Large scale synthesis of cyclodiphospho-D-glycerate

Full text of DOI:10.1021/jo9601472

J . Org. Chem. 1996, 61, 5697-5700
5697
Sch em e 1. Syn th esis of n -Bu tyl D-Glycer a te (5)
La r ge Sca le Syn th esis of
Cyclod ip h osp h o-D-glycer a te
fr om Ma n n itol (1)
Martyn J . Earle, Asiya Abdur-Rashid, and
Nigel D. Priestley*,†
College of Pharmacy, The Ohio State University,
Columbus, Ohio 43210-1291
Received J anuary 23, 1996
In tr od u ction
Thermophilic microorganisms require significant ad-
aptations of their cellular constituents and biochemistry
to live at high temperatures. Thermostabilizing features
in thermophiles include changes in their protein and
nucleic acid stability as well as chemical differences in
their intracellular environments. In archaeal (formerly
archaebacterial) thermophilic species that produce meth-
ane, such as the methanogen species Methanothermus
fervidus (optimal growth at 83 °C) and Methanopyrus
kandleri (optimal growth at 98 °C), very high concentra-
tions of the potassium salt of cyclo-2,3-diphospho-D-
glycerate (cDPG) appear to be important for in vivo
enzyme thermostability.1-4 In methanogen species, cDPG
is widely distributed and has been shown in vitro to
increase the half life of thermal inactivation of enzymes
purified from thermophilic species by greater than 100-
fold.2-4 Since cDPG is present at concentrations of ca.
0.3 M in M. fervidus and ca. 1.0 M in M. kandleri this
compound is likely to play a major role in the thermo-
adaptation of thermophilic methanogen species.4
cyclo-2,3-Diphospho-D-glycerate can be extracted and
purified from M. fervidus or Methanobacterium ther-
moautotrophicum.2,5 Unfortunately, fermentative pro-
duction of cDPG is not practical as the organisms are not
easy to culture on a large scale (high temperature, growth
on H2) and limited amounts of the material can be
obtained. Previous synthetic approaches to cDPG are
also unsuitable because of low yield and expense of
starting material.6,7 For a complete series of biophysical
experiments, a synthesis capable of producing multigram
quantities of the target compound was required.
mized the effort spent in purification at the intermediate
stages and avoided, as much as possible, the need for
chromatography.
Previously, cDPG has been synthesized from 2,3-
bisphosphoglycerate in aqueous media using a water
soluble carbodiimide reagent.6 The yield of the process
was only 10% and a mere 50 µmol of product was
obtained. Considering the expense of 2,3-bisphospho-
glycerate this approach was not feasible. Nobel and
Potter have recently reported a synthesis of cyclic pyro-
phosphate compounds based on desulfurization of vicinal
bisphosphothioates.8 The process was shown to work
well in model systems although the high yields of the
bisphosphate make it less suitable in the present work.
Mannitol was the obvious choice of starting material
because it is readily available and possesses the absolute
stereochemistry at C-2 and C-5 required to give an
enantiospecific synthesis of the target (Schemes 1 and
2 ). The enantiomer, cyclodiphospho-L-glycerate, can also
be prepared as the conversion of ascorbic acid into the
acetonide of L-glyceraldehyde is a known procedure.9
D-Glyceraldehyde acetonide was prepared from man-
nitol essentially by the procedures of Schmid and Bry-
ant10 and Chittenden.11 In our hands, 2 of greater than
95% purity could be obtained in 75-80% yield represent-
ing an improvement over the published procedure: the
only change being that the 2,2-dimethoxypropane was
freshly distilled before use to ensure removal of acetone
and methanol. Traces of the starting material, the
mono-, and tris-acetonide were removed from the product
2 by virtue of their differential solubility in hexanes or
acetone. Periodate oxidation10 followed by direct per-
manganate oxidation,12 without purification of the alde-
We report here an inexpensive, relatively straightfor-
ward synthesis that can be applied to the large scale
production of cDPG. We have routinely been able to
make cDPG in greater than 30% overall yield from
mannitol.
