A highly efficient synthesis of [1-13C, 18O]- and [1-13C, 2H2]-glycerol for the elucidation of biosynthetic pathways

Article abstract of DOI:10.1016/S0040-4039(02)02697-7

Labeled glycerol is a widely used biochemical probe to investigate biosynthetic pathways. A highly efficient synthesis of [1-¹³C, ¹⁸O]- and [1-¹³C, ²H₂]-glycerol is described in which the ¹³C label is introduced using cyanide. The ¹⁸O label was introduced by a Pinner synthesis and reduction of the ester 5 allowed incorporation of the ²H labels.

Full text of DOI:10.1016/S0040-4039(02)02697-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 789–792
18
A highly efficient synthesis of [1-13C, O]- and [1-13C,
2H2]-glycerol for the elucidation of biosynthetic pathways
Alexandros P. Siskos and Alison M. Hill*,†
Department of Chemistry, King’s College London, Strand, London WC2R 2LS, UK
Received 3 September 2002; accepted 22 November 2002
Abstract—Labeled glycerol is a widely used biochemical probe to investigate biosynthetic pathways. A highly efficient synthesis
of [1-13C, 18O]- and [1-13C, 2H2]-glycerol is described in which the 13C label is introduced using cyanide. The 18O label was
2
introduced by a Pinner synthesis and reduction of the ester 5 allowed incorporation of the H labels. © 2003 Elsevier Science Ltd.
All rights reserved.
Glycerol is an ubiquitous primary metabolite with
numerous metabolic sources and is an important inter-
mediate of energy metabolism.1 Glycerol and glycerate
are often efficiently incorporated into primary and sec-
ondary metabolites via a number of biosynthetically
distinct pathways and isotopically labeled glycerols and
glycerates have been used to investigate numerous
biosynthetic pathways2–5 and in metabolism studies.6–9
Doubly isotopically labeled compounds enable bond-
breaking, bond-forming and rearrangement processes in
biosynthetic pathways to be investigated using NMR
spectroscopy.10 Detailed information on the oxidation
levels of intermediates and oxidative and reductive pro-
cessing steps of the metabolite can be obtained by using
A where we had observed incorporation of 13C and
deuterium from [1,3-13C2]- and [1-2H2]-glycerol, respec-
tively.11 To ascertain whether intact incorporation of
the 1-CH bond of glycerol into the 11-CH bond of
soraphen A was taking place, we needed to prepare
2
[1-13C, H2]-glycerol 1b.2 We were also interested in the
origin of the oxygen atoms in soraphen A and in
particular whether the 1-CO bond of glycerol was
incorporated intact into soraphen A and consequently
required [1-13C, 18O]-glycerol 1a.2 In this paper we
describe a highly efficient sequence to produce both of
these doubly isotopically labeled glycerols.
Benzyloxyacetaldehyde 2 was prepared from ozonolysis
of Z-dibenzyloxybut-2-ene.12 Introduction of the 13C
label was achieved by reaction of 2 with potassium
13C-cyanide to give the cyanohydrin 3 (Scheme 1).4,13,14
2
precursors labeled with 13C in conjunction with H and
18O. We were particularly interested in the incorpora-
tion of glycerol into the polyketide metabolite soraphen
2
Scheme 1. Synthesis of [1-13C, 18O]- and [1-13C, H2]-glycerol. Reagents and conditions: (i) K13CN, H2O, 95%; (ii) AcCl, EtOH;
(iii) H2ꢀ, THF, 92% (over two steps); (iv) LiAlH4, ether, 79%; (v) H2, Pd, EtOH, 95–98%; (vi) H2O, THF, 89% (over two steps);
13
18
’
(vii) LiAlD4, ether, 93%. Key: = C; ꢀ= O.
Keywords: glycerol; labeling; cyanohydrin; imidic acids and derivatives; biosynthesis.
* Corresponding author. Fax: +44 (0)1392 263434; e-mail: a.m.hill@ex.ac.uk
† Present address: School of Chemistry, University of Exeter, Stocker Road, Exeter, EX4 4QD.
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02697-7

