Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors

Article abstract of DOI:10.1016/j.chembiol.2018.07.005

The oligosaccharyltransferase (OST) is a multisubunit enzyme complex that N-glycosylates proteins in the secretory pathway and is considered to be constitutive and unregulated. However, small-molecule OST inhibitors such as NGI-1 provide a pharmacological approach for regulating N-linked glycosylation. Herein we design cell models with knockout of each OST catalytic subunit (STT3A or STT3B) to screen the activity of NGI-1 and its analogs. We show that NGI-1 targets the function of both STT3A and STT3B and use structure-activity relationships to guide synthesis of catalytic subunit-specific inhibitors. Using this approach, pharmacophores that increase STT3B selectivity are characterized and an STT3B-specific inhibitor is identified. This inhibitor has discrete biological effects on endogenous STT3B target proteins such as COX2 but does not activate the cellular unfolded protein response. Together this work demonstrates that subsets of glycoproteins can be regulated through pharmacologic inhibition of N-linked glycosylation. Inhibition of the oligosaccharyltransferase and N-linked glycosylation was previously demonstrated using the small molecule NGI-1. Here we show that changes in the functional groups of NGI-1 are able to regulate subunit specificity toward the oligosaccharyltransferase catalytic subunits, STT3A and STT3B, and identify a series of STT3B-specific inhibitors.

Full text of DOI:10.1016/j.chembiol.2018.07.005

Article
Editing N-Glycan Site Occupancy with Small-
Molecule Oligosaccharyltransferase Inhibitors
Graphical Abstract
Authors
Natalia Rinis, Jennifer E. Golden,
Caleb D. Marceau, Jan E. Carette,
Michael C. Van Zandt, Reid Gilmore,
Joseph N. Contessa
Correspondence
joseph.contessa@yale.edu
In Brief
Inhibition of the
oligosaccharyltransferase and N-linked
glycosylation was previously
demonstrated using the small molecule
NGI-1. Here we show that changes in the
functional groups of NGI-1 are able to
regulate subunit specificity toward the
oligosaccharyltransferase catalytic
subunits, STT3A and STT3B, and identify
a series of STT3B-specific inhibitors.
Highlights
d
d
d
d
NGI-1 is a new OST inhibitor targeting catalytic subunits
STT3A and STT3B
NGI-1 analogs were tested with reporters in STT3A- and
STT3B-deficient cells
Analysis revealed inhibitors with differential subunit profiles
toward STT3A and STT3B
A series of STT3B-specific inhibitors were identified
Rinis et al., 2018, Cell Chemical Biology 25, 1–11
October 18, 2018 ª 2018 Published by Elsevier Ltd.
https://doi.org/10.1016/j.chembiol.2018.07.005

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Cell Chemical Biology
Article
Editing N-Glycan Site Occupancy with Small-
Molecule Oligosaccharyltransferase Inhibitors
Natalia Rinis,1 Jennifer E. Golden,2 Caleb D. Marceau,3 Jan E. Carette,3 Michael C. Van Zandt,4 Reid Gilmore,5
and Joseph N. Contessa1,6,7,
*
1Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT 06511, USA
2University of Wisconsin, Medicinal Chemistry Center, Madison, WI 53705-2222, USA
3Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305, USA
4New England Discovery Partners, Branford, CT 06405, USA
5Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA 01605, USA
6Department of Pharmacology, Yale School of Medicine, New Haven, CT 06511, USA
7Lead Contact
*Correspondence: joseph.contessa@yale.edu
https://doi.org/10.1016/j.chembiol.2018.07.005
SUMMARY
units, STT3A or STT3B (Kelleher et al., 2003). Both OST com-
plexes share a set of accessory subunits including ribophorins
The oligosaccharyltransferase (OST) is a multisubu- 1 and 2, OST48, DAD1, OST4, and TMEM258 as well as subunits
that have catalytic subunit-specific interactions—DC2 and
KCP2 for STT3A (Roboti and High, 2012a, 2012b; Shibatani
nit enzyme complex that N-glycosylates proteins in
the secretory pathway and is considered to be
constitutive and unregulated. However, small-mole-
cule OST inhibitors such as NGI-1 provide a pharma-
cological approach for regulating N-linked glycosyla-
tion. Herein we design cell models with knockout of
each OST catalytic subunit (STT3A or STT3B) to
screen the activity of NGI-1 and its analogs. We
et al., 2005; Shrimal et al., 2017) and MagT1 or TUSC3 for
STT3B (Cherepanova et al., 2014; Mohorko et al., 2014).
Recently it has been demonstrated that OST catalytic subunits
have distinct functions for glycosylation. STT3A is associated
with the translocon, regulates co-translational glycosylation,
and is responsible for the majority of N-linked glycosylation in
mammalian cells. In contrast, STT3B glycosylates a smaller
number of sequons and carries out post-translocational glyco-
sylation or maximizes sequon site occupancy by glycosylating
show that NGI-1 targets the function of both STT3A
and STT3B and use structure-activity relationships
to guide synthesis of catalytic subunit-specific inhib- sites that are skipped by STT3A (Ruiz-Canada et al., 2009).
Biochemical studies have demonstrated that sites frequently
skipped by STT3A are efficiently glycosylated by STT3B (Shrimal
itors. Using this approach, pharmacophores that in-
crease STT3B selectivity are characterized and an
STT3B-specific inhibitor is identified. This inhibitor
has discrete biological effects on endogenous
et al., 2015), including sites within five residues of the signal
sequence cleavage site, closely spaced NXS sites (Shrimal and
Gilmore, 2013), internal acceptor sites with suboptimal sequons,
STT3B target proteins such as COX2 but does not
and sites within the C-terminal ꢀ65 amino acids of a protein
activate the cellular unfolded protein response.
(Shrimal et al., 2013).
Together this work demonstrates that subsets of gly-
coproteins can be regulated through pharmacologic
Structural studies have also contributed to our understanding
of the N-linked glycosylation mechanism. In photo-crosslinking
inhibition of N-linked glycosylation.
experiments STT3A but not STT3B was shown to be associated
with the translocon (Nilsson et al., 2003), whereas co-crystalliza-
tion of the bacterial OST subunit PglB with an acceptor peptide
provided insight into the spatial arrangement of peptide and LLO
INTRODUCTION
´
binding sites (Lizak et al., 2011; Napiorkowska et al., 2017).
Asparagine (N)-linked glycosylation is a co- and post-transla- Furthermore, cryoelectron tomography studies of canine OST
tional modification important for protein stability, folding, traf- (Pfeffer et al., 2014) demonstrated the physical relationships be-
ficking, and physiological function (Tannous et al., 2015). This tween the OST, SEC61, and TRAP (Pfeffer et al., 2017).
task begins with the stepwise assembly of a lipid-linked oligo-
Inhibition of N-linked glycosylation using small molecules was
saccharide by a series of glycosyltransferases located on long limited to the use of the highly toxic natural product tunica-
both the cytoplasmic and luminal sides of the ER (Stanley, mycin, which blocks the transfer of N-acetylglucosamine-phos-
2016). It then proceeds with the en bloc transfer of the preas- phate to phosphorylated dolichol and abolishes all N-linked
sembled oligosaccharide onto asparagine residues within glycosylation (Takatsuki et al., 1971). However, we recently re-
NXT/S (X s P) sequons by the oligosaccharyltransferase ported the results of a high-throughput screening effort that
(OST) complex (Kelleher and Gilmore, 2006).
identified a small-molecule inhibitor of the OST, N-linked glyco-
Mammalian OSTs are hetero-oligomeric membrane com- sylation inhibitor 1 (NGI-1) (Lopez-Sambrooks et al., 2016).
plexes and contain one of two separately encoded catalytic sub- NGI-1 was shown to engage and block the activity of the OST
Cell Chemical Biology 25, 1–11, October 18, 2018 ª 2018 Published by Elsevier Ltd.
