Gram-scale preparation of the antibiotic lead compound salicyl-AMS, a potent inhibitor of bacterial salicylate adenylation enzymes

Article abstract of DOI:10.1016/bs.mie.2020.04.051

Salicyl-AMS (1) is a potent inhibitor of salicylate adenylation enzymes used in bacterial siderophore biosynthesis and a promising lead compound for the treatment of tuberculosis. An optimized, multigram synthesis is presented, which provides salicyl-AMS as its sodium salt (1·Na) in three synthetic steps followed by a two-step salt formation process. The synthesis proceeds in 11.6% overall yield from commercially available adenosine 2′,3′-acetonide and provides highly purified material.

Full text of DOI:10.1016/bs.mie.2020.04.051

CHAPTER FOUR
Gram-scale preparation
of the antibiotic lead compound
salicyl-AMS, a potent inhibitor
of bacterial salicylate adenylation
enzymes
Nihar Kinarivalaa, Lisa C. Standkeb, Tezcan Guneya, Cheng Jia,
Naoyoshi Noguchic,†, Yasutomi Asanoc,‡, Derek S. Tana,b,d,
*
aChemical Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York,
NY, United States
bPharmacology Program, Weill Cornell Graduate School of Medical Sciences, Memorial Sloan Kettering
Cancer Center, New York, NY, United States
cTri-Institutional Therapeutics Discovery Institute, New York, NY, United States
dTri-Institutional Research Program, Memorial Sloan Kettering Cancer Center, New York, NY, United States
*Corresponding author: e-mail address: tand@mskcc.org
Contents
1. Introduction
2. Synthesis of salicyl-AMS, sodium salt
2.1 Equipment
2.2 Chemicals
3. Protocol
70
72
74
75
76
76
77
77
78
79
80
81
3.1 Preparation of sulfamoyl chloride (10)
3.2 Preparation of salicylic acid, N-hydroxysuccinimide ester (12)
3.3 Synthesis of 50-O-sulfamoyladenosine 20,30-acetonide (14)
3.4 Synthesis of 50-O-(N-salicylsulfamoyl)adenosine 20,30-acetonide (15)
3.5 Synthesis of salicyl-AMS (free acid) (1)
3.6 Conversion to triethylammonium salt (1ꢀEt3N)
3.7 Conversion to sodium salt (1ꢀNa)
3.8 LC-MS/MS analysis of salicyl-AMS, sodium salt (1ꢀNa) to quantify AMS
impurity
82
83
84
4. Concluding remarks
Acknowledgments
†
Present address: Research Division, Scohia Pharma, Inc., Shonan Health Innovation Park, 26-1,
Muraoka-Higashi 2-chome, Fujisawa, Kanagawa 251-8555, Japan.
‡
Present address: Discovery Science, Axcelead Drug Discovery Partners, Inc., Shonan Health
Innovation Park, 26-1, Muraoka-Higashi 2-chome, Fujisawa, Kanagawa 251-0012, Japan.
#
Methods in Enzymology, Volume 638
ISSN 0076-6879
https://doi.org/10.1016/bs.mie.2020.04.051
2020 Elsevier Inc.
All rights reserved.
69

70
Nihar Kinarivala et al.
Conflicts of interests
References
84
84
Abstract
Salicyl-AMS (1) is a potent inhibitor of salicylate adenylation enzymes used in bacterial
siderophore biosynthesis and a promising lead compound for the treatment of tuber-
culosis. An optimized, multigram synthesis is presented, which provides salicyl-AMS as
its sodium salt (1ꢀNa) in three synthetic steps followed by a two-step salt formation pro-
cess. The synthesis proceeds in 11.6% overall yield from commercially available aden-
osine 20,30-acetonide and provides highly purified material.
1. Introduction
Tuberculosis remains a global public health crisis, with an estimated
10 million new cases and 1.5 million deaths in 2018 (Furin, Cox, & Pai,
2019; World Health Organization, 2019). Current treatments are effective
against drug-sensitive infections, but multi-drug resistant and extensively-
drug resistant tuberculosis present major threats, and an estimated 484,000
new infections resistant to the first-line drug rifampicin were reported in
2018. Thus, there is a great need for new antibiotics to treat drug-resistant
tuberculosis infections. Mycobacterium tuberculosis, the bacteria that causes
tuberculosis, acquires the essential nutrient Fe3+ using iron-chelating mo-
lecules called siderophores (De Voss, Rutter, Schroeder, & Barry, 1999;
De Voss, Rutter, Schroeder, Su, et al., 2000; Quadri & Ratledge, 2005;
Ratledge, 2004). Siderophores bind iron with very high affinity and are used
to transport iron across the cell envelope. Accordingly, inhibition of side-
rophore biosynthesis to limit bacterial iron uptake has been advanced as
promising new antibacterial strategy (Lamb, 2015).
In collaboration with the laboratory of Luis Quadri, we have previously
reported the development of the first designed siderophore biosynthesis
inhibitor, salicyl-AMS (50-O-[N-salicylsulfamoyl]adenosine) (1) (Fig. 1A)
(Bythrow, Mohandas, Guney, Standke, et al., 2019; Ferreras, Ryu, Di
Lello, Tan, et al., 2005). Salicyl-AMS is a potent inhibitor of MbtA, a
salicylate adenylation enzyme that catalyzes the first step in the synthesis
of the M. tuberculosis mycobactin siderophores 2 (McMahon, Rush, &
Thomas, 2012; Quadri, Sello, Keating, Weinreb, et al., 1998). Salicyl-
AMS mimics a tightly-bound reaction intermediate, salicyl-AMP (4), and
potently inhibits MbtA biochemical activity as well as M. tuberculosis side-
rophore production and bacterial growth in cell culture. In collaboration
with the laboratory of William Bishai, we demonstrated promising

