Phosphonic acid analogs of GABA through reductive dealkylation of phosphonic diesters with lithium trialkylborohydrides

Article abstract of DOI:10.1016/j.bmcl.2007.04.026

Lithium trialkylborohydrides were found to effect rapid monodealkylation of phosphonic diesters, and this reaction was applied to the synthesis of alkylphosphonic acid 2-aminoethyl esters [H₂N(CH₂)₂OP(OH)R, 4], a little-ex

Full text of DOI:10.1016/j.bmcl.2007.04.026

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3745–3748
Phosphonic acid analogs of GABA through reductive
dealkylation of phosphonic diesters with lithium
trialkylborohydrides
Sarwat Chowdhury,a Niraj J. Muni,b,c Nicholas P. Greenwood,a
d,
David R. Pepperbergc,b, and Robert F. Standaert
*
*
aDepartment of Chemistry, University of Illinois at Chicago, 845 West Taylor Street, Chicago, IL 60607-7056, USA
bDepartment of Bioengineering, University of Illinois at Chicago, 851 S. Morgan St., Chicago, IL 60607-7052, USA
cDepartment of Ophthalmology and Visual Sciences, University of Illinois at Chicago, 1855 West Taylor Street,
Chicago, IL 60612-7243, USA
dBiosciences Division, Oak Ridge National Laboratory, PO Box 2008 MS 6123, Oak Ridge, TN 37831-6123, USA
Received 9 March 2007; revised 4 April 2007; accepted 5 April 2007
Available online 10 April 2007
Abstract—Lithium trialkylborohydrides were found to effect rapid monodealkylation of phosphonic diesters, and this reaction was
applied to the synthesis of alkylphosphonic acid 2-aminoethyl esters [H2N(CH2)2OP(OH)R, 4], a little-explored class of analogs of
the inhibitory neurotransmitter c-aminobutyric acid (GABA). Compound 4a (R = Me) proved to be a potent antagonist at human
q1 GABAC receptors (expressed in Xenopus laevis oocytes), with an IC50 of 11.1 lM, but is inactive at a1b2c2 GABAA receptors.
Ó 2007 Elsevier Ltd. All rights reserved.
c-Aminobutyric acid (GABA) is a major inhibitory neu-
rotransmitter in the central nervous system and has
three major classes of receptors, designated GABAA,
GABAB, and GABAC.1 Effectors of these receptors
(agonists, antagonists, and allosteric modulators) are
an important class of compounds as pharmaceuticals
and pharmacological probes. Compounds targeting
GABAA and GABAB have been extensively studied,2
and GABAC effectors are attracting increased interest.3
For this reason, and because of the important role of
GABAC receptors in vision,3d,4 we have sought to devel-
op new GABAC effectors. We report here the discovery
of a new reaction of lithium trialkylborohydrides, the
reductive monodealkylation of phosphonic diesters,
and its application to the synthesis of 2-aminoethyl
alkylphosphonates (4), a previously unexplored class
of GABAC receptor antagonists.
Phosphinic acids are the most prominent class of
GABAC antagonists.5 In the course of pursuing new
synthetic approaches to 3-aminopropyl alkyl phosphi-
nates (e.g., 2a–c, Fig. 1), we surveyed metal hydrides
for their ability to reduce phosphonic diesters to the cor-
responding H-phosphinates. Among the reagents tested,
only lithium trialkylborohyrides reacted cleanly, but
monodealkylation, rather than reduction at phosphorus,
was observed. Partial conversion was observed with so-
dium tri(s-butyl)borohydride, while little conversion was
observed with the corresponding potassium reagent,
suggesting a specific role for the lithium counterion.
The dealkylation reaction, which likely occurs via SN2
nucleophilic attack at carbon, was surprising in that
Keywords: GABA antagonists; GABAC receptors; Dealkylation;
Phosphonate esters; Lithium trialkylborohydrides.
