Difluoromethylene analogues of aspartyl phosphate: The first synthetic inhibitors of aspartate semi-aldehyde dehydrogenase

Article abstract of DOI:10.1039/b103988c

The difluoromethylene analogue of aspartyl phosphate 6 has been prepared by the fluoride catalysed coupling of diethyl trimethyisilyldifluoromethyl phosphonate with an appropriate aldehyde followed by Dess-Martin oxidation and deprotection; the deprotected compound inhibited (K₁ 95 μM) aspartate semi-aldehyde dehydrogenase, a key enzyme involved in bacterial amino acid and peptidoglycan biosynthesis.

Full text of DOI:10.1039/b103988c

View Article Online / Journal Homepage / Table of Contents for this issue
Difluoromethylene analogues of aspartyl phosphate: the first synthetic
inhibitors of aspartate semi-aldehyde dehydrogenase
Russell J. Cox,*a Andrea T. Hadfieldb and M. Belén Mayo-Martína
a School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, UK.
E-mail: r.j.cox@bris.ac.uk
b Department of Biochemistry, University of Bristol, Bristol BS8 1TD, UK
Received (in Cambridge, UK) 3rd May 2001, Accepted 12th July 2001
First published as an Advance Article on the web 10th August 2001
The difluoromethylene analogue of aspartyl phosphate 6 has
been prepared by the fluoride catalysed coupling of diethyl
trimethylsilyldifluoromethyl phosphonate with an appro-
priate aldehyde followed by Dess–Martin oxidation and
deprotection; the deprotected compound inhibited (KI 95
mM) aspartate semi-aldehyde dehydrogenase, a key enzyme
involved in bacterial amino acid and peptidoglycan bio-
synthesis.
Difluoromethylene phosphonates have found use as phos-
phate mimics in numerous applications.7 The difluoromethy-
lene unit effectively mimics oxygen and the pKa of difluoro-
methylene phosphonates closely matches that of the analogous
phosphates.8 Compounds such as 6 could form effective
inhibitors of ASA-DH. Additionally the difluoromethylene a to
the carbonyl would have the effect of making it an excellent
electrophile and likely target for thiol attachment.
Berkowitz has reacted difluoromethylene anions such as 7
Bacterial resistance to current classes of antibacterial com-
pounds is increasing alarmingly.1,2 Many successful antibiotics
target bacterial cell wall biosynthesis, and the efforts of our
group have concentrated on attempts to design and synthesise
inhibitors of enzymes from this pathway.3 We have focussed on
with methyl esters to prepare ketones directly.9 We attempted to
follow this methodology for the synthesis of 6 (Scheme 2). -
L
Aspartic acid 8 was selectively protected at the g-ester using
methanol and thionyl chloride to form the mono ester 9. BOC
protection to give 10 followed by formation of the a-tert-butyl
ester then gave the required methyl ester 11. However this ester
was unreactive under the reaction conditions of Berkowitz9 and
we considered that the mono-protected amine could be the
source of an acidic proton. Formation of the doubly BOC
protected aspartate 12 was achieved, but this also could not be
transformed to the required phosphonate. The bis(methyl ester)
13 was also synthesised by parallel methodology, but was
similarly unreactive towards anion 7.
inhibitors of
L
-lysine 1 biosynthesis since
L
-lysine and its
precursor, diaminopimelic acid (DAP), form the strength
bearing cross-links of the peptidoglycan layer of bacterial cell
walls (Scheme 1).4 One of the earliest enzymes on the pathway,
aspartate semi-aldehyde dehydrogenase (ASA-DH, EC
1.2.1.11) forms the key intermediate
L
-aspartate semi-aldehyde
-methionine
2 the precursor of DAP and also -threonine 3 and
L
L
4. Inhibitors of ASA-DH would therefore block dual key
bacterial metabolic processes forming proteins and peptido-
glycan.
We next attempted to add 7 to aldehydes (Scheme 3).10
Treatment of 13 with DIBAL-H yielded a mixture of the
corresponding alcohol 14 (22%) and aldehyde 15 (68%);11 14
could be converted to 15 by treatment with the Dess–Martin
periodinane.12,13 When 15 was treated with excess 7 (ca. 7.5
eq.) at 278 °C we were able to isolate the mixture of
diastereomers 16a and 16b in very low yield—no products were
The mechanism of ASA-DH is believed to involve an active
site nucleophilic cysteine.5 g-Aspartyl phosphate 5, is bound at
the active site. Thiolester formation then occurs with concomi-
tant loss of phosphate. NADPH then transfers its 4-proS hydride
to reduce the thiolester, releasing the aspartate semi-aldehyde 2
and regenerating the nucleophile (Scheme 1). The 3D crystal
structure of ASA-DH has recently been solved, but as no
substrates were bound in the structure little information is
available regarding the active site geometry and potential
binding sites.6 We decided to synthesise specific ASA-DH
inhibitors. Ideally these compounds would mimic the natural
substrate and contain an electrophillic centre for covalent
attachment to the active site nucleophile.
Scheme 2 Reagents and conditions: (i) SOCl2, MeOH, 210 °C, 78%; (ii)
BOC2O, NaHCO3–MeOH, 75%; (iii) (CH3)3COH, DMAP, EDCI, 64%;
(iv) 7, THF, 278 °C–0 °C; (v) NaH, BOC2O, THF, D, 86%; (vi) LDA, 278
°C–0 °C.
Scheme 1 Bacterial pathways to cell wall and protein components.
1710
Chem. Commun., 2001, 1710–1711
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b103988c

