Selective lewis acid catalyzed assembly of phosphonomethyl ethers: Three-step synthesis of tenofovir

Article abstract of DOI:10.1021/ol503612h

Described herein is a novel Lewis acid catalyzed rearrangement-coupling of oxygen heterocycles and bis(diethylamino)chlorophosphine that provides direct formation of the phosphonomethyl ether functionality found in several important antiretroviral agents. A wide range of dioxolanes and 1,3-dioxanes may be employed, furnishing the desired products in good yield. The utility of this method is demonstrated in a novel synthesis of tenofovir, an antiretroviral drug used in the treatment of HIV/AIDS and hepatitis B.

Full text of DOI:10.1021/ol503612h

Letter
pubs.acs.org/OrgLett
Selective Lewis Acid Catalyzed Assembly of Phosphonomethyl
Ethers: Three-Step Synthesis of Tenofovir
Charles E. Ocampo, Doris Lee, and Timothy F. Jamison*
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
S
* Supporting Information
ABSTRACT: Described herein is a novel Lewis acid catalyzed
rearrangement−coupling of oxygen heterocycles and bis-
(diethylamino)chlorophosphine that provides direct formation
of the phosphonomethyl ether functionality found in several
important antiretroviral agents. A wide range of dioxolanes and
1,3-dioxanes may be employed, furnishing the desired products
in good yield. The utility of this method is demonstrated in a
novel synthesis of tenofovir, an antiretroviral drug used in the
treatment of HIV/AIDS and hepatitis B.
ver 35 million people are infected with HIV/AIDS
worldwide, 28 million of whom are in developing
O
nations.1 Accessibility to effective treatments is limited in
these regions, even with help from organizations such as the Bill
and Melinda Gates Foundation (BMGF) and the Clinton
Health Access Initiative (CHAI).2 Tenofovir disoproxil
fumarate (1, TDF) has emerged as an effective treatment for
HIV/AIDS,3 and it is preferred over alternative HIV
medications because it is less toxic and has a lower rate of
resistance.4 Approved by the FDA in 2001, it is part of the first-
line treatment against HIV/AIDS in combination with other
antiretroviral drugs.5
Tenofovir (2), the active constituent of TDF, is a nucleoside
analogue reverse transcriptase inhibitor that is released in vivo
by phosphonate ester hydrolysis.6 It has been proposed that the
electronegative nature of the β-oxygen atom facilitates
phosphorylation by kinases.7 The phosphonomethyl ether
core found in tenofovir and other antiretrovirals such as
adefovir dipivoxil and cidofovir (Figure 1a) is typically
synthesized via multistep processes involving protecting groups,
a substitution reaction of a tosyloxymethyl ether of a
phosphonate diester, and the use of bases that are either
expensive or difficult to handle on industrial scale.8 Herein, we
report a novel catalytic rearrangement−coupling of readily
available starting materials that furnishes the critical phospho-
nomethyl ether functionality directly (Figure 1b) and enables a
three-step synthesis of tenofovir.
Figure 1. (a) Antiretroviral pharmaceuticals; (b) a novel Lewis acid
catalyzed rearrangement−coupling that assembles the common
phosphonomethyl ether core.
We initially envisioned generation of the phosphonomethyl
ether by effecting nucleophilic ring opening of a cyclic
phosphonate (phostone), in analogy to a carbohydrate-derived
phostone ring opening reported by Moravcova and co-workers
(Scheme 1).9,10 Phostone 3 (R = CH3) was generated from the
ring expansion of the corresponding dioxolane by treatment
with diethylchlorophosphite under Lewis acidic conditions.11
Use of adenine as the nucleophile in the substitution reaction of
3 would have been ideal, as this would directly furnish the core
structure of the antiretrovirals shown in Figure 1a. Unfortu-
nately, the desired reactivity was not observed using a variety of
different bases (e.g., Cs2CO3, K2CO3, NaH, NaOH). Other
nucleophiles such as iodide and organolithium reagents were
also ineffective.12
We were drawn, however, to the structure of the putative
phosphonium intermediate in this process. Specifically, we
Received: December 15, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/ol503612h
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Initial Design: Phosphonomethyl Ether
Formation via Nucleophilic Phostone Ring Opening
Scheme 2. Proposed Mechanism of the Reaction
envisioned an analogous intermediate starting from bis(diethyl-
amino)chlorophosphine (Figure 1b). In this case, reaction at
the ethyl groups would be unlikely; instead, we hypothesized
that direct chloride attack onto the phosphonium ring would
furnish the corresponding phosphonomethyl ether in a single
step.