*
Corresponding authors. Tel.: +1 865 574 2631; fax: +1 865 574 6210
(R.F.S.); tel.: +1 312 996 4262; fax: +1 312 996 7773 (D.R.P.); e-mail
addresses: davipepp@uic.edu; standaertrf@ornl.gov
Figure 1. Structures of GABA and phosphorus oxyacid analogs.
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.026

3746
S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3745–3748
other metal hydride reagents, such as lithium aluminum
hydride,6 lithium bis(methoxyethoxy)aluminum hy-
dride,7 sodium bis(methoxyethoxy)aluminum hydride,8
and sodium diethyl aluminum hydride,8 are known to
attack phosphonates at phosphorus. The examples de-
scribed herein are the first in which nucleophilic attack
on phosphonate esters by metal hydride reagents occurs
preferentially at carbon, leading to dealkylation.
sealed NMR tube fitted with a J. Young valve, a singlet
at d 0.81 ppm, consistent with ethane, appeared and
grew over the course of the reaction. No olefinic signals
were observed, excluding E2 elimination as the primary
mechanism. In the analogous reduction of dimethyl
methylphosphonate, a singlet at d 0.18 ppm, consistent
with methane, likewise emerged. Our findings therefore
suggest that lithium trialkylborohydrides can displace
substantially more basic leaving groups than has been
observed previously.
The initial substrate tested was diethyl difluorobenzyl-
phosphonate,9 which was subjected to reaction with
1.5 equiv of LiHBEt3 or LiHB(s-Bu)3 in THF at room
temperature. With both reagents, the diester was con-
sumed within 30 min, and the sole product was the
monoethyl ester, which could be isolated in 89 or 82%
yield, respectively. The reaction was repeated with
diethyl benzylphosphonate (5a), and again, clean mono-
dealkylation was observed, though in this instance, the
reaction required an hour to go to completion.
Nucleophilic displacement of alkyl groups in phospho-
nate esters occurs with a variety of other nucleophiles,
and reactions of this type are useful for preparative
deprotection reactions. Boron11 and silicon12 halides
are widely employed for complete dealkylation, though
recent work has led to the development of binuclear bor-
on complexes that catalyze removal of a single alkyl
group by BBr3.13
A set of additional diesters 5b–e was examined to test
the scope and selectivity of the reaction (Scheme 1). In
each case, treatment with 1.5–2.2 equiv of LiHBR3 led
to clean monodealkylation with high yield, and pure
products were obtained through a simple workup
involving only repeated evaporation from methanol to
remove borates and protonation via aqueous extraction
or ion exchange.
Monodealkylation of phosphonic diesters is commonly
effected with heteroatom-based nucleophilic reagents in
the absence of a strong Lewis acid. Methyl and benzyl
esters are cleaved most easily, and reagents used for
cleaving these groups include sodium iodide in refluxing
acetone or 2-butanone,14 lithium bromide in acetoni-
trile,15 tert-butylamine,16 quinuclidine or DABCO in
refluxing toluene,17 and potassium cyanide in DMF at
70 °C.18 With these reagents, it is often possible to cleave
a methyl group preferentially over benzyl16,18 or ethyl,19
and potassium cyanide appears to be effective only on
methyl groups.
Lithium triethylborohydride has been noted as an
exceptionally potent SN2 nucleophile that rapidly re-
duces primary alkyl sulfonates and halides.10 Consistent
with our postulate of an SN2 mechanism for the phos-
phonate dealkylation, the selectivity for methyl over
ethyl, and ethyl over isopropyl, was complete as judged
Cleavage of ethyl groups requires more potent nucleo-
philes, higher temperatures, or both. Sodium thiophen-
oxide and thioethoxide in ethanol at 70 °C are
effective,20 and refluxing morpholine has been used with
one substrate.21 More commonly employed are alkali
metal halides, such as lithium bromide in higher ketone
solvents (e.g., 2-hexanone or 2-pentanone at 80–
110 °C)22 or refluxing pyridine.23 At 100 °C in DMF,
both iodide and azide (as their lithium or sodium salts)
are effective, and azide also cleaves isopropyl groups.24
1
by H NMR, while selectivity for benzyl over ethyl was
94/6 with both reagents.
