7662
F. Lv et al. / Bioorg. Med. Chem. 23 (2015) 7661–7670
Thus enhancement of glycosylation of
nue to treat both MD and cancers. However, up to date, all the gene
therapies to enhance the FOG of -DG have been limited to animal
experiments. Furthermore, biosynthesis of the FOG of -DG is a
a
-DG is a potential ave-
reen HTS assay measures the FOG of a-DG-HIG secreted into the
cell culture media specifically, and allows for the screening of
compounds in a cost and time effective manner. The AlphaScreen
HTS assay is illustrated in Figure 1.
a
a
complicated process that requires the direct involvement of more
than a dozen of genes. The balance of the expression levels of these
genes is critical for the concerted biosynthesis process. Simply
overexpressing one of these genes may have adverse impacts on
the process. In fact, two independent studies demonstrated that
overexpression of the LARGE gene alone exacerbated the dys-
trophic phenotypes rather than mitigated the dystroglycanopathy
in the mouse models.29,30
2.2. Ultra HTS campaign, hit selection and validation
The Prebys Center completed the primary screening (PubChem
AID: 624168) of ꢀ364,168 compounds of the NIH MLSMR at a sin-
gle concentration of 10 lM in dimethyl sulphoxide (DMSO, 1% final
solvent concentration) with the AlphaScreen assay. The assay was
performed well in 1536-well (5
l
L) format with Z0-factor of 0.61,
robust Z0-factor of 0.59, signal-to-background ratio (S/B) of 63.3,
signal-to-noise ratio (S/N) of 328, and a signal window (SW) of
5.1. We chose a generous hit criteria of P8% enhancement based
on the control wells, or compounds that show a Z-score of P8.
In combining normalized and corrected data that meet the first
criterion and Z-scores based on raw data, 805 compounds were
selected as hits corresponding to 0.22% of screened compounds.
After filter of Pan Assay Interference Compounds (PAINS),31 649
compounds were requested as ‘cherry picks’ from the NIH MLSM
and 600 compounds were available from the MSLMR. The descrip-
tion of the detailed validation process with the AlphaScreen assay
by the Prebys Center can be found in the Pubchem databank
(AID651671), in which no any hits was revealed to be significant
Utilizing small molecules to specifically enhance the endoge-
nous biosynthesis of the FOG of
dles. We hypothesized that the cells can specifically enhance the
FOG of -DG in response to certain chemical stimuli. Thus we
developed a cell-based AlphaScreenÒ assay that can quantitatively
monitor the FOG of -DG in an HTS friendly manner to identify
compounds that enhance FOG of -DG. Through an NIH-funded
a-DG may circumvent these hur-
a
a
a
initiative of Molecular Libraries Program (MLP), and collaboration
with the Conrad Prebys Center for Chemical Genomics (CPCCG), a
comprehensive center in the Molecular Libraries Probe Production
Centers Network (MLPCN), 364,168 small compounds of the NIH’s
Molecular Libraries Small Molecules Repository (MLSMR) were
screened in an ultra-HTS AlphaScreen campaign, in which a set
of hits were selected, followed by intensive validation with a series
of assays. Furthermore, an iterative medicinal chemistry study was
conducted. These works led to discovery of a cluster of compounds
enhancers for FOG on a-DG-HIG.
2.3. Secondary validation of compounds
that can enhance the FOG of
the patient-derived primary myoblasts. The compounds enhanced
the FOG of -DG by 8–12 folds in an endogenous O-mannosylation
pathway-dependent manner. These potent (EC1.5 1.2 M) com-
a-DG in various cell lines including
As the AlphaScreen assay measured the FOG of
secreted into and accumulated in bulk cell media and with LARGE
co-expression as an enhancement control, an orthogonal direct
a-DG-HIG
a
l
pounds will be useful molecular probes to study the O-mannosyla-
tion pathway and offer potential starting point for the drug
development to treat dystroglycanopathy.
measurement of FOG of endogenous
a-DG on the cell surfaces
may yield a more biologically significance. We then validated
narrowed-down 600 compounds with the HTS assay using Pro5
cells and C2C12 mouse myoblasts coupled with the IIH6 antibody
as the probe for detecting the FOG of endogenous
surfaces. Cells were grown in 384-well microplates and treated
with compounds at a single concentration of 10 M in quadrupli-
a-DG on the cell
2. Results and discussion
l
2.1. Assay development for high-throughput screening (HTS)
We chose a cell-based assay platform to identify small-molecules
cate for 48 h. Unlike Pro5, C2C12 cells, did not give rise to any hits
enhancement of IIH6 immunofluorescent staining signal over 20%.
Interestingly, 101 compounds could increase fluorescent staining
signal on the cell surface more than 20% of untreated control, when
Pro5 cells were used in the validation experiments. Among the 101
hits the false positive hits that were often with nuclear, cytoplas-
mic staining patterns or auto fluorescence were eliminated and
the compounds that could induce cell surface staining patterns
and consistent among all four repeating wells were selected as true
positive hits. Two hits (Hit1, compound 1, PubChem CID: 650572;
Hit2, compound 8, PubChem CID: 653717; structures for two hits
that enhance the FOG of
a-DG on the cell surfaces. Pro5 cell, a
Chinese hamster ovary (CHO) epithelial cell line, was chosen after
screening a number of cell lines such as CHOK1, T293, and a series
of myoblasts from humans and rodents, since it gave the best HTS
results. Two assays were established. In the first assay, Pro5-LARGE
cell (P5-LG), in which mouse LARGE cDNA was tagged with an MYC
epitope was stably overexpressed, served as the positive control,
while the parent Pro5 cell served as the negative control and
screening cell. The IIH6 antibody and laminin-111 were used as
shown in Fig. 2a) could enhance the FOG of a-DG on the cell sur-
the probes to detect the FOG of
produced very similar results. This assay examines the FOG of
endogenous -DG and is easy to adapt to other adherent cells such
as myoblasts. It can also be adapted to a HTS assay via fluorescent
microscopy. The detailed assay information are shown in Figures
S1–S4 and Tables S1 and S2. However, it does require wash steps
in HTS process, thus it is not ideal for a primary HTS assay. In order
to effectively screen a very large amount of compounds (>360,000),
we reconfigured the first assay into a wash step-free HTS-friendly
AlphaScreen format, and miniaturized samples into 1536-well
a-DG on the cell surfaces, and
faces by 37% and 56%, respectively, compared to the untreated con-
trol. The fluorescent microscopy images of the two hits were
shown in Figure 2b. Furthermore, Laminin-111, the ligand of
a
the FOG of
a-DG, was employed as another probe to examine the
FOG of -DG on Pro5 cell surface induced by the two hits. The
a
results clearly indicated that the two hits can enhance the lami-
nin-111 binding on cell surfaces (Fig. 2c). Since there are other
receptors for laminin-111 such as integrin, which are also highly
glycosylated, in order to check whether the hits increase the IIH6
and laminin binding specifically through the enhancement of the
microplates. In this assay, Pro5-LARGE/
LARGE and a secreted form of -DG-HIG were stably overex-
pressed, served as the positive control. Pro5/ -DG-HIG cell, in
which a secreted form of -DG-HIG was stably overexpressed,
served as the negative control and the screening cell. This AlphaSc-
a-DG-HIG cell, in which
FOG on
at a series of concentrations as indicated in Figure 2d for 72 h. After
accumulation the secreted form of -DG-HIG in serum free media,
the laminin overlay assay to analyze the media was established,
a-DG, the Pro5/a-DG-HIG cells were incubated with Hit2
a
a
a
a