A Stereocontrolled Synthesis of a Phosphorothioate Cyclic Dinucleotide-Based STING Agonist

Dinucleotide-Based STING Agonist

Article

A Stereocontrolled Synthesis of a Phosphorothioate Cyclic

James Kempson,* Huiping Zhang, Xiaoping Hou, Lyndon Cornelius, Rulin Zhao, Bei Wang,

Zhenqiu Hong, Martins S. Oderinde, Joseph Pawluczyk, Dauh-Rurng Wu, Dawn Sun, Peng Li,

Shiuhang Yip, Aaron Smith, Janet Caceres-Cortes, Darpandeep Aulakh, Amy A. Sarjeant, Peter K. Park,

Lalgudi S. Harikrishnan, Lan-ying Qin, Dharmpal S. Dodd, Brian Fink, Gregory Vite, and Arvind Mathur

Cite This: J. Org. Chem. 2021, 86, 8851 −8861

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: We describe a stereodeﬁned synthesis of the newly

identiﬁed non-natural phosphorothioate cyclic dinucleotide (CDN)

STING agonist, BMT-390025. The new route avoids the low-

yielding racemic approach using P(III)-based reagents, and the

stereospeciﬁc assembly of the phosphorothioate linkages are forged

via the recently invented P(V)-based platform of the so-called PSI

(Ψ) reagent system. This P(V) approach allows for the complete

control of chirality of the P-based linkages and enabled conclusive

evidence of the absolute conﬁguration. The new approach oﬀers

robust procedures for preparing the stereodeﬁned CDN in eight

steps starting from advanced nucelosides, with late-stage direct drop

isolations and telescoped steps enabling an eﬃcient scale-up that

proceeded in an overall 15% yield to produce multigram amounts of

the CDN.

INTRODUCTION

■

The stimulator of interferon genes (STING) targeted

immunotherapies have gained much attention in drug

discovery since the adaptor protein was discovered in 2008.

′,3′-Cyclic guanosine monophosphate-adenosine monophos-

phate (2′,3′-cGAMP) (Figure 1) is the representative structure

of one of the four known human STING natural agonists and

corresponds to the cyclic dinucleotide (CDN) class of natural

Figure 2. BMT-390025.

BMT-390025 is characterized as a CDN bearing two bridging

phosphorothioate linkages, which impart four potential

phosphorus diastereoisomers. Seven additional carbon-based

stereocenters come from the “linear” tricyclic guanosine-

products.

derived nucleoside and the lobucavir-inspired cyclobutane,

which constitute the novel design elements of this complex

molecule.

With the successful identiﬁcation of BMT-390025, further

supplies of material were required to support toxicology

studies, with access to multigram quantities needed. The

original synthesis had been developed by relying on the

Figure 1. 2′,3′-cGAMP.

Much of the medicinal chemistry eﬀort over the past decade

has been characterized by modiﬁcations originating from these

known CDNs with novel base, sugar, and phosphorus linkages

explored in an eﬀort to produce new CDNs that inﬂuence

STING-pathway agonism in addition to exhibiting favorable

Received: April 3, 2021

Published: June 14, 2021

−5

drug-like properties.

One such optimization eﬀort has been

successful at the identiﬁcation of the potent non-natural

STING agonist, BMT-390025 (Figure 2). Interestingly,

2021 American Chemical Society

851

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

Scheme 1. Construction of the Phosphorothioate via the Original P(III) Route

classical phosphoramidite chemistry that ﬁrst appeared in the

load and couple) from starting nucleoside 6 to construct the

stereopure phosphorothioate linkages of the CDN. We decided

to use a stepwise approach to CDN synthesis as this has

previously been reported to ensure the complete transfer of

stereochemical information from the Ψ-reagent by forging the

CDN macrocycle one P−O bond at a time. As such, we

proposed loading of the 2′-position in nucleoside 6 with the

(R)-Ψ reagent to give the (S)-loaded 7. Coupling of the

second nucleoside 5 was then envisioned to deliver the ﬁrst

phosphorothioate linkage 8 with a stereochemically pure (R)-

conﬁguration (Scheme 2a). A second loading and iteration

using the (S)-Ψ reagent through a 5′-loading would then

generate the second stereodeﬁned phosphorus with an (R)-

conﬁguration 10 (Scheme 2b). Subjecting 10 to macro-

cyclization and a deprotection sequence would provide

enantiomerically and diastereomerically pure BMT-390025.

At the outset of this proposed new synthesis, the

stereochemistry of the phosphorothioate linkages in BMT-

literature during the 1970s and which has more recently found

application in the synthesis of c-di-GMP thiophosphate. Our

strategy involved the loading of an appropriately protected

nucleoside to give the low valent phosphoramidite 2, as

depicted in Scheme 1. Subsequent coupling of advanced

nucleoside 5 followed by oxidative sulfurization generated the

ﬁrst phosphorothioate linkage as a mixture of epimers.

Protecting group manipulations to allow for an analogous

P(III)-sulfurization sequence then closed the CDN ring and

delivered a mixture of four phosphorus-centered diaster-

eoisomers that were separated by HPLC chromatography.

These eﬀorts produced an initial milligram amount of BMT-

90025 having unknown stereochemistry at the phosphorus

centers. In addition to the challenges associated with the

diastereomeric phosphorothioate construction, the synthesis

suﬀered from poor overall yields with the protected

phosphoramidite monomer 2 and cyclobutane coupling

partner 4, which proved to be diﬃcult to synthesize on a

scale. Despite the P(III)-based coupling being rapid, it was

highly moisture-sensitive, leading to variable results even after

repeated azeotropes. The use of hazardous reagents such as

H S gas, 1H-tetrazole, and CCl , together with the signiﬁcant

90025 arising from the P(III) synthesis was unknown; a

tentative assignment was made based on the phosphorothioate

stereochemistry of related CDN compounds having single-

crystal X-ray data. The order of assembly and selection of Ψ-

reagents would not only allow for the interception of common

nucleoside intermediates used in the previous P(III) synthesis

but would also allow us to conﬁrm the proposed stereo-

chemical assignment via this programmed approach.

puriﬁcation diﬃculties, made this route unlikely to be scaled

beyond milligram quantities.

Due to the aforementioned challenges of phosphorothioate

synthesis via low valent P(III), we were intrigued by the recent

discovery of a P(V)-based reagent platform called the PSI (Ψ)

Nucleoside Monomer Synthesis. “Linear” Tricyclic

Guanosine. The advanced tricyclic guanosine 6 was prepared

11,6

in 6 steps according to literature precedent.

Orchestrating

reagent. This reagent has already been utilized in the

the appropriate protecting group strategy for this triply

protected fragment presented several challenges when

performed on-scale. The nonselective OTBS protection of

diol 11 using the standard conditions of TBSCl and imidazole

in DMF at room temperature gave a ∼1:1 regioisomeric

mixture of protected 2′- and 3′-alcohols, 12 and 6, respectively

stereocontrolled synthesis of various nucleotidic architectures,

including antisense oligonucleotides (ASOs) and CDNs, and

most recently has found further application in the stereo-

speciﬁc synthesis of chiral phosphines and methylphosphonate

nucleotides. As such, BMT-390025 presented itself as an

ideal candidate on which to test the scalability of such an

approach. Reported herein are our eﬀorts culminating in a

stereocontrolled and scalable synthesis of this non-natural

CDN utilizing this novel reagent system.

(

Scheme 3). Separation by silica gel chromatography was

suﬃcient for accessing the desired 3′-protected compound 6

on a gram scale, albeit in a modest yield of 31%. Increasing the

scale of this reaction to >10 g led to the observation of

signiﬁcant (5−10%) silyl group migration back to 12, which

was occurring post puriﬁcation during the evaporation of

puriﬁed fractions.

This undesired scrambling hampered our ability to control

key impurities and exacerbated downstream puriﬁcation

diﬃculties when this mixture was used in the subsequent Ψ-

RESULTS AND DISCUSSION

■

Synthesis Design Using Ψ-Reagents. Scheme 2 shows

the proposed modular stereocontrolled CDN synthesis, which

entails a marked divergence from the P(III) approach,

requiring only four steps ((a) 2′-load and couple; (b) 5′-

852

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

Scheme 2. Proposed Stepwise Ψ-Loading and Coupling Steps for Stereocontrolled Phosphorothioate Synthesis

Scheme 3. Nonselective Mixture of 2′- and 3′-Protected Nucleosides

NPE = p-nitrophenethyl.

loading step. Attempts to leverage existing chemistry around

selective protection methods included the use of silver nitrate

reported by Ogilvie and the use of a scaﬀolding catalyst

reported by Tan; however, both failed to show any

which they had subjected a 3′-silyl isomer to 1% Et N in

MeOH in order to isomerize to a 1:1 mixture of 2′- and 3′-

isomers, we decided to focus on surveying solvents for the

already observed partial migration. Methanol alone was a

suitable solvent to aﬀect the migration, although extended

3,14

improvement.