1

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
A
B
C
D
E
F
Figure 1. Screen for Small-Molecule Inhibition of OST Catalytic Subunits with the ERLucT Reporter
(A) Diagram of ERLucT with the N-terminal ER signal sequence and three glycosylation sites.
(B) Parental, STT3A-ko, and STT3B-ko HEK293 cells stably transfected with ERLucT were treated with 1 mM tunicamycin (T) or 10 mM NGI-1 (N) for 24 hr and
analyzed by western blot. Arrows on the right indicate the different glycoforms of ERLucT (0n–3n).
(legend continued on next page)
2
Cell Chemical Biology 25, 1–11, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
catalytic subunits and have a higher specificity toward STT3B (WD) (Marceau et al., 2016; Puschnik et al., 2017). Our results
compared with STT3A. Surprisingly, NGI-1 was sufficient to show that similar to wild-type Huh7 cells with endogenous
partially block glycosylation of multiple proteins without the STT3B, both STT3B mutants are thermostabilized by NGI-1
generalized toxicity observed with tunicamycin, suggesting (Figure S1). These results indicate that neither loss of catalytic
that N-linked glycan site occupancy can be pharmacologically activity nor mutation of the peptide binding site disrupts the inter-
modified in mammalian cells.
In this paper, we investigated strategies for synthesizing and
action of NGI-1 with the OST catalytic subunit.
We have previously reported the synthesis of an aminobenza-
screening a series of NGI-1 analogs in order to establish their midosulfonamide chemical series that partially blocks OST func-
biological effects on N-linked glycosylation and to improve their tion (Lopez-Sambrooks et al., 2016). To test the STT3A- and
specificity toward individual OST catalytic subunits. Our results STT3B-specific effects of small-molecule NGI-1 analogs, we per-
reveal structure-activity relationships for catalytic subunit inhibi- formed luciferase assays in the STT3A-ko and STT3B-ko HEK293
tion that translate into discrete biological effects at the protein cells after treatment with each of 17 unique NGI-1 derivatives
and cellular level. Overall this work provides evidence that phar- (Table S1, C1–C17) at a concentration of 10 mM for 24 hr. Results
macologic inhibitors of the OST will provide a platform for editing from the STT3B-ko cell line showed that the inhibitors had quan-
protein glycosylation and for enabling insight into OST biology.
titatively different effects on STT3A dependent N-linked glycosyl-
ation as measured by luminescence (Figure 1D) or mean glycan
occupancy on western blot (Figures 1E and 1F). Compounds
that generated the biggest changes in relative luciferase activity,
e.g., C2 and C13–C16, caused the biggest molecular weight
shifts on luciferase by western blot and were characterized by
RESULTS
Small-Molecule Inhibition of STT3A and STT3B Catalytic
Subunits
The ERLucT is a bioluminescent reporter developed to detect in- the lowest mean glycan occupancy values (approximately 0.5
hibition of N-linked glycosylation (Contessa et al., 2010). It con- for C2 and 0.6–0.8 for C13–C16). In STT3A-ko cells luminescence
sists of an ER signal sequence followed by a luciferase gene was not significantly induced by NGI-1 or its analogs, coincident
containing three N-linked sequons (Figure 1A). N-linked glyco- with high basal activity (Figure S2), indicating that the enzymatic
sylation of the ERLucT reporter inhibits luciferase activity, and in- activity of luciferase in our model system is largely regulated by
hibition of glycosylation thus increases luciferase activity and a single STT3A-dependent site. These experiments show that
measureable bioluminescence. To definitively assign inhibitory NGI-1 and its analogs have different potencies for STT3A and,
activities of small-molecule inhibitors of the OST to STT3A, coupled with our previous work (Lopez-Sambrooks et al., 2016),
STT3B, or both catalytic subunits, we stably transfected also suggest that the degree of STT3A inhibition can be regulated
HEK293 cells with a CRISPR/Cas9-mediated knockout (ko) of by the structure of the small molecule.
one catalytic subunit (either STT3A or STT3B) (Cherepanova
The shortcomings of the ERLucT reporter and the potential for
and Gilmore, 2016) with the ERLucT reporter (Figure 1B). Com- variability in catalytic subunit usage of sequons in different pro-
parison of luciferase glycoforms detected in parental, STT3A- teins led us to develop a second model system to determine
ko, and STT3B-ko cells revealed that one glycosylation site is the effects of OST inhibition on N-linked glycosylation. We
partially dependent on the presence of STT3A. Mean glycosyla- selected the HaloTag (Promega) because it is a protein of bacte-
tion of ERLucT dropped from 3.0 in parental cells to 2.2 in rial origin that can be easily detected by western blot, is ꢀ33 kDa
STT3A-ko cells but was unchanged in STT3B-ko cells (Fig- with superior resolution of glycoforms (compared with the
ure 1C). Treatment of cells with 10 mM NGI-1 had differential ꢀ60 kDa luciferase protein), and is relatively resistant to degra-
effects on ERLucT glycosylation in STT3A-ko and STT3B-ko dation (Neklesa et al., 2011). The Halo3N reporter was con-
backgrounds. In STT3A-ko cells NGI-1 blocked all glycosylation, structed by the addition of the signal sequence of epidermal
similar to tunicamycin treatment, whereas NGI-1 reduced growth factor receptor (EGFR), introduction of three consensus
ERLucT glycosylation in STT3B-ko cells to levels similar to those sequons by site-directed mutagenesis, and addition of a C-ter-
in the HEK293 parental controls. This finding suggests that minal ER-retention sequence (KDEL; Figure 2A). Three sites
NGI-1 is a more potent inhibitor of STT3B. The isolation of cata- were used to improve protein glycoform separation on western
lytic subunit activity by CRISPR/Cas9 knockout also definitively blot analysis and allow for band-intensity quantitation. Of note,
demonstrates that NGI-1 alters the function of both STT3A and one site was engineered into the C-terminal tail and predicted
STT3B containing OST complexes.
to be STT3B dependent due to the requirement of STT3B to
This understanding that NGI-1 is a partial inhibitor of both glycosylate sequons located within the C-terminal ꢀ65 amino
STT3A and STT3B led us to test the effects of NGI-1 with mutant acids of a protein (Shrimal et al., 2013).
OST catalytic subunits using the cellular thermal shift assay
The Halo3N reporter was stably expressed in parental, STT3A-
(Lopez-Sambrooks et al., 2016). For these experiments we ko, and STT3B-ko cells, and demonstrated near complete glyco-
used Huh7 cells with CRISPR/Cas9 knockout of STT3B and sylation in parental and STT3A-ko cells but complete loss of
reconstitution with (1) an STT3B catalytically inactive mutant glycosylation for one site in STT3B-ko cells, consistent with
(CM) or (2) an STT3B mutation to the WWD peptide binding motif the inability of these cells to glycosylate the C-terminal site
(C) Quantification of mean glycosylation site occupancy in (B).