1
Observations from preliminary H NMR experiments
are also consistent with an SN2 mechanism. When the
dealkylation of 5a with LiHBEt3 was performed in a
The nucleophilicity of lithium trialkylborohydrides is
such that dealkylations proceed quickly to completion
at room temperature even with ethyl phosphonoesters
and modest substrate concentrations (e.g., 0.1–0.2 M).
For preparative applications, this high reactivity may
be advantageous with refractory substrates and when
short reaction times or lower reaction temperatures are
desired. The use of THF in place of more toxic, high-
er-boiling solvents may also be a benefit in some cases.
High reactivity is also the principal drawback of the
reagents, as it makes them incompatible with easily
reduced groups. In other respects, such as high yield,
selectivity, and simplicity of workup, the lithium trial-
kylborohydride procedure compares favorably with the
alternatives.
Scheme 1. Monodealkylation of phosphonic diesters. aSelectivities
were the same for LiHBEt3 and LiHB(s-Bu)3. bIsolated yields (6 + 7)
for LiHBEt3 and, in parentheses, LiHB(s-Bu)3.
Scheme 2. Synthesis of phosphonic analogs of GABA. Reagents and
conditions: (a) LiHBEt3, THF, rt, 90% (R = Me), 90% (R = Bn); (b)
BocNH(CH2)2OH, EtO2CN@NCO2Et, PPh3, 75% (R = Me), 94%
(R = Bn); (c) LiHB(s-Bu)3 THF, rt, 87% (R = Me), 74% (R = Bn); (d)
TFA, 92% (4a) 94% (4b).
To demonstrate the suitability of the dealkylation reac-
tion for slightly more complex substrates in a multi-step
reaction sequence, we employed it in the synthesis of two

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3745–3748
3747
2-aminoethyl alkylphosphonate analogs (4a and 4b,
Scheme 2) of GABA which had not been studied as
GABA effectors. Our approach exploited the high selec-
tivity of the reaction for methyl over primary substitu-
ents to allow preparation of the targets from readily
available dimethyl phosphonates. These were first
mono-demethylated with LiHBEt3 and then realkylated
via the Mitsunobu reaction25 with BocNH(CH2)2OH.
The remaining methyl groups were cleaved (with 63%
attack at the primary carbon) using LiH(s-Bu)3, and
the Boc group was removed by treatment with TFA to
afford 4a and 4b in 92 and 94% yields, respectively.
The stability of the Boc urethane to the hydride reagent
may result from protective deprotonation at nitrogen.
Figure 3. Dose–response curves of antagonists versus 1 lM GABA at
homopentameric q1 GABAC receptors expressed in X. laevis oocytes.
Data are shown for TPMPA (squares), 4a (circles), and 2c (triangles),
and reflect the averages obtained from two determinations in each of
6–8 oocytes. Curves show the best fit of the Hill equation to the data.
The biological activity of the compounds was assessed
by measuring chloride ion currents at homopentameric
human q1 GABAC receptors expressed in Xenopus laevis
oocytes. Neither compound activated the receptor, and
benzylphosphonate 4b was also inactive as an antago-
nist. However, methylphosphonate 4a proved to be a
full antagonist of GABA-induced currents, reducing
them in a dose-dependent fashion to near-baseline val-
ues (Fig. 2). Hill analysis of averaged data from six dif-
ferent oocytes led to the determination of an IC50 of
11.1 lM and a Hill coefficient nH of 1.94 (Fig. 3).
was also tested at heteropentameric (a1b2c2) GABAA
receptors expressed in oocytes. At 400 lM, it neither
activated the receptor nor substantially reduced the cur-
rent elicited by 40 lM GABA in four different oocytes.