Upon the basis of a report by Usman, in

853

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

times (48 h) were needed to achieve 50% migration (Table 1).

an initial route was developed that comprised 5 steps from 13

to the advanced intermediate 18, which proceeded in 11%

overall yield (Scheme 4). The lowest yielding reaction was the

Following the Usam protocol, the addition of 1% Et N further

Table 1. Solvent Screen Used for TBDMS Migration and the

Recycling of Undesired Regioisomer 12

nonselective LiBH -mediated deprotection of the benzoylated

alcohol proximal to the hydroxycyclobutane, 16 → 17.

Attempts to improve on this deprotection centered around

invoking the previously reported

bicyclic syn-Alanate

transition states encountered during the ketone reduction

step, although all attempts at selective deprotection, including

LiAlH and BH , failed.

Our focus to avoid this problematic step led us to the work

of Lee-Ruﬀ on the regioselective porcine pancreatic lipase

conditions

results

(

PPL) enzyme-catalyzed acylation and hydrolysis of cyclo-

dichloromethane, rt, 48 h

methanol, rt, 48 h

acetonitrile, rt, 48 h

ethyl acetate, rt, 48 h

THF, rt, 48 h

no signiﬁcant migration observed

50% conversion

butanone derivatives. Initial investigations began with the

PPL enzyme-catalyzed hydrolysis performed on diacetate, 19,

which was formed from the bis-acetylation of diol 13 with

acetyl chloride in pyridine (Scheme 5). Well-aligned with Lee-

Ruﬀ’s observation, the PPL-catalyzed bioconversion resulted in

selective hydrolysis of the C-3 acetoxymethyl group leaving the

C-2 acetoxymethyl intact, although isolated yields of 20

remained low due to emulsion issues encountered during

workup, particularly with gram-scale reactions (Scheme 5).

Attempts to mitigate these issues included the addition of

MeOH to denature the enzyme but always resulted in a loss of

the product to the aqueous layer during the subsequent

extractive workup. Salting-out (K CO or Na SO ) and

10% conversion

20% conversion

methanol, Et N, 50 °C, 6 h

50% conversion, decomposition evident

50% conversion, clean proﬁle

complex mixture

IPA, Et N, 50 °C, 4 h

IPA, Et N, 80 °C, 1 h

accelerated the migration, with 50% Et N in isopropanol (IPA)

at 50 °C ultimately selected as the optimal combination to

eﬀect clean migration within 4 h. In the event, we developed a

practical supercritical ﬂow chromatography (SFC) method

for this separation, which allowed for rapid puriﬁcation on a

multigram scale and was combined with continuous and

immediate evaporation postpuriﬁcation, resulting in <1%

observed migration on a scale greater than 100 g. The

undesired regioisomer was subjected to the IPA-Et N

conditions to eﬀectively “recycle” into a 1:1 mixture within 5

h on a 100 g scale. Evaporation and resubjecting the material

to SFC allowed the eﬀective yield for this transformation to be

increased from the aforementioned 31% on a gram scale to

16c

ﬁltration-based (addition of Celite) workup procedures failed

to improve on the low isolated yield.

A nonaqueous, selective enzyme-mediated acylation strategy

was the next option. To our delight, PPL-mediated acetylation

of diol 13 with vinyl acetate and toluene as the solvent mixture

resulted in the regioselective acetylation of the C-3

hydroxymethyl group, giving monoacetate 21 (Scheme 6).

Workup now consisted of a simple ﬁltration to remove the

enzyme and evaporation of the ﬁltrate prior to puriﬁcation by

silica gel chromatography. Optimal reaction conditions

necessitated the use of 150 mol % PPL in order to aﬀord

100% conversion in 12 h. Yields of >90% were obtained after

isolation, and the reaction scaled well (>100 g), utilizing the

same conditions. To complete the synthesis of nucleoside 5, a

modiﬁed order of previously utilized transformations allowed

the interception of the appropriately protected cyclobutane 4

(Scheme 6). This new four-step sequence delivered 4 in an

overall 67% yield from diol 13 and represented a signiﬁcant

improvement over the previous ﬁve steps, 11% overall yield of

5−75% on a scale greater than100 g, dependent on the

number of times the SFC puriﬁcation−recycle protocol was

repeated.

Cyclobutane Fragment. A diastereoselective [2+2] cyclo-

addition reported on a kilogram scale was successfully

employed to deliver the known intermediate 13 on a scale

greater than 500 g. However, in contrast to the synthesis of

lobucavir, this key fragment now necessitated a late-stage

orthogonal protecting group strategy for the successful

incorporation into the CDN class of molecules. To that end,

Scheme 4. Original Protecting Group Strategy Developed from the Lobucavir Common Intermediate

854

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

Scheme 5. PPL-Mediated Hydrolysis of Diacetate, 19

Scheme 6. Selective Enzyme-Mediated Acylation and Advancement to a Key Nucleoside, 5

Scheme 7. Stereoselective Phosphorothioate Construction via First Load and Couple Sequence with (R)-Ψ Reagent

8 from the same diol intermediate (Scheme 4). An analogous

End-Game Synthesis Enabled by Ψ-Reagents. With

nucleosides 5 and 6 in hand, we commenced with the

construction of the ﬁrst stereocontrolled phosphorothioate

linkage by loading the 2′-alcohol of 6 with the (R)-Ψ reagent

Mitsunobu and aminolysis sequence delivered the advanced

nucleoside 5 in 75% yield for the two-step procedure on a

multigram scale. The regiochemical and stereochemical

structural integrity of 5 was conﬁrmed by a single-crystal X-

ray structure determination.

(

Scheme 7). Rapid addition of DBU to an acetonitrile solution

of 6 and the (R)-Ψ reagent yielded S -loaded building block 7

within 5 min. Quenching with acetic acid followed by

855

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

Scheme 8. Deprotection and Second Loading Step with (S)-Ψ Reagent

Scheme 9. End Game to Achieve BMT-390025

evaporation and puriﬁcation by column chromatography

evident that, over prolonged reaction times (>10 min) and

during the evaporation process, the impurity proﬁle of the

reaction mixture accumulated byproducts related to the P(V)-

aﬀorded 7 in 90% isolated yield as a single S diastereoisomer

as conﬁrmed by P NMR. By virtue of the rapid loading step,

large-scale reactions (>30 g) performed equally well, with no

observation of the previously observed OTBS migration. The

subsequent coupling step between 7 and 5 proved equally as

facile and was complete within 10 min to give dinucleotide 8 as

reagent, which contributed to product decomposition.

Methanolysis was also observed postcolumn chromatography,

so eﬀorts were focused on rapid, nonchromatographic isolation

of loaded intermediate 10. A direct drop isolation method was

developed in which the reaction was dropped onto a 10-fold

volume excess of MTBE to eﬀectively precipitate the desired

product. Removal of the supernatant and a citric acid wash of

the DCM-dissolved product was eﬀective at removing the

residual DBU. This protocol gave consistent results and

allowed for a 100% isolated yield of 10 that was suitable for use

in the next deprotection step.

Trityl deprotection of 10 was complete within 1 h using a 2-

fold excess of DCA (Scheme 9). Given the previously observed

sensitivity of the auxiliary toward ring-opening, we decided to

forego puriﬁcation and telescope 24 into the macrocyclization

step through a simple solvent switch from DCM to MeCN

followed by the addition of DBU. Accompanying the desired

cyclization to 25 was a signiﬁcant quantity (∼35−40%) of

desulfurized product 26. Beaucage has reported similar

a single R diastereoisomer in an excellent 84% isolated yield

following column chromatography.

Selective deprotection of the 5′-DMTr in 8 was achieved

under standard dichloroacetic acid (DCA) and triethylsilane

(

TES) conditions, which gave a 70% isolated yield of 9 after 6

h at rt along with unreacted starting material (Scheme 8).

