Mass Spectrometry-Guided Discovery of Novel GCPII Inhibitor Scaffolds

Robyn Wiseman¹Ajit G. Thomas¹John Janiszewski²Nate Hoxie²Chae Bin Lee¹Ying Wu¹Takashi Tsukamoto¹Xin Hu²Michael Ronzetti²Ganesha Rai²Rana Rais¹Jesse Alt¹Matt Hall²Stephen Kales²Barbara Slusher^1*

• ¹Johns Hopkins University, Baltimore, United States
• ²National Center for Advancing Translation Sciences, Rockville, United States

There is an unmet need for therapeutics with a novel mechanism to address symptoms associated with conditions where aberrant glutamatergic neurotransmission is presumed pathogenic. One enzyme of potential relevance is glutamate carboxypeptidase II (GCPII), a brain metallopeptidase with significantly upregulated activity in nervous tissues following neurodegeneration or injury. Current inhibitors are too polar and charged leading to minimal brain penetration necessitating high systemic doses or direct brain injection. Our efforts are focused on identifying new inhibitor scaffolds with favorable brain penetration. Herein, we used a newly developed dual-stream liquid chromatography mass spectrometry (LC/MS/MS) substrate cleavage assay to screen two small molecule libraries. The two top confirmed hits were cefsulodin (IC50 = 2 ± 0.1 µM) and amaranth (IC50 = 0.3 ± 0.01 µM). A nano differential scanning fluorimetry (DSF) thermal shift assay was used to examine the interactions cefsulodin and amaranth with GCPII. A first derivative analysis revealed a shift in Tm of Δ 0.76 °C (± 0.04) for amaranth at 25 μM and 80.41 °C (± 0.05) for cefsulodin at 250 μM, suggesting both compounds are acting as stabilizers for the enzyme. Increasing concentrations of cefsulodin increased the Km of N-acetyl-aspartyl-glutamate (NAAG) as a substrate with no change in Vmax, suggesting active site competitive inhibition. In contrast, increasing concentrations of amaranth led to reductions in Vmax while the Km remained constant, suggesting a non-competitive mode of inhibition. Results from in silico docking studies complemented this mode of inhibition data, suggesting cefsulodin likely binds in the active site while amaranth likely binds in an allosteric site. As cefsulodin is an antibiotic previously used clinically to treat bacterial meningitis, we tested for its brain pharmacokinetics in mice. Using a sensitive LC/MS method we developed, we demonstrated that administration of cefsulodin (100 mg/kg IP) led to a Cmax of 4 µM in the brain, exceeding its GCPII IC50 value. Moreover, following confirmation and characterization of cefsulodin and amaranth as viable hits an SAR investigation was conducted with analogs of both compounds. Our new screening approaches identified novel inhibitors of GCPII that could serve as molecular templates for further structural optimization.