α-Fluorinated phosphonates as substrate mimics for glucose 6-phosphate dehydrogenase: The CHF stereochemistry matters

Article abstract of DOI:10.1021/jo000220v

Reported is a systematic study of the 'fitness' (in terms of k(cat)/K(m)) of a series of phosphonate mimics of glucose 6-phosphate (G6P) as unnatural substrates for G6P dehydrogenase from Leuconostoc mesenteroides. The four G6P analogues (9, 10, 15a, and 15b) differ only in the degree of fluorination at the 'bridging' phosphonate carbon. All have been synthesized from benzyl 6-O-trifluoromethane-sulfonyl-2,3,4-tri-O-benzyl β-D-glucopyranoside (6). The phosphonates with bridging CH₂ (9) and CF₂ (10) groups are cleanly obtained by direct displacements with the appropriate LiX₂CP(O)(OEt)₂ reagents (X = H, F) in 15 min at -78 °C. For the (α-monofluoro)alkylphosphonates (15a/b), homologation of 6 is achieved via lithiodithiane-mediated triflate displacement, followed by aldehyde unmasking [CaCO₃, Hg(ClO₄)₂, H₂O]. Addition of diethyl phosphite anion produces diastereomeric, (α-hydroxy)phosphonates 13a/b (1.4:1 ratio) which may be readily separated by chromatography. The stereochemistry of the minor diastereomer was established as 7(S) via X-ray crystallographic structure determination of its p-bromobenzoate derivative, 16b. Treatment of the major 7(R) diastereomer with DAST produces α-fluorinated phosphonate 14a, in modest yield, with inversion of configuration, as established, again, by X-ray crystallography. To our knowledge, this is first example of DAST-mediated fluorination of a (nonbenzylic, nonpropargylic) secondary (α-hydroxy)-phosphonate and thus establishes the stereochemical course of this transformation. α-Deprotonation/kinetic quenching of 14a provides access to the 7(R)-epimer (14b). For all four protected phosphonates (7, 8, 14a, and 14b), diethyl phosphonate ester deprotection was carried out with TMSBr, followed by global hydrogenolytic debenzylation to produce the free phosphonates, as α/β anomeric mixtures. Titrations of G6P itself and the free phosphonic acids provides second pK(a) values of 6.5 (1, bridging-O), 5.4 (10, bridging-CF₂), 6.2 (14a, bridging-CHF), and 7.6 (9, bridging-CH₂). Leuconostoc mesenteroides G6PDH-mediated oxidation and Lineweaver-Burk analysis yields normalized k(cat)/K(m) values of 0.043 (14b, bridging-7(R)-CHF), 0.11 (10, bridging-CF₂), 0.23 (14b, bridging-CH₂), and 0.46 (14a, bridging-7(S)-CHF) relative to G6P itself, largely reflecting differences in K(m). The fact that k(cat)/K(m) increases by more than an order of magnitude in going from the 7(R)-α-monofluoroalkyl phosphonate (worst substrate) to the 7(S)-diastereomer (best substrate) is especially notable and is discussed in the context of the known phosphate binding pocket of this enzyme as revealed by X-ray crystallography (Adams, M. J. et al. Structure 1994, 2, 1073-1087).

Full text of DOI:10.1021/jo000220v

4498
J . Org. Chem. 2000, 65, 4498-4508
r-F lu or in a ted P h osp h on a tes a s Su bstr a te Mim ics for Glu cose
6-P h osp h a te Deh yd r ogen a se: th e CHF Ster eoch em istr y Ma tter s
David B. Berkowitz,* Mohua Bose, Travis J . Pfannenstiel, and Tzanko Doukov
Department of Chemistry, University of Nebraska, Lincoln, Nebraska 68588-0304
dbb@unlserve.unl.edu
Received February 16, 2000
Reported is a systematic study of the “fitness” (in terms of kcat/Km) of a series of phosphonate mimics
of glucose 6-phosphate (G6P) as unnatural substrates for G6P dehydrogenase from Leuconostoc
mesenteroides. The four G6P analogues (9, 10, 15a , and 15b) differ only in the degree of fluorination
at the “bridging” phosphonate carbon. All have been synthesized from benzyl 6-O-trifluoromethane-
sulfonyl-2,3,4-tri-O-benzyl â-D-glucopyranoside (6). The phosphonates with bridging CH2 (9) and
CF2 (10) groups are cleanly obtained by direct displacements with the appropriate LiX2CP(O)-
(OEt)2 reagents (X ) H, F) in 15 min at -78 °C. For the (R-monofluoro)alkylphosphonates (15a /b),
homologation of 6 is achieved via lithiodithiane-mediated triflate displacement, followed by aldehyde
unmasking [CaCO3, Hg(ClO4)2, H2O]. Addition of diethyl phosphite anion produces diastereomeric,
(R-hydroxy)phosphonates 13a /b (1.4:1 ratio) which may be readily separated by chromatography.
The stereochemistry of the minor diastereomer was established as 7(S) via X-ray crystallographic
structure determination of its p-bromobenzoate derivative, 16b. Treatment of the major 7(R)
diastereomer with DAST produces R-fluorinated phosphonate 14a , in modest yield, with inversion
of configuration, as established, again, by X-ray crystallography. To our knowledge, this is first
example of DAST-mediated fluorination of a (nonbenzylic, nonpropargylic) secondary (R-hydroxy)-
phosphonate and thus establishes the stereochemical course of this transformation. R-Deprotonation/
kinetic quenching of 14a provides access to the 7(R)-epimer (14b). For all four protected
phosphonates (7, 8, 14a , and 14b), diethyl phosphonate ester deprotection was carried out with
TMSBr, followed by global hydrogenolytic debenzylation to produce the free phosphonates, as R/â
anomeric mixtures. Titrations of G6P itself and the free phosphonic acids provides second pKa values
of 6.5 (1, bridging-O), 5.4 (10, bridging-CF2), 6.2 (14a , bridging-CHF), and 7.6 (9, bridging-CH2).
Leuconostoc mesenteroides G6PDH-mediated oxidation and Lineweaver-Burk analysis yields
normalized kcat/Km values of 0.043 (14b, bridging-7(R)-CHF), 0.11 (10, bridging-CF2), 0.23 (14b,
bridging-CH2), and 0.46 (14a , bridging-7(S)-CHF) relative to G6P itself, largely reflecting differences
in Km. The fact that kcat/Km increases by more than an order of magnitude in going from the 7(R)-
R-monofluoroalkyl phosphonate (worst substrate) to the 7(S)-diastereomer (best substrate) is
especially notable and is discussed in the context of the known phosphate binding pocket of this
enzyme as revealed by X-ray crystallography (Adams, M. J . et al. Structure 1994, 2, 1073-1087).
In tr od u ction
Simple phosphonates, in which the labile P-O-C bond
is replaced by a stable P-CH2-C bond, have been
explored as phosphate surrogates for some time now.2
Yet, new members of this parent phosphonate class
continue to appear on a regular basis and often emerge
as useful bioorganic tools.3 A more recent effort, inspired
by Blackburn,4 has been in the area of R-halogenated
phosphonates, as potentially better “isosteric and isopo-
lar” mimics of the phosphates themselves. Particular
attention has been given to the R,R-difluorinated phos-
phonates, in recent years.4-6 Among the parameters that
potentially favor R-fluorinated phosphonates7 over their
nonfluorinated congeners are (a) reduced pKa2, (b) in-
creased C-CF2-P dihedral angle, (c) increased polarity
of the bridging group, and perhaps even (d) the possibility
for C-F‚‚‚H-X hydrogen bonding.8,9 Sterics, on the other
hand, tend to favor the simple phosphonate as literally
Phosphate esters and anhydrides are among the most
common functional groups found in Nature’s repertoire,
particularly as displayed in her metabolic intermediates.1
Among other features, the phosphate ester provides a
convenient handle for substrate binding that has pre-
sumably led to the evolution of the many well-defined
phosphate binding pockets. An important challenge for
the bioorganic/medicinal chemist remains, however, to
develop phosphate mimics of natural metabolites that
retain high affinity for targeted enzymatic phosphate
binding pockets, but are themselves resistant to phos-
phatase-mediated cleavage. As part of a project directed
toward this goal, we report here the synthesis and
enzymatic characterization of a series of (fluorinated)
phosphonate analogues of D-glucose 6-phosphate.
(2) For reviews on the synthesis and study of phosphonates, see:
(a) Wiemer, D. F. Tetrahedron 1997, 53, 16609-16644. (b) Engel, R.
Chem. Rev. 1977, 77, 349-367.
(1) For a thoughtful discussion on “Why Nature chose phosphates?”
see: Westheimer, F. H. Science 1987, 235, 1173-1178.
10.1021/jo000220v CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/01/2000

R-Fluorinated Phosphonates
J . Org. Chem., Vol. 65, No. 15, 2000 4499
the best isostere. (In R-fluorinated phosphonates, the
C-F bond is typically 1.4-1.5 Å, 30-50% longer than
the corresponding C-H bond.) Indeed, several instances
have now been documented in which R,R-difluorinated
phosphonates behave as excellent pyrophosphate10 or
phosphate ester mimics11 in target enzyme active sites.
of these studies were individual (CHF)P stereoisomers
studied, although O’Hagan reports differential reactivity
(based upon changes in the 19F NMR spectrum of the
mixture) of his diastereomeric (though unassigned) glyc-
erol 3PDH substrates.13b
Although a fair number of (R-monofluoro)alkylphos-
phonates have been reported,12-14 the biological activity
of this class of phosphate mimics remains much less
explored than that of either their simple phosphonate or
R,R-difluorinated congeners. This is surprising in light
of their iso-acidity: the second pKa of the CHF phospho-
nate is often nearly identical to that of the phosphate
group (pKa ≈ 6.4).4 We are aware of only a handful of
studies in which phosphonates have been compared, as
enzyme substrates or inhibitors, across the entire range
of R-fluorination {(CH2)Pf(CHF)Pf(CF2)P}.13 In none
(3) For recent examples, see: (a) Vidil, C.; More`re, A.; Garcia, M.;
Barragan, V.; Hamdaoui, B.; Rochefort, H.; Montero, J .-L. Eur. J . Org.
Chem. 1999, 447-450 (mannose 6P analogues as ligands for the M6P/
IGF II receptor). (b) Viera de Almeida, M.; Cleophax, J .; Gateau-
Olesker, A.; Prestat, G.; Dubreuil, D.; Gero, S. D. Tetrahedron 1999,
55, 12997-13010 (inositol 1,4,5-triphosphate analogues). (c) Billault,
I.; Vasella, A. Helv. Chim. Acta 1999, 82, 1137-1149 (glucosyl
transferase inhibitors). (d) Kakinuma, H.; Shimazaki, K.; Takahashi,
N.; Takahashi, K.; Niihata, S.; Aoki, Y.; Hamada, K.; Matsushita, H.;
Nishi, Y. Tetrahedron 1999, 55, 25559-2572 (phosphonate transition
state analogues for esterolytic, catalytic Ab development). (e) Sach-
patzidis, A.; Dealwis, C.; Lubetsky, J . B.; Liang, P.-H.; Anderson, K.
S.; Lolis, E. Biochemistry 1999, 38, 12665-12674 (transition state
analogues for the trp synthase R-reaction/X-ray studies). (f) Arni, R.
K.; Watanabe, L.; Ward, R. J .; Kreitman, R. J .; Kumar, K.; Walz, F.
G., J r. Biochemistry 1999, 38, 2452-2461 (RNAse T1 substrate
analogue/X-ray studies). (g) Bedard, J .; May, S.; Lis, M.; Tryphonas,
L.; Drach, J .; Huffman, J .; Sidwell, R.; Chan, L.; Bowlin, T.; Rando, R.
Antimicrob. Agents Chemotherapy 1999, 43, 557-567 (anti-HCMV
agents).
G6PDH occupies an important place in metabolism and
in human health. The enzyme catalyzes the committing
step to the pentose phosphate shunt, through which
ribose 5-phosphate and NADPH reducing equivalents are
produced. The former is involved in nucleic acid biosyn-
thesis and the latter in the production of reduced energy
stores, particularly fatty acids. But, perhaps most im-
portantly, in cells lacking mitochondria, such as red blood
cells, G6PDH deficiency or modification (∼400 variants
are known in humans) alters the cellular redox balance
and can lead to anemia,15a,b osteoarthritis,15c vulnerability
to oxidative stress,15d and even teratogenesis.15e Males
are particularly susceptible as the G6PDH gene is carried
on the X-chromosome.
(4) Blackburn, G. M. Chem. Ind. (London) 1981, 134-138. (b)
Blackburn, G. M.; Kent, D. E.; Kolkmann, F. J . Chem. Soc., Perkin
Trans. 1 1984, 1119-1125.