Attempts to further increase the conversion to 9 with longer

reaction times or increased DCA concentrations led to

concomitant Trt deprotection. The second loading step with

the (S)-Ψ reagent produced R -loaded 10 in a similar manner

to the ﬁrst loading step (Scheme 8). However, despite an

HPLC proﬁle for the reaction showing complete and clean

transformation to 10 within minutes of DBU addition, workup

via an AcOH quench and evaporation followed by column

chromatography gave only a 35% isolated yield. It became

856

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

desulfurization of oligonucleoside phosphorothioates as a

problem emanating from the secondary decomposition

products of trichloroacetic acid (TCA) used in their DMTr

25 were as follows: BEH 2-ethylpyridine (25 cm × 5 cm, 5 μm), 40%

MeOH/H O (95:5) with 10 mM NH HCO in CO , 35 °C, 350

mL/min, 220 nm, 100 bar BPR, 16.81 mg/mL, 3.5 mL/2.3 min. SFC

conditions for compound 27 were as follows: BEH 2-ethylpyridine

deprotections. This prompted us to perform subsequent

trityl deprotections of 10 using only fresh DCA solutions with

an intermediate isolation of 24 implemented from diethyl ether

to give 24 as a crude white solid in 90% yield from 10.

Gratifyingly, cyclization using this material signiﬁcantly

reduced the observed desulfurization product 26 to only 1−

(

25 × 5 cm, 5 μm), 40% MeOH/H O (95:5) with 10 mM

NH HCO in CO , 35 °C, 350 mL/min, 220 nm, 100 bar BPR, 12.56

mg/mL, 3 mL/2 min. SFC conditions for BMT-390025 were as

follows: BEH 2-EP (25 cm × 5 cm, 5 μm), 65% MeOH/ACN/water

(

47.5:47.5:5) with 20 mM ammonium acetate in CO , 230 mL/min,

220 nm, 47 °C, 100 bar BPR, 40 mg/mL, 3.75 mL/3.3 min. NMR

samples were dissolved in approximately 0.65 mL of a suitable

%, which was readily removed by SFC to give 25 as a single

S_p-diastereoisomer at the newly formed phosphorothioate

linkage. Further optimization allowed for NPE deprotection to

occur in one pot directly after cyclization by the addition of 5

equiv of DBU and gentle heating (40 °C). SFC puriﬁcation

allowed penultimate 27 to be isolated, with the eﬀective crude

isolation and telescoping of 4 reactions generating a single

deuterated solvent (dimethyl sulfoxide-d , methanol-d , choroform-d,

6 4

deuterium oxide). NMR spectra were recorded on a 400 MHz Avance

III HD spectrometer equipped with a 5 mm BBFO probe, a 500 MHz

Bruker Avance III HD NMR spectrometer equipped with a 5 mm

BBO Prodigy cryoprobe, or a 700 MHz Bruker Avance III HD NMR

spectrometer equipped with a 5 mm TCI cryoprobe (Bruker, Billerica,

MA). One- and two-dimensional (1D and 2D) NMR spectra for

structure elucidation were collected at 27 °C and include 1D proton,

proton decoupled carbon, proton decoupled phosphorus, 2D COSY

(correlation spectroscopy) multiplicity-edited heteronuclear single

quantum coherence (1H-13C edited HSQC), and heteronuclear

R ,S -diastereoisomer in 46% yield from 9.

Final deprotection using triethylamine trihydrogen ﬂuoride

Et N·3HF) provided clean conversion to BMT-390025;

(

however, diﬃculties associated with the workup and

puriﬁcation of the resulting triethylammonium salt contributed

to signiﬁcant product loss. Optimal conditions using the

readily available and less toxic ammonium ﬂuoride resulted in

complete deprotection after heating in anhydrous MeOH

multiple-bond correlation ( H− C HMBC and H− P HMBC)

experiments. ACD/NMR prediction software (ACD/LABORATO-

RIES Release 2015 Pack 2) and ACD NMR Spectrus Workbook

(

ACD/NMR Workbook 2019.1.1) from Advanced Chemistry

Development Inc. (Toronto, ON, Canada) were used. NMR spectra

are referenced to residual undeuterated solvents. Data for H NMR

overnight. SFC puriﬁcation conditions were developed using

matched” counterion ammonium acetate as the mobile phase,

“

are reported as follows: chemical shift (δ ppm), multiplicity (s =

singlet, d = doublet, dd = doublet of doublet, ddd = doublet of

doublet of doublets, t = triplet, dt = doublet of triplets, td = triplet of

doublets, q = quartet, quint = quintet, m = multiplet, b = broad),

and direct injection of the ﬁnal deprotection solution led to an

eﬃcient ﬁnal puriﬁcation delivering the target compound in

0% yield.

CONCLUSION

integration, and coupling constant (Hz). Data for C NMR and

NMR are reported in terms of chemical shift, and no special

nomenclature is used for equivalent carbons. High-resolution mass

spectra (HRMS) were acquired on a Thermo Fisher QExactive mass

spectrometer operating in electrospray ionization (ESI) mode

■

In summary, we have developed an eﬃcient and multigram

synthesis of the novel CDN, BMT-390025. Key features

included leveraging an observed silyl group migration to

maximize the throughput of advanced nucleoside 6 and the use

of an enzyme-catalyzed selective acylation to streamline the

protecting group strategy to access the nucleoside coupling

partner 5. This route avoided the low-yielding racemic

approach using P(III)-based reagents and the associated

reaction sensitivity, multiple impurities, and intensive use of

chromatographic separation, which further complicated the use

of the original synthetic route on a scale. The P(V) approach

with the recently disclosed Ψ-reagents allowed for the

complete control of chirality of the P-based linkages and

enabled conclusive evidence of the absolute conﬁguration. The

new approach oﬀers robust procedures for preparing the

stereodeﬁned CDN in eight steps starting from the advanced

nucleosides, with late-stage direct drop isolations and tele-

scoped steps enabling an eﬃcient scale-up that proceeded in an

overall 15% yield.

(

positive or negative ion). Compounds 13, 21, 22, 23, and

are known compounds and were prepared according to the

literature-reported procedures. Compound 26 was assigned based on

mass analysis but was not isolated and characterized. Compound 5

was obtained as crystalline material suitable for X-ray diﬀraction

following recrystallization from absolute ethanol.

-((2R,3R,4S,5R)-5-((Bis(4-methoxyphenyl)(phenyl)methoxy)-

methyl)-4-((tert-butyldimethylsilyl)oxy)-3-hydroxytetrahydrofuran-

-yl)-5-(4-nitrophenethyl)-3,5-dihydro-9H-imidazo[1,2-a]purin-9-

one (6). To a 5 L round-bottom ﬂask were added compound 11 (240

g, 288 mmol), anhydrous DMF (1300 mL), 1H-imidazole (58.8 g,

863 mmol), and TBSCl (52.1 g, 345 mmol). The reaction mixture

was stirred at room temperature, and the reaction progress was

monitored by HPLC. After completion of the reaction (14 h), ice−

water (1500 mL) was charged, and the resulting supernatant was

decanted. EtOAc (2000 mL) was charged to dissolve the precipitated

gum and gave a solution after stirring the mixture for 30 min. It was

washed with aqueous LiCl solution (1000 mL) and then water (1000

mL), dried (Na SO ), and evaporated to give the crude product as a

dark red gum (308 g). The crude material was puriﬁed by column

chromatography (silica gel, eluting with 0−20% MeOH in CH Cl ) to

EXPERIMENTAL SECTION

■

General Experimental Information. All reactions were

performed under a nitrogen atmosphere using anhydrous techniques

unless otherwise noted. Reagents were used as received from the

vendors, unless otherwise noted. Quoted yields are for isolated

material. Reactions were monitored by reversed-phase HPLC on a

Shimadzu system using CH CN/H O/MeOH as the mobile phase

give a mixture of two isomers 12 and 6 (204 g). The mixture was

further puriﬁed by supercritical ﬂuid chromatography to give

compound 6 as a yellow foam (79 g, 89 mmol, yield 31%). Data

for compound 6: H NMR (400 MHz, CDCl

) δ 8.09 (d, J = 8.7 Hz,

2H), 7.88 (s, 1H), 7.56 (d, J = 2.7 Hz, 1H), 7.39 (d, J = 8.6 Hz, 2H),

7.29 (m, 4H), 7.25 (d, J = 8.8 Hz, 2H), 7.23 (m, 2H), 7.17 (m, 1H),

6.77 (d, J = 8.3 Hz, 4H), 6.66 (d, J = 2.7 Hz, 1H), 5.97 (d, J = 4.8 Hz,

1H), 4.47 (q, J = 5.4 Hz, 1H), 4.40 (m, 1H), 4.28 (m, 2H), 4.15 (q, J

= 4.0 Hz, 1H), 3.74 (s, 3H), 3.74 (s, 3H), 3.46 (dd, J = 3.3, 10.6 Hz,

1H), 3.28 (dd, J = 4.2, 10.6 Hz, 1H), 3.18 (t, J = 7.1 Hz, 2H), 0.87 (s,

(

containing either 0.05% TFA or 0.1% NH OAc). Supercritical ﬂuid

chromatography (SFC) conditions for compound 6 were as follows:

