Immunoproteasome inhibition and bioactivity of thiasyrbactins

Article abstract of DOI:10.1016/j.bmc.2017.11.048

A family of macrodilactam natural products, the syrbactins, are known proteasome inhibitors. A small group of syrbactin analogs was prepared with a sulfur-for-carbon substitution to enhance synthetic accessibility and facilitate modulation of their solubility. Two of these compounds surprisingly proved to be inhibitors of the trypsin-like catalytic site, including of the immunoproteasome. Their bound and free conformations suggest special properties of the thiasyrbactin ring are responsible for this unusual preference, which may be exploited to develop drug-like immunoproteasome inhibitors. These compounds show greater selectivity than earlier compounds used to infer phenotypes of immunoproteasome inhibition, like ONX-0914.

Full text of DOI:10.1016/j.bmc.2017.11.048

Accepted Manuscript
Immunoproteasome Inhibition and Bioactivity of Thiasyrbactins
Nicole A. Bakas, Chad R. Schultz, Lisette P. Yco, Christopher C. Roberts, Chia-
en A. Chang, André S. Bachmann, Michael C. Pirrung
PII:
S0968-0896(17)31871-0
https://doi.org/10.1016/j.bmc.2017.11.048
BMC 14101
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
22 September 2017
21 November 2017
29 November 2017
Please cite this article as: Bakas, N.A., Schultz, C.R., Yco, L.P., Roberts, C.C., Chang, C.A., Bachmann, A.S.,
Pirrung, M.C., Immunoproteasome Inhibition and Bioactivity of Thiasyrbactins, Bioorganic & Medicinal
Chemistry (2017), doi: https://doi.org/10.1016/j.bmc.2017.11.048
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Immunoproteasome Inhibition and Bioactivity of
Thiasyrbactins
Nicole A. Bakas,1 Chad R. Schultz,2 Lisette P. Yco,2 Christopher C. Roberts,1 Chia-en A. Chang,1 André
S. Bachmann,*,2,4 and Michael C. Pirrung*1,3,4
1Department of Chemistry, University of California, Riverside, CA 92521, USA.
2Department of Pediatrics and Human Development, College of Human Medicine, Michigan State
University, Grand Rapids, MI 49503, USA.
3Department of Pharmaceutical Sciences, University of California, Irvine, CA 92697 USA.
4Equal contribution.
ABSTRACT
A family of macrodilactam natural products, the syrbactins, are known proteasome inhibitors. A small
group of syrbactin analogs was prepared with a sulfur-for-carbon substitution to enhance synthetic
accessibility and facilitate modulation of their solubility. Two of these compounds surprisingly proved
to be inhibitors of the trypsin-like catalytic site, including of the immunoproteasome. Their bound and
free conformations suggest special properties of the thiasyrbactin ring are responsible for this unusual
preference, which may be exploited to develop drug-like immunoproteasome inhibitors. These
compounds show greater selectivity than earlier compounds used to infer phenotypes of
immunoproteasome inhibition, like ONX-0914.
Keywords: proteasome; macrolactam; conformational analysis; trypsin-like site; neuroblastoma
2

Corresponding
Authors.
*Tel.:
*Tel.:
(951)
(616)
827-2722.
234-2841.
Fax:
Fax:
(951)
(616)
827-2749.
234-2838.
E-mail:
E-mail:
Michael.pirrung@ucr.edu.
andre.bachmann@hc.msu.edu
1. Introduction
The syrbactins are a family of bacterial, macrocyclic, non-ribosomal peptide natural products active
against the mammalian proteasome.1 Their -unsaturated lactam functionality reacts covalently with
the catalytic Thr1 residue of proteasome catalytic subunits. Given the clinical success of the proteasome
inhibitors bortezomib2 (BTZ), ixazomib, and carfilzomib3 against cancers such as multiple myeloma and
mantle cell lymphoma,4 intense investigations into the syrbactins followed their initial discovery.5,6
Unlike some other proteasome inhibitors, even some used clinically, the syrbactin syringolin A (SylA) is
quite selective for proteasome  subunits over other protein targets, as demonstrated by affinity-based
protein profiling.7,8,9 Several syntheses of the natural syrbactins themselves have been reported, and
significant work to prepare analogs has been pursued.10-13 We recently reported the compound TIR-199
(Chart 1),14 the first syrbactin to demonstrate activity against tumor cell lines in animal studies. It is most
potent against the chymotrypsin-like activity of the 5 constitutive proteasome subunit, which can be
ascribed to structural features derived from other syrbactins including the natural product glidobactin A,
which was discovered through its intrinsic activity against tumor cell lines.15 Its key structural features
include the macrolactam methyl group and the long unbranched, unfunctionalized side chain. A straight-
chain alkyl urea side chain similar to that of glidobactin A was originally applied to syrbactin analogs by
Kaiser.11 Maintaining this feature in the analogs from which TIR-199 emerged enabled SAR for
syrbactin macrolactam cores to be readily discerned. Pre-clinical profiling assays of TIR-199 showed it
has low aqueous solubility without co-solvents, hampering studies of bioavailability and cellular
transport. Approaches to enhance the solubility of syrbactins were therefore of high priority. Both this
goal and the desire for synthetic simplification from the 10-step synthesis of TIR-199 drove the study
described herein.
3

Chart 1. TIR-119, the approved drug bortezomib (Velcade), and the natural product glidobactin A.
The design of the current family of syrbactins referred to informally as NAMs was based on our
concise syringolin B analog synthesis13 that can be applied to any lysine analog. Here it exploits the
commercial compound -aminoethylcysteine, or thialysine. Its sulfur was envisioned to provide a
synthetic handle to enable late-stage strategies to enhance solubility. Its TIR-199-based macrodilactam
can be accessed synthetically in only a few steps. The specific compounds prepared in this study are
shown in Chart 2. Compounds 1 and 2 are inspired by TIR-199, including most of its structure but
replacing the isolated alkene. The sulfoxide in 2 is at the same position as the alcohol of glidobactin A
and could replicate its stereochemistry. Compounds 3 and 4 are inspired by bortezomib, in that it is a
dipeptide with an electrophilic C-terminus and an N-terminal pyrazine. The same spatial relationship
between the electrophilic carbon and the pyrazine seen in bortezomib is found in the unsaturated amide
of 3 and 4. Compound 5 retains the core of 1 and 2 but adds a branched, saturated, chiral carbon in the
side chain, one strategy recommended to enhance solubility.16-18 Compound 6 likewise retains the
macrolactam-valine but terminates it with the more polar pyrazinamide. All were evaluated
computationally for their physicochemical properties that affect drug-likeness.19,20 These results are
summarized in Table 1.
4

Chart 2. Synthetic syrbactin analogs prepared in this study.
Table 1. Properties affecting thiasyrbactin drug-likeness including solubility, hydrophobicity, and
molecular weight.
a
Compound NAM # Mr (Da) log Pa Log Sb Log Sw
TIR-199
-
533
520
553
569
5.95 -4.51 -5.91
3.12 -3.89 -4.24
5.51 -4.65 -5.75
3.89 -4.36 -4.83
GlbA
-
105
135
1
2
5

41
111
93
349
365
531
448
-0.76 -3.03 -1.50
-2.38 -2.91 -0.58
3.86 -4.52 -5.09
-0.05 -3.56 -2.33
3
4
5
6
95
aCalculated using the method in ref. 19.bCalculated using the method in ref. 20.
2. Results
2.1. Synthesis
The synthesis of the core macrodilactam common to all of the targets prepared here is summarized in
Scheme 1. The commercial aminoethylcysteine hydrochloride (7, thialysine), was converted in one pot
to the -phosphonoacetatamide/-Boc derivative 9 in 83% yield. Small quantities of the doubly Boc-
protected and -phosphonoacetatamide/-Boc amino acids were also formed, and experimental
optimization was necessary to reduce their levels. Conventional carbodiimide coupling with alaninol
gave 10 in 68% yield and set the stage for macrocycle formation. Dess-Martin oxidation of 10 gave an
intermediate aldehyde that was neither purified nor handled unnecessarily. We have applied the Horner-
Wadsworth-Emmons condensation as modified by Helquist21 to the preparation of dozens of
macrolactam analogs in the syrbactin family. This is a highly reliable process that gives superior yields
in favorable circumstances. Here, macrolactam 11 was delivered in 58% yield.
6