(D–F) STT3B-ko HEK293 cells stably transfected with ERLucT were treated with 1 mM tunicamycin, 10 mM NGI-1, or 10 mM of C1-C17 and luciferase activity was
measured (D), or ERLucT glycosylation was analyzed by western blot (E) and quantified by calculating the mean glycan occupancy (F) (for details see STAR
Methods). Error bars represent the range of values from two independent experiments. D, DMSO; Luc, luciferase.
Cell Chemical Biology 25, 1–11, October 18, 2018
3

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
A
B
C
D
E
(legend on next page)
4
Cell Chemical Biology 25, 1–11, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Figure 3. Inhibitory Activity of STT3B-Spe-
cific Inhibitors on ERLucT and Halo3N
Glycosylation
Percent inhibition corresponds to relative reduc-
tion in mean glycans upon treatment.
shown to have inhibitory effects on
STT3B-dependent glycosylation but no
effects on STT3A-dependent glycosyla-
tion. This screen therefore provides evi-
dence that the chemical structure of the
NGI-1 series can be modified to produce
inhibitors that are specific for either
STT3A or STT3B glycosylation sites.
Structure-Activity Relationships for
STT3B Inhibition
A summary of the structure-activity rela-
tionships for STT3B-specific inhibitors
is presented in Figure 3. Initially, NGI-1
was identified as the most potent small-
molecule OST inhibitor and, as expected,
there is erosion of overall potency for both
STT3A- and STT3B-dependent glycosyl-
ation for each of these analogs. However,
we found that substitutions at all three
side groups of the heterocyclic com-
pound could reduce effects on STT3A
for both ERLucT and Halo3N reporters
while preserving an inhibitory effect to-
ward STT3B. These chemical modifica-
tions included (1) replacing the 5-methyl-
aminothiazole group at the R1 position
with a 4-methoxy-aniline (C6), (2) replacing the dimethylamine
group at the R2 position with a diethylamine (C1), N-piperazine
(C3), or benzylamine (C9), or (3) replacing the N-pyrrolidine at
the R3 position with a diethylamine (C11) or N-piperidine (C12).
Together these findings suggest that the NGI-1 binding site of
STT3B is more tolerant to changes in this chemical series
compared with the binding site of STT3A.
(Figure 2B). Treatment with NGI-1 again had a larger effect on
Halo3N glycosylation in STT3A-ko compared with STT3B-ko cells
(Figure 2B). Surprisingly, unlike luciferase, NGI-1 did not
completely block STT3B-dependent glycosylation of the Halo3N
reporter. These results therefore demonstrate that not all sequons
glycosylated by STT3B are susceptible to inhibition by NGI-1.
We next tested the 17 NGI-1 analogs on STT3A and STT3B
function by screening for effects on Halo3N glycosylation in
HEK293 cell lines. We found that several compounds have inhib-
itory activity that is comparable with NGI-1 in STT3A-ko cells
such as C2, C5, and C14–C16 (Figure 2D). Likewise, C2 and
C14–C16 cause reduction of mean glycan site occupancy in
STT3B-ko cells, similar to effects of NGI-1 (Figure 2E). These re-
sults confirm those from both luciferase activity and mean glycan
site occupancy experiments in STT3B-ko cells (Figure 1). Inter-
estingly, seven analogs (C1, C3, C6, C9, C11, and C12) were
The presence of a heterocyclic amine at the R1 position is crit-
ical for STT3A inhibitory activity. We therefore investigated the
role of the amide bond by reversing the amide at the R1 position
and generating two new compounds with either a methylthiazole
or phenyl substituent, respectively (Figures 4A and 4B). C18 and
C19 were then tested for inhibition of STT3A or STT3B using the
Halo3N model system. We found that C18 partially inhibited
STT3A activity (compare with C2, Figure 2E) and that C19
did not show any STT3A inhibitory activity at all (Figures 4C
Figure 2. Screen for Small-Molecule Inhibition of OST Catalytic Subunits with the Halo3N Reporter
(A) Diagram of Halo3N with the N-terminal ER signal sequence, three glycosylation sites, and the C-terminal KDEL ER-retention signal.
(B) Parental, STT3A-ko, and STT3B-ko HEK293 cells stably transfected with Halo3N were treated with 1 mM tunicamycin (T) or 10 mM NGI-1 (N) for 24 hr and
analyzed by western blot. Arrows on the right indicate the different glycoforms of Halo3N (0n–3n).
(C) Quantification of mean glycosylation site occupancy in (B).
(D and E) Western blot and mean glycan occupancy of Halo3N in STT3A-ko (D) and STT3B-ko cells treated for 24 hr with 1 mM tunicamycin, 10 mM NGI-1, or 10 mM
of C1–C17 (E). Error bars represent the range of values from two independent experiments. D, DMSO.
Cell Chemical Biology 25, 1–11, October 18, 2018
5

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Figure 4. C18 and C19 and Their Inhibitory
Activity on STT3B
Structures of C18 (A) and C19 (C). STT3A-ko and
STT3B-ko HEK293 cells stably transfected with
Halo3N were treated for 24 hr with 10 mM NGI-1 (N)
or 10 mM of C18 (B) or C19 (D), and Halo3N
glycosylation was analyzed by western blot. Ar-
rows on the right indicate the different glycoforms
of Halo3N. D, DMSO.
the receptor. Together these data show
that EGFR is a largely STT3A-dependent
glycoprotein that is only marginally
affected by either STT3B inhibition or
knockout.
We also evaluated the contributions of
STT3A and STT3B to the function of cy-
clooxygenase-2 (COX2), an enzyme that
catalyzes the rate-limiting step in prosta-
glandin synthesis (Smith et al., 2000).
and 4D). These findings demonstrate the additional importance COX2 is a glycoprotein with four N-linked glycosylation sites
of the central amide bond for STT3A-specific inhibition and (Figure 6B), and glycosylation of its most C-terminal glycosyla-
expand our understanding of the pharmacophore’s structural tion site at N580 has been shown to reduce the stability and ac-
constraints for blocking OST catalytic subunit function.
tivity of the protein (Sevigny et al., 2006). As N580 lies within the
After identifying these structure-activity relationships for last 65 amino acids of the protein we hypothesized that an inhib-
STT3A and STT3B inhibition, we next performed full dose- itor selectively targeting STT3B would decrease glycosylation at
response and time-course experiments for NGI-1 and analogs this site and therefore increase COX2 stability and activity. To
C2, C4, and C19 to determine STT3B specificity at maximum test this hypothesis, we evaluated COX2 levels in HEK293 cells
concentrations and times of exposure (Figure 5). NGI-1 and C2 with STT3A-ko or STT3B-ko, both with and without OST inhibitor
had comparable inhibitory activity toward STT3A and STT3B at treatment (Figure 6C). Similar to EGFR, STT3A-ko reduced over-
doses up to 25 mM (Figure 5B) and with treatment times up to all levels of COX2 protein. In contrast, we found that COX2 was
48 hr (Figure 5C). These results confirm that morpholine substi- increased in both STT3B-ko cells and parental cells treated with
tution at the R2 position does not significantly affect inhibitory C19. These effects were not observed with NGI-1 treatment due
activity. C19, which contains the reverse amide and phenyl to its inhibition of both STT3A- and STT3B-dependent glycosyl-
group at R1 and a morpholine at R2, displayed no STT3A inhibi- ation sites. In STT3B-ko cells, C19 could not further increase
tion whatsoever. These experiments demonstrate that even in COX2 expression, consistent with the loss of C-terminal glyco-
the absence of OST structural data, catalytic subunit-specific in- sylation in these cells. U87 glioma cells have high COX2 levels
hibitors can be generated. In summary, this work therefore pro- and were therefore also tested with NGI-1 and C19 (Figure 6D).