BEH 2-EP (25 cm × 5 cm, I.D. 5 μm), 35% CAN/IPA [1:1 (v/v)]

with 0.1% NH OH in CO , 35 °C, 280 mL/min, 238 nm, 100 bar

BPR, crude sample concentration at 312.3 mg/mL, 1.4 mL injection

volume, 2.1 min injection cycle time. SFC conditions for compound

9H), 0.07 (s, 3H), −0.02 (s, 3H); C{ H} NMR (101 MHz, CDCl )

δ 158.6 (s, 2C), 151.8, 149.9, 147.1, 144.8, 144.8, 144.2, 136.54,

857

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

35.6, 135.4, 130.0 (s, 2C), 130.0 (s, 2C), 129.7 (s, 2C), 128.1 (s,

C), 127.9 (s, 2C), 127.0, 123.9 (s, 2C), 117.4, 116.8, 113.2 (s, 4C),

O-(((1S,2R,3R)-3-(6-Amino-9H-purin-9-yl)-2-((trityloxy)methyl)-

cyclobutyl)methyl) O-((2R,3R,4R,5R)-5-((Bis(4-methoxyphenyl)-

(

phenyl)methoxy)methyl)-4-((tert-butyldimethylsilyl)oxy)-2-(5-(4-

07.0, 88.3, 86.6, 83.9, 74.8, 71.9, 63.0, 55.2 (s, 2C), 46.1, 34.9, 25.7

nitrophenethyl)-9-oxo-5,9-dihydro-3H-imidazo[1,2-a]purin-3-yl)-

tetrahydrofuran-3-yl) O-Hydrogen Phosphorothioate (8). To a 1 L

round-bottom ﬂask were added 7 (31.0 g, 27.7 mmol), 5 (16.3 g, 33.2

mmol), and anhydrous 1:1 MeCN/THF (200 mL). The mixture was

stirred for 30 min at 20 °C to give a homogeneous solution, which

was then evaporated. This azeotroping procedure was repeated two

more times before the addition of anhydrous THF (310 mL). DBU

(

s, 3C), 18.0, −4.7, −4.8; HRMS (ESI-TOF) m/z [M + H] calcd for

C H N O Si 873.3638, found 873.3622. The undesired compound

2 (95 g) was heated in triethylamine and isopropanol at 50 °C

overnight to give a mixture of the two isomers, which were again

separated by SFC. The recycle procedure was repeated again to give a

total of 84 g of the desired isomer from the recycling of undesired

product and an overall 65% yield of 6 (163 g, 186 mmol).

(12.52 mL, 83 mmol) was added dropwise over the course of 10 min.

(

(1S,2R,3R)-3-(6-Amino-9H-purin-9-yl)-2-((trityloxy)methyl)-

cyclobutyl)methanol (5). A 1 L steel bomb was charged with

(1S,2R,3R)-3-(6-chloro-9H-purin-9-yl)-2-((trityloxy)methyl)-

The initial slurry completely dissolved after the addition of DBU, and

the color of the reaction mixture changed to a light yellow solution.

The reaction mixture was stirred at room temperature with progress

monitored by HPLC. After completion of the reaction (15 min), the

crude solution was dropped onto MTBE (1000 mL) over the course

of 10 min. The resulting precipitate was collected and puriﬁed by

column chromatography (silica gel, eluting with 0−10% MeOH in

CH Cl ) to give 8 as a white foam (33.6 g, 23.3 mmol, 84% yield).

(

cyclobutyl)methyl acetate (27g, 48.8 mmol) and 1,4-dioxane (135

mL), and the mixture was stirred for 20 min until all starting material

dissolved. NH OH (279 mL, 2148 mmol) was added before sealing

the bomb and heating using an oil bath at 70 °C for 2 days. After

completion of the reaction, the temperature was allowed to cool to 20

°C before concentrating to dryness. The crude material was dissolved

Data for 8: H NMR (500 MHz, CDCl ) δ 8.18 (s, 1H), 8.16 (s, 1H),

in EtOH (100 mL) and evaporated; this process was repeated two

more times. The resulting crude material was dissolved in minimal

EtOH before adding CH Cl (500 mL) and the precipitate collected

.06 (d, J = 8.7 Hz, 2H), 7.88 (s, 1H), 7.46 (d, J = 2.7 Hz, 1H), 7.39

m, 2H), 7.28−7.22 (m, ov, 12H), 7.19−7.05 (m, 12H), 6.70 (br d, J

8.9 Hz, 2H), 6.69 (br d, J = 8.9 Hz, 2H), 6.50 (d, J = 2.7 Hz, 1H),

.22 (d, J = 7.1 Hz, 1H), 5.69 (ddd, J = 4.8, 7.2, 12.4 Hz, 1H), 4.69

q, J = 8.7 Hz, 1H), 4.60 (dd, J = 2.1, 4.6 Hz, 1H), 4.24−4.06 (m,

H), 3.69−3.65 (m, 8H), 3.35 (m, 2H), 3.17−3.01 (m, 4H), 2.84 (m,

1H), 2.38 (td, J = 8.1, 10.7 Hz, 1H), 2.15 (m, 1H), 2.05 (m, 1H),

(

by ﬁltration, and EtOH (500 mL) was added to the cake (34 g). The

suspension was sonicated for 30 min, and the insoluble solid was

removed by ﬁltration. The ﬁltrate was concentrated to dryness to give

compound 5 as a white solid (21.7 g, 90% yield). Data for compound

: H NMR (400 MHz, DMSO-d ) δ 8.35 (s, 1H), 8.13 (s, 1H),

.88 (s, 9H), 0.19 (s, 3H), 0.11 (s, 3H); C{ H} NMR (126 MHz,

.28−7.15 (m, 18H), 4.78 (q, J = 9.0 Hz, 1H), 4.58 (t, J = 5.3 Hz,

H), 3.48 (t, J = 5.5 Hz, 2H), 3.10 (m, 2H), 2.97 (m, 1H), 2.45 (td, J

CDCl ) δ 158.6, 158.4, 158.4, 154.b (br s, 1C), 151.9, 150.5, 149.7,

47.0, 145.1, 144.6, 144.5, 143.7 (s, 3C), 140.1 (br s, 1C), 138.5,

35.8, 135.6, 130.1 (s, 2C), 130.0 (s, 2C), 129.8 (s, 2C), 128.6 (s,

C), 128.1 (s, 2C), 127.8 (s, 2C), 127.7 (s, 6C), 126.9 (s, 3C), 126.7,

23.9 (s, 2C), 119.3, 117.6, 116.8, 113.1 (br s, 2C), 113.1 (br s, 2C),

06.6, 86.5, 86.4, 85.5 (br d, J = 4.5 Hz, 1C), 85.3, 75.2 (br d, J = 5.4

7.9, 10.3 Hz, 1H), 2.21 (q, J = 10.1 Hz, 1H), 2.09 (m, 1H);

C{ H} NMR (101 MHz, DMSO-d ) δ 156.0, 152.2, 149.5, 143.7 (s,

C), 139.4, 128.1 (s, 6C), 127.7 (s, 6C), 126.8 (s, 3C), 119.0, 85.6,

4.0, 63.5, 47.7, 45.9, 33.3, 30.2; HRMS (ESI-TOF) m/z [M + H]

calcd for C H N O 492.2394, found 492.2388.

Hz, 1C), 72.6 (br d, J = 3.6 Hz, 1C), 67.0 (br d, J = 5.4 Hz, 1C), 64.0,

3.4, 55.1, 47.5, 46.1, 45.8, 34.7, 31.64 (d, J = 7.3 Hz, 1C), 30.0, 25.9

-((2R,3R,4R,5R)-5-((Bis(4-methoxyphenyl)(phenyl)methoxy)-

methyl)-4-((tert-butyldimethylsilyl)oxy)-3-(((2S,3aR,6S,7aR)-3a-

methyl-6-(prop-1-en-2-yl)-2-sulﬁdohexahydrobenzo[d][1,3,2]-

oxathiaphosphol-2-yl)oxy)tetrahydrofuran-2-yl)-5-(4-nitrophe-

nethyl)-3,5-dihydro-9H-imidazo[1,2-a]purin-9-one (7). To a 1 L

round-bottom ﬂask was added compound 6 (33.0 g, 37.8 mmol),

dissolved in anhydrous CH CN (150 mL), and the mixture was

concentrated. This was repeated twice, and the residual solid was

placed under a high vacuum for 60 min. The solid was dissolved in

(s, 3C), 18.2, −4.0, −4.9; P NMR (202 MHz, CDCl ) δ 58.20 (s,

P); HRMS (ESI-TOF) m/z [M + H] calcd for C H N O PSSi

77 81 11 12

442.5288, found 1442.5284.