Scheme 1. Synthesis of key macrodilactam.
Oxidation of 11 with sodium periodate gave sulfoxide 12 as a single stereoisomer (Chart 3) (89%).
The stereochemistry shown is supported by spectroscopic analysis (vide infra).
Chart 3. Thiasyrbactin sulfoxide R configuration assigned to compound 12 derived from 11.
Table 2. NMR Data for Macrodilactams 11 and 12
Compound 11
Compound 12
b
b
H , mult.
H , mult.
a
C
a
C
(J, Hz)
(J, Hz)
Position
COSY
NOESY
6
COSY
NOESY
6
1-S
2
32.4
42.0
a: 2.47, m
2b, 3a, 3b
2a, 3a, 3bd
54.5c
37.4
a: 3.29, m
2b, 3a, 3b
2a, 3a, 3b
b: 2.70, ddd
(5.1, 8.6, 14.0)
b: 2.96, ddd
(7.2, 8.9, 13.9)
3
a: 3.50, m
2a, 2b, 3b
12a
a: 3.70, m
2a, 2b, 3b
12a
7

b: 3.22d, m
2a, 2b, 3a
12b
b: 3.48, ddd
(4.5, 6.9, 16.1)
5
6
7
170.8
121.3 6.45, d (15.7)
169.7
120.0 6.15, d (15.4)
7
2a, 12a
7, 8
6, 8
2a, 12a
146.0
6, 8
149.5
6.72, dd (4.9,
15.6)
6.92, dd (4.8,
15.4)
8
8-Me
10
48.0
18.8
172.7
53.8
34.1
4.52, m
7, 8-Me
47.9
18.7
4.57, m
6, 7, 8-Me
8
1.33, d (7.1)
1.33, d (7.1)
170.9
52.9
11
4.43, m
12a, 12b
11, 12b
4.77, t (3.5)
12a, 12b
11, 12b
12
3a, 6
54.5c
3a, 6
3b
a: 3.00, dd
(6.1, 14.4)
a: 3.14, dd
(3.8, 14.4)
b: 3.22d, m
11, 12a
b: 3.77, dd
(3.2, 14.4)
2’
4’
5’
157.2
81.1
28.8
157.1
81.4
28.8
1.45, s
8
1.45, s
8
b
c
d
aRecorded at 126 MHz. Recorded at 700 MHz. The carbon signals 2 and 12 of compound 12 overlap. The
proton signals of 3b and 12b in compound 11 overlap.
Stereochemical/conformational analysis of the macrolactam relied on extensive NMR spectroscopy.
Data collected on 11 and 12 included high field carbon, proton, COSY and NOESY spectra. They are
reported in Table 2. The C2-C3 protons of 11 form an AA'BB' system and the C11-C12 protons form an
ABB' system. The three NH signals were not observed because spectra were obtained in d4-methanol.
We developed a conformational model of 11 to aid in understanding its chemical and biological
properties and assigned the sulfoxide configuration of 12 using this data. NOESY data were used to
identify proximal transannular hydrogens, which exerted substantial constraint on the available
conformations. Based on molecular modeling (Spartan MMFF), there are two main low-energy
conformations available to 11, with the sulfide pointing above or below the nominal plane of the ring.
The macrodilactam conformation observed (Figure 1) has the sulfide below the plane and has NOE
8

correlations between H-6 and both H-2a and H-12a. The similar coupling constants between H-11 and
both H-12a and H-12b in 11 further support this conformation. If the sulfide was above the plane, the
protons on carbon 12 would point down, which would cause the dihedral angles between H-11 and H-
12a/12b to be closer to 135°. This would result in a larger coupling constant than is observed.
Additionally, this structure would lack NOE correlations between the protons on carbons 3 and 12.
1
Analysis of the H NMR spectra of 12 enabled the conformation and sulfoxide configuration to be
confirmed. All adjacent protons are deshielded in the sulfinyl derivative, but the relationships among
hydrogens, both in terms of dipolar and scalar couplings, are very similar to 11, supporting a similar
conformation. Table S1 compares the modeled dihedral angles and those implied from J couplings.
Compound 12 showed additional NOE correlations between H-3b and H-12b, and H-3a and H-12a,
strengthening the case for the ‘sulfide below’ conformation with slight inward rotation of C-3. The
greatest deshielding is experienced by H-2a, which has a dihedral angle with the sulfoxide of 162°. This
relationship provides the greatest downfield shift when sulfides are converted to sulfoxides.22 Least
hindered approach of the reagent from the equatorial direction, as is commonly observed in medium
rings, would yield the sulfinyl orientation depicted.
9

Figure 1. The conformation of macrodilactam 11 consistent with nOes and dihedral angles. [color
figure]
Figure 2. The conformation of macrodilactam 12 consistent with nOes and dihedral angles. [color
figure]
Continuing the synthesis from 11, the Boc group was removed with acid and the resulting
hydrochloride salt was neutralized with a carbonate ion-exchange resin, in line with our past work.13 The
free base was then coupled with a valine active ester to provide 13. (60%) Removal of its Fmoc group
and urea formation with dodecyl isocyanate provided the first target 1. (75%) Periodate oxidation
yielded the sulfoxide 2 in a single stereoisomeric form, which based on spectral similarity to 12 and the
precedent of its formation in the same type of reaction is assigned the R configuration.
10

Scheme 2. Synthesis of TIR-199 analogs 1 and 2
The other targets required active esters 14 and 15, which were prepared conventionally from the
corresponding acids.
Chart 3. Side chain active esters.
Macrolactam 11 was again deprotected and converted to the free base. It was coupled with pyrazine
1
carboxylate to yield 3 (48%) that was oxidized to yield 4 as a single stereoisomer. The H NMR
properties of 4 were very similar to those of 12, and therefore its sulfoxide configuration was likewise
assigned the R stereochemistry. Valine-substituted macrolactam 12 was the starting point for the
preparation of 5 and 6 using straightforward methods.
Scheme 3. Synthesis of TIR-199 analogs 3-6
2.2. Biological properties
11

The thiasyrbactins were evaluated for their ability to inhibit each of the catalytic subunits of the
constitutive and immunoproteasomes as well as for cytotoxic activity against neuroblastoma cancer
cells. As 2-4 had no substantial activity in these assays, they will not be discussed.
To test the inhibitory activity of 1, 5, and 6, we performed an in vitro assay and measured the three
sub-catalytic activities (i.e., the CT-L, C-L, and T-L activities) of the constitutive proteasome and the
immunoproteasome (Table 3). For control comparison, we included the FDA-approved peptide boronic
acid proteasome inhibitor bortezomib (BTZ) and the immunoproteasome inhibitor ONX-0914 (a peptide
epoxyketone related to carfilzomib) using identical experimental conditions.
While 1, 5, and 6 predominantly inhibit the CT-L and T-L activities of the constitutive proteasome,
these compounds more potently and more selectively inhibit the T-L activity of the immunoproteasome,
with no effect on the CT-L or C-L activities. It was initially surprising that inhibition of the T-L activity
by 5 and 6 appears to stimulate the CT-L activity of the immunoproteasome (Figure S1). However, such
behavior has been previously observed, including in actual patient samples.23 In contrast, BTZ inhibits
the CT-L and C-L sub-catalytic activities indiscriminately, with weaker activities against the T-L
activity. The immunoproteasome inhibitor ONX-0914 exhibits higher potency against the
immunoproteasome CT-L activity compared to the CT-L activity of the constitutive proteasome, but
also inhibited the C-L and T-L activities of both proteasome types at higher concentrations. Thus, 1, 5,
and 6 more selectively target the T-L site with Ki-50 values at 1.39 μM, 0.74, and 0.75, respectively,
with no effect at ≤10 μM on the C-L and CT-L activities of the immunoproteasome. In summary, all
three compounds show comparable behavior with single digit micromolar inhibition of both 2 sites in
addition to the 5 site of the constitutive proteasome that was expected based on the properties of TIR-
199.
Table 3. In vitro effects of thiasyrbactins on sub-catalytic activities of the constitutive proteasome and
the immunoproteasome.
12