View Article Online
Scheme 3 Reagents and conditions: (i) DIBAL-H, THF, 278 °C, 68%; (ii)
Dess–Martin periodinane, CH2Cl2, RT, 75%; (iii) (EtO)2P(O)CF2SiMe3,
THF, 10 mol% TBAF, 260 °C RT, 55%; (iv) Dess–Martin periodinane,
CH2Cl2, RT, 69%; (v) TMSI, then aq. KOH, then Dowex AG50 WX8,
95%.
formed with lower excesses or stoichiometric amounts of 7.
Under milder conditions treatment of 15 with (EtO)2P(O)CF2-
SiMe3 in THF at 0 °C in the presence of a catalytic fluoride
source such as CsF, TBAF or TBAT (Bu4N+SiPh3F22) yielded
the required products. The most favourable conditions required
the use of 1.3 eq. of (EtO)2P(O)CF2SiMe3, 10 mol% TBAF and
one equivalent of 15. This reaction yielded a mixture of
diastereomers 16a (2S,4S) and 16b (2S,4R)14 in 3+1 ratio in
overall 55% (Scheme 3) with the remaining mass balance being
unreacted aldehyde. The alcohols were smoothly converted to
the ketone 17. A two-step deprotection involving treatment of
17 with 5 eq. of TMSI (Me3SiI) followed by KOH treatment to
remove the methyl ester yielded 6 in excellent yield as a 45+55
mixture of keto- and hydrate forms.15 After purification by ion
exchange chromatography, 6 was used in inhibition assays
using recombinant E. coli ASA-DH.
Fig. 1 A. Inhibition of ASA-DH by preincubation with 6 at 0.66mM. 1
uninhibited reaction; + preincubation with 6 for 0 min; 8 preincubation
with 6 for 12 min; 5 preincubation with 6 for 25 min; - preincubation with
6 for 68 min; X preincubation with 6 for 138 min. B. Rate of inhibition of
ASA-DH at varying 6 concentration (1 0.66; 8 2.5; - 5.0 mM) and at 5
2.5 mM 6 in the presence of phosphate (15 mM).
site geometry of ASA-DH. The inhibited enzyme is currently
undergoing crystallisation trials.
We thank Alessia Rossi for technical assistance and the
University of Bristol for financial support (University Scholar-
ship to MBM-M).
Notes and references
The ASA-DH assay (Scheme 4) requires that the enzyme be
run in the ‘reverse’ direction as aspartyl phosphate is unavaila-
ble whereas aspartate semi-aldehyde is obtained by reductive
ozonolysis of allyl glycine in 1 M aqueous HCl. In the assay,
reaction is observed at 340 nm as NADPH is formed from
NADP+ (150 mM). In order to ensure satisfactory reaction rates
the other substrates (PO432 and ASA) must be present in large
excess (15 and 0.35 mM respectively). In initial inhibition
reactions under these conditions 6 showed little observable
effect. However in pre-incubation reactions in which ASA-DH
was incubated with 6 prior to addition of the other assay
components clear inhibition was observable varying with time
(Fig. 1A) and concentration (Fig 1B). Based on this assay a KI
of 95 mM was measured.16 The rate of inhibition of ASA-DH by
