ation of the residue as above afforded (2RS)-[2-3H1]glycolic
acid (77 mg, 88%), identified by comparison with an authentic
sample.
it (100 mg) for 45 min at room temperature in a mixture (0.5
cm) prepared from acetonitrile (9.5 cm3) and 40% aqueous
hydrogen fluoride (0.5 cm3). Saturated aqueous sodium
hydrogen carbonate (2 cm3) was added and the mixture was
extracted thrice with dichloromethane to give benzyl glycolate
(45 mg) identical to the material above. The benzyl group was
then removed by hydrogenolysis as already described to yield
glycolic acid which was sublimed at 90 ЊC/0.1 mmHg (19 mg,
88%).
Brief description of trial experiments and preparation of simple
derivatives of glycolic acid
(a) Silylations. The alcohol6 22 (97 mg) was dissolved in a
mixture (5 cm3) prepared from aqueous 0.6 M KOH (50 cm3)
and THF (25 cm3) and stirred at room temperature for 3 h.
After the THF had been evaporated, the solution was adjusted
to pH 10.2 with aqueous 6% perchloric acid then filtered and
freeze dried to yield a solid (162 mg). Part of the residue (44
mg) containing the sodium salt of 20 was dissolved in dimethyl-
formamide, DMF, (0.5 cm3) and treated first with Hünig’s base
(150 mg) then with tert-butyldimethylsilyl chloride (132 mg).
The mixture was stirred under argon at room temperature for
10 min then the addition of base and silyl chloride was repeated
and after 1 min more, ethyl acetate (10 cm3) was added. After
the solution had been washed with water (3 × 5 cm3), it was
dried and evaporated to yield the silyl derivative 21 as an oil
(29 mg, 52%) which showed only minor impurities by NMR.
δH (CD2Cl2, 100 MHz) 0.09 (6 H, s, SiMe2), 0.22 and 0.24 (each
6 H, s, CO2SiMe2), 0.90 (9 H, s, But), 0.94 (18 H, s, CO2SiBut),
2.62 (4 H, m, CH2CH2CO), 3.41 (2 H, s, CH2CO), 4.66 (2 H, s,
CH2O), 6.50 (1 H, s, 2-H pyrrole), 8.0 (1 H, br s, NH): m/z 569,
C28H55NO5Si requires 569.
Enzymic conversion of (11R)-[11-3H1]PBG 1a and the (11S)-
isomer 1b into the bilanes 5a and 5b and determination of their
configurations
(d) Immobilised hydroxymethylbilane synthase. Sepharose,
activated by cyanogen bromide (Sigma, 2 g) was swollen in 1 M
hydrochloric acid (50 cm3) and washed on a sinter with 1 mM
hydrochloric acid (400 cm3) then with a solution (300 cm3)
containing sodium hydrogen carbonate (0.1 M) and sodium
chloride (0.5 M). Purified hydroxymethylbilane synthase4 was
dialysed against the second of the washing solutions above and
part of it (2 × 105 units) was diluted to 25 cm3 with the dialysis
solution and cooled to 5 ЊC. Unlabelled PBG was then added to
ca. 2.2 mol PBG per mol of enzyme and after 20 min, the
above washed Sepharose was added. The mixture was shaken at
5 ЊC for 23 h at such a rate that the gel remained suspended and
it was then collected; assay of the filtrate showed the absence
of enzyme. The gel was resuspended in aqueous 0.2 M glycine
(30 cm3, pH 8.2) and shaken at 5 ЊC as before for 23 h to block
any remaining reactive groups on the Sepharose.
The pyrrole 22 (50 mg) in dry DMF (0.5 cm3) was silylated
using Hünig’s base (0.1 g) and tert-butyldimethylsilyl chloride
(60 mg) as above and worked up in a similar way to give the silyl
ether 23 as an oil (61 mg, 82%) which was essentially pure by
NMR. δH (CD2Cl2, 100 MHz) 0.10 (6 H, s, SiMe2), 0.93 (9 H, s,
But), 2.65 (4 H, m, CH2CH2CO), 3.43 (2 H, s, CH2CO), 3.66
(6 H, s, 2 × OMe), 4.67 (2 H, s, CH2O), 6.51 (1 H, d, J 2, 2-H
pyrrole), 8.1 (1 H, br s, NH).
After the gel had been washed with the above dialysis solu-
tion (300 cm3), it was equilibrated with 50 mM Tris-HCl buffer,
pH 8.0 and packed into a jacketed column.