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
71
A
B
C
Fig. 1 (A) Salicyl-AMS (1) is a potent inhibitor of the salicylate adenylation enzyme
MbtA, and mimics a tightly-bound salicyl-AMP intermediate (4), blocking production
of the mycobactin siderophores (2). R represents various fatty acyl chains (mycobactin
variants) or chains terminating in a carboxylate or methyl ester (carboxymycobactin
variants). (B) Structure of highly toxic fragment 50-O-sulfamoyladenosine (AMS, 6).
(C) Potential decomposition of salicyl-AMS (1) leading to N3–50-cycloadenosine (7)
and N-salicylsulfamate (8).
in vivo efficacy of salicyl-AMS in a mouse model of tuberculosis (Lun, Guo,
Adamson, Cisar, et al., 2013). Aldrich and coworkers have also reported
extensive structure–activity relationship studies of salicyl-AMS analogs
(Dawadi, Boshoff, Park, Schnappinger, et al., 2018; Dawadi, Kawamura,
Rubenstein, Remmel, et al., 2016; Dawadi, Viswanathan, Boshoff, Barry
3rd, et al., 2015; Duckworth, Nelson, & Aldrich, 2012; Engelhart &
Aldrich, 2013; Gupte, Boshoff, Wilson, Neres, et al., 2008; Krajczyk,
Zeidler, Januszczyk, Dawadi, et al., 2016; Nelson, Viswanathan, Dawadi,
Duckworth, et al., 2015; Neres, Labello, Somu, Boshoff, et al., 2008;
Qiao, Gupte, Boshoff, Wilson, et al., 2007; Qiao, Wilson, Bennett, &
Aldrich, 2007; Somu, Boshoff, Qiao, Bennett, et al., 2006; Somu, Wilson,
Bennett, Boshoff, et al., 2006; Vannada, Bennett, Wilson, Boshoff,
et al., 2006).

72
Nihar Kinarivala et al.
To enable further preclinical evaluation of salicyl-AMS, we report
herein an optimized, scalable protocol for synthesis, purification, and analysis
of salicyl-AMS. Importantly, this protocol provides salicyl-AMS that is free
of the des-salicyl fragment 50-O-sulfamoyladenosine (AMS, 6) (Fig. 1B)
(<0.0006mol%, below the lower limit of detection), which is known to
be highly cytotoxic (NCI-60 mean GI50 ¼1.8nM; NSC133114)
(Bloch & Coutsogeorgopoulos, 1971; Jaffe, McCormack, & Meymerian,
1970; Rengaraju, Narayanan, Ganju, Amin, et al., 1986), as well as an
N3–50-cycloadenosine byproduct (7), which forms under basic conditions
via a precedented pathway (Fig. 1C) (Duckworth, Geders, Tiwari,
Boshoff, et al., 2011; Shi, Tiwari, Wilson, Seiler, et al., 2013; Shuman,
Robins, & Robins, 1969). Moreover, salicyl-AMS is highly acidic at the
acylsulfamate moiety (pKa ꢁ1.5) (Cisar & Tan,unpublished results;
Somu, Boshoff, et al., 2006), and Aldrich has reported previously that con-
version to the triethylammonium salt (1ꢀEt3N) provides increased storage
stability (Somu, Boshoff, et al., 2006). Our protocol affords salicyl-AMS
as the corresponding sodium salt (1ꢀNa), which is more commonly used
in pharmaceutical formulations (Paulekuhn, Dressman, & Saal, 2007), to
enable its use in preclinical and potentially clinical studies.
2. Synthesis of salicyl-AMS, sodium salt
Salicyl-AMS (1) was synthesized in three steps from commercially
available adenosine 20,30-acetonide (13), in up to 10-g scale and an overall
yield of 14.4% (Fig. 2). The route incorporates numerous improvements to
our original route (Ferreras et al., 2005), based on subsequent reports from
our group and Aldrich and coworkers. Notably, only a single protect-
ing group manipulation is required. Sulfamoyl chloride (10) was prepared
by treatment of commercially available chlorosulfonylisocyanate (9) with
formic acid (Fig. 2A) (Appel & Berger, 1958; Brodsky & Du Bois,
2005; Heacock, Forsyth, Shiba, & Musier-Forsyth, 1996). Salicylic acid,
N-hydroxysuccinimide ester (12) was prepared by activation of commercially
available salicylic acid (11) with dicyclohexylcarbodiimide and coupling with
N-hydroxysuccinimide (Van Brussel & Van Sumere, 1978) (Fig. 2B).
Adenosine acetonide 13 reacted with sulfamoyl chloride (10) selecti-
vely at the 50-hydroxyl in N,N-dimethylacetamide (Cisar, Ferreras, Soni,
Quadri, et al., 2007; Ferreras, Stirrett, Lu, Ryu, et al., 2008; Okada,
Iwashita, & Koizumi, 2000) to form 50-O-sulfamoyladenosine 20,30-
acetonide (14) in 62.5% yield (Fig. 2C). Acylation with salicylic acid,
N-hydroxysuccinimide ester (12) using cesium carbonate base in

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
73
B
A
C
Fig. 2 (A) Preparation of sulfamoyl chloride (10). (B) Preparation of salicylic acid,
N-hydroxysuccinimide ester (12). (C) Synthesis of salicyl-AMS (1) and conversion
to the corresponding sodium salt (1ꢀNa). CSA, camphorsulfonic acid; DCC,
N,N0-dicyclohexylcarbodiimide;
DMA,
N,N-dimethylacetamide;
DMF,
N,N-
dimethylformamide; NHS, N-hydroxysuccinimde.
N,N-dimethylformamide (Somu, Boshoff, et al., 2006) provided salicyl-
AMS 20,30-acetonide (15) in 84.5% yield. Finally, deprotection was achieved
most efficiently using camphorsulfonic acid in methanol to afford salicyl-
AMS (1) in 27.3% yield (14.4% overall yield over three steps). This
deprotection method provided higher yields compared to a variety of other
conditions evaluated (5:2 trifluoroacetic acid [TFA]/H2O; 3 equiv. TFA
in 1:1 MeOH/H2O; 3 equiv. toluenesulfonic acid in MeOH; 1 equiv.
BF3ꢀEt2O in CH2Cl2; 2 equiv. FeCl3 in MeOH).
Salicyl-AMS was then converted to the corresponding sodium salt via a
two-step procedure. Aldrich and coworkers have reported formation of
the triethylammonium salt of salicyl-AMS directly from silica flash chroma-
tography in the presence of 1% triethylamine (3:1 ethyl acetate/methanol)
(Somu, Boshoff, et al., 2006). However, in our hands this led to incomplete
conversion to the triethylammonium salt. Thus, we first attempted direct
conversion of the free acid of salicyl-AMS (1) to the sodium salt 1ꢀNa
using 1 equivalent of aqueous sodium hydroxide or 1 equivalent of