As our model, we chose to study G6PDH from Leu-
conostoc mesenteroides, as this enzyme has been well-
(5) For reviews that include discussions of the R-fluorinated phos-
phonates, see: (a) Tozer, M. J .; Herpin, T. F. Tetrahedron 1996, 52,
8619-8683. (b) Burton, D. J .; Yang, Z.-Y.; Qiu, W. Chem. Rev. 1996,
96, 1641-1715. (c) Welch, J . T. Fluorine in Bioorganic Chemistry;
Wiley: New York, 1991.
(6) For recent leading references, see: Berkowitz, D. B.; Bhuniya,
D.; Peris, G. Tetrahedron Lett. 1999, 40, 1869-1872.
(12) For examples of (monofluoro)alkylphosphonate analogues of
biological phosphates, see: (a) Wnuk; S. F.; Robins, M. J . J . Am. Chem.
Soc. 1996, 118, 2519-2520 (UMP). (b) Chen, W.; Flavin, M. T.; Filler,
R.; Xu, Z.-Q. Tetrahedron Lett. 1996, 37, 8975-8978 (acyclic phospho-
nate nucleosides). (c) Yokomatsu, T.; Yamagishi, T.; Matsumoto, K.;
Shibuya, S. Tetrahedron 1996, 52, 11725-11738 (pTyr). (d) Burke, T.
R., J r.; Smyth, M. S.; Nomizu, M.; Otaka, A.; Roller, P. P. J . Org. Chem.
1993, 58, 1336-1340 (pTyr). (e) Blackburn, G. M.; Langston, S. P.
Tetrahedron Lett. 1991, 32, 6425-6428 (dTTP, dATP.; R-â bridge).
(f) Campbell, A. S.; Thatcher, G. R. J . Tetrahedron Lett. 1991, 32,
2207-2210 (inositol phosphate). (g) Blackburn, G. M.; Guo, M.-J .;
Langston, S. P.; Taylor, G. E. Tetrahedron Lett. 1990, 31, 5637-5640
(Ap3A). (h) Blackburn, G. M.; Rashid, A. J . C. S. Chem. Commun. 1988,
317-319 (2-phosphoglycerate). (i) Blackburn, G. M.; Perre´e, T. D.;
Rashid; A.; Bisbal; C.; Lebleu, B. Chem. Scripta 1986, 26, 21-24
(glyceraldehyde 3-phosphate; 3-phosphoglycerate, TMP). (j) Blackburn,
G. M.; Eckstein, F.; Kent, D. E.; Perre´e, T. D. Nucleosides Nucleotides
1985, 4, 165-167 (ATP; â-γ bridge). (k) McKenna, C. E.; Shen, P. J .
Org. Chem. 1981, 46, 4573-4576 (pyrophosphate). (l) Blackburn, G.
M.; England, D. A.; Kolkmann, F. J . Chem. Soc. Chem. Commun. 1981,
930-932 (pyrophosphate).
(13) In related work, Blackburn has recently introduced the novel
tris(phosphonomethyl) group as a supercharged, surrogate for the
pyrophosphate group: (a) Liu, X.; Zhang, X.-R.; Blackburn, G. M. J .
Chem. Soc. Chem. Commun. 1997, 87-88 (ATP). (b) Liu, X.; Adams,
H.; Blackburn, G. M. J . C. S. Chem. Commun. 1998, 2619-2620
(pyrophosphate; includes the R-fluorinated analogue).
(14) (a) Caplan, N. A.; Pogson, C. I.; Hayes, D. J .; Blackburn, G. M.
Bioorg. Med. Chem. Lett. 1998, 8, 515-520 [(1, 3-bis-phosphonom-
ethyl)arene inhibitors of phosphoglycerate kinase]. (b) Nieschalk, J .;
Batsanov, A. S.; O’Hagan, D.; Howard, J . A. K. Tetrahedron 1996, 52,
165-176 (glycerol 3-phosphate). (c) Burke, T. R., J r.; Smyth, M. S.;
Nomizu, M.; Otaka, A.; Nomizu, M.; Roller, P. P.; Wolf, G.; Case, R.;
Shoelson, S. E. Biochemistry 1994, 33, 6490-6494 (pTyr). (d) For an
example in the PCXYP series (pyrophosphate analogues), see: Gura-
nowski, A.; Starzynska, E.; Gzik, L.; Langston, S. P.; Brown, P.;
Blackburn, G. M. Nucleosides Nucleotides 1995, 14, 731-734 (Ap3A
analogues as substrates/inhibitors of enzymatic hydrolysis or phos-
phorolysis).
(7) For discussions on the possible effects of fluorination R to
phosphorus, see: (a) O’Hagan, D.; Rzepa, H. S. J . C. S. Chem.
Commun. 1997, 645-652. (b) Chen, Li.; Wu, Li.; Otaka, A.; Smyth,
M. S.; Roller, P. P.; Burke, T. R., J r.; den Hertog, J .; Zhang, Z.-Y.
Biochem. Biophys. Res. Commun. 1995, 216, 976-984. (c) Thatcher,
G. R. J .; Campbell, A. S. J . Org. Chem. 1993, 58, 2272-2281.
(8) For examples of C-F‚‚‚H-X hydrogen bonds based upon X-ray
crystal structures, see: (a) Brady, K.; Wei, D.; Ringe, D.; Abeles, R.
H. Biochemistry 1990, 29, 7600-7607. (b) Takahashi, L. H.; Radhakrish-
nan, R.; Rosenfield, R. E., J r.; Meyer, E. F., J r.; Trainor, D. A. J . Am.
Chem. Soc. 1989, 111, 3368-3374. (c) Murray-Rust, P.; Stallings, W.
C.; Monti, C. T.; Preston R. K.; Glusker, J . P. J . Am. Chem. Soc. 1983,
105, 3206-3214.
(9) For recent discussions of the possibility of C-F‚‚‚H-X hydrogen
bonds in conjunction with surveys of crystallographic databases, see:
(a) Dunitz, J . D.; Taylor, R. Chem. Eur. J . 1997, 3, 89-98. (b) Plenio,
H.; Diodone, R. Chem. Ber./ Recueil 1997, 130, 633-640. (c) Howard,
J . A. K.; Hoy, V. J .; O’Hagan, D.; Smith, G. T. Tetrahedron 1996, 52,
12613-12622.
(10) (a) Arabshahi, L.; Khan, N. N.; Butler, M.; Noonan, T.; Brown,
N. C.; Wright, G. E. Biochemistry 1990, 29, 8620-8626 [â,γ-CF2-dGTP
as a substrate for Bacillus subtilis DNA pol III; Km < Km(dGTP); (b)
Stremler, K. E.; Poulter, C. D. J . Am. Chem. Soc. 1987, 109, 5542-
5544 (geranyl diphosphate analogues as FDP synthase pseudo-
substrates).
(11) (a) Halazy, S.; Ehrhard, A.; Eggenspiller, A.; Berges-Gross, V.;
Danzin, C. Tetrahedron 1996, 51, 177-184 (PNP inhibitors). (b) Burke,
T. R., J r.; Kole, H. K.; Roller, P. P. Biochem. Biophys. Res. Commun.
1994, 204, 129-134. (c) Kole, H. K.; Smyth, M. S.; Russ, P. L.; Burke,
T. R., J r. Biochem. J . 1995, 311, 1025-1031 (PTP inhibitors). (d)
Phillion, D. P.; Cleary, D. G. J . Org. Chem. 1992, 57, 2763-2764 (EPSP
synthase inactivator). (e) Hebel, D.; Kirk, K. L.; Kinjo, J .; Kovacs, T.;
Lesiak, K.; Balzarini, J .; De Clercq, E.; Torrence, P. F. Bioorg. Med.
Chem. Lett. 1991, 1, 357-360 (RT inhibitors).

4500 J . Org. Chem., Vol. 65, No. 15, 2000
Berkowitz et al.
F igu r e 1. Active site of Leuconostoc mesenteroides G6PDH (PDB: 1DPG; subunit A).
studied16 and a crystal structure is now available.17
Although no cocrystal structure with G6P (or any other
carbohydrate substrate mimic) is yet available, G6PDH
readily incorporates inorganic phosphate into its active
site under the conditions of enzyme crystallization. As
can be seen in Figure 1, the Adams structure17a reveals
a very well defined phosphate binding pocket surround-
ing this conserved phosphate. Specific phosphate contacts
include a likely electrostatic interaction with His-178 and
presumed H-bonds with Tyr-415 and Glu-147, as well as
with the backbone carbonyl of Ile-176. There is also a
water of crystallization occupying a portion of the prob-
able sugar binding pocket.
Sch em e 1. Syn th esis of P ivota l Tr ifla te 6a
aReagents: (a) TBDPSCl, imidazole (98%); (b) NaH, BnBr (87%);
(c) TBAF, THF (88%); (d) Tf2O, DTBMP (99%).
tematically evaluate phosphonate binding affinity, across
the entire range of R-fluorination. Our synthetic routes
to all four target phosphonates (bridging CH2,18 (R)-CHF,
(S)-CHF, and CF2) pass through glucopyranose triflate
6, taking advantage of our triflate displacement meth-
odology.19 This triflate is readily prepared in high yield
from benzyl â-D-glucopyranoside20 (Scheme 1). The TB-
DPS group figures prominently in this sequence as it very
selectively protects the lone primary hydroxyl group,21
among four total, in 4.
Resu lts a n d Discu ssion
This clearly defined phosphate binding pocket seemed
a most appropriate model receptor with which to sys-
(15) (a) Tian, W.-N.; Braunstein, L. D.; Apse, K.; Pang, J .; Rose, M.;
Tian, X.; Stanton, R. C. Am. J . Physiol. 1999, 276, C1121-C1131. (b)
Venkataram, M. N.; Al-Suwaid, A. R. Int. J . Dermatol. 1999, 38, 730-
734. (c) Fioravanti, A.; Marcolongo, R.; Carcassi, A. J . Rheumatol. 1999,
26, 1425. (d) Roos, D.; van Zwieten, R.; Wijnen, J . T.; Go´mez-Gallego,
F.; de Boer, M.; Stevens, D.; Pronk-Admiraal, de Rijk, T.; van Noorden,
C. J . F.; Weening, R. S.; Vulliamy, T. J .; Ploem, J . E.; Mason, P. J .;
Bautista, J . M.; Khan, P. M.; Beutler, E. Blood 1999, 94, 2955-2962.
(e) Nicol, C. J .; Zielinski, J .; Tsui, L.-C.; Wells, P. G. FASEB J . 2000,
14, 111-127.
Construction of the C6-C7 bond is the key step in
convergent syntheses of both the nonfluorinated phos-
phonate and its R,R-difluorinated congener (Scheme 2).
Thus, displacement of triflate 6 with either (dieth-
ylphosphonomethyl)lithium (n-BuLi as base) or [dieth-
ylphosphono(difluoromethyl)]lithium (LDA as base) pro-
ceeds cleanly in 74-83% yield within 15 min at -78 °C.
(16) (a) Levy, H. R. Biochem. Soc. Trans. 1989, 17, 313-315. (b)
Viola, R. E. Arch. Biochem. Biophys. 1984, 228, 415-424. (c) Levy, H.
R.; Christoff, M.; Ingulli, J .; Ho, E. M. L. Arch. Biochem. Biophys. 1983,
222, 473-488. (d) Hermes, J . D.; Roeske, C. A.; O’Leary, M. H.; Cleland,
W. W. Biochemistry 1982, 21, 5106-5114.
(17) (a) Rowland, P.; Basak, A. K.; Gover, S.; Levy, H. R.; Adams,
M. J . Structure 1994, 2, 1073-1087. (b) Cosgrove, M. S.; Naylor, C.;
Paludan, S.; Adams, M. J .; Levy, H. R. Biochemistry 1998, 37, 2257-
2267.
(18) For a previous synthesis and characterization of the simple,
nonfluorinated analogue of G6P, see: Adams, P. R.; Harrison, R.; Inch,
T. D.; Rich, P. Biochem. J . 1976, 155, 1-4 {(Km(G6P) ) 46 µM; Km-
(CH2-phosphonate) ) 192 µM; Vrel ) 0.46 at pH 8 with yeast G6PDH}.
(19) Berkowitz, D. B., Eggen, M.; Shen, Q.; Sloss, D. G. J . Org. Chem.
1993, 58, 6174-6176.
(20) Slotta, K. H.; Heller, H. Chem. Ber. 1930, 63, 1024-1028.
(21) Hanessian, S.; Lavallee, P. Can. J . Chem. 1975, 53, 2975-2977.

R-Fluorinated Phosphonates
J . Org. Chem., Vol. 65, No. 15, 2000 4501
stereochemistry at the center R to phosphorus.24 Subse-
quently, the DAST reaction was repeated under exactly
the same conditions with the individual diastereomers
13a and 13b. While 13a gives 14a in yields comparable
to those observed with the diastereomeric mixture, 13b
decomposes under the reaction conditions.