6 was diminished in the presence of phosphate (Fig. 1B)
indicating that inhibition occurs at the active site. The inhibition
appears to be slowly reversible as indicated by the regeneration
of activity upon dilution of the inhibited enzyme into the assay
cuvette. This behaviour is consistent with that expected for a
slow binding model of inhibition and is most likely caused by
covalent bond formation between 6 and the active site thiol of
ASA-DH.
1 House of Lords Select Committee on Science and Technology, Seventh
Report, 1998.
2 The Management and Control of Hospital Acquired Infections in Acute
NHS Trusts in England and Wales Report by the Comptroller and
Auditor General, House of Commons, 14th February 2000.
3 R. J. Cox, A. Sutherland and J. C. Vederas, Bioorg. Med. Chem., 2000,
8, 843.
4 R. J. Cox, Nat. Prod. Rep., 1996, 13(1), 29.
5 W. E. Karsten and R. E. Viola, Biochim. Biophys. Acta, 1991, 1077,
209.
6 A. Hadfield, G. Kryger, J. Ouyang, G. A. Petsko, D. Ringe and R. Viola,
J. Mol. Biol., 1999, 289(4), 991.
7 M. J. Tozer and T. F. Herpin, Tetrahedron, 1996, 52(26), 8619.
8 G. M. Blackburn, D. E. Kent and F. Kolkmann, J. Chem. Soc., Perkin
Trans. 1, 1984, 1119.
9 D. B. Berkowitz, M. Eggen, Q. Shen and R. K. Shoemaker, J. Org.
Chem., 1996, 61, 4666.
10 S. F. Martin, D. W. Dean and A. S. Wagman, Tetrahedron Lett., 1992,
33, 1839.
11 A. Sutherland, J. F. Caplan and J. C. Vederas, J. Chem. Soc., Chem.
Commun., 1999, 555.
12 D. B. Dess and J. C. Martin, J. Am. Chem. Soc., 1991, 113, 7277.
13 R. E. Ireland and L. B. Liu, J. Org. Chem., 1993, 58(10), 2899.
14 R. J. Cox, M. Murray and M. B. Mayo-Martin, unpublished.
3
15 Selected data for 6. Hydrate: dH(400 MHz, D2O) 4.05 (1 H, dd, JHH
4.1, 8.05, aCH), 2.35 (1H, dd, 2JHH 15.4, 3JHH 4.4, bCH), 2.17 (1H, dd,
3
2
2JHH 15.4, JHH 8.2, bCH); dF(283 MHz, D2O) 2122.6 (1F, dd, JFF
301.8, 2JFP 90.6), 2123.5 (1F, dd, 2JFF 300.0, 2JFP 90.3); dP(122 MHz,
D2O) 3.26 (dd, 2JPF 86.6, 89.1). Ketone: dH(400 MHz, D2O) 4.05 (1H,
dd, 3JHH 4.02, 8.05, aCH), 3.39 (1H, dd, 2JHH 19.8, 3JHH 4.1, bCH),
3.32 (1H, dd, 2JHH 20.1, 3JHH 7.8, bCH); dF(283 MHz, D2O) 2120.2 (br
d, 2JFP 84.1); dP(122 MHz, D2O) 0.67 (t, 2JPF 85.3); m/z (ESMS,
CH3CN–H2O) 247.9 ([M]H+, 100%), 265.9 ([M + H2O]H+, 75%),
270.0 ([M]Na+, 50%), 289.9 ([M + CH3CN]H+, 45%), 494.8 ([2M]H+,
20%).
Scheme 4 Assay of ASA-DH. In the reverse reaction phosphate and ASA
are present in large excess.
Thus we have shown that a rational approach can success-
fully be used for the design and synthesis of ASA-DH
inhibitors. The route should enable the synthesis of a range of
related compounds designed to probe the mechanism and active
16 Fundamentals of Enzyme Kinetics, ed. A. Cornish-Bowden, Portland
Press, London, 1995.
Chem. Commun., 2001, 1710–1711
1711