Synthesis of a novel cyclopropyl phosphonate nucleotide as a phosphate mimic

Article abstract of DOI:10.1039/d1cc02328d

The inherentin vivoinstability of oligonucleotides presents one of many challenges in the development of RNAi-based therapeutics. Chemical modification to the 5′-terminus serves as an existing paradigm which can make phosphorylated antisense strands less prone to degradation by endogenous enzymes. It has been recently shown that installation of 5′-cyclopropyl phosphonate on the terminus of an oligonucleotide results in greater knockdown of a targeted protein when compared to its unmodified phosphate derivative. In this paper we report the synthesis of a 5′-modified uridine.

Full text of DOI:10.1039/d1cc02328d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Synthesis of a novel cyclopropyl phosphonate
nucleotide as a phosphate mimic†
Cite this: Chem. Commun., 2021,
57, 6808
Erich F. Altenhofer, *a Michael J. Lawler,a Pankaj Kumar,a Leo A. Joyce,
a
Matthew Fowler-Watters,a Tao Peia and Zhen Lib
Received 30th April 2021,
Accepted 15th June 2021
DOI: 10.1039/d1cc02328d
rsc.li/chemcomm
The inherent in vivo instability of oligonucleotides presents one of expression can be silenced. The RNAi trigger loads onto the
many challenges in the development of RNAi-based therapeutics. RNA-induced silencing complex (RISC) where the duplex is
Chemical modification to the 50-terminus serves as an existing unwound, and the guide strand binds to exonuclease
paradigm which can make phosphorylated antisense strands less Argonaute-2 (Ago2) protein via 50-phosphate interaction.4 The
prone to degradation by endogenous enzymes. It has been recently RISC complex hybridizes with a complementary, targeted mes-
shown that installation of 50-cyclopropyl phosphonate on the senger RNA (mRNA) and exonuclease Ago2 cleaves the targeted
terminus of an oligonucleotide results in greater knockdown of a mRNA. Before this can successfully happen a 50-phosphoryla-
targeted protein when compared to its unmodified phosphate tion of the externally delivered RNAi trigger must occur to
derivative. In this paper we report the synthesis of a 50-modified enable RISC loading. Phosphorylation of the 50-alcohol occurs
uridine.
through kinase Clp1.5 The in vivo stability of the newly formed
50-phosphate plays a key role in the propagation of the RNAi
RNAi-based therapeutics have received increasing attention in pathway.
the last few decades, especially as drug candidates for diseases
Upon phosphorylation, the guide strand is prone to enzy-
that are difficult to access with small molecules. Chemical matic degradation. The labile P–O bond at the 50-terminus of
modification of oligonucleotides is an effective strategy to the guide strand (Fig. 1, 1a) is prone to cleavage by phospha-
increase selectivity, in vivo stability, and accumulation in tases, leading to degradation before entering the RNAi
tissue.1 To direct the therapeutic to a cell of interest, small pathway.6 Additionally, exonucleases cleave monomeric units
molecules serving as targeting ligands can be conjugated to a of the oligonucleotide which also reduces efficacy of the
single strand RNA (ss-RNA). For example, N-acetylgalacto- therapeutic. Chemical modification of the terminal phosphate
samine (GalNAc or NAG) has been shown to be highly efficient serves as an existing approach to prevent these two degradation
in delivering exogenous RNA to hepatocyte cells.2 Additionally, pathways.7
chemical modifications to nucleotides can greatly impact
A cyclopropyl phosphonate moiety in place of the 50-ter-
the stability of ss-RNA. Specifically, the use of 20-fluoro and minal phosphate addresses this shortcoming by removing the
20-methoxy modified nucleotides has been shown to enhance necessity for intracellular phosphorylation of the guide strand
stability of antisense (guide) strands in vivo.3 This work pri- and increasing the metabolic stability of the strand (Fig. 1, 1b).
marily focuses on the chemical modification of the oligo- In a head-to-head study, the 50-cyclopropyl phosphonate was
nucleotide at the 50-terminus as a means to increase the
metabolic stability.
Once an exogenously administered double-stranded RNAi
trigger (e.g., an siRNA or dsRNA) is internalized in the cell type
of interest, a cascade of processes must occur before gene
a Arrowhead Pharmaceuticals, Inc., 502 S Rosa Rd, Madison, WI 53705, USA.