Sch em e 2. Tr ifla te Disp la cem en t En tr y to
P h osp h on o-Su ga r sa
To investigate the stereochemical course of this DAST-
mediated transformation, we were able to establish the
stereochemistry of the starting alcohols 13a /b, by obtain-
ing an X-ray crystal structure of 16b, the p-bromoben-
zoate ester of (R-hydroxy)phosphonate 13b (the minor
diastereomer).25 The structure reveals that 13b possesses
the 7(S) stereochemistry. This means that 13a must have
the 7(R) stereochemistry, and therefore its transforma-
tion to 14a proceeds with inversion of configuration,
albeit in modest yield.
a
Reagents: (a) MeP(O)(OEt)2, BuLi (74%); (b) HF2CP(O)(OEt)2,
LDA (83%); (c) TMSBr, CH2Cl2; then (d) H2, Pd(OH)2/C (69% for
9; 91% for 10; 2 steps).
(24) Crystals of 14a are at -75 ( 2 °C, triclinic, space group P1-
C1 (No. 1) with a ) 8.5887(7) Å, b ) 14.5270(11) Å, c ) 16.9171(13)
1
Å, R ) 109.512(2)°, â ) 97.058(2)°, γ ) 107.051(2)°, V ) 1844.1(2) Å3
and Z ) 2 molecules {dcalcd ) 1.248 g cm-3; µa(Mo KR) ) 0.130 µµ-1}.
A full hemisphere of diffracted intensities (omega scan width of 0.20°)
was measured for a two-domain nonmerohedrally twinned specimen
using graphite-monochromated Mo KR radiation on a Bruker SMART
CCD Single-Crystal Diffraction System. The two domains were related
by a 180° rotation about the a axis and the major domain accounted
for 74.6(2)% of the total sample volume. X-rays were provided by a
normal-focus sealed X-ray tube operated at 50 kV and 40 mA. The
lattice constants reported herein were determined with the Bruker
SAINT software package using peak centers for 1659 reflections of the
major domain. Similar lattice constants were obtained for the minor
domain. Data frames were integrated with the Bruker program SAINT
using the orientation matrixes for both domains and the resulting
19526 reflections (∼9760 with 2Θ(Mo KR) < 50.28° for each domain)
were then merged with the Bruker program TWHKL to eliminate 4439
reflections which were only partially overlapped for both both domains.
A total of 9584 reflections from both domains were totally overlapped,
and 5500 were not overlapped. These reflections were then combined
to give a final data set which contained 10292 independent reflections;
4792 of these were totally overlapped pairs of reflections and 5500 were
nonoverlapped reflections from both domains. This data set was
eventually used to simultaneously refine the common structure for both
domains which is reported herein. The Siemens/Bruker SHELXTL-
PC software package was used to solve the structure using direct
methods techniques with the data set obtained from frame integration
for the major domain. This data set was discarded after using it to
solve the structure and perform several cycles of preliminary least-
squares structure refinement. All stages of subsequent weighted full-
Our approach to the diastereomeric, R-monofluorinated
phosphonates involves initial terminal sugar homologa-
tion through coupling of triflate 6 with lithiodithiane22
(Scheme 3). Aldehyde unmasking proceeds smoothly, via
selective S,S-acetal cleavage in the presence of the
anomeric O,O-acetal, to provide 12. The Pudovik reac-
tion23 is then employed to fashion the C7-P bond. Diethyl
phosphite anion adds into aldehyde 12 at -78 °C to
produce a pair of chromatographically separable, (R-
hydroxy)phosphonates, 13a /b in excellent overall yield.
A slight preference (1.4:1) in favor of the 7(R)-diastere-
omer is observed.
Sch em e 3. In sta lla tion of th e r-F lu or op h osp h on o
F u n ction a litya
2
matrix least-squares refinement were conducted using Fo data with
the SHELXTL-PC Version 5 software package for nonoverlapping
reflections of individual domains or totally overlapping and nonover-
lapping reflections of both domains. Final agreement factors at
convergence for refinement of both domains simultaneously are: R1-
(unweighted, based on F) ) 0.080 for 4918 independent reflections
having 2Θ (Mo KR) < 50.28° and I > 2σ(I); wR2 (weighted, based on
F2) ) 0.235 for 8523 independent reflections having 2Θ(Mo KR) <
50.28° and I > 0. Final agreement factors for all of the data are R1-
(unweighted, based on F) ) 0.203 and wR2 (weighted, based on F2) )
0.312 for all 10 292 independent reflections having 2Θ (Mo KR) <
50.28°. Parallel refinements with ∼2750 nonoverlapping reflections for
each individual domain gave similar (although less precise) structural
parameters. The structural model incorporated isotropic thermal
parameters for all hydrogen atoms as well as carbon atom C14 which
could not be refined satisfactorily as an unrestrained anisotropic atom.
The remaining nonhydrogen atoms were incorporated into the struc-
tural model with anisotropic thermal parameters. Hydrogen atoms
were located from difference Fourier syntheses and included in the
structural model as fixed isotropic atoms (using idealized sp2- or sp3-
hybridized geometry and C-H bond lengths of 0.95-1.00 Å) riding on
their respective carbon atoms. Methyl groups were given idealized
staggered orientations. The isotropic thermal parameters of all hy-
drogen atoms were fixed at values 1.2 (nonmethyl) or 1.5 (methyl) times
the equivalent isotropic thermal parameter of the carbon atom to which
they are covalently bonded. The phosphonate ethoxy groups for the
second crystallographically independent molecule are disordered and
one of these did not refine satisfactorily when all of its atoms were
allowed to vary as independent atoms. The O18-C48, C48-C49 ethoxy
bond lengths and the O18‚‚‚C49 separation were therefore restrained
to values which were 0.940, 1.000, and 1.633 times a common C-C
distance that was included as a free variable in the least-squares
refinement cycles. This free variable refined to a final value of 1.31(2)
Å.
a
Reagents: (a) dithiane, BuLi (94%); (b) Hg(ClO4)2, CaCO3,
THF, H2O (88%); (c) HP(O)(OEt)2, LiHMDS; THF (93%; 13a :13b
) 1.4:1); (d) DAST, CH2Cl2 (32-50%; 14a :14b ) 10:1); (e) LDA,
HOAc quench (98%; 14a :14b ) 1:1); (f) TMSBr, CH2Cl2; (g) H2,
Pd(OH)2/C (60% for 15a ; 70% for 15b; 2 steps).
Initially, the diastereomeric mixture 13a /b was treated
directly with DAST, which reproducibly gave predomi-
nantly one diastereomer (10:1 selectivity) of the corre-
sponding (R-monofluoro)phosphonate (14a ) in modest
yield. The crystal structure of 14a establishes the (S)
(22) Seebach, D.; Corey, E. J . J . Org. Chem. 1975, 40, 231-237.
(23) Pudovik,, A. N.; Konovalova, I. V. Synthesis 1979, 81-96.

4502 J . Org. Chem., Vol. 65, No. 15, 2000
Berkowitz et al.
The reaction of DAST with secondary alcohols to
produce secondary alkyl fluorides is well-known26 and
proceeds with inversion of configuration, in most cases.26b-e
This stereochemical course has been attributed to an SN2-
like mechanism wherein fluoride displaces an intermedi-
ate DAST-derived SIV-ester.26b However, overall retention
has sometimes been observed, particularly in certain
steroidal systems, presumably due either to neighboring
group participation, or to an SN1-like pathway, wherein
one face of the carbocationic intermediate is well
shielded.27a,b For other secondary alcohols in functional-
ized environments, the DAST reaction is known to be
problematic resulting in rearranged products27c or elimi-
nation.27d
panding literature on the preparation of enantiomerically
enriched (R-hydroxy)phosphonates.30
For our purposes here, the failure of 13b to undergo
R-fluorination with DAST meant that another route to
the 7(R) R-monofluorophosphonate analogue of G6P (14b)
would be needed. Low-temperature deprotonation (LDA)
of 14a , followed by a kinetic quench (HOAc) provides for
facile epimerization at the fluorine-bearing center (Scheme
3). Though an equimolar mixture of 14a and 14b is
obtained, the two diastereomers are easily separated by
flash chromatography, and, importantly, no decomposi-
tion occurs under the conditions of the epimerization.
All four protected phosphonate analogues, 7, 8, 14a ,
and 14b may be deprotected smoothly by a two-step
procedure involving phosphonate ester cleavage under
the Rabinowitz-McKenna-J ung conditions (TMSBr,
CH2Cl2),31 followed by hydrogenolysis (Pearlman’s cata-
lyst) of all benzyl ether protecting groups (Schemes 2 and
3). After considerable experimentation, the best condi-
tions for phosphonate purification were found to involve
silica gel chromatography with a mixed organic (MeCN/
i-PrOH), aqueous (ammonium bicarbonate buffer) eluent
similar to one reported by Poulter and co-workers.32 One
obtains the free phosphonates as their ammonium salts
(R/â anomeric mixtures). If the free acids are desired, they
may be obtained by a couple of cycles of lyophilization
(see Experimental Section, Titrations).
To our knowledge, this is the first reported example of
use of DAST to convert a simple, unactivated secondary
(R-hydroxy)phosphonate to the corresponding (R-fluoro)-
phosphonate. For propargylic28 and benzylic12c,d,29 (R-
hydroxy)phosphonates this transformation is known.
However, allylic systems give rearrangement (formally
an SN2′ displacement or a [3,3]-sigmatropic rearrang-
ment).29 As far as we are aware, only for benzylic
systems, has the stereochemical course of the DAST-
mediated transformation of (R-hydroxy)phosphonates to
(R-fluoro)phosphonates been investigated. Initial reports
suggested that this reaction goes with inversion of
configuration.12c However, this has been corrected through
a careful study by Shibuya which reveals that, in general,
benzylic (R-hydroxy)phosphonates give scrambling of the
R-stereochemistry upon treatment with DAST.12d
This is suggestive of an SN1-like mechanism, and
indeed, there has been speculation that, in general,
phosphoryl groups may stabilize adjacent carbocations
through resonance structures of the C+--PtO+ variety.29b
The observation of inversion of configuration in the
transformation of 13a to 14a here is suggestive of an SN2-
like pathway. However, given the failure of 13b to react
via this manifold and the limited efficiency of the
transformation, further studies are clearly needed to
establish generality for this synthetic operation. Such
studies are certainly warranted given the rapidly ex-
En zym a tic Stu d ies. Pleasingly, all four phosphonate
analogues of G6P are accepted as substrates for G6PDH
from L. mesenteroides. This enzyme utilizes either NAD+
or NADP+ as cofactor. We chose to employ the latter, and
performed a Michaelis-Menten steady-state kinetic analy-
sis under conditions of cofactor saturation. When one
superimposes the kinetic data for G6P itself and each of
the phosphonate analogues, in double-reciprocal form
(Lineweaver-Burk plots, Figure 2), one is immediately
struck by the near intersection of all lines at the y-
intercept (1/Vmax). In other words, as measured by kcat
,
we observe only very small differences in turnover
efficiency across the entire range of phosphate mimics
studied. Indeed, as can be seen from Table 1, the highest
values of kcat (CF2 analogue and G6P) are only a factor
of 1.3 greater than the lowest kcat (7(R)-CHF analogue).
(25) Preliminary results {R1 (unweighted, based on F) ) 0.166 for
2761 independent absorption-corrected reflections having 2Θ (Mo KR)
< 41.6° and I > 2σ(I) and wR2 (weighted, based on F2) ) 0.492 for all
4962 independent absorption-corrected reflections having 2Θ (Mo KR)
One reason for the similarity in kcat values observed,
is surely rooted in Viola’s finding that below pH 8.7,
product dissociation is partially rate-limiting for this
enzyme.16b Nonetheless, careful kinetic isotope effect
studies carried out by O’Leary and Cleland and co-
< 41.6°} from
a low-temperature (-80 ( 2 °C) X-ray structure
determination for small needle-shaped single crystals {a ) 30.447(2),
b ) 11.077(1), and c ) 14.636(1) Å, â ) 97.583(2)° and Z ) 4} of 16b
are shown in the Supporting Information. Since the small size and
poor quality of the present crystals prevented a high precision structure
determination, attempts are underway to grow larger and higher
quality crystals of this and closely related compounds for an eventual
high-precision X-ray structural study.
(30) For recent asymmetric approaches to R-hydroxyphosphonates,
see: (a) Sharpless, K. B.; Thomas, A. A. J . Org. Chem. 1999, 64, 8379-
8385. (b) Yamagishi, T.; Yokomatsu, T.; Suemune, K.; Shibuya, S.
