The Discovery of Two Novel Classes of 5,5-Bicyclic Nucleoside-Derived PRMT5 Inhibitors for the Treatment of Cancer

pubs.acs.org/jmc

Article

The Discovery of Two Novel Classes of 5,5-Bicyclic Nucleoside-

Derived PRMT5 Inhibitors for the Treatment of Cancer

Ryan V. Quiroz,* Michael H. Reutershan, Sebastian E. Schneider, David Sloman, Brian M. Lacey,

Brooke M. Swalm, Charles S. Yeung, Craig Gibeau, Daniel S. Spellman, Danica A. Rankic, Dapeng Chen,

David Witter, Doug Linn, Erik Munsell, Guo Feng, Haiyan Xu, Jonathan M. E. Hughes, Jongwon Lim,

Josep Saur í , Kristin Geddes, Murray Wan, My Sam Mansueto, Nicole E. Follmer, Patrick S. Fier,

Phieng Siliphaivanh, Pierre Daublain, Rachel L. Palte, Robert P. Hayes, Sandra Lee, Shuhei Kawamura,

Steven Silverman, Sulagna Sanyal, Timothy J. Henderson, Yingchun Ye, Yuanwei Gao,

Benjamin Nicholson, and Michelle R. Machacek

Cite This: J. Med. Chem. 2021, 64, 3911 −3939

Read Online

sı

*

Metrics & More

Article Recommendations

Supporting Information

ACCESS

ABSTRACT: Protein arginine methyltransferase 5 (PRMT5) is a type II arginine methyltransferase that catalyzes the post-

translational symmetric dimethylation of protein substrates. PRMT5 plays a critical role in regulating biological processes including

transcription, cell cycle progression, RNA splicing, and DNA repair. As such, dysregulation of PRMT5 activity is implicated in the

development and progression of multiple cancers and is a target of growing clinical interest. Described herein are the structure-based

drug designs, robust synthetic eﬀorts, and lead optimization strategies toward the identiﬁcation of two novel 5,5-fused bicyclic

nucleoside-derived classes of potent and eﬃcacious PRMT5 inhibitors. Utilization of compound docking and strain energy

calculations inspired novel designs, and the development of ﬂexible synthetic approaches enabled access to complex chemotypes

with ﬁve contiguous stereocenters. Additional eﬀorts in balancing bioavailability, solubility, potency, and CYP3A4 inhibition led to

the identiﬁcation of diverse lead compounds with favorable proﬁles, promising in vivo activity, and low human dose projections.

DNA repair, cell cycle progression, transcriptional regulation,

and RNA splicing.²⁻⁷In turn, PRMT5 has been implicated as a

potential oncogene whose upregulation results in tumor cell

proliferation and invasiveness for numerous diﬀerent can-

cers,²⁻⁷including colorectal, lung, ovarian, prostate, and

pancreatic cancers, and lymphoma, leukemia, and glioblasto-

ma.^7,8This multipathway regulation culminating in broad-

ranging tumorigenesis suggests that PRMT5 inhibition could

INTRODUCTION

■

Even with the tremendous advances made in oncology

therapies over the past several years, most notably in

immunotherapies such as checkpoint inhibitors,¹cancer still

remains a daunting global health challenge. While eﬀorts to

harness and augment immune responses to tumors remain

promising and worthwhile, new targets and approaches are still

crucially needed to continue expanding the patient populations

successfully managed with anticancer drugs. One epigenetic

target of emerging and promising clinical potential is protein

arginase methyltransferase 5 (PRMT5), a type II arginine

methyltransferase that post-translationally modiﬁes histone and

other protein targets via symmetric dimethylation. This

methylation of protein substrates catalyzed by PRMT5 plays

a critical role in regulating key cellular processes including

Received: December 1, 2020

Published: March 23, 2021

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

3911

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 1. Binding of SAM (green, PDB 4X61) to PRMT5 overlaid with the H4 peptide (salmon, PDB 4GQB). Main interactions highlighted with

yellow dashed lines.

Figure 2. (A) Binding mode of SAM (1, PDB 4X61). Prominent interactions with the protein are shown with yellow dashed lines. (B) Binding

mode of the H4 peptide arginine residue in the PRMT5 catalytic site (pink, PDB 5FA5) overlaid with the SAM binding mode (green, PDB 4X61).

The arginine is in close proximity to the methyl group of SAM, anchored by Glu444.

serve as a powerful anticancer strategy. As such, numerous

preclinical inhibitor discovery eﬀorts have been pursued,⁸⁻¹⁹

and several pharmaceutical companies have recently initiated

clinical trials with small-molecule PRMT5 inhibitors.^4,6

Active PRMT5 forms a hetero-octameric complex with

adapter protein MEP50 and utilizes S-adenosylmethionine

(SAM, 1, Figure 1) as a cofactor to transfer a methyl group to

the substrate arginine.^20,21The active site of the enzyme

consists of a conserved SAM binding pocket and a nearby

protein substrate binding pocket, which are connected by a

short, narrow channel in the enzyme.²⁰In addition to SAM,

the endogenous ligand methylthioadenosine (MTA) also binds

to and inhibits PRMT5 in a SAM-competitive manner. For

some cancers, gene deletion of the enzyme responsible for

metabolizing MTA, methylthioadenosine phosphorylase

(MTAP), results in a natural partial inhibition of PRMT5 by

accumulated MTA.²²⁻²⁵

Given these catalytic site structural features of PRMT5,

three of the most commonly pursued classes of inhibitors

include SAM-competitive, substrate-competitive, and dual

SAM/substrate-competitive inhibitors (Figure 1). Such

structurally diverse PRMT5 inhibitors with these mechanisms

of action (MoAs) pursued preclinically and clinically have been

extensively reviewed.^4,6Interestingly, another unique class of

substrate-competitive inhibitors represented by EPZ015666 is

SAM-uncompetitive inhibitors, which still require SAM to

achieve PRMT5 inhibition.^9,16Allosteric inhibition,²⁶covalent

inhibition,²⁷and PROTAC-mediated inhibition²⁸have also

been reported but are beyond the scope of this paper. Of

interest to us, dual SAM/substrate-competitive inhibitors

occupy the SAM pocket and extend into the narrow groove

that connects the substrate protein pocket, precluding both

SAM and substrate binding, a mechanism that could be utilized

for inhibitor design to perhaps enhance potency and prolong

compound residence time.⁴

Given this backdrop and the signiﬁcant potential of PRMT5

inhibition for the treatment of cancer, we initiated a robust

medicinal chemistry eﬀort to develop potent, selective, orally

bioavailable, and low projected human dose nucleoside-derived

PRMT5 inhibitors. Speciﬁcally, we designed, synthesized, and

biologically evaluated multiple classes of novel 5,5-bicyclic

nucleoside inhibitors, including a fused ribose-carbocycle core

exempliﬁed by the lead compound 34 and an all-carbon 5,5-

fused core exempliﬁed by 72. These complex bicyclic

compounds were exquisitely potent, eﬃcacious, and structur-

ally unique dual-competitive inhibitors compared to other

known PRMT5 inhibitors. Furthermore, the development of

modular and reliable synthetic routes enabled rapid exploration

of SAR, which was critical to modulating solubility and

CYP3A4 inhibition proﬁles alongside aﬃnity and pharmaco-

kinetic (PK) proﬁles. These eﬀorts culminated in the discovery

of two new classes of PRMT5 inhibitors with potential utility

as cancer therapeutics due to their balanced overall properties,

3912

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 3. (A) Inspired by 3, designing a lower-strain 5,5-bicyclic system by fusing 3′ and 4′ led to the more potent inhibitor 4. (B) Retention of the

key nucleoside binding pharmacophore for 4 was validated by X-ray crystallography (PDB 7KIB). H-bonds shown in dashed yellow lines.

demonstration of in vivo tumor regression, and low predicted

once daily (QD) human doses.

within both the SAM and peptide binding sites, we

hypothesized that optimizing the conformational preference

of the structure used to link the nucleobase, diol, and

aminoquinoline moieties would be critical to achieving high-

aﬃnity binders.³⁰Accordingly, we utilized docking and

conformational analysis to prioritize the designs of geometri-

cally privileged cores. In particular, the calculation of strain

energy incurred between ground state and bound state analog

conformations was conducted.³¹In combination with binding

free energy estimations, these strain energy calculations of

bioactive conformations helped rank new designs and predict/

explain potency trends (see the Supporting Information for

details).

The initial key ﬁnding utilizing these computational tools

stemmed from representative compound 3 (Figure 3A). While

3 demonstrated that methyl substitution was tolerated at the 3′

position of a simple nucleoside, yielding an active inhibitor in

our cellular PRMT5 target engagement (TE) assay, the

introduction of this methyl group did induce strain in the

conformation adopted by the linker connecting the ribose core

and an aminoquinoline moiety. With this insight and the

subsequent goal to lower the overall strain energy to better

favor a desired bioactive conformation, we designed a 5,5-fused

bicyclic ribose inhibitor wherein 3′ and 4′ were connected via a

cyclopentane ring (Figure 3A, 4). Gratifyingly, not only did the

calculations predict that such a cyclization would reduce the

linker strain, but also compound 4 proved to be more potent in

vitro (Figure 3A and Figure S1). This data helped solidify our

conﬁdence that modeling and strain energy calculations could

guide designs to improve target aﬃnity.

Additional insight into our novel 5,5-fused PRMT5 inhibitor

design came from the X-ray crystal structure of compound 4

(Figure 3B). The nucleoside diol and nucleobase retained the

critical hydrogen bonding interactions that mimicked the

adenosine moiety in SAM. Additionally, the aminoquinoline

moiety could indeed extend into the channel connecting the

SAM and substrate pockets, making a face-to-face pi stacking

interaction with Phe327 and a bidentate hydrogen bond with

Glu444, conﬁrming 4’s function as a dual-competitive

inhibitor. With these key binding interactions, the 5,5-fused

ring system could be used to design compounds with exquisite

potency in a favorable low-strain bioactive conformation.

Capitalizing more generally on these computational tools,

we leveraged docking and strain energy calculations for

RESULTS AND DISCUSSION

■

Modeling, Energy Calculations, and Initial Com-

pound Design. Our approach to design PRMT5 inhibitors

took inspiration from the natural cofactor SAM (1), as well as

the known PRMT5 nucleoside-based inhibitors referenced

above.^4,6,8−19Furthermore, computational modeling was

leveraged to investigate the intricate interactions between

these ligands and their respective binding pockets to help

enable analog design.

Speciﬁcally, the crystal structure of SAM bound to

PRMT5¹⁶revealed several critical interactions in the active

site between the SAM cofactor and the receptor. Notably, the

adenosine core diol engaged in a hydrogen bonding network

with Tyr324 and Glu392, the purine C6-amino group

hydrogen-bonded with Asp419, and the purine N1 nitrogen

hydrogen-bonded with the main chain amide −NH of Met420

(Figure 2A).

While the SAM pharmacophore highlighted the main

nucleoside binding interactions essential to analog design, we

also took an interest in dual-competitive compounds that

engaged key residues in the channel connecting the cofactor

and substrate binding pockets, speciﬁcally the catalytic residue

Glu444, which anchors the protein substrate arginine and is

part of the active site of PRMT5 (Figure 2B). Along these

lines, we took inspiration from previously developed SAM-

mimetic compounds like JNJ64619178,^4,29which contains an

aminoquinoline moiety that modeling suggested could indeed

extend from the SAM binding pocket to the peptide binding

site channel. The aminoquinoline was speciﬁcally modeled to

engage in a bidentate hydrogen bond with Glu444 similarly to

the substrate arginine residue, thereby putatively preventing

both SAM and substrate binding.

With this knowledge gleaned from SAM and previous

PRMT5i chemical matter, we proposed that nucleoside-

derived small molecules represented the most attractive

approach toward utilizing the SAM pocket as a key anchor

for the design of potent, dual-competitive inhibitors. As such,

our design campaign sought to develop novel nucleoside cores

and Glu444-interacting substituents to generate compounds

that would functionally and optimally serve this goal. Given the

speciﬁc geometry required to eﬀectively bridge key interactions

3913

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 4. Analysis of predicted protein−ligand interaction energy (MMGBSA) and predicted global strain identiﬁed diﬀerent 5,5-fused cores as

having a high POS for delivering high-aﬃnity inhibitors, which was validated by TE potencies. Unﬁlled shapes were designed but not made.

Scheme 1. Representative Synthetic Route to Access Ether-Linked Bicyclic Ribose Inhibitors

subsequent 5,5-bicyclic designs toward the goal of identifying

compounds with predicted high aﬃnity. Speciﬁcally, the

docked binding aﬃnity was estimated using a docking and

reﬁnement protocol including minimization steps and

MMGBSA calculations, and the minimized pose was used as

an input for the strain calculations, which was conducted using

freeform (see the Supporting Information). Ultimately, these

strain energy calculations and aﬃnity predictions were used in

concert to rank order several 5,5-fused designs and compare

them against 3′-methyl-substituted, 3′/4′-dimethyl-substituted,

and hypothetical 4,5-, 6,5-, and 5,6-fused cores (Figure 4).

Gratifyingly, this analysis broadly supported the 5,5-fused

designs as having a high probability of success (POS) for

3914

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Scheme 2. Representative Synthetic Route to Access Methylene-Linked Bicyclic Ribose Inhibitors

delivering high-aﬃnity inhibitors, which was validated by the

TE potencies for the designs that were ultimately synthesized.

With these calculations establishing conﬁdence in our 5,5-

fused inhibitor designs, one modiﬁcation that could be readily

assessed was the change in the 5′ linker identity and

stereochemistry. Four possible combinations were analyzed,

namely, an ether linker vs a methylene linker and S- or R-

stereochemistry at that linker position. Utilizing this analysis,

we were able to prioritize the S-ether (“down”) linker and the

S-methylene (“up”) linker for the bicyclic ribose systems

(Figure S1). These combinations were modeled to adopt the

most preferred bioactive conformation to allow the nucleoside

core to interact with the SAM pocket and the aminoquinoline

to engage with the catalytic residue Glu444. In all, due to the

relatively lengthy syntheses of these complex inhibitors with

ﬁve contiguous stereocenters, these modeling-informed in-

sights were valuable tools for eﬃciently driving novel design

progression.

Synthetic Chemistry for 5,5-Bicyclic Ribose Com-

pounds. To enable the expansion of SAR and the

optimization of compound properties to develop projected

low-dose PRMT5 inhibitors, new synthetic routes needed to

be developed to access the complex core of 4. Few synthetic

protocols have been reported for this class of bicyclic ribose

scaﬀolds with a similar substitution pattern.³²⁻³⁴Leveraging

this synthetic insight, our approach relied on a 14-step route

beginning with the commercially available glucofuranose

derivative 5, as highlighted with the representative synthesis

of ether-linked scaﬀolds shown below (Scheme 1).

Oxidation of the 3′ alcohol of the protected glucofuranose 5

with IBX aﬀorded ketone 6, to which vinyl Grignard was added

to aﬀord tertiary allylic alcohol 7. After protection of the 3′

alcohol, periodic acid was used for selective acetonide cleavage

to aﬀord 4′ aldehyde 9. Following literature precedent,³²an

Rh-mediated intramolecular cyclization between the 3′ oleﬁn

and 4′ aldehyde furnished the bicycle core 10. After ketone

reduction and acetonide deprotection, 3′ benzyloxy hydro-

genation aﬀorded triol 13. Subsequent 2′/3′ acetonide

protection and tosylation of the 5′ secondary alcohol aﬀorded

tosylate 15, which was primed for an S_N2 with phenol 16 to

aﬀord ether 17 with the desired stereochemistry suggested by

modeling. Acid-mediated hydrolysis yielded triol 18, which

underwent an epoxide-opening addition³⁵of nucleobase 19 to

aﬀord nucleoside 20. When needed, ﬁnal amination of the

purine chloride aﬀorded the generic inhibitor 21. One

beneﬁcial feature of this route, besides its general ability to

access complex chiral structures, was its late-stage modularity.

By simply changing the left-hand-side quinoline moiety or the

nucleobase added to the core, we could furnish numerous

diverse analogs to modulate potency, physical properties, oﬀ-

target proﬁles, and PK.

In addition to these ether-linked compounds, we were also

interested in accessing methylene-linked 5,5-bicyclic ribose

analogs to expand the chemical space tested and tune linker

strain. Importantly, the opposite stereochemistry at 5′ was

desired for the methylene linker compared to the ether linker

based on the strain energy and docking discussed above

(Figure S1). A slightly altered synthetic route was adopted to

access such methylene-linked scaﬀolds based on the common

intermediate 14 (Scheme 2).

Brieﬂy, oxidation of the 5′ alcohol of 14 aﬀorded ketone 22,

which was oleﬁnated under Wittig conditions to give exocyclic

oleﬁn 23. A hydroboration with 9-BBN gratifyingly provided

the desired facial selectivity to aﬀord a terminal alkyl borane,

which was immediately subjected to Suzuki conditions with

quinoline halide 24 (V = −I or −Br) to furnish methylene-

linked bicycle 25. Acid-mediated deprotection aﬀorded triol

26, which again underwent epoxide-mediated 7-deazapurine

addition to aﬀord nucleoside 27. When necessary, a ﬁnal

amination aﬀorded the generic inhibitor 28. The relative

stereochemical assignment of both ether- and methylene-

linked representative analogs was conﬁrmed by 2D NMR

(Figure S2).

5,5-Bicyclic Ribose Inhibitor In Vitro Proﬁling. To

rigorously proﬁle our PRMT5 inhibitors in vitro, we developed

three tiers of assays to assess compound potency and

proliferation inhibition. First, analogs were proﬁled in a simple

biochemical (BC) assay containing a PRMT5/MEP50

complex in solution. Second, the cell-based TE assay measured

levels of histone methylation. Finally, a B-cell Z138 lymphoma

assay measured the proliferation inhibition eﬃcacy of the

inhibitors (details in Experimental Section). Since most

3915

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 1. 5,5-Bicyclic Ribose SAR

compounds in this series demonstrated tight binding kinetics

and were of higher aﬃnity than could be reliably measured in

our standard biochemical assay format, cell-based TE and

proliferation EC₅₀values were the main drivers of SAR (see

Figure S3, which also contains representative SPR data to help

support the bisubstrate MoA). Utilizing the modular synthetic

routes discussed above, a series of diverse analogs were

synthesized and screened for PRMT5 potency (Table 1;

structural elements for comparison are colored).

As can be seen in Table 1 (4 and 29−33), both ether-linked

and methylene-linked 5,5-dual inhibitors had single-digit

nanomolar EC₅₀values in the cellular TE and proliferation

inhibition assays. However, compounds 40 and 41 supported

the importance of stereochemistry (and therefore linker strain)

for compound potency, as was computationally predicted

(Figure S1). There was a signiﬁcant loss of potency (>2000×

and >8× for ether- and methylene-linked compounds,

respectively, in the TE assay) when more strained linker

conformations were adopted. Furthermore, subtle potency

eﬀects could be seen when the aminoquinoline halogen was

exchanged. For 4 and 29−33, −Br-substituted analogs were

consistently the most potent in the proliferation assay, with an

emerging trend of −Br > −Cl > −F. The −Br and −Cl

compounds were equipotent in the TE assay, with both being

more potent than −F. While modeling indicated that the larger

size of the halogen atom could help ﬁll a pocket to increase

potency, other likely contributing factors were the impact of

the halogen atom on the basicity of the quinoline nitrogen

(Table S1), aﬀecting the strong bidentate hydrogen bond, and

its impact on the electronic properties of the quinoline

aromatic ring, which engaged in a face-to-face pi stacking

interaction with Phe327. Ultimately, changes to either the

linker atom identity or the halogen atom were still well

tolerated, and analogs with <10 nM TE potency were possible

in all combinations.

In addition to the C6-amino group on the nucleobase,

compounds 34−39 demonstrated that a simple C6-methyl

group was also tolerated for these inhibitors. While the amino-

containing compounds in general demonstrated higher aﬃnity,

this data suggested that the purported hydrogen bond between

3916

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

this nucleobase C6-amino group and PRMT5 was not critical

for potency and that, perhaps, interacting with the N1 atom

was more important in that region. X-ray crystallography of the

representative compound 34 also conﬁrmed that all other

major interactions were maintained with this class of inhibitors

(Figure 5). Finally, with the C6-methyl nucleobase, a similar

(compound 42), a 27× loss in TE potency was observed

relative to 4, potentially due to the lack of ﬁlling of a crucial

pocket or the change in the basicity of the adjacent N1

pyrimidine nitrogen. However, clear size limitations were also

revealed. A larger cyclopropyl group (43) was ∼18× less

potent, while an intermediate C6-aminomethyl group (46 and

47) was tolerated, albeit with an ∼4−8× loss in potency

relative to the amine matched pairs.

Conversely, on the other side of the core, removing the

aminoquinoline halogen (44) resulted in the most potent

inhibitor within this series. Speciﬁcally, 44 demonstrated

signiﬁcant eﬀects in the TE and proliferation assays at the

lowest concentrations tested. This data bolstered the

hypothesis that enhancing the quinoline amine basicity

(Table S1) and adjusting the overall aromatic electronics

were more inﬂuential than ﬁlling the space in this region.

Accordingly, replacing the halogen with a methyl group (45)

was also advantageous, exhibiting subnanomolar potency.

Finally, other vectors on both the quinoline and nucleobase

rings could be leveraged to generate potent inhibitors with

unique properties. Compound 49 demonstrated that while a

−F substitution on the C7-position of the purine was well

tolerated, a methyl substitution on the C2-position of this ring

(48) resulted in ∼2.5× reduced TE potency relative to 31,

albeit with a proliferation EC₅₀still <10 nM. Compounds 50

and 51 highlighted a ﬂuorine walk around the quinoline, with

compound 50 showing the more preferred position for the

added substituent. Overall, these synthesis- and design-enabled

compounds aﬀorded exquisitely potent PRMT5 inhibitors with

a range of tolerated modiﬁcations.

Figure 5. C6-methyl nucleobase inhibitor 34 retains key binding

interactions (PDB 7KIC). H-bonds shown in dashed yellow lines.

potency trend for the quinoline halogen atoms was also seen,

with −Br > −Cl > −F; in fact, this trend was generally more

distinct for these methyl nucleobase compounds relative to

their amino matched pairs for both the TE and Z138 assays.

While amino and methyl groups were well tolerated at C6 of

the nucleobase, the SAR in this region of the inhibitor revealed

some limitations. With no substituent on the nucleobase

5,5-Bicyclic Ribose Inhibitor PK and Solubility

Proﬁling. As we were developing these promising PRMT5

Table 2. PK and Solubility Proﬁles for C6-Amino Nucleobase Analogs

3917

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 3. PK and Solubility Proﬁles for C6-Methyl Nucleobase Analogs

a

Not determined due to modest potency preventing compound progression.

inhibitors, the next tier of compound progression en route to a

projected low-dose QD candidate was in vivo PK proﬁling. For

these analogs, various linker, halogen atom, and substituent

combinations could be leveraged to modulate physical

chemical properties and PK. However, what became apparent

early on was the fact that many of the C6-amino-substituted

nucleobase compounds suﬀered from low bioavailability (%F)

in rats (Table 2).

In fact, for every matched pair (compounds 34−39 vs 4 and

29−33), the compounds demonstrated a substantial boost in

%F with the methyl swap. The largest increase was seen from

matched pairs 29 and 35, which saw an increase from 2 %F to

71 %F (note that the higher LogD of 29 vs 35 is likely due to

unique ionization eﬀects, while the LogP still follows the trend,

with values of 1.92 vs 2.41, respectively). Unbound clearance

values and in vitro hepatocyte clearance values (Table S3) were

not drastically diﬀerent across the matched pairs (and, in fact,

were generally higher for 34−39), supporting the hypothesis

that improvements in oral bioavailability were partly driven by

LogD and F_a, as opposed to fraction escaping liver metabolism

and excretion (F_h). Finally, compounds 33, 36, and 39

demonstrated that −F-substituted quinolines achieved the best

solubility proﬁle across all tested media (>100 μM in pH 2 and

pH 7 buﬀers and FaSSIF), regardless of C6-amino or methyl

substitution. This cumulative SAR provided key insights that

helped us reﬁne the balance between %F, Cl, MRT, and

solubility across a discrete LogD range and substitution

pattern.

To further test these insights and SAR/PK/solubility sweet

spots, compounds with more unique substitutions and physical

chemical properties were additionally proﬁled (Table 4).

Compounds that replaced the quinoline halogen with a

hydrogen or alkyl group showed exquisite potency (44 and

45), but extra-hepatic clearance emerged as a severe liability

for compound progression. To try further balancing the

lipophilicity necessary to improve %F, nucleobase analogs with

methylated amino groups were synthesized (46 and 47).

Compound 46 showed a dramatic improvement in %F relative

to its amino matched pair 4 (68% vs 10%), but a very high

unbound clearance dampened the signiﬁcance of that increase.

Compounds 4 and 29−33 demonstrated this ﬁnding that

regardless of linker or quinoline halogen identity, most 5,5-

bicyclic ribose inhibitors with a C6-amino nucleobase showed

reasonable rat clearance and MRT values but prohibitively low

%F (≤10%), with the exception of compound 33 (which did

not translate to dogs; data not shown). Additionally, while

permeability was high for all compounds tested (Table S2), the

less ﬂexible aryl ether-linked analogs (4 and 29−30) had

negligible pH 7 kinetic solubility, which improved with the

methylene-linked matched pairs (31−33, Table 2).^36,37

Nonetheless, %F for 31−33 still remained low (Table 2).

Therefore, given that the C6-amino-substituted nucleobase

was the one common factor among analogs with low %F, we

wanted to test C6-Me-substituted analogs in vivo. We

hypothesized that an increase in the lipophilicity (LogD) of

the compounds could boost bioavailability through modulation

of fraction absorbed (F_a) in spite of the potential that

increasing lipophilicity could contribute to poor solubility.^38,39

Gratifyingly, this simple C6-exchange actually revealed a

physical property sweet spot, with a dramatic improvement

in %F as well as a maintenance, and at times, an improvement,

in kinetic solubility accompanying the general increase in LogD

(Table 3).

3918

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 4. PK and Solubility Proﬁles for 5,5-Bicyclic Nucleoside Analogs

a

b

Not determined due to high plasma clearance. Not determined due to assay failure.

Inhibitor 47 had a more modest improvement in %F over

compound 31 (20% vs 10%). Compound 48 with a C2-methyl

group on the deazapurine, balanced with a C6-amine,

unfortunately had a mere 11 %F, demonstrating that the C2-

methyl did not have a similar eﬀect as the C6-methyl.

Compound 49 was designed with the hypothesis that an

intramolecular hydrogen bond between the C7-F and C6-

amine groups on the nucleobase could potentially mask PSA

and one HBD to increase permeability without lowering the

solubility. However, 49 showed the equivalent level of

bioavailability compared to matched pair 33 (24% vs 28%,

respectively), and it suﬀered from high clearance as well.

Finally, additional −F-substituted quinoline inhibitors 50 and

51 highlighted another way we achieved increased LogD values

relative to 33. They both showed promising PK proﬁles with

good MRT values (>1.5 h) and high solubility across the pH

range, but %F remained modest at ∼30% (which got lower for

50 in dogs; data not shown).