74
Nihar Kinarivala et al.
aqueous sodium bicarbonate at various temperatures (25°C, 0°C, ꢂ20°C).
However, all of these reactions resulted in formation of varying amounts of
an N3–50-cycloadenosine byproduct (7), which has been reported previ-
ously for related sulfamoyladenosines (Fig. 1C) (Duckworth et al., 2011;
Shi et al., 2013; Shuman et al., 1969).
Accordingly, we turned to a two-step salt formation process involving
initial conversion of salicyl-AMS to the stable, isolable triethylammonium
salt, followed by salt exchange to the desired sodium salt. Thus, the free acid
of salicyl-AMS (1) was treated with excess, freshly distilled triethylamine
at ꢂ20°C to form the triethylammonium salt 1ꢀEt3N. Purification by silica
flash chromatography (4:1 ethyl acetate/methanol with 1% Et3N) removed
the cycloadenosine byproduct 7 and afforded the triethylammonium salt of
salicyl-AMS in 85.8% yield. Notably, increasing the reaction tempera-
ture from ꢂ20°C to 0°C or 25°C resulted in increased formation of the
cycloadenosine byproduct 7.
Next, conversion of the triethylammonium salt 1ꢀEt3N to the sodium
salt 1ꢀNa was achieved using cation exchange on a Dowex 50WX8 column,
which was washed rigorously with water, methanol, and acetonitrile, then
pretreated with aqueous sodium hydroxide to generate the sodium form.
A solution of salicyl-AMS, triethylammonium salt (1ꢀEt3N, 0.1M)
was passed through the column and the collected solution was lyophilized to
provide the salicyl-AMS, sodium salt (1ꢀNa) in 94.4% yield (81.0% overall
yield from free acid, 11.6% overall yield from adenosine acetonide 13). This
salt formation protocol has been applied successfully to both small-scale
(100mg) and large-scale syntheses (2.5g). Elemental analysis indicated a
sulfur/sodium ratio of 1.43:1, comparable to the theoretical value of 1.39:1.
LC-MS/MS analysis of the salicyl-AMS free acid (1) indicated 0.31mol
% AMS as an impurity. In contrast, analysis of the final salicyl-AMS, sodium
salt (1ꢀNa) indicated <0.0006mol% AMS, below the lower limit of
detection. With appropriate minor modifications, this protocol has been
applied to sodium salt formation with other salicyl-AMS analogs
(Bythrow et al., 2019), and may also be useful for the preparation of acyl-
AMS inhibitors of other targets in the adenylate-forming enzyme superfam-
ily (Lux, Standke, & Tan, 2019).
2.1 Equipment
•
•
•
Round bottom flask
Stir bar
Magnetic hot plate

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
75
•
•
•
•
•
•
•
•
•
•
•
•
Septum
Vacuum pump
Vacuum/argon manifold
Thermometer
Sonicator (water bath)
Chromatography columns (glass, various sizes)
Distillation apparatus
Rotary evaporator (Buchi)
Lyophilizer
Automated chromatography instrument (CombiFlash; Teledyn Isco)
NMR spectrometer (400 or 500MHz; Bruker)
High-performance liquid chromatograph–mass spectrometer (LC-MS;
Waters)
•
High-performance liquid chromatograph–triple quadrupole mass spec-
trometer (LC-MS/MS; Agilent)
2.2 Chemicals
Reagents were obtained from Aldrich Chemical (www.sigma-aldrich.com),
Acros Organics (www.fishersci.com), Alfa Aesar (www.alfa.com), or TCI
America (www.tcichemicals.com) and used without further purification
unless otherwise indicated. Optima or HPLC grade solvents were obtained
from Fisher Scientific (www.fishersci.com), degassed with Ar, and purified
on a solvent drying system unless otherwise indicated.
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Chlorosulfonylisocyanate
Formic acid
Petroleum ether
Salicylic acid
N-Hydroxysuccinimide
1,4-Dioxane
N,N0-Dicyclohexylcarbodiimide
N,N-Dimethylacetamide
Sodium bicarbonate
Sodium sulfate
N,N-Dimethylformamide
Cesium carbonate
(1S)-(+)-10-Camphorsulfonic acid
Triethylamine
Dowex 50WX8 (hydrogen form)

76
Nihar Kinarivala et al.
•
•
•
•
•
•
•
1N Sodium hydroxide solution
Guanosine
Methanol
Ethyl acetate
Water
Acetonitrile
Silica gel 60 thin-layer chromatography (TLC) plates with fluorescent
indicator (F254)
•
•
•
•
Silica gel 60 (230–400 mesh)
Celite
Phosphate buffered saline
Dimethylsulfoxide
3. Protocol
All reactions were performed in flame-dried glassware under positive
Ar pressure with magnetic stirring unless otherwise noted. Liquid reagents
and solutions were transferred thru rubber septa via syringes flushed with Ar
prior to use. Cold baths were generated as follows: 0°C, wet ice/water;
ꢂ25°C, dry ice/acetone.
TLC was performed on 0.25mm E. Merck silica gel 60 F254 plates and
visualized under UV light (254nm) or by staining with potassium perman-
ganate (KMnO4), or cerium ammonium molybdenate (CAM). Silica flash
chromatography was performed manually on E. Merck 230–400 mesh silica
gel 60 or on an ISCO CombiFlash Rf+instrument with RediSep silica gel
normal phase columns or RediSep Gold silica gel normal phase columns
with UV detection at 254nm.
NMR spectra were recorded on a Bruker UltraShield Plus 400MHz
Avance III NMR or UltraShield Plus 600MHz Avance III NMR with
DCH CryoProbe at 24°C in methanol-d4 or DMSO-d6.
3.1 Preparation of sulfamoyl chloride (10)
1. In a 100-mL round bottom flask, chlorosulfonylisocyanate (9) (20g,
141.31mmol, 12.27mL, 1.0 equiv.) was cooled to 0°C.
2. Formic acid (6.5g, 141.31mmol, 5.33mL, 1.0 equiv.) was added
dropwise to the reaction mixture at 0°C.
3. The reaction mixture was allowed to warm to 25°C and stirred for 1h.