Tetrahedron 1999, 55, 12125-12136. (c) Duxbury, J . P.; Cawley, A.;
Thorton-Pett, M.; Wantz, L.; Warne, J . N. D.; Greatrex, R.; Brown,
D.; Kee, T. P. Tetrahedron Lett. 1999, 40, 4403-4406. (d) Groaning,
M. D.; Rowe, B. J .; Spilling, C. D. Phosphorus, Sulfur Silicon 1999,
147, 1217. (e) Groaning, M. D.; Rowe, B. J .; Spilling, C. D. Tetrahedron
Lett. 1998, 39, 5485-5488. (f) Cravaotto, G.; Giovenzana, G. B.;
Pagliarin R.; Sisti, M. Tetrahedron Asymm. 1998, 9, 745-748. (g)
Pogatchnik, D. M.; Wiemer, D. F. Tetrahedron Lett. 1997, 38, 3495-
3498. (h) Sasai, H.; Bougauchi, M.; Arai, T.; Shibasaki, M. Tetrahedron
Lett. 1997, 38, 2717-2720. (i) Yokomatsu, T.; Yamagishi, T.; Shibuya
S. J . Chem. Soc., Perkin Trans. 1 1997, 1527-1533. (j) Drescher, M.;
Hammerschmidt, F.; Kahlig, H. Synthesis 1995, 1267-1272.
(31) (a) Rabinowitz, R. J . Org. Chem. 1963, 28, 2975-2978. (b) J ung,
M. E.; Lyster, M. A. J . Am. Chem. Soc. 1977, 99, 968-969. (c)
McKenna, C. E.; Higa, M. T.; Cheung, N. H.; McKenna, M.-C.
Tetrahedron Lett. 1977, 18, 155-158.
(26) (a) Middleton, W. J . J . Org. Chem. 1975, 40, 574-578. (b)
Tewson, T. J .; Welch, M. J . J . Org. Chem. 1978, 43, 1090-1092. (c)
Lowe, G.; Potter, B. V. L. J . Chem. Soc. Perkin Trans. 1 1980, 2029-
2032. (d) Focella, A.; Bizzarro, F.; Exon, C. Synth. Commun. 1991, 21,
2165-2170. (e) Scharer, O. D.; Verdine, G. L. J . Am. Chem. Soc. 1995,
117, 10781-10782.
(27) (a) Bird, T. G. C.; Fredericks, P. M.; J ones, E. R. H.; Meakins,
G. D. J . C. S. Chem. Commun. 1979, 65-66. (b) Rozen, S.; Faust, Y.;
Ben-Yakov, H. Tetrahedron Lett. 1979, 20, 1823-1826. (c) Albert, R.;
Dax, K.; Katzenbeisser, U.; Sterk, H.; Stu¨tz, A. E. J . Carbohydrate
Chem. 1985, 4, 521-528. (d) Pansare, S. V.; Vederas, J . C. J . Org.
Chem. 1987, 52, 4804-4810.
(28) (a) Benayoud, F.; de Mendonca D. J .; Digits, C. A.; Moniz, G.
A.; Sanders, T. C.; Hammond. G. B. J . Org. Chem. 1996, 61, 5159-
5164. (b) Sanders, T. C.; Hammond, G. B. J . Org. Chem. 1993, 58,
5598-5599.
(29) (a) Blackburn, G. M.; Kent D. E. J . Chem. Soc., Chem. Commun.
1981, 511-513. (b) Blackburn, G. M.; Kent D. E. J . Chem. Soc., Perkin
Trans. 1 1986, 913-917.
(32) Woodside, A. G.; Huang, Z.; Poulter, C. D. Org. Synth. 1987,
66, 211-219.

R-Fluorinated Phosphonates
J . Org. Chem., Vol. 65, No. 15, 2000 4503
Ta ble 1. P h osp h on a tes a s G6P DH P seu d o-Su bstr a tes
bridging
group
kcat/Km
a
substrate
Km (mM)a
kcat (s-1
)
(mM-1 s-1
)
(kcat/Km)rel
glucose 6-phosphate (1)
O
CH2
(S)-CHF
0.12 ( 0.01
0.49 ( 0.04
0.23 ( 0.02
241 ( 14
224 ( 14
212 ( 6
2.01 ( 0.17 × 103
4.57 ( 0.27 × 102
9.22 ( 0.80 × 102
8.60 ( 1.0 × 101
1.84 ( 0.12 × 102
1.00
0.23
0.46
glucose-CH2-phosphonate (9)
glucose-CHF-phosphonate (15a )
{(7S)-diastereomer}
glucose-CHF-phosphonate (15b)
{(7R)-diastereomer}
factor of 11
}
(R)-CHF
2.26 + 0.26
1.35 ( 0.06
195 ( 14
248 ( 16
0.043
0.11
glucose-CF2-phosphonate (10)
CF2
a
All assays were at pH 7.8 as described in the Experimental Section. Km and kcat values were obtained by least-squares fitting of the
1/v0 vs 1/[S] data to a line, wherein substrate concentrations have been determined by UV (observation of the NADPH formed upon
complete G6PDH-mediated oxidation to the corresponding lactone) or 19F NMR (trifluoroethanol standard) in the case of 15b. Experimental
uncertainties were determined by comparing the kinetic constants obtained from several independent runs (2-3 data sets) on the same
substrate.
Ta ble 2. p Ka Va lu es vs Degr ee of r-F lu or in a tion
bridging
group
% dianion
at pH 7.8
substrate
pKa2
1
9
15a
10
O
6.5
7.6
6.2
5.4
95
62
98
CH2
(S)-CHF
CF2
>99
The second pKa values for G6P and for each type of
phosphonate analogue were determined by titration. As
expected, R-monofluorination produces a phosphonate
functionality that is most nearly “iso-ionic” (pKa2 ) 6.2)
with the actual phosphate ester (pKa2 ) 6.5). As can be
seen in Table 2, all of the phosphonates are predomi-
nantly in their dianionic forms at the pH of the assay.
Furthermore, there is really no correlation between pKa2
and Km here. Clearly factors other than phosphonate
ionization state are contributing to the differential bind-
ing to this phosphate binding pocket at this pH.
With an eye toward stimulating discussion on possible
origins of the striking effect of CHF stereochemistry upon
enzyme binding, we elected to superimpose the crystal-
lographic coordinates of our best G6P mimic upon those
of G6PDH itself, as revealed from Adams’ original crystal
structure.17a,33 Accordingly, 15a [as approximated by the
coordinates of 14a sans benzyl ether and ethyl ester
protecting groups] was placed in the active site of G6PDH
(PDB: 1DPG, subunit A).34 The phosphorus atom and
all three oxygen atoms of the phosphonate were super-
imposed upon the corresponding atoms of the conserved
phosphate (HPO4-2000). Care was taken to locate the
â-C1-O(H) at an appropriate distance (3.4 Å, see Figure
3) from the ꢀ-nitrogen of His240, the presumed catalytic
base.17b Note that the δ-nitrogen of this same histidine
donates a hydrogen bond to Asp177 as part of the catalytic
diad that is postulated to operate for G6PDH. To accom-
modate the substrate, two waters of crystallization (H2O-
1079 and H2O-1244) were deleted. Also, the second
(nonconserved) phosphate of crystallization (HPO4-2002)
has been removed for clarity.
F igu r e 2. Superimposed Lineweaver-Burk plots for G6P (1)
and its phosphonate analogues 9, 10, 15a , and 15b as
substrates for G6PDH from L. mesenteroides.
workers on G6PDH from L. mesenteroides revealed that
kH/kD(Vmax/Km) ) 3.0 and kH/kD(Vmax) ) 1.8 at pH 8.16c All
of our kinetic experiments have been performed at pH
7.8. So, it is clear that kcat does at least partially reflect
the rate of hydride transfer here. As such, our results
indicate that alterations in the bridging group (from O
to CX2, X ) H, F) do not have a dramatic effect on L.
mesenteroides G6PDH-mediated hydride transfer. This
is interesting as the perturbations (at least in terms of
group electronegativity) are rather significant and not
particularly distant from the site of enzymatic oxidation.
After all, only a relatively short three-atom bridge (-C6-
C5-O-) separates the bridging group from C1.
On the other hand, there are notable differences in Km.
Indeed, the most striking observation is that the phos-
phate mimics with the lowest Km (7(S)-CHF, 0.23 mM)
and the highest Km (7(R)-CHF, 2.3 mM) are both mem-
bers of the (R-monofluoro)phosphonate family. These
results clearly indicate that close attention must be paid
to stereochemistry when examining this class of phos-
phonates. In other words, our data show that introduction
of a fluorine atom R to a phosphoryl group in a biological
phosphate mimic must be regarded as a vectorial opera-
tion rather than a scalar one. Furthermore, in contrast
to common practice heretofore,14 these results underscore
the advantage of examining single diastereomers when
deriving enzyme kinetic parameters for phosphate mim-
ics of this variety.
In this way, the three phosphonate oxygens retain
favorable interactions with His178 (δN), Tyr415 (OH), Ile176
(33) This original L. mesenteroides G6PDH crystal structure con-
tains inorganic phosphate bound to the sugar phosphate site, but no
nicotinamide cofactor.17a A more recent structure of the same enzyme
does include a bound NADP+.17b However, in that structure, the
nicotinamide ring of the cofactor is apparently so significantly disor-
dered that its coordinates could not be defined.
(34) The coordinates of 14a and of 1DPG were modeled in XtalView.
This software was designed by Duncan McRee and is free for nonprofit
use. See: http://www.sdsc.edu/CCMS/Packages/XTALVIEW/xtalvie-
w.html. For visualization purposes, the Insight II package (version
98.0, Molecular Simulations, Inc., San Diego, CA) was employed.

4504 J . Org. Chem., Vol. 65, No. 15, 2000
Berkowitz et al.
F igu r e 3. Compound 15a (â anomer, as derived from the X-ray coordinates of 14a ) placed in the active site of G6PDH (PDB:
1DPG; subunit A).
(backbone carbonyl) and Glu147 (CO2H). In this hypo-
thetical view of 15a bound at the G6PDH active site, the
C-F bond of this 7(S)-diastereomer could engage in a
favorable ion-dipole interaction with the ꢀ-ammonium
nitrogen of Lys148 (N-F distance ) 3.4 Å, as modeled).
This is particularly attractive in light of recent model
studies with fluorinated piperidines in which the surpris-
ing predominance of conformers with opposing axially
disposed fluorine substituents is ascribed to such ion-
dipole interactions.35,36 However, it is important to note
here that Lys148 is apparently quite mobile.37 So, a more
precise positioning of this lysine, relative to sugar phos-
ph(on)ate, must await the crystallographic solution of
either a well-ordered binary complex (E-NADP+ or E-
sugar phosphate), or perhaps, of an abortive ternary
complex [e.g., E-NADPH-sugar phosph(on)ate or
E-NADP+-phosphogluconolactone], should such be forth-
coming.
tion and C-F stereochemistry, across a complete set of
fluorinated phosphonate mimics of G6P. The syntheses
of all four phosphonates diverge from glucopyranose
6-triflate (6) and highlight the utility of the triflate
displacement approach for sugar homologation and func-
tionalization. In the course of the synthesis of 14a , it was
established that DAST-mediated conversion of a non-
benzylic, secondary (R-hydroxy)phosphonate to the (R-
monofluoro)phosphonate proceeds with inversion of con-
figuration. Steady-state enzyme kinetic analysis with L.
mesenteroides G6PDH yields kcat/Km values of 0.043 (15b,
bridging-7(R)-CHF), 0.11 (10, bridging-CF2), 0.23 (14b,
bridging-CH2), and 0.46 (15a , bridging-7(S)-CHF) relative
to G6P itself, largely reflecting differences in Km. Though
the best phosphate mimic, 15a , (Km ) 0.23 mM vs 0.12
mM for G6P) is of the iso-acidic (pKa2 ) 6.2 vs 6.5 for
G6P) (R-monofluoro)phosphonate class, so also is the
worst phosphate mimic, 15b (Km ) 2.3 mM). These
results suggest that the vectorial disposition of a C-F
bond in an enzymatic phosphate binding pocket can
contribute up to an order of magnitude in binding affinity
in (R-monofluoro)phosphonates. Clearly further such
studies are warranted to establish the generality, range,
and origin of the effects of C-F stereochemistry upon
receptor binding in this iso-acidic class of hydrolytically
stable phosphate mimics.
Con clu sion s
This report details the first study of enzyme binding
affinity/turnover, as a function of both degree of fluorina-
(35) Snyder, J . P.; Chandrakumar, N. S.; Sato, H.; Lankin, D. C. J .
Am. Chem. Soc. 2000, 122, 544-545.
(36) Several authors have recently critically examined both the CSD
(Cambridge Structural Database)9a-c and the PDB (Protein Databank)9a
and concluded that bona fide C-F‚‚‚H-N hydrogen bonds are ex-
tremely rare. Dunitz9a has pointed out, however, that systems bearing
a negative charge in the vicinity of the C-F bond (such as -SerO-
addition complexes to R-fluorinated ketonic inhibitors) are more likely
to serve as H-bond acceptors at fluorine. If indeed 15a binds to G6PDH
in its dianionic form, this factor could make such a hydrogen bond (to
Lys338, say) more favorable here.
