA53T mice did not show any noticeable effects in BK50118-C ( Figure 5I

Discussion
The current research demonstrates that pharmacological inhibition of USP13 lowers neurotoxic protein levels, including alpha-synuclein and improves neurodegenerative pathology. We previously showed that shRNAs which target USP13 expression can reduce alpha-synuclein via autophagy and or the proteasome in vivo and in vitro [20,21]. The current results are in agreement with our previous findings that knockdown via shRNA and pharmacological inhibition of USP13 increase clearance of ubiquitinited proteins [20,21]. We synthesized novel small molecules and potent inhibitors of USP13 that effectively reduce neurotoxic protein levels and protect against neuronal death. Mechanistically, these novel USP13 inhibitors mitigate the detrimental activity of USP13 that de-ubiquitinates alpha-synuclein. USP13 inhibition may antagonize de-ubiquitination of neurotoxic proteins and prevent their aggregation into LBs and tangles via promotion of ubiquitination to facilitate autophagy and proteasome protein clearance [20,21]. Collectively these findings suggest that USP13 is a therapeutic target in neurodegenerative diseases [28].
The novel USP inhibitors show very high potency to inhibit USP13, and to our knowledge, they represent the first library of small molecule, brain penetrant USP13 inhibitors. Spautin-1 does not cross the BBB, and it inhibits both USP10 and USP13 and affects the proteasome and autophagy at much higher concentration than the novel USP13 library [23]. Our novel USP13 inhibitors are non-toxic at the concentrations used, but more work is needed to determine their activity against other USPs. Therefore, further investigation to determine the specificity of the novel molecules to other USPs and additional in vivo studies will be performed to determine PK parameters of a wider range of drug concentrations and their safety and efficacy. Our future investigation will focus on mechanistic aspects beyond de-ubiquitination to include specific mechanisms of protein clearance via autophagy and the proteasome in USP13 knockout mice.

Transgenic Mice, Stereotaxic Surgery and Treatment
Transgenic TgA53T mice were used in the experiment. rTg4510 mice are hemizygous and have the Tet-responsive element as well as a mouse prion protein promoter sequence directing expression of the human P301L tau. Hemizygous rTg4510 mice are bred to CaMKIIa-tTA mice resulting in the rTg4510 bi-transgenic model with conditional P301L tau expression in the forebrain. Transgenic A53T mice harbor the arginine to threonine (A53T) mutation of human alpha-synuclein under the control of prion promoter and have abundant alpha-synuclein in the striatum as early as 3 months of age [26].
BK50118-C is one of novel small molecule inhibitors of USP13 synthesized in our laboratory.
For the synthesis of BK50118-A, BK50118-B and BK50118-C ( Figure S4), an 8 mL pressure vessel was charged with a heteroaryl chloride and two equivalents of either aniline (2) or benzyl alcohol (4) in DMSO unless noted otherwise. The pressure vessel was then placed in a 100 • C oil bath, and the mixture was stirred for 24 h. After full conversion was achieved based on 1 H NMR analysis, the reaction mixture was quenched with NaHCO 3 and extracted with EtOAc. The combined organic layers were extracted with water and dried over sodium sulfate, and the solvent was removed in vacuo. The crude product was purified by flash chromatography on silica gel using with hexanes-ethyl acetate as mobile phase as described below.

Cell LINES, Transfection and Treatment
Human SH-SY5Y neuroblastoma cells were procured from the Tissue Culture and Biobanking Shared Resource at Georgetown University Lombardi Comprehensive Cancer Center. Stably transfected SH-SY5Y cells expressing human wild-type alpha-synuclein were developed via selective resistance to G418 disulfate salt (Sigma-Aldrich, A1720, St. Louis, MO, USA). Cell identity was not further authenticated. SH-SY5Y cells were cultured in DMEM with Ham's F12 (1:1) (ThermoFisher, 11765054, Waltham, MA, USA) with 10% FBS, 1% PenStrep and 1% L-glutamine. Cells were plated at a density tailored to reach 70-80% confluence at the beginning of every experiment.

Immunoprecipitation (IP)
Mouse brain tissues were homogenized in 1x STEN buffer, and the soluble fraction was isolated as indicated above. The lysates were pre-cleaned with immobilized recombinant protein A/G agarose (Santa Crutz, sc-2003, Dallas, TX, USA) and centrifuged at 2500× g for 1 min at 4 • C. The supernatant was recovered and quantified by protein assay, and a total of 300 µg protein was incubated overnight at 4 • C with primary anti-alpha-synuclein (1:200, Thermofisher, AHB0261) mouse antibodies or anti-ubiquitin (1:100) (Thermo Fisher, PA3-16717, Rockford, IL, USA) antibodies in the presence of sepharose G and an IgG control with primary antibodies. The immunoprecipitates were collected by centrifugation at 2500× g for 3 min at 4 • C, washed 5× in PBS, with spins of 3 min, 2500× g using detergent-free buffer for the last washing step, and the proteins were eluted according to Pierce instructions (Pierce #20365, Rockford, IL, USA). After IP, the samples were size-fractionated on 4-12% SDS-NuPAGE and transferred onto 0.45 µm nitrocellulose membranes. WB detection was then performed using horseradish peroxidase (HRP)-conjugated secondary antibodies.