O-(((1S,2R,3R)-3-(6-Amino-9H-purin-9-yl)-2-((trityloxy)methyl)-

cyclobutyl)methyl) O-((2R,3R,4R,5R)-4-((tert-Butyldimethylsilyl)-

oxy)-5-(hydroxymethyl)-2-(5-(4-nitrophenethyl)-9-oxo-5,9-dihydro-

H-imidazo[1,2-a]purin-3-yl)tetrahydrofuran-3-yl) O-Hydrogen

(

R)-Phosphorothioate (9). To a 500 mL round-bottom ﬂask were

anhydrous CH CN (329 mL), and the (R)-Ψ-reagent (33.8 g, 76.0

added 8 (31.5 g, 21.8 mmol), anhydrous CH Cl (150 mL), and

mmol) and DBU (11.4 mL, 76.0 mmol) were added to the reaction

mixture. The reaction mixture was stirred at room temperature with

progress monitored by HPLC. After completion of the reaction (20

min), acetic acid (5.41 mL, 94.0 mmol) was charged, and the mixture

was concentrated. The crude solid was puriﬁed by column

triethylsilane (10.5 mL, 65.5 mmol). Dichloroacetic acid (5.4 mL,

5.5 mmol) was charged dropwise over a course of 10 min. The

reaction mixture was stirred at room temperature with progress

monitored by HPLC. After completion of the reaction (14 h), the

crude solution was dropped onto MTBE (200 mL) over the course of

10 min. The resulting precipitate was collected and puriﬁed by

column chromatography (silica gel, eluting with 0−20% MeOH in

CH Cl ) to give 9 as a light yellow foam (17.3 g, 15.2 mmol, 70%

chromatography (silica gel, eluting with 0−50% EtOAc in CH Cl )

to give 7 as a yellow foam (38.2 g, 90% yield). Data for compound 7:

H NMR (500 MHz, CDCl ) δ 8.13 (d, J = 8.7 Hz, 2H), 7.70 (s, 1H),

.57 (d, J = 2.7 Hz, 1H), 7.36 (d, J = 8.7 Hz, 2H), 7.32−7.31 (ov,

yield). Data for 9: H NMR (500 MHz, DMSO-d ) δ 8.40 (s, 1H),

H), 7.15 (d, J = 8.9 Hz, 4H), 6.83−6.78 (m, 4H), 6.60 (d, J = 2.7

8.14 (s, 1H), 8.10 (s, 1H), 8.06 (d, J = 8.8 Hz, 2H), 7.54 (d, J = 2.7

Hz, 1H), 7.49 (d, J = 8.8 Hz, 2H), 7.32 (d, J = 2.7 Hz, 1H), 7.26−

7.17 (m, 14H), 7.17−7.12 (m, 3H), 5.97 (d, J = 5.9 Hz, 1H), 5.34 (q,

J = 5.7 Hz, 1H), 4.71 (q, J = 8.8 Hz, 1H), 4.45 (dd, J = 3.9, 4.6 Hz,

1H), 4.32 (m, 2H), 3.98 (q, J = 3.9 Hz, 1H), 3.84 (m, 1H), 3.72 (m,

1H), 3.69 (br dd, J = 3.9, 11.9 Hz, 1H), 3.60 (br dd, J = 4.0, 11.9 Hz,

1H), 3.24 (m, 2H), 3.01 (m, 2H), 2.88 (m, 1H), 2.36 (m, 1H), 2.22−

Hz, 1H), 6.13 (ddd, J = 4.8, 8.0, 16.1 Hz, 1H), 5.94 (d, J = 8.1 Hz,

H), 5.02 (s, 1H), 4.89 (s, 1H), 4.60 (m, 1H), 4.53 (d, J = 4.8 Hz,

H), 4.39 (td, J = 3.2, 12.3 Hz, 1H), 4.23 (br s, 1H), 4.20 (td, J = 7.6,

4.5 Hz, 1H), 3.93 (dd, J = 1.9, 12.7 Hz, 1H), 3.77 (s, 6H), 3.73 (m,

H), 3.28 (m, 2H), 2.53 (m, 1H), 2.20 (m, 1H), 1.97 (m, 1H),

.91(m, 1H), 1.82 (m, 1H), 1.77 (ov, 4H), 1.68 (m, 1H), 1.52 (s,

H), 0.94 (s, 9H), 0.17 (s, 3H), 0.16 (s, 3H); C{ H} NMR (126

2.15 (m, 2H), 0.83 (s, 9H), −0.05 (s, 6H); C{ H} NMR (126

MHz, CDCl ) δ 158.6, 151.6, 149.0, 147.3, 147.1, 145.2, 144.9, 144.3,

MHz, DMSO-d ) δ 154.4 (br s, 1C), 151.0, 150.3 (br s, 1C), 149.5,

39.4 (s, 2C), 139.1, 130.0 (s, 2C), 129.1 (s, 4C), 127.8 (s, 2C),

27.7 (s, 2C), 127.0, 123.9 (s, 2C), 118.4, 118.3, 113.1 (s, 4C), 111.8,

07.2, 88.8, 87.8 (d, J = 5.5 Hz, 1C), 86.7, 81.4, 76.1 (d, J = 5.4 Hz,

C), 72.8 (d, J = 4.5 Hz, 1C), 66.2, 62.8, 55.2, 46.8, 38.9, 35.1, 33.8

149.0, 146.2, 146.0 (s, 2C), 144.2, 143.6 (s, 3C), 140.1, 138.2, 130.2

(s, 2C), 128.0 (s, 6C), 127.7 (s, 6C), 126.7 (s, 3C), 123.3 (s, 2C),

119.3, 118.6, 115.8, 105.9, 86.6 (br d, J = 10.9 Hz, 1C), 85.5, 85.3,

76.2 (br s, 1C), 70.0, 67.3 (br s, 1C), 63.2, 61.5, 47.2, 45.8, 45.2, 33.8,

31.1 (br d, J = 7.3 Hz, 1C), 30.4, 25.7 (s, 3C), 17.7, −3.2 (s, 2C); P

NMR (202 MHz, DMSO-d ) δ 57.0 (br s, 1P); HRMS (ESI-TOF)

P); HRMS (ESI-TOF) m/z [M + H] calcd for C H N O PS Si

m/z [M + H] calcd for C H N O PSSi 1140.3987, found

56 63 11 10

1140.3981.

858

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

Article

O-(((1S,2R,3R)-3-(6-Amino-9H-purin-9-yl)-2-((trityloxy)methyl)-

cyclobutyl)methyl) O-((2R,3R,4R,5R)-4-((tert-Butyldimethylsilyl)-

oxy)-5-((((2R,3aR,6S,7aR)-3a-methyl-6-(prop-1-en-2-yl)-2-

sulﬁdohexahydrobenzo[d][1,3,2]oxathiaphosphol-2-yl)oxy)-

methyl)-2-(5-(4-nitrophenethyl)-9-oxo-5,9-dihydro-3H-imidazo-

[1,2-a]purin-9-one (12) and 3-((1R,3R,6S,8R,9R,12S,15R,17R,18R)-8-

(6-Amino-9H-purin-9-yl)-18-((tert-butyldimethylsilyl)oxy)-3,12-di-

hydroxy-3,12-disulﬁdo-2,4,11,13,16-pentaoxa-3,12-

diphosphatricyclo[13.2.1.0 ]octadecan-17-yl)-3,5-dihydro-9H-

imidazo[1,2-a]purin-9-one (27). To a 2 L round-bottom ﬂask were

added 24 (16.0 g, 14.0 mmol) and anhydrous 1:1 MeCN/THF (200

mL). The mixture was stirred for 30 min at 20 °C to give a

homogeneous solution, which was evaporated. This azeotroping

procedure was repeated two more times, and the solid was placed

under a high vacuum for 60 min. Anhydrous THF (1400 mL) was

charged followed by DBU (21.1 mL, 140.0 mmol) in one portion.

The reaction mixture was stirred at room temperature with progress

monitored by HPLC. After completion of the reaction (2 min) to give

[1,2-a]purin-3-yl)tetrahydrofuran-3-yl) O-Hydrogen Phosphoro-

thioate (10). To a 1 L round-bottom ﬂask were added 9 (8.7 g, 7.6

mmol) and anhydrous 1:1 MeCN/THF (160 mL). The mixture was

stirred for 30 min at 20 °C to give a homogeneous solution, which

was then evaporated. This azeotroping procedure was repeated two

more times, and the residual solid was placed under a high vacuum for

0 min. The (S)-Ψ-reagent (5.1 g, 11.4 mmol) and anhydrous THF

(

100 mL) were then charged, and the slurry was cooled to 0 °C. DBU

(

5.7 mL, 37.9 mmol) was added over the course of 1 min. The

5, the reaction mixture was evaporated to remove the THF, and the

reaction mixture was stirred at 0 °C with progress monitored by

HPLC. After completion of the reaction (5 min), acetic acid (4.3 mL,

ﬂask was charged with CH CN (1400 mL) and heated to 40 °C using

a heating mantle. The deprotection reaction progress was monitored

by HPLC, and after completion of the reaction (14 h), the mixture

was evaporated and puriﬁed by supercritical ﬂuid chromatography to

give 27 as a white solid (5.9 g, 7.1 mmol, 51% yield over two steps).