The closest analogs of 4a which have been studied at
GABAC receptors are (3-aminopropyl)methylphosphi-
nic acid 2a (IC50 = 0.75 lM, Kb = 0.58 lM)5d and
phosphonic acid 3 (Kb = 10 lM).5a The closest analog
of 4b is phosphinic acid 2b, which is likewise inactive
at GABAC.5d Comparison with the present results sug-
gests that structure–activity relationships at GABAC
are conserved between the isosteric 3-aminopropyl phos-
phinates and 2-aminoethyl phosphonates. Almost all
known phosphonic analogs of GABA incorporate ter-
minal phosphono groups. We have found only one other
example of a 2-aminoethyl phosphonate that has been
studied at GABA receptors. Cates et al. demonstrated
that phenylphosphonic acid derivative 4c weakly inhib-
ited the binding of [3H]GABA to GABAA and GABAB
receptors but did not assess effects (agonism or antago-
nism) on receptor function.27
For reference, two well-characterized GABAC receptor
antagonists, (1,2,5,6-tetrahydropyridine-4-yl)methylphos-
phinic acid (TPMPA) and (3-aminopropyl)-n-butylphos-
phinic acid (2c), were examined under the same
conditions. The IC50 values determined for these
compounds were in good agreement with literature
values (IC50 of 0.67 lM vs. a reported binding constant
Kb of 2.1 lM for TPMPA5b and IC50 of 68.2 lM vs. a
reported value of 62.5 lM for 2c5d).26 Compound 4a
Because of their potent activity, the 2-aminoethyl phos-
phonates are a promising new class of GABAC antago-
nists. In this regard, the ease with which additional
analogs with varying side chains at phosphorus can be
made from readily available phosphonic diesters facili-
tates a further exploration of structure–activity relation-
ships. The novel dealkylation described should prove of
value in the synthesis of these and other asymmetrically
substituted phosphonic acid derivatives.
Figure 2. Representative responses of a X. laevis oocyte expressing
homopentameric human q1 GABAC receptors to 1 lM GABA in the
presence of different concentrations of 4a. Oocytes were superfused
with Ringer solution at a flow rate of approximately 1 mL/min while
chloride ion currents across the membrane were monitored with a two-
microelectrode voltage clamp. Over a fixed period, denoted by the bar
labeled Application of Compounds, the superfusing medium was
switched to Ringer solution supplemented with compounds as
indicated. After this treatment period, the superfusing medium was
switched back to Ringer solution, leading to washout of the
compounds and recovery of the membrane current, as denoted by
the bar labeled Washout/Recovery.
Acknowledgments
´ `
We thank Drs. Haohua Qian and Helene A. Gussin for
helpful discussions, Ms. Tiffany C. Chen for technical
assistance and Dr. John S. Harwood for assistance with
the pressure-tube NMR experiments. This research was
supported by Grants EY013693, EY016094, and
EY001792 from the National Institutes of Health; by
an unrestricted departmental award from Research to

3748
S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3745–3748
6. Horner, L.; Hoffmann, H.; Beck, P. Chem. Ber. 1958, 91,
1583.
7. Wetzel, R. B.; Kenyon, G. L. J. Am. Chem. Soc. 1972, 94,
1774.
8. Wetzel, R. B.; Kenyon, G. L. J. Org. Chem. 1974, 39,
1531.
9. Qiu, W.; Burton, D. J. Tetrahedron Lett. 1996, 37, 2745.
10. (a) Brown, H. C.; Krishnamurthy, S. J. Am. Chem.
Soc. 1973, 95, 1669; (b) Krishnamurthy, S.; Brown, H. C.
J. Org. Chem. 1976, 41, 3064.
11. (a) Gauvry, N.; Mortier, J. Synthesis 2001, 553; (b)
Mortier, J. Rec. Res. Dev. Org. Chem. 2002, 6, 169.
12. (a) McKenna, C. E.; Higa, M. T.; Cheung, N. H.;
McKenna, M. C. Tetrahedron Lett. 1977, 155; (b) Gross,
H.; Boeck, C.; Costisella, B.; Gloede, J. J. Prakt. Chem.