With the complexity of these compounds, understanding

trends in potency, PK, and solubility were critical in

prioritizing the best analogs to design, make, and test. The

SAR gathered on these 5,5-bicyclic ribose dual-competitive

PRMT5 inhibitors revealed three crucial discoveries that drove

compound progression. First, methylene-linked analogs were

generally more potent than the ether-linked matched pairs, and

bromoquinolines were the most potent of the halogen series.

Second, bioavailability emerged as a potential limiting factor

for many of the compounds, but C6-methyl-substituted

nucleobase inhibitors were far superior to the corresponding

C6-amino-substituted nucleobases and other substitution

patterns in achieving high %F. LogD and F_a(potentially also

modulated by solubility) were possible contributing factors to

oral bioavailability. Additionally, for 4 and 29−39, the ether-

linked compounds demonstrated lower clearance and higher

MRT values. Third, ﬂuoroquinoline analogs consistently had

the best kinetic solubility across the three tested media for the

halogen matched pairs, underscoring a delicate and, at times,

mutually exclusive balance between potency, %F, clearance,

MRT, and solubility for this 5,5-ribose series (Figure 6,

includes compounds not exempliﬁed in Table 1).

Given the above SAR, compound 34 emerged as a promising

and balanced compound whose potency and PK proﬁle could

potentially lead to a low-dose prediction and good POS for

conventional formulation (due in part to the development of a

more soluble maleate salt form). Therefore, to assess the

overall selectivity proﬁle for these 5,5-bicyclic ribose PRMT5

inhibitors, analog 34 was thoroughly explored in a panel of 4

DNA methyltransferases and 36 methyltransferases, including

PRMT 1, 3, 4, 6, 7, and 8. Signiﬁcantly, 34 exhibited >4000-

fold selectivity for PRMT5 compared to every other oﬀ-target

tested, indicating that this class of bicyclic, dual-competitive

inhibitors was highly selective for the desired target (Table

S4).

3919

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

alter physical properties. Speciﬁcally, we designed a class of all-

carbon 5,5-bicyclic nucleoside inhibitors. One potential

advantage of these compounds came from the possibility of

combining a more lipophilic core with more polar moieties like

the ﬂuoroquinoline to achieve a more well-rounded proﬁle.

Overall, these types of bicyclic nucleoside inhibitors are

unprecedented in the literature, and we proﬁled these unique

and structurally complex inhibitors to compare them to the

5,5-bicyclic ribose compounds toward progression of the best

chemical matter. Similar to the bicyclic ribose compounds,

these all-carbon compounds with a methylene linker were

modeled with low strain energy and favorable binding energy.

In contrast, the ether linker did not appear to be similarly well

tolerated (Figure S1), and so we focused on the methylene-

linked variants.

Synthetic Chemistry for All-Carbon 5,5-Bicyclic

Inhibitors. Given the novelty of these substituted all-carbon

5,5-bicyclic nucleoside-derived PRMT5 inhibitors, a new

synthetic route had to be developed that would allow for the

same modularity that was so advantageous for the bicyclic

ribose series. Fortunately, there is a wealth of diverse strategies

to access the general [3.3.0]octane skeleton to draw inspiration

from, including, but not limited to, ring contraction,⁴¹⁻⁴⁴ring

expansion,⁴⁵intramolecular Wittig cyclizations,^46,47intra-

molecular aldol cyclizations,^48,49Pinacol/radical cou-

plings,⁵⁰⁻⁵²photorearrangement/addition,⁵³C-H insertion

reactions,⁵⁴[3 + 2] cycloadditions,^55,56and the Pauson−

Khand carbonylative [2 + 2 + 1] cycloaddition.⁵⁷⁻⁶³Due to

the ability to rapidly access an all-carbon 5,5-fused core with

the necessary functionality to achieve the desired substitution

pattern, we chose to pursue a Pauson−Khand approach for our

synthesis. Therefore, a novel route was developed that enabled

access to this unique class of PRMT5 inhibitors (Scheme 3).

First, the simple alkynol 52 was oxidized to aﬀord aldehyde

53. Vinyl Grignard addition followed by triphenylsilyl

protection of the resulting alcohol aﬀorded enyne 54.

Compound 54 was primed for an intramolecular Co-catalyzed

Pauson−Khand reaction, which aﬀorded bicyclo[3.3.0]-

octenone 56. This reaction provided a mixture of cis and

trans isomers as racemates, which could be readily separated by

standard/chiral chromatography with absolute conﬁguration

conﬁrmed by ECD (Figure S4). The trans isomer drawn in

Scheme 3 (56) was carried forward to mimic the preferred

stereochemistry of the 5,5-ribose system, which was also

supported by modeling. Then, a subsequent stereospeciﬁc

Luche reduction aﬀorded enol 57, which was acetate-protected

to aﬀord the protected enol 58. Due to the lack of an anomeric

center that is present in the ribose compounds, traditional

nucleobase installation chemistry could not be easily applied to

this all-carbon 5,5-bicyclic core. Therefore, the allyl acetate

moiety was purposefully installed to enable Tsuji−Trost AAA

Figure 6. Analysis of the balance between rat %F, LogD, rat MRT,

and Z138 potency for 5,5-bicyclic ribose compounds.

Additionally, we leveraged human dose predictions based on

rat and dog allometry and Z138 potency to comprehensively

assess progress regarding compound optimization and to triage

compound progression. Gratifyingly, 34 indeed demonstrated

a balanced proﬁle and PK consistencies across species that led

to a low QD human dose prediction (Table 5).

Finally, due to the potential safety and eﬃcacy risks

associated with oncology therapy DDI,⁴⁰we monitored the

reversible inhibition (RI) and time-dependent inhibition

(TDI) of CYP3A4 for these analogs. While the C6-amino

compounds exhibited little to no RI or TDI risk, C6-methyl

inhibitors showed an increased risk, with 34 having a TDI ratio

> 2.4 (Table S5). While it appeared that one way to mitigate

TDI was to lower the LogD of the analogs via appropriate

halogen substitution (38 and 39, Table S5), dialing out the

DDI risk appeared diﬃcult without sacriﬁcing other favorable

compound properties. Taken together, the 5,5-bicyclic ribose

inhibitors were very promising lead scaﬀolds, and while the

low-dose prediction (and low predicted C_max) of 34 conferred

a low to moderate DDI risk, there was still a need for

continued optimization and diﬀerentiation.

All-Carbon 5,5-Bicyclic Inhibitors. While the novel 5,5-

bicyclic ribose series achieved numerous favorable qualities

including exquisite potency, tractable PK, and good solubility,

we recognized that there was still room for improvement in

balancing aspects like MRT, solubility, and CYP3A4 RI and

TDI. Much of the SAR thus far seemed to indicate that

achieving high aﬃnity, good PK (high %F and long MRT),

adequate solubility (≥50 μM), and low risk for CYP3A4-

mediated DDI at odds with one another. Therefore, to expand

the chemical space sampled by these 5,5-bicyclic PRMT5

inhibitors, we considered core changes to more signiﬁcantly

Table 5. Human Dose Prediction for the Lead Bicyclic Ribose PRMT5 Inhibitor 34

a

Target coverage = Z138 EC₅₀.

3920

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Scheme 3. Representative Synthetic Route to Access All-Carbon 5,5-Bicyclic Inhibitors

Table 6. All-Carbon 5,5-Bicyclic SAR

chemistry to add the nucleobase, a powerful strategy that has

been employed on monocyclic cyclopentene compounds in the

literature.⁶⁴Using a simple catalyst system with dppf as the

ligand, the AAA chemistry was successfully employed to

generate compound 59. A stereospeciﬁc dihydroxylation then

aﬀorded diol 60 to complete the nucleoside core. An acid-

mediated acetonide protection, which concomitantly depro-

tected the silyl group, aﬀorded alcohol 61. Oxidation and

Wittig homologation then aﬀorded exocyclic oleﬁn 62. Similar

to the bicyclic ribose compounds, hydroboration of the

exocyclic oleﬁn followed by an immediate Suzuki coupling

with 63 installed the methylene-linked quinoline. 2D NMR

analysis conﬁrmed the overall relative stereochemistry of the

isolated inhibitor (Figure S5). When necessary, an amination

was conducted to convert the pyrimidine chloride to the amino

group, and for every compound, a ﬁnal TFA deprotection of

the 2′/3′ acetonide aﬀorded the generic analog 64. This

synthetic protocol represents a novel and modular strategy to

generate these unique, structurally complex PRMT5 inhibitors.

All-Carbon 5,5-Bicyclic Inhibitor In Vitro Proﬁling.

Informed by the SAR of the 5,5-bicyclic ribose series, several

select analogs were synthesized for the all-carbon core, which

were also proﬁled in the cellular TE and Z138 assays (Table

6).

As mentioned, strain energy calculations and the initial SAR

allowed prioritization of methylene-linked analogs, and

excitingly, these compounds proved to be equipotent and, at

times, even more potent (e.g.,39 vs 68) than their 5,5-ribose

matched pairs. In fact, every compound listed above

demonstrated robust inhibition of proliferation in our Z138

cellular assay (EC₅₀< 10 nM), underscoring the exquisite

aﬃnity of these novel inhibitors. Accordingly, for the C6-

3921

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 7. Key binding interactions were retained with all-carbon inhibitor 72 (PDB 7KID). H-bonds shown in dashed yellow lines.

Table 7. All-Carbon 5,5-Bicyclic Analogs PK and Solubility Proﬁles

methyl-substituted nucleobase analogs (66−68), there ap-

peared to be a less pronounced diﬀerence in potency for the

−Br, −Cl, and −F quinoline derivatives compared to the

ribose series, an advantage for identifying compounds with a

range of PK and physical properties that would still yield low

human dose predictions. Similarly, C7-substituted (69 and 70)

and C2-substituted (71) deazapurine derivatives were very well

tolerated in this all-carbon series, along with −F-substituted

quinoline 72, which was designed with the hopes of translating

the PK SAR seen with bicyclic ribose inhibitor 50. Finally, the

ligand pharmacophore and bioactive conformation of these

inhibitors were conﬁrmed by X-ray crystallography, closely

matching those of the bicyclic ribose series, as seen in the

structure of 72 in the PRMT5 binding pocket (Figure 7).

All-Carbon 5,5-Bicyclic Inhibitor PK and Solubility

Proﬁling. While the potency of the all-carbon bicyclic

inhibitors was in line with, and at times superior to, the 5,5-

bicyclic ribose compounds, it was critical to also assess their

PK and solubility proﬁles to determine if they still had a viable

path forward. Table 7 highlights the relevant data for these

compounds.

Comparing inhibitors 66−68 with the other analogs in

Table 7 revealed the same SAR trend seen with the bicyclic

ribose compounds, where the C6-methyl nucleobase was truly

3922

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 8. Human Dose Prediction for the Lead All-Carbon Bicyclic PRMT5 Inhibitor 72

a

Target coverage = Z138 EC₅₀.

Figure 8. In vivo antitumor eﬃcacy of PRMT5 inhibitor 34. (A) Tumor growth inhibition curves for compound 34 at four diﬀerent doses (PO)

relative to vehicle control revealed signiﬁcant TGI and even tumor regression. (B) Tumor SDMA-SMB/B′ levels were used as in vivo markers of

target engagement, with 34 demonstrating a robust reduction in PRMT5 inhibition over the duration of the study.

critical for achieving high levels of bioavailability. Modifying

C2/C7 on the nucleobase (69−71) or the quinoline (72) with

additional substituents could only provide a modest increase in

%F, up to 24% for compound 72 compared to the 77% seen

with 68. Additionally, LogD and unbound clearance tracked

with the identity of the halogen on the quinoline similarly to

the 5,5-ribose analogs, with −F < −Cl < −Br (66−68).

Gratifyingly, solubility also remained high with this class of all-

carbon inhibitors, again matching the trend of −F > −Cl >

−Br (particularly at pH 7). Clearance and MRT values

remained reasonable for the majority of compounds as well. In

fact, although the %F remained modest, compounds like 71

and 72 demonstrated the potential of this series with MRT

values ≥2.5 h (not seen with the bicyclic ribose compounds),

and in general, 7 of 10 compounds in Table 7 had MRT values

≥1.5 h. Finally, comparing the SAR of compounds 72 and 50

to their matched pairs (69 and 33, respectively) indeed

suggested that while that speciﬁc −F substitution on the

quinoline ring did not help boost %F, it may have blocked a

site of metabolism, leading to lower clearance and higher MRT

values. Taken together, the single-digit nanomolar potency, the

ability to achieve high %F and solubility, and promising PK

proﬁles established this all-carbon bicyclic core as a uniquely

favorable scaﬀold. Additionally complementing the promising

features of this core, representative potent compound 68 also

demonstrated a good oﬀ-target selectivity proﬁle in a broad

panel analysis (Figure S6).

Given the low nanomolar potency, as well as its attractive PK

(low clearance, >2 h MRT) consistent across species, inhibitor

72 achieved a favorable, balanced compound proﬁle with a low

QD human dose prediction (Table 8), low-dose number, and

minimal DDI risk.

5,5-Bicyclic Inhibitor In Vivo Eﬃcacy. Ultimately,

compounds 34 and 72 were identiﬁed as novel 5,5-fused

nucleoside-derived PRMT5 inhibitors with attractive, balanced

proﬁles, low human dose predictions, and diﬀerentiated DDI

risk. We elected to pursue the representative inhibitor 34 in a

mouse Z138 in vivo tumor eﬃcacy model to provide

translational validation for these compounds (see Experimental

Section for details). In this model, dose-dependent tumor

growth inhibition (TGI) was monitored (Figure 8A), as were

tumor levels of symmetric dimethylarginine (SDMA) markers

on small-molecule ribonucleoprotein-associated protein B and

B′ (SMB/B′), which served as a biomarker for PRMT5 target

engagement (Figure 8B).

Gratifyingly, compound 34 demonstrated robust, potent

antitumor eﬃcacy in a dose-dependent manner, even causing

signiﬁcant tumor regression at the highest dose (Figure 8A, 60

mg QD, blue curve). To validate that the TGI and regression

were due to PRMT5 inhibition, levels of SDMA-SMB/B′ were

measured in the tumors, and at all doses during the course of

the 14 day experiment (days 7 and 14), a dramatic reduction of

the PRMT5 dimethylarginine mark was observed relative to

the vehicle control (Figure 8B). In addition, compound 34 was

well tolerated in the study, with minimal to no body weight

change relative to the control at all doses (Figure S7). Taken

together, this data suggested that these 5,5-bicyclic PRMT5

inhibitors exhibit signiﬁcant, target-selective antitumor eﬃcacy

Further bolstering this encouraging data, all compounds in

the series also carried a reduced risk of CYP3A4-mediated DDI

given their reduced propensity for reversible or time-depend-

ent inhibition (Table S6). In particular, compound 72 emerged

as a standout, diﬀerentiated compound (TDI ratio = 1.0).

3923

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

of (3aR,5S,6S,6aR)-5-((R)-2,2-dimethyl-1,3-dioxolan-4-yl)-2,2-

dimethyltetrahydrofuro[2,3-d][1,3]dioxol-6-ol (5, 500 g, 1.9 mol,

1.0 equiv) in MeCN (3 L) at 25 °C was slowly added IBX (807 g, 2.9

mol, 1.5 equiv) at 20−25 °C. The reaction mixture was then stirred at

85−90 °C for 3 h. The mixture was ﬁltered and concentrated under

reduced pressure. The product (3aR,5R,6aS)-5-((R)-2,2-dimethyl-1,3-

dioxolan-4-yl)-2,2-dimethyldihydrofuro[2,3-d][1,3]dioxol-6(5H)-one

and tumor regression, representing a highly promising

approach to the potential clinical treatment of cancer.

CONCLUSIONS

■

Through its regulation of numerous pathways involved in

tumorigenesis, including DNA repair, RNA splicing, and cell

cycle progression, PRMT5 is an exciting new target in

continued eﬀorts to develop more eﬃcacious and more

broad-reaching anticancer therapies. Our strategy to develop

the most optimized, projected low-dose chemical matter for

the inhibition of PRMT5 relied on computational docking,

strain energy calculations, X-ray crystallography, robust

synthetic strategies, and balancing SAR trends, ultimately

leading to the design and evaluation of unique 5,5-bicyclic,

nucleoside-based inhibitors. Speciﬁcally, we pursued a 5,5-

bicyclic ribose core and an all-carbon 5,5-fused core, both of

which were able to demonstrate exquisite potency, good PK,

high solubility, low risk for DDI, and low predicted human

doses. Lead compounds 34 and 72 were identiﬁed with overall

balanced proﬁles and diﬀerentiated DDI risk. When taken in

vivo, the representative compound 34 demonstrated robust

TGI and even tumor regression, underscoring the highly

promising potential of bringing these unique series of 5,5-fused

PRMT5 inhibitors to the clinic. Further, in line with this paper,

our work on other 5,5-bicyclic inhibitors, in vitro and in vivo

models used to further proﬁle the eﬃcacy of these compounds,

and the complete safety proﬁle of these and other PRMT5

inhibitors we developed will be reported in other publications

in due course.

1

(6) was used crude in the next step without further puriﬁcation. H

NMR (400 MHz, CDCl₃) δ 6.14 (d, J = 4.40 Hz, 1H), 4.45−4.31 (m,

3H), 4.06−4.00 (m, 2H), 1.46 (s, 3H), 1.43 (s, 3H), 1.34 (s, 6H).

Step 2: Synthesis of (3aR,5R,6R,6aR)-5-((R)-2,2-Dimethyl-1,3-

dioxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro[2,3-d][1,3]-

dioxol-6-ol (7). To a solution of (3aR,5R,6aS)-5-((R)-2,2-dimethyl-

1,3-dioxolan-4-yl)-2,2-dimethyldihydrofuro[2,3-d][1,3]dioxol-6(5H)-

one (6, 500 g, 1.9 mol, 1.0 equiv) in dry THF (2.4 L) cooled down to

0−5 °C was added vinylmagnesium bromide (1 M, 3.87 L, 3.9 mol,

2.0 equiv) dropwise, maintaining the temperature at 0−5 °C. The

reaction was then warmed to 15−20 °C and stirred for 30 min. Eight

reactions run in parallel were combined. The reaction mixture was

quenched by pouring into aqueous NH₄Cl (10 L) at 0−5 °C. The

aqueous phase was extracted with EtOAc (3 L × 3). The combined

organic phase was washed with brine (2 L), dried with anhydrous

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The

residue was puriﬁed by silica gel column chromatography (1:0 to 5:1

petroleum ether/EtOAc) to aﬀord (3aR,5R,6aR)-5-((R)-2,2-dimeth-

yl-1,3-dioxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro[2,3-d][1,3]-

1

dioxol-6-ol (7, 2.94 kg, 55%) as a white solid. H NMR (400 MHz,

DMSO-d₆) δ 5.81 (d, J = 3.6 Hz, 1H), 5.77−5.67 (m, 1H), 5.40−5.36

(m, 1H), 5.29−5.26 (m, 1H), 5.24−5.19 (m, 1H), 4.20−4.16 (m,

1H), 4.08−4.06 (m, 1H), 4.02−3.96 (m, 1H), 3.79−3.69 (m, 2H),

1.49 (s, 3H), 1.36−1.20 (m, 9H).

Step 3: Synthesis of (3aR,5R,6R,6aR)-6-(Benzyloxy)-5-((R)-2,2-

dimethyl-1,3-dioxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro-

[2,3-d][1,3]dioxole (8). To a solution of NaH (140 g, 3.5 mol, 60%

dispersion in mineral oil, 2.5 equiv) in DMF (2 L) at 15−20 °C was

dropwise added a solution of (3aR,5R,6aR)-5-((R)-2,2-dimethyl-1,3-

dioxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro[2,3-d][1,3]dioxol-

6-ol (7, 400 g, 1.4 mol, 1.0 equiv) in DMF (0.4 L) at 15−20 °C. The

reaction mixture was then stirred at 55−60 °C for 1 h before BnBr

(574 g, 3.4 mol, 398 mL, 2.4 equiv) was added. The reaction mixture

was stirred at 15−20 °C for another 5 h. The reaction was quenched

by pouring the mixture into ice water (3 L). The resultant mixture

was extracted with EtOAc (1 L × 3). The combined organic phase

was washed with aqueous LiCl (1.5 L), dried with anhydrous Na₂SO₄,

ﬁltered, and concentrated under reduced pressure. The resulting

product (3aR,5R,6R,6aR)-6-(benzyloxy)-5-((R)-2,2-dimethyl-1,3-di-

oxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro[2,3-d][1,3]dioxole

EXPERIMENTAL SECTION

■

Chemistry General Materials and Methods. All reagents and

solvents were purchased from commercial sources and used as is

without further puriﬁcation. Reaction progress and synthetic

intermediate analysis was assessed by LCMS (UV detection with

ESI mass detection) when applicable using an Agilent instrument with

a MeCN/water gradient with either TFA or NH₄HCO₃modiﬁer. All

reported yields are isolated yields unless otherwise stated. Silica gel

chromatography and reverse-phase ﬂash column chromatography

were conducted with Teledyne ISCO CombiFlash instruments and

commercially available prepacked columns. Reverse-phase preparative

HPLC puriﬁcation of ﬁnal analogs was performed on an Agilent

preparative HPLC instrument with UV and MS detection using a

MeCN/water gradient with either TFA or NH₄OH modiﬁer. All

reported compounds tested in the assays were ≥95% pure as

1

(8) was used crude in the next step without further puriﬁcation. H

NMR (300 MHz, DMSO-d₆) δ 7.40−7.24 (m, 5H), 5.89 (d, J = 3.3

Hz, 1H), 5.86−5.76 (m, 1H), 5.49−5.36 (m, 2H), 4.79 (d, J = 3.6 Hz,

1H), 4.55−4.46 (m, 2H), 4.14−4.02 (m, 2H), 3.90−3.85 (m, 1H),

3.74−3.69 (m, 1H), 1.50 (s, 3H), 1.30 (s, 3H), 1.27 (s, 3H), 1.24 (s,

3H).

1

determined by LCMS or HPLC UV analysis. The H NMR spectra

were collected at room temperature. Chemical shifts are reported in

ppm relative to the listed deuterated solvent, and multiplicities,

coupling constants (where applicable), and signal integrations are

listed parenthetically. Final compounds were isolated as either TFA

salts or the free base after puriﬁcation. High-resolution mass

spectrometry (HRMS) was performed on a Water Xevo G2 QTof

instrument, with ESI+ ionization and 120 °C source temperature. The

syntheses of 3 and novel nucleobases and aminoquinolines are

described in the patent submitted for this work (WO2020/33288).

Safety statement: given the highly potent nature of the ﬁnal

compounds whose preparations are detailed in this section and

their demonstrated ability to inhibit PRMT5 in vitro and in vivo,

proper procedures for the handling of these highly potent compounds

should be followed at all times in accordance with institutional

policies.

Step 4: Synthesis of (3aR,5S,6R,6aR)-6-(Benzyloxy)-2,2-dimethyl-

6-vinyltetrahydrofuro[2,3-d][1,3]dioxole-5-carbaldehyde (9). To a

solution of (3aR,5R,6R,6aR)-6-(benzyloxy)-5-((R)-2,2-dimethyl-1,3-

dioxolan-4-yl)-2,2-dimethyl-6-vinyltetrahydrofuro[2,3-d][1,3]dioxole

(8, 2 kg, 5.3 mol, 1.0 equiv) in EtOAc (10 L) at 15−20 °C was added

periodic acid (1.25 kg, 5.5 mol, 1.03 equiv). The resultant mixture was

stirred for 1 h. The reaction was ﬁltered, and the solvent was removed

under reduced pressure. The residue was puriﬁed by silica gel column

chromatography (20:1 to 0:1 petroleum ether/EtOAc) to aﬀord

(3aR,5S,6R,6aR)-6-(benzyloxy)-2,2-dimethyl-6-vinyltetrahydrofuro-

[2,3-d][1,3]dioxole-5-carbaldehyde (9, 1.68 kg, 52%) as a yellow

1

solid. H NMR (400 MHz, CDCl₃) δ 9.58 (s, 1H), 7.26−7.43 (m,

5,5-Bicyclic Ribose Inhibitor Procedures. Representative

Synthetic Procedure for (2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromo-

quinolin-7-yl)oxy)-2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (4). Step 1: Synthesis

of (3aR,5R,6aS)-5-((R)-2,2-Dimethyl-1,3-dioxolan-4-yl)-2,2-

dimethyldihydrofuro[2,3-d][1,3]dioxol-6(5H)-one (6). To a solution

5H), 5.97 (d, J = 3.20 Hz, 1H), 5.77 (dd, J = 11.20, 11.20 Hz, 1H),

5.38−5.54 (m, 2H), 4.59−4.73 (m, 4H), 1.62 (s, 3H), 1.40 (s, 3H).

Step 5: Synthesis of (3aR,4aS,7aS,7bR)-7a-(Benzyloxy)-2,2-

dimethylhexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-

one (10). A mixture of [Rh(nbd)₂]BF₄(9.8 g, 26.3 mmol, 0.08 equiv)

3924

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

and 1,2-bis(diphenylphosphino)benzene (14.6 g, 32.7 mmol, 0.1

equiv) in DCE (300 mL) at 15−20 °C under N₂was degassed three

times with H₂using a vacuum/purge cycle. Then, H₂was bubbled

through the solution for 15 min. The reaction mixture was ﬂushed

again with N₂for 15 min to remove H₂. Then, a solution of

(3aR,5S,6R,6aR)-6-(benzyloxy)-2,2-dimethyl-6-vinyltetrahydrofuro-

[2,3-d][1,3]dioxole-5-carbaldehyde (9, 100 g, 329 mmol, 1.0 equiv)

in DCE (400 mL) was added dropwise to the above catalyst solution

at 15−20 °C under N₂. The mixture was stirred at 75−80 °C for 12 h.

The mixture was ﬁltered, and the solvent was removed under reduced

pressure to aﬀord (3aR,4aS,7aS,7bR)-7a-(benzyloxy)-2,2-dimethyl-

hexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-one (10),

which was used crude in the next step. ¹H NMR (300 MHz,

CDCl₃) δ 7.40−7.31 (m, 5H), 5.96 (d, J = 3.6 Hz, 1H), 4.77 (d, J =

10.8 Hz, 1H), 4.67−4.61 (m, 2H), 4.19 (s, 1H), 2.58−2.43 (m, 3H),

1.82−1.68 (m, 1H), 1.66 (s, 3H), 1.42 (s, 3H).