Resu lts a n d Discu ssion
We originally set ourselves the task of developing a
synthesis of cDPG that could be done on a large scale
(ca. 100 g of target compound) and that was relatively
inexpensive and simple to carry out. The main constraint
imposed was to devise a synthetic scheme which mini-
* EMail: priestle@dendrite.pharmacy.ohio-state.edu.
† This paper is dedicated to the memory of Conor R. O’Hanlon,
friend, colleague, graduate student.
(1) Seeley, R. J .; Fahrney, D. E. Curr. Microbiol. 1984, 10, 85.
(2) Hensel, R.; Ko¨nig, H. FEMS Microbiol Lett. 1988, 49, 75.
(3) Hensel, R.; J acob, I. Syst. Appl. Microbiol. 1994, 16, 242.
(4) Lehmacher, A.; Hensel, R. Mol. Gen. Genet. 1994, 242, 163.
(5) Kanodia, S.; Roberts, M. F. Proc. Natl. Acad. Sci. U.S.A. 1983,
80, 5217.
(8) Noble, N. J .; Potter, B. V. L. J . Chem. Soc., Chem. Commun.
1989, 1194.
(9) J ung, M. E.; Shaw, T. J . J . Am. Chem. Soc. 1980, 102, 6304;
Takano, S.; Numata, H.; Ogasawara, K. Heterocycles 1982, 19, 327.
(10) Schmid, C. R.; Bryant, J . D. Org. Synth. 1993, 68, 6.
(11) Chittenden, G. J . F. Carbohydr. Res. 1980, 87, 219.
(12) Tankaka, A.; Yamashita, K. Agric. Biol. Chem. 1980, 44, 199.
(6) Gorris, L. G. M.; Korteland, J .; Derksen, R. J . A. M.; Van Der
Drift, C.; Vogels, G. D. J . Chromatogr. 1990, 504, 421.
(7) Tohidi, M.; Orgal, L. E. J . Mol. Evol. 1990, 30, 97.
S0022-3263(96)00147-8 CCC: $12.00 © 1996 American Chemical Society

5698 J . Org. Chem., Vol. 61, No. 16, 1996
Notes
Sch em e 2. Syn th esis of cDP G (9) fr om n -Bu tyl
The critical procedure in the proposed synthesis was
the method chosen to introduce protected forms of
phosphate into the sequence. Various strategies were
attempted based on the use of either phosphorus(III) or
phosphorus(V) chemistry and either acid- or base-
catalyzed reactions. All trials involving either pyrophos-
phoryl chloride (Cl2P(O)OP(O)Cl2), POCl3, tetrabenzyl
pyrophosphate, or dibenzyl chlorophosphonate with 5 met
with the same fate: destruction of the carbon skeleton
of 5. With hindsight, the plethora of alternate leaving
groups and the ease with which elimination reactions
could occur doomed this particular strategy.
D-Glycer a te (5)
The same fate befell similar reactions of 1-alkyl and
1-acyl substituted glycerol derivatives.
Initially, benzyl H-phosphonate14 was evaluated as a
P(III) phosphorylation reagent; however, these efforts
met with little success for either 5 or 1-alkyl and acyl
glycerol derivatives. After much trial, dibenzyl diiso-
propylphosphoramidite15 emerged as the reagent of choice
for phosphorylation of 5. Dibenzyl diisopropylphosphor-
amidite can be made from PCl3 in two steps conveniently
on a molar scale.15-17
Optimization of the coupling procedure was done to
determine both the choice of acid catalyst and method of
oxidation to the corresponding phosphorus(V) compound.
Pyridinium hydrochloride was most effective as an acid
catalyst giving the derivative 6 in 75-90% yield as
compared to ethereal HCl (<20%) or trichloroacetic acid
(50-75%). 1H-Tetrazole was not investigated due to its
relative expense and the observation that the use of
pyridinium hydrochloride already gave acceptable yields.