E-mail: ealtenhofer@arrowheadpharma.com
b 5871 Oberlin Drive, San Diego, CA 92121, USA
† Electronic supplementary information (ESI) available: Detailed experimental
procedures and compound characterization data. CCDC 2048176. For ESI and
crystallographic data in CIF or other electronic format see DOI: 10.1039/
Fig. 1 Chemical modification of the 50-terminus of ss-RNA.
d1cc02328d
6808 | Chem. Commun., 2021, 57, 6808–6811
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
With compound 6 in hand, the key step to this route relied
upon a cyclopropanation of the olefin adjacent to the phos-
phonate. Although the cyclopropanation of diethyl vinylpho-
sphonate is known,11 it was unclear whether chemoselectivity
issues would arise from the neighboring nucleobase. Side
reactions such as N-methylation or cyclopropanation of the
vinyl group on uracil would be problematic.
After initial screening experiments failed, it became clear
that the cyclopropanation of a 50-vinylphosphonate was not
trivial. Simmons–Smith chemistry12 and other carbene reagents
were either unreactive with vinylphosphonate 6 or resulted in
substrate decomposition. The most promising methodology
involved Corey–Chaykovsky conditions13 using sulfoxonium
salts which were observed to give clean conversion of 6 to a
more polar product, though unfortunately not the desired
cyclopropyl derivative 7. This methodology was investigated
further since it was the only set of conditions that gave clean
conversion of compound 6. Upon closer examination of the
crude reaction mixture by 1H NMR, the vinylphosphonate
protons of 6 were no longer present (see ESI†), indicating that
reaction of the a,b-unsaturated system had indeed occurred.
The reaction was found to be highly chemoselective, as all
Scheme 1 Original route to access 50-modified uridine 8.
observed to achieve a greater knockdown of a targeted protein uracil protons were still intact.
than its 50-phosphate derivative.8 Herein, we describe the
Mass spectrometry (MS) data were inconclusive, however, as
synthesis of a 50-modified uridine which was achieved through the base peak observed was [M] + 78 rather than the expected
a novel cyclopropanation reaction.
[M] + 1. It was unclear whether this was a result of a DMSO
Synthesis of oligonucleotides is typically carried out on solid solvent adduct of the desired product 7, or the molecular ion
1
phase, coupling a 30-phosphoramidite in solution with a resin for a different higher molecular weight species. The H NMR
bound 50-OH nucleoside. Using this traditional phosphorami- spectrum showed an absence of signals around 1 ppm where
dite chemistry to incorporate a 50-cyclopropyl phosphonate cyclopropyl protons would be expected to resonate. Given these
onto the terminus of the antisense strand would require data, it was hypothesized that the cyclopropanation stalled at
synthesis of 8 containing a phosphoramidite moiety at the intermediate 7a (Scheme 2). This intermediate was found to be
30-position of the ribose (Scheme 1). Commercially available highly stable and could not be converted to the desired product
20-OMe uridine 2 was converted to 30-OTBS 3 through a bis-TBS even at elevated temperatures. Attempts to isolate and char-
protection, followed by selective TBS mono-deprotection at the acterize the intermediate were unsuccessful as it could not be
50-position.9 Oxidation of resulting 50-alcohol with Dess–Martin extracted out of the DMSO reaction solvent.
periodinane (DMP) resulted in clean conversion to aldehyde 4.
The electron deficient character of the olefin that made 1,
Running the oxidation at 0 1C with an equimolar amount of 4-addition proceed very rapidly unfortunately also seemed to
DMP was key to completely consuming 3, while avoiding over- stabilize intermediate 7a and prevent conversion to desired
oxidation of the aldehyde to a carboxylic acid.
product 7. It was hypothesized that replacing the oxygen atom
Unfortunately, the aldehyde was found to decompose during with less electronegative sulfur would further delocalize the
aqueous workup and could not be isolated. The crude solution electron density across the olefin and allow displacement of the
of aldehyde was then treated with ylide 5, affording the dimethylsulfoxonium adduct to give the desired cyclopropane
E-vinylphosphonate 6 exclusively in 60% yield over two steps. product. We therefore envisioned the synthesis of vinyl thio-
This sequence is an improvement over Horner–Wadsworth– phosphonate derivative 10 (Scheme 3).