vides a lead compound (C19) for probing the biological effects of In this second cell line, we show that NGI-1 reduces COX2 glyco-
STT3B inhibition.
sylation without affecting protein levels but that C19 again
increases COX2 protein concomitant with inhibition of glycosyl-
ation. To understand the functional consequences of disrupting
STT3B-dependent COX2 glycosylation, we measured levels of
Regulation of Protein Function by STT3A and STT3B
Glycosylation
The EGFR is a heavily glycosylated receptor tyrosine kinase, and prostaglandin E2 (PGE2), the product of COX2 activity. C19 in-
we have previously demonstrated that disrupting EGFR N-linked creases PGE2 levels after 48 hr of treatment (Figure 6E). Together
glycosylation with NGI-1 alters receptor localization and function these data demonstrate that selective inhibition of the STT3B
(Lopez-Sambrooks et al., 2016). We therefore evaluated EGFR catalytic subunit can specifically regulate the biological activity
expression and function in the setting of STT3A-ko and of an STT3B-dependent protein such as COX2.
STT3B-ko, alone or in combination with small-molecule OST in-
hibitor treatment (Figure 6A). Both EGFR protein levels and EGF- Activation of the Unfolded Protein Response by Small-
stimulated kinase activation were reduced in STT3A-ko cells; Molecule OST Inhibitors
however, there was no appreciable effect on EGFR in STT3B- Glycosylation of proteins is required for appropriate folding and
ko cells. NGI-1, which partially blocks both STT3A and STT3B, transit inside the cell. Proteins that are not properly glycosylated
reduced EGFR protein levels and ligand stimulation of kinase ac- often fail to reach their native conformation and do not pass the
tivity in parental cells. C19 did not reduce glycosylation and ER-glycoprotein quality control (Helenius and Aebi, 2004). Accu-
ligand stimulation of EGFR in parental and STT3B-ko cells. How- mulation of misfolded proteins activates the unfolded protein
ever, in STT3A-ko cells, where EGFR glycosylation is dependent response (UPR) pathway, which helps cells adapt to perturba-
on STT3B alone, C19 did reduce ligand stimulated activation of tions in protein folding (Hetz, 2012). To investigate the effects
6
Cell Chemical Biology 25, 1–11, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
A
B
C
Figure 5. STT3A and STT3B Catalytic Subunit Inhibition by Small Molecules
(A) Structures of NGI-1, C2, C4, and C19.
(B and C) Dose response (B) and time course (C) for inhibition of glycosylation of Halo3N in stably transfected STT3A-ko and STT3B-ko cells. Cells in (B) were
treated with the indicated concentrations for 24 hr and cells in (C) were treated with 10 mM of respective compounds for the indicated periods of time before
western blot analysis. Arrows on the right indicate the different glycoforms of Halo3N (0n–3n).
of OST inhibitors on UPR activation, we conducted an analysis of icity of NGI-1 analogs by measuring induction of cell membrane
BiP induction and XBP1 mRNA splicing (Figures 7A and 7B). permeability with propidium iodide (PI) staining by flow cytome-
Tunicamycin applied at 1 mM for up to 24 hr induced BiP and try. Treatment of HEK293 cells for 24–48 hr with 10 mM C2, C4,
XBP1 mRNA splicing, and NGI-1 moderately activated both C18, or C19 showed no significant increase of cell membrane
BiP induction and XBP1 mRNA splicing. In contrast, C19 did permeability over DMSO-treated controls (Figure 7C). In com-
not significantly activate any markers of the UPR. None of these parison, tunicamycin treatment induced toxicity in ꢀ25% of cells
inhibitors stimulated ATF6 cleavage, although changes in glyco- after 48 hr.
sylation were noted for tunicamycin and NGI-1 treatment (Fig-
ure S3). Together these results demonstrate the graded effects DISCUSSION
of aberrant glycosylation on the UPR and suggest that inhibition
of subsets of N-glycosylation sites can be achieved without N-linked glycosylation is a major co- and post-translational
global induction of the UPR. We also directly examined cytotox- modification of the secretory compartment in eukaryotic cells.
Cell Chemical Biology 25, 1–11, October 18, 2018
7

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Figure 6. Regulation of EGFR and COX2
A
Function by STT3A- and STT3B-Dependent
Glycosylation
(A) Parental HEK293 cells, STT3A-ko, and STT3B-
ko HEK293 cells were treated for 48 hr with 10 mM
NGI-1 (N) or 10 mM C19 and stimulated for 10 min
with 0–25 ng/mL EGF. EGFR expression and
activation were analyzed by western blot.
(B) Diagram of COX2 with its four N-glycosylation
sites. Glycosylation of C-terminal N580 regulates
COX2 turnover.
(C and D) Parental HEK293 cells, STT3A-ko cells,
and STT3B-ko HEK293 cells (C), and U87 cells (D)
were treated for 48 hr with 10 mM NGI-1 (N) or
10 mM C19, and COX2 expression and glycosyla-
tion were analyzed by western blot. Arrows on the
right indicate the different glycoforms of COX2.
(E) U87 cells were treated for 48 hr with 10 mM
NGI-1 or 10 mM C19 and PGE2 levels of the su-
pernatant were quantified by ELISA. The graph
represents the mean and SD from three indepen-
dent experiments. D, DMSO.
B
C
or both OST catalytic subunits (Cherepa-
nova and Gilmore, 2016). Importantly, the
demonstrated inhibitory effect of NGI-1
on glycosylation in the STT3A-ko and
STT3B-ko lines provides definitive evi-
dence that the small molecule targets
both OST complexes. Because NGI-1 re-
duces glycosylation in both STT3A-ko
and STT3B-ko cell lines, the results
eliminate the possibility that the partial
inhibitory effect of NGI-1 is due to complex
interactionscausedbyinhibitionofasingle
catalytic subunit. Furthermore, the use of
multiple glycoprotein substrates in this
STT3A-ko and STT3B-ko model has pro-
D
E
Inhibition of N-linked glycosylation by either gene deletion or vided the insight that NGI-1 does not completely block catalytic
pharmacologic inhibition was long considered to be an all-or- subunit activity. While previous data had suggested that NGI-1
nothing lethal event. However, more recently the demonstration completely blocks STT3B and partially blocks STT3A, the data
of partially overlapping functions of the OST catalytic subunits provided herein show that NGI-1 is a partial inhibitor of both cat-
(STT3A and STT3B [Ruiz-Canada et al., 2009]) and the incom- alytic subunits. The use of OST catalytic subunit-deficient cells
plete effects of a small-molecule inhibitor of the OST (NGI-1 has therefore provided a new mechanistic understanding of the
[Lopez-Sambrooks et al., 2016]) have demonstrated that site inhibitor; NGI-1 increases OST sequon skipping.
occupancy of N-linked glycosylation can be regulated through
In the absence of a crystal structure of the mammalian OST
partial and selective inhibition of OST function. In this work, complex, the inhibitor’s mechanism of action at the atomic level
we investigated the structure-activity relationships for analogs cannot be delineated. However, our work does allow us to spec-
of the NGI-1 chemical series on the OST catalytic subunits using ulate on how NGI-1 could be affecting OST enzymatic activity.