6.0 mmol) was added, and the crude solution was dropped onto cold

0 °C) MTBE (500 mL) over the course of 10 min. The resulting

precipitate was allowed to settle for 10 min before decanting the

MTBE solution. The precipitate was dissolved in CH Cl (600 mL)

(

Data for 25 after SFC puriﬁcation of an aliquot: H NMR (500 MHz,

and washed with citric acid (20% w/v aq solution, 3 × 200 mL). The

CD OD) δ 8.47 (s, 1H), 8.28 (br s, 1H), 8.17 (s, 1H), 8.07 (d, J = 8.7

organic layer was dried (MgSO ) and evaporated to give a solid,

Hz, 2H), 7.50 (br d, J = 1.7 Hz, 1H), 7.41 (d, J = 8.7 Hz, 2H), 7.16

which was slurried in MTBE (200 mL) and ﬁltered to give 10 as a

(

br d, J = 1.7 Hz, 1H), 6.17 (d, J = 8.2 Hz, 1H), 5.34 (m, 1H), 4.82−

.74 (m, 2H), 4.49−4.33 (m, 3H), 4.29−4.20 (m, 3H), 4.12−4.03

(m, 2H), 3.89 (m, 1H), 3.35 (m, 1H), 3.26 (m, 1H), 2.60 (td, J = 8.1,

white powder (10.5 g, 7.6 mmol, 100% yield). Data for 10: H (500

MHz), ¹³C{ H} (126 MHz), P (202 MHz), CDCl . The structure

of 10 in this crude sample was not fully elucidated. Many resonances

in the various 1D spectra acquired were broad. Two-dimensional

NMR allowed the assignment of some resonances in the ribose and

cyclobutyl rings, the tert-butyldimethylsilyl ester, ethyl nitrobenzene,

and limonine moiety. Please see Supporting Information for spectra.

0.4 Hz, 1H), 2.47 (m, 1H), 2.29 (q, J = 10.0 Hz, 1H), 1.04 (s, 9H),

.35 (s, 3H), 0.30 (s, 3H); C{ H} NMR (126 MHz, CD OD) δ

55.8 (br s, 1C), 153.6 (br s, 1C), 152.6, 151.5 (br s, 1C), 150.6,

48.3, 147.6, 146.3, 142.4 (br s, 1C), 140.0 (br s, 1C), 131.3 (s, 2C),

24.5 (s, 2C), 120.6, 120.1, 116.4 (br s, 1C), 107.1, 87.7 (br d, J = 9.1

HRMS (ESI-TOF): m/z [M + H] calcd for C H N O P S Si

78 11 11

Hz, 1C), 86.2 (br s, 1C), 77.3 (br s, 1C), 75.1 (br d, J = 3.6 Hz, 1C),

386.4283, found 1386.4295.

O-(((1S,2R,3R)-3-(6-Amino-9H-purin-9-yl)-2-(hydroxymethyl)-

9.5 (br d, J = 3.6 Hz, 1C), 68.2 (d, J = 3.3 Hz, 1C), 66.6 (br s, 1C),

1.0 (br s, 1C), 47.4, 36.0, 33.5 (br d, J = 8.2 Hz, 1C), 30.2 (br s, 1C),

cyclobutyl)methyl) O-((2R,3R,4R,5R)-4-((tert-butyldimethylsilyl)-

oxy)-5-((((2R,3aR,6S,7aR)-3a-methyl-6-(prop-1-en-2-yl)-2-

sulﬁdohexahydrobenzo[d][1,3,2]oxathiaphosphol-2-yl)oxy)-

methyl)-2-(5-(4-nitrophenethyl)-9-oxo-5,9-dihydro-3H-imidazo-

6.7 (s, 3C), 19.3, −3.5, −4.6; P NMR (202 MHz, CD OD) δ 58.0

(

br s, 1P), 55.8 (br s, 1P); HRMS (ESI-TOF) m/z [M + H] calcd

for C H N O P S Si 976.2215, found 976.2208. Data for 27: H

37 48 11 11

2 2

NMR (500 MHz, DMSO-d ) δ 12.51 (s, 1H), 8.37 (s, 1H), 8.24 (s,

[

1,2-a]purin-3-yl)tetrahydrofuran-3-yl) O-Hydrogen (R)-Phosphor-

othioate (24). To a 1 L round-bottom ﬂask were added 10 (20.1 g,

4.5 mmol), anhydrous CH Cl (240 mL), and triisopropylsilane

1H), 8.15 (s, 1H), 7.58 (d, J = 2.2 Hz, 1H), 7.41 (br d, J = 2.2 Hz,

2H), 5.97 (d, J = 8.6 Hz, 1H), 5.35 (m, 1H), 4.63 (br s, 1H), 4.67−

4.59 (m, 2H), 4.14 (dt, J = 4.9, 10.7 Hz, 1H), 4.08 (m, 1H), 4.04 (m,

1H), 3.89 (m, 1H), 3.85 (m, 1H), 3.76 (m, 1H), 3.71 (m, 2H), 3.02

(m, 2H), 2.49 (m, 1H), 2.24 (m, 1H), 2.12 (q, J = 9.9 Hz, 1H), 0.93

(

31.5 mL, 145.0 mmol). The mixture was stirred for 30 min at 20 °C

to give a homogeneous solution. Dichloroacetic acid (7.62 mL, 116

mmol) in CH Cl (39.8 mL) was added over 5 min, during which

time the mixture went from light yellow to orange to pale yellow. The

reaction progress was monitored by HPLC, and after completion of

the reaction (1.25 h), the mixture was evaporated. The residue was

(s, 9H), 0.20 (s, 3H), 0.19 (s, 3H); C{ H} NMR (126 MHz,

DMSO-d ) δ 155.4 (br s, 1C), 151.6 (br s, 1C), 151.3, 151.1, 149.3,

146.0, 139.7 (br s, 1C), 137.3 (br s, 1C), 118.9, 116.4, 114.1, 106.8,

85.5 (br d, J = 7.3 Hz, 1C), 83.0 (br s, 1C), 74.7 (br s, 1C), 73.8 (br s,

1C), 67.3 (br s, 1C), 66.8 (br s, 1C), 65.5 (br s, 1C), 48.9 (br s, 1C),

47.2 (br s, 1C), 33.3 (br s, 1C), 28.8 (br s, 1C), 26.0 (s, 3C), 18.1,

treated with Et O (1000 mL), and the resulting precipitate was

collected to give 24 as a white powder (14.9 g, 13.1 mmol, 90% yield).

Data for compound 24: H NMR (500 MHz, DMSO-d ) δ 8.40 (s,

H), 8.23 (s, 1H), 8.12 (s, 1H), 8.08 (m, 2H), 7.55 (d, J = 2.7 Hz,

H), 7.52 (br d, J = 8.8 Hz, 2H), 7.31 (d, J = 2.7 Hz, 1H), 6.09 (d, J =

.6 Hz, 1H), 5.72 (td, J = 6.1, 11.8 Hz, 1H), 4.72 (br s, 2H), 4.56 (m,

H), 4.55 (m, 1H), 4.42 (m, 1H), 4.40 (m, 1H), 4.31 (m, 1H), 4.27

−4.0, −5.2; P NMR (202 MHz, DMSO-d ) δ 55.6 (br s, 1P);