Emmons chemistry that gives an inseparable 9 : 1 E/Z mixture of
Direct conversion of vinylphosphonate 6 to the corres-
the vinylphosphonate. During the scale up synthesis of dimethyl ponding vinyl thiophosphonate 10 would appear to be a
hydroxymethylphosphonate, an intermediate to compound 5, a
runaway exotherm was observed. The original procedure heated a
neat mixture of reagents until the reaction initiated, causing an
uncontrollable exotherm.10 This necessitated modification of the
process to improve safety parameters. To that end, methanol was
added as a solvent to aid in heat distribution and eliminate hot
spots. Furthermore, a dose-controlled addition of dimethylpho-
sphite at 50 1C prevented reagent accumulation, providing a more
controlled process (see ESI†).
Scheme 2 Formation of locked 1,4-adduct.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 6808–6811 | 6809

View Article Online
ChemComm
Communication
Fig. 2 X-Ray structure of compound 13 (atoms: carbon is gray, hydrogen
is white, oxygen is red, nitrogen is blue, and phosphorus is orange).
13 was observed in under four hours, which to the best of our
knowledge, constitutes a novel use of Oxone.
After workup, compound 13 was found to be crystalline,
providing an opportunity to further purify the crude material
without using silica. The penultimate compound was screened
for solubility in a variety of solvents and appeared to be
insoluble in ethyl acetate. Analysis of the undissolved solids
by 31P NMR showed enrichment of the major diastereomer.
After some development, the major diastereomer could be
enriched with an ethyl acetate trituration giving diastereomeric
purities greater than 90%. To further improve the diastereo-
meric purity, using a 4 : 6 mixture of ethanol and heptane, the
solids could be recrystallized allowing for isolation of 13 as a
single diastereomer. The absolute stereochemistry around the
cyclopropyl group was then confirmed by X-ray crystallography
(Fig. 2).
To complete the synthesis, only the major diastereomer was
carried through to the final step. Using optimized phosphityla-
tion conditions, the penultimate alcohol 13 underwent clean
conversion to phosphoramidite 14 as a 1 : 1 mixture of diaster-
eomers resulting from installation of the new phosphorus
stereocenter.
In conclusion, chemical modification of the 50-terminus of
oligonucleotides constitutes a promising methodology to impart
increased stability and promote enhanced accumulation in
tissues. The present work improves upon existing synthetic
procedures towards the synthesis of 50-vinylphosphonate precur-
sors, allowing access to these substrates. Most importantly, a
novel cyclopropanation reaction on a vinyl thiophosphonate has
been described. Cyclopropanation has been accomplished by
tuning the electronics of the phosphonate, which nicely illustrates
the power of fundamental chemical knowledge in solving syn-
thetic challenges.
Scheme 3 Revised route to access cyclopropyl phosphoramidite 14.
reasonable solution. However, using Lawesson’s reagent on
large scale was not desirable. Having a well-defined Wittig
olefination procedure optimized for aldehyde 4, a route which
used a thiophosphonate analogue of ylide 5 was explored. Due
to limited commercial availability of O,O0-dimethyl chlorothio-
phosphate, the ethyl thioester analogue was used to make
compound 9. To our knowledge, this is the first reported
synthesis and application of this unknown ylide, although a
similar phenyl derivative14 has been reported. The stabilized
ylide was found to exist as an amorphous solid which compli-
cated any attempts at purification. Nevertheless, the ylide was
found to be stable for several months at 5 1C and the crude
amorphous solids were used to carry out the next step.
To circumvent the previously mentioned instability of
aldehyde 4, it was generated in situ and reacted with the corres-
ponding ylide, leading to clean conversion to E-vinyl thiophos-
phonate 10. Similar to our previous findings, 1H NMR did
not show evidence of Z-isomer of vinyl thiophosphonate 10.
Subjecting E-vinyl thiophosphonate 10 to Corey–Chaykovsky con-
ditions led to clean conversion to the desired cyclopropyl thiopho-
sphonate 11 in 65% yield with a 3 : 2 diastereomeric ratio.
Previously mentioned side reactions such as N-methylation and
cyclopropanation on uracil were not detected during analysis of
1
crude H NMR spectra. After a search of current literature, this
appears to be the only documented cyclopropanation of a vinyl-
phosphonate using a sulphur ylide. Efforts to increase the dia-
stereoselectivity, using different trialkylsulfoxonium reagents were
explored, but in all cases, the 1,4-addition did not occur.
Deprotection of the 30OTBS group with hydrochloric acid
resulted in clean conversion of 11 to corresponding alcohol 12.