Step 6: Synthesis of (3aR,4aR,5R,7aR,7bR)-7a-(Benzyloxy)-2,2-

dimethylhexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-ol

(11). To a mixture of (3aR,4aS,7aS,7bR)-7a-(benzyloxy)-2,2-dime-

thylhexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-one (10,

400 g, 1.3 mol, 1.0 equiv) in MeOH (2.4 L) at 0−5 °C was added

NaBH₄(99.5 g, 2.6 mol, 2.0 equiv). The mixture was stirred at 0−5

°C for 1 h. Four reactions run in parallel were combined. The reaction

was poured into ice water (3 L), and the aqueous phase was extracted

with EtOAc (1 L × 3). The combined organic phase was washed with

brine (500 mL), dried with anhydrous Na₂SO₄, ﬁltered, and

concentrated under reduced pressure. The residue was puriﬁed by

silica gel column chromatography (20:1 to 0:1 petroleum ether/

EtOAc) to aﬀord (3aR,4aR,5R,7aR,7bR)-7a-(benzyloxy)-2,2-dime-

thylhexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-ol (11,

400 g, 25%) as a yellow oil. ¹H NMR (400 MHz, DMSO-d₆) δ

7.16−7.14 (m, 4H), 7.10−7.05 (m, 1H), 5.66 (d, J = 3.6 Hz, 1H),

4.52 (d, J = 6.4 Hz, 1H), 4.43 (d, J = 3.6 Hz, 1H), 4.30−4.27 (m,

2H), 3.92−3.91 (m, 1H), 3.82−3.76 (m, 1H), 1.75−1.62 (m, 2H),

1.54−1.33 (m, 2H), 1.29 (s, 3H), 1.13 (s, 3H).

Step 9: Synthesis of (3aR,5aR,6R,8aR)-4-Methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-ol (14).

To a solution of (3R,3aS,6R,6aR)-2-methoxyhexahydro-3aH-

cyclopenta[b]furan-3,3a,6-triol (13, 30 g, 158 mmol, 1.0 equiv) in

acetone (210 mL) were added TsOH·H₂O (1.5 g, 7.9 mmol, 0.05

equiv) and 2,2-dimethoxypropane (41.1 g, 48.3 mL, 394 mmol, 2.5

equiv). The reaction was stirred at RT for 1 h. Two reactions run in

parallel were combined. The reaction mixture was poured into water

(1.5 L) and neutralized with aqueous Na₂CO₃(300 mL). The

aqueous phase was extracted with DCM (400 mL × 4). The

combined organic phase was washed with brine (300 mL), dried over

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The

residue was puriﬁed by silica gel column chromatography (20:1 to 0:1

petroleum ether/EtOAc) to aﬀord (3aR,5aR,6R,8aR)-4-methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-ol (14,

53.6 g, 74%) as a mixture of anomers as a yellow oil. ¹H NMR

(400 MHz, CDCl₃) δ 5.08−4.98 (m, 1H), 4.43−4.29 (m, 1H), 4.28−

4.21 (m, 1H), 4.09−4.03 (m, 1H), 3.51−3.47 (m, 3H), 2.53−2.07

(m, 1H), 2.08−1.86 (m, 4H), 1.57−1.38 (m, 6 H).

Step 10: Synthesis of (3aR,5aR,6R,8aR)-4-Methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl 4-

methylbenzenesulfonate (15). To a solution of (3aR,5aR,6R,8aR)-

4-methoxy-2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]-

dioxol-6-ol (14, 320 g, 1.4 mol, 1.0 equiv) in DCM (1.2 L) was added

DMAP (424 g, 3.5 mol, 2.5 equiv) at RT. Then, TsCl (265 g, 1.4 mol,

1.0 equiv) was added. The reaction mixture was stirred overnight at

RT. The resulting mixture was treated with saturated aqueous NH₄Cl

(1.8 L) and extracted with DCM (1 L × 3). The combined organic

phase was washed with brine (1 L), dried over anhydrous Na₂SO₄,

ﬁltered, and concentrated under reduced pressure. The resulting

residue was puriﬁed by silica gel column chromatography (10−20%

EtOAc/petroleum ether) to aﬀord (3aR,5aR,6R,8aR)-4-methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl 4-

methylbenzenesulfonate (15, 508 g, 95% yield) as a mixture of

anomers. ESI MS m/z = 402 [M + NH₄]⁺.

Step 11: Synthesis of 3-Bromo-7-(((3aR,5aR,6S,8aR)-4-methoxy-

2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl)-

oxy)quinolin-2-amine (17, X = Br). To a solution of (3aR,5a-

R,6R,8aR)-4-methoxy-2,2-dimethylhexahydrocyclopenta[2,3]furo-

[3,4-d][1,3]dioxol-6-yl 4-methylbenzenesulfonate (15, 508 g, 1.3 mol,

1.0 equiv) and 2-amino-3-bromoquinolin-7-ol (16 (X = Br), 331 g,

1.4 mol, 1.05 equiv) in NMP (15.2 L) was added Cs₂CO₃(1.29 kg,

4.0 mol, 3.0 equiv). The reaction mixture was stirred at 110 °C for ∼3

days under an argon atmosphere. The reaction mixture was diluted

with water (6 L) and extracted with EtOAc (6 L × 3). The combined

organic phase was washed with brine (6 L × 3), dried over anhydrous

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The

residue was puriﬁed by reverse-phase ﬂash chromatography (20−55%

MeCN in H₂O (with 0.5% TFA)). The product-containing fractions

were collected, evaporated under reduced pressure, and extracted with

EtOAc (22 L × 3) to aﬀord 3-bromo-7-(((3aR,5aR,6S,8aR)-4-

methoxy-2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]-

dioxol-6-yl)oxy)quinolin-2-amine (17 (X = Br), 361 g, 61%) as a

mixture of anomers. ESI MS m/z = 451/453 [M + H]⁺.

Step 7: Synthesis of (3R,3aS,6R,6aR)-3a-(Benzyloxy)-2-methox-

yhexahydro-2H-cyclopenta[b]furan-3,6-diol (12). To a solution of

TsOH·H₂O (87.4 g, 460 mmol, 0.64 equiv) in MeOH (1.2 L) at 15−

20 °C was added (3aR,4aR,5R,7aR,7bR)-7a-(benzyloxy)-2,2-dime-

thylhexahydro-5H-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-ol (11,

220 g, 718 mmol, 1.0 equiv). The mixture was stirred at 15−20 °C

for 12 h. The reaction was poured into ice water (600 mL) and

neutralized with aqueous Na₂CO₃(300 mL). The aqueous phase was

extracted with EtOAc (600 mL × 4). The combined organic phase

was washed with brine (600 mL), dried over anhydrous Na₂SO₄,

ﬁltered, and concentrated under reduced pressure. The residue was

puriﬁed by silica gel column chromatography (20:1 to 0:1 petroleum

ether/EtOAc) to aﬀord (3R,3aS,6R,6aR)-3a-(benzyloxy)-2-methox-

yhexahydro-2H-cyclopenta[b]furan-3,6-diol (12, 150.4 g, 75%) as a

1

mixture of anomers as a yellow oil. H NMR (400 MHz, CDCl₃) δ

7.26−7.40 (m, 5H), 4.93−5.03 (m, 1H), 4.52−4.76 (m, 1H), 4.33−

4.45 (m, 1H), 4.00−4.19 (m, 1H), 3.78−3.97 (m, 1H), 3.46 (d, J =

7.60 Hz, 3H), 2.98−3.04 (m, 1H), 2.20−2.34 (m, 1H), 1.82−2.12

(m, 4H).

Step 12: Synthesis of (3R,3aS,6S,6aR)-6-((2-Amino-3-bromoqui-

nolin-7-yl)oxy)hexahydro-3aH-cyclopenta[b]furan-2,3,3a-triol (18,

X = Br). A solution of 3-bromo-7-(((3aR,5aR,6S,8aR)-4-methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl)oxy)-

quinolin-2-amine (17 (X = Br), 361 g, 800 mmol, 1.0 equiv) in HCl

(4 L, 0.4 M, 1.6 mol, 2.0 equiv) and MeCN (7.22 L) was stirred at 90

°C for 16 h. The reaction mixture was quenched with saturated

NaHCO₃(5 L) and extracted with EtOAc (10 L × 5). The combined

organic phase was washed with brine (5 L), dried over anhydrous

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The

residue was slurried with MeCN (5 volumes, 60 °C to RT, overnight)

to aﬀord (3R,3aS,6S,6aR)-6-((2-amino-3-bromoquinolin-7-yl)oxy)-

hexahydro-2H-cyclopenta[b]furan-2,3,3a-triol (18 (X = Br), 269.1 g,

85%) as a mixture of anomers. ESI MS m/z = 397/399 [M + H]⁺.

Step 13: Synthesis of (2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromo-

quinolin-7-yl)oxy)-2-(4-chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (20, X = Br, Y = Z =

Step 8: Synthesis of (3R,3aS,6R,6aR)-2-Methoxyhexahydro-3aH-

cyclopenta[b]furan-3,3a,6-triol (13). Pd(OH)₂/C (5 g, 7.1 mmol, 20

wt %, 0.04 equiv) was added to a solution of (3R,3aS,6R,6aR)-3a-

(benzyloxy)-2-methoxyhexahydro-2H-cyclopenta[b]furan-3,6-diol

(12, 50 g, 178 mmol, 1.0 equiv) in MeOH (350 mL) at 15−20 °C

under N₂. Then, acetic acid (8.8 g, 8.37 mL, 146 mmol, 0.82 equiv)

was added. The suspension was degassed under reduced pressure and

purged with H₂several times. The mixture was then stirred under H₂

(50 psi) at 50−55 °C for 12 h. Four reactions run in parallel were

combined. The reaction mixture was ﬁltered and concentrated under

reduced pressure to aﬀord (3R,3aS,6R,6aR)-2-methoxyhexahydro-

3aH-cyclopenta[b]furan-3,3a,6-triol (13) as a mixture of anomers,

1

which was used crude in the next reaction. H NMR (400 MHz,

CDCl₃) δ 5.01−4.95 (m, 1H), 4.24−4.16 (m, 2H), 3.81−3.80 (m,

1H), 3.49−3.47 (m, 3H), 2.39−2.07 (m, 3H), 2.06−1.94 (m, 2H),

1.73−1.64 (m, 2H).

3925

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

H, W = Cl). To a stirred solution of (3R,3aS,6S,6aR)-6-((2-amino-3-

bromoquinolin-7-yl)oxy)hexahydro-2H-cyclopenta[b]furan-2,3,3a-

triol (18 (X = Br), 140 mg, 0.35 mmol, 1.0 equiv) in dry MeCN (8

mL) under argon at 0 °C was added a solution of (E)-diazene-1,2-

diylbis(piperidin-1-ylmethanone) (133 mg, 0.53 mmol, 1.5 equiv) in

MeCN (0.5 mL) dropwise (over 2 min), followed by tributylphos-

phine (0.141 mL, 0.56 mmol, 1.6 equiv) over 5 min at 0 °C. The

reaction mixture was stirred at RT for 5 min, followed by heating to

35 °C for 45 min. Concurrently, to a stirred suspension of 4-chloro-

7H-pyrrolo[2,3-d]pyrimidine (19 (Y = Z = H, W = Cl), 101 mg, 0.66

mmol, 1.9 equiv) in dry MeCN (8 mL) was added DBU (0.1 mL,

0.66 mmol, 1.9 equiv) at RT. The reaction was stirred at RT for 30

min, and then the solution was transferred to the aforementioned

epoxide-containing reaction mixture by means of a syringe. The

resulting mixture was stirred at 35 °C for 2 h. The reaction mixture

was treated with brine (50 mL) and extracted with EtOAc (100 mL ×

3). The combined organic phase was dried over anhydrous Na₂SO₄,

ﬁltered, and concentrated under reduced pressure. The residue was

puriﬁed by silica gel column chromatography (1:0 to 9:1 DCM/

MeOH) to aﬀord (2R,3R,3aS,6S,6aR)-6-((2-amino-3-bromoquinolin-

7-yl)oxy)-2-(4-chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-

2H-cyclopenta[b]furan-3,3a-diol (20 (X = Br, Y = Z = H, W = Cl),

130 mg, 69%) as an oﬀ-white solid. Note that we found that NaH was

also an eﬀective base for this reaction. ESI MS m/z = 532/534 [M +

synthesized according to the general procedure described for

compound 4 using the appropriately substituted 16 and 19.

Step 11: 67%, ESI MS m/z = 391 [M + H]⁺; step 12: 73%, ESI MS

m/z = 337 [M + H]⁺; step 13: 43%, ESI MS m/z = 472 [M + H]⁺;

1

step 14: 45%, H NMR (400 MHz, DMSO-d₆) δ 8.09 (s, 1H), 7.73

(d, J = 11.6 Hz, 1H), 7.57−7.52 (m, 2H), 7.11 (br s, 2H), 6.87−6.85

(m, 2H), 6.69−6.66 (m, 3H), 6.02 (d, J = 8.4 Hz, 1H), 5.39 (d, J =

7.2 Hz, 1H), 5.33 (s, 1H), 4.59 (d, J = 4.8 Hz, 1H), 4.40 (t, J = 7.6

Hz, 1H), 4.08 (s, 1H), 2.50−2.45 (m, 1H), 2.07−1.97 (m, 3H).

HRMS (ESI+) m/z [M + H]⁺calculated for C₂₂H₂₁FN₆O₄: 453.1686;

found: 453.1688.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)oxy)-2-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (34). Compound 34 was similarly

synthesized according to the general procedure described for

compound 4 using the appropriately substituted 16 and 19.

1

Step 13: 46% (NaOtBu as a base). H NMR (400 MHz, DMSO-

d₆) δ 8.70 (s, 1H), 8.26 (s, 1H), 7.95 (d, J = 4.0 Hz, 1H), 7.57 (d, J =

8.8 Hz, 1H), 6.88−6.83 (m, 3H), 6.52 (br s, 2H), 6.15 (d, J = 8.4 Hz,

1H), 5.45 (d, J = 7.2 Hz, 1H), 5.39 (s, 1H), 4.64 (d, J = 4.8 Hz, 1H),

4.48 (t, J = 8.0 Hz, 1H), 4.14 (s, 1H), 2.69 (s, 3H), 2.56−2.50 (m,

1H), 2.10−1.99 (m, 3H). HRMS (ESI+) m/z [M + H]⁺calculated

for C₂₃H₂₂BrN₅O₄: 512.0933; found: 512.0926.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-chloroquinolin-7-yl)oxy)-2-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (35). Compound 35 was similarly

synthesized according to the general procedure described for

compound 4 using the appropriately substituted 16 and 19 (see 29

for the synthesis of triol).

1

H]⁺. H NMR (300 MHz, DMSO-d₆) δ 8.72 (s, 1H), 8.27 (s, 1H),

8.18 (d, J = 3.9 Hz, 1H), 7.57 (d, J = 8.7 Hz, 1H), 6.88−6.83 (m,

3H), 6.61 (br s, 2H), 6.17 (d, J = 8.4 Hz, 1H), 5.53 (d, J = 7.2 Hz,

1H), 5.44 (s, 1H), 4.67 (d, J = 4.8 Hz, 1H), 4.48 (t, J = 7.8 Hz, 1H),

4.15 (s, 1H), 2.51−2.50 (m, 1H), 2.10−1.95 (m, 3H).

Step 13: 24%. ¹H NMR (400 MHz, DMSO-d₆) δ 8.70 (s, 1H), 8.09

(s, 1H), 7.96 (d, J = 3.6 Hz, 1H), 7.57 (d, J = 8.8 Hz, 1H), 6.88−6.84

(m, 3H), 6.62 (br s, 2H), 6.15 (d, J = 8.4 Hz, 1H), 5.46 (d, J = 7.2 Hz,

1H), 5.40 (s, 1H), 4.64 (d, J = 5.6 Hz, 1H), 4.48 (t, J = 7.6 Hz, 1H),

4.13−4.12 (m, 1H), 2.69 (s, 3H), 2.56−2.54 (m, 1H), 2.08−2.01 (m,

3H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₃H₂₂ClN₅O₄:

468.1438; found: 468.1448.

Step 14: Synthesis of (2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromo-

quinolin-7-yl)oxy)-2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (4). A solution of

(2R,3R,3aS,6S,6aR)-6-((2-amino-3-bromoquinolin-7-yl)oxy)-2-(4-

chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta-

[b]furan-3,3a-diol (20 (X = Br, Y = Z = H, W = Cl), 55 mg, 0.10

mmol, 1.0 equiv) in 1,4-dioxane (15 mL) and NH₃·H₂O (15 mL) was

stirred in a sealed tube at 90 °C for 10 h. The reaction mixture was

concentrated under reduced pressure, and the residue was diluted

with saturated NH₄Cl (30 mL). The mixture was extracted with

DCM (25 mL × 5). The combined organic phase was washed with

brine (50 mL), dried over anhydrous Na₂SO₄, ﬁltered, and

concentrated under reduced pressure. The residue was puriﬁed by

silica gel ﬂash chromatography (8% MeOH/DCM) to aﬀord

(2R,3R,3aS,6S,6aR)-6-((2-amino-3-bromoquinolin-7-yl)oxy)-2-(4-

amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta-

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroquinolin-7-yl)oxy)-2-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (36). Compound 36 was similarly

synthesized according to the general procedure described for

compound 4 using the appropriately substituted 16 and 19 (see 30

for the synthesis of triol).

Step 13: 28%. ¹H NMR (400 MHz, DMSO-d₆) δ 8.70 (s, 1H), 7.95

(d, J = 4.0 Hz, 1H), 7.74−7.72 (m, 1H), 7.57−7.54 (m, 1H), 6.89−

6.83 (m, 3H), 6.67 (br s, 2H), 6.15 (d, J = 8.4 Hz, 1H), 5.46 (d, J =

6.8 Hz, 1H), 5.39 (s, 1H), 4.63 (d, J = 5.2 Hz, 1H), 4.50−4.46 (m,

1H), 4.13 (s, 1H), 2.69 (s, 3H), 2.53−2.50 (m, 1H), 2.09−1.98 (m,

3H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₃H₂₂FN₅O₄:

452.1734; found: 452.1743.

1

[b]furan-3,3a-diol (4, 16 mg, 30%) as a white solid. H NMR (400

MHz, DMSO-d₆) δ 8.26 (s, 1H), 8.09 (s, 1H), 7.57 (d, J = 8.8 Hz,

1H), 7.50 (d, J = 3.6 Hz, 1H), 7.05 (br s, 2H), 6.87−6.83 (m, 2H),

6.66 (d, J = 4.0 Hz, 1H), 6.53 (br s, 2H), 6.02 (d, J = 8.8 Hz, 1H),

5.36 (d, J = 7.2 Hz, 1H), 5.31 (s, 1H), 4.60 (d, J = 4.8 Hz, 1H), 4.40

(t, J = 7.6 Hz, 1H), 4.08 (s, 1H), 2.52−2.50 (m, 1H), 2.07−1.99 (m,

3H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₂H₂₁BrN₆O₄:

513.0886; found: 513.0891.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-chloroquinolin-7-yl)oxy)-2-(4-

amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (29). Compound 29 was similarly

synthesized according to the general procedure described for

compound 4 using the appropriately substituted 16 and 19.

Step 11: 79%, ESI MS m/z = 407 [M + H]⁺; step 12: 91%, ESI MS

m/z = 353 [M + H]⁺; step 13: 60%, ESI MS m/z = 488 [M + H]⁺;

step 14: 27%. ¹H NMR (300 MHz, DMSO-d₆) δ 8.09 (s, 2H), 7.59−

7.50 (m, 2H), 7.06 (br s, 2H), 6.86−6.83 (m, 2H), 6.65−6.62 (m,

3H), 6.01 (d, J = 8.1 Hz, 1H), 5.38 (d, J = 6.9 Hz, 1H), 5.33 (s, 1H),

4.61−4.59 (m, 1H), 4.41 (t, J = 7.8 Hz, 1H), 4.06 (s, 1H), 2.08−1.97

(m, 4H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₂H₂₁ClN₆O₄:

469.1391; found: 469.1382.

(2R,3R,3aS,6R,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)oxy)-2-

(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (40). Compound 40 was synthesized

from the common intermediate 14.

Step 1: To a solution of (3aR,5aR,6R,8aR)-4-methoxy-2,2-

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-ol (14,

2.0 g, 8.7 mmol, 1.0 equiv, mixture of anomers) in DMF (30 mL)

was added NaH (0.869 g, 60 wt % in mineral oil, 21.7 mmol, 2.5

equiv) at 0 °C. The mixture was stirred at 0 °C for 1 h. Then,

(bromomethyl)benzene (2.08 mL, 17.4 mmol, 2.0 equiv) was added

at 0 °C. The reaction was stirred at RT overnight. The mixture was

then diluted with water and extracted with EtOAc. The organic phase

was washed with aqueous NaCl (10%), followed by brine, and then

dried over Na₂SO₄, ﬁltered, and concentrated under reduced pressure.

The resulting crude orange oil containing (3aR,5aR,6R,8aR)-6-

(benzyloxy)-4-methoxy-2,2-dimethylhexahydrocyclopenta[2,3]furo-

[3,4-d][1,3]dioxole as a mixture of anomers was carried directly to the

next step.

Step 2: To a solution of (3aR,5aR,6R,8aR)-6-(benzyloxy)-4-

methoxy-2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]-

dioxole (2.8 g, 8.7 mmol, 1.0 equiv) in water (36.1 mL) and MeCN

(51.4 mL) was added HCl (2.08 mL, 25.3 mmol, 2.9 equiv). The

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroquinolin-7-yl)oxy)-2-(4-

amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (30). Compound 30 was similarly

3926

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

reaction was stirred at 80 °C for 4 h. The mixture was quenched with

NaHCO₃(2.13 g, 25.3 mmol, 2.9 equiv) at 0 °C and then extracted

with EtOAc. The organic phase was washed with brine, dried over

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The

resulting residue was puriﬁed by silica gel column chromatography

(0−60% (3:1 EtOAc/EtOH)/hexanes) to aﬀord (3R,3aS,6R,6aR)-6-

(benzyloxy)hexahydro-3aH-cyclopenta[b]furan-2,3,3a-triol (2.1 g,

91% (two steps)) as a mixture of anomers. ESI MS m/z = 289 [M

+ Na]⁺.

amine (64 mg, 60%) as a white solid. ESI MS m/z = 702/704 [M +

H]⁺.

Step 7: To a solution of 3-bromo-N-(2,4-dimethoxybenzyl)-7-

(((3aR,4R,5aR,6R,8aR)-2,2-dimethyl-4-(4-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-

yl)oxy)quinolin-2-amine (64 mg, 0.091 mmol, 1.0 equiv) in DCM

(0.91 mL) were added TFA (1.4 mL, 18.2 mmol, large excess) and

one drop of water. The reaction was then heated to 50 °C for 4 h. The

solution was concentrated under reduced pressure, and the residue

was puriﬁed by reverse-phase preparative HPLC (MeCN/H₂O) to

aﬀord (2R,3R,3aS,6R,6aR)-6-((2-amino-3-bromoquinolin-7-yl)oxy)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

Step 3: To a solution of 4-methyl-7H-pyrrolo[2,3-d]pyrimidine

(160 mg, 1.2 mmol, 2.0 equiv) in anhydrous DMF (1.72 mL) was

added NaH (72.1 mg, 60 wt % in mineral oil, 1.8 mmol, 3.0 equiv)

quickly at 0 °C under argon. The mixture was stirred for 30 min at

RT. To a second ﬂame-dried round-bottom ﬂask was added a solution

of (3R,3aS,6R,6aR)-6-(benzyloxy)hexahydro-3aH-cyclopenta[b]-

furan-2,3,3a-triol (160 mg, 0.6 mmol, 1.0 equiv) in anhydrous

MeCN (9 mL) under argon. This solution was cooled down to 0 °C,

and then tributylphosphine (231 mg, 1.1 mmol, 1.9 equiv) and 1,1′-

(azodicarbonyl)dipiperidine (273 mg, 1.1 mmol, 1.8 equiv) in MeCN

(1 mL) were added. This second mixture was stirred for 30 min at

RT. Then, the ﬁrst mixture containing the nucleobase was transferred

via a cannula into the second mixture at 0 °C. The resulting reaction

was stirred for 1 h at RT. The mixture was then quenched with

saturated NH₄Cl (10 mL) at 0 °C and diluted with EtOAc/H₂O (100

mL/30 mL). The organic layer was separated and washed with H₂O

(50 mL) and brine (50 mL). The organic phase was dried over

anhydrous Na₂SO₄, ﬁltered, and concentrated under reduced

pressure. The residue was puriﬁed by reverse-phase CombiFlash

chromatography (0−40-95% MeCN/H₂O) to aﬀord (2R,3R,3aS,6-

R,6aR)-6-(benzyloxy)-2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-2H-cyclopenta[b]furan-3,3a-diol (95 mg, 42%) as a white

solid. ESI MS m/z = 382 [M + H]⁺.

Step 4: To a mixture of (2R,3R,3aS,6R,6aR)-6-(benzyloxy)-2-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta-

[b]furan-3,3a-diol (2.4 g, 6.3 mmol, 1.0 equiv) in 2,2-dimethox-

ypropane (50 mL) was added 4-methylbenzenesulfonic acid (0.108 g,

0.63 mmol, 0.1 equiv) at RT under argon. The reaction was stirred at

70 °C for 48 h. The mixture was concentrated under reduced

pressure, and the residue was puriﬁed by silica gel column

chromatography (0−60% EtOAc/petroleum ether) to aﬀord 7-

( ( 3 a R , 4 R , 5 a R , 6 R , 8 a R ) - 6 - ( b e n z y l o x y ) - 2 , 2 -

dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-

methyl-7H-pyrrolo[2,3-d]pyrimidine (2.2 g, 83%) as a colorless oil.

ESI MS m/z = 422 [M + H]⁺.

1

cyclopenta[b]furan-3,3a-diol (40, 28 mg, 60%) as a white solid. H

NMR (500 MHz, DMSO-d₆) δ 8.63 (s, 1H), 8.24 (s, 1H), 7.78 (d, J =

3.8 Hz, 1H), 7.51 (d, J = 8.9 Hz, 1H), 6.94 (d, J = 2.2 Hz, 1H), 6.86

(dd, J = 8.8, 2.4 Hz, 1H), 6.77 (d, J = 3.7 Hz, 1H), 6.54 (s, 2H), 6.14

(d, J = 8.2 Hz, 1H), 5.46 (d, J = 7.0 Hz, 1H), 5.39 (s, 1H), 4.86−4.81

(m, 1H), 4.41 (d, J = 5.3 Hz, 1H), 4.33 (t, J = 7.6 Hz, 1H), 2.64 (s,

3H), 2.29−2.23 (m, 1H), 2.21−2.12 (m, 1H), 2.12−2.03 (m, 1H),

1.73−1.64 (m, 1H). HRMS (ESI+) m/z [M + H]⁺calculated for

C₂₃H₂₂BrN₅O₄: 512.0933; found: 512.0933.

(2R,3R,3aS,6R,6aR)-6-((2-Amino-3-chloroquinolin-7-yl)methyl)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (41). Compound 41 was synthesized

from the intermediate after step 5 for the synthesis of compound 40.

Step 1: To a mixture of (3aR,4R,5aR,6R,8aR)-2,2-dimethyl-4-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydrocyclopenta[2,3]-

furo[3,4-d][1,3]dioxol-6-ol (2.0 g, 6.0 mmol, 1.0 equiv) in DCM (60

mL) was added Dess−Martin periodinane (4.61 g, 10.9 mmol, 1.8

equiv) at RT under argon. The resulting mixture was stirred at RT for

1.5 h. The reaction was then quenched with saturated aqueous

NaHCO₃(150 mL) and extracted with EtOAc (250 mL × 3). The

combined organic phase was dried over anhydrous Na₂SO₄, ﬁltered,

and concentrated under reduced pressure. The residue was puriﬁed by

silica gel column chromatography (0−100% EtOAc/petroleum ether)

to aﬀord (3aR,4R,5aS,8aS)-2,2-dimethyl-4-(4-methyl-7H-pyrrolo[2,3-

d]pyrimidin-7-yl)tetrahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-

(5aH)-one (1.9 g, 96%) as a white solid. ESI MS m/z = 330 [M +

H]⁺.