790
A. P. Siskos, A. M. Hill / Tetrahedron Letters 44 (2003) 789–792
Excess aldehyde (1.4 equiv. or more) was critical for the
complete consumption of the cyanide (monitored using
the cyanide sedimentation test with silver nitrate).13
Workup of the reaction gave a mixture of the cyanohy-
drin 3 (78%) and aldehyde (22%) which could be used
without any purification in the next step; separation
over silica gave the pure product if required. We
wanted to introduce the 18O label using H128O which is
very expensive and so we required a method which
utilised the minimum amount of labeled water while
simultaneously giving a product that was isotopically
enriched. Conversion of 3 into the corresponding acid
was not straightforward due to its reactive nature.
Hydrolysis under basic and acidic reaction conditions
failed to give a product. Hence, a different strategy was
adopted: the cyanohydrin 3 was converted to the ester
5 using the Pinner synthesis.15 Treatment of 3 with
acetyl chloride and ethanol gave the imino ethyl ester 4
as its hydrochloride salt which was unstable and hygro-
scopic. Consequently, it was important to use the imi-
date salt 4 immediately in the next reaction, especially
when 18O was being introduced. The advantage of using
this method is that introduction of the 18O label is
highly efficient as only the carboxylate oxygen atom is
labeled (which is retained in the subsequent reduction
step) and hence only 4.5 equiv. of H128O were required
to give the ethyl ester 5a in 92% yield over two steps.
Reduction and deprotection gave the [1-13C, 18O]-glyc-
obtained using the Bruker AMX 400 University of
London Intercollegiate Research Service for NMR at
King’s College. This work was financially supported by
funding from Syngenta (Research Triangle Park, North
Carolina, USA) and the State Scholarship Foundation
(Republic of Greece) (A.P.S.).
References
1. Brisson, D.; Vohl, M.-C.; St-Pierre, J.; Hudson, T. J.;
Gaudet, D. BioEssays 2001, 23, 534–542.
2. Hill, A. M.; Siskos, A. P.; Harris, J. P. Synth. Appl.
Isotop. Lablld. Cmpds. 2001, 7, 626–630.
3. Pitlik, J.; Townsend, C. A. Chem. Commun. 1997, 225–
226.
4. Thirkettle, J. E.; Baldwin, J. E.; Edwards, J.; Griffin, J.
P.; Schofield, C. J. Chem. Commun. 1997, 1025–1026.
5. Challis, G. L.; Chater, K. F. Chem. Commun. 2001,
935–936.
6. Ackermans, M. T.; Ruiter, A. F. C.; Endert, E. Anal.
Biochem. 1998, 258, 80–86.
7. McIntosh, T. S.; Davis, H. M.; Matthews, D. E. Anal.
Biochem. 2001, 300, 163–169.
8. Ficher, J. S.; Hickner, R. C.; Racette, S. B.; Binder, E. F.;
Landt, M.; Kohrt, W. M. J. Clin. Endocrinol. Metab.
1999, 84, 3726–3731.
9. Lindau, B. R. Proc. Nutr. Soc. 1999, 58, 973–978.
10. Simpson, T. J. Top. Curr. Chem. 1998, 195, 1–48.
11. Hill, A. M.; Harris, J. P.; Siskos, A. P. Chem. Commun.
1998, 2361–2362.
12. Denmark, S. E.; Herbert, B. J. Org. Chem. 2000, 65,
2887–2896.
13. Thirkettle, J. E. Ph.D. Thesis, University of Oxford,
1997.
2
erol 1a. To prepare [1-13C, H2]-glycerol 1b, the imidate
salt 4 was converted to the ethyl ester 5b and the
deuterium atoms were introduced using LiAlD4 in the
reduction step. Overall, the five step sequence gave a 66
and 77% yield of the doubly isotopically labeled glyc-
erols 1b and 1a, respectively, based on K13CN.16
The synthetic route is highly versatile allowing the
incorporation of any one, two or three of the isotopic
labels to be introduced at position 1 of glycerol. The
route could also be adapted to prepare glycerate with
one or both of the carbonyl atoms isotopically labeled.
The incorporation of 18O is highly efficient as only 4.5
equiv. of 18O water were used; this compares
favourably with 10 equiv. used to prepare [1-13C, 18O2]-
labeled glycerate.4,13 The synthesis of small molecules
such as glycerol is difficult as they partition into
organic and aqueous solutions making isolation
difficult. The use of a benzyl protecting group for the
3-hydroxyl moiety made the intermediates UV active
and organic soluble which made isolation and handling
of the intermediates straightforward. The benzyl group
was removed efficiently and cleanly by hydrogenolysis
and the labeled glycerol 1 could be isolated by a simple
procedure of filtration and removal of the solvent
thereby eliminating the need for an extraction step. In
conclusion, a highly efficient five-step procedure has
been developed to produce doubly labeled glycerol.
14. Seriani, A. S.; Nunez, H. A.; Barker, R. Carbohydr. Res.
1979, 72, 71–78.
15. Neilson, D. G. In The Chemistry of the Amidines and
Imidates; Patai, S., Ed.; John Wiley: New York, 1991;
Vol. 2, pp. 425–483.
16. Experimental data. Unless otherwise indicated all reac-
tions were carried out under argon or nitrogen. THF and
ether were dried and freshly distilled from sodium/ben-
zophenone. Dichloromethane was freshly distilled over
CaH2. Flash chromatography was conducted on Merck
silica gel 60 (40–63 mm). 1H, 2H and 13C NMR spectra
were recorded on a Bruker AM360 or AMX400 spec-
trometers. Chemical shifts were reported in ppm with
1
reference to TMS (0.0 ppm) for H NMR, natural abun-
dance signal of CH2Cl2 (5.30 ppm) or H2O (4.79 ppm) for
2H NMR and to dideuteriomethane (53.52 ppm) for 13C
NMR. Mass spectra were recorded on a Jeol AX505W
and a Kratos MS890MS spectrometer.
[1-13C]-3-Benzyloxy-2-hydroxy-propionitrile 3. To a solu-
tion of potassium [13C]cyanide (2.00 g, 30.3 mmol, CIL,
99% 13C) in water (140 mL) at pH 8.5, a solution of
benzyloxyacetaldehyde 2 (6.36 g, 42.3 mmol, 1.4 equiv.)
in THF (90 mL) was added dropwise over 1 h and the
resultant solution stirred for a further 1 h at room
temperature. The pH of the reaction was continuously
monitored and maintained between pH 7–9 using sodium
hydroxide (3 M) and hydrochloric acid (3 M). The reac-
tion was monitored by TLC (Rf aldehyde 2 0.56, nitrile 3
Acknowledgements
We would like to thank Mrs. J. Hawkes and Mr. J.
Cobb for NMR support and Mr. A Cakebread and Mr.
R. Tye for mass spectra. Deuterium NMR spectra were