Constitutive Proteasome
Immunoproteasome


Ki-50 (µM)
0.28
>10
SD
Ki-50 (µM)
>10
SD
-
0.07
-
1
CT-L (5)
C-L (1)
T-L (2)
CT-L (5)
C-L (1)
T-L (2)
CT-L (5)
C-L (1)
T-L (2)
CT-L (5)
C-L (1)
T-L (2)
CT-L (5)
C-L (1)
T-L (2)
>10
-
1.76
0.91
>10
0.78
0.18
-
1.39
>10
0.11
-
5
>10
-
1.07
1.22
>10
0.19
0.18
-
0.74
>10
0.06
-
6
>10
-
0.92
0.56
2.07
1.11
0.01
0.02
1.37
0.12
0.19
0.30
0.06
0.00
0.00
0.11
0.75
0.11
1.32
1.34
0.01
0.03
0.78
0.03
0.02
0.21
0.22
0.00
0.01
0.03
ONX-0914
BTZ
To study the biological effects of these thiasyrbactins, we tested their action on the viability of human
neuroblastoma cells. As shown in Figure 3, 5 and 6 exhibit minimal effects on three neuroblastoma cell
lines MYCN2, SK-N-Be(2)c, and SK-N-SH at the highest concentration tested (1 μM) after 24 hours
exposure. In contrast, 1 significantly inhibits the viability of MYCN2 and SK-N-SH cells at ~30-50% of
control (DMSO). TIR-199 strongly inhibits viability of MYCN2 and SK-N-SH cells as previously
reported14 with little effect on SK-N-Be(2)c, a cell line originally derived from a highly metastatic,
MYCN-amplified neuroblastoma tumor that exhibits chemoresistance. BTZ exhibits strong cytotoxic
effects most prominently in MYCN2 and SK-N-SH cells at 0.1-1 μM.
13

14

Figure 3. Effect of thiasyrbactins on the viability of human neuroblastoma cancer cells. (A) MYCN2,
(B) SK-N-Be(2)c, and (C) SK-N-SH neuroblastoma cells were treated with 1, 5, or 6 for 24 hours at
three different concentrations (0.01, 0.1, and 1 μM). Bortezomib (BTZ) and TIR-199 were included as
positive controls and compound 2 as a negative control. Cell viability was measured using the MTS
reagent as described in the Methods section. Data shown are representative of two independent
experiments, each performed in triplicate wells (n=6).
2.3. Molecular dynamics simulation
The actions of these compounds against the trypsin-like 2 subunits were interesting, particularly of
the immunoproteasome, as selective inhibition of this sub-site has been difficult to achieve. We
therefore aimed to gain structural understanding of their properties via modeling. Using the same
methods from our earlier work,14 binding of a thiasyrbactin to the proteasome was evaluated using
molecular mechanics/generalized Born/surface area energetic post-simulation analysis of molecular
dynamics trajectories. Due to significant folding of the ring, the conformation of 5 bound to the 2 site
has an even more exaggerated preference for the ‘sulfide below’ conformation than the free compound
(Figure 4). Binding of SylA in this site was also simulated, and it binds similar to the way it binds to the
15

5 site of the constitutive proteasome, with the macrolactam rather flat like in its free conformation.
The conformational change is quite dramatic with 5, resulting in several differences in the bound
conformation compared to the natural product. While the C5 amide is a cis amide in both, in 5 the NH
lies significantly above where it is in bound SylA. This initiates a 3-atom excursion above the area of the
corresponding ring atoms in SylA. Whereas the N9 NH in 5 projects upward, in SylA it projects
downward. Whereas the C10 carbonyl in 5 projects downward, in SylA it projects upward. The amide of
5 is thus ‘flipped’.
Figure 4. Comparison of the modeled macrolactam conformations of syringolin A (SylA) and 5 bound
to the 2 site of the immunoproteasome. SylA is rendered in thin tubes and 5 in thick tubes. The Thr1
residue of the protein is shown in wire frame. The C11 side chains were removed from each molecule
after modeling because they are quite different and obscure the rings. The inset shows a structural key
with atom numbering. For 5, R=Me, for SylA R=i-Pr. [color figure]
Examining this conformation regarding proximal residues of the proteasome (Figure 5), the ring
structure of 5 that is very different from SylA allows the compound to rearrange itself in the binding
16

pocket. The three-atom excursion mentioned above makes direct van der Waals contacts with side
chains of Ala49, Met222, and Leu344. In these MD simulations, the N9 NH of 5 forms a stable
hydrogen bond with the backbone carbonyl oxygen of Gly47 of the protein. This is the same residue that
has been assigned the role of polarizing the unsaturated amide carbonyl of the syrbactins for the
conjugate addition of the protein’s catalytic Thr1. The C11 side chain makes extensive non-polar protein
contacts. The benzene ring of the ibuprofen is also proximal to the macrolactam that may influence the
folding of this side chain against the ring. The dense network of atoms in this folded form of the
compound accounts for loss of target-site potency upon conversion of the sulfide to the sulfoxide.
Figure 5. Modeled 5 bound to the 2 site of the immunoproteasome. The compound is rendered in
Licorice, nearby protein residues are in VDW, and the remainder of the protein is in Cartoon. [color
figure]
3. Discussion
An unexpected observation was the ability of 1, 5 and 6 to inhibit the trypsin-like2 subunit of the
immunoproteasome, whereas their parent compound, TIR-199, is a selective5 subunit inhibitor of the
constitutive proteasome. While there are several reports of selective inhibitors of the immunoproteasome
17

trypsin-like subunit,24-26 most are not drug-like. The desired goals of enhanced solubility and a
streamlined synthesis for syrbactin analogues were achieved in this work. Yet, the compounds intended
to be more soluble were far less active because their protein target was unexpected and the structural
change used to increase solubility affected the interface with the protein.
In our hands, ONX-0914 is not very selective but is most potent on the immunoproteasome
chymotrypsin-like subunit. It has been used broadly in cell-based and even animal models to ascertain
the consequences of immunoproteasome inhibition for a variety of conditions.27-29 The NAMs appear to
already offer greater selectivity as well as substantial potential to be developed into very selective tool
compounds to assess the role of the immunoproteasome trypsin-like subunit in many immune disorders.
Given the preference of 1 and 5 for the immunoproteasome catalytic activity that is not found in most
cells, it was not very surprising that thiasyrbactins 5 and 6 have weak biological activity against
neuroblastoma cells. Compounds 5 and 6 are more drug-like and soluble than TIR-199, though their
current potency likely limits their utility to tool compounds that can elucidate the phenotypes of
selectively inhibiting the 2 subunit of the proteasome. However, epoxyketone compounds are known
that have selectivity somewhat like this.30 Even though they are not very potent on their own, they show
interesting synergistic activities with other proteasome inhibitors.
An important conformational change occurs in a thiasyrbactin when bound to the 2 proteasomal
subunit compared to past syrbactins. The ‘flipped’ C10 amide and the highly folded ring results in its
contacting different protein residues.
The immunoproteasome is regarded as a next-generation target for proteasome inhibitors,31 for both
autoimmune disease32,33 and hematologic malignancies.34,34 The actions of the thiasyrbactins on immune
cells should prove interesting, including those associated with tumor cell growth. Unlike bortezomib,
syrbactins block the proteasome irreversibly and, importantly, do not inhibit normal proteases,5 the latter
of which being the main reason for the reported bortezomib-induced neuropathies in patients. The
18