Early investigations of this reaction employed tert-butyl
hydroperoxide as an oxidant. We were, however, un-
comfortable with the use of tert-butyl hydroperoxide on
the scale required to produce significant quantities of our
target compound. Cumene hydroperoxide was investi-
gated as an alternate oxidizing agent. The reaction
proceeds more cleanly and in higher yield. Currently,
yields of 90-95% are consistently achieved when the
phosphorylation is done on batches of 40-50 mmol of 5.
Other products of this reaction include tribenzyl phos-
phite, tribenzyl phosphate, and dibenzyl phosphite. This
is the only stage of the synthetic sequence where signifi-
cant chromatography is required. Intermediate 6 must
be free of all P(III) species or the efficiency of the
subsequent catalytic hydrogenation is severely compro-
mised.
hyde, gave the known potassium D-glycerate in 64%
overall yield from mannitol. Overall yields were sub-
stantially lower when purification of the intermediate
aldehyde was attempted. Purification of 3 could be
conveniently achieved at this time by recrystallization
from ethanol.
The chloride 4 was obtained from 3 by the procedure
of Tanaka and Yamashita.12 The acyl chloride, itself, or
esters readily derived from it, represent chiral synthons
comparable in utility to D-glyceraldehyde acetonide.
Purification of the chloride was done by Ku¨gelrohr
distillation. As with glyceraldehyde acetonide, however,
higher overall yields were achieved when 4 was used
directly in the next step without purification.
The chloride 4 was derivatized using (R)-R-methyl-
benzylamine. The diastereomeric amide product was
1
analyzed by H NMR. No resonances corresponding to
the alternate diastereoisomer were observed showing
that no racemization of the R-carbon had occurred up to
this point.13
Formation of the butyl ester 5 was carried out without
addition of any base to neutralize the acid produced in
the reaction. In this way esterification and deprotection
of the acetonide could be conveniently carried out in one
step. The butyl ester, 5, was purified by short-path
distillation.
The butyl ester, rather than, for example, the methyl
or benzyl ester, was used in this sequence as it repre-
sented the best compromise between hydrophobicity and
stability. Use of a benzyl ester would conflict with later
protection of the phosphate groups. Use of a methyl ester
gave intermediate compounds that were very hydrophilic
and, therefore, difficult to analyze by TLC and in the
latter stages of the synthesis almost impossible to recover
from aqueous solution.
The last stages of the synthesis can be done without
purification of the intermediates although some care
must be exercised. The hydrogenation proceeds smoothly
in methanol to give the free acid form of the phosphate.
1
The reaction is best followed by H NMR, observing the
loss of the benzylic proton resonances at 5.15 ppm. If
the product is left too long, hydrolysis and transesteri-
fication of the butyl ester occurs. Attempts to do the
reaction under neutral conditions or in 1-butanol met
with limited success. The butyl ester obtained after the
hydrogenation is soluble in THF. The coupling to give
the cyclic pyrophosphate was successfully carried out
using 1,3-dicyclohexylcarbodiimide in THF. The conver-
sion was quantitative and no bisphosphate species could
be detected by 31P NMR. On completion of the coupling
(13) To determine if resonances from the minor diastereoisomer
could be resolved and detectable in such an NMR experiment, a sample
of the authentic diastereoisomer was mixed with the amide derived
from 4 and (S)-R-methylbenzylamine and analyzed. The minor dias-
tereoisomer was clearly detectable.
(14) Larson, E.; Lu¨ning, B. Tetrahedron Lett. 1994, 35, 2737.
(15) Yu, K.-L.; Fraser-Reid, B. Tetrahedron Lett. 1988, 29, 979.
(16) Uhlmann, E.; Engels, J . Tetrahedron Lett. 1986, 27, 1023.
(17) Tanaka, T.; Tamatsukuri, S.; Ikehara, T. Tetrahedron Lett.
1983, 24, 3171.

