Identifying Small Molecule Therapies by HTS assays using LSD patient cells

We are developing robust high-throughput screening (HTS) cell-based assays using directly patient cell lines to screen selected libraries of drug-like compounds that enhance the residual activity of specific misfolded mutant enzymes. Recently, we developed quantitative HTS screening assay for metachromatic leukodystrophy (MLD), a LSD caused by deficiency of arylsulfatase A (ASA) resulting sulfite accumulation in nervous system. Sulfatides play essential role in myelin stability (Walkins P et al 2004). Clinically, MLD presents a wide spectrum of manifestations secondary to central and peripheral white matter degeneration including cognitive impairment, seizures, progressive neuropathies and psychiatric disturbances. Since this is a cell-based assay followed by biochemical assay (measuring of enzyme ASA activity), small molecules that assist the folding of mutant ASA directly and others that do so indirectly will be identified. A novel fluorescence ASA assay was developed in high-density 1,536-well plates using the traditional colorimetric pNCS substrate, whose product (pNC) acts as “plate fluorescence quencher” in white solid-bottom plates (Geng et al 2011). In collaboration with Marc Ferrer PhD and Wei Zheng PhD from the National Center for Advancing Translational Sciences (NCATS), Rockville, MD, we showed that quantitative cell-based HTS assay for ASA generated strong statistical parameters when tested against a diverse small molecule collection. we tested the developed assay against small molecule library and confirmed the robustness of the assay. Currently we are preparing this quantitative cell-based HTS assay to screen the Molecular Libraries Production Network Center (ML), which comprehends a collection of 400,000 small molecules.

Fig.1. Development of a fluorescence-quench absorbance assay for ASA using a colorimetric substrate in white solid-bottom 1,536-well plates

Legend to Fig.1. Development of a fluorescence-quench absorbance assay for ASA using a colorimetric substrate in white solid-bottom 1,536-well plates. (A) Basis of the fluorescence-quench absorbance assay. The desulfation of pNCS (substrate) by ASA generates pNC (product), a colorimetric product, whose optimal absorbance is detectable at 515 nm. Thus, the reaction is performed in clear-bottom (transparent) microplates to measure trans-absorbance. Since white plastic solid-bottom plates have fluorescence properties (Zuck P et al. 2005), when an excitation light is directed to white solid-bottom microplates, emission light can be detected by epi-absorbance. At excitation/emission pair (525/598 nm), pNCS does not quenches the fluorescence from white solid-bottom plate (yellow well with increased green arrow), demonstrating an increasing OD signal with the increase of %pNC and decreased %pNCS (substrate) in the mixed solution of pNCS+pNC. This pNC+pNCS solution (yellow line) overlaps with the solution containing only pNC at 0.2 mM (pink line). (B) The disposition of SV40t fibroblasts from MLD patient (ASA-I179S – blue) and from control (ASA-WT – yellow) in a 1,536-well plate. (C) The time-line of events of this cell-based HTS assay for ASA. A library of 1,280 small-molecules at seven different concentrations from 7×10-3 to 114.2 mM was used to perform a pilot HTS assay. (D) Scatter plot from the quantitative cell-based HTS assay for ASA using the LOPAC library. In each panel, columns 2, 3, 46 and 47 depict OD values from wells with control cells (ASA-WT), which were not exposed to small molecules. OD values from MLD patient fibroblasts (ASA-I179S) treated only with DMSO were located in columns 4 and 45. SV40t MLD patient fibroblasts were seeded in columns 5 to 44. The 1,536-well plates treated with DMSO (0.57%) (A), lowest (7 x 10-3 mM-left panel) and highest (114.2 mM-right panel) concentrations of LOPAC are shown. (E) Concentration-response curve analysis obtained from the quantitative cell-based HTS assay for ASA against LOPAC library. More information at Geng H. et al. 2011. This research was funded by NINDS-NIH and National Tay-Sachs Association (NTSAD).