Immunohistology
Animals were deeply anesthetized with a mixture of xylazine and ketamine (1:8), washed with normal saline for 1 min and then perfused with 4% paraformaldehyde (PFA) for 15-20 min. Brains were quickly dissected out and immediately stored in 4% PFA for 24 h at 4 • C and then transferred to 30% sucrose at 4 • C for 48 h. Brains were cut using a cryostat microtome into 20 µm thick coronal sections and stored at −20 • C.
Immunohistochemistry was performed on the 20 µm thick brain sections for evaluation of ubiquitination of alpha-synuclein. The antibodies used were human alpha-synuclein monoclonal antibody (Thermo Fisher, AHB0261, Rockford, IL, USA) and ubiquitin polyclonal antibody (Thermo Fisher, PA3-16717, Rockford, IL, USA). The optic densitometry of co-localization of ubiquitin with alpha-synuclein was measured using Image J. Tyrosine hydroxylase (TH) is the limiting enzyme in DA synthesis, so probing for TH+ neurons will help to evaluate the status of DA producing neurons in the SN and their terminals in the striatum. We performed fluorescent staining of TH+ in striatum and conducted optic densitometry measurement of striatal DA terminals. Nuclear staining with 4, 6-diamidino-2-phenylindole was performed according to manufacturer's protocols (Life Technologies, Rockford, IL, USA).
DAB staining was performed for alpha-synuclein (Thermo Fisher, AHB0261, Rockford, IL, USA) on the 20 µm thick mouse brain sections, and stereological counting of alpha-synuclein+ neurons counterstained with Nissl was conducted.

Pharmacokinetics Studies
C57BL/6 mice received a single intraperitoneal dose (10 mg/kg) of BK50118-C. Brain and serum samples were collected at 1, 2, 3, 4, 8 and 12 h (n = 3 per time point). Animals injected with vehicle (DMSO) were used for background subtraction. Stock solution of the drug (1 mg/mL) and internal standards were prepared in methanol. Intermediate solutions used for the calibrators and control samples were serial-diluted in methanol/water (1:1). Preparation of the calibration curve standards and quality samples (QC) were performed by mixing the intermediate dilutions in blank samples (brain homogenates, serum). The internal standard working solution contained deuterium labeled BK50118-C-d7 at the concentration of 5 ng/mL diluted in acetonitrile (ACN)/ethyl acetate (4:1). Serum and brain samples were stored at −80 • C and then thawed to room temperature prior to preparation. The brains were homogenized in MilliQ water (1 mg brain: 10 µL water). Proteins were precipitated in both brain and serum samples by mixing 25 µL aqueous sample with 75 µL internal standard working solution. The mixture was centrifuged at 12,300× g for 5 min. Thereafter, 75 µL of each supernatant and 25 µL of MilliQ water were pipetted into a 96-well PCR plate (Fisher Scientific, Dawsonville, GA, USA).
The concentrations of BK50118-C in the brain tissue and serum samples were measured by ultrahigh performance liquid chromatography tandem mass spectrometry (UHPLC-MS/MS). Briefly, the UHPLC-MS/MS system included an Elute HTG binary gradient UHPLC pump, an Elute column oven and an EVOQ Elite triple quadrupole mass spectrometer (all from Bruker Daltonik GmbH, Bremen, Germany) equipped with an electrospray ionization (ESI) source operating in a positive mode. The samples were injected by use of a PAL auto sampler (CTC Analytics, Zwingen, Switzerland) equipped with a 10-µL sample loop; the samples were kept in a PAL stack cooler for 6 microtiter plates and operating at +6 • C. The system was controlled by a Compass 2.0/HyStar 4.0 software (Bruker); the compound screening and quantitation was performed by a TASQ 2.2 data acquisition and processing software (Bruker). The mass spectrometer was supplied by nitrogen and air generated by a Genius 3045 nitrogen/air generator (Peak Scientific Instruments, Inchinnan, Scotland, UK). The ESI parameters were as follows: probe gas flow 50, nebulizer gas flow 60, probe temperature +400 • C, cone gas flow 20, cone temperature +350 • C, CID gas Ar 1.5 mTorr. The mass spectra were scanned in the MRM mode to find the optimal collision energies for the test compounds and their respective precursor ions.

Statistical and Data Analysis
All statistical analyses were performed using GraphPad Prism version 5.0 (GraphPad software, Inc, San Diego, CA, USA). Two-tailed student's t-test was used in comparison of means of two groups. Ordinary one-way analysis of variance (ANOVA) followed by Newman-Keuls or Dunnett comparison post hoc tests were used in comparison of means of multiple groups. Asterisks denote actual p-value significances (* < 0.05, ** < 0.01, *** < 0.001 and **** < 0.0001), and N is the number of animals or the number of independent experiments (cell culture) per group. Unless otherwise indicated, data are expressed as Mean ± SD. IC50 to USP13 levels was also estimated using GraphPad Prism version 5.0 (GraphPad software, Inc, San Diego, CA, USA).

Conclusions
In conclusion, these findings suggest that USP13 inhibition may oppose protein aggregation and inclusion formation. The role of USP13 in regulating the ubiquitination and de-ubiquitination cycle and initiation and control of autophagy are integral to cell homeostasis and survival. USP13 inhibition via our novel analogues may provide balance of ubiquitination and de-ubiquitination for toxic protein degradation in neurodegenerative diseases.