Exp er im en ta l Section
Gen er a l. All reactions were conducted under argon atmo-
sphere using oven-dried glassware. Methylene chloride and
diisopropylamine were distilled from CaH2. THF and Et2O
were distilled from sodium benzophenone ketyl. HMPA was
distilled from Na under reduced pressure. Methanol was
distilled from Mg. n-Butyllithium in hexanes (nominally 1.6
M) was purchased from Aldrich and titrated before each use.
(37) For example, in going from subunit A to subunit B of 1DPG,
the ꢀ-nitrogen of Lys148 moves 2.9 Å. This movement is associated
with a change from a trans-Lys148-Pro149 amide bond (subunit A) to a
cis-amide bond (subunit B).

R-Fluorinated Phosphonates
J . Org. Chem., Vol. 65, No. 15, 2000 4505
Other reagents were obtained from commercial sources and
used without further purification. 31P and 19F NMR spectral
lines are with respect to the internal (capillary) standards
triphenylphosphine (-5.80 ppm) and trifluoroacetic acid (-76.5
ppm), respectively. Mass spectra were acquired at the Ne-
braska Center for Mass Spectrometry (University of Nebraska-
Lincoln). Elemental analyses were carried out by MHW Labs
(Phoenix, AZ).
1 H), 7.28 (m, 20 H); 13C NMR (125 MHz, CDCl3) δ 62.1, 71.6
74.9, 75.0, 75.1, 75.6, 77.6, 82.3, 84.6, 102.8, 127.58, 127.62,
127.82, 127.86, 127.89, 127.9, 128.0, 128.1, 128.3 (2 C), 128.5
(2 C), 137.3, 138.0, 138.3, 138.5; IR (film) 3360 (br), 3030, 2910
cm-1; HRMS (FAB, 3-NOBA/Na2CO3) calcd for C34H36O6Na (M
+ Na+) 563.2410, obsd 563.2416. Anal. Calcd for C34H36O6: C,
75.53; H, 6.71. Found: C, 75.76; H, 6.71.
Ben zyl 2,3,4-Tr i-O-ben zyl-6-O-tr iflu or om eth a n esu lfo-
n yl-â-D-glu cop yr a n osid e (6). To a solution of alcohol 17 (1.00
g, 1.9 mmol) and 2,6-di-tert-butyl-4-methylpyridine (513 mg,
2.50 mmol) in CH2Cl2 (10 mL) at -40 °C was added triflic
anhydride (0.39 mL, 2.3 mmol), dropwise, via syringe. After
being stirred for 1.5 h at -40 °C, the volatiles were evaporated,
and the residue was immediately subjected to flash chroma-
tography (30% EtOAc-hexane) to give 6 (1.23 g, 99%) as a
Ben zyl 6-O-ter t-Bu tyld ip h en ylsilyl-â-D-glu cop yr a n o-
sid e (17). To a solution of benzyl â-D-glucopyranoside 420 (10.0
g, 39 mmol) and imidazole (5.27 g, 77.4 mmol) in DMF (100
mL) at -30 °C was added dropwise, via cannula, a solution of
tert-butyldiphenylsilyl chloride (10.1 mL, 39 mmol) in DMF
(50 mL). After 3.5 h the reaction mixture was poured into
NaHCO3 (aq.) and Et2O. The aqueous layer was further
extracted with Et2O. The combined organics were dried
(MgSO4), filtered, and evaporated. Flash chromatography (5%
MeOH-CH2Cl2) provided 17 (20.0 g, 98%) as a foam: mp 51-
white solid: mp 71-72 °C; [R]22 +7.42 (c 1.00, CDCl3); 1H
D
NMR (500 MHz, (CDCl3) δ 3.45 (dd, J ) 9, 10 Hz, 1 H), 3.52
(dd, J ) 8, 9 Hz, 1 H), 3.57 (ddd, J ) 2, 6, 10 Hz, 1 H), 3.67
(app t, J ) 9 Hz, 1 H), 4.30 (dd, J ) 6, 10.5 Hz, 1 H), 4.52 (d,
J ) 8 Hz, 1 H), 4.55 (d, J ) 11 Hz, 2 H), 4.58 (dd, J ) 2, 10.5
Hz, 1 H), 4.65 (d, J ) 12 Hz, 1 H), 4.70 (d, J ) 11 Hz, 1 H),
4.77 (d, J ) 11 Hz, 1 H), 4.89 (d, J ) 12 Hz, 1 H), 4.94 (d, J )
11 Hz, 1 H), 4.95 (d, J ) 11 Hz, 1 H), 7.29 (m, 20 H); 13C NMR
(125 MHz, (CDCl3) δ 71.1, 72.3, 74.7, 74.9, 75.0, 75.7, 76.4,
82.0, 84.5, 102.1, 127.7, 127.8, 128.0, 128.1, 128.17, 128.20 (2
C), 128.3, 128.38, 128.42, 128.5, 128.6, 136.9, 137.3, 138.19,
138.22; IR (ATR) 3038, 2875 cm-1; HRMS (FAB, 3-NOBA/NaI)
calcd for C35H35SO8F3Na 695.1902, obsd 695.1903.
52 °C; [R]22 +41 (c 1.00, CHCl3); 1H NMR (500 MHz, (CDCl3)
D
δ 1.06 (s, 9 H), 2.78 (br s, 3 H), 3.41 (m, 2 H), 3.54 (t, J ) 9
Hz, 1 H), 3.61 (app t, J ) 8, 9 Hz, 1 H), 3.91 (dd, J ) 5, 11 Hz,
1 H), 3.94 (dd, J ) 5, 11 Hz, 1 H), 4.34 (d, J ) 8 Hz, 1 H), 4.55
(d, J ) 12 Hz, 1 H), 4.86 (d, J ) 12 Hz, 1 H), 7.38 (m, 11 H),
7.70 (d, J ) 8 Hz, 4 H); 13C NMR (125 MHz, CDCl3) δ 19.2,
26.8 (3 C), 64.6, 70.8, 71.9, 73.6, 75.0, 76.3, 101.2, 127.78,
127.80, 128.2, 128.5, 129.9, 132.9, 133.1, 135.6, 135.7, 136.9;
IR (film) 3380 (br), 2930, 2860 cm-1; HRMS (FAB, 3-NOBA/
LiI) calcd for C29H36SiO6Li (M + Li+) 515.2442, obsd 515.2443.
Anal. Calcd for C29H36SiO6: C, 68.47; H, 7.13. Found: C, 68.24;
H, 7.21.
Ben zyl 6-Deoxy-6-(d ieth ylp h osh on om eth yl)-2,3,4-tr i-
O-ben zyl-â-D-glu cop yr a n osid e (7). All solutions were deox-
ygenated by freezing (liquid nitrogen) and subjecting to five
cycles of evacuation and purging with Ar. To a solution of
diethyl methylphosphonate (0.32 mL, 2.22 mmol) and HMPA
(0.39 mL, 2.22 mmol) in THF (2.8 mL) at -78 °C was added
n-butyllithium (1.59 mL of a 1.4 M solution in hexane, 2.22
mmol). The resulting solution was allowed to stir for 30 min
at -78 °C. To this solution was then added, via cannula, a
precooled (-78 °C) solution of the triflate 6 (50 mg, 0.74 mmol)
in THF (4.2 mL). After 10 min at -78 °C, the reaction was
quenched by adding NH4Cl (aq.) and Et2O. The aqueous layer
was further extracted with Et2O, and the combined organic
extracts were dried (MgSO4), filtered, and evaporated. Flash
chromatography (30% EtOAc-hexane) gave 7 (369 mg, 74%)
Ben zyl 6-O-ter t-Bu tyld ip h en ylsilyl-2,3,4-tr i-O-ben zyl-
â-D-glu cop yr a n osid e (5). To a solution of triol 16 (8.11 g, 16
mmol) in DMF (47 mL) at 0 °C was added NaH (2.23 g of a
60% dispersion in mineral oil, 55.9 mmol). The reaction
mixture was allowed to reach rt over a period of 20 min. Benzyl
bromide (6.26 mL, 52.7 mmol) then was added dropwise, via
syringe. After 7 h, the reaction mixture was poured into
NaHCO3 (aq.) and EtOAc. The aqueous layer was further
extracted with EtOAc, and the combined organics were dried
(MgSO4), filtered, and evaporated. Flash chromatography (10%
Et2O-hexane) gave 5 (10.8 g, 87%) as a white solid: mp 74-
76 °C; [R]22D - 4.98 (c 1.00, CDCl3); 1H NMR (500 MHz, (CDCl3)
δ 1.14 (s, 9 H), 3.42 (dt, J ) 3, 9 Hz, 1 H), 3.62 (app t, J ) 8,
9 Hz, 1 H), 3.72 (t, J ) 9 Hz, 1 H), 3.83 (t, J ) 9 Hz, 1 H), 4.02
(s, 2 H), 4.58 (d, J ) 8 Hz, 1 H), 4.73 (d, J ) 12 Hz, 1 H), 4.75
(d, J ) 11 Hz, 1 H), 4.81 (d, J ) 11 Hz, 1 H), 4.87 (d, J ) 11
Hz, 1 H), 4.95 (d, J ) 11 Hz, 1 H), 4.98 (s, J ) 11 Hz, 1 H),
5.04 (d, J ) 12 Hz, 1 H), 5.06 (d, J ) 11 Hz, 1 H), 7.36 (m, 26
H), 7.77 (d, J ) 9 Hz, 2 H), 7.83 (d, J ) 8 Hz, 2 H); 13C NMR
(125 MHz, CDCl3) δ 19.3, 26.8 (3 C), 62.8, 70.6, 74.9, 75.1, 75.8
(2 C), 77.8, 82.6, 84.8, 102.3, 127.60, 127.61, 127.7, 127.74,
127.8, 127.87, 127.93, 127.98, 128.1, 128.3, 128.37 (2 C),
128.42, 133.2, 133.7, 134.8, 135.6, 135.9, 137.6, 138.3, 138.61,
138.64; IR (film) 3030, 2855, 2930 cm-1; HRMS (FAB, 3-NOBA/
LiI) calcd for C50H54SiO6Li (M + Li+) 785.3850, obsd 785.3857.
Anal. Calcd for C50H54SiO6: C, 77.09; H, 6.99. Found: C, 77.23;
H, 7.01.
as a white solid: mp 56-58 °C; [R]22 +2.44 (c 1.00, CHCl3);
D
1H NMR (500 MHz, CDCl3) δ 1.32 (app dt, J ) 3.5, 7 Hz, 6 H),
1.78 (m, 2 H), 1.98 (m, 1 H), 2.19 (m, 1 H), 3.28 (m, 2 H), 3.49
(dd, J ) 8, 9 Hz, 1 H), 3.63 (app t, J ) 9 Hz, 1 H), 4.09 (m, 4
H), 4.49 (d, J ) 8 Hz, 1 H), 4.62 (d, J ) 11 Hz, 1 H), 4.67 (d,
J ) 12 Hz, 1 H), 4.73 (d, J ) 11 Hz, 1 H), 4.78 (d, J ) 11 Hz,
1 H), 4.88 (d, J ) 11 Hz, 1 H), 4.92 (d, J ) 12 Hz, 1 H), 4.93
(d, J ) 11 Hz, 1 Hz),4.96 (d, J ) 1 Hz, 1 H),7.32 (m, 20 H); 13
C
NMR (125 MHz, CDCl3) δ 16.5 (br, 2 C), 21.7(d, J ) 144 Hz),
24.9 (br), 61.5 (br, 2 C), 71.3, 74.3 (d, J ) 17 Hz), 74.8,75.2,-
75.6, 81.4, 82.5, 84.6, 102.6, 127.58, 127.61, 127.79, 127.83,
127.9, 128.0, 128.1, 128.29, 128.33, 128.37, 128.4, 137.3, 138.0,
138.4, 138.5; 31P NMR (202 MHz, CDCl3) δ 31.07; IR (ATR)
2972, 2904 cm-1; HRMS (FAB, 3-NOBA/NaI) calcd for C39H47O8-
PNa (M + Na+) 697.2906, obsd 697.2914.