greater target selectivity of syrbactins compared to bortezomib implies reduced off-target activity and
more dose flexibility in animal models and patients.
4. Experimental Section
General. All reagents were purchased from commercial sources and were used without further
purification, unless otherwise stated. Chromatographic purification of products was accomplished using
flash column chromatography with silica gel 60. All melting point data were measured on a Büchi
Melting Point B-545 instrument and are uncorrected. IR spectra were obtained on a Perkin Elmer
Spectrum One FT-IR spectrometer or a Bruker Alpha FT-IR spectrometer using the ATR accessory and
are reported in absorption frequency (cm-1). The specific rotation data were measured on a Rudolph
Research Analytical AutoPol IV polarimeter. Proton (1H) and carbon (13C) NMR spectra were obtained
on Varian Inova 300 (300 MHz and 75 MHz, respectively), Varian Inova 400 (400 MHz and 101 MHz,
respectively), Varian Inova 500 (500 MHz and 126 MHz, respectively), or Bruker Avance 700 (700
MHz) spectrometers as noted, and were internally referenced to residual deuterated solvent signals and
reported in terms of chemical shift (δ ppm). Mass spectra were obtained on an Agilent G3250AA LCMS
instrument. Based on the criteria provided by Baell & Holloway,36 we believe the new compounds
assayed in this work have low potential for pan-assay interference and that assays to establish they do
not show it were unnecessary. All tested compounds exhibited >95% purity based on clean NMR spectra
without extraneous peaks.
Synthetic methods
(S)-2-(3-Dodecylureido)-3-methyl-N-((8S,11R,E)-8-methyl-5,10-dioxo-1-thia-4,9-diazacyclo dodec-6-
en-11-yl)butanamide 1. Compound 13 (19 mg, 33.6 μmol) was dissolved in 1 mL dry
dimethylformamide and treated with piperidine (6 μL). The solution was stirred for 40 min and the
volatiles were removed in vacuo. The reaction mixture was diluted with 1 mL dry dimethylformamide
and dodecylisocyanate (16 μL, 66.4 μmol) was added. The resulting mixture was stirred at room
temperature for 22 h at room temperature and concentrated in vacuo to obtain a solid that was purified
19

by column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to 1:9) to afford a white
solid (0.0140 g, 75%). Mp: 231-233 °C. []D25 -28.4 (c 0.02, MeOH). IR (neat): 3311, 3270, 3064, 2958,
2923, 2853, 1626, 1544, 1466, 1386, 1283, 1239, 1219, 1168, 1095, 1017, 961, 915, 848, 721, 643 cm-1.
1H NMR (400 MHz, CD3OD): δ 6.74 (dd, J = 15.6, 4.8 Hz, 1H), 6.49 (d, J = 15.6 Hz, 1H), 4.72-4.66 (m,
1H), 4.57 – 4.47 (m, 1H), 4.10 (d, J = 5.3 Hz, 1H), 3.59 – 3.45 (m, 1H), 3.16 – 3.06 (m, 3H), 2.77 – 2.61
(m, 1H), 2.55 – 2.41 (m, 1H), 2.23 – 2.10 (m, 1H), 1.52-1.41 (m, 2H), 1.36 – 1.22 (m, 20H), 1.05 – 0.72
13
(m, 12H). C NMR (126 MHz, CD3OD): δ 145.9, 121.5, 78.4, 71.5, 60.7, 56.4, 53.0, 48.0, 41.8, 41.1,
33.9, 33.2, 32.7, 32.0, 31.4, 30.9, 30.6, 28.1, 23.9, 20.1, 18.8, 18.0, 14.6. HRMS Calcd. for C28H52N5O4S
[M+H]+ 554.3740, Found 554.3746.
(2S)-2-(3-Dodecylureido)-3-methyl-N-((8S,11R,E)-8-methyl-1-oxido-5,10-dioxo-1-thia-4,9-
diazacyclododec-6-en-11-yl)butanamide 2. Compound 1 (12.7 mg, 22.9 μmol) was dissolved in 1.5 mL
of methanol and set to stir at 0 °C. Sodium periodate (5.7 mg, 26.6 μmol) was dissolved in 0.2 mL of
water and added dropwise; the reaction mixture was stirred for 24 h and concentrated in vacuo. The
crude solid was purified by column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to
1:9) to give a white solid (7.8 mg, 60%). Mp: 204-206 °C. []D25 -36.3 (c 0.024, MeOH). IR (neat): 3316,
2956, 2921, 2852, 1708, 1657, 1622, 1553, 1522, 1457, 1377, 1340, 1293, 1238, 1205, 1166, 1091,
1042, 1013, 914, 853, 815, 766, 721, 648, 580, 541, 474, 456 cm-1. 1H NMR (400 MHz, CD3OD): δ 6.87
(dd, J = 15.3, 5.0 Hz, 1H), 6.16 (d, J = 15.5 Hz, 1H), 4.97 (t, J = 4.2 Hz, 1H), 4.57 – 4.33 (m, 1H), 4.08
(d, J = 5.4 Hz, 1H), 3.76 (dd, J = 14.6, 3.2 Hz, 1H), 3.71 – 3.64 (m, 1H), 3.49 – 3.42 (m, 1H), 3.17 (dd, J
= 14.4, 4.4 Hz, 1H), 3.09 (t, J = 6.5 Hz, 1H), 2.97 – 2.88 (m, 1H), 2.19 – 2.12 (m, 1H), 1.47 – 1.42 (m,
13
2H), 1.39 – 1.12 (m, 22H), 1.06 – 0.73 (m, 12H). C NMR (126 MHz, CD3OD): δ 149.0, 120.2, 78.4,
60.6, 54.3, 54.1, 51.5, 47.9, 41.1, 37.2, 33.2, 32.0, 31.4, 30.9, 30.6, 28.1, 23.9, 20.1, 18.7, 18.0, 14.6.
HRMS calcd. for C28H52N5O5S [M+H]+ 570.3689, found 570.3703.
N-((8S,11R,E)-8-Methyl-5,10-dioxo-1-thia-4,9-diazacyclododec-6-en-11-yl)pyrazine-2-carboxamide
3. The macrolactam 11 (60.1 mg, 0.175 mmol) was treated with hydrochloric acid in ethyl acetate (3 N,
20