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
77
4. The reaction mixture was diluted with petroleum ether (50mL),
then filtered.
5. The filter cake was collected and dried under vacuum to give sulfamoyl
chloride (10) (15g) as a white solid, which was used without further
purification.
6. Note: Chlorosulfonylisocyanate and sulfamoyl chloride are highly sensi-
tive to moisture and react violently with water.
3.2 Preparation of salicylic acid, N-hydroxysuccinimide
ester (12)
1. In a 2-L round bottom flask, salicylic acid (11) (42g, 304.08mmol, 1.0
equiv.) and N-hydroxysuccinimide (35g, 304.08mmol, 1.0 equiv.) were
dissolved in 1,4-dioxane (700mL) and cooled to 0°C under nitrogen.
2. N,N0-Dicyclohexylcarbodiimide (62.74g, 304.08mmol, 61.51mL,
1.0 equiv.) was dissolved in 1,4-dioxane (300mL) and the solution
was added to the reaction mixture.
3. The reaction mixture was allowed to warm to 25°C and stirred for 12h.
4. The reaction mixture was filtered to provide salicylic acid,
N-hydroxysuccinimide ester (12) (76g, crude) as a white solid, which
was used without further purification.
5. 1H NMR (DMSO-d6, 400MHz): δ 10.63 (br.s, 1H), 7.85 (dd, J1 ¼
1.6Hz, J2 ¼6.4Hz, 1H), 7.56 (td, J1 ¼1.6Hz, J2 ¼5.6Hz, 1H), 7.06
(d, J¼8.4Hz, 1H), 6.98–6.96 (m, 1H), 2.87 (s, 4H).
3.3 Synthesis of 50-O-sulfamoyladenosine 20,30-acetonide (14)

78
Nihar Kinarivala et al.
1. In a 1-L round bottom flask, adenosine 20,30-acetonide (13) (46g,
149.69mmol, 1.0 equiv.) was dissolved in N,N-dimethylacetamide
(500mL) and cooled to 0°C under nitrogen.
2. Sulfamoyl chloride (10) (34.59g, 299.38mmol, 2.0 equiv., from
Section 3.1) was added to the reaction mixture at 0°C.
3. The reaction mixture was allowed to warm to 25°C and stirred for 12h
with monitoring by LC-MS analysis of aliquots.
4. Upon completion, the reaction was quenched by addition of saturated
aqueous sodium bicarbonate until the pH reached 7–8.
5. The mixture was extracted with ethyl acetate (5ꢃ500mL), then the
organic layers were combined, dried over sodium sulfate, filtered, and
concentrated by rotary evaporation to provide a yellow residue.
6. The residue was purified by silica flash chromatography (ethyl acetate !
20:1 ethyl acetate/methanol) to afford 50-O-sulfamoyladenosine 20,30-
acetonide (14) (43g, 84% purity: 93.48mmol, 62.5% yield [adjusted
for purity]) as a yellow oil.
7. LC-MS: tR ¼1.092min, purity: 84%, m/z 387.0 [M+H]+.
8. 1H NMR (CD3OD, 400MHz): δ 8.27 (s, 1H), 8.22 (s, 1H), 6.25
(d, J¼2.4Hz, 1H), 5.43 (dd, J1 ¼2.0Hz, J2 ¼6.0Hz, 1H), 5.14 (dd, J1 ¼
2.8Hz, J2 ¼6.0Hz, 1H), 4.52–4.50 (m, 1H), 4.32–4.27 (m, 2H), 1.61
(s, 3H), 1.39 (s, 3H).
3.4 Synthesis of 50-O-(N-salicylsulfamoyl)adenosine
20,30-acetonide (15)
1. In a 1-L round bottom flask, 50-O-sulfamoyladenosine 20,30-acetonide
(14) (43g, 84% purity: 93.48mmol, 1.0 equiv. [adjusted for purity], from
Section 3.3) was dissolved in N,N-dimethylformamide (450mL) and
cooled to 0°C under nitrogen.

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
79
2. Salicylic acid, N-hydroxysuccinimide ester (12) (52.35g, 222.58mmol,
2.38 equiv., from Section 3.2) and cesium carbonate (108.78g,
333.87mmol, 3.57 equiv.) were added to the reaction mixture at 0°C.
3. The reaction mixture was allowed to warm to 25°C and stirred for 12h
with monitoring by LC-MS analysis of aliquots.
4. Upon completion, the reaction mixture was filtered and concentrated by
rotary evaporation to provide a white residue.
5. The residue was purified by silica flash chromatography (1:1 petroleum
ether/ethyl acetate!20:1 ethyl acetate/methanol) to afford 50-O-
(N-salicylsulfamoyl)adenosine 20,30-acetonide (15) (40.2g, 99.5% purity:
78.97mmol, 84.5% yield [adjusted for purity]) as an off-white solid.
6. LC-MS: tR ¼1.173min, purity: 99.5%, m/z 507.1 [M+H]+.
7. 1H NMR (CD3OD, 400MHz): δ 8.46 (s, 1H), 8.15 (s, 1H), 7.91 (dd,
J1 ¼1.6Hz, J2 ¼7.6Hz, 1H), 7.33–7.29 (m, 1H), 6.82–6.79 (m, 2H),
6.25 (d, J¼2.8Hz, 1H), 5.39 (dd, J1 ¼2.8Hz, J2 ¼6.0Hz, 1H), 5.16
(dd, J1 ¼2.0Hz, J2 ¼6.0Hz, 1H), 4.59–4.58 (m, 1H), 4.34 (dd,
J1 ¼2.8Hz, J2 ¼4.0Hz, 2H), 1.60 (s, 3H), 1.27 (s, 3H).
3.5 Synthesis of salicyl-AMS (free acid) (1)
1. In a 1-L round bottom flask, 50-O-(N-salicylsulfamoyl)adenosine 20,30-
acetonide (15) (40.2g, 99.5% purity: 78.97mmol, 1.0 equiv. [adjusted
for purity], from Section 3.4) was dissolved in methanol (500mL) and
cooled to 0°C under nitrogen.
2. Camphorsulfonic acid (55.04g, 236.93mmol, 3.0 equiv.) was added at
0°C in one portion.
3. The reaction mixture was allowed to warm to 25°C and stirred for 12h
with monitoring by LC-MS analysis of aliquots.
4. Upon completion, the reaction mixture was quenched by addition of
solid sodium bicarbonate until the pH reached 7–8, then concentrated
by rotary evaporation.