Conflicts of interest
Oxidation of thiophosphonate 12 was first explored with Oxone, Arrowhead Pharmaceuticals, Inc. has filed a patent application
as it is a relatively safe oxidant and its application spans a wide covering the process: U.S. patent application no. PCT/US2017/
range of functional groups.15 Conversion of 12 to phosphonate 036108.
6810 | Chem. Commun., 2021, 57, 6808–6811
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
150, 883; (c) W. Chang, Y. Pei, E. N. Guidry, D. Zewge, C. A. Parish,
E. C. Sherer, J. DiMuzio, H. Zhang, V. J. South, W. R. Strapps, L. Sepp-
Lorenzino, S. L. Colletti and M. G. Stanton, Bioorg. Med. Chem. Lett.,
2016, 26, 4513; (d) T. P. Prakash, W. F. Lima, H. M. Murray, W. Li,
G. A. Kinberger, A. E. Chappell, H. Gaus, P. P. Seth, B. Bhat,
S. T. Crooke and E. E. Swayze, Nucleic Acids Res., 2015, 43, 2993.
8 Z. Li, T. Pei and M. Lawler, PCT International Publication, n.d.,
WO2017/214112A1, 2017.
9 A. V. Kel’in, I. Zlatev, J. Harp, M. Jayarman, A. Bisbe, J. O’Shea,
N. Taneja, R. M. Monoharan, S. Khan, K. Charisse, M. A. Maier,
M. Egli, K. G. Rajeev and M. Monoharan, J. Org. Chem., 2016,
81, 2261.
10 (a) J. Gajewiak, Y. Xu, S. A. Lee, T. G. Kutateladze and
G. D. Prestwich, Org. Lett., 2006, 8, 2811; (b) D. P. Phillion and
S. S. Andrew, Tetrahedron Lett., 1986, 27, 1477.
11 L. Tokes and G. Jones, Org. Mass Spectrom., 1975, 10, 241.
12 H. E. Simmons and R. D. Smith, J. Am. Chem. Soc., 1958, 80,
5323.
Notes and references
1 (a) R. Haraszti, L. Roux, A. H. Coles, A. A. Turanov, J. F. Alterman,
D. Echeverria, B. M. D. C. Godinho, N. Aronin and A. Khvorova,
Nucleic Acids Res., 2017, 45, 7581; (b) G. F. Deleavey and M. J. Damha,
Chem. Biol., 2012, 19, 937.
2 A. D. Springer and S. F. Dowdy, Nucleic Acid Ther., 2018, 28, 109.
3 C. R. Allerson, N. Sioufi, R. Jarres, T. P. Prakash, N. Naik, A. Berdeja,
L. Wanders, R. H. Griffey, E. E. Swayze and B. Bhat, J. Med. Chem., 2005,
48, 901.
4 (a) S. M. Hammond, E. Bernstein, D. Beach and G. J. Hannon,
Nature, 2000, 404, 293; (b) A. Khvorova, A. Reynolds and
S. D. Jayasena, Cell, 2003, 115, 209; (c) A. L. Jackson, S. R. Bartz,
J. Schelter, S. V. Kobayashi, J. Burchard, M. Mao, B. Li, G. Cavet and
P. S. Linsley, Nat. Biotechnol., 2003, 21, 635.
5 M. Manoharan, Curr. Opin. Chem. Biol., 2004, 8, 570.
6 A. Khvorova and J. K. Watts, Nat. Biotechnol., 2017, 35, 238.
7 (a) R. Parmar, J. L. S. Willoughby, J. Liu, D. J. Foster, B. Brigham,
C. S. Theile, K. Charisse, A. Akinc, E. Guidry, Y. Pei, W. Strapps,
M. Cancilla, M. G. Stanton, K. G. Rajeev, L. Sepp-Lorenzino,
M. Manoharan, R. Meyers, M. A. Maier and V. Jadhav, ChemBioChem,
2016, 17, 985; (b) W. F. Lima, T. P. Prakash, H. M. Murray,
G. A. Kinberger, W. Li, A. E. Chappell, C. S. Li, S. F. Murray,
13 E. J. Corey and M. Chaykovsky, J. Am. Chem. Soc., 1965, 87,
1353.
14 H. Run-Qiu, D. She-Shiang and Y. Shi-Xian, Acta Chim. Sin., 1989,
7, 64.
H. Gaus, P. P. Seth, E. E. Swayze and S. T. Crooke, Cell, 2012, 15 A. V. Narsaiah, Synlett, 2002, 1178.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 6808–6811 | 6811