a genetically defined cell system with deletion of either STT3A or The observation that glycosylation of individual sequons is differ-
STT3B. Our results demonstrate the ability to regulate N-linked entially inhibited and that neither STT3A nor STT3B can be
glycosylation by altering the chemical structure of the inhibitor completely blocked by NGI-1 suggests that the inhibitor either
such that glycosylation of target sequons can be pharmacolog- disrupts a domain required for sequon recognition or modifies
ically controlled. Furthermore, we define structural characteris- enzyme kinetics through an allosteric mechanism. Chemical
tics of STT3B-specific inhibitors and demonstrate the functional thermostabilization experiments using a catalytically inactive
consequences for reduced glycosylation of a STT3B target STT3B or an STT3B with mutation to the peptide acceptor site
protein, COX2.
were no different from endogenous STT3B, showing that neither
The knockout of STT3A or STT3B in HEK293 cells is the ideal catalytic activity nor an intact peptide binding site are required
model system to investigate the specificity of small-molecule for NGI-1 binding and OST inhibition. The recent observation
OST inhibitors as it provides cellular systems with either a single that NGI-1 also blocks flavivirus replication that is dependent
8
Cell Chemical Biology 25, 1–11, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
2006). Often designated as the subunit responsible for post-
translational glycosylation, STT3B glycosylates sites skipped
by STT3A including sites within five residues of the signal
sequence cleavage site, closely spaced NXS sites, internal
acceptor sites with suboptimal sequons, and sites within the
C-terminal ꢀ65 residues of a protein (Shrimal et al., 2013). C19
was used to examine the effects of STT3B-specific inhibition
on two glycoproteins, EGFR and COX2. In contrast to NGI-1,
which reduces EGFR glycosylation and cell surface expression,
C19 does not interfere with EGFR glycosylation or activation un-
less analyzed in STT3A-deficient cells. These results suggest
that EGFR is highly dependent on STT3A to achieve proper
folding and cellular function. For COX2, which harbors a C-termi-
nal N-linked glycosylation site dependent on STT3B and
required for protein stability, C19 effectively reduces N-linked
glycosylation, stabilizes the protein, and leads to an overall in-
crease in COX2 activity. Taken together, these data strengthen
the concept that the function of endogenous proteins can be
modified through OST subunit-specific inhibition.
As protein glycosylation and folding are strongly intercon-
nected, we looked to elucidate the effects of small-molecule
OST inhibition on the UPR through analysis of BiP, ATF6,
and XBP1s. The results confirm no detectable UPR induction
by C19 and subtle but measureable induction of BiP and
XBP1 mRNA splicing by NGI-1. These results are consistent
Figure 7. UPR Induction by C19 and Compound Toxicity Verification
with the graded effects on protein misfolding caused by each
small molecule from this chemical series, and are dissimilar
to the marked UPR induction and cell toxicity caused by
inhibition of glycan precursor synthesis with tunicamycin.
Together the results show that small molecules such as C19
can reduce N-linked glycosylation without significant UPR acti-
vation or cell toxicity. However, it is also important to note that
loss of N-linked glycosylation to subsets of proteins, such as
those that are targets of STT3B, could be physiologically dele-
(A and B) Parental HEK293 cells were treated with 1 mM tunicamycin (T), 10 mM
NGI-1 (N), or 10 mM C19 for increasing amounts of time, and cells were
analyzed by western blot for BiP levels (A) or by semiquantitative RT-PCR for
XBP1s mRNA levels (B).
(C) HEK293 cells were treated for 24 and 48 hr with 10 mM NGI-1, C2, C4, C18,
and C19, and toxicity was assessed by verification of percentage of PI-positive
cells. As a control, HEK293 cells were incubated for 30 min in ice-cold 70%
ethanol ((+)-control) or were treated with 1 mM tunicamycin. The graph rep-
resents the mean and SD from three independent experiments. D, DMSO.
on catalytically inactive STT3A (Marceau et al., 2016; Puschnik terious; thus the effects of selective inhibition of N-linked
et al., 2017) is consistent with these findings. Together this glycosylation at the tissue and organism levels requires further
work suggests that NGI-1’s mechanism of action could be allo- investigation.
steric inhibition; however, this hypothesis will require further
investigation and identification of the NGI-1 binding site.
In conclusion, we report on a series of OST inhibitors with dif-
ferential inhibitory effects toward STT3A and STT3B. Using
The differential effect of NGI-1 analogs on the two mammalian cellular models containing OST catalytic subunit knockout, we
OST catalytic subunits presented the possibility that specific in- also further characterize the mechanism of action for this
hibitors for both STT3A and STT3B could be generated. Indeed, small-molecule series. Most importantly, we demonstrate struc-
our results show that erosion of STT3A inhibitory activity can be ture-activity relationships for modifying the small molecule’s
achieved without loss of STT3B-specific inhibition. Analysis of target, altering glycosylation of a limited number of sequons,
chemical structure-activity relationships demonstrated that all and producing discrete biological effects in the cell. Together
three R-group substituents could be modified to cause a greater this work provides a pharmacologic approach for editing
loss of STT3A inhibitory activity compared with that of STT3B. N-linked glycosylation in mammalian cells.
Examples include the substitution of 4-methoxy-aniline at R1
(C6), the introduction of bulky amines at R2 (C1, C3, and C9), SIGNIFICANCE
and the substitution of a dimethylamine (C11) or N-piperidine
(C12) for the R3 N-pyrrolidine. These findings led us to also The existing paradigm suggests that pharmacologic
explore synthesis of a reverse amide for the R1 position, a struc- inhibition of N-linked glycosylation induces ER stress, the
tural change that also affected STT3A inhibitory activity (C18) unfolded protein response, and cellular toxicity. This
and led to testing of C19. This final analog contains both the concept is largely based on over 45 years of work reported
reverse amide and a phenyl group substitution, and demon- with tunicamycin, a natural product that blocks all N-linked
strates STT3B-specific inhibition with relative preservation of glycosylation. However, the recent discovery of an oligosac-
potency.
charyltransferase (OST) inhibitor, NGI-1, has challenged this
According to phylogenetic analysis, STT3B is considered to paradigm. OST inhibition is now being investigated as a ther-
be the oldest of the OST catalytic subunits (Kelleher and Gilmore, apeutic approach for cancer therapy as well as an antiviral
Cell Chemical Biology 25, 1–11, October 18, 2018
9

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
REFERENCES
for flavivirus infection. Herein we use genetic models and a
glycoprotein reporter system to probe the activity of NGI-1
Cherepanova, N.A., and Gilmore, R. (2016). Mammalian cells lacking either the
analogs for blocking the partially redundant STT3A and
cotranslational or posttranslocational oligosaccharyltransferase complex
STT3B catalytic subunits. We demonstrate structure-activ-
display substrate-dependent defects in asparagine linked glycosylation. Sci.
ity relationships for catalytic subunit-specific inhibition
Rep. 6, 20946.
and deliver an STT3B-specific small-molecule inhibitor
Cherepanova, N.A., Shrimal, S., and Gilmore, R. (2014). Oxidoreductase activ-
(C19). We show that this small molecule reduces glycosyla-
ity is necessary for N-glycosylation of cysteine-proximal acceptor sites in gly-
tion of STT3B-dependent glycosylation sites and regulates
target glycoprotein function without induction of ER stress.