80
Nihar Kinarivala et al.
5. The residue was purified by silica gel chromatography (ethyl acetate
!5:1 ethyl acetate/methanol) to afford 25g of partially purified product.
6. The partially purified product was further purified by preparative
HPLC (column: Phenomenex Gemini C18 250ꢃ50mm, 10μm;
mobile phase: water (0.05% v/v ammonia hydroxide) with 1–18% ace-
tonitrile, 28min). Lyophilization afforded salicyl-AMS (free acid) (1)
(10.05g, 99.9% purity: 21.53mmol, 27.3% yield [adjusted for purity])
as a white solid.
7. 1H NMR (CD3OD, 400MHz): δ 8.57 (s, 1H), 8.17 (s, 1H), 7.93 (dd,
J1 ¼1.6Hz, J2 ¼8.0Hz, 1H), 7.29 (td, J1 ¼1.6Hz, J2 ¼7.2Hz, 1H),
6.80–6.75 (m, 2H), 6.09 (d, J¼6.4Hz, 1H), 4.72 (t, J¼5.6Hz, 1H),
4.42–4.40 (m, 2H), 4.37 (dd, J1 ¼3.2Hz, J2 ¼11.2Hz, 1H),
4.33–4.31 (m, 1H).
3.6 Conversion to triethylammonium salt (1ꢀEt3N)
1. In a 250-mL round bottom flask, salicyl-AMS (free acid) (1) (2.5g,
5.36mmol, from Section 3.5) was dissolved in 200mL methanol
(27mM).
2. The mixture was sonicated in a water bath for 5–10min to increase
solubilization.
3. The mixture was filtered using a 0.2μM syringe filter.
4. The filtered solution was transferred to a 500-mL round bottom flask
and cooled to ꢂ20°C with stirring for 10min.
5. Note: Salicyl-AMS may precipitate at temperatures below ꢂ25°C.
6. Triethylamine (11.2mL, 80.4mmol, 15 equiv.) was added dropwise
at ꢂ20°C and the solution was stirred for 30min at ꢂ20°C.
7. Note: Freshly distilled triethylamine was used.
8. The reaction mixture was allowed to warm to room temperature and
concentrated by rotary evaporation.

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
81
9. The residue was purified by silica flash chromatography (4:1 ethyl
acetate/methanol+1% triethylamine) to afford salicyl-AMS, tri-
ethylammonium salt (1ꢀEt3N) (2.61g, 4.60mmol, 85.8% yield) as
a white solid.
10. Note: The crude product was dry-loaded onto the silica gel column
by dissolving in 200mL methanol, addition of 10g Celite, rotary
evaporation, and transfer of the Celite onto the column.
11. 1H NMR (CD3OD, 600MHz): δ 8.41 (s, 1H), 8.13 (s, 1H), 7.94
(d, J¼7.8Hz, 1H), 7.28 (t, J¼7.2Hz, 1H), 6.72–6.87 (m, 2H), 6.08
(d, J¼6.4Hz, 1H), 4.69 (t, J¼5.6Hz, 1H), 4.24–4.34 (m, 4H), 3.17
(q, J¼7.2Hz, 6H), 1.33 (t, J¼7.2Hz, 9H) (Somu, Wilson,
et al., 2006).
3.7 Conversion to sodium salt (1ꢀNa)
Table 1 Elemental analysis of salicyl-AMS, sodium salt (1ꢀNa).
Sample
%C
%H
%N
%S
%Na
4.47
4.71
Experimental
Calculated
38.65
41.81
3.89
3.51
15.54
17.21
6.39
6.56
1. Dowex 50WX8 (hydrogen form) (1000g) was placed in a glass column
(3 in inside diameter ꢃ 12 in length, with 2000mL reservoir).
2. The resin on the column was sequentially washed with water
(5ꢃ1000mL), methanol (5ꢃ1000mL), water (5ꢃ1000mL), acetoni-
trile (5ꢃ1000mL), water (5ꢃ1000mL) and 1N aqueous sodium
hydroxide (5ꢃ1000mL).
3. The resin on the column was washed repeatedly with water until the
eluent reached pH 7.