Ben zyl 2,3,4-Tr i-O-ben zyl-â-D-glu cop yr a n osid e (18). To
a solution of silyl ether 5 (523 mg, 0.67 mmol) in THF (5.5
mL) at 0 °C was added, dropwise, a solution of tetrabutylam-
monium fluoride (0.739 mmol) in THF (0.739 mL). After being
stirred 9 h at room temperature, the reaction was quenched
with NaHCO3 (aq.) and EtOAc. The aqueous layer was further
extracted with EtOAc. The combined organic layers were dried
(MgSO4), filtered, and evaporated. Flash chromatography (25%
EtOAc-hexane) gave 18 (321 mg, 88%) as a white solid. [Note:
Scale-up does not compromise yield. Thus, 3.90 g of 5 gives
Ben zyl 6-Deoxy-6-[(dieth ylph osph on o)diflu or om eth yl]-
2,3,4-tr i-O-ben zyl-â-D-glu cop yr a n osid e (8). All solutions
were deoxygenated as in the synthesis of 7. To a solution of
diisopropylamine (83 µL, 0.63 mmol) in THF (0.5 mL) at -78
°C was added n-butyllithium (490 µL of 1.3 M solution in
hexane, 0.63 mmol). The resulting solution was allowed to stir
for 30 min at 0 °C and then cooled to -78 °C. To this solution
of LDA at -78 °C were added, dropwise via cannula, a
precooled (-78 °C) solution of diethyl R,R-difluoromethylphos-
phonate [99 µL, 0.63 mmol (d ) 1.198)] in THF (0.25 mL) and,
2 min later, a solution of triflate 6 (85 mg, 0.13 mmol) in THF
(0.5 mL). After being stirred 15 min at -78 °C, the reaction
was quenched by addition of NH4Cl (aq.) and Et2O. The
aqueous layer was further extracted with Et2O. The combined
organics were dried (MgSO4), filtered, and evaporated. Flash
chromatography (50% EtOAc-hexane) gave 8 (73.6 mg, 83%)
as a white solid: mp 75-76 °C; [R]22D +1.6 (c 1.25, CHCl3); 1H
2.33 g of 18 (87%)] mp 102-104 °C; [R]22 -5.63 (c 1.00,
D
1
CDCl3); H NMR (300 MHz, (CDCl3) δ 1.79 (dd, J ) 6, 7 Hz,
1 H), 3.33 (m, 1 H), 3.45 (app t, J ) 9 Hz, 1 H), 3.53 (app t, J
) 9 Hz, 1 H), 3.63 (app t, J ) 9 Hz, 1 H), 3.67 (m, 1 H), 3.83
(ddd, J ) 3, 6, 12 Hz, 1 H), 4.53 (d, J ) 8 Hz, 1 H), 4.60 (d, J
) 11 Hz, 1 H), 4.66 (d, J ) 12 Hz, 1H), 4.69 (d, J ) 11 Hz,
1H), 4.77 (d, J ) 11 Hz, 1H), 4.82 (d, J ) 11 Hz 1H), 4.88 (d,
J ) 12 Hz, 1 H), 4.89 (d, J ) 10 Hz, 1 H), 4.91 (d, J ) 10 Hz,

4506 J . Org. Chem., Vol. 65, No. 15, 2000
Berkowitz et al.
NMR (500 MHz, (CDCl3) δ 1.34 (app t, J ) 7 Hz, 3 H), 1.35
(app t, J ) 7 Hz, 3 H), 2.17 (m, 1 H), 2.56 (m, 1 H), 3.27 (app
t, J ) 9 Hz, 1 H), 3.51 (app t, J ) 9 Hz, 1 H), 3.65 (app t, J )
9 Hz, 1 H), 3.75 (app t, J ) 9 Hz, 1 H), 4.24 (m, 4 H), 4.50 (d,
J ) 8 Hz, 1 H), 4.59 (d, J ) 11 Hz, 1 H), 4.66 (d, J ) 12 Hz,
1 H), 4.70 (d, J ) 11 Hz, 1 H), 4.75 (d, J ) 11 Hz, 1 H), 4.92
(m, 3 H), 4.96 (d, J ) 11 Hz, 1 H), 7.31 (m, 20 H); 13C NMR
(125 MHz, (CDCl3) δ 16.3, 16.4, 35.4-35.8 (m), 64.38, 64.42,
65.7, 68.9, 70.7, 74.7, 74.9, 80.4, 82.3, 84.7, 101.8, 116.9-122.7
(app dt, J ) 126, 261 Hz), 127.57, 127.6, 127.7, 127.8, 127.86,
127.92, 128.1, 128.2, 128.3, 128.35 (2C), 128.39, 137.3, 137.8,
138.43 (2 C); 19F NMR (188 MHz; (CDCl3) δ -113.7 to -111.6
(dd, J ) 108, 303 Hz, 1 F), -110.6 to -108.3 (ddd, J ) 22,
108, 303 Hz, 1 F); 31P NMR (81 MHz, (CDCl3) δ 6.28 (app t,
J F,P ) 108 Hz); IR (film) 2910 cm-1; HRMS (FAB, 3-NOBA/
Na2CO3) calcd for C39H45PF2O8Na (M + Na+) 733.2718, obsd
733.2726. Anal. Calcd for C39H45PF2O8: C, 65.91; H, 6.38.
Found: C, 66.13; H, 6.49.
1 M solution in THF, 1.23 mmol). The resulting solution was
stirred for 5 min at -78 °C, followed by the addition, via
cannula, of a solution of aldehyde 12 (618 mg, 1.12 mmol) in
THF (5 mL) at -78 °C. After 10 min, the reaction was
quenched with NH4Cl (aq.) and Et2O. The aqueous layer was
further extracted with Et2O. The combined organics were dried
(MgSO4), filtered, and evaporated. One obtains 13a /b (650 mg,
85% total yield) as a mixture of diastereomers (1.4:1 ratio)
separable by flash chromatography (50% EtOAc-hexane). On
a smaller scale (93 mg of 12), the same procedure gives 13a /b
in 93% (108 mg) total yield. Major diastereomer 13a (first-
eluting): mp 122-123 °C; [R]22D +2.16 (c 1.6, CHCl3); 1H NMR
(500 MHz, CDCl3) δ 1.32 (t, J ) 7 Hz, 6 H), 1.94 (ddd, J ) 9,
14.5, 21 Hz, 1 H), 2.41 (m, 1 H), 3.18 (dd, J ) 3, 16 Hz, 1 H),
3.38 (app t, J ) 9 Hz, 1 H), 3.5 (dd, J ) 8,9 Hz, 1 H), 3.58 (dt,
J ) 3, 9 Hz, 1 H), 3.62 (app t, J ) 9 Hz, 1 H), 4.18 (m, 5 H),
4.55 (d, J ) 8 Hz, 1 H), 4.63 (d, J ) 12 Hz, 1 H), 4.65 (d, J )
11 Hz, 1 H), 4.72 (d, J ) 11 Hz, 1 H), 4.78 (d, J ) 11 Hz, 1 H),
4.88 (d, J ) 12 Hz, 1 H), 4.89 (d, J ) 11 Hz, 1 H), 4.93 (d, J )
11 Hz, 1 H), 4.94 (d, J ) 11 Hz, 1 H), 7.3 (m, 20 H); 13C NMR
(125 MHz, CDCl3) δ 16.5 (br, 2 C), 33.3, 62.6 (br, 2 C), 67.2 (d,
J ) 165 Hz), 71.5, 74.8 (d, J ) 15 Hz), 74.9, 75.2, 75.7, 81.5,
82.2, 84.5, 102.7, 127.6, 127.7, 127.8 (2 C), 127.89, 127.93,
128.0, 128.1, 128.3, 128.36, 128.4, 128.5, 136.9, 137.9, 138.3,
138.4; 31P NMR (202 MHz, CDCl3) δ 22.93; IR (ATR) 3251-
3283 (br), 3030, 2863, 1218, 1061, 1020 cm-1; HRMS (FAB,
3-NOBA/NaI) calcd for C39H47O9PNa (M + Na+) 713.2855, obsd
713.2877. Minor diastereomer 13b (second-eluting): mp 120-
Ben zyl 6-Deoxy-6-(2′-d ith ia n yl)-2,3,4-tr i-O-ben zyl-â-D-
glu cop yr a n osid e (11). All solutions were deoxygenated as
in the preparation of 7. To a solution of dithiane (556 mg, 4.62
mmol) and HMPA (0.80 mL, 4.6 mmol) in THF (9 mL) at -78
°C was added n-BuLi (3.8 mL of a 1.22 M solution in hexane,
4.6 mmol). The resulting solution was allowed to stir for 5 min
at -78 °C. To this solution was then added, via cannula, a
precooled (-78 °C) solution of the triflate 6 (889 mg, 1.32
mmol) in THF (9 mL). After 10 min at -78 °C, the reaction
was quenched by adding NH4Cl (aq.) and Et2O. The aqueous
layer was further extracted with Et2O, and the combined
organic extracts were dried (MgSO4), filtered, and evaporated.
Flash chromatography (20% EtOAc-hexane) gave 11 (800 mg,
122 °C; [R]22 +5.09 (c 0.60, CHCl3); 1H NMR (500 MHz,
D
CDCl3) δ 1.32 (app dt, J ) 7, 4 Hz, 6 H), 1.83 (m, 1 H), 2.24
(m, 1 H), 2.95 (br s, 1H), 3.3 (app t, J ) 9, 10 Hz, 1 H), 3.49
(app t, J ) 9 Hz, 1 H), 3.65 (app t, J ) 9 Hz, 1 H), 3.69 (dt, J
) 2, 10 Hz, 1 H), 4.15 (dq, J ) 3, 7 Hz, 5 H), 4.54 (d, J ) 8 Hz,
1 H), 4.60 (d, J ) 11 Hz, 1 H), 4.69 (d, J ) 12 Hz, 1 H), 4.71
(d, J ) 11 Hz, 1 H), 4.76 (d, J ) 11 Hz, 1 H), 4.87 (d, J ) 11
Hz, 1 H), 4.88 (d, J ) 12 Hz, 1 H), 4.93 (app t, J ) 11 Hz, 2
H), 7.32 (m, 20 H); 13C NMR (125 MHz, (CDCl3) 16.5 (br, 2 C),
33.1, 62.6 (2 C), 64.6 (d, J ) 163 Hz), 71 (d, J ) 15 Hz), 71.6,
74.9, 75.1, 75.7, 81.3, 82.5, 84.7, 102.8, 127.6, 127.7, 127.8,
127.9, 127.96, 128.0, 128.1, 128.32, 128.37 (2 C), 128.43, 137.4,
138.0, 138.4, 138.5; 31P NMR (202 MHz, CDCl3) δ 24.23; IR
(ATR) 3293 (br), 3030, 2905 cm-1; HRMS (FAB, 3-NOBA/NaI)
calcd for C39H47O9PNa 713.2855, obsd 713.2832.
94%) as a white solid: mp 110-111 °C; [R]22 +2.67 (c 1.00,
D
1
CHCl3); H NMR (500 MHz, CDCl3) δ 1.86 (ddd, J ) 4.5, 10,
14 Hz, 1 H), 1.92 (m, 1 H), 2.08 (m, 1 H), 2.32 (ddd, J ) 2.4,-
10,14 Hz, 1 H), 2.75 (ddd, J ) 3, 10, 14 Hz, 1 H), 2.84 (m, 3
H), 3.29 (t, J ) 9 Hz, 1 H),3.5 (dd, J ) 4, 10 Hz, 1 H), 3.58 (dt,
J ) 2, 10 Hz, 1 H), 3.64 (app t, J ) 9 Hz, 1 H), 4.18 (dd, J )
4, 10 Hz, 1 H), 4.49 (d, J ) 8 Hz, 1 H), 4.62 (d, J ) 11 Hz, 1
H), 4.72 (d,J ) 11 Hz, 2 H), 4.77 (d, J ) 11 Hz, 1 H), 4.88 (d,
J ) 11 Hz, 1 H), 4.92 (d, J ) 11 Hz, 2 H), 4.96 (d, J ) 11 Hz,
1 H),7.34 (m, 20 H); 13C NMR (125 MHz, CDCl3) δ 26.0, 29.2,
29.7, 37.4, 43.3, 71.2, 74.8, 75.2, 75.7, 81.6, 82.5, 84.8, 102.4,
127.6 (2 C), 127.8, 127.9, 128.0 (2 C), 128.09, 128.1, 128.29,
128.34, 128.4 (2 C), 137.5, 138.0, 138.5, 138.6; IR (ATR) 2901,
2891 cm-1. Anal. Calcd for C38H42O5S2: C, 70.99; H, 6.58.
Found: C, 70.86; H, 6.68.