Several transition-metal- and lanthanide-based Lewis acid
catalysts were evaluated for the rearrangement−coupling of
dioxolane 4a in the presence of excess bis(diethylamino)-
chlorophosphine.13 Zinc(II) catalysts, particularly ZnCl2 and
Zn(OTf)2, gave optimal results. 4-Methyl-1,3-dioxolane 4a was
chosen as a development substrate for several reasons: it would
probe group selectivity, does not display a strong electronic or
steric bias, and, following the rearrangement, would form the
phosphonomethyl ether core of tenofovir. Indeed, the predicted
product 5a was generated in 10:1 selectivity over minor isomer
6a (Table 1). Increasing the temperature to 130 °C gave
to generate cyclic phosphonium intermediate 9. Finally,
chloride attack furnishes the phosphonomethyl ether product.
In this framework, the minor isomer is generated by initial
reaction at the more sterically hindered oxygen.
To probe the initial Lewis acid−chlorophosphine interaction,
we conducted 31P NMR studies.15 Mixing a 1:1 ratio of
chlorophosphine and ZnCl2 (chosen for solubility reasons) in
THF-d8 at ambient temperature resulted in no detectable
change in chemical shift. In order to enhance the putative
interaction, we conducted further experiments in CD2Cl2, a
noncoordinating solvent. Indeed, we found a dramatic
downfield shift in the 31P NMR spectrum (158.8 to 234.6
ppm). These results support the formation of a (Et2N)2PCl−
ZnX2 complex and is consistent with previous NMR studies.16
We evaluated the rearrangement conditions for a variety of
substituted dioxolanes and 1,3-dioxanes (Table 2). In the case
of the 4-substituted substrates, even a methyl group was
sufficient in biasing reaction to the less sterically hindered
oxygen, yielding the desired product in a 10:1 ratio (entries 1
and 11). Higher selectivity was observed for the substrates
bearing a phenyl or geminal dimethyl group at the 4-position
(entries 3, 12, and 13), possibly a result of the larger size of the
phenyl ring or geminal dimethyl group (relative to methyl),
decreasing the degree of coordination of the more sterically
encumbered oxygen to the chlorophosphine. Interestingly, a
decrease in selectivity was observed for 4-((benzyloxy)methyl)-
1,3-dioxolane (entry 4). It is postulated that the benzyloxy
group can direct reaction at the more sterically hindered oxygen
of the dioxolane by coordinating to phosphorus. Electronic
effects may also play a significant role in this case.
Generally, we observed higher yields for the 2-substituted
dioxolanes (entries 5−8), possibly due to the groups in this
position stabilizing the oxonium intermediate. Additionally, we
obtained higher yields at lower concentration (1.0 M vs 2.5 M)
for these substrates; no significant concentration effect was
observed for the 4-substituted substrates.13 A lower yield was
observed with 2-vinyl dioxolane 4i due to partial alkene
isomerization in the product, giving an internal olefin (entry 9).
Highly sterically hindered substrates, bearing geminal dimethyl
groups, were also tolerated under these conditions (entries 6
and 12).
Table 1. Evaluation of Lewis Acids and Reaction
a
Temperatures for the Dioxolane Rearrangement
b
entry
Lewis acid
temp (°C)
yield (%)
c
1
2
3
ZnCl2
100
100
120
120
120
120
120
130
130
140
140
21
ZnCl2
33
41
51
53
57
40
50
61
52
59
ZnCl2
d
4
ZnCl2
e
5
ZnCl2
6
Zn(OTf)2
Zn(OTf)2
ZnCl2
d
7
8
9
Zn(OTf)2
ZnCl2
10
11
Zn(OTf)2
a
Dioxolane (0.8 mmol), chlorophosphine (1.9 mmol), ZnX2 (20 mol
% relative to dioxolane, unless otherwise noted), THF (0.3 mL); see
b
the Supporting Information. Isolated yield after purification (SiO2
c
d
e
chromatography). 1.2 equiv of phosphine. 60 mol % ZnX2. 100 mol
% ZnX2.
significant improvement in yield; however, no further improve-
ment was observed at or above 140 °C. These studies thus
established our standard conditions (130 °C, 20 mol %
Zn(OTf)2).