82
Nihar Kinarivala et al.
4. Salicyl-AMS, triethylammonium salt (1ꢀEt3N) (2.61g, 4.60mmol, from
Section 3.6) was dissolved in minimal amount of 1:1 water/acetonitrile
(ꢁ50mL).
5. The solution was loaded onto the Dowex column and incubated with
the resin for 10min.
6. The column was eluted with water (1000mL) and 25-mL fractions were
collected and analyzed by spotting on a TLC plate and visualizing under
UV light (254nm).
7. Fractions that were UV active were combined and lyophilized to afford
salicyl-AMS, sodium salt (1ꢀNa) (2.12g, 4.34mmol, 94.4% yield, 99.9%
purity) as a fluffy white solid.
8. 1H NMR (CD3OD, 600MHz): δ 8.55 (s, 1H), 8.19 (s, 1H), 7.97
(d, J¼7.6Hz, 1H), 7.31 (t, J¼7.3Hz, 1H), 8.70–6.84 (m, 2H), 6.12
(d, J¼6.4Hz, 1H), 4.73 (t, J¼5.5Hz, 1H), 4.34–4.44 (m, 4H).
9. Elemental analysis was performed to determine the percentage of car-
bon, hydrogen, nitrogen, sulfur, and sodium (Robertson Microlit).
The sulfur:sodium ratio was calculated at 1.43:1, comparable to the
theoretical ratio of 1.39:1 (1:1molar ratio) (Table 1).
3.8 LC-MS/MS analysis of salicyl-AMS, sodium salt (1ꢀNa)
to quantify AMS impurity
3.8.1 Preparation of samples
1. Preparation of salicyl-AMS test sample solution: salicyl-AMS, sodium
salt (1ꢀNa) (2mg) was dissolved in 0.4mL of 9:1 phosphate buffered
saline (PBS)/dimethylsulfoxide (DMSO) to generate a 5mg/mL stock
solution.
2. Note: The salicyl-AMS, sodium salt (1ꢀNa) solution should be freshly
prepared before use.
3. Preparation of guanosine internal standard solution: guanosine (2.5mg)
was dissolved in 0.5mL DMSO to generate a 5mg/mL stock solution.
The stock solution was diluted 1000-fold with PBS to generate a
5μg/mL working solution.
4. A calibration curve of AMS was prepared as follows:
a. 50-O-Sulfamoyladenosine (AMS, 6) (Ferreras et al., 2005) (2mg) was
dissolved in 0.4mL DMSO to generate a 5mg/mL stock solution.
b. The AMS stock solution was diluted 100-fold with PBS to generate
a 50μg/mL working solution.
c. The 50μg/mL AMS working solution was serially diluted twofold
with PBS 12 times to generate calibration curve solutions at 50,

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
83
25, 12.5, 6.25, 3.13, 1.56, 0.781, 0.391, 0.195, 0.0977, 0.0488,
0.0244, 0.0122μg/mL. A 0μg/mL blank was also generated with
PBS, for a total of 14 samples.
3.8.2 Preparation of 96-well plate
1. 100μL of each salicyl-AMS calibration curve solution (from
Section 3.8.1, step 4) was added to each well of 96-well plate, along with
100μL of the internal standard working solution (from Section 3.8.1,
step 3). The experiment was conducted in triplicates.
2. Final concentrations:
a. Guanosine (internal standard): 2.5μg/mL in all wells.
b. AMS calibration curve: 0.061–25μg/mL.
3. Note: When calculating the concentrations of the internal standard and
of AMS in the samples, a dilution factor of 2 needs to be applied
compared to the working solutions.
3.8.3 Method for LC-MS/MS analysis
LC-MS/MS analysis was carried out on an Agilent Technologies 6410
triple quadrupole LC-MS/MS system with autosampler in electrospray ion-
ization (ESI) mode, with an Agilent Zorbax Eclipse XBD-C18 reverse phase
column (50ꢃ4.6mm, 5μm) using a flow rate of 0.5mL/min and an
isocratic mobile phase of 10% acetonitrile in 0.1% aqueous formic acid over
10min. Positive electrospray in multiple reaction monitoring (MRM) mode
was employed to quantify AMS, using the ion transitions indicated below.
Compound
AMS (6)
Precursor ion
347.1
Product ion
136.1
Fragmentor
Collision energy
128
71
20
8
Guanosine
284.1
152.1
4. Concluding remarks
We have developed an efficient protocol for the multigram-scale
preparation of the anti-tuberculosis lead compound salicyl-AMS as its
sodium salt (1ꢀNa). The process proceeds in three synthetic steps, followed
by a two-step salt formation protocol, to afford the target compound in

84
Nihar Kinarivala et al.
11.6% overall yield from commercially available adenosine 20,30-acetonide
(13). The two-step salt formation process includes an additional chromato-
graphic purification and removes both a cycloadenosine byproduct (7) and
a toxic trace impurity AMS (6), providing highly purified material suitable
for preclinical evaluation. This protocol has been applied to milligram to
multigram-scale syntheses, and with minor modifications to other salicyl-
AMS analogs (Bythrow et al., 2019). We envision that this protocol
may also be useful in the preparation of other acyl-AMS inhibitors of
the adenylate-forming enzyme superfamily (Lux et al., 2019).
Acknowledgments
We thank Kazuyoshi Aso (Tri-Institutional Therapeutics Discovery Institute and Axcelead
Drug Discovery Partners) and Makoto Fushimi (Tri-Institutional Therapeutics Discovery
Institute and Takeda Pharmaceuticals) for helpful discussions and for coordinating large-
scale synthesis of salicyl-AMS (free acid) at WuXi AppTec. Generous financial support
was provided by the National Institutes of Health (R01 AI118224 to D.S.T.; F31
AI129244 and T32 GM073546-Gross to L.C.S.; and CCSG P30 CA008748 to C.B.
Thompson) and the Tri-Institutional Therapeutics Discovery Institute.
Conflicts of interests
D.S.T. is an inventor on U.S. Patents 8,461,128 and 8,946,188, and D.S.T. and L.C.S. are
inventors on U.S. Patent Application PCT/US2019/068107 based on this work.
References
€
Appel, R., & Berger, G. (1958). Hydrazinsulfons€aure-amide, I. Uber das Hydrazodisulfamid.
Chemische Berichte, 91(6), 1339–1341.
Bloch, A., & Coutsogeorgopoulos, C. (1971). Inhibition of protein synthesis by 50-sul-
famoyladenosine. Biochemistry, 10(24), 4395–4398. https://doi.org/10.1021/bi00800a007.
Brodsky, B. H., & Du Bois, J. (2005). Oxaziridine-mediated catalytic hydroxylation of
unactivated 3° C-H bonds using hydrogen peroxide. Journal of the American Chemical
Society, 127(44), 15391–15393. https://doi.org/10.1021/ja055549i.
Bythrow, G. V., Mohandas, P., Guney, T., Standke, L. C., et al. (2019). Kinetic analyses of
the siderophore biosynthesis inhibitor salicyl-AMS and analogues as MbtA inhibitors and
antimycobacterial agents. Biochemistry, 58(6), 833–847. https://doi.org/10.1021/acs.
biochem.8b01153.
Cisar, J. S., Ferreras, J. A., Soni, R. K., Quadri, L. E. N., et al. (2007). Exploiting ligand
conformation in the selective inhibition of non-ribosomal peptide synthetase amino acid
adenylation with designed macrocyclic small molecules. Journal of the American Chemical
Society, 129, 7752–7753. https://doi.org/10.1021/ja0721521.
Cisar, J. S., & Tan, D. S., unpublished results.
Dawadi, S., Boshoff, H. I. M., Park, S. W., Schnappinger, D., et al. (2018). Conformationally
constrained cinnolinone nucleoside analogues as siderophore biosynthesis inhibitors for
tuberculosis. ACS Medicinal Chemistry Letters, 9(4), 386–391. https://doi.org/10.1021/
acsmedchemlett.8b00090.