Ben zyl
6-Deoxy-6-[(d iet h ylp h osp h on o)-(S)-flu or o-
m eth yl]-2,3,4-tr i-O-ben zyl-â-D-glu cop yr a n osid e (14a ). To
a solution of DAST (0.08 mL, 0.58 mmol) in CH2Cl2 (0.24 mL)
at -78 °C was added rapidly down the sides of the flask, via
cannula, a precooled (-78 °C) solution of the diastereomeric
mixture (1.4: 1) ïf the (R-hydroxy)phosphonates 13a and 13b
(270 mg, 0.4 mmol) in CH2Cl2 (2.5 mL). The mixture was
stirred at -78 °C for 3-4 min and then raised immediately to
room temperature. After 50-60 min at room temperature, the
reaction was quenched with NaHCO3 (aq.) followed by extrac-
tion with CH2Cl2. The combined organics were dried (MgSO4),
filtered, and evaporated. 19F NMR of the crude showed that it
contained a 10:1 mixture of diastereomers. Flash chromatog-
raphy (50% EtOAc-hexane) gave 14a [88 mg, 32% (40% based
on recovered starting material)] and 14b (6 mg; 2%). The same
procedure, when repeated on a sample of pure 13a (36 mg,
0.05 mmol), gave 14a [14 mg, 40% (50% based on recovered
starting material)]. Crystals suitable for X-ray diffraction could
be obtained by recrystallization from Et2O/hexane: mp 75-
76 °C; [R]22D +12.1 (c 1.00, CHCl3); 1H NMR (500 MHz, CDCl3)
δ 1.35 (app dt, J ) 7, 5 Hz, 6H), 1.79 (br app dt, J ) 13, 44
Hz, 1 H), 2.47 (m, 1H), 3.28 (t, J ) 9 Hz, 1H), 3.5 (dd, J ) 8,
9 Hz, 2 H), 3.65 (t, J ) 9 Hz, 1 H), 4.18 (app quintet, J ) 7.5
Hz, 4 H), 4.52 (d, J ) 8 Hz, 1 H), 4.61 (d, J ) 11 Hz, 1 H), 4.69
(d, J ) 12 Hz, 1 H), 4.72 (d, J ) 11 Hz, 1 H), 4.77 (d, J ) 11
Hz, 1 H), 4.88 (d, J ) 11 Hz, 1 H), 4.89 (d, J ) 12 Hz, 1 H),
4.92 (d, J ) 11 Hz, 1 H), 4.94 (app ddt, J ) 2, 13, 47 Hz, 1 H),
4.95 (d, J ) 11 Hz, 1 H), 7.31 (m, 20 H); 13C NMR (125 MHz,
CDCl3) δ 16.45 (br, 2 C), 32.6 (br), 63.0 (br, 2 C), 69.4 (d, J )
11 Hz), 71.6, 74.9, 75.1, 75.7, 81.5, 82.4, 84.7, 102.6, 127.6 (2
C), 127.8, 127.89, 127.92, 127.98, 128.0, 128.1, 128.3, 128.36,
128.41, 128.5, 137.2, 137.9, 138.4, 138.5; 19F NMR δ -214
Ben zyl 6-Deoxy-6-for m yl-2,3,4-t r i-O-b en zyl-â-D-glu -
cop yr a n osid e (12). To a solution of the dithiane derivative
11 (964 mg, 1.50 mmol) and CaCO3 (2.7 g, 27 mmol) in THF
(70 mL) and water (15 mL) was added dropwise a 2 M aqueous
solution of Hg(ClO4)2 (1.88 mL, 3.75 mmol). After 5 h or
stirring at room temperature, the reaction mixture was diluted
with Et2O (50 mL) and filtered through a plug of neutral
alumina. The filtrate was washed with water (45 mL) and
extracted with Et2O (3 × 30 mL). The combined organics were
dried (MgSO4), filtered, and evaporated. Flash chromatography
(30% EtOAc-hexane) gave 12 (724 mg, 88%) as a white solid:
1
mp 108-109 °C; [R]22 +2.30 (c 1.00, CHCl3); H NMR (500
D
MHz, CDCl3) δ 2.82 (ddd, J ) 2, 9, 15 Hz, 1 H), 2.74 (ddd, J
) 2, 4, 12 Hz, 1 H), 3.33 (app t, J ) 9 Hz, 1 H), 3.51 (dd, J )
8, 9 Hz, 1 H), 3.66 (app t, J ) 9 Hz, 1 H), 3.83 (dt, J ) 4,9 Hz,
1 H), 4.54 (d, J ) 8 Hz, 1 H), 4.57 (d, J ) 11 Hz, 1 H),4.63 (d,
J ) 12 Hz, 1 H), 4.71 (d, J ) 11 Hz, 1 H) 4.77 (d, J ) 11 Hz,
1 H),4.85 (d, J ) 12 Hz, 1 H), 4.88 (d, J ) 11 Hz, 1 H), 4.94 (d,
J ) 11 Hz, 1 H), 4.95 (d, J ) 11 Hz, 1 H), 7.3 (m, 20 H), 9.72
(t, J ) 2 Hz, 1 H); 13C NMR δ 45.8, 70.0, 71.2, 74.9, 75.0, 75.6,
80.6, 82.4, 84.6, 102.4, 127.7 (2 C), 127.86, 127.91, 128.0,
128.07, 128.1 (2 C), 128.35, 128.4, 128.45, 128.51, 137.1, 137.7,
138.3, 138.4, 199.6; IR (ATR) 3028, 2905, 1722 cm-1. Anal.
Calcd for C35H36O6: C, 76.04; H, 6.56. Found: C, 75.90; H,
6.82.
Ben zyl
6-Deoxy-6-[d iet h ylp h osp h on o(1′-h yd r oxy)-
m et h yl]-2,3,4-t r i-O-b en zyl-â-D-glu cop yr a n osid e (13a &
13b). To a solution of diethyl phosphite (0.16 mL, 1.23 mmol)
in THF (0.7 mL) at -78 °C was added LiHMDS (1.23 mL of a

R-Fluorinated Phosphonates
J . Org. Chem., Vol. 65, No. 15, 2000 4507
(dddd, J ) 12, 44, 46, 73 Hz); 31P NMR (202 MHz, CDCl3) δ
anomer), 112.85 (dddd, J ) 10, 30, 87, 281 Hz; â anomer),
113.69 (dddd, J ) 10, 30, 87, 281 Hz; R anomer); HRMS (FAB,
TEA) calcd for C7H13O8F2P [(M-H)-] 293.0238, obsd 293.0241.
6-Deoxy-6-[(S)-m on oflu or op h osp h on om et h yl]-D-glu -
cop yr a n osid e, Bis-Am m on iu m Sa lt (15a ). From 14a (186
mg, 0.27 mmol) following the deprotection procedure for 7, was
obtained 15a (50 mg, 60%; white solid) as a mixture of
anomers: 1H NMR (300 MHz, D2O) â:R ≈ 2:1 as determined
by integration of anomeric H signals, δ 4.65 (d, J ) 8 Hz, â
anomer), 5.21 (d, J ) 4 Hz, R anomer) (see Supporting
Information for a copy of the entire NMR spectrum); 13C NMR
(75.5 MHz, D2O) δ 92.7 (R anomer), 96.6 (â anomer); 31P NMR
(121 MHz, D2O); 12.4 (d, J ) 62 Hz; â anomer), 12.9 (d, J )
62 Hz; R anomer); 19F NMR (470 MHz, D2O) δ -205.01 to
-205.77 (overlapping app ddt, J ) 12, 46, 66 Hz, 1 F from
each anomer); HRMS (FAB, TEA): calcd for C7H14O8PF (M -
H)- 275.0332, obsd 275.0336.
6-Deoxy-6-[(R)-m on oflu or op h osp h on om et h yl]-D-glu -
cop yr a n osid e, Bis-Am m on iu m Sa lt (15b). From 14b (90
mg, 0.13 mmol), following the deprotection procedure for 7,
was obtained 15b (28 mg, 70%; white solid) as a mixture of
anomers: 1H NMR (300 MHz, D2O) â:R ≈ 2:1 as determined
by integration of anomeric H signals, δ 4.66 (d, J ) 8 Hz, â
anomer), 5.25 (d, J ) 4 Hz, R anomer) (see Supporting
Information for a copy of the entire NMR spectrum); 13C NMR
(75.5 MHz, D2O) δ 92.8 (â anomer), 96.9 (R anomer); 31P NMR
(121 MHz, D2O) 12.7 (d, J ) 67 Hz, 1F; both anomers); 19F
NMR (470 MHz, D2O) δ -199.7 to -200.4 (m, 1 F from each
anomer); HRMS (FAB, TEA) calcd for C7H14O8PF [(M - H)-]
275.0332, obsd 275.0338.
17.47 (d, J
)73 Hz); IR (ATR) 2905 cm-1; HRMS (FAB,
F,P
3-NOBA/NaI) calcd for C39H46O8PFNa (M + Na+) 715.2812,
obsd 715.2816.
Ben zyl
6-d eoxy-6-[(d iet h ylp h osp h on o)-(R)-flu or o-
m eth yl]-2,3,4-tr i-O-ben zyl-â-D-glu cop yr a n osid e (14b). To
a solution of diisopropylamine (150 µL, 1.13 mmol) in THF (3
mL) at -78 °C was added n-butyllithium (1 mL of 1.11 M
solution in hexanes, 1.13 mmol). The resulting solution was
allowed to stir for 30 min at 0 °C and then cooled to -78 °C.
To this solution of LDA at -78 °C was added dropwise via
cannula a precooled (-78 °C) solution of 14a (154 mg, 0.23
mmol) in THF (1.2 mL). After 1 h, a solution of HOAc (0.13
mL, 2.25 mmol) in THF (0.52 mL) cooled to -78 °C was
cannulated into the reaction mixture. Upon an additional 10
min of stirring at that temperature, the reaction was quenched
with NaHCO3 (aq.) and Et2O. The aqueous layer was further
extracted with Et2O. The combined organics were dried
(MgSO4), filtered, and evaporated. Flash chromatography (50%
EtOAc-hexane) gave a white solid 14b [73 mg, 47% (98%
based on recovered starting material)]: [R]22 -6.82 (c 0.8,
D
1
CHCl3); H NMR (500 MHz, CDCl3) δ 1.35 (app dt, J ) 7, 8
Hz, 6 H), 2.39 (m, 2 H), 3.47 (t, J ) 9 Hz, 1 H), 3.56 (m, 2 H),
3.65 (t, J ) 9 Hz, 1 H), 4.21 (m, 4 H), 4.56 (d, J ) 8 Hz, 1 H),
4.66 (d, J ) 11 Hz, 1 H), 4.70 (d, J ) 12 Hz, 1 H), 4.75 (d, J )
11 Hz, 1 H), 4.72 (d, J ) 11 Hz, 1 H), 4.77 (d, J ) 11 Hz, 1 H),
4.88 (d, J ) 11 Hz, 1 H), 4.89 (d, J ) 12 Hz, 1 H), 4.80 (d, J )
11 Hz, 1 H), 4.93 (d, J ) 11 Hz, 1 H), 4.95 (d, J ) 12 Hz, 1 H),
4.96 (d, J ) 11 Hz, 1 H), 4.99 (d, J ) 11 Hz, 1 H), 5.12 (dddd,
J ) 3, 5, 9, 47 Hz), 7.34 (m, 20 H); 13C NMR (125 MHz, CDCl3)
δ 16.4 (br, 2 C), 32.1 (d, J ) 18 Hz), 63.0 (dd, J ) 7, 31 Hz, 2
C),71.2, 71.4 (dd, J ) 4, 11 Hz), 74.8, 75.0, 75.6, 81.2, 82.4,
84.7, 86.1 (dd, J ) 171, 180 Hz), 102.5, 127.6 (2 C), 127.7,
127.74, 127.80, 127.9 (2 C), 128.0, 128.3, 128.32 (2 C), 128.4,
137.3, 138.1, 138.4, 138.5; 19F NMR (470 MHz, CDCl3) δ
-205.5 (dddd, J ) 18, 32, 47, 73 Hz); 31P NMR (202 MHz,
CDCl3) δ 17.22 (d, J F,P ) 73 Hz); HRMS (FAB, 3-NOBA/LiI)
calcd for C39H46O8PFLi (M + Li+) 699.3062, obsd 699.3058.
6-Deoxy-6-(p h osp h on om eth yl)-D-glu cop yr a n osid e, Bis-
Am m on iu m Sa lt (9). To a stirred solution of 7 (300 mg, 0.45
mmol) in CH2Cl2 (5.2 mL) at room temperature was added
TMSBr (0.18 mL, 1.39 mmol), dropwise via syringe. The
reaction mixture was stirred at room temperature for the
required period of time (1H NMR monitoring, 2 days in this
case) and then concentrated by rotary evaporation. [Note: It
is advantageous to use an inert cap for this reaction vessel
(plastic Capplugs or glass) as common laboratory septa are
unstable to TMSBr.] Several cycles of washing (MeOH) and
evaporation then gave the free phosphonic acid. A suspension
of this crude acid and 20% Pd(OH)2/C (55 mg) in EtOAc (2.5
mL) and MeOH (2.5 mL) was shaken (Parr Hydrogenator)
under hydrogen pressure (50 psi) for 1 day. Filtration, evapo-
ration of the volatiles, and flash chromatography (50% i-PrOH/
25% CH3CN/25% 50 mM NH4HCO3) gave 9 (90 mg, 69%) as a
mixture of anomers: 1H (300 MHz, D2O) â:R ≈ 3:1 as
determined by integration of the anomeric H signals, δ 4.67
(d, J ) 8 Hz, 1 H; â anomer), 5.24 (d, J ) 4 Hz, 1 H; R anomer)
(see Supporting Information for a copy of the entire NMR
spectrum); 13C NMR (75.5 MHz, D2O, characteristic signals)
δ 24.1 (d, J ) 135 Hz; â anomer), 24.2 (d, J ) 135 Hz, R
anomer), 92.6 (R anomer), 96.6 (â anomer); 31P NMR (202 MHz,
D2O) δ 24.97 (â anomer), 25.03 (R anomer); HRMS (FAB, TEA)
calcd for C7H15O8P [(M - H)-] 257.0426, obsd 257.0418.