Using a similar approach a quantitative cell-based HTS assay for globoid-cell leukodystrophy (GLD), known mostly as Krabbe disease, another quantitative cell-based HTS assay have been developed b-galactocerebrosidase (GALC). GLD is rapidly progressive, invariably fatal LSDs manifested mostly in infancy (Wenger D et al 2001). Clinically, GLD initially presents with hyperactive reflexes and seizures in the early stages, evolving to hypotonia, blindness and deafness. Peripheral neuropathy is present in most patients. GLD may manifest as late onset, including juvenile and adult onset forms presenting with developmental delay, spastic paraparesis, neuropathies and psychiatric symptoms (Furuya H et al 1997; Fuimara A et al. 2010). Deficiency of GALC results in increased levels of galactosylceramide and galactosylsphingosine (psychosine) as natural substrates. In brain specimens of GLD patients, psychosine levels are substantially elevated (Svennerholm L et al 1980). At elevated concentrations, psychosine becomes cytotoxic to neuronal cells, particularly to oligodendrocytes (OLG), neural cells responsible for myelin formation. Therefore, the biochemical disturbances in GLD are manifested by different degrees of brain demyelination resulting in a broad neurological spectrum of GLD (Suzuki K 1998). A pilot study using a cell-based assay against a small 1,248-small-molecule library produced robust results (Fig.2). Dr. Furuya kindly donated the cell line (GALC-G270D) used in this assay. Cellular assays using patient cells provide a unique opportunity to assess a target protein and/or pathway in the context of potentially disrupted biochemical and/or signaling pathways secondary to the disease process. In addition, these assays permit the evaluation of multiple intervention points, which are potentially altered in the disease, as opposed to commonly used specific protein target or a predefined step of a purified or recombinant protein-based assay. Using psychosine, a glycosphingolipid that is extremely cytotoxic at elevated concentrations, we are currently developing a throughput assay using oligodendocyte-precursor cells (OPCs) from mouse model for GLD (GALCtwi/twi) (Fig.2). The primary glial cultured cells from which OPCs were transformed and isolated were kindly donated by Dr. Ernesto Bongarzonne PhD, University of Illinois, Chicago, IL. In collaboration with Ann Moser from Kennedy Krieger Institute, Johns Hopkins Medicine, we were able to show thorough a developed LC-MS/MS method that we could detect increased levels of psychosine in cultured OPCs. Using a specific column, we were also able to distinguish psychosine (galactosylsphingosine) from glucosylsphingosine (have same molecular weight). We were able to show robustness of measurements at 96-well plate format. We are currently optimizing this assay for testing in small compound collections to test the robustness of this assay.

Fig.2. Two different approaches to identify therapeutic small molecules for Krabbe disease (GLD)


Legend to Fig.2. Two different approaches to identify therapeutic small molecules for Krabbe disease (GLD). 
One of the approaches is targeting the mutant misfolded GALC using a cell-based quantitative HTS assay to identify small molecules that can assist mutant GALC folding enhancing its residual enzymatic activity. (A) Scatter plot of GALC assay performed in 384-well plate containing cultured skin fibroblasts from GLD patient GALC-G270D mutant and a control with GALC wild type (WT) using a specific synthetic fluorescent substrate, 6-hexadecanoylamino-4-methylumbelliferyl-b-D-galactoside (HMUbGal). (B) Quantitative cell-based HTS assay was adapted to 1,536-well plates and was tested against the 1,248 small molecules from LOPAC at seven concentrations ranging 4.6-230 μM concentrations. (C) The three scatter plots are depicted with results from 1,536-well plates with cells only treated with DMSO (A), and LOPAC at 230 μM (B). Columns 1 and 2 contained only culture medium, 3 and 4 contained SV40t control cells (GALC-WT) and 5-48 contained SV40t GLD patient cells (GALC-G270D). (D) Nineteen “hits” were selected based on concentration-response curves (CRCs): 2 molecules showed class 1 CRCs, 11 CRCs class 2 and 6 others showed class 3 CRCs (Inglese J et al. 2005). (E) Another approach is to target the cytotoxic molecule, galactosylsphingosine, known as psychosine, that is accumulate din GLD. Here is a characterization of OPC cells obtained from primary glial cultures from brain of mouse model for GLD, Twitcher GALCtwi/twi. These cell demonstrated expression of A2B5 cell marker and accumulation of galactosylceramide (immediate substrate also increased in GLD). (E) In collaboration with Ann Moser and Walter Hubbard, we developed a robust LC-MS/MS assay able to detect elevated levels of galactosyl-sphingosine (SPG) in OPC from GALCtwi/twi in comparison with GALCwt/wt. Using a specific BEH glycan column, we were able to distinguish the galactosphongosine (psychosine) from glucosyl-phingosine (two lower spectra). (F) Using the LC-MS/MS assay with a X-Terra C8 reverse column with a short run (4min), OPCs from GALCtwi/twi in 96-well plates demonstrated increase psychosine levels indicating the throughput potential of the developed assay.

The HTS sketch flow for small molecule discovery campaigns. Here the HTS for psychosine-reducing agents from the primary screening and characterization assays is depicted.

Legend to Figure 3. The HTS sketch flow for small molecule discovery campaigns. Here the HTS for psychosine-reducing agents from the primary screening and characterization assays is depicted.

The characterization of small molecule candidates from primary screenings are examined in robust animal models. We have been validating the psychosine-reducing agents in the Twitcher mouse (galctwi/twi) and performing immunohistochemistry, LC-MS/MS for sphingolipids analysis, ultrastructural studies and neurobehavioral assessments.

Legend to Figure 4. The characterization of small molecule candidates from primary screenings are examined in robust animal models. We have been validating the psychosine-reducing agents in the Twitcher mouse (galctwi/twi) and performing immunohistochemistry, LC-MS/MS for sphingolipids analysis, ultrastructural studies and neurobehavioral assessments.