Step 2: To a mixture of bromo(methyl)triphenylphosphorane (5.77

g, 16.2 mmol, 2.8 equiv) in THF (30 mL) was added nBuLi (6.0 mL,

2.5 M in hexane, 15 mmol, 2.6 equiv) at −10 °C under argon. The

resulting mixture was stirred at −10 °C for 30 min. Then,

(3aR,4R,5aS,8aS)-2,2-dimethyl-4-(4-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)tetrahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-

(5aH)-one (1.9 g, 5.8 mmol, 1.0 equiv) in THF (30 mL) was added

dropwise at −10 °C. The reaction mixture was stirred at −10 °C for 1

h. The reaction was quenched with saturated aqueous NH₄Cl (150

mL) and extracted with DCM (200 mL × 3). The combined organic

phase was dried over anhydrous Na₂SO₄, ﬁltered, and concentrated

under reduced pressure. The residue was puriﬁed by silica gel column

chromatography (0−60% EtOAc/petroleum ether) to aﬀord 7-

((3aR,4R,5aR,8aR)-2,2-dimethyl-6-methylenehexahydrocyclopenta-

[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-methyl-7H-pyrrolo[2,3-d]-

pyrimidine (1.04 g, 55%) as a white solid. ESI MS m/z = 328 [M +

H]⁺.

Step 5: To a solution of 7-((3aR,4R,5aR,6R,8aR)-6-(benzyloxy)-

2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-

methyl-7H-pyrrolo[2,3-d]pyrimidine (3.5 g, 8.3 mmol, 1.0 equiv) in

anhydrous MeOH (100 mL) was added Raney Ni (24 g, 50 wt % in

water) quickly at RT under argon. The resulting mixture was stirred at

60 °C for 5 h. The mixture was ﬁltered, and the ﬁltrate was

concentrated under reduced pressure. The crude residue was puriﬁed

by silica gel column chromatography (0−100% EtOAc/petroleum

ether) to aﬀord (3aR,4R,5aR,6R,8aR)-2,2-dimethyl-4-(4-methyl-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)hexahydrocyclopenta[2,3]furo[3,4-d]-

[1,3]dioxol-6-ol (2.15 g, 78%) as a white solid. ESI MS m/z = 332 [M

+ H]⁺.

Step 3: To a mixture of chloro(1,5-cyclooctadiene)iridium(I)

dimer (18 mg, 0.035 mmol, 0.05 equiv) and dppe (28 mg, 0.07 mmol,

0.1 equiv) in DCM (3 mL) was added 4,4,5,5-tetramethyl-1,3,2-

dioxaborolane (281 μL, 1.8 mmol, 2.5 equiv) under N₂. The mixture

was degassed and backﬁlled with N₂(three times). After stirring for

20 min at RT, a solution of 7-((3aR,4R,5aR,8aR)-2,2-dimethyl-6-

methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-

methyl-7H-pyrrolo[2,3-d]pyrimidine (230 mg, 0.7 mmol, 1.0 equiv)

in DCM (3 mL) was added under N₂. The mixture was degassed and

backﬁlled with N₂(three times), and the reaction was stirred at RT

for 15 h. The reaction was then cooled down to 0 °C, quenched with

MeOH (4 mL), and concentrated under reduced pressure. The crude

residue was puriﬁed by silica gel column chromatography (10−100%

EtOAc/DCM) to aﬀord 7-((3aR,4R,5aR,8aR)-2,2-dimethyl-6-

Step 6: (3aR,4R,5aR,6R,8aR)-2,2-Dimethyl-4-(4-methyl-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)hexahydrocyclopenta[2,3]furo[3,4-d]-

[1,3]dioxol-6-ol (50 mg, 0.15 mmol, 1.0 equiv), 3-bromo-N-(2,4-

dimethoxybenzyl)-7-iodoquinolin-2-amine (113 mg, 0.226 mmol, 1.5

equiv), 4-(pyrrolidin-1-yl)pyridine (27 mg, 0.18 mmol, 1.2 equiv),

cuprous iodide (3 mg, 0.015 mmol, 0.1 equiv), and potassium

phosphate (128 mg, 0.6 mmol, 4.0 equiv) were added to a microwave

vial charged with toluene (0.75 mL). The reaction was then heated at

120 °C overnight. The reaction was ﬁltered, and the ﬁltrate was

concentrated under reduced pressure. The crude residue was puriﬁed

by silica gel column chromatography (10−100% EtOAc/DCM) to

aﬀord 3-bromo-N-(2,4-dimethoxybenzyl)-7-(((3aR,4R,5aR,6R,8aR)-

2,2-dimethyl-4-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl)oxy)quinolin-2-

3927

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

((4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)methyl)-

hexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-methyl-7H-

pyrrolo[2,3-d]pyrimidine (338 mg) as a still impure clear resin, which

was used directly in the next reaction without further puriﬁcation. ESI

MS m/z = 456 [M + H]⁺.

8.26 (s, 1H), 8.18 (s, 1H), 7.58−7.49 (m, 3H), 6.87−6.83 (m, 2H),

6.65 (d, J = 3.6 Hz, 1H), 6.53 (br s, 2H), 6.03 (d, J = 8.0 Hz, 1H),

5.36 (d, J = 7.2 Hz, 1H), 5.31 (s, 1H), 4.61 (d, J = 5.2 Hz, 1H), 4.42

(t, J = 8.0 Hz, 1H), 4.09 (s, 1H), 2.98 (d, J = 4.4 Hz, 3H), 2.47−2.45

(m, 1H), 2.08−1.98 (m, 3H). HRMS (ESI+) m/z [M + H]⁺

calculated for C₂₃H₂₃BrN₆O₄: 527.1042; found: 527.1042.

Step 4: To a vial containing 7-((3aR,4R,5aR,8aR)-2,2-dimethyl-6-

((4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)methyl)-

hexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-4-methyl-7H-

pyrrolo[2,3-d]pyrimidine (75 mg, 0.17 mmol, 1.0 equiv), 7-bromo-3-

chloroquinolin-2-amine (51 mg, 0.198 mmol, 1.2 equiv), and

PdCl₂(dppf)·CH₂Cl₂(27 mg, 0.033 mmol, 0.2 equiv) were added

THF (2.75 mL) and water (0.55 mL). The mixture was degassed

under N₂(three times) and then charged with thallium(I) ethoxide

(35.0 μL, 0.49 mmol, 3.0 equiv). The reaction was heated for several

days at 65 °C. The reaction was then diluted with EtOAc and ﬁltered.

The ﬁltrate was washed with brine, dried over Na₂SO₄, ﬁltered, and

concentrated under reduced pressure. The crude residue was puriﬁed

by silica gel column chromatography (10−100% EtOAc/DCM) to

aﬀord 3-chloro-7-(((3aR,4R,5aR,8aR)-2,2-dimethyl-4-(4-methyl-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)hexahydrocyclopenta[2,3]furo[3,4-d]-

[1,3]dioxol-6-yl)methyl)quinolin-2-amine (16 mg, 15%), which was

still impure but was used directly in the next step. ESI MS m/z = 506

[M + H]⁺.

Representative Synthetic Procedure for (2R,3R,3aS,6S,6aR)-6-((2-

Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-(4-amino-7H-pyrrolo[2,3-

d]pyrimidin-7-yl)hexahydro-3aH-cyclopenta[b]furan-3,3a-diol

(33). Step 1: Synthesis of (3aR,5aS,8aS)-4-Methoxy-2,2-

dimethyltetrahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6(5aH)-

one (22). To a solution of oxalyl chloride (27.4 mL, 313 mmol, 4.0

equiv) in anhydrous DCM (360 mL) under N₂was added a solution

of DMSO (55.5 mL, 782 mmol, 10.0 equiv) in DCM (54 mL) at −78

°C. The solution was stirred at −78 °C for 30 min. Then, a solution of

(3aR,5aR,6R,8aR)-4-methoxy-2,2-dimethylhexahydrocyclopenta[2,3]-

furo[3,4-d][1,3]dioxol-6-ol (14, 18 g, 78 mmol, 1.0 equiv) in DCM

(18 mL) was added. The reaction was stirred at −78 °C for 30 min

before triethylamine (109 mL, 782 mmol, 10.0 equiv) was added. The

solution was stirred at −78 °C for 30 min. The reaction was then

quenched with H₂O (100 mL) at 0 °C and then extracted with DCM

(50 mL × 2). The combined organic phase was dried over anhydrous

Na₂SO₄, ﬁltered, and concentrated under reduced pressure. The crude

material was coevaporated with toluene (10 mL × 3) and then

puriﬁed by silica gel column chromatography (0−40% EtOAc/

h e x a n e s ) t o aﬀ o r d (3 aR , 5a S, 8a S) -4 -m et h o x y -2 , 2 -

dimethyltetrahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6(5aH)-

one (22, 14.03 g, 79%), which was isolated as two separate anomers

(7.93 g of the less polar anomer as a light yellow solid and 6.1 g of the

more polar anomer as a yellow orange oil). Note that we found that

Dess−Martin periodinane was also eﬀective for this transformation.

Additionally, the anomers could be carried forward together or

separately as there is an ultimate convergence to a common

intermediate (triol 26), but more commonly, one isomer was isolated

and carried forward. ¹H NMR (less polar isomer) (600 MHz, DMSO-

d₆) δ 4.97 (s, 1H), 4.39 (s, 1H), 4.15 (s, 1H), 3.09 (s, 3H), 2.49−2.41

Step 5: To a solution of 3-chloro-7-(((3aR,4R,5aR,8aR)-2,2-

dimethyl-4-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl)methyl)-

quinolin-2-amine (10 mg, 0.012 mmol, 1.0 equiv) in DCM (3 mL)

was added one drop of water followed by TFA (0.46 mL, 5.9 mmol,

large excess). The reaction was heated at 50 °C for 2 h. The reaction

was concentrated under reduced pressure, and the crude residue was

puriﬁed by SFC (Whelk-O (R,R), achiral prep, 35% MeOH (with

0.1% NH₄OH) mobile phase) to aﬀord (2R,3R,3aS,6R,6aR)-6-((2-

amino-3-chloroquinolin-7-yl)methyl)-2-(4-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (41, 3.7

mg, 67%) as a white solid. ¹H NMR (500 MHz, DMSO-d₆) δ 8.64 (s,

1H), 8.14 (s, 1H), 7.80 (d, J = 3.8 Hz, 1H), 7.58 (d, J = 8.2 Hz, 1H),

7.32 (s, 1H), 7.10 (d, J = 8.1 Hz, 1H), 6.77 (d, J = 3.7 Hz, 1H), 6.66

(s, 2H), 6.01 (d, J = 8.1 Hz, 1H), 5.47 (s, 1H), 5.31 (s, 1H), 4.33 (d, J

= 8.1 Hz, 1H), 3.93−3.90 (m, 1H), 2.90−2.85 (m, 1H), 2.73−2.69

(m, 1H), 2.66 (s, 3H), 2.35−2.29 (m, 1H), 2.05−1.95 (m, 2H),

1.90−1.85 (m, 1H), 1.63−1.57 (m, 1H). HRMS (ESI+) m/z [M +

H]⁺calculated for C₂₄H₂₄ClN₅O₃: 466.1646; found: 466.1654.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)oxy)-2-

(7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-cyclopenta[b]-

furan-3,3a-diol (42). Compound 42 was similarly synthesized

according to the general procedure described for compound 4 using

the appropriately substituted 19 (see 4 for the synthesis of triol).

Step 13: 18%. ¹H NMR (400 MHz, DMSO-d₆) δ 9.08 (s, 1H), 8.85

(s, 1H), 8.26 (s, 1H), 8.02 (s, 1H), 7.57 (d, J = 8.8 Hz, 1H), 6.83 (t, J

= 17.5 Hz, 3H), 6.51 (s, 2H), 6.19 (d, J = 8.2 Hz, 1H), 5.47 (d, J = 6.7

Hz, 1H), 5.40 (s, 1H), 4.65 (s, 1H), 4.51 (d, J = 8.4 Hz, 1H), 4.15 (s,

1H), 2.45−2.15 (m, 1H), 2.06 (br s, 3H). HRMS (ESI+) m/z [M +

H]⁺calculated for C₂₂H₂₀BrN₅O₄: 498.0777; found: 498.0777.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)oxy)-2-(4-

(methylamino)-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (46). Compound 46 was similarly

synthesized according to the general procedure described for

compound 4 (see 4 for the synthesis of −Cl-substituted nucleobase).

Step 14: A sealed tube containing a solution of (2R,3R,3aS,6S,6aR)-

6-((2-amino-3-bromoquinolin-7-yl)oxy)-2-(4-chloro-7H-pyrrolo[2,3-

d]pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-diol (20 (X

= Br, Y = Z = H, W = Cl), 30 mg, 0.056 mmol, 1.0 equiv) and

methylamine (2.5 mL, 2 M in THF, 5.0 mmol, large excess) in 1,4-

dioxane (2.5 mL) was stirred at 70 °C for 4 h. The mixture was then

concentrated under reduced pressure. The residue was puriﬁed by

reverse-phase CombiFlash (0−50−95% MeCN/5 mM aqueous

NH₄HCO₃gradient) to aﬀord (2R,3R,3aS,6S,6aR)-6-((2-amino-3-

bromoquinolin-7-yl)oxy)-2- (4-(methylamino)-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-diol (47, 16.7

1

(m, 2H), 2.39−2.29 (m, 2H), 1.38 (s, 3H), 1.36 (s, 3H). H NMR

(more polar isomer) (600 MHz, DMSO-d₆) δ 4.77 (d, J = 3.6 Hz,

1H), 4.53 (d, J = 3.6 Hz, 1H), 4.09 (s, 1H), 3.38 (s, 3H), 2.56−2.52

(m, 1H), 2.44−2.34 (m, 2H), 2.27−2.22 (m, 1H), 1.42 (s, 3H), 1.38

(s, 3H).

Step 2: Synthesis of (3aR,5aR,8aR)-4-Methoxy-2,2-dimethyl-6-

methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxole (23).

To a solution of methyltriphenylphosphonium bromide (26.9 g, 75

mmol, 2.2 equiv) in anhydrous THF (79 mL) under N₂was added

nBuLi (32 mL, 2.5 M in hexane, 80 mmol, 2.3 equiv) dropwise at 0

°C. The heterogeneous mixture was stirred at RT for 30 min. The

system was sonicated and cooled down to 0 °C. Then, a solution of

(3aR,5aS,8aS)-4-methoxy-2,2-dimethyltetrahydrocyclopenta[2,3]furo-

[3,4-d][1,3]dioxol-6(5aH)-one (22 (less polar anomer), 7.93 g, 34.7

mmol, 1.0 equiv) in anhydrous THF (79 mL) was added dropwise at

0 °C via an addition funnel over 20 min. The reaction mixture was

allowed to warm slowly to RT. The reaction was then quenched with

brine (20 mL) at RT. The mixture was extracted with EtOAc (50 mL

× 2). The combined organic phase was dried over anhydrous Na₂SO₄,

ﬁltered, and concentrated under reduced pressure. The residue was

puriﬁed by silica gel column chromatography (0−10% EtOAc/

hexanes) to aﬀord (3aR,5aR,8aR)-4-methoxy-2,2-dimethyl-6-

methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxole (23, 6 g,

76%). ¹H NMR (500 MHz, DMSO-d₆) δ 5.07 (s, 1H), 5.02 (s, 1H),

4.94 (s, 1H), 4.49 (s, 1H), 4.30 (s, 1H), 3.16 (s, 3H), 2.56−2.48 (m,

1H), 2.31−2.27 (m, 1H), 2.21−2.16 (m, 1H), 1.96−1.89 (m, 1H),

1.38 (s, 3H), 1.32 (s, 3H).

Step 3: Synthesis of 3-Fluoro-7-(((3aR,5aR,6S,8aR)-4-methoxy-

2,2-dimethylhexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-6-yl)-

methyl)quinolin-2-amine (25, A = B = C = H, X = F). To

( 3 a R , 5 a R , 8 a R ) - 4 - m e t h o x y - 2 , 2 - d i m e t h y l - 6 -

methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxole (23, 2 g,

8.8 mmol, 1.0 equiv) was added 9-BBN solution (88 mL, 0.5 M in

THF, 44.2 mmol, 5.0 equiv) at RT under argon. The reaction was

then stirred at 50 °C for 1 h. To this reaction solution was then added

1

mg, 56%) as an oﬀ-white solid. H NMR (400 MHz, DMSO-d₆) δ

3928

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a solution of K₃PO₄(9.39 g, 44.2 mmol, 5.0 equiv) in water (8 mL) at

0 °C under argon. The reaction was stirred at RT for 30 min. Then, a

solution of 7-bromo-3-ﬂuoroquinolin-2-amine (24 (A = B = C = H, V

= Br, X = F), 2.345 g, 9.73 mmol, 1.1 equiv) in THF (10 mL) and

Pd(dppf)Cl₂(0.976 g, 1.3 mmol, 0.15 equiv) was added at RT. The

ﬁnal reaction mixture was stirred at 75 °C for 2 h. The mixture was

diluted with water (100 mL) and extracted with EtOAc (250 mL ×

3). The combined organic phase was washed with brine (250 mL),

dried over anhydrous Na₂SO₄, ﬁltered, and concentrated under

reduced pressure. The residue was puriﬁed by silica gel column

chromatography (0−40% EtOAc/petroleum ether) to aﬀord 3-ﬂuoro-

7-(((3aR,5aR,6S,8aR)-4-methoxy-2,2-dimethylhexahydrocyclopenta-

[2,3]furo[3,4-d][1,3]dioxol-6-yl)methyl)quinolin-2-amine (25 (A = B

= C = H, X = F), 2.5 g, 73%) as a yellow solid. ESI MS m/z = 389 [M

+ H]⁺. ¹H NMR (300 MHz, DMSO-d₆) δ 7.77 (d, J = 12.0 Hz, 1H),

7.58 (d, J = 8.4 Hz, 1H), 7.36 (s, 1H), 7.14 (d, J = 8.1 Hz, 1H), 6.70

(s, 2H), 4.98 (s, 1H), 4.23−4.21 (m, 1H), 4.18 (d, J = 5.7 Hz, 1H),

3.45 (s, 3H), 2.85−2.81 (m, 1H), 2.67−2.61 (m, 1H), 2.17−2.00 (m,

2H), 1.87−1.76 (m, 1H), 1.58−1.51 (m, 2H), 1.35 (s, 3H), 1.30 (s,

3H).

Step 4: Synthesis of (3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroquino-

lin-7-yl)methyl)hexahydro-3aH-cyclopenta[b]furan-2,3,3a-triol

(26, A = B = C = H, X = F). A solution of 3-ﬂuoro-7-

(((3aR,5aR,6S,8aR)-4-methoxy-2,2-dimethylhexahydrocyclopenta-

[2,3]furo[3,4-d][1,3]dioxol-6-yl)methyl)quinolin-2-amine (25 (A = B

= C = H, X = F), 600 mg, 1.6 mmol, 1.0 equiv) in MeCN (14.3 mL),

H₂O (9.6 mL), and HCl (0.82 mL) was stirred at 90 °C for 1 h. Then,

the pH of the resulting solution was adjusted to pH 7 with 1 M

NaOH at 0 °C. The solvent was removed under reduced pressure,

and the crude residue was puriﬁed by silica gel column

chromatography (0−15% MeOH/DCM (with 0.1% NH₃·H₂O)) to

aﬀord (3R,3aS,6S,6aR)-6-((2-amino-3-ﬂuoroquinolin-7-yl)methyl)-

hexahydro-3aH-cyclopenta[b]furan-2,3,3a-triol (26 (A = B = C =

H, X = F), 365 mg, 71%) as a mixture of anomers as a light yellow

solid. ESI MS m/z = 335 [M + H]⁺.

Step 6: Synthesis of (2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroqui-

nolin-7-yl)methyl)-2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (33). A mixture of

(2R,3R,3aS,6S,6aR)-6-((2-amino-3-ﬂuoroquinolin-7-yl)methyl)-2-(4-

chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta-

[b]furan-3,3a-diol (27 (A = B = C = R = Y = Z = H, X = F, W = Cl),

50 mg, 0.11 mmol, 1.0 equiv) and NH₃·H₂O (2 mL, 28 wt %, large

excess) in dioxane (0.5 mL) was heated at 80 °C in a sealed tube

overnight. The crude material was puriﬁed by reverse-phase column

chromatography (15−35% MeCN/H₂O) to aﬀord (2R,3R,3aS,6-

S,6aR)-6-((2-amino-3-ﬂuoroquinolin-7-yl)methyl)-2-(4-amino-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-

diol (33, 34 mg, 71%) as a white solid. ¹H NMR (400 MHz, DMSO-

d₆) δ 8.09 (s, 1H), 7.74 (d, J = 12.0 Hz, 1H), 7.53 (d, J = 8.4 Hz, 1H),

7.44 (d, J = 3.6 Hz, 1H), 7.29 (s, 1H), 7.09−7.05 (m, 3H), 6.68−6.66

(m, 3H), 5.88 (d, J = 8.4 Hz, 1H), 5.25−5.20 (m, 1H), 5.07−5.05 (m,

1H), 4.13 (d, J = 8.0 Hz, 1H), 3.95 (d, J = 5.6 Hz, 1H), 2.84−2.79

(m, 1H), 2.67−2.58 (m, 1H), 2.29−2.13 (m, 1H), 1.96−1.92 (m,

1H), 1.74−1.68 (m, 2H), 1.57−1.49 (m, 1H). HRMS (ESI+) m/z

[M + H]⁺calculated for C₂₃H₂₃FN₆O₃: 451.1894; found: 451.1895.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (31). Compound 31 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24.

Step 3: 50% (24, A = B = C = H, X = Br, V = I), ESI MS m/z =

449/451 [M + H]⁺; step 4: 95%, ESI MS m/z = 395/397 [M + H]⁺;

1

step 5: 56%, ESI MS m/z = 530/532 [M + H]⁺; step 6: 62%. H

NMR (400 MHz, DMSO-d₆) δ 8.30 (s, 1H), 8.09 (s, 1H), 7.54 (d, J =

8.4 Hz, 1H), 7.43 (d, J = 4.0 Hz, 1H), 7.28 (s, 1H), 7.09 (d, J = 8.4

Hz, 1H), 7.03 (br s, 2H), 6.67 (d, J = 3.6 Hz, 1H), 6.54 (br s, 2H),

5.88 (d, J = 8.0 Hz, 1H), 5.23 (d, J = 7.2 Hz, 1H), 5.04 (s, 1H), 4.13

(t, J = 7.6 Hz, 1H), 3.95 (d, J = 6.0 Hz, 1H), 2.82 (dd, J = 13.2, 8.0

Hz, 1H), 2.62 (dd, J = 14.4, 7.2 Hz, 1H), 2.24−2.22 (m, 1H), 1.96−

1.92 (m, 1H), 1.70−1.68 (m, 2H), 1.57−1.51 (m, 1H). HRMS (ESI

+) m/z [M + H]⁺calculated for C₂₃H₂₃BrN₆O₃: 511.1093; found:

511.1099.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-chloroquinolin-7-yl)methyl)-

2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (32). Compound 32 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24.

Step 3: 78% (24, A = B = C = H, X = Cl, V = Br), ESI MS m/z =

405 [M + H]⁺; step 4: 77%, ESI MS m/z = 351 [M + H]⁺; step 5:

44%, ESI MS m/z = 486 [M + H]⁺; step 6: 64%. ¹H NMR (300 MHz,

DMSO-d₆) δ 8.12 (s, 1H), 8.09 (s, 1H), 7.56 (d, J = 8.1 Hz, 1H), 7.45

(d, J = 3.6 Hz, 1H), 7.28 (s, 1H), 7.10−7.00 (m, 3H), 6.68−6.65 (m,

3H), 5.90 (d, J = 8.1 Hz, 1H), 5.26 (d, J = 6.9 Hz, 1H), 5.07 (s, 1H),

4.13 (t, J = 7.8 Hz, 1H), 3.96 (d, J = 5.4 Hz, 1H), 2.85−2.81 (m, 1H),

2.65−2.58 (m, 1H), 2.27−2.24 (m, 1H), 1.97−1.93 (m, 1H), 1.75−

1.60 (m, 2H), 1.59−1.54 (m, 1H). HRMS (ESI+) m/z [M + H]⁺

calculated for C₂₃H₂₃ClN₆O₃: 467.1598; found: 467.1593.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (37). Compound 37 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24 (24

amine was 4-methoxybenzyl (-PMB)-protected for the synthesis of

37, but reaction conditions were the same as described in the general

procedure, with the exception of a ﬁnal deprotection step).

Step 3: 86% (24 (-PMB-protected), A = B = C = H, X = Br, V = I),

ESI MS m/z = 569/571 [M + H]⁺; step 4: 63%, ESI MS m/z = 515/

517 [M + H]⁺; step 5: 65%, ESI MS m/z = 630/632 [M + H]⁺.