Notes
J . Org. Chem., Vol. 61, No. 16, 1996 5699
KOH (128 g, 2.0 mol) was dissolved in water (3 L) in a 10 L
beaker cooled in a large ice-salt bath. Crude 2,3-isopropy-
lideneglyceraldehyde and ice (1.0 kg) were added with manual
stirring. KMnO4 (230 g, 1.5 mol) in water (2 L) was added
cautiously, with constant manual stirring. The reaction tem-
perature was kept below 10 °C by the addition of ice to the
reaction vessel. The brown sludge obtained was neutralized to
pH 8 by the cautious addition of 50% (v/v) aqueous H2SO4. The
water was removed by evaporation in vacuo (60 °C, 20 Torr) to
give, on cooling, a white solid. The solid was extracted with
boiling ethanol (2 × 750 mL) and separated from the inorganic
potassium salts by filtration. The solution was concentrated and
the crude 3 (199 g, 63%) dried (70 °C at 0.1 Torr). 3 was readily
recrystallized from ethanol:13 1H NMR (D2O) δ 1.23 (s, 3H), 1.29
(s, 3H), 3.75 (dd, 1H, J ) 6.9, 8.3 Hz), 4.12 (dd, 1H, J ) 7.6, 8.3
Hz), 4.35 (dd, 1H, J ) 6.9, 7.6 Hz); 13C NMR (D2O) δ 23.3, 23.7,
65.8, 74.1, 109.1, 176.2; IR (KI disc) 1597 s cm-1; HRMS calcd
for C6H9O4 (M+) 145.0523, found 145.0501.
(2S)-2,3-Isop r op ylid en eglycer yl Ch lor id e (4). 3 (50 g,
0.27 mol) was suspended in anhydrous Et2O (500 mL) and
anhydrous pyridine (0.5 mL). The mixture was cooled with an
ice bath, and then oxalyl chloride (49.5 mL, 0.34 mol) was added
dropwise by syringe over a period of 30 min. The ice bath was
removed, and the contents of the flask were stirred for 20 h at
room temperature. The residue of KCl was removed by filtration
and the solvent removed (35 °C, 20 Torr) to give 4 (38.2 g, 85%).
Further purification was achieved by Ku¨gelrohr distillation (bp
70-75 °C, 20 Torr). 1H NMR δ 1.36 (s, 3H), 1.44 (s, 3H), 4.25
(m, 2H), 4.77 (t, 1H, J ) 5.7 Hz); 13C NMR δ 25.2, 25.5, 66.8,
81.1, 112.9, 173.1; IR (liquid film) 1772 cm-1; HRMS calcd for
C5H9O2 (M - COCl) 101.0586, found 101.0603.
reaction, water was added and the solution neutralized
to pH 4 by the addition of lithium hydroxide. Organic
impurities were removed at this stage by filtration and
washing with dichloromethane. Hydrolysis at pH 12
using aqueous lithium hydroxide gave the trilithium salt
of the target compound in approximately 90% yield from
6.
The potassium salt of cDPG is required for biophysical
studies yet the synthesis has been developed to make the
lithium salt. The final cleavage reaction, salt formation,
and crystallization of the cDPG proceed more reliably
with the lithium salt than the potassium salt. Once final
purification has been secured, conversion of the cDPG
to its potassium salt by cation exchange chromatography
is a simple matter.
The cDPG made in this study has been characterized
fully and is identical to material isolated from the natural
source. Higher quality spectra are obtainable for the
synthesized material than isolated material largely due
to sample availability and these should prove to be useful
standards for biophysical studies. It is difficult to record
good 13C spectra with the small amounts of cDPG that
can be recovered from fermentation sources. Not only
are all the carbon resonances split into doublets, the
presence of phosphorus prevents the usual NOE derived
1
signal enhancement caused by H-decoupling.
In conclusion, cDPG can be made on a multigram scale
in 30% overall yield from mannitol. The compound will
find wide use in studies of the thermal stability of
proteins and nucleic acids from methanogen species as
well as more general applications to protein thermosta-
bility. One distinct advantage to the synthesis procedure
is in its reliance on well developed and tested chemistry,
a feature crucial to the successful repetition of this work.