We propose that the mechanism of the reaction initially
involves interaction of Zn(OTf)2 with the chlorine atom to
increase the electrophilicity at phosphorus (Scheme 2). The
less sterically hindered oxygen of the dioxolane then reacts with
the activated chlorophosphine to generate oxonium inter-
mediate 7. Phosphorus attack onto oxocarbenium 814 in an
intramolecular Michaelis−Arbuzov-type reaction then proceeds
To demonstrate the utility of this method, we developed a
novel and efficient synthesis of tenofovir.17 The first step of the
route involves reaction of enantiomerically pure (R)-4a,
B
DOI: 10.1021/ol503612h
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
prepared from (R)-(−)-1,2-propanediol and paraformaldehyde,
under the standard rearrangement conditions to afford
enantiomerically pure phosphonomethyl ether (R)-5a in 72%
yield (Scheme 3).18
Table 2. Zn(OTf)2-Catalyzed Rearrangement of Substituted
Dioxolanes and 1,3-Dioxanes
a
Scheme 3. Synthesis of Tenofovir via Zn(OTf)2-Catalyzed
Dioxolane Rearrangement
Alkylation with adenine using potassium phosphate (K3PO4),
followed by hydrolysis of the phosphonamide under mild
conditions (2 N HCl, 80 °C), affords tenofovir monohydrate
(2·H2O) in 41% yield over two steps after recrystallization. The
overall yield of the synthesis starting from (R)-4a is 27%. The
advantages of this synthetic route include rapid formation of
the phosphonomethyl ether core, a late-stage introduction of
adenine, mild hydrolysis conditions, and recrystallization of the
final product by simple pH adjustment. Moreover, the final two
steps require no chromatographic purification.
In summary, we have developed a protocol that allows for
direct access to a variety of phosphonomethyl ether
compounds. This novel rearrangement−coupling utilizes
reagents that either are commercially available or can be
accessed using inexpensive starting materials. Moreover, all of
the atoms contained in the starting materials are transferred to
the product. Dioxolane and 1,3-dioxane substrates bearing
substituents at either the 2- or 4-position are tolerated,
generating the corresponding products in good yield. Finally,
this transformation enables a new, three-step synthesis of
tenofovir, demonstrating the utility of this method and its
potential for accessing other antiretroviral compounds contain-
ing the phosphonomethyl ether motif.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and spectral data (1H, 13C, 31P NMR;
HRMS) for all new compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: tfj@mit.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
a
This research was supported by the Bill and Melinda Gates
Foundation (“Medicines for All” Initiative). D.L. would like to
thank the NSERC for a postdoctoral fellowship. We thank B.
Frank Gupton (VCU), D. Tyler McQuade (FSU), and Bryan J.
Marquardt (University of Washington) for stimulating
Dioxolane (0.8 mmol), chlorophosphine (1.9 mmol), Zn(OTf)2 (20
b
mol %), THF (0.3 mL); see the Supporting Information. Isolated
yield after purification (SiO2 chromatography). Values in parentheses
c
d
represent isomeric ratios. 0.8 mL of THF. 23% of internal alkene
isomer also isolated.
C
DOI: 10.1021/ol503612h
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
discussions. Li Li (MIT) is acknowledged for HRMS data,
which were obtained on an instrument purchased with the
assistance of NSF Grant CHE-0234877. NMR spectroscopy
was carried out on instruments purchased in part with funds
provided by the NSF (CHE-9808061 and CHE-8915028).
REFERENCES
■
(1) World Health Organization. HIV/AIDS. http://www.who.int/
hiv/en/ (accessed Oct 14, 2014).
(2) (a) Bill and Melinda Gates Foundation. HIV Strategy Overview.
http://www.gatesfoundation.org/What-We-Do/Global-Health/HIV
(accessed Nov 10, 2014). (b) Clinton Health Access Initiative. HIV/
AIDS. http://www.clintonhealthaccess.org/program-areas/HIV-AIDS
(accessed Nov 10, 2014).
(3) (a) Naessens, L.; Bischofberger, N.; Augustijns, P.; Annaert, P.;
Van Den Mooter, G.; Arimilli, M. N.; Kim, C. U.; De Clercq, E.
Antimicrob. Agents Chemother. 1998, 42, 1568−1573. (b) Robbins, B.