Step 6 (-PMB deprotection): A solution of (2R,3R,3aS,6aR)-6-((3-

bromo-2-((4-methoxybenzyl)amino)quinolin-7-yl)methyl)-2-(4-

methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta-

[b]furan-3,3a-diol (150 mg, 0.21 mmol, 1.0 equiv) in TFA (2 mL)

was stirred at 60 °C for 1 h. The reaction was then quenched with

saturated aqueous NaHCO₃(40 mL) and extracted with EtOAc (40

mL × 3).The combined organic phase was washed with brine (40

mL), dried over anhydrous Na₂SO₄, ﬁltered, and concentrated under

Step 5: Synthesis of (2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroqui-

nolin-7-yl)methyl)-2-(4-chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (27, A = B = C = Y =

Z = H, X = F, W = Cl). To a solution of (3R,3aS,6S,6aR)-6-((2-amino-

3-ﬂuoroquinolin-7-yl)methyl)hexahydro-2H-cyclopenta[b]furan-

2,3,3a-triol (26 (A = B = C = H, X = F), 0.5 g, 1.5 mmol, 1.0 equiv) in

dry NMP (17 mL) was added tributylphosphine (0.862 g, 4.3 mmol,

2.8 equiv), followed by a solution of (E)-diazene-1,2-diylbis-

(piperidin-1-ylmethanone) (1 g, 4 mmol, 2.7 equiv) in NMP (4

mL) under N₂at RT. The reaction mixture was stirred at RT for 1.5

h. Concurrently, into a separate ﬂask under N₂was added DBU

(0.676 mL, 4.49 mmol, 3.0 equiv) to a solution of 4-chloro-7H-

pyrrolo[2,3-d]pyrimidine (19 (Y = Z = H, W = Cl), 0.46 g, 3 mmol,

2.0 equiv) in dry NMP (10 mL) at 0 °C, and then the suspension was

stirred at RT for 1.5 h. This suspension was then transferred to the

aforementioned crude epoxide-containing solution via a syringe at

RT. The resulting mixture was stirred at RT overnight. The reaction

was then quenched with saturated aqueous NH₄Cl (80 mL) and

extracted with EtOAc (60 mL × 4). The combined organic phase was

dried over anhydrous MgSO₄, ﬁltered, and concentrated under

reduced pressure. The crude residue was puriﬁed by silica gel column

chromatography (15:1 DCM/MeOH). The product fractions were

isolated, and this material was further puriﬁed by reverse-phase

column chromatography (15−35% MeCN/H₂O) to aﬀord

(2R,3R,3aS,6S,6aR)-6-((2-amino-3-ﬂuoroquinolin-7-yl)methyl)-2-(4-

chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-cyclopenta-

[b]furan-3,3a-diol (27 (A = B = C = Y = Z = H, X = F, W = Cl),

0.315 g, 45%) as a yellow solid. Note that we also found that NaH was

1

an eﬀective base for this reaction. ESI MS m/z = 470 [M + H]⁺. H

NMR (400 MHz, DMSO-d₆) δ 8.72 (s, 1H), 8.12 (d, J = 4.0 Hz, 1H),

7.75 (d, J = 12.0 Hz, 1H), 7.54 (d, J = 8.0 Hz, 1H), 7.30 (s, 1H),

7.11−7.09 (m, 1H), 6.83 (d, J = 4.0 Hz, 1H), 6.68 (s, 2H), 6.03 (d, J

= 8.0 Hz, 1H), 5.40 (d, J = 7.2 Hz, 1H), 5.17 (s, 1H), 4.24 (t, J = 7.2

Hz, 1H), 4.05 (d, J = 5.6 Hz, 1H), 2.86−2.81 (m, 1H), 2.67−2.62 (m,

1H), 2.34−2.27 (m, 1H), 1.99−1.95 (m, 1H), 1.83−1.42 (m, 3H).

3929

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

reduced pressure. The residue was puriﬁed by preparative TLC (8:1

DCM/MeOH) followed by reverse-phase CombiFlash (0−31−95%

MeCN/5 mM aqueous NH₄HCO₃gradient) to aﬀord (2R,3R,3aS,6-

S,6aR)-6-((2-amino-3-bromoquinolin-7-yl)methyl)-2-(4-methyl-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-

diol (37, 60 mg, 55%) as a white solid. ¹H NMR (400 MHz, DMSO-

d₆) δ 8.69 (s, 1H), 8.30 (s, 1H), 7.87 (d, J = 3.6 Hz, 1H), 7.54 (d, J =

8.4 Hz, 1H), 7.27 (s, 1H), 7.08 (d, J = 8.4 Hz, 1H), 6.82 (d, J = 3.6

Hz, 1H), 6.53 (br s, 2H), 6.01 (d, J = 8.0 Hz, 1H), 5.30 (d, J = 6.8 Hz,

1H), 5.11 (s, 1H), 4.22 (t, J = 7.6 Hz, 1H), 4.00 (d, J = 5.6 Hz, 1H),

2.83 (dd, J = 13.6, 7.6 Hz, 1H), 2.70 (s, 3H), 2.64 (dd, J = 13.6, 7.6

Hz, 1H), 2.24−2.35 (m, 1H), 1.97 (dd, J = 12.4, 5.6 Hz, 1H), 1.79−

1.68 (m, 2H), 1.58−1.52 (m, 1H). HRMS (ESI+) m/z [M + H]⁺

calculated for C₂₄H₂₄BrN₅O₃: 510.1141; found: 510.1158.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-chloroquinolin-7-yl)methyl)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (38). Compound 38 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24 (see

32 for details on triol synthesis).

reduced pressure. The residue was puriﬁed by preparative HPLC

(20−58% MeCN/10 mM aqueous NH₄HCO₃) to aﬀord

(2R,3R,3aS,6S,6aR)-6-((2-aminoquinolin-7-yl)methyl)-2-(4-methyl-

7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-

1

3,3a-diol (43, 7.4 mg, 44%) as a white solid. H NMR (400 MHz,

DMSO-d₆) δ 8.69 (s, 1H), 7.87 (d, J = 3.6 Hz, 1H), 7.79 (d, J = 8.8

Hz, 1H), 7.48 (d, J = 8.0 Hz, 1H), 7.22 (s, 1H), 7.00 (d, J = 8.0 Hz,

1H), 6.82 (d, J = 3.6 Hz, 1H), 6.66 (d, J = 8.8 Hz, 1H), 6.31 (br s,

2H), 6.01 (d, J = 8.0 Hz, 1H), 5.29 (d, J = 7.2 Hz, 1H), 5.11 (s, 1H),

4.22 (t, J = 7.6 Hz, 1H), 4.01 (d, J = 5.6 Hz, 1H), 2.84−2.78 (m, 1H),

2.69 (s, 3H), 2.63−2.58 (m, 1H), 2.28−2.25 (m, 1H), 1.98−1.94 (m,

1H), 1.80−1.68 (m, 2H), 1.58−1.53 (m, 1H). HRMS (ESI+) m/z

[M + H]⁺calculated for C₂₄H₂₅N₅O₃: 432.2035; found: 432.2039.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-methylquinolin-7-yl)methyl)-

2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (45). Compound 44 was synthesized in

one step from compound 31.

Step 1: To a microwave tube were added (2R,3R,3aS,6S,6aR)-6-

((2-amino-3-bromoquinolin-7-yl)methyl)-2-(4-amino-7H-pyrrolo-

[2,3-d]pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-diol

(31, 40 mg, 0.078 mmol, 1.0 equiv), 2,4,6-trimethyl-1,3,5,2,4,6-

trioxatriborinane (19.6 mg, 0.16 mmol, 2.0 equiv), Pd(dppf)Cl₂(14.4

mg, 0.020 mmol, 0.26 equiv), K₂CO₃(0.96 mL, 2 M in water, 1.92

mmol, 25.0 equiv), and DMF (2 mL) at room temperature under

argon. The reaction mixture was irradiated for 1 h at 130 °C. The

reaction mixture was cooled down to room temperature and

combined with a previous batch. The solvent was removed under

reduced pressure. The residue was puriﬁed by reverse-phase

CombiFlash (0−45−95% MeCN/5 mM aqueous NH₄HCO₃

gradient) followed by preparative HPLC (28−49% MeCN/10 mM

aqueous NH₄HCO₃) to aﬀord (2R,3R,3aS,6S,6aR)-6-((2-amino-3-

methylquinolin-7-yl)methyl)-2-(4-amino-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-diol (22.6 mg,

Step 5: 32%. ¹H NMR (400 MHz, DMSO-d₆) δ 8.69 (s, 1H), 8.12

(s, 1H), 7.88 (d, J = 3.6 Hz, 1H), 7.54 (d, J = 8.0 Hz, 1H), 7.28 (s,

1H), 7.08 (dd, J = 8.4, 1.6 Hz, 1H), 6.82 (d, J = 3.6 Hz, 1H), 6.64 (br

s, 2H), 6.00 (d, J = 8.4 Hz, 1H), 5.31 (d, J = 6.8 Hz, 1H), 5.12 (s,

1H), 4.22 (t, J = 8.0 Hz, 1H), 4.00 (d, J = 6.0 Hz, 1H), 2.85−2.79 (m,

1H), 2.69 (s, 3H), 2.65−2.60 (m, 1H), 2.33−2.25 (m, 1H), 1.98−

1.93 (m, 1H), 1.80−1.67 (m, 2H), 1.58−1.51 (m, 1H). HRMS (ESI

+) m/z [M + H]⁺calculated for C₂₄H₂₄ClN₅O₃: 466.1646; found:

466.1654.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (39). Compound 39 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24 (see

33 for details on triol synthesis).

1

37%) as a white solid. H NMR (400 MHz, DMSO-d₆) δ 8.10 (s,

Step 5: 11%. ¹H NMR (400 MHz, DMSO-d₆) δ 8.69 (s, 1H), 7.87

(d, J = 4.0 Hz, 1H), 7.74 (d, J = 11.6 Hz, 1H), 7.52 (d, J = 8.0 Hz,

1H), 7.28 (s, 1H), 7.07 (d, J = 8.0 Hz, 1H), 6.82 (d, J = 3.6 Hz, 1H),

6.66 (br s, 2H), 6.01 (d, J = 8.0 Hz, 1H), 5.31 (d, J = 7.2 Hz, 1H),

5.12 (s, 1H), 4.22 (d, J = 7.6 Hz, 1H), 4.01 (d, J = 6.0 Hz, 1H), 2.84−

2.79 (m, 1H), 2.69 (s, 3H), 2.67−2.59 (m, 1H), 2.28−2.22 (m, 1H),

1.98−1.94 (m, 1H), 1.76−1.69 (m, 2H), 1.58−1.53 (m, 1H). HRMS

(ESI+) m/z [M + H]⁺calculated for C₂₄H₂₄FN₅O₃: 450.1941; found:

450.1952.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-cyclopropyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-

cyclopenta[b]furan-3,3a-diol (43). Compound 43 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24 (see

31 for triol synthesis).

1H), 7.65 (s, 1H), 7.47−7.45 (m, 2H), 7.25 (s, 1H), 7.06 (br s, 2H),

7.00 (dd, J = 8.0, 1.2 Hz, 1H), 6.68 (d, J = 4.0 Hz, 1H), 6.16 (br s,

2H), 5.90 (d, J = 8.4 Hz, 1H), 5.25 (d, J = 7.2 Hz, 1H), 5.07 (s, 1H),

4.14 (t, J = 7.6 Hz, 1H), 3.97 (d, J = 5.6 Hz, 1H), 2.84−2.79 (m, 1H),

2.63−2.58 (m, 1H), 2.28−2.22 (m, 1H), 2.19 (s, 3H), 1.95 (dd, J =

12.0, 4.4 Hz, 1H), 1.75−1.70 (m, 2H), 1.56−1.51 (m, 1H). HRMS

(ESI+) m/z [M + H]⁺calculated for C₂₄H₂₆N₆O₃: 447.2144; found:

447.2136.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-(methylamino)-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-

3aH-cyclopenta[b]furan-3,3a-diol (47). Compound 47 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24. This

compound was carried through with a -PMB protecting group on the

aminoquinoline similarly to compound 37, and the reaction

conditions were the same as described in the general procedure

(see 37 for the synthesis of -PMB triol).

Step 5: 10%. ¹H NMR (500 MHz, DMSO-d₆) δ 8.62 (s, 1H), 8.31

(s, 1H), 7.86 (d, J = 3.8 Hz, 1H), 7.55 (d, J = 8.3 Hz, 1H), 7.28 (s,

1H), 7.09 (dd, J = 8.2, 1.4 Hz, 1H), 6.96 (d, J = 3.8 Hz, 1H), 6.55 (s,

2H), 6.00 (d, J = 8.1 Hz, 1H), 5.31 (d, J = 7.0 Hz, 1H), 5.12 (s, 1H),

4.22 (t, J = 7.6 Hz, 1H), 4.00 (d, J = 5.7 Hz, 1H), 2.86−2.79 (m, 1H),

2.67−2.60 (m, 1H), 2.58−2.53 (m, 1H), 2.33−2.24 (m, 1H), 2.00−

1.94 (m, 1H), 1.83−1.67 (m, 2H), 1.60−1.51 (m, 1H), 1.24−1.20

(m, 2H), 1.16−1.12 (m, 2H). HRMS (ESI+) m/z [M + H]⁺

calculated for C₂₆H₂₆BrN₅O₃: 536.1297; found: 536.1311.

(2R,3R,3aS,6S,6aR)-6-((2-Aminoquinolin-7-yl)methyl)-2-(4-meth-

yl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-3aH-cyclopenta[b]-

furan-3,3a-diol (44). Compound 44 was synthesized in one step from

compound 37.

Step 1: To a solution of (2R,3R,3aS,6S,6aR)-6-((2-amino-3-

bromoquinolin-7-yl)methyl)-2-(4-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydro-2H-cyclopenta[b]furan-3,3a-diol (37, 20

mg, 0.039 mmol, 1.0 equiv) in MeOH (5 mL) were added

triethylamine (4.0 mg, 0.039 mmol, 1.0 equiv) and anhydrous Pd/

C (10 mg, 10% Pd/C) under argon. The resulting mixture was stirred

at RT under an atmosphere of H₂(∼1.2 atm) for 30 min. The

mixture was ﬁltered, and the ﬁltrate was then concentrated under

Step 5: 67%, ESI MS m/z = 650/652 [M + H]⁺; step 6 (similar to

step 14 for compound 46): ND (material used crude without

puriﬁcation), ESI MS m/z = 645/647 [M + H]⁺; step 7 (-PMB

1

deprotection, similar to step 6 for 37): 68% (two steps). H NMR

(600 MHz, DMSO-d₆) δ 8.31 (s, 1H), 8.19 (s, 1H), 7.60−7.53 (m,

2H), 7.45 (d, J = 3.6 Hz, 1H), 7.28 (s, 1H), 7.09 (dd, J = 8.2, 1.4 Hz,

1H), 6.67 (d, J = 3.6 Hz, 1H), 6.57 (s, 2H), 5.90 (d, J = 8.1 Hz, 1H),

5.25 (d, J = 7.1 Hz, 1H), 5.07 (s, 1H), 4.15 (t, J = 7.4 Hz, 1H), 3.96

(d, J = 5.7 Hz, 1H), 2.99 (d, J = 4.5 Hz, 3H), 2.85−2.79 (m, 1H),

2.65−2.60 (m, 1H), 2.30−2.22 (m, 1H), 1.97−1.92 (m, 1H), 1.77−

1.66 (m, 2H), 1.58−1.50 (m, 1H). HRMS (ESI+) m/z [M + H]⁺

calculated for C₂₄H₂₅BrN₆O₃: 525.1250; found: 525.1262.

(2R,3R,3aS,6S,6aR)-2-(4-Amino-2-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)-6-((2-amino-3-bromoquinolin-7-yl)methyl)-

hexahydro-3aH-cyclopenta[b]furan-3,3a-diol (48). Compound 48

was similarly synthesized according to the general procedure

described for compound 33 using the appropriately substituted 19

and 24 (see 31 for triol synthesis).

3930

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Step 5: ND, ESI MS m/z = 544/546 [M + H]⁺; step 6: 24% (two

steps). H NMR (500 MHz, DMSO-d₆, TFA salt) δ 8.60 (s, 1H),

at RT. Then, tributylphosphine (7.6 mL, 30.5 mmol, 1.6 equiv) was

added via a syringe over 5 min at RT. The reaction was stirred at 46

°C for 3 h. The resultant mixture was used in the next step directly. In

parallel, to a solution of 4-chloro-7H-pyrrolo[2,3-d]pyrimidine (19 (Y

= Z = H, W = Cl), 5.6 g, 36.2 mmol, 1.9 equiv) in dry MeCN (30

mL) under argon was added DBU (5.18 mL, 34.3 mmol, 1.8 equiv) at

RT. The resulting mixture was stirred at RT for 30 min. Then, this

solution was transferred to the aforementioned epoxide-containing

solution by means of a syringe. The resulting mixture was stirred at 46

°C for 2 h and then concentrated under reduced pressure. The

residue was puriﬁed by silica gel column chromatography (0−30%

EtOAc/petroleum ether) to aﬀord (2R,3R,3aS,6aR)-3a-(benzyloxy)-

2-(4-chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-6-methylenehexahy-

dro-2H-cyclopenta[b]furan-3-ol (4.0 g, 53%) as an oﬀ-white solid. ESI

MS m/z = 398 [M + H]⁺. ¹H NMR (400 MHz, DMSO-d₆) δ 8.71 (s,

1H), 7.96 (d, J = 4.0 Hz, 1H), 7.45−7.28 (m, 5H), 6.80 (d, J = 3.6

Hz, 1H), 6.30 (d, J = 8.0 Hz, 1H), 5.83 (d, J = 6.8 Hz, 1H), 5.14 (d, J

= 16.0 Hz, 2H), 4.92 (d, J = 12.0 Hz, 1H), 4.71 (d, J = 11.6 Hz, 1H),

4.67 (s, 1H), 4.61−4.58 (m, 1H), 2.84−2.78 (m, 1H), 2.56−2.51 (m,

1H), 2.19−2.13 (m, 1H), 2.09−2.04 (m, 1H).

Step 4: To a solution of (2R,3R,3aS,6aR)-3a-(benzyloxy)-2-(4-

chloro-7H-pyrrolo[2,3-d] pyrimidin-7-yl)-6-methylenehexahydro-2H-

cyclopenta[b]furan-3-ol (690 mg, 1.7 mmol, 1.0 equiv) in DCM (10

mL) was dropwise added trichloroborane (3.47 mL, 1 M in DCM, 3.5

mmol, 2.0 equiv) at −78 °C under argon. The solution was then

stirred at −78 °C for 3 h. The reaction mixture was quenched with

triethylamine (0.967 mL, 6.9 mmol, 4.0 equiv) and stirred at −78 °C

for 30 min. Then, the reaction was poured into saturated aqueous

NaHCO₃(150 mL) at 0 °C and vigorously stirred for 30 min. The

ﬁnal mixture was extracted with EtOAc (200 mL × 3). The combined

organic phase was dried over anhydrous Na₂SO₄, ﬁltered, and

concentrated under reduced pressure. The residue was puriﬁed by

silica gel column chromatography (0−10% MeOH/DCM) to aﬀord

(2R,3R,3aS,6aR)-2-(4-chloro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-6-

methylene hexahydro-2H-cyclopenta[b]furan-3,3a-diol (500 mg, 80%,

85% purity by LCMS) as a yellow solid. ESI MS m/z = 308 [M + H]⁺.

¹H NMR (400 MHz, DMSO-d₆) δ 8.72 (s, 1H), 7.95 (d, J = 4.0 Hz,

1H), 6.78 (d, J = 4.0 Hz, 1H), 6.21 (d, J = 8.0 Hz, 1H), 5.52 (d, J =

7.2 Hz, 1H), 5.38 (s, 1H), 5.12−5.07 (m, 2H), 4.44−4.34 (m, 2H),

2.78−2.69 (m, 1H), 2.51−2.42 (m, 1H), 2.08−2.03 (m, 1H), 1.72−

1.64 (m, 1H).

1

7.74 (d, J = 3.7 Hz, 1H), 7.68 (d, J = 8.1 Hz, 1H), 7.39 (s, 1H), 7.24

(d, J = 7.8 Hz, 1H), 6.99 (d, J = 3.7 Hz, 1H), 5.90 (d, J = 8.1 Hz, 1H),

4.10 (d, J = 8.1 Hz, 1H), 4.03 (d, J = 5.8 Hz, 1H), 2.91−2.84 (m,

1H), 2.73−2.63 (m, 1H), 2.57 (s, 3H), 2.38−2.26 (m, 1H), 1.99−

1.91 (m, 1H), 1.78−1.68 (m, 2H), 1.59−1.50 (m, 1H). Integration

accounts for 19 of the 25 protons as exchangeable protons are not

observed. HRMS (ESI+) m/z [M + H]⁺calculated for

C₂₄H₂₅BrN₆O₃: 525.1250; found: 525.1261.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-

3aH-cyclopenta[b]furan-3,3a-diol (49). Compound 49 was similarly

synthesized according to the general procedure described for

compound 33 using the appropriately substituted 19 and 24 (see

33 for triol synthesis).

Step 5: 54%, ESI MS m/z = 488 [M + H]⁺; step 6: 42%. ¹H NMR

(300 MHz, DMSO-d₆) δ 8.12 (s, 1H), 7.78 (d, J = 12.0 Hz, 1H),

7.56−7.51 (m, 2H), 7.32 (s, 1H), 7.12−7.06 (m, 3H), 6.70 (br s,

2H), 5.97−5.94 (m, 1H), 5.29 (d, J = 6.9 Hz, 1H), 5.09 (s, 1H), 4.03

(t, J = 7.8 Hz, 1H), 3.97 (d, J = 5.7 Hz, 1H), 2.84 (dd, J = 13.8, 8.1

Hz, 1H), 2.63 (dd, J = 12.9, 6.9 Hz, 1H), 2.28−2.22 (m, 1H), 1.96−

1.91 (m, 1H), 1.78−1.65 (m, 2H), 1.57−1.46 (m, 1H). HRMS (ESI

+) m/z [M + H]⁺calculated for C₂₃H₂₂F₂N₆O₃: 469.1799; found:

469.1805.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3,5-diﬂuoroquinolin-7-yl)-

methyl)-2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-

3aH-cyclopenta[b]furan-3,3a-diol (50). Compound 50 was synthe-

sized using a modiﬁed route compared to Scheme 2, wherein the

nucleobase was installed before the quinoline. That synthesis is

described here from the common intermediate 10.

Step 1: To

a stirred mixture of bromo(methyl)-

triphenylphosphorane (28.3 g, 79 mmol, 2.8 equiv) in THF (109

mL) was added nBuLi (28.3 mL, 2.5 M in hexane, 70.6 mmol, 2.5

equiv) dropwise at −60 °C under argon. The resulting mixture was

stirred for 0.5 h at RT. Then, (3aR,4aS,7aS,7bR)-7a-(benzyloxy)-2,2-

dimethyltetrahydro-3aH-cyclopenta[4,5]furo[2,3-d][1,3]dioxol-5-

(4aH)-one (8.6 g, 28.3 mmol, 1.0 equiv) in THF (109 mL) was

added at −60 °C. The reaction mixture was stirred for 2 h at RT. The

mixture was quenched with brine (200 mL) at 0 °C and extracted

with EtOAc (300 mL × 3). The combined organic phase was dried

over anhydrous Na₂SO₄, ﬁltered, and concentrated under reduced

pressure. The residue was puriﬁed by silica gel column chromatog-

raphy (0−30% EtOAc/petroleum ether) to aﬀord (3aR,4a-

R,7aR,7bR)-7a-(benzyloxy)-2,2-dimethyl-5-methylenehexahydro-

3aH-cyclopenta[4,5]furo[2,3-d][1,3]dioxole (6.8 g, 80%) as an oﬀ-

Step 5: To a mixture of (2R,3R,3aS,6aR)-2-(4-chloro-7H-pyrrolo-

[2,3-d]pyrimidin-7-yl)-6-methylenehexahydro-2H-cyclopenta[b]-

furan-3,3a-diol (720 mg, 2.3 mmol, 1.0 equiv) in 2,2-dimethox-

ypropane (2 mL) was added 4-methylbenzenesulfonic acid (40.3 mg,

0.23 mmol, 0.1 equiv) at RT under argon. The mixture was stirred for

16 h at RT. The reaction was then quenched with NaHCO₃(200 mg)

at RT. The mixture was concentrated under reduced pressure, and the

residue was puriﬁed by silica gel column chromatography (0−30%

EtOAc/petroleum ether) to aﬀord 4-chloro-7-((3aR,4R,5aR,8aR)-2,2-

dimethyl-6-methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]-

dioxol-4-yl)-7H-pyrrolo[2,3-d]pyrimidine (740 mg, 91%) as a color-

less oil. ESI MS m/z = 348 [M + H]⁺. ¹H NMR (400 MHz, DMSO-

d₆) δ 8.72 (s, 1H), 7.87 (d, J = 4.0 Hz, 1H), 6.79 (d, J = 4.0 Hz, 1H),

6.36 (d, J = 3.6 Hz, 1H), 5.30 (d, J = 3.6 Hz, 1H), 5.15−5.14 (m,

2H), 4.68 (s, 1H), 2.58−2.41 (m, 3H), 2.04−1.93 (m, 1H), 1.57 (s,

3H), 1.39 (s, 3H).

1

white solid. H NMR (400 MHz, DMSO-d₆) δ 7.34−7.25 (m, 5H),

5.87 (d, J = 4.0 Hz, 1H), 5.23−5.22 (m, 1H), 5.10−5.09 (m, 1H),

4.68 (d, J = 3.6 Hz, 1H), 4.59 (d, J = 11.2 Hz, 1H), 4.51 (d, J = 11.2

Hz, 1H), 4.44 (s, 1H), 2.49−2.39 (m, 2H), 2.22−2.16 (m, 1H),

1.63−1.55 (m, 1H), 1.51 (s, 3H), 1.32 (s, 3H).

Step 2: To neat (3aR,4aR,7aR,7bR)-7a-(benzyloxy)-2,2-dimethyl-5-

methylenehexahydro-3aH-cyclopenta[4,5]furo[2,3-d][1,3]dioxole

(6.8 g, 22.5 mmol, 1.0 equiv) was added a solution of TFA (45.3 mL)

in water (11.33 mL) at 0 °C. The resulting mixture was then stirred

for 15 min at RT. The mixture was neutralized with 2 M NaOH and

extracted with EtOAc (200 mL × 4). The combined organic phase

was dried over anhydrous Na₂SO₄, ﬁltered, and concentrated under

reduced pressure. The residue was puriﬁed by silica gel column

chromatography (0−70% EtOAc/petroleum ether) to aﬀord

(3R,3aS,6aR)-3a-(benzyloxy)-6-methylenehexahydro-2H-cyclopenta-

[b]furan-2,3-diol (5 g, 85%) as a colorless solid as a mixture of

anomers in a 5:4 ratio. ¹H NMR (400 MHz, DMSO-d₆) δ 7.37−7.24

(m, 5H), 6.51−6.06 (m, 1H), 5.25−4.87 (m, 4H), 4.68−4.36 (m,

3H), 3.87−3.76 (m, 1H), 2.57−2.33 (m, 2H), 2.10−1.72 (m, 2H).

Step 3: To a stirred solution of (3R,3aS,6aR)-3a-(benzyloxy)-6-

methylenehexahydro-2H- cyclopenta[b]furan-2,3-diol (5.0 g, 19.1

mmol, 1.0 equiv) in dry MeCN (63 mL) under argon was added

dropwise a solution of (E)-diazene-1,2-diylbis(piperidin-1-ylmetha-

none) (7.2 g, 28.6 mmol, 1.5 equiv) in MeCN (63 mL) via a syringe

Step 6: An amination was conducted similar to that of step 14 for

compound 4 to aﬀord 7-((3aR,4R,5aR,8aR)-2,2-dimethyl-6-

methylenehexahydrocyclopenta[2,3]furo[3,4-d][1,3]dioxol-4-yl)-7H-

pyrrolo[2,3-d]pyrimidin-4-amine (82%). ESI MS m/z = 329 [M +

H]⁺. ¹H NMR (400 MHz, DMSO-d₆) δ 8.09 (s, 1H), 7.29 (d, J = 3.6

Hz, 1H), 7.10 (s, 2H), 6.64 (d, J = 3.6 Hz, 1H), 6.20 (d, J = 4.4 Hz,

1H), 5.19 (d, J = 4.0 Hz, 1H), 5.13−5.11 (m, 2H), 4.55 (s, 1H),

2.63−2.42 (m, 3 H), 2.01−1.96 (m, 1 H), 1.55 (s, 3H), 1.38 (s, 3H).