2.0 mL) and stirred for 30 min at room temperature, then concentrated in vacuo. The resulting white
solid hydrochloride salt was dissolved in 2 mL dimethylformamide and MP-carbonate resin (2.94
mmol/g, 0.1801 g, 0.529 mmol) was added at 0 °C, followed by pyrazine carboxylic acid N-
hydroxysuccinimide ester (14) (49.0 mg, 0.222 mmol). The resulting mixture was stirred at 0 °C for 15
min and 14 h at room temperature and concentrated in vacuo to obtain a solid that was purified by
column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to 1:9) to give a white solid
(0.0293 g, 48%). Mp: 256-258 °C. []D - 13.9 (c 0.13, MeOH). IR (neat): 3450, 3387, 3374, 3265,
25
3044, 2986, 2932, 2918, 1721, 1684, 1666, 1651, 1628, 1582, 1543, 1514, 1464, 1454, 1404, 1364,
1350, 1335, 1308, 1293, 1268, 1241, 1218, 1167, 1156, 1094, 1047, 1021, 964, 947, 907, 893, 875, 852,
1
843, 795, 778, 734, 716, 688, 669 cm-1. H NMR (300 MHz, CD3OD): δ 9.25 (d, J = 1.4 Hz, 1H), 8.82
(d, J = 2.5 Hz, 1H), 8.71 (dd, J = 2.4, 1.5 Hz, 1H), 6.77 (dd, J = 15.6, 4.9 Hz, 1H), 6.49 (d, J = 15.7 Hz,
1H), 4.95 (dd, J = 5.8, 1.9 Hz, 1H), 4.61 – 4.49 (m, 1H), 3.58 (ddd, J = 14.7, 8.8, 4.1 Hz, 1H), 3.35 (d, J
= 2.8 Hz, 1H), 3.28 – 3.17 (m, 2H), 2.77 (ddd, J = 14.4, 9.0, 5.1 Hz, 1H), 2.48 (ddd, J = 14.9, 9.1, 6.0
13
Hz, 1H), 1.37 (d, J = 7.1 Hz, 3H). C NMR (126 MHz, CD3OD): δ 171.7, 170.7, 164.3, 149.2, 146.0,
145.7, 145.1, 144.8, 121.2, 52.5, 48.1, 42.0, 33.7, 32.0, 18.8. HRMS calcd. for C15H20N5O3S [M+H]+
350.1287, found 350.1298.
N-((8S,11R,E)-8-Methyl-1-oxido-5,10-dioxo-1-thia-4,9-diazacyclododec-6-en-11-yl)pyrazine-2-
carboxamide 4. Compound 3 (6.2 mg, 17.7 μmol) was dissolved in 1 mL of methanol and set to stir at 0
°C. Sodium periodate (4.3 mg, 20.1 μmol) was dissolved in 0.1 mL of water and added drop-wise, and
the reaction mixture was stirred for 24 hours and concentrated in vacuo. The white crude solid was
purified by column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to 1:9) to give a
white solid (6.4 mg, 98%). Mp: 250-252 °C. []D -32.0 (c 0.05, MeOH). IR (neat): 3370, 3332, 3267,
25
3067, 2956, 2922, 2852, 1674, 1661, 1646, 1629, 1583, 1529, 1472, 1454, 1407, 1366, 1354, 1282,
1267, 1238, 1213, 1169, 1155, 1091, 1050, 1021, 1006, 971, 945, 908, 883, 852, 775, 734, 717, 677,
631, 582, 553, 463, 432 cm-1. 1H NMR (400 MHz, d6-DMSO): δ 9.21 (s, 1H), 8.91 (d, J = 2.1 Hz, 1H),
21

8.75 (d, J = 7.5 Hz, 1H), 6.83 (dd, J = 15.3, 4.6 Hz, 1H), 6.06 (d, J = 15.0 Hz, 1H), 5.16 (dd, J = 7.5, 3.6
Hz, 1H), 4.51 (dd, J = 12.6, 7.3 Hz, 1H), 4.04 – 3.95 (m, 1H), 3.55 – 3.45 (m, 1H), 3.25 – 3.14 (m, 3H),
13
2.75 – 2.65 (m, 1H), 1.24 (d, J = 6.4 Hz, 3H). C NMR (126 MHz, d6-DMSO): δ 168.0, 165.5, 162.0,
148.2, 147.5, 143.7, 143.5, 143.3, 118.5, 55.3, 52.9, 50.5, 45.8, 36.5, 18.5. HRMS calcd. for
C15H19N5O4SNa [M+Na]+ 388.1055, found 388.1040.
(S)-2-((S)-2-(4-Isobutylphenyl)propanamido)-3-methyl-N-((8S,11R,E)-8-methyl-5,10-dioxo-1-thia-4,9-
diazacyclododec-6-en-11-yl)butanamide 5. Compound 13 (22.2 mg, 39.3 μmol) was dissolved in 1 mL
dry dimethylformamide and treated with piperidine (10 μL) and the solution was stirred for 30 min and
volatiles were removed in vacuo. The residue was dissolved in 1 mL dimethylformamide and (S)-(+)-
ibuprofen N-hydroxysuccinimide ester (15) (27.8 mg, 91.6 μmol) was added. The resulting mixture was
stirred at 0 °C for 30 min and 24 h at room temperature and concentrated in vacuo to obtain a white solid
that was purified by column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to 1:9) to
give a white solid (15.2 mg, 73%). Mp: 189-191 °C. []D25 -37.1 (c 0.07, MeOH). IR (neat): 3274, 3055,
2957, 2925, 2870, 1706, 1630, 1535, 1459, 1382, 1282, 1220, 1166, 1088, 1030, 908, 848, 815, 777,
718, 659, 639, 577, 459 cm-1. 1H NMR (300 MHz, CD3OD): δ 7.25 (d, J = 8.1 Hz, 2H), 7.08 (d, J = 7.9
Hz, 2H), 6.72 (dd, J = 15.7, 4.9 Hz, 1H), 6.46 (d, J = 15.4 Hz, 1H), 4.60 (dd, J = 6.1, 1.9 Hz, 1H), 4.54 –
4.45 (m, 1H), 4.21 (d, J = 7.4 Hz, 1H), 4.19 – 4.11 (m, 1H), 3.75 (dd, J = 14.4, 7.2 Hz, 1H), 3.55 – 3.42
(m, 1H), 3.26 – 3.16 (m, 1H), 3.12 (dd, J = 14.5, 2.2 Hz, 1H), 2.95 (dd, J = 14.5, 6.2 Hz, 1H), 2.44 (d, J
= 7.1 Hz, 2H), 2.41 – 2.33 (m, 1H), 2.12 – 2.03 (m, 1H), 1.88 – 1.78 (m, 1H), 1.46 (d, J = 7.1 Hz, 3H),
13
1.37 – 1.27 (m, 6H), 0.97 – 0.88 (m, 9H). C NMR (126 MHz, CD3OD): δ 177.7, 176.0, 173.1, 171.6,
170.9, 145.9, 141.6, 139.9, 130.5, 128.5, 121.5, 60.4, 53.1, 48.0, 47.0, 46.2, 42.0, 33.7, 32.7, 31.8, 31.6,
30.8, 30.1, 28.6, 26.4, 22.9, 19.9, 19.0, 18.8. HRMS calcd. for C28H42N4O4SNa [M+Na]+ 553.2825, found
553.2802.
N-((S)-3-Methyl-1-(((8S,11R,E)-8-methyl-5,10-dioxo-1-thia-4,9-diazacyclododec-6-en-11-yl) amino)-
1-oxobutan-2-yl)pyrazine-2-carboxamide 6. Compound 13 (59.7 mg, 0.106 mmol) was dissolved in 2
22