1978, 320, 344; (c) Morita, T.; Okamoto, Y.; Sakurai, H.
Tetrahedron Lett. 1978, 2523; (d) McKenna, C. E.;
Schmidhauser, J. Chem. Commun. (Cambridge, U.K.)
1979, 739; (e) Blackburn, G. M.; Ingleson, D. J. Chem.
Soc., Perkin Trans. 1 1980, 1150; (f) Morita, T.; Okamoto,
Y.; Sakurai, H. Bull. Chem. Soc. Jpn. 1981, 54, 267; (g)
Issleib, K.; Balszuweit, A.; Stiebitz, B. Z. Anorg. Allg.
Chem. 1987, 546, 147.
Prevent Blindness (New York, NY); by a Macular
Degeneration Research Grant from the American
Health Assistance Foundation (Clarksburg, MD); by a
grant from the Daniel F. and Ada L. Rice Foundation
(Skokie, IL); by a Grant in Aid of Research from Sigma
Xi; by the Laboratory Directed Research and Develop-
ment Program of Oak Ridge National Laboratory,
managed by UT-Battelle, LLC, for the U.S. Department
of Energy under Contract No. DE-AC05-00OR22725;
and by the University of Illinois at Chicago. D.R.P. is
a Senior Scientific Investigator of Research to Prevent
Blindness.
Supplementary data
Experimental procedures, 1H NMR spectra for all
products from Schemes 1 and 2, and representative
electrophysiological recordings (for TPMPA and 2c at
GABAC receptors along with 4a at GABAA and
GABAC receptors). Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.bmcl.2007.04.026.
13. Keizer, T. S.; De Pue, L. J.; Parkin, S.; Atwood, D. A.
J. Am. Chem. Soc. 2002, 124, 1864.
14. (a) O’Brien, T. A.; Kluger, R.; Pike, D. C.; Gennis, R. B.
Biochim. Biophys. Acta 1980, 613, 10; (b) Hoffmann, M.
J. Prakt. Chem. 1988, 330, 820; (c) Jacques, J.; Leclercq,
M.; Brienne, M.-J. Tetrahedron Lett. 1981, 37, 1727; (d)
Boutevin, B.; Hervaud, Y.; Jeanmaire, T.; Boulahna, A.;
Elasri, M. Phosphorus Sulfur Silicon Relat. Elem. 2001,
174, 1.
15. Karaman, R.; Goldblum, A.; Breuer, E. J. Chem. Soc.,
Perkin Trans. 1 1989, 765.
16. Gray, M. D. M.; Smith, D. J. H. Tetrahedron Lett. 1980,
21, 859.
17. Benzyl esters: (a) Saady, M.; Valleix, A.; Lebeau, L.;
Mioskowski, C. J. Org. Chem. 1995, 60, 3685; Methyl and
p-nitrobenzyl esters: (b) Hoffmann, M. Phosphorus Sulfur
Silicon Relat. Elem. 1998, 134, 109; Carbomethoxymethyl
esters: (c) Smith, W. W.; Bartlett, P. A. J. Am. Chem. Soc.
1998, 120, 4622.
18. Saady, M.; Lebeau, L.; Mioskowski, C. Helv. Chim. Acta
1995, 78, 670.
19. Taylor, S. D.; Mirzaei, F.; Bearne, S. L. Org. Lett. 2006, 8,
4243.
References and notes
1. Reviews: (a) Bormann, J. Trends Pharmacol. Sci. 2000, 21,
16; (b) Chebib, M.; Johnston, G. A. R. Clin. Exp.
Pharmacol. Physiol. 1999, 26, 937.
2. Recent reviews: (a) Ong, J.; Kerr, D. I. CNS Drug Rev.
2005, 11, 317; (b) Marshall, F. H. J. Mol. Neurosci. 2005,
26, 169; (c) Johnston, G. A. Curr. Pharm. Des. 2005, 11,
1867; (d) Iversen, L. Biochem. Pharmacol. 2004, 68, 1537;
(e) Whiting, P. J. Curr. Opin. Drug Discov. Dev. 2003, 6,
648; (f) Vacher, C. M.; Bettler, B. Curr. Drug Targets CNS
Neurol. Disord. 2003, 2, 248; (g) Ashton, H.; Young, A. H.