6-Deoxy-6-(d iflu or op h osp h on om eth yl)-D-glu cop yr a n o-
sid e, Bis-Am m on iu m Sa lt (10). From 8 (44 mg, 0.06 mmol),
following the deprotection procedure for 7, was obtained 10
(18 mg, 91%; white solid) as a mixture of anomers: 1H NMR
(500 MHz, D2O) â:R ≈ 3:1 as determined by integration of
anomeric H signals, δ 4.71 (d, J ) 8 Hz, 1 H; â anomer), 5.23
(d, J ) 4 Hz, 1 H; R anomer) (see Supporting Information for
a copy of the entire NMR spectrum); 13C NMR (125 MHz, D2O)
δ 92.7 (R anomer), 96.6 (â anomer); 31P NMR (202 MHz, D2O)
δ 4.73 (t, J ) 86 Hz; â anomer), 4.81 (t, J ) 86 Hz; R anomer);
19F NMR (470 MHz, D2O) δ -110.52 (dddd, J ) 5, 35, 87, 281
Hz; â anomer), -110.75 (dddd, J ) 5, 35, 87, 281 Hz; R
Ben zyl 6-Deoxy-6-[d iet h ylp h osp h on o(1′-(4′′-b r om o)-
ben zoyloxy)m eth yl]-2,3,4-tr i-O-ben zyl-â-D-glu cop yr a n o-
sid e (16a a n d 16b). To a solution of the diastereomeric
mixture of (R-hydroxy)phosphonates 13a and 13b (114 mg,
0.17 mmol) in THF (1.65 mL) at -78 °C was added n-BuLi
(0.14 mL of a 1.15 M solution in hexane, 0.17 mmol). After 5
min, a precooled solution of p-bromobenzoyl chloride (54.4 mg,
0.25 mmol) in THF (1 mL) was added, via cannula, and the
resulting reaction mixture was stirred at -78 °C for 30 min.
The reaction was quenched by addition of NH4Cl (aq., 10 mL)
and Et2O (10 mL). The aqueous layer was further extracted
with Et2O. The combined organics were dried (MgSO4), filtered,
and evaporated. One obtains 16a /b (126 mg, 88% total yield,
white solid) as a mixture (1.5:1) of diastereomers separable
by flash chromatography (30%f50% EtOAc-hexane). Major
diastereomer 16a (first-eluting, 76 mg): [R]22 -22.9 (c 1.00,
D
1
CHCl3); H NMR (500 MHz, CDCl3) δ 1.27 (dt, J ) 3, 7 Hz, 6
H), 2.07 (m, 1H), 2.65 (dddd, J ) 2, 5, 11, 15 Hz, 1 H), 3.27 (t,
J ) 9 Hz, 1 H), 3.46 (dd, J ) 8, 9 Hz, 1 H), 3.62 (app t, J ) 9
Hz, 1 H), 3.64 (app t, J ) 9 Hz, 1 H), 4.14 (m, 4 H), 4.41 (d, J
) 12 Hz, 1 H), 4.48 (d, J ) 8 Hz, 1 H), 4.56 (d, J ) 11 Hz, 1
H), 4.64 (d, J ) 11 Hz, 1 H), 4.67 (d, J ) 11 Hz, 1 H), 4.77 (d,
J ) 11 Hz, 1 H), 4.91 (t, J ) 11 Hz, 3 H), 5.76 (dt, J ) 5,9 Hz,
1 H), 7.11 (m, 2 H), 7.31 (m, 18 H), 7.54 (d, J ) 9 Hz, 2 H),
7.93 (d, J ) 9 Hz, 2 H); 13C NMR (125 MHz, CDCl3) δ 16.4 (d,
J ) 5 Hz), 16.5 (d, J ) 5 Hz), 32.0, 62.8 (d, J ) 20 Hz), 62.9
(d, J ) 18 Hz), 67.5 (d, J ) 167 Hz), 71.1, 73.8 (d, J ) 13 Hz),
74.7, 75.1, 75.6, 81.4, 82.3, 84.5, 102.7, 127.5, 127.6 (2 C), 127.7,
127.8, 127.9, 128.0, 128.17, 128.24 (2 C), 128.3, 128.4, 128.6,
131.3, 131.9, 137.3, 137.9, 138.4, 138.5, 164.6; 31P NMR (202
MHz, (CDCl3) δ 19.33 (s); IR (ATR) 3031, 2907, 1724 cm-1
;
HRMS (FAB, 3-NOBA/NaI) calcd for C46H50O10BrPNa (M+Na+)
895.2222, obsd 895.2244; Minor diastereomer 16b (second-
eluting, 50 mg): mp 78°-79 °C; [R]22 -7.59 (c 0.50, CHCl3);
D
1H NMR (500 MHz, CDCl3) δ 1.29 (dt, J ) 7, 13 Hz, 6 H), 2.00
(m, 1 H), 2.56 (m, 1 H), 3.21 (app t, J ) 9,10 Hz, 1 H), 3.31 (t,
J ) 9 Hz, 1 H), 3.46 (app t, J ) 8,9 Hz, 1 H), 3.54 (t, J ) 9 Hz,
1 H), 4.20 (m, 4 H), 4.30 (d, J ) 8 Hz, 1H), 4.63 (d, J ) 11 Hz,
2 H), 4.67 (d, J ) 11 Hz, 1 H), 4.71 (d, J ) 11 Hz, 1 H), 4.87
(d, J ) 11 Hz, 2H), 4.92 (d, J ) 11 Hz, 1H), 4.95 (d, J ) 12 Hz,
1 H), 5.91 (dt, J ) 2,10 Hz, 1 H), 7.34 (m, 20 H), 7.58 (d, J )
10 Hz, 2 H), 7.91 (d, J ) 10 Hz, 2 H); 13C NMR (202 MHz,
(CDCl3) δ 16.5 (br, 2 C), 31.6, 62.8 (d, J ) 9 Hz), 62.9 (d, J )
11 Hz), 64.5 (d, J ) 170 Hz), 69.7 (d, J ) 24 Hz), 71.1, 74.8,
75.3, 75.7, 81.3, 82.4, 84.6, 102.3, 127.60, 127.65, 127.69,

4508 J . Org. Chem., Vol. 65, No. 15, 2000
Berkowitz et al.
127.89, 127.95, 128.01 (2 C), 128.1, 128.22, 128.30, 128.40,
128.48, 128.7, 129.9, 131.4, 131.9, 137.5, 137.8, 138.3, 138.4,
164.6; 31P NMR (202 MHz, CDCl3) δ 20.03; IR (ATR) 2920,
2856, 1727 cm-1; HRMS (FAB, 3-NOBA/NaI) calcd for
C46H50O10BrPNa (M + Na+) 897.2202 obsd 897.2204. To
determine the relative stereochemistry at C-7, crystals of 16b
were grown from Et2O/hexane and employed for X-ray diffrac-
tion.
carried out in duplicate for phosphonates 9 (at 0.3, 0.5, 0.8,
1.0, 2.0, 3.0, 5.0 mM), 10 (at 0.3, 0.5, 0.7, 0.9, 1.5, 2.5, 3.5 mM),
15a (at 0.1, 0.2, 0.3, 0.5, 0.8, 1.0, 2.0 mM), and 15b (at 0.2,
0.3, 0.7, 1.0, 1.9, 2.9, 4.0, 5.9, 10 mM). In each case, initial
rates were estimated from the observed absorbance over the
first 5 min of reaction, for which excellent linearity was
observed. Linear least-squares fitting of 1/vo vs 1/[S] provided
Km and Vmax values.
En zym e Sta n d a r d iza tion . G6PDH solutions were stan-
dardized (in terms of U/mL) on a regular basis to ensure that
all kinetics experiments were carried out with a fixed amount
of enzyme. G6PDH (L. mesenteroides, 200 U/mg protein,
Calbiochem), â-NADP+ (sodium salt, Sigma), and â-D-glucose-
6-phosphate (monosodium salt, Aldrich) were used in the
assay. All UV absorbance data were recorded on a Shimadzu
model UV-2101PC scanning spectrophotometer. A standard
assay contained 1 mM glucose 6-phosphate and 0.6 mM
NADP+ and Tris-HCl buffer (0.1 M, pH ) 7.8) in a total volume
of 1 mL.
Con cen tr a tion Deter m in a tion s, P h osp h on a te Solu -
tion s. Concentrations of the phosphonate substrate stock
solutions (for 9, 10, 15a , and G6P itself) were determined
spectrophometrically, by complete oxidation with G6PDH and
quantitation of the NADPH thereby formed. At pH 7.8, the
reaction is functionally irreversible as the 6-phosphoglucono-
lactone enzymatic product is presumably converted to 6-phos-
phogluconate, at least on the time scale of these measurements
(ca. 1-2 h). Typically, 3 µL of the stock solution (10-75 mM)
are diluted in buffer (100 mM Tris-HCl, pH 7.8; 1 mL total
volume; 1 cm path length) containing 1 mM NADP+ and
G6PDH (L. mesenteroides, 25 mU). The total increase in optical
density at 340 nm is then monitored at 25 °C, whereby
additional enzyme (typically two 25 mU portions) is added over
the time course of the assay until the absorbance readings level
off. Division of ∆O.D.340(total) by 6.22 mM-1 gives the total
concentration of NADPH formed, reflecting the concentration
of G6P or analogue initially present. For 15b, the stock
solution concentration was determined by 19F NMR after
diluting 1:1 with D2O. A solution of trifluoroethanol in D2O
was employed as concentration standard.
p Ka Deter m in a tion s. To obtain the free phosphonic acids
from the ammonium salts of 9, 10, and 15a , two cycles of
dissolution (in distilled, deionized water), freezing and lyo-
philization were carried out. That ammonia had, in fact, been
removed by lyophilization could be seen by monitoring the
change in pH of an aqueous solution of these phosphonates
(ca. pH 8 prior to lyophilization and ca. pH 3 thereafter).
Titrations were carried out on 1 mg/mL solutions of the
phosphonic acids, by adding a solution of base (14 mM NaOH)
in 10 µL aliquots and monitoring pH with an Orion glass
electrode (model 81-15) and a Corning pH meter (model 240).
Following titration, the pH profile for each analogue was
plotted vs amount of base added, from which the second pKa
was determined.
Ack n ow led gm en t . Financial support from the
American Heart Association is gratefully acknowledged.
D.B.B. is an Alfred P. Sloan Research Fellow. T.J .P. was
a participant in the NSF REU (Research Experience for
Undergraduates) program at the University of Ne-
braska. We thank Prof. Victor W. Day for performing
the crystal structure analyses of 14a and 16b, and Prof.
J ohn J . Stezowski for access to his SGI workstation.
MariJ ean Eggen and Darby G. Sloss are acknowledged
for early experiments with 10. This research was
facilitated by grants for NMR and GC/MS instrumenta-
tion from the NIH (SIG 1-S10-RR06301) and the NSF
(CHE-93000831), respectively.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
compounds 5-12, 13a , 13b, 14a , 14b, 15a , 15b, 16a , 16b, 17
and 18, as well as perspective drawings of the solid-state
structures for 14a and 16b and a summary of crystallographic
parameters and tables of fractional coordinates, anisotropic
thermal parameters, bond angles, bond lengths and modeled
hydrogen atom coordinates for the X-ray crystal structure of
14a . This material is available free of charge via the Internet
at http://pubs.acs.org.
En zym e Kin etic An a lysis (G6P DH). The assay cuvette
(total volume 1 mL) containing Tris-HCl buffer (0.1 M, pH 7.8),
NADP+ (0.6 mM) and varying concentrations of glucose
6-phosphate (0.05, 0.1, 0.2, 0.3, 0.5, 0.8 and 1 mM) was
incubated at 25 °C for 2 min. Each reaction was then initiated
by the addition of enzyme (25 mU per cuvette). Initial velocities
were determined by monitoring ∆O.D.340 (against a buffer
blank) vs time. (Note, control experiments showed no reduction
of NADP+ in the absence of enzyme.) Each assay was per-
formed in duplicate. This same dehydrogenase assay was
J O000220V