This route also allows the for large scale production of
2,3-bisphosphoglycerate via LiOH hydrolysis of the butyl
ester 7.
n -Bu tyl (2S)-Glycer a te (5). 1-Butanol (900 mL) was placed
in a 2 L round bottom flask, equipped with a reflux condenser
and magnetic stirrer. 4 (50.0 g, 0.27 mol) was added, and the
contents of the flask were heated at reflux for 1 h. The flask
was cooled and the butanol removed (60 °C, 20 Torr). The
residue was distilled (90 °C, 0.1 Torr) to give 5 (46.7 g, 95%). 1H
NMR δ 0.94 (t, 3H, J ) 7.5 Hz), 1.44 (m, 2H, J ) 7.5 Hz), 1.66
(quin, 2H, J ) 7.5 Hz), 3.19 (br m, 2H), 3.87 (dq, 2H, J ) 3.8,
8.9 Hz), 4.22 (t, 2H, J ) 6.7 Hz), 4.27 (t, 1H, J ) 3.8 Hz); 13C
NMR δ 3.6, 19.1, 30.6, 64.3, 65.7, 72.0, 173.1; IR (liquid film)
3402 br s and 1739 cm-1; HRMS calcd for C7H14O4 (M+)
162.0892, found 162.0894.
n -Bu tyl (2S)-2,3-Bis(d iben zylp h osp h o)glycer a te (6).
A
solution of pyridinium hydrochloride (19.75 g, 155 mmol) in CH2-
Cl2 (750 mL) was added to a solution of 5 (8.1 g, 50 mmol) and
dibenzyl diisopropylphosphoramidate (51.9 g, 150 mmol) in CH2-
Cl2 (500 mL). The solution was stirred at room temperature
for 1 h and then cooled to 0 °C. Cumene hydroperoxide (80%)-
(28.8 mL, 155 mmol) was cautiously added to the cooled solution.
After 1 h at rt the solvent was concentrated. Chromatography
on silica gel, eluting with EtOAc-hexanes (1:2), gave 6 (32.5 g,
95.3%). 1H NMR δ 0.85 (t, 3H, J ) 7.3 Hz), 1.25 (sextet, 2H, J
) 7.3 Hz), 1.53 (quin, 2H, J ) 7.3 Hz), 4.10 (t, 2H, J ) 6.7 Hz),
4.32 (m,), 4.95-5.15 (m, 8H), 7.22-7.34 (m, 20H); 13C NMR δ
13.6, 19.0, 30.5, 66.1, 67.1 (br t, J 5.7), 69.7 (m), 128.0, 128.2,
128.2, 128.4, 128.6, 135.6 (t, J ) 7.0 Hz), 166.9 (d, J ) 3.6 Hz);
Exp er im en ta l Section
Gen er a l. The solvents CH2Cl2 (CaH2), Et2O (Na/K-benzo-
phenone), n-BuOH (CaH2), and THF (Na/K-benzophenone) were
dried and distilled prior to use. Solutions were concentrated
by evaporation in vacuo.
1,2:5,6-Diisop r op ylid en em a n n itol (2) Mannitol (200 g, 1.1
mol) was placed in a 2 L three-neck flask equipped with a
mechanical stirrer. Anhydrous dimethoxyethane (500 mL) and
freshly distilled, methanol free 2,2-dimethoxypropane (320 mL,
2.6 mol) were added and the slurry stirred vigorously. SnCl2
(0.2 g, 1 mmol) was added and the slurry was heated at reflux.
After 16 h the resulting clear solution was left to cool to rt, and
then pyridine (1.0 mL) was added. The solvent was removed
by evaporation in vacuo (80 °C, 20 Torr). Upon cooling, a white
solid formed. The solid was suspended in hexanes (1 L) and
the insoluble 2 collected by filtration. 2 was dissolved in acetone
(1.5 L), and the residual solid was removed by filtration. The
acetone extract was evaporated in vacuo, and the resulting
crystals of 2 (227 g, 75%) were dried under vacuum (0.1 Torr,
16 h). The crude product was identified by comparison with
spectral data reported for the known compound.12
P ota ssiu m (2S)-2,3-Isop r op ylid en eglycer a te (3). In a 3
L, two neck flask, equipped with a mechanical stirrer and reflux
condenser, 2 (227 g, 0.85 mol) was dissolved in a mixture of CH2-
Cl2 (2 L) and a saturated aqueous solution of NaHCO3 (100 mL).