Step 7/8: A Suzuki coupling was conducted similar to that of step 3

for compound 33 (borylation was run at RT overnight and used 24

(B = C = H, A = X = F, V = Br), followed by a TFA-mediated 2′/3′

acetonide deprotection similar to that shown in step 2 to aﬀord 50

1

(56%; two steps). H NMR (500 MHz, DMSO-d₆) δ 8.09 (s, 1H),

3931

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

7.79 (d, J = 11.4 Hz, 1H), 7.45 (d, J = 3.7 Hz, 1H), 7.13 (s, 1H), 7.06

(s, 2H), 6.98−6.90 (m, 3H), 6.68 (d, J = 3.6 Hz, 1H), 5.90 (d, J = 8.1

Hz, 1H), 5.26 (d, J = 7.1 Hz, 1H), 5.07 (s, 1H), 4.14 (t, J = 7.6 Hz,

1H), 3.97 (d, J = 5.7 Hz, 1H), 2.85−2.78 (m, 1H), 2.65−2.59 (m,

1H), 2.30−2.21 (m, 1H), 1.97−1.91 (m, 1H), 1.77−1.66 (m, 2H),

1.58−1.50 (m, 1H). HRMS (ESI+) m/z [M + H]⁺calculated for

C₂₃H₂₂F₂N₆O₃: 469.1799; found: 469.1809.

material (1000 g) was then subjected to chiral prep-SFC (Chiralcel

OJ column, 25% MeOH mobile phase, concentrated in a timely

manner at 30 °C to prevent any decomposition) to aﬀord (6S,6aS)-6-

((triphenylsilyl)oxy)-4,5,6,6a-tetrahydropentalen-2(1H)-one (56,

peak 1, 380 g, 38% recovery oﬀ the chiral column) as a light purple

solid. ESI MS m/z = 397 [M + H]⁺. ¹H NMR (600 MHz, CDCl₃) δ

7.65−7.61 (m, 6H), 7.47−7.43 (m, 3H), 7.42−7.37 (m, 6H), 5.80−

5.78 (m, 1H), 3.93 (q, J = 8.8 Hz, 1H), 3.18−3.12 (m, 1H), 2.83−

2.74 (m, 1H), 2.49−2.41 (m, 1H), 2.37 (dd, J = 18.1, 6.2 Hz, 1H),

2.21−2.10 (m, 2H), 1.73 (dd, J = 18.1, 3.1 Hz, 1H). [α]²_D⁷= −131.2°

(c = 0.99 g/100 mL in MeOH). See the Supporting Information

(Figure S3) for details on absolute structure determination of the

undesired R,R-trans isomer, which therefore enabled us to determine

which peak was the desired S,S-trans isomer.

Step 5: Synthesis of (2R,6S,6aS)-6-((Triphenylsilyl)oxy)-

1,2,4,5,6,6a-hexahydropentalen-2-ol (57). A solution of (6S,6aS)-

6-((triphenylsilyl)oxy)-4,5,6,6a-tetrahydropentalen-2(1H)-one (56,

236 g, 595 mmol, 1.0 equiv) and CeCl₃·7H₂O (222 g, 595 mmol,

1.0 equiv) in EtOH (5 L) and THF (3.5 L) was cooled down to −30

to −35 °C. Then, NaBH₄(24.7 g, 655 mmol, 1.1 equiv) was added in

portions. The reaction was stirred for 1 h at −30 to −35 °C. The

reaction was quenched by pouring into saturated aqueous NH₄Cl (10

L). The mixture was extracted with DCM (5 and 2 L), and the

combined organic phase was concentrated under reduced pressure to

aﬀord crude (2R,6S,6aS)-6-((triphenylsilyl)oxy)-1,2,4,5,6,6a-hexahy-

dropentalen-2-ol (57, 222 g) as a white solid. This material was used

directly in the next reaction. ESI MS m/z = 421 [M + Na]⁺.

(2R,3R,3aS,6S,6aR)-6-((2-Amino-3,6-diﬂuoroquinolin-7-yl)-

methyl)-2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydro-

3aH-cyclopenta[b]furan-3,3a-diol (51). Compound 51 was similarly

synthesized according to the general procedure described for

compound 50 using the appropriately substituted 24.

Step 7: 38%. ESI MS m/z = 509 [M + H]⁺; step 8 (using HCl in

1

MeOH): 38%. H NMR (500 MHz, DMSO-d₆) δ 8.09 (s, 1H), 7.76

(d, J = 11.8 Hz, 1H), 7.44 (d, J = 3.7 Hz, 1H), 7.41−7.35 (m, 2H),

7.06 (s, 2H), 6.76−6.64 (m, 3H), 5.89 (d, J = 8.1 Hz, 1H), 5.25 (d, J

= 7.1 Hz, 1H), 5.09 (s, 1H), 4.15 (t, J = 7.6 Hz, 1H), 3.97 (d, J = 5.7

Hz, 1H), 2.82−2.77 (m, 1H), 2.73−2.68 (dd, J = 13.9, 7.3 Hz, 1H),

2.34−2.24 (m, 1H), 1.99−1.93 (m, 1H), 1.78−1.66 (m, 2H), 1.58−

1.51 (m, 1H). HRMS (ESI+) m/z [M + H]⁺calculated for

C₂₃H₂₂F₂N₆O₃: 469.1799; found: 469.1812.

All-Carbon 5,5-Bicyclic Inhibitor Procedures. Representative

Synthetic Procedure for (1S,2R,3aR,4S,6aR)-4-((2-Amino-3,5-di-

ﬂuoroquinolin-7-yl)methyl)-2-(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydropentalene-1,6a(1H)-diol (72). Step 1:

Synthesis of Pent-4-ynal (53). To a solution of pent-4-yn-1-ol (52,

1.0 kg, 11.9 mol, 1.0 equiv) in DME (10 L) was added IBX (3.99 kg,

14.2 mol, 1.2 equiv). The reaction was stirred at 80 °C for 5 h. Four

reactions conducted in parallel were combined, and the resulting

mixture was then ﬁltered. The ﬁltrate (44 L) containing crude pent-4-

ynal (53) was used directly in the next reaction. Note that on a

smaller scale, this reaction was also typically run with DMP in DCM

at RT.

Step 2: Synthesis of Hept-1-en-6-yn-3-ol (54). A solution of pent-

4-ynal (53, 900 g, 10.9 mol, 1.0 equiv) in DME (11 L) was added to

vinylmagnesium bromide (16.4 L, 1 M in THF, 16.4 mol, 1.5 equiv)

at 0−5 °C. The resulting solution was stirred for 1 h at 0−5 °C. Four

reactions run in parallel were combined, and the mixture was poured

into saturated aqueous NH₄Cl (20 L) and then diluted with DCM (8

L). The organic layer was separated, and the aqueous layer was re-

extracted with DCM (4 L). The combined organic phase was washed

with brine (10 L), dried over anhydrous Na₂SO₄, ﬁltered, and

concentrated under reduced pressure to aﬀord crude hept-1-en-6-yn-

3-ol (54, 4.0 kg) as a yellow oil, which was used directly in the next

reaction.

Step 3: Synthesis of (Hept-1-en-6-yn-3-yloxy)triphenylsilane

(55). To a solution of hept-1-en-6-yn-3-ol (54, 1.0 kg, 9.08 mol, 1.0

equiv), pyridine (1.44 kg, 1.47 L, 18.1 mol, 2.0 equiv), and DMAP

(1.33 kg, 10.8 mol, 1.2 equiv) in DCM (10 L) was added

chlorotriphenylsilane (2.94 kg, 9.99 mol, 1.1 equiv) at 0−5 °C. The

reaction was stirred at RT for 2 h. Four reactions run in parallel were

combined, and the mixture was concentrated under reduced pressure.

The resulting crude residue was puriﬁed by silica gel column

chromatography (50:1 to 10:1 petroleum ether/EtOAc) to aﬀord

(hept-1-en-6-yn-3-yloxy)triphenylsilane (55, 9.0 kg, 51% over three

steps) as a yellow oil. ¹H NMR (600 MHz, CDCl₃) δ 7.65−7.61 (m,

6H), 7.45−7.41 (m, 3H), 7.39−7.35 (m, 6H), 5.85−5.78 (m, 1H),

5.06−4.99 (m, 2H), 4.42 (q, J = 6.2 Hz, 1H), 2.27−2.15 (m, 2H),

1.87−1.78 (m, 2H), 1.77−1.70 (m, 1H).

Step 6: Synthesis of (2R,6S,6aS)-6-((Triphenylsilyl)oxy)-

1,2,4,5,6,6a-hexahydropentalen-2-yl acetate (58). To a solution of

(2R,6S,6aS)-6-((triphenylsilyl)oxy)-1,2,4,5,6,6a-hexahydropentalen-2-

ol (57, 222 g crude, 557 mmol, 1.0 equiv) in DCM (2 L) were added

DMAP (136 g, 1.1 mol, 2.0 equiv) and pyridine (132 g, 1.7 mol, 3.0

equiv) at RT. The solution was then cooled down to 0−5 °C, and

Ac₂O (114 g, 1.11 mol, 2.0 equiv) was added dropwise. The reaction

was stirred at RT for 2 h. The reaction was quenched by pouring into

saturated aqueous NH₄Cl (2 L). The mixture was extracted with

DCM (500 and 200 mL). The combined organic phase was

concentrated under reduced pressure. The crude product was puriﬁed

by silica gel column chromatography (1:0 to 5:1 petroleum ether/

EtOAc) to aﬀord (2R,6S,6aS)-6-((triphenylsilyl)oxy)-1,2,4,5,6,6a-

hexahydropentalen-2-yl acetate (58, 231 g, 88% over two steps) as

1

a colorless oil. H NMR (600 MHz, CDCl₃) δ 7.66−7.62 (m, 6H),

7.46−7.42 (m, 3H), 7.41−7.37 (m, 6H), 5.82−5.77 (m, 1H), 5.24−

5.20 (m, 1H), 3.93 (q, J = 8.0 Hz, 1H), 2.98−2.91 (m, 1H), 2.49−

2.43 (m, 1H), 2.40−2.32 (m, 1H), 2.20−2.12 (m, 1H), 2.09−2.04

(m, 2H), 1.99 (s, 3H), 1.14−1.08 (m, 1H).

Step 7: Synthesis of 4-Chloro-5-ﬂuoro-7-((2R,6S,6aS)-6-

((triphenylsilyl)oxy)-1,2,4,5,6,6a-hexahydropentalen-2-yl)-7H-

pyrrolo[2,3-d]pyrimidine (59, Y = F, W = Cl, Z = H). To a solution of

4-chloro-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidine (40.8 g, 238 mmol,

1.5 equiv), dppf (52.8 g, 95.3 mmol, 0.6 equiv), KOtBu (26.7 g, 238

mmol, 1.5 equiv), and allylpalladium(II) chloride dimer (2.91 g, 7.95

mmol, 0.05 equiv) in THF (1.4 L) at 10−20 °C was added a solution

of (2R,6S,6aS)-6-((triphenylsilyl)oxy)-1,2,4,5,6,6a-hexahydropenta-

len-2-yl acetate (58, 70 g, 158 mmol, 1.0 equiv) in THF (700 mL)

under N₂. The reaction was heated at 40 °C for 3 h under N₂. Four

reactions that were run in parallel were combined, and the mixture

was poured into water (2 L) and diluted with EtOAc (3 L). The

organic layer was separated, and the aqueous phase was re-extracted

with EtOAc (1 L). The combined organic phase was washed with

brine (3 L), dried over anhydrous Na₂SO₄, ﬁltered, and concentrated

under reduced pressure. The residue was puriﬁed by silica gel column

chromatography (1:0 to 1:1 petroleum ether/EtOAc) to aﬀord 4-

chloro-5-ﬂuoro-7-((2R,6S,6aS)-6-((triphenylsilyl)oxy)-1,2,4,5,6,6a-

hexahydropentalen-2-yl)-7H-pyrrolo[2,3-d]pyrimidine (59 (Y = F, W

= Cl, Z = H), 230 g, 66%) as a yellow solid. ESI MS m/z = 552 [M +

Step 4: Synthesis of (6S,6aS)-6-((Triphenylsilyl)oxy)-4,5,6,6a-

tetrahydropentalen-2(1H)-one (56). A solution of (hept-1-en-6-yn-

3-yloxy)triphenylsilane (55, 900 g, 2.4 mol, 1.0 equiv) and dicobalt

octacarbonyl (900 g, 2.6 mol, 1.08 equiv) in MeCN (9 L) was stirred

at 80 °C for 3 h. Ten reactions were run in parallel and the crude

mixtures were combined and concentrated under reduced pressure.

The crude material was puriﬁed by silica gel column chromatography

(1:0 to 5:1 petroleum ether/EtOAc) to aﬀord trans-6-

[(triphenylsilyl)oxy]-1,2,4,5,6,6a-hexahydropentalen-2-one (1700 g,

18%) as a purple solid. (Note that on a smaller scale (4.7 g of 55)

with a similar procedure, separate trans and cis isomer peaks were

isolated in a 1:1 ratio for a combined yield of 57%). The racemic trans

1

H]⁺. H NMR (600 MHz, CDCl₃) δ 8.59 (s, 1H), 7.66−7.63 (m,

6H), 7.44−7.41 (m, 3H), 7.40−7.37 (m, 6H), 6.78 (d, J = 2.4 Hz,

1H), 6.18 (s, 1H), 5.18 (s, 1H), 3.98 (q, J = 8.1 Hz, 1H), 3.16−3.12

3932

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

(m, 1H), 2.62−2.54 (m, 1H), 2.51−2.44 (m, 1H), 2.32−2.23 (m,

1H), 2.23−2.15 (m, 2H), 1.12−1.05 (m, 1H).

[1,6a-d][1,3]dioxol-6-one (20 g, 55 mmol, 1.0 equiv) was added at

approximately −45 to −50 °C. The reaction was then stirred at 0−10

°C for 1 h. Three reactions run in parallel were combined, and the

mixture was poured into saturated NH₄Cl (1 L), cooled down in an

ice bath, and diluted with EtOAc (800 mL). The organic layer was

separated, and the aqueous phase was re-extracted with EtOAc (400

mL). The combined organic phase was washed with brine (1 L), dried

over anhydrous Na₂SO₄, ﬁltered, and concentrated under reduced

pressure. The residue was puriﬁed by silica gel column chromatog-

raphy (50:1 to 10:1 petroleum ether/EtOAc) to aﬀord 4-chloro-7-

((3aS,4R,5aR,8aR)-2,2-dimethyl-6-methylenehexahydro-5H-

pentaleno[1,6a-d][1,3]dioxol-4-yl)-5-ﬂuoro-7H-pyrrolo[2,3-d]-

pyrimidine (62 (Y = F, W = Cl, Z = H), 27 g, 45%) as a light yellow

solid. ESI MS m/z = 364 [M + H]⁺.

Step 12: Synthesis of 7-(((3aS,4R,5aR,6S,8aR)-4-(4-Chloro-5-

ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-

5H-pentaleno[1,6a-d][1,3]dioxol-6-yl)methyl)-3,5-diﬂuoroquinolin-

2-amine. To a solution of 4-chloro-7-((3aS,4R,5aR,8aR)-2,2-

dimethyl-6-methylenehexahydro-5H-pentaleno[1,6a-d][1,3]dioxol-4-

yl)-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidine (62 (Y = F, W = Cl, Z =

H), 9.0 g, 25 mmol, 1.0 equiv) in THF (90 mL) was added 9-BBN

(148 mL, 0.5 M, 74 mmol, 3.0 equiv) at 10−20 °C. The reaction was

stirred at 10−20 °C for 12 h under argon. Then, K₃PO₄(124 mL, 1

M aqueous, 124 mmol, 5.0 equiv) was added, and the mixture was

stirred at 10−20 °C for 30 min. Then, t-BuOK (3.46 mL, 1 M, 3.46

mmol, 0.14 equiv), cataCXium A-Pd-G2 (1.82 g, 2.72 mmol, 0.11

equiv), and 7-bromo-3,5-diﬂuoroquinolin-2-amine (63 (X = A = F, V

= Br), 7.69 g, 29.7 mmol, 1.2 equiv) were added. The reaction was

then stirred for 4 h at 50 °C under argon. Three reactions run in

parallel were combined, and the mixture was concentrated under

reduced pressure. The residue was puriﬁed by silica gel column

chromatography (20:1 to 1:2 petroleum ether/EtOAc) to aﬀord 7-

(((3aS,4R,5aR,6S,8aR)-4-(4-chloro-5-ﬂuoro-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)-2,2-dimethylhexahydro-5H-pentaleno[1,6a-d][1,3]-

dioxol-6-yl)methyl)-3,5-diﬂuoroquinolin-2-amine (21 g, 52%) as a

yellow solid. ESI MS m/z = 544 [M + H]⁺.

Step 13: Synthesis of 7-(((3aS,4R,5aR,6S,8aR)-4-(4-Amino-5-

ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-

5H-pentaleno[1,6a-d][1,3]dioxol-6-yl)methyl)-3,5-diﬂuoroquinolin-

2-amine. A mixture of 7-(((3aS,4R,5aR,6S,8aR)-4-(4-chloro-5-ﬂuoro-

7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-5H-

pentaleno[1,6a-d][1,3]dioxol-6-yl)methyl)-3,5-diﬂuoroquinolin-2-

amine (3.0 g, 5.5 mmol, 1.0 equiv) in NH₃/MeOH (30 mL, 7 M, 210

mmol, 38 equiv) was heated at 120 °C for 3 h in a microwave reactor.

Seven reactions run in parallel were combined, and the mixture was

concentrated under reduced pressure. The crude product was puriﬁed

by silica gel column chromatography (20:1 to 1:2 petroleum ether/

EtOAc) and SFC (Chiralcel OD-3 column, 10−40% MeOH (with

0.05% isopropylamine) mobile phase) to aﬀord 7-(((3aS,4R,5aR,6-

S,8aR)-4-(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-di-

methylhexahydro-5H-pentaleno[1,6a-d][1,3]dioxol-6-yl)methyl)-3,5-

diﬂuoroquinolin-2-amine (12.5 g, 60%) as a yellow solid. Note that on

a small scale, the crude product after removing the solvent was

typically taken directly to the ﬁnal deprotection step and then

puriﬁed. ESI MS m/z = 525 [M + H]⁺. ¹H NMR (400 MHz, CDCl₃)

δ 8.25 (s, 1H), 7.76−7.73 (m, 1H), 7.23 (s, 1H), 6.85−6.83 (m, 1H),

6.75 (s, 1H), 5.38−5.34 (m, 2H), 5.16 (s, 2H), 4.83−4.76 (m, 1H),

4.70−4.67 (m, 1H), 2.76−2.71 (m, 2H), 2.69−2.48 (m, 2H), 2.29−

2.15 (m, 3H), 1.98−1.92 (m, 2H), 1.70−1.55 (m, 1H), 1.52 (s, 3H),

1.33 (s, 3H).

Step 14: Synthesis of (1S,2R,3aR,4S,6aR)-4-((2-Amino-3,5-di-

ﬂuoroquinolin-7-yl)methyl)-2-(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)hexahydropentalene-1,6a(1H)-diol (72). To a sol-

ution of 7-(((3aS,4R,5aR,6S,8aR)-4-(4-amino-5-ﬂuoro-7H-pyrrolo-

[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-5H-pentaleno[1,6a-d]-

[1,3]dioxol-6-yl)methyl)-3,5-diﬂuoroquinolin-2-amine (42.1 mg, 0.08

mmol, 1.0 equiv) in DCM (1.22 mL) and water (0.385 mL) was

added TFA (0.928 mL, 12.0 mmol, 150 equiv). The reaction was

stirred at RT for 4 h. The mixture was then concentrated under

reduced pressure, and the crude material was puriﬁed by reverse-

phase column chromatography (MeCN/H₂O with 0.1% NH₄OH

Step 8: Synthesis of (1S,2R,3aR,4S,6aR)-2-(4-Chloro-5-ﬂuoro-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)-4-((triphenylsilyl)oxy)-

hexahydropentalene-1,6a(1H)-diol (60, Y = F, W = Cl, Z = H). To a

solution of 4-chloro-5-ﬂuoro-7-((2R,6S,6aS)-6-((triphenylsilyl)oxy)-

1,2,4,5,6,6a-hexahydropentalen-2-yl)-7H-pyrrolo[2,3-d]pyrimidine

(59 (Y = F, W = Cl, Z = H), 57.5 g, 104 mmol, 1.0 equiv) in THF

(1.1 L) and H₂O (550 mL) were added OsO₄(2.65 g, 10.4 mmol,

0.54 mL, 0.1 equiv) and NMO (28.1 g, 208 mmol, 2.0 equiv) at 0−5

°C. The reaction was stirred at 0 °C for 3 h. Four reactions run in

parallel were combined, and the mixture was poured into saturated

aqueous Na₂SO₃(1 L) and then diluted with EtOAc (1.5 L). The

organic layer was separated, and the aqueous phase was extracted with

EtOAc (800 mL). The combined organic phase was concentrated

under reduced pressure to aﬀord crude (1S,2R,3aR,4S,6aR)-2-(4-

chloro-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-4-((triphenylsilyl)-

oxy)hexahydropentalene-1,6a(1H)-diol (60 (Y = F, W = Cl, Z = H),

220 g crude) as a black oil, which was used directly in the next

reaction. ESI MS m/z = 586 [M + H]⁺.

Step 9: Synthesis of (3aS,4R,5aR,6S,8aR)-4-(4-Chloro-5-ﬂuoro-

7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-5H-

pentaleno[1,6a-d][1,3]dioxol-6-ol (61, Y = F, W = Cl, Z = H). To a

solution of (1S,2R,3aR,4S,6aR)-2-(4-chloro-5-ﬂuoro-7H-pyrrolo[2,3-

d]pyrimidin-7-yl)-4-((triphenylsilyl)oxy)hexahydropentalene-1,6a-

(1H)-diol (60 (Y = F, W = Cl, Z = H), 55 g, 94 mmol, 1.0 equiv) in

acetone (800 mL) at 10−20 °C was added H₂SO₄(9.20 g, 5.0 mL, 94

mmol, 1.0 equiv). The reaction was stirred at 20 °C for 4 h. Four

reactions run in parallel were combined, and the mixture was poured

into saturated aqueous Na₂CO₃(1 L) and diluted with EtOAc (1 L).

The organic phase was separated, and the aqueous phase was re-

extracted with EtOAc (500 mL). The combined organic phase was

concentrated under reduced pressure. The residue was puriﬁed by

silica gel column chromatography (10:1 to 1:2 petroleum ether/

EtOAc) to aﬀord (3aS,4R,5aR,6S,8aR)-4-(4-chloro-5-ﬂuoro-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-5H-pentaleno-

[1,6a-d][1,3]dioxol-6-ol (61 (Y = F, W = Cl, Z = H), 80 g, 52% over

two steps) as a yellow oil. Note that on a smaller scale, this reaction

was typically run with p-TsOH and 2,2-dimethoxypropane in DCM.

ESI MS m/z = 368 [M + H]⁺. ¹H NMR (600 MHz, CDCl₃) δ 8.62 (s,

1H), 7.08 (d, J = 2.6 Hz, 1H), 4.93−4.87 (m, 1H), 4.72 (d, J = 7.2

Hz, 1H), 4.13−4.10 (m, 1H), 2.61−2.54 (m, 2H), 2.43−2.36 (m,

1H), 2.27−2.22 (m, 1H), 2.15−1.95 (m, 3H), 1.61 (s, 3H), 1.40 (s,

3H). Integration accounts for 18 of the 19 protons as the

exchangeable OH proton is not observed.

Step 10: Synthesis of (3aS,4R,5aS,8aR)-4-(4-Chloro-5-ﬂuoro-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-6H-

pentaleno[1,6a-d][1,3]dioxol-6-one. A solution of (3aS,4R,5aR,6-

S,8aR)-4-(4-chloro-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-di-

methylhexahydro-5H-pentaleno[1,6a-d][1,3]dioxol-6-ol (61 (Y = F,

W = Cl, Z = H), 20 g, 54 mmol, 1.0 equiv) and IBX (28.2 g, 101

mmol, 1.5 equiv) in MeCN (400 mL) was stirred at 80 °C for 5 h.

Four reactions run in parallel were combined, and the mixture was

ﬁltered. The ﬁltrate was concentrated under reduced pressure. The

residue was puriﬁed by silica gel column chromatography (1:0 to 1:1

petroleum ether/EtOAc) to aﬀord (3aS,4R,5aS,8aR)-4-(4-chloro-5-

ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-6H-

pentaleno[1,6a-d][1,3]dioxol-6-one (60 g, 75%) as a light yellow

solid. Note that DMP at RT was typically employed on a smaller

scale. ESI MS m/z = 366 [M + H]⁺. ¹H NMR (400 MHz, CDCl₃) δ

8.61 (s, 1H), 6.99−6.98 (m, 1H), 5.00−4.96 (m, 1H), 4.95−4.90 (m,

1H), 2.88−2.75 (m, 1H), 2.75−2.72 (m, 2H), 2.56−2.44 (m, 4H),

1.57 (s, 3H), 1.41 (s, 3H).

Step 11: Synthesis of 4-Chloro-7-((3aS,4R,5aR,8aR)-2,2-dimeth-

yl-6-methylenehexahydro-5H-pentaleno[1,6a-d][1,3]dioxol-4-yl)-5-

ﬂuoro-7H-pyrrolo[2,3-d]pyrimidine (62, Y = F, W = Cl, Z = H). To a

solution of methyltriphenylphosphonium bromide (64.4 g, 180 mmol,

3.3 equiv) in THF (200 mL) was added nBuLi (54.6 mL, 2.5 M, 136

mmol, 2.5 equiv) dropwise at 0−5 °C. The reaction was then stirred

at 30 °C for 3 h. Then, (3aS,4R,5aS,8aR)-4-(4-chloro-5-ﬂuoro-7H-

pyrrolo[2,3-d]pyrimidin-7-yl)-2,2-dimethylhexahydro-6H-pentaleno-

3933

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

modiﬁer) to aﬀord (1S,2R,3aR,4S,6aR)-4-((2-amino-3,5-diﬂuoroqui-

nolin-7-yl)methyl)-2-(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-

7-yl)hexahydropentalene-1,6a(1H)-diol (72, 27.3 mg, 70% over two

(ESI+) m/z [M + H]⁺calculated for C₂₅H₂₆ClN₅O₂: 464.1853;

found: 464.1855.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydropentalene-

1,6a(1H)-diol (68). Compound 68 was similarly synthesized

according to the general procedure described for compound 72

using the appropriately substituted 19 and 63 (see 66 for the

synthesis of exo-oleﬁn).

1

steps) as a white solid. H NMR (600 MHz, DMSO-d₆) δ 8.30 (s,

1H), 7.99 (d, J = 11.1 Hz, 1H), 7.88 (s, 1H), 7.20 (s, 1H), 7.06 (d, J =

10.9 Hz, 1H), 4.84−4.78 (m, 1H), 3.76 (d, J = 10.3 Hz, 1H), 2.79−

2.66 (m, 2H), 2.48−2.37 (m, 1H), 2.20−2.13 (m, 1H), 1.87−1.80

(m, 1H), 1.79−1.70 (m, 2H), 1.69−1.64 (m, 1H), 1.64−1.55 (m,

1H), 1.50−1.41 (m, 1H). Integration accounts for 17 of the 23

protons as exchangeable protons are not considered. HRMS (ESI+)

m/z [M + H]⁺calculated for C₂₄H₂₃F₃N₆O₂: 485.1913; found:

485.1911.