mL dry dimethylformamide and treated with piperidine (20 μL). The solution was stirred for 30 min and
the volatiles were removed in vacuo. The residue was dissolved in 2 mL dimethylformamide and
pyrazine carboxylic acid N-hydroxysuccinimide ester (14) (48.0 mg, 0.217 mmol) was added. The
resulting mixture was stirred for 20 h at room temperature and concentrated in vacuo to obtain a solid
that was purified by column chromatography (SiO2, methanol/dichloromethane gradient, 0:10 to 1:9) to
give a white solid (19.8 mg, 42%). Mp: 241-243 °C. []D25 -35.2 (c 0.05, MeOH). IR (neat): 3289, 3273,
3058, 2962, 2929, 2874, 1634, 1513, 1450, 1389, 1336, 1291, 1218, 1164, 1095, 1047, 1019, 961, 889,
1
859, 847, 775, 731, 645, 577, 439 cm-1. H NMR (400 MHz, CD3OD): δ 9.24 (d, J = 1.1 Hz, 1H), 8.81
(d, J = 2.4 Hz, 1H), 8.71 (dd, J = 2.3, 1.5 Hz, 1H), 6.72 (dd, J = 15.6, 4.8 Hz, 2H), 6.49 (d, J = 15.6 Hz,
1H), 4.71 (dd, J = 19.0, 4.1 Hz, 1H), 4.59 – 4.55 (m, 1H), 4.51 (dd, J = 13.9, 7.5 Hz, 1H), 3.55 – 3.46
(m, 1H), 3.27 – 3.16 (m, 2H), 3.05 (dd, J = 14.3, 6.4 Hz, 2H), 2.73 – 2.65 (m, 1H), 2.54 – 2.45 (m, 1H),
13
2.30 – 2.18 (m, 1H), 1.33 (d, J = 7.1 Hz, 3H), 1.05 – 0.99, m, 6H). C NMR (101 MHz, CD3OD): δ
173.7, 172.7, 171.7, 170.9, 165.1, 149.0, 145.9, 145.0, 121.5, 60.5, 59.9, 53.3, 53.1, 48.0, 41.9, 33.9,
32.8, 31.8, 29.7, 20.1, 19.9, 18.8. HRMS calcd. for C20H28N6O4SNa [M+Na]+ 471.1790, found 471.1796.
tert-Butyl
((S)-3-((2-(2-(diethoxyphosphoryl)acetamido)ethyl)thio)-1-(((S)-1-hydroxypropan-2-
yl)amino)-1-oxopropan-2-yl)carbamate 10. L-4-Thialysine hydrochloride salt (7) (0.2965 g, 1.48 mmol)
was dissolved in dioxane/water (1:1, 4 mL) and cooled to 0 °C. Sodium carbonate (0.3149 g, 2.97
mmol) was added to the mixture at 0 °C. After 30 min, phosphonoacetic acid N-hydroxysuccinimide
ester (8) (0.4352 g, 1.48 mmol) was added and the mixture was stirred for 1 h at 0 °C and for 5 days at
room temperature. Boc anhydride (0.6058 g, 2.78 mmol) dissolved in a minimum amount of dioxane
was added drop-wise to the reaction mixture, which was stirred for 20 h at room temperature. The
reaction mixture was washed with ethyl acetate (10 mL) to remove organic impurities. The aqueous
phase was acidified to pH 2 using 1N hydrochloric acid, and extracted with ethyl acetate (4  15 mL).
The combined organic extract was dried over anhydrous magnesium sulfate and concentrated to afford 9
as a colorless oil. This compound was used in the subsequent step without further purification.
23

The crude material 9 (0.5112 g, 1.16 mmol) was dissolved in 11 mL dry dichloromethane and set to
stir at 0 °C. N,N'-Dicyclohexylcarbodiimide (0.2904 g, 1.41 mmol) and N-hydroxysuccinimide (0.1347
g, 1.17 mmol) were added at 0 °C, followed by L-alaninol (0.18 mL, 2.31 mmol). The mixture was
stirred at 0 °C for 30 min, then for 20 h at room temperature and the N,N'-dicyclohexylurea was filtered
off. The filtrate was concentrated to afford a yellowish oil that was purified by column chromatography
(SiO2, acetone/dichloromethane gradient, 1:4 to 4:1) to afford the title compound 10 as a colorless oil
(0.3987 g, 54%). []D25 -14.99 (c 1.0, MeOH). IR (neat): 3289, 3282, 2978, 2931, 2875, 1711, 1651,
1
1531, 1454, 1411, 1392, 1365, 1239, 1164, 1097, 1048, 1020, 968, 865, 839, 781 cm-1. H NMR (300
MHz, CDCl3): δ 7.81 (s, 1H), 7.57 (d, J = 5.7 Hz, 1H), 5.59 (d, J = 8.4 Hz, 1H), 4.37 – 4.24 (m, 1H),
4.20 – 4.07 (m, 4H), 4.05 – 3.96 (m, 1H), 3.66 (dd, J = 11.4, 3.4 Hz, 2H), 3.48 (dd, J = 11.1, 5.8 Hz,
1H), 3.27 (br s, 1H), 2.90 – 2.78 (m, 4H), 2.67 – 2.54 (m, 1H), 1.42 (s, 9H), 1.32 (q, J = 7.0 Hz, 6H),
1.16 (d, J = 6.8 Hz, 3H). 13C NMR (75 MHz, CDCl3): δ 170.7, 164.6, 155.9, 80.1, 66.1, 63.1, 62.9, 54.6,
47.9, 39.7, 36.7, 36.0, 34.3, 33.7, 28.4, 16.8, 16.5. HRMS calcd. for C19H39N3O8PS [M+H]+ 500.2195,
found 500.2208.
tert-Butyl ((8S,11R,E)-8-methyl-5,10-dioxo-1-thia-4,9-diazacyclododec-6-en-11-yl) carbamate 11.
Dess-Martin periodinane (0.4123 g, 0.972 mmol) was added to a stirring solution of phosphono-alcohol
10 (0.4051 g, 0.811 mmol) in 8 mL of dry dichloromethane at room temperature and stirred for 45 min.
The reaction mixture was diluted with 12 mL dichloromethane and 24 mL of a 1:1 mixture of sat
NaHCO3 and 2% sodium thiosulfate was added. The mixture was stirred vigorously until the organic
phase became clear, then the phases were separated. The organic phase was dried over sodium sulfate,
filtered, and concentrated in vacuo without warming to obtain an aldehyde that was used without
purification in the subsequent step.
Tetramethylethylenediamine (0.15 mL, 1.00 mmol) and triethylamine (0.46 mL, 3.30 mmol) were
added to a stirring suspension of zinc triflate (0.6566 g, 1.81 mmol) in 120 mL dry tetrahydrofuran at
room temperature, and stirred for 45 minutes under an inert atmosphere. The crude phosphono-aldehyde
24

was dissolved in 70 mL dry tetrahydrofuran and added drop-wise to the suspension over 2.5 h. The
reaction mixture was stirred at room temperature for 21 h and concentrated in vacuo to ca. 10 mL. The
residue was diluted with 100 mL ethyl acetate and washed with 50 mL each of brine and 1%
hydrochloric acid. The organic phase was dried over magnesium sulfate and concentrated to afford a
yellowish oil that was purified by chromatography (SiO2, acetone/dichloromethane gradient, 1:4 to 4:1)
to afford the title compound as a white solid film (0.1624 g, 58%). Mp: 176-179 °C. []D -55.38 (c 0.30,
25
MeOH). IR (neat): 3308, 2978, 2933, 1706, 1663, 1635, 1494, 1455, 1393, 1367, 1247, 1225, 1159,
1095, 1057, 1030, 974, 913, 885, 848, 759, 737 cm-1. 1H NMR (700 MHz, CD3OD): δ 6.72 (dd, J = 15.6,
4.9 Hz, 1H), 6.45 (d, J = 15.6 Hz, 1H), 4.54 – 4.48 (m, 1H), 4.43 (d, J = 4.4 Hz, 1H), 3.54 – 3.46 (m,
2H), 3.26 – 3.18 (m, 2H), 3.00 (dd, J = 14.4, 6.1 Hz, 1H), 2.70 (ddd, J = 14.0, 8.6, 5.8 Hz, 1H), 2.51 –
13
2.43 (m, 1H), 1.45 (s, 9H), 1.33 (d, J = 7.1 Hz, 3H). C NMR (101 MHz, CD3OD): δ 172.7, 170.8,
157.2, 146.0, 121.3, 81.1, 53.8, 48.0, 42.0, 34.1, 32.4, 28.8, 18.8. HRMS calcd. for C15H26N3O4S [M+H]+
344.1644, found 344.1650.
tert-Butyl
((8S,11R,E)-8-methyl-1-oxido-5,10-dioxo-1-thia-4,9-diazacyclododec-6-en-11-yl)
carbamate 12. The macrolactam 11 (10.5 mg, 30.6 μmol) was dissolved in 1 mL of methanol and set to
stir at 0 °C. Sodium periodate (8.2 mg, 38.3 μmol) was dissolved in 0.1 mL of water and added drop-
wise, and the reaction mixture was stirred for 24 hours and concentrated in vacuo. The white crude solid
was purified by column chromatography (SiO2, 1:9 methanol/dichloromethane) to give a white solid (9.8
mg, 89%). Mp: decomposition at 210 °C. []D -108.76 (c 0.3, MeOH). IR (neat): 3417, 3295, 3218,
25
3094, 2975, 2927, 2467, 2387, 2263, 2204, 1710, 1671, 1632, 1535, 1469, 1366, 1249, 1162, 1018, 986,
852, 789, 668, 637, 569, 514, 432 cm-1. 1H NMR (700 MHz, CD3OD): δ 6.92 (dd, J = 15.4, 4.8 Hz, 1H),
6.15 (d, J = 15.4 Hz, 1H), 4.77 (t, J = 3.5 Hz, 1H), 4.59 – 4.55 (m, 1H), 3.77 (dd, J = 14.4, 3.2 Hz, 1H),
3.73 – 3.67 (m, 1H), 3.48 (ddd, J = 16.1, 6.9, 4.5 Hz, 1H), 3.30 – 3.27 (m, 1H), 3.14 (dd, J = 14.4, 3.8
13
Hz, 1H), 2.96 (ddd, J = 13.9, 8.9, 7.2 Hz, 1H), 1.45 (s, 9H), 1.33 (d, J = 7.1 Hz, 3H). C NMR (126
25