HRMS (ESI-TOF) m/z [M + H] calcd for C29H N O P S Si

41 10 9 2 2

827.1738, found 827.1742

3-((1R,3R,6S,8R,9R,12S,15R,17R,18R)-8-(6-Amino-9H-purin-9-yl)-

,12,18-trihydroxy-3,12-disulﬁdo-2,4,11,13,16-pentaoxa-3,12-

(m, 1H), 4.15 (dt, J = 2.5, 5.7 Hz, 1H), 3.49 (m, 2H), 3.40 (m, 1H),

diphosphatricyclo[13.2.1.0 ]octadecan-17-yl)-3,5-dihydro-9H-

imidazo[1,2-a]purin-9-one, BMT-390025. To a 1 L round-bottom

ﬂask were added 27 (14.0 g, 16.9 mmol), anhydrous MeOH (170

.33 (m, 1H), 3.32 (m, 2H), 3.30 (m, 4H), 2.62 (m, 1H), 2.18 (m,

H), 2.13 (m, 1H), 1.98 (q, J = 9.9 Hz, 1H), 1.91 (m, 1H), 1.87 (m,

H), 1.84 (m, 1H), 1.78 (m, 1H), 1.71 (m, 1H), 1.63 (s, 3H), 1.58 (s,

_H₎_,₀_.₈₉₍_s_,₉_H₎_,₀_.₁₈₍_s_,₃_H₎_,₀_.₁₇₍_s_,₃_H₎_;1 C{ H} NMR (126

mL), and NH F (3.1 g, 85.0 mmol). The mixture was heated to 60 °C

using a heating mantle, and the reaction progress was monitored by

HPLC. After completion of the reaction (14 h), the mixture was

puriﬁed by supercritical ﬂuid chromatography to give BMT-390025 as

MHz, DMSO-d ) δ 151.1, 149.7, 148.7 (m, 1C), 148.7 (br s, 1C),

46.3, 146.1, 144.9, 144.2, 141.0 (br s, 1C), 139.1, 130.2 (s, 2C),

23.4 (s, 2C), 119.4 (br s, 1C), 118.6, 116.2, 111.3, 105.9, 86.2 (br d,

a white solid (7.6 g, 10.1 mmol, 60% yield): H NMR (700 MHz,

J = 5.5 Hz, 1C), 85.8, 83.0 (br d, J = 9.1 Hz, 1C), 74.0 (br s, 1C), 71.7

br s, 1C), 65.8, 60.75, 47.5, 47.00, 45.4 (br s, 1C), 38.2, 33.8, 33.1,

D O) δ 8.21 (s, 2H), 8.15 (s, 1H), 7.52 (d, J = 2.5 Hz, 1H), 7.02 (d, J

(

2.5 Hz, 1H), 6.16 (d, J = 7.9 Hz, 1H), 5.46 (dt, J = 4.5, 8.6 Hz, 1H),

0.5 (br d, J = 6.4 Hz, 1C), 29.2 (br s, 1C), 25.8 (s, 3C), 22.6, 22.3,

4.81 (d, J = 4.3 Hz, 1H), 4.75 (ov, 1H), 4.53 (m, 1H), 4.31 (m, 1H),

4.28−4.21 (m ov, 3H), 4.04 (m, 1H), 3.86 (td, J = 4.9, 10.2 Hz, 1H),

7.9, −4.3, −5.2; P NMR (202 MHz, DMSO-d ) δ 100.8 (s, 1P),

7.94 (br s, 1P); HRMS (ESI-TOF) m/z [M + 2H] /2 calcd for

.01 (m, 1H), 2.65 (td, J = 8.5, 11.2 Hz, 1H), 2.48 (m, 1H), 2.31 (m,

C H N O P S Si/2 572.6636, found 572.6623.

65 11 11 2 3

H); C{ H} NMR (176 MHz, D O) δ 153.2 (br s, 1C), 153.1,

-((1R,3R,6S,8R,9R,12S,15R,17R,18R)-8-(6-Amino-9H-purin-9-yl)-

50.5, 148.9 (br s, 1C), 148.5, 145.3, 141.76, 140.0 (br s, 1C), 118.3,

]

8-((tert-butyldimethylsilyl)oxy)-3,12-dihydroxy-3,12-disulﬁdo-

,4,11,13,16-pentaoxa-3,12-diphosphatricyclo[13.2.1.0 ]-

octadecan-17-yl)-5-(4-nitrophenethyl)-3,5-dihydro-9H-imidazo-

116.7, 115.7, 107.0, 86.7 (br d, J = 6.4 Hz, 1C), 84.4 (d, J = 10.2 Hz,

1C), 75.8 (d, J = 7.6 Hz, 1C), 71.2 (d, J = 2.5 Hz, 1C), 68.1 (d, J = 6.4

859

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

https://pubs.acs.org/doi/10.1021/acs.joc.1c00784.

Article

Hz, 1C), 67.3 (d, J = 5.1 Hz, 1C), 64.9 (d, J = 5.1 Hz, 1C), 49.1, 46.6

Peng Li − Research and Early Development, Bristol-Myers

Squibb Company, Princeton, New Jersey 08543-4000, United

States

Shiuhang Yip − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Aaron Smith − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Janet Caceres-Cortes − Research and Early Development,

Bristol-Myers Squibb Company, Princeton, New Jersey

08543-4000, United States

(

d, J = 7.6 Hz, 1C), 31.4 (d, J = 11.4 Hz, 1C), 28.0; P NMR (202

MHz, D O) δ 55.9 (s, 1P), 52.3 (s, 1P); HRMS (ESI-TOF) m/z [M

H] calcd for C H N O P S 713.0879, found 713.0897; [α]

23 27 10 9 2 2

−

41.71 (c 0.556, H O)

ASSOCIATED CONTENT

sı Supporting Information

■

The Supporting Information is available free of charge at

H and C{ H} NMR spectra for all new compounds;

P NMR spectra for compounds 7, 8, 9, 10, 24, 25, 27,

Darpandeep Aulakh − Materials Science & Engineering,

Bristol-Myers Squibb Company, New Brunswick, New Jersey

and BMT-390025; X-ray crystallography data for

compound 5 (PDF)

08903, United States

Accession Codes

Amy A. Sarjeant − Materials Science & Engineering, Bristol-

Myers Squibb Company, New Brunswick, New Jersey 08903,

United States

Peter K. Park − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

CCDC 2074384 contains the supplementary crystallographic

data for this paper. These data can be obtained free of charge

via www.ccdc.cam.ac.uk/data_request/cif, or by emailing

data_request@ccdc.cam.ac.uk, or by contacting The Cam-

bridge Crystallographic Data Centre, 12 Union Road,

Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

Lalgudi S. Harikrishnan − Research and Early Development,

Bristol-Myers Squibb Company, Princeton, New Jersey

8543-4000, United States

AUTHOR INFORMATION

Corresponding Author

James Kempson − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States; orcid.org/0000-0002-9540-3886;

Email: james.kempson@bms.com

■

Lan-ying Qin − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Dharmpal S. Dodd − Research and Early Development,

Bristol-Myers Squibb Company, Princeton, New Jersey

08543-4000, United States

Brian Fink − Research and Early Development, Bristol-Myers

Squibb Company, Princeton, New Jersey 08543-4000, United

States

Gregory Vite − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Arvind Mathur − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Authors

Huiping Zhang − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Xiaoping Hou − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States; orcid.org/0000-0001-6169-3443

Lyndon Cornelius − Research and Early Development,

Bristol-Myers Squibb Company, Princeton, New Jersey

8543-4000, United States; orcid.org/0000-0001-6736-

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.joc.1c00784

351

Rulin Zhao − Research and Early Development, Bristol-Myers

Squibb Company, Princeton, New Jersey 08543-4000, United

States

Notes

The authors declare no competing ﬁnancial interest.

Bei Wang − Research and Early Development, Bristol-Myers

Squibb Company, Princeton, New Jersey 08543-4000, United

States

Zhenqiu Hong − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

ACKNOWLEDGMENTS

■

We thank our Chemical Development colleagues, Michael A.

Schmidt, Richard Fox, and Amanda Rogers, for insightful

discussions during the course of this work. We thank Linping

Wang, Ying Li, Jenny Cutrone, and Yingru Zhang for their help

with the analytical experiments including HRMS and optical

rotation measurements. We would also like to thank our

colleagues in the Department of Discovery Synthesis at the

Biocon-BMS R&D Center (Bengaluru, India) for the synthesis

of intermediate 13.

Martins S. Oderinde − Research and Early Development,

Bristol-Myers Squibb Company, Princeton, New Jersey

8543-4000, United States; orcid.org/0000-0002-6858-

21X

Joseph Pawluczyk − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Dauh-Rurng Wu − Research and Early Development, Bristol-

Myers Squibb Company, Princeton, New Jersey 08543-4000,

United States

Dawn Sun − Research and Early Development, Bristol-Myers

Squibb Company, Princeton, New Jersey 08543-4000, United

States

REFERENCES

■

1) Ishikawa, H.; Barber, G. N. STING is an endoplasmic reticulum

adaptor that facilitates innate immune signalling. Nature 2008, 455,

2) (a) Gomelsky, M. cAMP, c-di-GMP, c-di-AMP and now cGMP:

74−678.

bacteria use them all! Mol. Microbiol. 2011, 79, 562. (b) Ablasser, A.;

Goldeck, M.; Cavlar, T.; Deimling, T.; Witte, G.; Röhl, I.; Hopfner,

860

J. Org. Chem. 2021, 86, 8851−8861

The Journal of Organic Chemistry

(16) (a) Kalíková, K.; Slechtová, T.; Vozka, J.; Tesarová, E.