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
85
Dawadi, S., Kawamura, S., Rubenstein, A., Remmel, R., et al. (2016). Synthesis and phar-
macological evaluation of nucleoside prodrugs designed to target siderophore biosynthe-
sis in Mycobacterium tuberculosis. Bioorganic and Medicinal Chemistry, 24(6), 1314–1321.
https://doi.org/10.1016/j.bmc.2016.02.002.
Dawadi, S., Viswanathan, K., Boshoff, H. I., Barry, C. E., 3rd, et al. (2015). Investigation and
conformational analysis of fluorinated nucleoside antibiotics targeting siderophore bio-
synthesis. Journal of Organic Chemistry, 80(10), 4835–4850. https://doi.org/10.1021/acs.
joc.5b00550.
De Voss, J. J., Rutter, K., Schroeder, B. G., & Barry, C. E., 3rd (1999). Iron acquisition and
metabolism by mycobacteria. Journal of Bacteriology, 181(15), 4443–4451.
De Voss, J. J., Rutter, K., Schroeder, B. G., Su, H., et al. (2000). The salicylate-derived
mycobactin siderophores of Mycobacterium tuberculosis are essential for growth in macro-
phages. Proceedings of the National Academy of Sciences of the United States of America, 97(3),
1252–1257. https://doi.org/10.1073/pnas.97.3.1252.
Duckworth, B. P., Geders, T. W., Tiwari, D., Boshoff, H. I., et al. (2011). Bisubstrate
adenylation inhibitors of biotin protein ligase from Mycobacterium tuberculosis. Chemistry
and Biology, 18(11), 1432–1441. https://doi.org/10.1016/j.chembiol.2011.08.013.
Duckworth, B. P., Nelson, K. M., & Aldrich, C. C. (2012). Adenylating enzymes in
Mycobacterium tuberculosis as drug targets. Current Topics in Medicinal Chemistry, 12(7),
766–796. https://doi.org/10.2174/156802612799984571.
Engelhart, C. A., & Aldrich, C. C. (2013). Synthesis of chromone, quinolone, and ben-
zoxazinone sulfonamide nucleosides as conformationally constrained inhibitors of
adenylating enzymes required for siderophore biosynthesis. Journal of Organic Chemistry,
78(15), 7470–7481. https://doi.org/10.1021/jo400976f.
Ferreras, J. A., Ryu, J. S., Di Lello, F., Tan, D. S., et al. (2005). Small-molecule inhibition of
siderophore biosynthesis in Mycobacterium tuberculosis and Yersinia pestis. Nature Chemical
Biology, 1(1), 29–32. https://doi.org/10.1038/nchembio706.
Ferreras, J. A., Stirrett, K. L., Lu, X., Ryu, J.-S., et al. (2008). Mycobacterial phenolic gly-
colipid virulence factor biosynthesis: Mechanism and small-molecule inhibition of poly-
ketide chain initiation. Chemistry and Biology, 15(1), 51–61. https://doi.org/10.1016/
j.chembiol.2007.11.010.
Furin, J., Cox, H., & Pai, M. (2019). Tuberculosis. Lancet, 393(10181), 1642–1656. https://
doi.org/10.1016/S0140-6736(19)30308-3.
Gupte, A., Boshoff, H. I., Wilson, D. J., Neres, J., et al. (2008). Inhibition of siderophore
biosynthesis by 2-triazole substituted analogues of 5´-O-[N-(salicyl)sulfamoyl]adeno-
sine: Antibacterial nucleosides effective against Mycobacterium tuberculosis. Journal of
Medicinal Chemistry, 51(23), 7495–7507. https://doi.org/10.1021/jm8008037.
Heacock, D., Forsyth, C. J., Shiba, K., & Musier-Forsyth, K. (1996). Synthesis and
aminoacyl-tRNA synthetase inhibitory activity of prolyl adenylate analogs. Bioorganic
Chemistry, 24(3), 273–289. https://doi.org/10.1006/bioo.1996.0025.
Jaffe, J. J., McCormack, J. J., & Meymerian, E. (1970). Trypanocidal properties of 5´-O-sul-
famoyladenosine, a close structural analog of nucleocidin. Experimental Parasitology,
28(3), 535–543. https://doi.org/10.1016/0014-4894(70)90121-9.
Krajczyk, A., Zeidler, J., Januszczyk, P., Dawadi, S., et al. (2016). 2-Aryl-8-aza-3-
deazaadenosine analogues of 5´-O-[N-(salicyl)sulfamoyl]adenosine: Nucleoside antibi-
otics that block siderophore biosynthesis in Mycobacterium tuberculosis. Bioorganic and
Medicinal Chemistry, 24(14), 3133–3143. https://doi.org/10.1016/j.bmc.2016.05.037.
Lamb, A. L. (2015). Breaking a pathogen’s iron will: Inhibiting siderophore production as an
antimicrobial strategy. Biochimica et Biophysica Acta, 1854(8), 1054–1070. https://doi.org/
10.1016/j.bbapap.2015.05.001.
Lun, S., Guo, H., Adamson, J., Cisar, J. S., et al. (2013). Pharmacokinetic and in vivo efficacy
studies of the mycobactin biosynthesis inhibitor salicyl-AMS in mice. Antimicrobial Agents
and Chemotherapy, 57(10), 5138–5140. https://doi.org/10.1128/aac.00918-13.