Step 12: 94% (63, V = Br, RuPhos Pd G3 catalyst), ESI MS m/z =

1

488 [M + H]⁺; step 13 (deprotection): 40%. H NMR (600 MHz,

DMSO-d₆) δ 8.60 (s, 1H), 7.82 (d, J = 3.6 Hz, 1H), 7.75 (d, J = 11.8

Hz, 1H), 7.55 (d, J = 8.2 Hz, 1H), 7.29 (s, 1H), 7.10 (d, J = 8.6 Hz,

1H), 6.71 (d, J = 3.6 Hz, 1H), 6.67 (s, 2H), 4.83−4.74 (m, 2H), 4.67

(s, 1H), 3.95 (dd, J = 10.2, 7.4 Hz, 1H), 2.76−2.66 (m, 2H), 2.65 (s,

3H), 2.48−2.41 (m, 1H), 2.18 (q, J = 9.4 Hz, 1H), 1.88−1.77 (m,

3H), 1.71−1.64 (m, 1H), 1.64−1.55 (m, 1H), 1.50−1.42 (m, 1H).

HRMS (ESI+) m/z [M + H]⁺calculated for C₂₅H₂₆FN₅O₂: 448.2149;

found: 448.2151.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-amino-5-ﬂuoro-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydropentalene-1,6a(1H)-diol (69). Compound 69 was sim-

ilarly synthesized according to the general procedure described for

compound 72 using the appropriately substituted 19 and 63.

Step 7: 44%, ESI MS m/z = 552 [M + H]⁺; step 8: ND (used

crude), ESI MS m/z = 586 [M + H]⁺; step 9: 78% (two steps), ESI

MS m/z = 368 [M + H]⁺; step 10: 98%, ESI MS m/z = 366 [M +

H]⁺; step 11: 27%, ESI MS m/z = 364 [M + H]⁺; step 12: 56% (63, V

= Br, RuPhos Pd G3 catalyst), ESI MS m/z = 526 [M + H]⁺; step 13:

ND (used crude), ESI MS m/z = 507 [M + H]⁺; step 14: 75% (two

steps). ¹H NMR (600 MHz, DMSO-d₆) δ 8.86−8.52 (m, 4H), 8.32−

8.24 (m, 2H), 7.88 (s, 1H), 7.77 (d, J = 8.2 Hz, 1H), 7.48 (s, 1H),

7.36 (d, J = 8.2 Hz, 1H), 4.85−4.78 (m, 1H), 3.76 (d, J = 10.3 Hz,

1H), 2.84−2.77 (m, 1H), 2.77−2.71 (m, 1H), 2.49−2.35 (m, 2H),

2.18−2.11 (m, 1H), 2.04−1.90 (m, 1H), 1.87−1.81 (m, 1H), 1.78−

1.72 (m, 2H), 1.71−1.66 (m, 1H), 1.65−1.57 (m, 1H), 1.48−1.41

(m, 1H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₄H₂₄F₂N₆O₂:

467.2007; found: 467.2020.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-amino-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydropentalene-

1,6a(1H)-diol (65). Compound 65 was similarly synthesized

according to the general procedure described for compound 72

using the appropriately substituted 19 and 63.

Step 7: 65%, ESI MS m/z = 534 [M + H]⁺; step 8: ND (used

crude), ESI MS m/z = 568 [M + H]⁺; step 9: 62% (two steps), ESI

MS m/z = 350 [M + H]⁺; step 10: 90%, ESI MS m/z = 348 [M +

H]⁺; step 11: 49%, ESI MS m/z = 346 [M + H]⁺; step 12: 54% (63, V

= I), ESI MS m/z = 568/570 [M + H]⁺; step 13: ND (used crude),

1

ESI MS m/z = 549/551 [M + H]⁺; step 14: 73% (two steps). H

NMR (600 MHz, DMSO-d₆, TFA salt) δ 8.73 (s, 1H), 8.35 (s, 1H),

7.82 (d, J = 3.6 Hz, 1H), 7.75 (d, J = 8.2 Hz, 1H), 7.43 (s, 1H), 7.33

(d, J = 8.3 Hz, 1H), 6.96 (d, J = 3.6 Hz, 1H), 4.77 (q, J = 10.3 Hz,

1H), 3.86 (d, J = 10.3 Hz, 1H), 2.82−2.71 (m, 2H), 2.48−2.43 (m,

1H), 2.17 (q, J = 9.4 Hz, 1H), 1.85 (dd, J = 12.3, 5.9 Hz, 1H), 1.82−

1.76 (m, 2H), 1.72−1.66 (m, 1H), 1.65−1.57 (m, 1H), 1.49−1.42

(m, 1H). Integration accounts for 19 of the 25 protons as

exchangeable protons are not observed. HRMS (ESI+) m/z [M +

H]⁺calculated for C₂₄H₂₅BrN₆O₂: 509.1300; found: 509.1312.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-bromoquinolin-7-yl)methyl)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydropentalene-

1,6a(1H)-diol (66). Compound 66 was similarly synthesized

according to the general procedure described for compound 72

using the appropriately substituted 19 and 63.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-amino-5-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)-

hexahydropentalene-1,6a(1H)-diol (70). Compound 70 was sim-

ilarly synthesized according to the general procedure described for

compound 72 using the appropriately substituted 19 and 63.

Step 7: 76%, ESI MS m/z = 548 [M + H]⁺; step 8: ND, ESI MS m/

z = 582 [M + H]⁺; step 9: 96% (two steps), ESI MS m/z = 364 [M +

H]⁺; step 10: 88%, ESI MS m/z = 362 [M + H]⁺; step 11: 20%, ESI

MS m/z = 360 [M + H]⁺; step 12: 51% (63, V = Br, RuPhos Pd G3

catalyst), ESI MS m/z = 522 [M + H]⁺; step 13: ND (used crude)

Step 7: 53%, ESI MS m/z = 514 [M + H]⁺(note that alternatively,

compounds with a C6-methyl nucleobase could be made by

converting the C6-Cl via a Negishi reaction with dimethyl zinc after

the Wittig homologation); step 8: ND (used crude), ESI MS m/z =

548 [M + H]⁺; step 9: 86% (two steps), ESI MS m/z = 330 [M +

H]⁺; step 10: ND (used crude), ESI MS m/z = 328 [M + H]⁺; step

11: 24% (two steps), ESI MS m/z = 326 [M + H]⁺; step 12: 97% (63,

V = I, racemic material 62 used for this step), ESI MS m/z = 548/550

[M + H]⁺; step 13 (deprotection): 70%. (The racemic material was

then separated by chiral SFC (AS-H column, 30% MeOH (with 0.1%

NH₄OH) mobile phase), and the data for the more active peak was

shown (peak 1).) ESI MS m/z = 508/510 [M + H]⁺. ¹H NMR (600

MHz, DMSO-d₆) δ 8.60 (s, 1H), 8.31 (s, 1H), 7.82 (d, J = 3.6 Hz,

1H), 7.56 (d, J = 8.2 Hz, 1H), 7.28 (s, 1H), 7.10 (dd, J = 8.2, 1.3 Hz,

1H), 6.71 (d, J = 3.6 Hz, 1H), 6.53 (s, 2H), 4.83−4.75 (m, 2H), 4.67

(s, 1H), 3.95 (dd, J = 10.3, 7.5 Hz, 1H), 2.75−2.66 (m, 2H), 2.65 (s,

3H), 2.48−2.42 (m, 1H), 2.18 (q, J = 9.4 Hz, 1H), 1.85 (dd, J = 12.3,

5.9 Hz, 1H), 1.80 (t, J = 9.6 Hz, 2H), 1.70−1.64 (m, 1H), 1.64−1.55

(m, 1H), 1.49−1.43 (m, 1H).

1

ESI MS m/z = 503 [M + H]⁺; step 14: 62% (two steps). H NMR

(500 MHz, DMSO-d₆) δ 8.11 (s, 1H), 7.78 (d, J = 11.8 Hz, 1H), 7.57

(d, J = 8.2 Hz, 1H), 7.36 (s, 1H), 7.29 (d, J = 12.1 Hz, 3H), 7.11 (d, J

= 8.3 Hz, 1H), 6.77 (s, 2H), 5.00−4.57 (m, 3H), 3.82 (d, J = 10.3 Hz,

1H), 2.76−2.62 (m, 2H), 2.49−2.43 (m, 1H), 2.41 (s, 3H), 2.14 (q, J

= 8.9 Hz, 1H), 1.83 (dd, J = 12.1, 5.7 Hz, 1H), 1.76−1.64 (m, 3H),

1.62−1.51 (m, 1H), 1.49−1.41 (m, 1H). HRMS (ESI+) m/z [M +

H]⁺calculated for C₂₅H₂₇FN₆O₂: 463.2258; found: 463.2268.

(1S,2R,3aR,4S,6aR)-2-(4-Amino-2-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-7-yl)-4-((2-amino-3-ﬂuoroquinolin-7-yl)methyl)-

hexahydropentalene-1,6a(1H)-diol (71). Compound 71 was sim-

ilarly synthesized according to the general procedure described for

compound 72 using the appropriately substituted 19 and 63.

(1S,2R,3aR,4S,6aR)-4-((2-Amino-3-chloroquinolin-7-yl)methyl)-

2-(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydropentalene-

1,6a(1H)-diol (67). Compound 67 was similarly synthesized

according to the general procedure described for compound 72

using the appropriately substituted 19 and 63 (see 66 for the

synthesis of exo-oleﬁn).

Step 7: 55%, ESI MS m/z = 548 [M + H]⁺; step 8: ND (used

crude), ESI MS m/z = 582 [M + H]⁺; step 9: 84% (two steps), ESI

MS m/z = 364 [M + H]⁺; step 10: 88%, ESI MS m/z = 362 [M +

H]⁺; step 11: 32% (two steps), ESI MS m/z = 360 [M + H]⁺; step 12:

51% (63, V = Br, RuPhos Pd G3 as a catalyst), ESI MS m/z = 522 [M

+ H]⁺; step 13: ND (used crude), ESI MS m/z = 503 [M + H]⁺; step

14: 38% (two steps). ¹H NMR (600 MHz, DMSO-d₆) δ 7.76 (d, J =

11.8 Hz, 1H), 7.56 (d, J = 8.2 Hz, 1H), 7.38 (s, 1H), 7.30 (s, 1H),

7.11 (d, J = 8.2 Hz, 1H), 6.59 (s, 1H), 4.70−4.65 (m, 1H), 3.87 (d, J

= 9.9 Hz, 1H), 2.77−2.64 (m, 2H), 2.46 (ddd, J = 12.9, 8.2, 5.3 Hz,

1H), 2.39 (s, 3H), 2.14 (q, J = 9.3 Hz, 1H), 1.84 (dd, J = 12.3, 5.9 Hz,

Step 12: 52% (63, V = Br, RuPhos Pd G3 catalyst), ESI MS m/z =

1

504 [M + H]⁺; step 13 (deprotection): 34%. H NMR (600 MHz,

DMSO-d₆) δ 8.60 (s, 1H), 8.13 (s, 1H), 7.82 (d, J = 3.6 Hz, 1H), 7.56

(d, J = 8.2 Hz, 1H), 7.28 (s, 1H), 7.10 (dd, J = 8.2, 1.2 Hz, 1H), 6.71

(d, J = 3.6 Hz, 1H), 6.62 (s, 2H), 4.83−4.75 (m, 2H), 4.67 (s, 1H),

3.95 (dd, J = 10.2, 7.5 Hz, 1H), 2.76−2.66 (m, 2H), 2.65 (s, 3H),

2.49−2.42 (m, 1H), 2.18 (q, J = 9.4 Hz, 1H), 1.88−1.78 (m, 3H),

1.71−1.64 (m, 1H), 1.64−1.55 (m, 1H), 1.50−1.42 (m, 1H). HRMS

3934

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

1H), 1.79−1.64 (m, 2H), 1.57 (qd, J = 12.4, 6.0 Hz, 1H), 1.45 (td, J =

12.5, 6.6 Hz, 1H). Integration accounts for 21 of the 27 protons as

exchangeable protons are not considered. HRMS (ESI+) m/z [M +

H]⁺calculated for C₂₅H₂₇FN₆O₂: 463.2258; found: 463.2251.

(1S,2R,3aR,4R,6aR)-4-((2-Amino-3-ﬂuoroquinolin-7-yl)methyl)-2-

(4-methyl-7H-pyrrolo[2,3-d]pyrimidin-7-yl)hexahydropentalene-

1,6a(1H)-diol (73). Compound 73 was synthesized from the

intermediate formed after step 12 (Suzuki coupling) for the synthesis

of compound 68. Under these Suzuki conditions, a minor

diastereomer product was formed with the (R)-stereochemistry at

the linker position (presumably resulting from incomplete facial

selectivity of the borylation). This product could be cleanly isolated

by chiral SFC (OJ-H column, 30% MeOH (with 0.1% NH₄OH)

modiﬁer) and subjected to TFA deprotection conditions (step 13) to

aﬀord the ﬁnal compound 73.

antibody at 4 °C overnight. The 1° antibody was removed, followed

by three washings with DPBS without Ca²⁺/Mg²⁺and 0.05% Tween

20. Hoechst (5 mg/mL), Cell Mask deep red stain (1:2000), and

Alexa488-conjugated goat anti-rabbit IgG (2 μg/mL) were added for

1 h at room temperature. A ﬁnal washing step (three washes) was

performed before sealing the plate for imaging on the In Cell Analyzer

2200 or Opera Phenix. Images from the analyzer were uploaded to

Columbus (at Merck & Co., Inc., West Point, PA, USA, or Merck &

Co., Inc., Boston, MA, USA) for image analysis. EC₅₀values were

determined by four-parameter robust ﬁt of percent ﬂuorescence units

vs (Log10) compound concentrations.

Z138 Cell Proliferation Assay. Compounds were assessed for their

eﬀect in a 10 day Z138 cell proliferation assay. In this assay, the cell

potency (EC₅₀) of each compound was determined from a 10-point

(1:3 serial dilution; top compound concentration of 10,000 nM)

titration curve using the following outlined procedure. Fifty nanoliters

of compound (or control) in DMSO was dispensed using Echo into

each well of a BD falcon black/clear-bottom 384-well plate, and then

500 Z138 cells were directly seeded in 50 μL of medium (IMDM

media, 10% Horse serum, and 1% 100× antibiotics) using a Hamilton

liquid handler, followed by 1 min mixing on a plate shaker. Cells were

allowed to grow for 10 days (37 °C CO₂incubator) with two changes

of fresh compound-medium (day 3 and day 7) to ensure that the

concentrations of the compound and nutrients are constant

throughout the experiment. For each compound-medium change,

30 μL of the culture medium in each well was replaced by 30 μL of

fresh medium containing the same amount of the drug. On day 10,

the living cell number from each well was estimated using the

luciferase assay (Cell Titer-Glo 2.0). EC₅₀values were determined by

four-parameter robust ﬁt of percent ﬂuorescence units vs (Log10)

compound concentrations.

In Vivo Antitumor Eﬃcacy Model General Materials and

Methods. CB17.SCID female mice (Charles River Laboratories)

were subcutaneously implanted with 5 × 10⁶Z138 cells. When

tumors reached an average size of approximately 100 mm³, mice were

pair-matched into the eﬃcacy study (n = 10 mice per group) and

daily PO dosing began (10, 30, or 60 mpk QD or 5 mpk BID). In

addition to eﬃcacy mice, satellite cohorts of animals were enrolled

into the study for an earlier day 7 collection time point (n = 4 mice

per group). The compound was formulated weekly using 0.5%

methylcellulose (Dow) as a vehicle. Tumors and mouse body weights

were captured twice weekly for the duration of the study and all

animal procedures were performed in accordance with IACUC

standards. Tumor volumes (TVs) were calculated as (L × W²)/2 and

%TGI was calculated at day 13 as [(ΔTV^vehicle− ΔTV^treatment)/

(ΔTV^vehicle)] × 100. P-values were calculated using two-sided Student

t-test.

1

Step 13 (TFA deprotection): 65%. H NMR (600 MHz, DMSO-

d₆) δ 8.60 (s, 1H), 7.75 (d, J = 11.7 Hz, 1H), 7.72 (d, J = 3.6 Hz, 1H),

7.56 (d, J = 8.1 Hz, 1H), 7.33 (s, 1H), 7.11 (d, J = 8.2 Hz, 1H), 6.77

(s, 2H), 6.67 (d, J = 3.6 Hz, 1H), 4.91 (td, J = 11.0, 7.0 Hz, 1H), 4.79

(s, 1H), 4.55 (s, 1H), 4.11 (d, J = 10.3 Hz, 1H), 2.77 (d, J = 7.5 Hz,

2H), 2.63 (s, 3H), 2.07 (q, J = 7.0 Hz, 1H), 1.97 (q, J = 8.0 Hz, 1H),

1.91 (dt, J = 11.8, 8.0 Hz, 1H), 1.85 (h, J = 6.1 Hz, 1H), 1.81−1.73

(m, 2H), 1.61 (dq, J = 15.6, 8.1 Hz, 1H), 1.47 (td, J = 12.0, 8.2 Hz,

1H). HRMS (ESI+) m/z [M + H]⁺calculated for C₂₅H₂₆FN₅O₂:

448.2149; found: 448.2160. The relative stereochemistry of the

compound was conﬁrmed by 2D NMR to enable assignment of the

(R)-isomer at the linker C5 carbon (Figure S8).

High-Throughput HPLC LogD pH 7 Measurements. High-

throughput LogD measurements at pH 7 were conducted from 10

mM DMSO stock solutions in a well plate format using a reverse-

phase HPLC retention time method with UV−vis detection.

High-Throughput Solubility Measurements at pH 2 and pH 7

and in FaSSIF (pH 6.5). High-throughput solubility measurements

were conducted in a well plate format by addition of 10 mM DMSO

stock solutions to 10 mM pH 2 and pH 7 potassium phosphate

buﬀers and FaSSIF (resulting total concentration of 200 μM)

followed by 24 h equilibration, ﬁltration, and reverse-phase HPLC

analysis using UV−vis detection.

In Vitro Pharmacology General Materials and Methods. PRMT5

Cell Target Engagement (TE) Assay. The PRMT5 TE assay is a

biomarker assay for identifying compounds that inhibit symmetric

dimethylation of arginine (SDMA) of PRMT5 substrates. Speciﬁcally,

symmetrically dimethylated nuclear proteins are detected using high-

content imaging technology. Detection of the expression of

symmetrically dimethylated nuclear proteins is through a mixture of

primary rabbit monoclonal antibodies to SDMA (CST 13222), which

in turn are recognized by an AlexaFluor 488 dye-conjugated anti-

rabbit IgG secondary antibody. The IN Cell Analyzer 2200 or Opera

Phenix measures nuclear AlexaFluor 488 ﬂuorescent dye intensity that

is directly related to the level of expression of symmetrically

dimethylated nuclear proteins at the single cell level. Nuclear

AF488 dye intensities are compared to the mean value for DMSO-

treated cells (MIN) to report the percent of inhibition for each

compound-treated well.

In this assay, the cell potency (EC₅₀) of each compound was

determined from a 10-point (1:3 serial dilution; top compound

concentration of 10,000 nM) titration curve using the following

outlined procedure. Each well of a BD falcon collagen-coated black/

clear-bottom 384-well plate was seeded with 4000 MCF-7 cells in 30

μL of media and allowed to attach for 5 h. The medium is ATCC-

formulated Eagle’s minimum essential medium (catalog no. 30-2003).

To make the complete growth medium, the following components

were added to the base medium: 0.01 mg/mL human recombinant

insulin and fetal bovine serum to a ﬁnal concentration of 10%.

Additional 30 μL of media containing 2× compounds was added to

each well. Cells were treated for 3 days in a 37 °C/5% CO₂incubator.

On day 3, cells were ﬁxed with Cytoﬁx, permeabilized with 0.4%

Triton-X-100/Cytoﬁx, and washed with D-PBS without Ca²⁺/Mg²⁺.

Cells were blocked with LI-COR Odyssey blocking reagent for 1 h at

room temperature, followed by incubation with anti-SDMA (1:1000)

The workﬂow to quantify SDMA SMB/B′ in tumors by LC−MS/

MS involved tissue homogenization, aﬃnity enrichment, protein

reduction, alkylation, and digestion, and analyses of representative

peptides by LC−MS/MS.

Frozen tumor samples were cut and placed in Bullet Blender 1.5

mL Navy RNA Lysis kit tubes (Next Advance, Inc., Troy, NY).

Homogenization buﬀer (50 mM Tris (pH 7.5), 200 mM NaCl, 2 mM

EDTA, and 0.0004% RapiGest (Waters Corp, Milford, MA), with

Halt protease inhibitor cocktail) was added to the tumor sample at a

ratio of 15:1 mL/g. Tumor samples were then homogenized in a

Bullet Blender 24 Gold (Next Advance, Inc., Troy, NY) for 10 min at

a speed setting of 12 until no visible tissue pieces were seen. Samples

were then centrifuged at 13,000g for 1 min at 4 °C. Eighty microliters

of homogenate was mixed with 25 μL of an in-house custom-made

anti-SMB/B′ antibody and 145 μL of cell lysis buﬀer (50 mM Tris

(pH 7.5) and 200 mM NH₄Cl) in an Eppendorf LoBind 500 μL 96-

well plate (Eppendorf North America, Hauppauge, NY). The plate

was incubated overnight at 4 °C with gentle shaking.

The SMB/B′/anti-SMB/B′ complex was immunoprecipitated with

Protein A/G magnetic beads (Pierce). Beads (62.5 μL) were prepared

for each tumor homogenate in a 96-well 500 μL microtiter plate

(Eppendorf LoBind, Eppendorf North America, Hauppauge, NY).

Using a plate magnet, the beads were pulled to one side of the well

3935

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

and the storage buﬀer was removed. The beads were then washed

three times with a wash buﬀer (50 mM Tris (pH 7.5), 200 mM NaCl,

2 mM EDTA, and 0.0004% RapiGest), removing the supernatant

after every wash. The supernatant was removed after the third wash

and the homogenate/anti-SMB/B′ antibody mixture (250 μL in total

volume) was added to the beads. The samples were incubated at

room temperature for 1.5 h with shaking to keep the magnetic beads

in the solution, and beads were then pulled to the side of the wells

with a magnet and the supernatant was removed. The beads were

washed three times with the lysis buﬀer and three times with a mass

spectrometry-friendly buﬀer (50 mM Tris (pH 7.5) and 200 mM

NH₄Cl), discarding the supernatant after every wash. Fifty microliters

of 0.1% TFA was added to the wells and the plate was incubated for

10 min at room temperature with shaking to elute the captured

protein from the magnetic beads. The beads were subsequently pulled

to the side of the wells with a magnet and the supernatant containing

the target antigen was collected.

For sample digestion, 30 μL of digestion buﬀer (50 mM

triethylammonium bicarbonate and 0.2% RapiGest (w/v) in water)

was added to each well. To this, 20 μL of 50 mM DTT was added and

the samples were incubated at 60 °C for 45 min with shaking. The

samples were cooled down to room temperature, 3 μL of 500 mM

iodoacetamide (Millipore Sigma, St. Louis, MO) was added, and the

samples were incubated at room temperature in the dark for 30 min.

Finally, 2 μg of trypsin (gold, mass spectrometer grade from Promega,

Madison, WI) was added and the samples were incubated overnight at

37 °C with shaking. To precipitate RapiGest, 1.5 μL of TFA was

added to each sample and incubated at 37 °C for 45 min with shaking.

The plate was then spun down at maximum speed for 15 min at room

temperature.

Synthetic heavy-labeled and unlabeled peptide standards were

obtained from New England Peptide (Gardner MA) for four unique

SMB/B′ peptides. The VLGLVLLR peptide represents the total

amount of SMB/B′. The peptide VPLAGAAGGPGIGR represents

the amount of SMB/B′ that has not been methylated and, if both

arginine residues have been methylated, then the peptide

V P L A G A A G G P G I G R ( d i m e t h y l ) A A G R ( d i m e t h y l ) -

GIPAGVPMPQAPAGLAGPVR represents the amount of modiﬁed

SMB/B′ in the sample. The peptide IFIGTFK is detectable for SMB/

B′ but not related to the methylation site. All four peptides were

combined into a single standard curve ranging from 0.25 to 100 fmol/

μL. Since SMB/B′ is an endogenous protein, the standard curve was

prepared using the unlabeled peptides in water. Each of the heavy-

labeled peptides was combined at a concentration of 200 fmol/μL.

Five microliters of the internal standard was added after digestion

before injection on the mass spectrometer.

Two microliters of the sample was analyzed with a Thermo Q

Exactive HF Hybrid Quadrupole-Orbitrap mass spectrometer in

nanospray mode coupled to a Waters nanoACQUITY LC system

equipped with a vented trap column. The samples were ﬁrst trapped

on a trap column for 10 min at a ﬂow rate of 10 μL/min (Waters

nanoACQUITY UPLC 2G-V/Mtrap 5 μm Symmetry C18 180 μm ×

20 mm) at 99.5% mobile phase A (0.1% formic acid in water) and

then eluted oﬀ the trap column onto an analytical column (PicoChip

Reprosil-Pur C18-AQ 3 μm 120A; 105 mm; 75 μm ID (New

Objective, Woburn, MA)) with a 50 min gradient from 0.5% mobile

phase B (0.1% formic acid in acetonitrile) to 30% B at a ﬂow rate of

0.5 μL/min. After the gradient, the trap and analytical columns were

re-equilibrated for 30 min at the starting condition with 99.5% A and

0.5% B. Data was collected in parallel reaction monitoring (PRM)

mode and analyzed using Skyline (www.skyline.ms).

Supplementary ﬁgures, tables, and plots; description of

modeling methods; crystallography methods; LC traces

for compounds 34 and 72 (PDF)

Representative docking models for compounds 4, 31,

41, 65, and 73 (PDB)

Molecular formula strings (CSV)

AUTHOR INFORMATION

Corresponding Author

■

Ryan V. Quiroz − Merck & Co., Inc., Boston, Massachusetts

02115, United States; orcid.org/0000-0001-6840-9636;

Email: ryan.quiroz@merck.com

Authors

Michael H. Reutershan − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Sebastian E. Schneider − Merck & Co., Inc., Boston,

Massachusetts 02115, United States; orcid.org/0000-

0003-1233-8984

David Sloman − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Brian M. Lacey − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Brooke M. Swalm − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Charles S. Yeung − Merck & Co., Inc., Boston, Massachusetts

02115, United States; orcid.org/0000-0001-7320-7190

Craig Gibeau − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Daniel S. Spellman − Merck & Co., Inc., Lansdale,

Pennsylvania 19446, United States; orcid.org/0000-

0001-7155-2329

Danica A. Rankic − Merck & Co., Inc., Boston, Massachusetts

02115, United States; orcid.org/0000-0002-7326-8900

Dapeng Chen − Merck & Co., Inc., Boston, Massachusetts

02115, United States

David Witter − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Doug Linn − Merck & Co., Inc., Boston, Massachusetts 02115,

United States

Erik Munsell − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Guo Feng − Merck & Co., Inc., Boston, Massachusetts 02115,

United States

Haiyan Xu − Merck & Co., Inc., Boston, Massachusetts 02115,

United States

Jonathan M. E. Hughes − Merck & Co., Inc., Rahway, New

Jersey 07065, United States

Jongwon Lim − Merck & Co., Inc., Boston, Massachusetts

02115, United States; orcid.org/0000-0003-3402-7810

Josep Saurí − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Kristin Geddes − Merck & Co., Inc., Lansdale, Pennsylvania

19446, United States

Murray Wan − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Unless speciﬁed, all reagents were purchased from Thermo Fisher

Scientiﬁc, Waltham, MA.