MHz, CD3OD): δ 171.0, 169.7, 157.1, 149.5, 120.0, 81.4, 54.5, 54.4, 52.9, 47.9, 37.4, 28.8, 18.7. HRMS
calcd. for C15H26N3O5S [M+H]+ 360.1588, found 360.1596.
(9H-Fluoren-9-yl)methyl
((S)-3-methyl-1-(((8S,11R,E)-8-methyl-5,10-dioxo-1-thia-4,9-diaza
cyclododec-6-en-11-yl)amino)-1-oxobutan-2-yl)carbamate 13. The macrolactam 11 (0.1231 g, 0.358
mmol) was treated with hydrochloric acid in ethyl acetate (3N, 3.0 mL) and stirred for 30 min at room
temperature, then concentrated in vacuo. The resulting white solid hydrochloride salt was dissolved in 4
mL dry dimethylformamide and set to stir at 0 °C. MP-Carbonate resin (2.94 mmol/g, 0.3663 g, 1.08
mmol) was added, followed by N-Fmoc-L-valine N-hydroxysuccinimide ester (0.1952 g, 0.447 mmol).
The reaction mixture was stirred for 30 min at 0 °C, then 24 h at room temperature. The reaction mixture
was concentrated in vacuo to obtain a white solid that was purified by column chromatography (SiO2,
methanol/dichloromethane gradient, 0:10 to 1:9) to afford a white solid (0.1217 g, 60%). Mp: 207-209
°C. []D -55.43 (c 0.05, MeOH). IR (neat): 3282, 3063, 2959, 2925, 2872, 2854, 2463, 2409, 2375,
25
2353, 2338, 1692, 1635, 1538, 1450, 1390, 1342, 1290, 1249, 1237, 1219, 1163, 1135, 1101, 1084,
1
1033, 992, 960, 852, 795, 758, 739, 730, 668 cm-1. H NMR (400 MHz, CD3OD): δ 7.80 (d, J = 7.5 Hz,
2H), 7.71 – 7.64 (m, 2H), 7.39 (t, J = 7.4 Hz, 2H), 7.31 (t, J = 7.3 Hz, 2H), 6.72 (dd, J = 15.6, 4.9 Hz,
1H), 6.47 (d, J = 15.8 Hz, 1H), 4.71 (d, J = 5.1 Hz, 1H), 4.54 – 4.47 (m, 1H), 4.37 (t, J = 7.1 Hz, 1H),
4.25 (t, J = 6.8 Hz, 1H), 3.99 (d, J = 6.7 Hz, 1H), 3.55 – 3.46 (m, 1H), 3.20 (d, J = 14.9 Hz, 1H), 3.07
(dd, J = 14.3, 6.1 Hz, 1H), 2.73 – 2.62 (m, 1H), 2.52 – 2.42 (m, 1H), 2.19 – 2.06 (m, 1H), 1.32 (d, J =
13
7.0 Hz, 3H), 0.97 – 0.93 (m, 6H). C NMR (126 MHz, CD3OD): δ 146.0, 128.9, 128.4, 126.4, 121.1,
13
68.2, 62.0, 53.0, 48.0, 41.8, 33.8, 32.7, 31.8, 30.9, 19.9, 18.8. (Partial C NMR data due to lack of
compound solubility.) HRMS calcd. for C30H37N4O5S [M+H]+ 565.2485, found 565.2497.
General procedure for synthesizing NHS esters. The N-hydroxysuccinimide esters of
pyrazinecarboxylic acid and (S)-(+)-ibuprofen were prepared by dissolving the starting acid (0.200 g, 1.0
equiv) in dry tetrahydrofuran and setting to stir at 0 °C. N-Hydroxysuccinimide (1.0 equiv) was added
followed by N,N'-dicyclohexylcarbodiimide (1.0 equiv). The mixture was warmed to room temperature
26

and stirred for 24 h. The N,N'-dicyclohexylurea was removed by filtration and the filtrate was
concentrated to afford a solid that was purified by column chromatography.
2,5-Dioxopyrrolidin-1-yl pyrazine-2-carboxylate 14. A crude white solid was purified by column
chromatography (SiO2, acetone/dichloromethane, 1:4) to give a white solid (0.2757 g, 77%). Mp: 161 –
164 °C. 1H NMR (300 MHz, CDCl3): δ 9.40 (d, J = 1.4 Hz, 1H), 8.90 (d, J = 2.3 Hz, 1H), 8.83 (dd, J =
13
2.4, 1.5 Hz, 1H), 2.95 (s, 4H). C NMR (101 MHz, CDCl3): δ 168.9, 159.6, 149.3, 147.1, 145.2, 140.2,
25.8. IR (neat): 3501, 3324, 3276, 3078, 2929, 2851, 1875, 1806, 1785, 1703, 1653, 1626, 1571, 1533,
1468, 1449, 1420, 1399, 1360, 1306, 1260, 1243, 1203, 1183, 1153, 1075, 1060, 1048, 1025, 1011, 995,
951, 892, 876, 853, 811, 782, 764, 713 cm-1. HRMS calcd. for C9H8N3O4 [M+H]+ 222.0509, found
222.0502.
2,5-Dioxopyrrolidin-1-yl (S)-2-(4-isobutylphenyl)propanoate 15. A crude white solid was purified by
column chromatography (SiO2, ethyl acetate/hexane gradient, 1:4 to 2:3) to give a white solid (0.2742 g,
93%). Mp: 74-76 °C. []D25 60.2 (c 1.0, MeOH). 1H NMR (500 MHz, CDCl3): δ 7.27 (d, J = 8.0 Hz, 2H),
7.15 (d, J = 7.9 Hz, 2H), 4.04 (q, J = 7.2 Hz, 1H), 2.79 (s, 4H), 2.47 (d, J = 7.2 Hz, 2H), 1.87 (dp, J =
13
13.4, 6.7 Hz, 1H), 1.64 (d, J = 7.2 Hz, 3H), 0.91 (d, J = 6.6 Hz, 6H). C NMR (126 MHz, CDCl3) δ
170.2, 169.2, 141.4, 135.6, 129.9, 129.7, 127.5, 127.3, 45.2, 42.8, 42.7, 30.3, 25.8, 22.6, 19.2, 19.1. IR
(neat): 3054, 2959, 2904, 2851, 1862, 1839, 1800, 1765, 1741, 1631, 1513, 1446, 1432, 1380, 1363,
1342, 1313, 1261, 1238, 1207, 1182, 1144, 1088, 1047, 991, 962, 924, 871, 850, 814, 769, 728, 688,
669, 641, 560, 507, 464 cm-1. HRMS calcd. for C17H21NO4Na [M+Na]+ 326.1368, found 326.1368.
Biological methods
In Vitro Proteasome Activity Assay. To determine the anti-proteasome activity of the thiasyrbactins
in the in vitro environment, we measured the three catalytic activities (β1, β2, β5) of the proteasome as
previously described.13 Bortezomib (BTZ) (LC Laboratories, Woburn, MA) and the immunoproteasome
inhibitor ONX-0914 (UBP Bio, Aurora, CO) were used as controls. Purified 20S constitutive
proteasome from human erythrocytes or immunoproteasome from human peripheral blood mononuclear
27