NaIO4 (280 g, 1.3 mol) was added over 20 min without cooling.
After 4 h, anhydrous MgSO4 (400 g) was added in several small
portions with vigorous stirring. The resulting suspension was
filtered and the filter cake washed with CH2Cl2 (2 L). The
combined organic fractions were evaporated in vacuo at less than
30 °C to give crude 2,3-isopropylideneglyceraldehyde which was
used immediately in the next reaction.
31P NMR δ -1.32 (s) and -1.80 (s); IR (liquid film) 1759 cm-1
;
HRMS calcd for C35H40O10P2 (M+) 682.2098, found 682.2110.
Cyclo-(2S)-2,3-d ip h osp h oglycer a te, Tr ilith iu m Sa lt (9)
Note: 6 must be free of phosphines or the catalyst will be
poisoned. The rate of this reaction varies considerably and must
be monitored by TLC or NMR. Palladium on carbon (1.0 g, 5%
Pd) was added to a solution of 6 (10.0 g, 14.7 mmol) in MeOH
(200 mL) and stirred under an atmosphere of H2 for between
1.5 and 10 h. The course of the reaction was followed by 1H
NMR. On completion, the suspension was filtered through
Celite and the solvent removed to give crude 7 (4.6 g, 95%
approximately): 1H NMR (D2O-CD3OD) 0.88 (t, 3H, J ) 6.6
Hz), 1.32 (sextet, 2H, J ) 7.3 Hz), 1.56 (pent, 2H, J ) 7.3 Hz),
4.05-4.40 (m, 3H), 4.19 (t, 2H, J ) 6.6 Hz) and 4.96 (br s, 4H);
13C NMR (D2O) 13.51, 18.98, 30.39, 66.50 (t, J ) 4.9 Hz), 67.10,
73.94 (dd, J ) 4.6, 8.2 Hz), and 172.85 (d, J ) 4.2 Hz); 31P NMR
3.06 (s) and 3.51 (s);IR (liquid film) 3650-2100 br s and 1736 s
cm-1
. Residual methanol in the crude 7 was removed by
azeotropic evaporation with anhydrous THF (2 × 100 mL). The
crude acid was dissolved in THF (200 mL), and a solution of
DCC (3.64 g, 17.6 mmol) in THF (20 mL) was added. The

5700 J . Org. Chem., Vol. 61, No. 16, 1996
Additions and Corrections
suspension was stirred for 16 h at 22 °C. Water (300 mL),
followed by LiOH (sufficient to raise the pH to 4.0), was added
and the mixture filtered. The aqueous solution was washed with
CH2Cl2 (2 × 200 mL). The aqueous phase was evaporated to
dryness (50 °C, 20 Torr) to give 8 as a hydrate: 1H NMR (D2O)
0.85 (t, 3H, J ) 7.3 Hz), 1.32 (sextet, 2H, J ) 7.3 Hz), 1.60 (pent,
this work. We also thank Professor J ohn Reeve and Dr.
Rowan A. Grayling (Department of Microbiology, The
Ohio State University) for providing a sample of the
naturally occurring cyclo-D-diphosphoglycerate isolated
from M. fervidus. Advice concerning the synthesis of
H-phosphonates and their use, from Dr. Elizabeth
Larson and Professor Bjo¨rn Lu¨ning, is gratefully ac-
knowledged.
2H, J ) 7.3 Hz), 4.18 (t, 2H, J ) 6.5 Hz), 4.15-4.45 (m, 3H); 13
C
NMR (D2O) 13.3, 18.7, 68.0 (d, J ) 6.6 Hz), 75.6 (d, J ) 5.0 Hz),
and 169.5 (d, J ) 10.3 Hz; 31P NMR (D2O-H2O) -11.13 (d, J )
17.2 Hz), -9.85 (d, J ) 17.2 Hz); IR (KI disc) 3467 bs s, 1749 s,
and 1644 s cm-1; HRMS calcd for C7H11O9P2 (M+ - 2 Li)
300.9878, found 300.9884. LiOH monohydrate was added
portionwise to a solution of 8 in water (20 mL) until a pH of 12
was maintained, and the solution was stirred at rt for 30 min.