(e) Enzymic production of bilanes 5a and 5b. The foregoing
column was held at 36–37 ЊC while a solution (300 cm3) of the
(11R)-PBG 1a (0.20 mg cmϪ3) in 50 mM Tris-HCl buffer, pH 8
was passed through at 1.1 cm3 minϪ1 into stirred aqueous 2 M
potassium hydroxide (15 cm3) containing sodium dithionite
(300 mg) under argon at 0 ЊC. The preparation of 1a from the
lactam 13a was as usual.1 Assay26 showed that hydroxymethyl-
bilane 5a was present (6.4 mg). The alkaline solution was evap-
orated on a high vacuum rotatory evaporator (water bath
< 30 ЊC) to ca. 10 cm3 and run onto a Sephadex G-10 column
(4 × 14 cm) eluting with 0.1 M aqueous KOH to separate the
Tris. The Ehrlich-positive fractions (30 cm3) were combined,
freeze dried and the residue was assayed again which showed
that sufficient hydroxymethylbilane 5a (1.6 mg) had survived. A
solution of this solid in water (10 cm3) was carefully adjusted to
pH 12 by dropwise addition of 0.6 M aqueous perchloric acid.
The filtered solution was then freeze dried and the residue dis-
solved under argon in dry DMF (1.8 cm3) containing Hünig’s
base (870 µl). The following additions were made at the indi-
cated times: tert-butyldimethylsilyl chloride (525 mg, start),
Hünig’s base (870 µl, 10 min), tert-butyldimethylsilyl chloride
(525 mg, 13 min), Hünig’s base (670 µl) and the silyl chloride
(525 mg) together at 20 min. After 30 min, ethyl acetate (40 cm3)
containing Hünig’s base (0.5 cm3) was added and the solution
was extracted twice with half saturated aqueous sodium hydro-
gen carbonate then with brine, dried and evaporated to an oil
(250 mg).
(b) Ozonolysis. The silyl ether 23 (37 mg) was spread on silica
(2 g) as earlier but using dry diethyl ether as solvent and the
ozonolysis was carried out in the same way but for 30 min at
Ϫ70 ЊC and 1 h at room temperature. The products were eluted
from the silica with ethyl acetate and the residue from evapor-
ation in methanol (3 cm3) was treated with an excess of phenyl-
diazomethane. Purification of the product by PLC on silica
using diethyl ether–pentane (1:6) gave the benzyl ester 25 (2.8
mg, 10%) identified by comparison with the standard sample
prepared below.
(c) Protection and deprotection of glycolic acid 16. The benzyl
ester of glycolic acid was prepared in the way above. Alter-
natively, glycolic acid (665 mg) in diethyl ether (40 cm3) was
stirred while a solution (40 cm3) of phenyldiazomethane (pre-
pared25 in diethyl ether from 1.86 g azibenzil) was added drop-
wise over 30 min. The small excess of reagent was removed by
addition of a little glycolic acid and the pale yellow solution
was washed with saturated aqueous sodium hydrogen carbon-
ate, dried and evaporated. Chromatography on silica (3 × 15
cm) eluting with ethyl acetate–hexane (1:9) removed yellow
material and further elution with ethyl acetate–hexane (1:1)
afforded benzyl glycolate. Distillation (Kugelrohr) at 170 ЊC/2
mmHg gave the pure ester as an oil (1.3 g, 67%). δH 2.38 (1 H,
br s, OH), 4.19 (2 H, s, CH2CO), 5.22 (2 H, s, CH2Ar) and 7.36
(5 H, s, ArH).
A solution of this ester (1.0 g) in dry DMF (7 cm3) was
stirred for 12 h with imidazole (1.0 g) and tert-butyldimethyl-
silyl chloride (1.17 g) and then mixed with ethyl acetate (50
cm3). The solution was washed thrice with brine, dried (MgSO4)
and evaporated. The residue by Kugelrohr distillation at
120 ЊC/0.3 mmHg afforded the silylated ester 25 as an oil (1.67
g, 92%). δH 0.08 (6 H, s, SiMe2), 0.90 (9 H, s, But), 4.28 (2 H, s,
CH2CO), 5.17 (2 H, s, CH2Ph) and 7.35 (5 H, s, Ar ).
(f) Degradation of 5a and 5b to glycolic acids followed
by stereochemical assay. A solution of the foregoing oil in
dichloromethane (50 cm3) was mixed with silica (20 g) as
for earlier examples and the solvent was evaporated. The
dissolution and evaporation were repeated twice from
dichloromethane (2 × 30 cm3) and this solid was dried at 0.1
mmHg. The ozonolysis was carried out as above but for 1 h at
Ϫ78 ЊC and 15 min at room temperature. Then the silica was
The silyl protecting group was removed from 25 by stirring
J. Chem. Soc., Perkin Trans. 1, 1999, 2691–2698
2697