This work provides a pharmacological approach for regu-
lating glycoprotein function in mammalian cells.
coproteins. J. Cell Biol. 206, 525–539.
Contessa, J.N., Bhojani, M.S., Freeze, H.H., Ross, B.D., Rehemtulla, A., and
Lawrence, T.S. (2010). Molecular imaging of N-linked glycosylation suggests
glycan biosynthesis is a novel target for cancer therapy. Clin. Cancer Res.
16, 3205–3214.
Helenius, A., and Aebi, M. (2004). Roles of N-linked glycans in the endoplasmic
STAR+METHODS
reticulum. Annu. Rev. Biochem. 73, 1019–1049.
Hetz, C. (2012). The unfolded protein response: controlling cell fate decisions
Detailed methods are provided in the online version of this paper
and include the following:
under ER stress and beyond. Nat. Rev. Mol. Cell Biol. 13, 89–102.
€
Jafari, R., Almqvist, H., Axelsson, H., Ignatushchenko, M., Lundback, T.,
d KEY RESOURCES TABLE
Nordlund, P., and Martinez Molina, D. (2014). The cellular thermal shift assay
for evaluating drug target interactions in cells. Nat. Protoc. 9, 2100–2122.
d CONTACT FOR REAGENTS AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B Cell Lines
Kelleher, D.J., and Gilmore, R. (2006). An evolving view of the eukaryotic oligo-
saccharyltransferase. Glycobiology 16, 47R–62R.
d METHOD DETAILS
Kelleher, D.J., Karaoglu, D., Mandon, E.C., and Gilmore, R. (2003).
Oligosaccharyltransferase isoforms that contain different catalytic STT3 sub-
units have distinct enzymatic properties. Mol. Cell 12, 101–111.
B Generation of Stable Cell Lines
B Stimulation, Preparation of Cell Lysates, SDS-PAGE
and Immunoblotting
Lizak, C., Gerber, S., Numao, S., Aebi, M., and Locher, K.P. (2011). X-ray struc-
ture of a bacterial oligosaccharyltransferase. Nature 474, 350–355.
B Luciferase, Proliferation, and Viability Assays
B Semiquantitative RT-PCR
Lopez-Sambrooks, C., Shrimal, S., Khodier, C., Flaherty, D.P., Rinis, N.,
Charest, J.C., Gao, N., Zhao, P., Wells, L., Lewis, T.A., et al. (2016).
Oligosaccharyltransferase inhibition induces senescence in RTK-driven tumor
cells. Nat. Chem. Biol. 12, 1023–1030.
B Assessment of PGE2 Levels
B Cellular Thermal Shift Assay (CETSA)
B Chemical Synthesis
Marceau, C.D., Puschnik, A.S., Majzoub, K., Ooi, Y.S., Brewer, S.M., Fuchs,
G., Swaminathan, K., Mata, M.A., Elias, J.E., Sarnow, P., et al. (2016).
Genetic dissection of Flaviviridae host factors through genome-scale
CRISPR screens. Nature 535, 159–163.
d QUANTIFICATION AND STATISTICAL ANALYSIS
B Mean Glycan Calculation
SUPPLEMENTAL INFORMATION
ꢀ ꢁ
Mohorko, E., Owen, R.L., Malojcic, G., Brozzo, M.S., Aebi, M., and
Glockshuber, R. (2014). Structural basis of substrate specificity of human oli-
gosaccharyl transferase subunit N33/Tusc3 and its role in regulating protein
N-glycosylation. Structure 22, 590–601.
Supplemental Information includes three figures and one table and can be
found with this article online at https://doi.org/10.1016/j.chembiol.2018.
07.005.
´
Napiorkowska, M., Boilevin, J., Sovdat, T., Darbre, T., Reymond, J.-L., Aebi,
M., and Locher, K.P. (2017). Molecular basis of lipid-linked oligosaccharide
recognition and processing by bacterial oligosaccharyltransferase. Nat.
Struct. Mol. Biol. 24, 1100–1106.
ACKNOWLEDGMENTS
This work was funded by US NIH R01CA172391 and R01GM127383 and in
Neklesa, T.K., Tae, H.S., Schneekloth, A.R., Stulberg, M.J., Corson, T.W.,
Sundberg, T.B., Raina, K., Holley, S.A., and Crews, C.M. (2011). Small-mole-
cule hydrophobic tagging-induced degradation of HaloTag fusion proteins.
Nat. Chem. Biol. 7, 538–543.
part by
a Research Scholar grant from the American Cancer Society
(J.N.C.), by NIH RO1GM43768 (R.G.), and by NIH grants DP2 AI104557 and
U19 AI109662 (J.E.C.).
Nilsson, I., Kelleher, D.J., Miao, Y., Shao, Y., Kreibich, G., Gilmore, R., Heijne,
von, G., and Johnson, A.E. (2003). Photocross-linking of nascent chains to the
STT3 subunit of the oligosaccharyltransferase complex. J. Cell Biol. 161,
715–725.
AUTHOR CONTRIBUTIONS
N.R., J.N.C., and R.G. designed the study. C.D.M. and J.E.C. generated
STT3B mutant Huh7 cell lines. N.R. performed all other experiments. J.E.G.
and M.C.V.Z. synthesized the compounds. N.R. and J.N.C. wrote the
manuscript.
Pfeffer, S., Dudek, J., Gogala, M., Schorr, S., Linxweiler, J., Lang, S., Becker,
T., Beckmann, R., Zimmermann, R., and Fo¨ rster, F. (2014). Structure of the
mammalian oligosaccharyl-transferase complex in the native ER protein trans-
locon. Nat. Commun. 5, 3072.
DECLARATION OF INTERESTS
Pfeffer, S., Dudek, J., Schaffer, M., Ng, B.G., Albert, S., Plitzko, J.M.,
Baumeister, W., Zimmermann, R., Freeze, H.H., Engel, B.D., et al. (2017).
Dissecting the molecular organization of the translocon-associated protein
complex. Nat. Commun. 8, 14516.
J.N.C. and J.E.G. are listed as inventors in a patent application for inhibitors
described in this paper.
Received: January 17, 2018
Revised: April 12, 2018
Accepted: July 3, 2018
Published: August 2, 2018
Puschnik, A.S., Marceau, C.D., Ooi, Y.S., Majzoub, K., Rinis, N., Contessa,
J.N., and Carette, J.E. (2017). A small-molecule oligosaccharyltransferase in-
hibitor with pan-flaviviral activity. Cell Rep. 21, 3032–3039.
10 Cell Chemical Biology 25, 1–11, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Roboti, P., and High, S. (2012a). Keratinocyte-associated protein 2 is a bona
fide subunit of the mammalian oligosaccharyltransferase. J. Cell. Sci. 125,
220–232.
Shrimal, S., Cherepanova, N.A., and Gilmore, R. (2015). Cotranslational and
posttranslocational N-glycosylation of proteins in the endoplasmic reticulum.
Semin. Cell Dev. Biol. 41, 71–78.
Shrimal, S., Cherepanova, N.A., and Gilmore, R. (2017). DC2 and KCP2
mediate the interaction between the oligosaccharyltransferase and the ER
translocon. J. Cell Biol. 216, 3625–3638.
Roboti, P., and High, S. (2012b). The oligosaccharyltransferase subunits
OST48, DAD1 and KCP2 function as ubiquitous and selective modulators of
mammalian N-glycosylation. J. Cell. Sci. 125, 3474–3484.