86
Nihar Kinarivala et al.
Lux, M. C., Standke, L. C., & Tan, D. S. (2019). Targeting adenylate-forming enzymes with
designed sulfonyladenosine inhibitors. Journal of Antibiotics (Tokyo), 72(6), 325–349.
https://doi.org/10.1038/s41429-019-0171-2.
McMahon, M. D., Rush, J. S., & Thomas, M. G. (2012). Analyses of MbtB, MbtE, and MbtF
suggest revisions to the mycobactin biosynthesis pathway in Mycobacterium tuberculosis.
Journal of Bacteriology, 194(11), 2809–2818. https://doi.org/10.1128/JB.00088-12.
Nelson, K. M., Viswanathan, K., Dawadi, S., Duckworth, B. P., et al. (2015). Synthesis and
pharmacokinetic evaluation of siderophore biosynthesis inhibitors for Mycobacterium
tuberculosis. Journal of Medicinal Chemistry, 58(14), 5459–5475. https://doi.org/10.1021/
acs.jmedchem.5b00391.
Neres, J., Labello, N. P., Somu, R. V., Boshoff, H. I., et al. (2008). Inhibition of siderophore
biosynthesis in Mycobacterium tuberculosis with nucleoside bisubstrate analogues:
Structure–activity relationships of the nucleobase domain of 5´-O-[N-(salicyl)sul-
famoyl]adenosine. Journal of Medicinal Chemistry, 51(17), 5349–5370. https://doi.org/
10.1021/jm800567v.
Okada, M., Iwashita, S., & Koizumi, N. (2000). Efficient general method for sulfamoylation
of a hydroxyl group. Tetrahedron Letters, 41(36), 7047–7051.
Paulekuhn, G. S., Dressman, J. B., & Saal, C. (2007). Trends in active pharmaceutical ingre-
dient salt selection based on analysis of the Orange Book database. Journal of Medicinal
Chemistry, 50(26), 6665–6672. https://doi.org/10.1021/jm701032y.
Qiao, C., Gupte, A., Boshoff, H. I., Wilson, D. J., et al. (2007). 50-O-[(N-acyl)sulfamoyl]
adenosines as antitubercular agents that inhibit MbtA: An adenylation enzyme required
for siderophore biosynthesis of the mycobactins. Journal of Medicinal Chemistry, 50(24),
6080–6094. https://doi.org/10.1021/jm070905o.
Qiao, C., Wilson, D. J., Bennett, E. M., & Aldrich, C. C. (2007). A mechanism-based aryl
carrier protein/thiolation domain affinity probe. Journal of the American Chemical Society,
129(20), 6350–6351. https://doi.org/10.1021/ja069201e.
Quadri, L. E. N., & Ratledge, C. (2005). Iron metabolism in the tubercle bacillus and other
mycobacteria. In S. T. Cole, K. D. Eisenach, & D. N. McMurrayet al.Tuberculosis and the
tubercle bacillus (pp. 341–357). Washington, DC: ASM Press.
Quadri, L. E., Sello, J., Keating, T. A., Weinreb, P. H., et al. (1998). Identification of a
Mycobacterium tuberculosis gene cluster encoding the biosynthetic enzymes for assembly
of the virulence-conferring siderophore mycobactin. Chemistry and Biology, 5(11),
631–645. https://doi.org/10.1016/s1074-5521(98)90291-5.
Ratledge, C. (2004). Iron, mycobacteria and tuberculosis. Tuberculosis (Edinburgh), 84(1–2),
110–130. https://doi.org/10.1016/j.tube.2003.08.012.
Rengaraju, S., Narayanan, S., Ganju, P., Amin, M., et al. (1986). 50-O-Sulfamoyladenosine
(defluoronucleocidin) from a Streptomyces. Meiji Seika Kenkyu Nenpo, 25, 49–55.
Shi, C., Tiwari, D., Wilson, D. J., Seiler, C. L., et al. (2013). Bisubstrate inhibitors of biotin
protein ligase in Mycobacterium tuberculosis resistant to cyclonucleoside formation. ACS
Medicinal Chemistry Letters, 4(12), 1213–1217. https://doi.org/10.1021/ml400328a.
Shuman, D., Robins, R. K., & Robins, M. J. (1969). The synthesis of adenine 5´-O-
sulfamoyl nucleosides related to nucleocidin. Journal of the American Chemical Society,
91(12), 3391–3392. https://doi.org/10.1021/ja01040a062.
Somu, R. V., Boshoff, H., Qiao, C., Bennett, E. M., et al. (2006). Rationally designed nucle-
oside antibiotics that inhibit siderophore biosynthesis of Mycobacterium tuberculosis. Journal
of Medicinal Chemistry, 49(1), 31–34. https://doi.org/10.1021/jm051060o.
Somu, R. V., Wilson, D. J., Bennett, E. M., Boshoff, H. I., et al. (2006). Antitubercular
nucleosides that inhibit siderophore biosynthesis: SAR of the glycosyl domain. Journal
of Medicinal Chemistry, 49(26), 7623–7635. https://doi.org/10.1021/jm061068d.

Gram-scale preparation of the antibiotic lead compound salicyl-AMS
87
Van Brussel, W., & Van Sumere, C. F. (1978). N-Acylamino acids and peptides. VI. A simple
synthesis of N-acylglycines of the benzoyl and cinnamyl type. Bulletin des Societes
Chimiques Belges, 87(10), 791–797. https://doi.org/10.1002/bscb.19780871008.
Vannada, J., Bennett, E. M., Wilson, D. J., Boshoff, H. I., et al. (2006). Design, synthesis,
and biological evaluation of beta-ketosulfonamide adenylation inhibitors as potential
antitubercular agents. Organic Letters, 8(21), 4707–4710. https://doi.org/10.1021/
ol0617289.
World Health Organization. (2019). Global tuberculosis report 2019. https://www.who.int/tb/
publications/global_report/en/.