The pH was adjusted to 7.0 by addition of 1.0 M H2SO4. The
solvent was removed. The product obtained was digested with
water-methanol (1:9), and the residual solid removed by filtra-
tion. The solvent was again removed to give the desired product
(3.5 g, 90%): [R]D +7.7° (c 4.49, water); 1H NMR (D2O) δ 4.10-
4.40 (m, 3H); 13C NMR (D2O) δ 70.0 (d, J ) 6.75 Hz), 78.6 (d, J
) 7.4 Hz), 174.6 (d, J ) 6.7 Hz); 31P NMR (D2O-H2O) δ -11.05
(d, J ) 17.5 Hz), -9.80 (d, J ) 17.6 Hz).
(2S)-2,3-Bisp h osp h oglycer a te, Dilith iu m Sa lt. This com-
pound was made from 6 in the same way that 9 was made from
6 (see above) with the exception that after the hydrogenation
step, water and LiOH‚H2O were added directly to saponify the
butyl ester. The dilithium salt was obtained as a glassy solid
in yields averaging 90%. The product was shown to be identical
to a commercially obtained sample of (2S)-2,3-bisphosphoglyc-
erate.
Su p p or tin g In for m a tion Ava ila ble: 13C and 31P NMR
1
spectra of the synthesized cDPG and H and 13C NMR spectra
of 6 (7 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
Note Ad d ed in P r oof. Shortly after original submis-
sion of this manuscript a synthesis of cDPG from D-
mannitol was published by the group of Berkessel. They
achieved the gram-scale conversion of D-mannitol into the
methyl ester of the acyclic bisphosphoglycerate in 9%
overall yield. Subsequently, cDPG which was 80% pure
was obtained in 20% yield from the methyl ester.
Berkessel, A.; Geisel, U.; He´rault, D. A. Tetrahedron Lett.
1996, 37, 355.
Ack n ow led gm en t. We wish to thank The College
of Pharmacy, The Ohio State University for support of
J O9601472
Additions and Corrections
Vol. 60, 1995
earlier work on the use of the alkoxycarbonyl moiety as
a protective (“blocking”) group in the malonic ester
synthesis and was inadvertently omitted from our paper.
Padgett, H. C.; Csendes, I. G.; Rapoport, H. J . Org. Chem.
1979, 44, 3492.
F a n g-J ie Zh a n g, Gu o-Qia n g Lin ,* a n d Qi-Ch en Hu a n g .
Synthesis, Resolution, and Absolute Configuration of Optically
Pure 5,5′′-Dihydroxy-4′,4′′,7,7′′-tetramethoxy-8,8′′-biflavone and
Its Derivatives.
Page 6428, Scheme 2, first line, (+) should be (().
Page 6429, Figure 2, left side, first line, should read
“The dashed lines refer to (-)-1, (-)-5, and (-)-6. The
solid lines refer to (+)-1, (+)-5, and (+)-6.”
J O964009O
S0022-3263(96)04009-1
An d r ew Str eitw ieser ,* F a r a j Abu -Ha sa n a yn ,
J O964006B
Ar n d t Neu h a u s, a n d F r a n k Br ow n . An ab Initio Study of
Some Phenyl- and (Halophenyl)alkali Compounds.
S0022-3263(96)04006-6
Page 3151. The correct spelling of second author’s
name is F a r a j Abu -Ha sa n a yn .
Ta k a n or i Ya m a za k i, Alek sa n d r Ka sa tk in ,
Ya su fu m i Ka w a n a k a , a n d F u m ie Sa to*. Allyl as Protective
Group for the Acidic Hydrogen of Malonic Ester.
J O9640130
Page 2266. The reference given below refers to some
S0022-3263(96)04013-3