cells (PBMCs) (Boston Biochem, Cambridge, MA) and luminogenic substrates Z-LRR-Glo, Z-nLPnLD-
Glo, and Suc-LLVY-Glo (Promega, Fitchburg, WI) specific for the β1, β2, and β5 (also referred to as C-
L, T-L, and CT-L) catalytic subunit activities, respectively, were used. Briefly, 2.2 ng/µl of constitutive
and immunoproteosome were incubated with increasing concentration of proteasome inhibitor (0 to 10
µM) in 10 mM HEPES, pH 7.3, for two hours at 37 ºC. Equal volume of the appropriate luminogenic
substrate was added to each reaction and luminescence was measured 10 minutes later using a Multi-
Mode Synergy (Biotek, Inc., Winooski, VT) plate reader. Assays were done in triplicate (n=3). The Ki-
50 values for each catalytic activity were determined for each of the inhibitors using Graphpad Prism 5
software (La Jolla, CA) and averaged using Microsoft Excel (Redmond, WA).
Cell Viability Assay. To determine the effect of the thiasyrbactins proteasome inhibitors on the
viability of cancer cells, three different human neuroblastoma (NB) cell lines were used. BTZ was
included as a control. SK-N-Be(2)c, SK-N-SH cells (ATCC, Manassas, VA) and MYCN2 cells
(provided by Dr. Jason Shohet, Texas Children’s Hospital)37 were plated overnight in RPMI media
containing 10% heat-inactivated fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA, USA),
supplemented with penicillin (100 U/ml) and streptomycin (100 µg/ml). Cells were cultured at 37 C in
a humidified atmosphere containing 5% CO2 and treated with 0.01, 0.1 and 1 µM proteasome inhibitor
for 24 hours. Cell viability was determined using the Cell Titer 96 AQueous One Cell Proliferation
Assay (Promega) reagent MTS by measuring absorbance at 490 nm using a Multi-Mode Synergy plate
reader as previously described.13 Assays were done in duplicate with triplicate wells of treatment per
experiment (n=6). Data were expressed as the average cell viability relative to untreated control cells
using Microsoft Excel.
Author Contributions
M.C.P., C.-E. C., and A.S.B. designed the study. N.A.B. performed the chemical synthesis of NAM
compounds. C.R.S and L.P.Y. performed the biological studies including the in vitro proteasome
28

activity assay and cell viability assay. C. C. R. performed the computational analysis. The manuscript
was written through contributions of all authors. All authors have given approval to the final version of
the manuscript.
Declaration of interests
M.C.P., N.A.B., and A.S.B. are named inventors on a US Patent application concerning the
thiasyrbactins.
Acknowledgments
Nicole A. (Mikulski) Bakas was supported by a US Department of Education Graduate Assistance in
Areas of National Need fellowship (award #P200A120170). Lisette P. Yco and Chad R. Schultz were
supported by MSU internal funds to A.S.B. We thank Dr. Jason Shohet (Texas Children’s Hospital,
Houston, TX, USA) for providing cell line MYCN2.
References and notes
1.
cepafungins (syrbactins). Nat. Prod. Rep. 2011, 28, 1854-67.
2. Barr, P.; Fisher, R.; Friedberg, J. The role of bortezomib in the treatment of lymphoma. Cancer
Invest. 2007, 25, 766-75.
Krahn, D.; Ottmann, C.; Kaiser, M. The chemistry and biology of syringolins, glidobactins and
3.
Sugumar, D.; Keller, J.; Vij, R. Targeted treatments for multiple myeloma: specific role of
carfilzomib. Pharmgenomics Pers Med. 2015, 8, 23-33.
4.
myeloma. Int. J. Clin. Oncol. 2015, 20, 423-30.
5. Groll, M.; Schellenberg, B.; Bachmann, A.S.; Archer, C.R.; Huber, R.; Powell, T.K.; Lindow, S.;
Torimoto, Y.; Shindo, M.; Ikuta, K.; Kohgo, Y. Current therapeutic strategies for multiple
Kaiser, M.; Dudler, R. A plant pathogen virulence factor inhibits the eukaryotic proteasome by a novel
mechanism. Nature 2008, 452, 755-8.
29

6.
Coleman, C. S.; Rocetes, J. P.; Park, D. J.; Wallick, C. J.; Warn-Cramer, B. J.; Michel, K.;
Dudler, R.; Bachmann, A. S. Syringolin A, a new plant elicitor from the phytopathogenic bacterium
Pseudomonas syringae pv. syringae, inhibits the proliferation of neuroblastoma and ovarian cancer cells
and induces apoptosis. Cell Prolif. 2006, 39, 599-609.
7.
in Sieber, S. A., Ed., Top. Curr. Chem. 2012, 324, Springer, Berlin pp 43-84.
8. Carmony, K. C.; Kim, K. B. Activity-based imaging probes of the proteasome. Cell Biochem.
Biophys. 2013, 67, 91-101.
Krysiak, J.; Breinbauer, R. Activity-based protein profiling for natural product target discovery
9.
Chiba, T.; Hosono, H.; Nakagawa, K.; Asaka, M.; Takeda, H.; Matsuda, A.; Ichikawa, S. Total
synthesis of syringolin A and improvement of its biological activity. Angew. Chem. Int. Ed. 2014, 53,
4836–4839.
10.
Clerc, J.; Groll, M.; Illich, D. J.; Bachmann, A. S.; Huber, R.; Schellenberg, B.; Dudler, R.;
Kaiser, M. Synthetic and structural studies on syringolin A and B reveal critical determinants of
selectivity and potency of proteasome inhibition. Proc. Natl. Acad. Sci. USA 2009, 106, 6507-12.
11.
Clerc, J.; Li, N.; Krahn, D.; Groll, M.; Bachmann, A. S.; Florea, B. I.; Overkleeft, H. S.; Kaiser,
M. The natural product hybrid of syringolin A and glidobactin A synergizes proteasome inhibition
potency with subsite selectivity. Chem. Commun. 2011, 47, 385-7.
12.
Clerc, J.; Schellenberg, B.; Groll, M.; Bachmann, A. S.; Huber, R.; Dudler, R.; Kaiser, M.
Convergent synthesis and biological evaluation of syringolin A and derivatives as eukaryotic 20S
proteasome inhibitors. Eur. J. Org. Chem. 2010, 3991-4003.
13.
and cytotoxicity of syringolin B-based proteasome inhibitors. Tetrahedron 2011, 67, 9950-6.
14. Bachmann, A. S.; Opoku-Ansah, J.; Ibarra-Rivera, T. R.; Yco, L. P.; Ambadi, S.; Roberts, C. C.;
Ibarra-Rivera, T. R.; Opoku-Ansah, J.; Ambadi, S.; Bachmann, A. S.; Pirrung, M. C. Synthesis
Chang, C. A.; Pirrung, M. C. The syrbactin structural analog TIR-199 blocks proteasome activity and
induces tumor cell death. J. Biol. Chem. 2016, 291, 8350-62.
30