My Sam Mansueto − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Nicole E. Follmer − Merck & Co., Inc., Boston, Massachusetts

02115, United States

ASSOCIATED CONTENT

■

sı

* Supporting Information

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c02083.

Patrick S. Fier − Merck & Co., Inc., Rahway, New Jersey

07065, United States; orcid.org/0000-0002-6102-815X

3936

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Phieng Siliphaivanh − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Pierre Daublain − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Rachel L. Palte − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Robert P. Hayes − Merck & Co., Inc., Lansdale, Pennsylvania

19446, United States

Sandra Lee − Merck & Co., Inc., Rahway, New Jersey 07065,

United States; orcid.org/0000-0003-1907-0263

Shuhei Kawamura − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Steven Silverman − Merck & Co., Inc., Rahway, New Jersey

07065, United States; orcid.org/0000-0003-0792-7071

Sulagna Sanyal − Merck & Co., Inc., Boston, Massachusetts

02115, United States

Timothy J. Henderson − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Yingchun Ye − Merck & Co., Inc., Boston, Massachusetts

02115, United States

cytochrome P450; DCM, dichloromethane; DDI, drug−drug

interaction; DMAP, N,N-dimethylaminopyridine; DME, dime-

thoxyethane; DMP, Dess−Martin periodinane; DMSO,

dimethyl sulfoxide; dppe, 1,2-bis(diphenylphosphino)ethane;

dppf, 1,1′-bis(diphenylphosphino)ferrocene; EC₅₀, eﬀective

concentration for 50% response; ECD, electronic circular

dichroism; EtOAc, ethyl acetate; %F, bioavailability; F_a,

fraction absorbed; F_h, fraction escaping hepatic metabolism

and excretion; H-bond, hydrogen bond; HBD, hydrogen bond

donor; IBX, 2-iodoxybenzoic acid; MeCN, acetonitrile; MoA,

mechanism of action; mpk, mg/kg; MRT, mean residence

time; MTA, methylthioadenosine; MTAP, methylthioadeno-

sine phosphorylase; nbd, norbornadiene; NMO, N-methyl-

morpholine N-oxide; NMP, N-methylpyrrolidine; NMR,

nuclear magnetic resonance; PD, pharmacodynamics; PK,

pharmacokinetics; PMB, 4-methoxybenzyl; POS, probability of

success; PRMT5, protein arginase methyl transferase 5; PSA,

polar surface area; QD, once a day; RT, room temperature;

SAM, S-adenosylmethionine; SAR, structure−activity relation-

ship; SDMA, symmetric dimethylarginine; SFC, supercritical

ﬂuid chromatography; SMB/B′, small-molecule ribonucleo-

protein-associated protein B and B′; TDI, time-dependent

inhibition; TE, target engagement; TEA, triethylamine; TFA,

triﬂuoroacetic acid; TGI, tumor growth inhibition; THF,

tetrahydrofuran; TLC, thin-layer chromatography; TPS,

triphenylsilyl; TsCl, 4-toluenesulfonyl chloride; TsOH, 4-

toluenesulfonic acid; Vd_u, unbound volume of distribution

Yuanwei Gao − Merck & Co., Inc., Lansdale, Pennsylvania

19446, United States

Benjamin Nicholson − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Michelle R. Machacek − Merck & Co., Inc., Boston,

Massachusetts 02115, United States

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.jmedchem.0c02083

REFERENCES

■

Author Contributions

The manuscript was prepared by R.V.Q. with contributions by

all authors to the research.

(1) Barrueto, L.; Caminero, F.; Cash, L.; Makris, C.; Lamichhane, P.;

Deshmukh, R. R. Resistance to checkpoint inhibition in cancer

immunotherapy. Transl. Oncol. 2020, 13, 100738.

(2) Xiao, W.; Chen, X.; Liu, L.; Shu, Y.; Zhang, M.; Zhong, Y. Role

of protein arginine methyltransferase 5 in human cancers. Biomed.

Pharmacother. 2019, 114, 108790.

Funding

This work was funded entirely by Merck Sharp & Dohme

Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA.

Notes

The authors declare no competing ﬁnancial interest.

PDB IDs of new crystal (X-ray) structures are as follows: 7KIB

(compound 4), 7KIC (compound 34), and 7KID (compound

72). The authors will release the atomic coordinates upon

article publication.

(3) Zhu, F.; Rui, L. PRMT5 in gene regulation and hematologic

malignancies. Genes Dis. 2019, 6, 247−257.

(4) Lin, H.; Luengo, J. I. Nucleoside protein arginine methyl-

transferase 5 (PRMT5) inhibitors. Bioorg. Med. Chem. Lett. 2019, 29,

1264−1269.

(5) Shailesh, H.; Zakaria, Z. Z.; Baiocchi, R.; Sif, S. Protein arginine

methyltransferase 5 (PRMT5) dysregulation in cancer. Oncotarget

2018, 9, 36705−36718.

(6) Wang, Y.; Hu, W.; Yuan, Y. Protein arginine methyltransferase 5

(PRMT5) as an anticancer target and its inhibitor discovery. J. Med.

Chem. 2018, 61, 9429−9441.

ACKNOWLEDGMENTS

■

We would like to thank Leo Joyce and the structural

determination team for solving the absolute structure of the

Pauson−Khand product. We would also like to thank the

Merck & Co., Inc., Boston, MA, USA, puriﬁcation team for

their rapid and eﬃcient work in purifying these complex

structures on small and large scales. We acknowledge external

chemistry members Lianyun Zhao and Wensheng Yu as well as

Zhibo Zhang and her team at Pharmaron for contributions to

compound synthesis. We thank Viva Biotech for help with

crystallography and data collection. We thank Hendrik Falk

and Ian Street from Cancer Therapeutics CRC (CTx) for

enabling the development of the cell-based assays.

̈

(7) Prabhu, L.; Wei, H.; Chen, L.; Demir, O.; Sandusky, G.; Sun, E.;

Wang, J.; Mo, J.; Zeng, L.; Fishel, M.; Safa, A.; Amaro, R.; Korc, M.;

Zhang, Z.-Y.; Lu, T. Adapting AlphaLISA high throughput screen to

discover a novel small-molecule inhibitor targeting protein arginine

methyltransferase 5 in pancreatic and colorectal cancers. Oncotarget

2017, 8, 39963−39977.

(8) Zhu, K.; Jiang, C.; Tao, H.; Liu, J.; Zhang, H.; Luo, C.

Identification of a novel selective small-molecule inhibitor of protein

arginine methyltransferase 5 (PRMT5) by virtual screening, resyn-

thesis and biological evaluations. Bioorg. Med. Chem. Lett. 2018, 28,

1476−1483.

(9) Duncan, K. W.; Rioux, N.; Boriack-Sjodin, P. A.; Munchhof, M.

J.; Reiter, L. A.; Majer, C. R.; Jin, L.; Johnston, L. D.; Chan-Penebre,

E.; Kuplast, K. G.; Porter Scott, M.; Pollock, R. M.; Waters, N. J.;

Smith, J. J.; Moyer, M. P.; Copeland, R. A.; Chesworth, R. Structure

and property guided design in the identification of PRMT5 tool

compound EPZ015666. ACS Med. Chem. Lett. 2016, 7, 162−166.

(10) Bonday, Z. Q.; Cortez, G. S.; Grogan, M. J.; Antonysamy, S.;

Weichert, K.; Bocchinfuso, W. P.; Li, F.; Kennedy, S.; Li, B.; Mader,

M. M.; Arrowsmith, C. H.; Brown, P. J.; Eram, M. S.; Szewczyk, M.

ABBREVIATIONS

■

AAA, asymmetric allylic alkylation; Ac₂O, acetic anhydride;

ADDP, 1,1′-(azodicarbonyl)dipiperidine; API, active pharma-

ceutical ingredient; 9-BBN, 9-borabicyclo[3.3.1]nonane; BLQ,

below the limit of quantiﬁcation; BnBr, benzyl bromide; Cl_p,

plasma clearance; Cl_u, unbound plasma clearance; CYP,

3937

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

M.; Barsyte-Lovejoy, D.; Vedadi, M.; Guccione, E.; Campbell, R. M.

LLY-283, a potent and selective inhibitor of arginine methyltransfer-

ase 5, PRMT5, with antitumor activity. ACS Med. Chem. Lett. 2018, 9,

612−617.

(11) Zhu, K.; Tao, H.; Song, J.-L.; Jin, L.; Zhang, Y.; Liu, J.; Chen,

Z.; Jiang, C.-S.; Luo, C.; Zhang, H. Identification of 5-benzylidene-2-

phenylthiazolones as potent PRMT5 inhibitors by virtual screening,

structural optimization and biological evaluations. Bioorg. Chem. 2018,

81, 289−298.

Yan-Neale, Y.; Yang, G.; Lin, F.; Yin, H.; Gao, H.; Kipp, D. R.; Zhao,

S.; McNamara, J. T.; Sprague, E. R.; Zheng, B.; Lin, Y.; Cho, Y. S.; Gu,

J.; Crawford, K.; Ciccone, D.; Vitari, A. C.; Lai, A.; Capka, V.; Hurov,

K.; Porter, J. A.; Tallarico, J.; Mickanin, C.; Lees, E.; Pagliarini, R.;

Keen, N.; Schmelzle, T.; Hofmann, F.; Stegmeier, F.; Sellers, W. R.

Disordered methionine metabolism in MTAP/CDKN2A-deleted

cancers leads to dependence on PRMT5. Science 2016, 351, 1208−

1213.

(24) Kryukov, G. V.; Wilson, F. H.; Ruth, J. R.; Paulk, J.; Tsherniak,

A.; Marlow, S. E.; Vazquez, F.; Weir, B. A.; Fitzgerald, M. E.; Tanaka,

M.; Bielski, C. M.; Scott, J. M.; Dennis, C.; Cowley, G. S.; Boehm, J.

S.; Root, D. E.; Golub, T. R.; Clish, C. B.; Bradner, J. E.; Hahn, W. C.;

Garraway, L. A. MTAP deletion confers enhanced dependency on the

PRMT5 arginine methyltransferase in cancer cells. Science 2016, 351,

1214−1218.

(25) Barbarino, M.; Cesari, D.; Bottaro, M.; Luzzi, L.; Namagerdi,

A.; Bertolino, F. M.; Bellan, C.; Proietti, F.; Somma, P.; Micheli, M.;

de Santi, M. M.; Guazzo, R.; Mutti, L.; Pirtoli, L.; Paladini, P.;

Indovina, P.; Giordano, A. PRMT5 silencing selectively affects

MTAP -deleted mesothelioma: in vitro evidence of a novel promising

approach. J. Cell. Mol. Med. 2020, 24, 5565−5577.

(26) Palte, R. L.; Schneider, S. E.; Altman, M. D.; Hayes, R. P.;

Kawamura, S.; Lacey, B. M.; Mansueto, M. S.; Reutershan, M.;

Siliphaivanh, P.; Sondey, C.; Xu, H.; Xu, Z.; Ye, Y.; Machacek, M. R.

Allosteric modulation of protein arginine methyltransferase 5

(PRMT5). ACS Med. Chem. Lett. 2020, 11, 1688−1693.

(27) Lin, H.; Wang, M.; Zhang, Y. W.; Tong, S.; Leal, R. A.; Shetty,

R.; Vaddi, K.; Luengo, J. I. Discovery of potent and selective covalent

protein arginine methyltransferase 5 (PRMT5) inhibitors. ACS Med.

Chem. Lett. 2019, 10, 1033−1038.

(28) Shen, Y.; Gao, G.; Yu, X.; Kim, H.; Wang, L.; Xie, L.; Schwarz,

M.; Chen, X.; Guccione, E.; Liu, J.; Bedford, M. T.; Jin, J. Discovery of

first-in-class protein arginine methyltransferase 5 (PRMT5) de-

graders. J. Med. Chem. 2020, 63, 9977−9989.

(29) Wu, T.; Brehmer, D.; Beke, L.; Boeckx, A.; Diels, G. S. M.;

Gilissen, R. A. H. J.; Lawson, E. C.; Pande, V.; Paradé, M. C. B. C.;

Schepens, W. B. G.; Thuring, J. W. J. F.; Viellevoye, M.; Sun, W.;

Meerpoel, L. 6-6 bicyclic aromatic ring substituted nucleoside

analogues for use as PRMT5 inhibitors. U.S. Patent 10,653,711 B2,

May 19, 2020.

(30) Zheng, Y.; Tice, C. M.; Singh, S. B. Conformational control in

structure-based drug design. Bioorg. Med. Chem. Lett. 2017, 27, 2825−

2837.

(31) Rai, B. K.; Sresht, V.; Yang, Q.; Unwalla, R.; Tu, M.;

Mathiowetz, A. M.; Bakken, G. A. Comprehensive assessment of

torsional strain in crystal structures of small molecules and protein-

ligand complexes using ab initio calculations. J. Chem. Inf. Model.

2019, 59, 4195−4208.

(32) Ravn, J.; Nielsen, P. Synthesis of bicyclic nucleosides by ring-

closing metathesis. J. Chem. Soc., Perkin Trans. 1 2001, 985−993.

(33) Haziri, A. I.; Silhar, P.; Renneberg, D.; Leumann, C. J. Synthesis

of the sugar building block of bicyclo-RNA. Synthesis 2010, 823−827.

(34) Haziri, A. I.; Leumann, C. J. Synthesis and pairing properties of

oligodeoxynucleotides containing bicyclo-RNA and bicyclo-ANA

modifications. J. Org. Chem. 2012, 77, 5861−5869.

(35) Downey, A. M.; Pohl, R.; Roithová, J.; Hocek, M. Synthesis of

nucleosides through direct glycosylation of nucleobases with 5-O -

monoprotected or 5-modified ribose: improved protocol, scope, and

mechanism. Chem. − Eur. J. 2017, 23, 3910−3917.

(36) Lovering, F.; Bikker, J.; Humblet, C. Escape from flatland:

increasing saturation as an approach to improving clinical success. J.

Med. Chem. 2009, 52, 6752−6756.

(37) Ritchie, T. J.; Macdonald, S. J. F. Physicochemical descriptors

of aromatic character and their use in drug discovery. J. Med. Chem.

2014, 57, 7206−7215.

(38) Arnott, J. A.; Planey, S. L. The influence of lipophilicity in drug

discovery and design. Expert Opin. Drug Discovery 2012, 7, 863−875.

(39) Augustijns, P.; Wuyts, B.; Hens, B.; Annaert, P.; Butler, J.;

Brouwers, J. A review of drug solubility in human intestinal fluids:

(12) Zhu, K.; Song, J.-L.; Tao, H.-R.; Cheng, Z.-Q.; Jiang, C.-S.;

Zhang, H. Discovery of new potent protein arginine methyltransferase

5 (PRMT5) inhibitors by assembly of key pharmacophores from

known inhibitors. Bioorg. Med. Chem. Lett. 2018, 28, 3693−3699.

(13) Zhu, K.; Shao, J.; Tao, H.; Yan, X.; Luo, C.; Zhang, H.; Duan,

W. Rational design, synthesis and biological evaluation of novel

triazole derivatives as potent and selective PRMT5 inhibitors with

antitumor activity. J. Comput.-Aided Mol. Des. 2019, 33, 775−785.

(14) Shao, J.; Zhu, K.; Du, D.; Zhang, Y.; Tao, H.; Chen, Z.; Jiang,

H.; Chen, K.; Luo, C.; Duan, W. Discovery of 2-substituted-N -(3-

(3,4-dihydroisoquinolin-2(1 H )-yl)-2-hydroxypropyl)-1,2,3,4-tetrahy-

droisoquinoline-6-carboxamide as potent and selective protein

arginine methyltransferases 5 inhibitors: design, synthesis and

biological evaluation. Eur. J. Med. Chem. 2019, 164, 317−333.

(15) Mao, R.; Shao, J.; Zhu, K.; Zhang, Y.; Ding, H.; Zhang, C.; Shi,

Z.; Jiang, H.; Sun, D.; Duan, W.; Luo, C. Potent, selective, and cell

active protein arginine methyltransferase 5 (PRMT5) inhibitor

developed by structure-based virtual screening and hit optimization.

J. Med. Chem. 2017, 60, 6289−6304.

(16) Chan-Penebre, E.; Kuplast, K. G.; Majer, C. R.; Boriack-Sjodin,

P. A.; Wigle, T. J.; Johnston, L. D.; Rioux, N.; Munchhof, M. J.; Jin,

L.; Jacques, S. L.; West, K. A.; Lingaraj, T.; Stickland, K.; Ribich, S. A.;

Raimondi, A.; Scott, M. P.; Waters, N. J.; Pollock, R. M.; Smith, J. J.;

Barbash, O.; Pappalardi, M.; Ho, T. F.; Nurse, K.; Oza, K. P.;

Gallagher, K. T.; Kruger, R.; Moyer, M. P.; Copeland, R. A.;

Chesworth, R.; Duncan, K. W. A selective inhibitor of PRMT5 with in

vivo and in vitro potency in MCL models. Nat. Chem. Biol. 2015, 11,

432−437.

(17) Ye, Y.; Zhang, B.; Mao, R.; Zhang, C.; Wang, Y.; Xing, J.; Liu,

Y.-C.; Luo, X.; Ding, H.; Yang, Y.; Zhou, B.; Jiang, H.; Chen, K.; Luo,

C.; Zheng, M. Discovery and optimization of selective inhibitors of

protein arginine methyltransferase 5 by docking-based virtual

screening. Org. Biomol. Chem. 2017, 15, 3648−3661.

(18) Kong, G.-M.; Yu, M.; Gu, Z.; Chen, Z.; Xu, R.-M.; O’Bryant,

D.; Wang, Z. Selective small-chemical inhibitors of protein arginine

methyltransferase 5 with anti-lung cancer activity. PLoS One 2017, 12,

No. e0181601.

(19) Pande, V.; Sun, W.; Beke, L.; Berthelot, D.; Brehmer, D.;

Brown, D.; Corbera, J.; Irving, S.; Meerpoel, L.; Nys, T.; Parade, M.;

Robinson, C.; Sommen, C.; Viellevoye, M.; Wu, T.; Thuring, J. W. A

chemical probe for the methyl transferase PRMT5 with a novel

binding mode. ACS Med. Chem. Lett. 2020, 11, 2227−2231.

(20) Antonysamy, S.; Bonday, Z.; Campbell, R. M.; Doyle, B.;

Druzina, Z.; Gheyi, T.; Han, B.; Jungheim, L. N.; Qian, Y.; Rauch, C.;

Russell, M.; Sauder, J. M.; Wasserman, S. R.; Weichert, K.; Willard, F.

S.; Zhang, A.; Emtage, S. Crystal structure of the human

PRMT5:MEP50 complex. Proc. Natl. Acad. Sci. U. S. A. 2012, 109,

17960−17965.

(21) Timm, D. E.; Bowman, V.; Madsen, R.; Rauch, C. Cryo-

electron microscopy structure of a human PRMT5:MEP50 complex.

PLoS One 2018, 13, No. e0193205.

(22) Marjon, K.; Cameron, M. J.; Quang, P.; Clasquin, M. F.;

Mandley, E.; Kunii, K.; McVay, M.; Choe, S.; Kernytsky, A.; Gross, S.;

Konteatis, Z.; Murtie, J.; Blake, M. L.; Travins, J.; Dorsch, M.; Biller,

S. A.; Marks, K. M. MTAP deletions in cancer create vulnerability to

targeting of the MAT2A/PRMT5/RIOK1 axis. Cell Rep. 2016, 15,

574−587.

(23) Mavrakis, K. J.; McDonald, E. R., III; Schlabach, M. R.; Billy,

E.; Hoffman, G. R.; deWeck, A.; Ruddy, D. A.; Venkatesan, K.; Yu, J.;

McAllister, G.; Stump, M.; deBeaumont, R.; Ho, S.; Yue, Y.; Liu, Y.;

3938

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

implications for the prediction of oral absorption. Eur. J. Pharm. Sci.

2014, 57, 322−332.

(60) Grossman, R. B. Pauson-khand approach to chiral, diaster-

eomerically pure group 4 ansa-metallocene complexes. Tetrahedron

1999, 55, 919−934.

(40) Conde-Estévez, D. Targeted cancer therapy: interactions with

other medicines. Clin. Transl. Oncol. 2017, 19, 21−30.

(41) Hodgson, D. M.; Galano, J.-M.; Christlieb, M. Enantioselective

total synthesis of (−)-xialenon A. Chem. Commun. 2002, 2436−2437.

(42) Cramer, N.; Buchweitz, M.; Laschat, S.; Frey, W.; Baro, A.;

Mathieu, D.; Richter, C.; Schwalbe, H. Total synthesis and NMR

investigations of cylindramide. Chem. − Eur. J. 2006, 12, 2488−2503.

(43) Hodgson, D. M.; Cameron, I. D. Functionalized bicyclo[3.3.0]-

octanes by enantioselective transannular desymmetrization. Org. Lett.

2001, 3, 441−444.

(61) Rodríguez Rivero, M.; Carretero, J. C. Intramolecular pauson-

khand reactions of α ,β -unsaturated esters and related electron-

deficient olefins. J. Org. Chem. 2003, 68, 2975−2978.

̈

(62) Blanco-Urgoiti, J.; Anorbe, L.; Pérez-Serrano, L.; Domínguez,

G.; Pérez-Castells, J. The pauson-khand reaction, a powerful synthetic

tool for the synthesis of complex molecules. Chem. Soc. Rev. 2004, 33,

32−42.

(63) Ma, K.; Martin, B. S.; Yin, X.; Dai, M. Natural product

syntheses via carbonylative cyclizations. Nat. Prod. Rep. 2019, 36,

174−219.

(44) O’Brien, P.; Rosser, C. M.; Caine, D. On the α -lithiation-

rearrangement of N -toluensulfonyl aziridines: mechanistic and

synthetic aspects. Tetrahedron 2003, 59, 9779−9791.

(64) Seley, K. L.; Schneller, S. W.; Rattendi, D.; Bacchi, C. J. (+)-7-

Deaza-5 ‘-noraristeromycin as an Anti-Trypanosomal Agent. J. Med.

Chem. 1997, 40, 622−624.

(45) Paquette, L. A.; Nakatani, S.; Zydowsky, T. M.; Edmondson, S.

D.; Sun, L.-Q.; Skerlj, R. Studies directed toward the synthesis of the

unusual antileukemic diterpene jatrophatrione. 1. A solution to the

problem of chirality merger during elaboration of the entire

carbotricyclic framework. J. Org. Chem. 1999, 64, 3244−3254.

(46) Trost, B. M.; Curran, D. P. On the stereochemistry of the bis-

nor-wieland-miescher ketome. Tetrahedron Lett. 1981, 22, 4929−

4932.

(47) Kanger, T.; Raudla, K.; Aav, R.; Muurisepp, A.-M.; Pehk, T.;

̈ ̈

Lopp, M. Synthesis and derivatization of bis-nor Wieland-Miescher

ketone. Synthesis 2005, 3147−3151.

(48) Tsantali, G. G.; Takakis, I. M. Expeditious copper-catalyzed

conjugate 1,4-addition of bromo[2-(1,3-dioxolan-2-yl)ethyl]magnesi-

um to α ,β -cycloalkenones and subsequent transformations. J. Org.

Chem. 2003, 68, 6455−6458.

(49) Paquette, L. A.; Liu, Z.; Ramsey, C.; Gallucci, J. C. Competitive

stereochemical control operative during conrotatory electrocyclization

of helically equilibrating diquinanyl-substituted 1,3,5,7-octatetraenyl

bisenolates. J. Org. Chem. 2005, 70, 8154−8161.

(50) Kawato, Y.; Takahashi, N.; Kumagai, N.; Shibasaki, M.

Catalytic asymmetric conjugate addition of α -cyanoketones for the

construction of a quaternary stereogenic center. Org. Lett. 2010, 12,

1484−1487.

(51) Corey, E. J.; Pyne, S. G. Conversion of ketones having δ , ϵ -π -

functions to cyclopentanols by zinc-trimethylchlorosilane. Tetrahedron

Lett. 1983, 24, 2821−2824.

(52) Peterson, T. H.; Carpenter, B. K. Activation parameters for 1,5-

hydrogen transfer and intramolecular cycloaddition in a thermally

generated cyclopentane-1,3-diyl. J. Am. Chem. Soc. 1993, 115, 5466−

5478.

(53) Yates, P.; Stevens, K. E. Bicyclopentanoid sesquiterpenes. The

synthesis of cedranoid sesquiterpenes via the photo-rearrangement of

bicyclo[2.2.2]octenones. Tetrahedron 1981, 37, 4401−4410.

(54) Kan, T.; Inoue, T.; Kawamoto, Y.; Yonehara, M.; Fukuyama, T.

A novel synthesis of bicyclo[3.3.0]octane ring system via a

desymmetric C-H insertion reaction. Synlett 2006, 1583−1585.

(55) Trost, B. M. [3+2] cycloaddition approaches to five-membered

rings via trimethylenemethane and its equivalents [New Synthetic

Methods (55)]. Angew. Chem., Int. Ed. 1986, 25, 1−20.

(56) Lee, T. V.; Boucher, R. J.; Porter, J. R.; Taylor, D. A. The

preparation and reactions with an o-silylated enolate of an allylsilane

bifunctional [3 + 2] annulating reagent. Tetrahedron 1988, 44, 4233−

4242.

(57) Hicks, F. A.; Kablaoui, N. M.; Buchwald, S. L. Titanocene-

catalyzed cyclocarbonylation of enynes to cyclopentenones. J. Am.

Chem. Soc. 1996, 118, 9450−9451.

(58) Hicks, F. A.; Berk, S. C.; Buchwald, S. L. A practical titanium-

catalyzed synthesis of bicyclic cyclopentenones and allylic amines. J.

Org. Chem. 1996, 61, 2713−2718.

(59) Krafft, M. E.; Bonaga, L. V. R.; Hirosawa, C. Modification and

limitations of the livinghouse catalytic pauson-khand reaction.

Tetrahedron Lett. 1999, 40, 9171−9175.

3939

https://doi.org/10.1021/acs.jmedchem.0c02083

J. Med. Chem. 2021, 64, 3911−3939

Article Doi

The Discovery of Two Novel Classes of 5,5-Bicyclic Nucleoside-Derived PRMT5 Inhibitors for the Treatment of Cancer

DOI: 10.1021/acs.jmedchem.0c02083

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.0c02083

Full text of DOI:10.1021/acs.jmedchem.0c02083

Products guided by the article

R&D Labs maybe for 1227269-86-4

Relevant to this article

Hot Product