L.; Srinivas, R. V.; Kim, C.; Bischofberger, N.; Fridland, A. Antimicrob.
Agents Chemother. 1998, 42, 612−617.
(4) Palacios, R.; Santos, J.; Camino, X.; Arazo, P.; Torres Perea, R.;
́
Echevarrfa, S.; Ribera, E.; Sanchez de la Rosa, R.; Moreno Guillen, S.
Curr. Ther. Res. 2005, 66, 117−129.
(5) (a) Feng, J. Y.; Ly, J. K.; Myrick, F.; Goodman, D.; White, K. L.;
Svarovskaia, E. S.; Borroto-Esoda, K.; Miller, M. D. Retrovirology 2009,
6, 44−59. (b) De Clercq, E. Biochem. Pharmacol. 2007, 73, 911−922.
(6) Friis, G. J.; Bundgaard, H. Eur. J. Pharm. Sci. 1996, 4, 49−59.
(7) Kim, C. U.; Luh, B. Y.; Misco, P. F.; Bronson, J. J.; Hitchcock, M.
J. M.; Ghazzouli, I.; Martin, J. C. M. J. Med. Chem. 1990, 33, 1207−
1213.
(8) (a) Holy, A.; Rosenberg, I. Collect. Czech. Chem. Commun. 1982,
́
47, 3447−3463. (b) Holy, A.; Masojídkova,
́
M. Collect. Czech. Chem.
Commun. 1995, 60, 1196−1212. (c) Holy, A.; Dvora ova, H.;
M. Collect. Czech. Chem. Commun. 1995, 60, 1390−
́
́
̌
k
́
́
Masojídkova,
́
1409. (d) Schultze, L. M.; Chapman, H. H.; Dubree, N. J. P.; Jones, R.
J.; Kent, K. M.; Lee, T. T.; Louie, M. S.; Postich, M. J.; Prisbe, E. J.;
Rohloff, J. C.; Yu, R. H. Tetrahedron Lett. 1998, 39, 1853−1856.
(9) Moravcova, J.; Heissigerova, H.; Kocalka, P.; Imberty, A.; Sykora,
D.; Fris, M. Tetrahedron Lett. 2003, 44, 8797−8800.
(10) For nucleophilic lactone ring openings, see: Zhengchun, L.;
Granata, A.; Shen, X.; Perlin, A. S. Can. J. Chem. 1992, 70, 2081−2088.
(11) Pav
I. Org. Lett. 2007, 9, 5469−5472.
́
́ ̌ ̌ ́
, O.; Barvík, I.; Budesínsky, M.; Masojídkova, M.; Rosenberg,
(12) Use of Cs2CO3, K2CO3, NaH, NaOH, organolithium reagents,
or iodide resulted in either recovery or decomposition of phostone
starting material.
(13) For a full evaluation of Lewis acid catalysts and other reaction
conditions, see the Supporting Information for details.
(14) We cannot rule out the possibility that oxonium 7 undergoes
chloride attack to generate the corresponding chloromethyl ether
intermediate; however, since gem-dimethyl groups are tolerated in the
2-position, it is more likely that the mechanism proceeds via
oxocarbenium intermediate 8.
(15) For full details, see the Supporting Information.
(16) Roussis, V.; Baenziger, N. C.; Wiemer, D. F. J. Heterocycl. Chem.
1996, 33, 979−981.
(17) For the current commercial synthesis of TDF, see: Ripin, D. H.
B.; Teager, D. S.; Fortunak, J.; Basha, S. M.; Bivins, N.; Boddy, C. N.;
Byrn, S.; Catlin, K. K.; Houghton, S. R.; Jagadeesh, S. T.; Kumar, K. A.;
Melton, J.; Muneer, S.; Rao, L. N.; Rao, R. V.; Ray, P. C.; Reddy, N.
G.; Reddy, R. M.; Shekar, K. C.; Silverton, T.; Smith, D. T.; Stringham,
R. W.; Subbaraju, G. V.; Talley, F.; Williams, A. Org. Process Res. Dev.
2010, 14, 1194−1201 and references therein.
(18) Enantioselectivity was determined by chiral HPLC of an
intermediate that was further functionalized. See the Supporting
Information for details.
D
DOI: 10.1021/ol503612h
Org. Lett. XXXX, XXX, XXX−XXX