Shrimal, S., Trueman, S.F., and Gilmore, R. (2013). Extreme C-terminal sites
are posttranslocationally glycosylated by the STT3B isoform of the OST.
J. Cell Biol. 201, 81–95.
Ruiz-Canada, C., Kelleher, D.J., and Gilmore, R. (2009). Cotranslational and
posttranslational N-glycosylation of polypeptides by distinct mammalian
OST isoforms. Cell 136, 272–283.
Smith, W.L., DeWitt, D.L., and Garavito, R.M. (2000). Cyclooxygenases: struc-
Sevigny, M.B., Li, C.-F., Alas, M., and Hughes-Fulford, M. (2006).
Glycosylation regulates turnover of cyclooxygenase-2. FEBS Lett. 580,
6533–6536.
tural, cellular, and molecular biology. Annu. Rev. Biochem. 69, 145–182.
Stanley, P. (2016). What have we learned from glycosyltransferase knockouts
in mice? J. Mol. Biol. 428, 3166–3182.
Shibatani, T., David, L.L., McCormack, A.L., Frueh, K., and Skach, W.R. (2005).
Proteomic analysis of mammalian oligosaccharyltransferase reveals multiple
subcomplexes that contain Sec61, TRAP, and two potential new subunits.
Biochemistry 44, 5982–5992.
Takatsuki, A., Arima, K., and Tamura, G. (1971). Tunicamycin, a new antibiotic.
I. Isolation and characterization of tunicamycin. J. Antibiot. (Tokyo) 24,
215–223.
Tannous, A., Pisoni, G.B., Hebert, D.N., and Molinari, M. (2015). N-linked
sugar-regulated protein folding and quality control in the ER. Semin. Cell
Dev. Biol. 41, 79–89.
Shrimal, S., and Gilmore, R. (2013). Glycosylation of closely spaced acceptor
sites in human glycoproteins. J. Cell. Sci. 126, 5513–5523.
Cell Chemical Biology 25, 1–11, October 18, 2018 11

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE
Antibodies
SOURCE
IDENTIFIER
Rabbit monoclonal anti-HaloTag
Goat polyclonal anti-Luciferase
Promega
Abcam
Cat#G9211; RRID: AB_2688011
Cat#ab81823; RRID: AB_1640671
Cat#3777; RRID: AB_2096270
Rabbit monoclonal anti-phospho-EGF Receptor
(Tyr1068) (D7A5)
Cell Signaling
Rabbit monoclonal anti-Cox2 (D5H5)
Rabbit monoclonal anti- BiP (C50B12)
Rabbit monoclonal anti-ATF6 (D4Z8V)
Rabbit monoclonal anti-Vinculin (E1E9V)
Rabbit polyclonal anti-EGF Receptor (1005)
Rabbit polyclonal anti-STT3B
Cell Signaling
Cell Signaling
Cell Signaling
Cell Signaling
Santa Cruz Biotechnology
Proteintech
Cat#12282; RRID: AB_2571729
Cat#3177; RRID: AB_2119845
Cat#65880
Cat#13901; RRID: AB_2728768
Cat#sc-03; RRID: AB_631420
Cat#15323-1-AP; RRID: AB_2198046
Cat#HRP-60004; RRID: AB_2107436
Mouse monoclonal anti-HRP
Proteintech
Chemicals, Peptides, and Recombinant Proteins
Geneticin
Tunicamycin
NGI-1
Gibco
Cat#11811031
Cat#654380
CID2519269
N/A
Calbiochem
NEDP
C18: 5-methyl-N-(5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)
phenyl)thiazole-2-carboxamide
NEDP
This paper
C19: N-(5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)phenyl)
benzamide
NEDP
N/A
This paper
Mouse EGF
Biomedical Technologies
Cat#BT-201
Critical Commercial Assays
ONE-Glo Luciferase Assay
Prostaglandin E2 ELISA Kit
Experimental Models: Cell Lines
Human: HEK293 STT3A-ko
Promega
Cat#E6120
Cat#514010
Cayman Chemical
Laboratory of R. Gilmore
N/A
N/A
N/A
N/A
N/A
Cherepanova and Gilmore 2016
Human: HEK293 STT3B-ko
Laboratory of R. Gilmore
Cherepanova and Gilmore 2016
Human: Huh7 STT3B-ko + STT3B-CM (D100A,
D218A, E402Q)
Laboratory of J. Carette
Marceau et al., 2016
Human: Huh7 STT3B-ko + STT3B-WD (mutation
of WWD peptide acceptor site to WAA)
Laboratory of J. Carette
Marceau et al., 2016
Human: U87
Laboratory of W. Cavanee
Oligonucleotides
XBP1s forward primer
50-ggtctgctgagtccgcagcagg- 30
IDT
IDT
IDT
IDT
N/A
N/A
N/A
N/A
XBP1s reverse primer
50-ggggcttggtatatatgtgg- 30
GAPDH forward primer
50-ggatgatgttctggagagcc- 30
GAPDH reverse primer
50-catcaccatcttccaggagc- 30
Recombinant DNA
pcDNA3-ER-LucT
pcDNA3-ER-Halo3N
Contessa et al., 2010
This paper
N/A
N/A
(Continued on next page)
e1 Cell Chemical Biology 25, 1–11.e1–e4, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Continued
REAGENT or RESOURCE
Software and Algorithms
Image Lab 6.0
SOURCE
IDENTIFIER
Bio-Rad
Yale ITS
http://www.bio-rad.com/de-de/product/
image-lab-software?ID=KRE6P5E8Z
Prism 6 Data Analysis Software
N/A
CONTACT FOR REAGENTS AND RESOURCE SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact and Cor-
responding Author, Joseph N. Contessa (joseph.contessa@yale.edu).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Cell Lines
STT3A-ko and STT3B-ko HEK293 cells were maintained in DMEM media supplemented with 5% FCS (Seradigm, VWR) and 1% Pen
Strep at 37ꢁC in a humidified incubator with 5% CO2. HEK293 cells stably transfected with ERLucT or Halo3N were maintained in
media containing 800 ng/ml Geneticin.
Huh7 STT3B-ko cells complemented with the catalytically inactive mouse STT3B-CM (mutations D100A, D218A, E402Q) or the
peptide binding deficient mouse STT3B-WD (mutation of WWD peptide acceptor site to WAA) were maintained in DMEM media sup-
plemented with 10% FCS (Gibco) and 1% Pen Strep at 37ꢁC in a humidified incubator with 5% CO2.
HEK293 cells are female whereas Huh7 and U87 cells were isolated from male patients. None of these cells have been
authenticated.
METHOD DETAILS
Generation of Stable Cell Lines
Transfections were performed with Lipofectamine 3000 (Thermo Fisher Scientific). Selection of HEK293 cells stably transfected with
ERLucT or Halo3N was performed with 1600 ng/ml Geneticin. Single clones were picked and analyzed for reporter expression
through western blot.
Stimulation, Preparation of Cell Lysates, SDS-PAGE and Immunoblotting
EGF stimulation was performed with EGF from the mouse submaxillary gland (Biomedical Technologies). Upon treatment, cells were
washed with ice-cold PBS and lysed with lysis buffer containing 1% Triton X-100, 25 mM Tris-HCl pH 7.4, 10 mM EDTA, 15% glyc-
erol, protease inhibitor tablet (Roche) and phosphatase inhibitor cocktails 2 and 3 (Sigma-Aldrich). Lysis was performed for 30 min on
ice. Protein concentration was measured using the BCA protein assay (Pierce). Lysates were subjected to SDS-PAGE, western blot
and immunodetection using the antibodies mentioned in the Key Resources Table. Horseradish-peroxidase conjugated secondary
antibodies were purchased from Millipore and proteins were detected by enhanced chemiluminescence (ECL, GE Healthcare).