Article

K.-P.; Ludwig, J.; Hornung, V. cGAS produces a 2 ′-5 ′-linked cyclic

dinucleotide second messenger that activates STING. Nature 2013,

Supercritical fluid chromatography as a tool for enantioselective

separation; A review. Anal. Chim. Acta 2014 , 821 , 1 −33. (b) West, C.

Current trends in supercritical fluid chromatography. Anal. Bioanal.

Chem. 2018, 410, 6441−6457. (c) Sun, D.; Wu, D.-R.; Li, P.; Yip, H.;

Wang, B.; Hou, X.; Zhao, R.; Zhang, H.; Kempson, J.; Mathur, A.,

Large-Scale Supercritical Fluid Chromatography Purification of

Unstable STING Agonist Intermediates, Journal of Chromatography

A, in press, 2021

498, 380−384. (c) Diner, E. J.; Burdette, D. L.; Wilson, S. C.;

Monroe, K. M.; Kellenberger, C. A.; Hyodo, M.; Hayakawa, Y.;

Hammond, M. C.; Vance, R. E. The Innate Immune DNA Sensor

cGAS Produces a Noncanonical Cyclic Dinucleotide that Activates

Human STING. Cell Rep. 2013, 3, 1355−1361. (d) Gao, P.; Ascano,

M.; Wu, Y.; Barchet, W.; Gaffney, B. L.; Zillinger, T.; Serganov, A. A.;

Liu, Y.; Jones, R. A.; Hartmann, G.; Tuschl, T.; Patel, D. J. Cyclic

(

17) Darrington, R. Accelerating laboratory chiral separation

workflows. Spec. Chem. Mag. 2009 , 29 , 30 −31.

18) (a) Ahmad, S. An efficient and diastereoselective [2+2]

[

G(2 ′,5 ′)pA(3 ′,5 ′)p] Is the Metazoan Second Messenger Produced

by DNA-Activated Cyclic GMP-AMP Synthase. Cell 2013 , 153 ,

094 −1107. (e) Gomelsky, M. Cyclic dimeric GMP-mediated

cycloaddition: convenient and enantioselective route to trans-2 ′,3 ′-

decisions in surface-grown Vibrio parahaemolyticus: a different kind

dihydroxymethylcyclobutane nucleoside analogs. Tetrahedron Lett.

of motile-to-sessile transition. J. Bacteriol. 2012, 194, 911−913.

991, 32, 6997−7000. (b) Singh, J.; Bisacchi, G. S.; Ahmad, S.;

(f) Zhang, X.; Shi, H.; Wu, J.; Zhang, X.; Sun, L.; Chen, C.; Chen, Z.

Godfrey, J. D.; Kissick, T. P.; Mitt, T.; Kocy, O.; Vu, T.; Papaioannou,

C. G.; Wong, M. K.; Heikes, J. E.; Zahler, R.; Mueller, R. H. A

Practical Asymmetric Synthesis of the Antiviral Agent Lobucavir,

BMS-180194. Org. Process Res. Dev. 1998, 2, 393−399.

J. Cyclic GMP-AMP Containing Mixed Phosphodiester Linkages Is

An Endogenous High-Affinity Ligand for STING. Mol. Cell 2013, 51,

3) Wu, J.-J.; Zhao, L.; Hu, H.-G.; Li, W.-H.; Li, Y.-M. Agonists and

26−235.

19) Islam, S.; Ahmad, S.; Liu, T.; Försterling, F. H.; Hossain, M. M.

inhibitors of the STING pathway: Potential agents for immunother-

First Unambiguous Characterization of a Novel RO-BH2 Inter-

apy. Med. Res. Rev. 2020, 40, 1117−1141.

13 11

mediate by H{ B} and H{ C, B} NMR Spectroscopy. Proof of

Boron −Carbonyl Interaction through Innovative Use of NMR

Techniques. Organometallics 2008, 27, 2354 −2355.

(4) Oost, T.; Kuttruff, C. A.; Nar, H. Sting agonists for immuno-

oncology. Med. Chem. Rev. 2019, 54, 197 −219.

5) Lian, Y.; Duffy, K. J.; Yang, J. STING Activation and its

Application in Immuno-Oncology. Curr. Top. Med. Chem. 2019, 19,

205−2227.

6) Fink, B. E.; Dodd, D. S.; Zhao, Y.; Qin, L.-Y.; Ruan, Z.;

(20) Salezadeh-Asl, R.; Lee-Ruff, E. Enantiomeric resolution of

cyclobutanones and related derivatives by enzyme-catalyzed acylation

(

and hydrolysis. Tetrahedron: Asymmetry 2005, 16, 3986−3991.

(

21) Cieslak, J.; Ausín, C.; Chmielewski, M. K.; Kauffman, J. S.;

Harikrishnan, L. S.; Kamau, M. G.; Walker, S. J.; Chen, L.; Park, P. K.

Cyclic dinucleotides as anticancer agents. U.S. Patent

WO2019160884A1, February 13, 2019.

Snyder, J.; Del-Grosso, A.; Beaucage, S. L. 31P NMR Study of the

Desulfurization of Oligonucleoside Phosphorothioates Effected by

“

Aged ” Trichloroacetic Acid Solutions. J. Org. Chem. 2005, 70, 3303−

306.

22) Zewge, D.; Gosselin, F.; Sidler, R.; DiMichele, L.; Cvetovich, R.

7) Mieczkowski, A.; Agrofoglio, L. A. Synthesis and Biological

Activity of Selected Carbocyclic Nucleosides. Modified Nucleosides

008, 393−424.

8) (a) Gaffney, B. L.; Veliath, E.; Zhao, J.; Jones, R. A. One-Flask

J. A Safe and Practical Procedure for Global Deprotection of

Oligoribonucleotides. J. Org. Chem. 2010, 75, 5305−5307.

Syntheses of c-di-GMP and the [Rp,Rp] and [Rp,Sp] Thiophosphate

Analogues. Org. Lett. 2010 , 12 , 3269 −3271. (b) Beaucage, S. L.;

Caruthers, M. H. Deoxynucleoside phosphoramidites A new class of

key intermediates for deoxypolynucleotide synthesis. Tetrahedron Lett.

(

981, 22, 1859−1862.

9) Knouse, K. W.; deGruyter, J. N.; Schmidt, M. A.; Zheng, B.;

Vantourout, J. C.; Kingston, C.; Mercer, S. E.; McDonald, I. M.;

Olson, R. E.; Zhu, Y.; Hang, C.; Zhu, J.; Yuan, C.; Wang, Q.; Park, P.;

Eastgate, M. D.; Baran, P. S. Unlocking P(V): Reagents for chiral

phosphorothioate synthesis. Science 2018, 361, 1234−1238.

(10) Xu, D.; Rivas-Bascón, N.; Padial, N. M.; Knouse, K. W.; Zheng,

B.; Vantourout, J. C.; Schmidt, M. A.; Eastgate, M. D.; Baran, P. S.

Enantiodivergent Formation of C −P Bonds: Synthesis of P-Chiral

Phosphines and Methylphosphonate Oligonucleotides. J. Am. Chem.

Soc. 2020, 142, 5785−5792.

(11) (a) Boryski, J. N2-Ethenoguanosine: Three Methods of

Synthesis. Nucleosides Nucleotides 1990 , 9 , 803 −813. (b) Boryski, J.;

Golankiewicz, B.; De Clercq, E. Synthesis and antiviral activity of 3-

substituted derivatives of 3,9-dihydro-9-oxo-5H-imidazo[1,2-a]-

purines, tricyclic analogs of acyclovir and ganciclovir. J. Med. Chem.

991, 34, 2380−2383.

12) Somoza, A. Protecting groups for RNA synthesis: an increasing

need for selective preparative methods. Chem. Soc. Rev. 2008, 37,

668−2675.

13) Hakimelahi, G. H.; Proba, Z. A.; Ogilvie, K. K. New catalysts

and procedures for the dimethoxytritylation and selective silylation of

ribonucleosides. Can. J. Chem. 1982, 60, 1106−1113.

(14) Blaisdell, T. P.; Lee, S.; Kasaplar, P.; Sun, X.; Tan, K. L.

Practical Silyl Protection of Ribonucleosides. Org. Lett. 2013, 15,

710−4713.

15) Matulic-Adamic, J.; Beigelman, L.; Portmann, S.; Egli, M.;

(

Usman, N. Synthesis and Structure of 1-Deoxy-1-phenyl-β -d-

ribofuranose and Its Incorporation into Oligonucleotides. J. Org.

Chem. 1996, 61, 3909−3911.

861