J. Psychopharmacol. 2003, 17, 174; (h) Krogsgaard-
Larsen, P.; Frølund, B.; Liljefors, T. Chem. Rec. 2002, 2,
419; (i) Frølund, B.; Ebert, B.; Kristiansen, U.; Liljefors,
T.; Krogsgaard-Larsen, P. Curr. Top. Med. Chem. 2002, 2,
817.
3. Recent reviews (a) Chebib, M. Clin. Exp. Pharmacol.
Physiol. 2004, 31, 800; (b) Johnston, G. A.; Chebib, M.;
Hanrahan, J. R.; Mewett, K. N. Curr. Drug Targets CNS
Neurol. Disord. 2003, 2, 260; (c) Johnston, G. A. Curr.
Top. Med. Chem. 2002, 2, 903; (d) Zhang, D.; Pan, Z. H.;
Awobuluyi, M.; Lipton, S. A. Trends Pharmacol. Sci.
2001, 22, 121.
20. Savignac, P.; Lavielle, G. Bull. Soc. Chim. Fr. 1974, 7–8,
1506; For related findings, see: Takeuchi, Y.; Demachi, Y.;
Yoshii, E. Tetrahedron Lett. 1979, 1231.
21. Reddy, G. V.; Jacobs, H. K.; Gopalan, A. S.; Barrans, R.
E.; Dietz, M. L.; Stepinski, D. C.; Herlinger, A. W. Synth.
Commun. 2004, 34, 331.
4. Recent reviews (a) Yang, X. L. Prog. Neurobiol. 2004, 73,
127; (b) Lukasiewicz, P. D.; Eggers, E. D.; Sagdullaev, B.
T.; McCall, M. A. Vision Res. 2004, 44, 3289; (c) Qian, H.;
Ripps, H. Prog. Brain Res. 2001, 131, 295; (d) Enz, R.
Biol. Chem. 2001, 382, 1111.
5. (a) Woodward, R. M.; Polenzani, L.; Miledi, R. Mol.
Pharmacol. 1993, 43, 609; (b) Murata, Y.; Woodward,
R. M.; Miledi, R.; Overman, L. E. Bioorg. Med. Chem.
Lett. 1996, 6, 2073; (c) Ragozzino, D.; Woodward, R.
M.; Murata, Y.; Eusebi, F.; Overman, L. E.; Miledi, R.
Mol. Pharmacol. 1996, 50, 1024; (d) Chebib, M.;
Vandenberg, R. J.; Froestl, W.; Johnston, G. A. R.
Eur. J. Pharmacol. 1997, 329, 223; (e) Chebib, M.;
Mewett, K. N.; Johnston, G. A. R. Eur. J. Pharmacol.
1998, 357, 227.
22. (a) Krawczyk, H. Synth. Commun. 1997, 27, 3151; (b)
Sikora, D.; Nonas, T.; Gajda, T. Tetrahedron 2001, 57,
1619.
23. Hutter, D.; Blaettler, M. O.; Benner, S. A. Helv. Chim.
Acta 2002, 85, 2777.
24. Holy´, A. Synthesis 1998, 381.
25. (a) Mitsunobu, O. Synthesis 1981, 1; (b) Campbell, D. A.
J. Org. Chem. 1992, 57, 6331.
26. IC50 and Kb are comparable though not identical mea-
sures. Among four phosphinic acid GABAC antagonists
for which both values were determined, they were the same
within a factor of two (Ref. 5d).
27. Cates, L. A.; Li, V. S.; Yakshe, C. C.; Fadeyi, M. O.;
Andree, T. H.; Karbon, E. W.; Enna, S. J. J. Med. Chem.
1984, 27, 654.