Luciferase, Proliferation, and Viability Assays
Luciferase assays were performed with the ONE-Gloꢀ Luciferase Assay (Promega) and proliferation assays with the CellTiter 96ꢁ
Non-Radioactive Cell Proliferation Assay (Promega) according to manufacturer’s instructions. Both assays were performed with an
initial number of 1000 cells/96well and measurements were performed in triplicate. For analysis of viability with propidium iodide (Pi)
staining, cells were detached from the plates using trypsin, washed 3 times with PBS supplemented with 5% FBS, resuspended in
200 ml of the same buffer containing 10 mg/ml Pi, and analyzed by flow cytometry on a BD LSRII. As a positive control cells were incu-
bated in 70% ice-cold ethanol for 30 min before performing Pi staining.
Semiquantitative RT-PCR
Upon treatment, RNA-isolation was performed with the RNeasy Plus Mini Kit (Qiagen). Reverse transcription was performed with the
ProtoScriptꢁ First Strand cDNA Synthesis Kit (NEB). Semiquantitative PCR was performed with Taq DNA polymerase
(ThermoFisher) and XBP1s/GAPDH-specific primers that are listed in the Key Resources Table.
Assessment of PGE2 Levels
For PGE2 level assessment 1.6*106 U87 cells were seeded on a 6 cm dish. Upon treatment, PGE2 levels of supernatants were
measured using the PGE2 ELISA Kit from Cayman Chemical according to manufacturer’s instructions. Measurements were per-
formed in quadruplicate.
Cell Chemical Biology 25, 1–11.e1–e4, October 18, 2018 e2

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
Cellular Thermal Shift Assay (CETSA)
Drug-target engagement was assessed using CETSA as previously described (Jafari et al., 2014). Briefly, cells were treated with
100 mM NGI-1 for 30 min, then harvested and resuspended in RPMI medium supplemented with 100 mM NGI-1. Resuspended cells
were subjected to thermal treatment for 3 min and were lysed subsequently. Lysis and western blot analysis of the soluble fraction
were performed as described above.
Chemical Synthesis
C18
Chemical name: 5-methyl-N-(5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)phenyl)thiazole-2-carboxamide.
A solution of 4-fluoro-3-nitrobenzene-1-sulfonyl chloride (500 mg, 2.40 mmol) and diisopropylethylamine (0.41 mL, 2.50 mmol) in
tetrahydrofuran (5.5 mL) was cooled to 0ꢁC and treated with morpholine (0.2 mL, 2.29 mmol). After stirring for 1 h, the cooling bath
was removed with stirring continuing at room temperature for 16 h. Once the reaction was complete, the mixture was diluted with
dichloromethane and washed successively with 0.5 N HCl and saturated aqueous sodium chloride. The organic phase was dried
over MgSO4, filtered and concentrated. Purification by flash column chromatography (10–40% ethyl acetate in hexanes) gave
4-((4-fluoro-3-nitrophenyl)sulfonyl)morpholine (381 mg, 63%) as a pale-yellow solid.
A solution of 4-((4-fluoro-3-nitrophenyl)sulfonyl)morpholine (185 mg, 0.637 mmol) and diisopropylethylamine (0.25 mL, 1.53 mmol)
in dimethylformamide (3.7 mL) was cooled to 0ꢁC and treated with pyrrolidine (0.057 mL, 0.70 mmol). After stirring for 1 h, the cooling
bath was removed with stirring continuing at room temperature for 16 h. Once the reaction was complete, the mixture was diluted
with dichloromethane and washed successively with 0.5 N HCl and saturated aqueous sodium chloride. The organic phase was dried
over MgSO4, filtered and concentrated. The crude product was triturated from diethyl ether to give 4-((3-nitro-4-(pyrrolidin-1-yl)
phenyl)sulfonyl)morpholine (136 mg, 63%) as a pale-yellow solid.
A solution of 4-((3-nitro-4-(pyrrolidin-1-yl)phenyl)sulfonyl)morpholine (140.7 mg, 0.412 mmol) and Pd-C (4.19 mg, 41.2 mmol) in
ethyl acetate / ethanol was degassed with a nitrogen stream and treated with hydrogen gas (balloon) with stirring at room temperature
overnight. Once complete, the reaction mixture was filtered through Celite and washed with ethyl acetate to give 5-(morpholinosul-
fonyl)-2-(pyrrolidin-1-yl)aniline (120 mg, 93%) as a yellow solid that was used in the subsequent step without further purification.
A
solution of 5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)aniline (96.1 mg, 0.339 mmol), 5-methyl-1,3-thiazole-2-carboxylic
acid (48.6 mg, 0.34 mmol), triethylamine (0.129 mL, 0.926mmol), 4’-N,N-dimethylaminopyridine (1.88 mg) and HOBt (2.08 mg) in di-
chloromethane (4 mL) was treated with EDC (118.3 mg, 0.617 mmol) and stirred at room temperature overnight. Once complete, the
e3 Cell Chemical Biology 25, 1–11.e1–e4, October 18, 2018

Please cite this article in press as: Rinis et al., Editing N-Glycan Site Occupancy with Small-Molecule Oligosaccharyltransferase Inhibitors, Cell Chem-
ical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.07.005
reaction mixture was diluted with dichloromethane and washed successively with saturated aqueous NaHCO3 and saturated
aqueous sodium chloride. The organic phase was dried over MgSO4, filtered and concentrated. The crude product was purified
by flash column chromatography (10-30% ethyl acetate in hexanes) to give 5-methyl-N-(5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)
phenyl)thiazole-2-carboxamide (18 mg, 13% yield) as an off-white solid.
C19
Chemical name: N-(5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)phenyl)benzamide
A solution of 5-(morpholinosulfonyl)-2-(pyrrolidin-1-yl)aniline (28.0 mg, 0.090 mmol) in pyridine (1.64 mL) was cooled to 0ꢁC and
treated with benzoyl chloride (0.010 mL, 0.090 mmol). After the addition was complete, the cooling bath was removed allowing
the reaction to warm to room temperature. After 3 h, the reaction mixture was diluted with dichloromethane (10 mL), washed
successively with 0.5 N HCl and saturated aqueous sodium chloride. The organic phase was dried over MgSO4, filtered and concen-
trated. Purification by flash column chromatography (10-50% ethyl acetate in hexanes) gave N-(5-(morpholinosulfonyl)-2-(pyrrolidin-
1-yl)phenyl)benzamide (15 mg, 40%) as an off-white solid.
QUANTIFICATION AND STATISTICAL ANALYSIS
Mean Glycan Calculation
Quantification of glycoforms was performed with the Image Lab software from Bio-Rad. Adjusted volume values (ADV) were
multiplied with the number of glycans present in each glycoform (0, 1, 2 or 3) and the resulting signal was divided by the overall signal
of glycoforms for each treatment. Addition of the relative signals from all glycoforms resulted in a compound-specific glycosylation
coefficient representing the glycosylation status of the reporter.
Cell Chemical Biology 25, 1–11.e1–e4, October 18, 2018 e4