A. P. Siskos, A. M. Hill / Tetrahedron Letters 44 (2003) 789–792
791
0.72, ether) and with the cyanide sedimentation test with
silver nitrate.13 Once all the cyanide had been consumed,
the pH was lowered to 4–5 and nitrogen was passed
through the reaction mixture for 10 min; the gas outlet
stream was connected to a solution of potassium hydrox-
ide (8 M) to remove any remaining hydrogen cyanide.
The reaction mixture was then extracted with
dichloromethane (6×25 mL) and the organic extracts
were washed with brine and dried (Na2SO4 and MgSO4).
The solvent was removed in vacuo to leave a pale yellow
oil (6.40 g) which was a mixture of the nitrile 3 (78%, 29.0
mmol, 95% based on K13CN, 69% based on benzyloxyac-
etaldehyde) and the aldehyde (22%, 8.3 mmol). The
nitrile 3 was used for the next step without any further
purification but can be purified by flash chromatography
on silica gel (hexane:ether, 1:2) to give pure nitrile 3.
Found M+ 178.0827, calculated for C913CH11O2N
178.0823.
gave the crude imidate salt 4 (4.378 g) which was dis-
solved in water (35 mL) and THF (25 mL). The title
compound was obtained as a pale yellow oil (2.791 g,
12.4 mmol, 89% over two steps). Found M+ 225.1085,
calculated for C1113CH16O4 225.1082.
1H NMR (CD2Cl2, 360 MHz) 1.25 (t, 3H, J 7.1 Hz,
CH3), 3.12 (br s, 1H, OH), 3.73 (m, 2H, 13CH2), 4.21 (q,
2H, J 7.1 Hz, CH2CH3), 4.28 (m, 1H, CH), 4.53 (m, 2H,
CH2Ph), 7.28–7.37 (m, 5H, ArH); 13C NMR (CD2Cl2, 90
3
2
MHz) 14.3 (d, CH3, J 21.9 Hz), 62.1 (d, CH2CH3, J 2.5
Hz), 71.1 (d, J 59.6 Hz, CH), 72.0 (CH2), 73.7 (CH2),
128.0, 128.0 (C4%), 128.7, 138.3 (C1%), 173.1 (13C=O);
wmax/cm−1 3472, 2990, 2867, 1697; m/z (EI) 225 (M+,
42%), 177 (53), 164 (22), 119 (45), 91 (75), 84 (100).
[1-13C,18O]-3-Benzyloxy-2-hydroxypropanol 6a. A solution
of the ester 5a (2.935 g, 12.9 mmol) in dry ether (20 mL)
was added slowly to a solution of lithium aluminium
hydride in ether (1 M, 27.2 mL, 27.2 mmol, 2.1 equiv.)
and the mixture stirred for 1.5 h. The reaction was
quenched with 50% aqueous THF (10 mL), filtered and
washed with dichloromethane (100 mL). The solid was
treated with a saturated solution of ammonium chloride
(60 mL), filtered and the filtrate extracted with
dichloromethane (6×10 mL). The organic extracts were
combined and dried (Na2SO4). The solvent was removed
in vacuo to give the crude alcohol which was purified by
flash silica chromatography (petrol 60–80°C/ethyl ace-
tate, 1:6) to give the alcohol 6a as an almost colourless oil
(1.88 g, 10.2 mmol, 79%). Found M+ 185.1026, calculated
for C913CH14O128O 185.1019.
1H NMR (CD2Cl2, 400 MHz) 3.30 (d, 1H, J 6.1 Hz,
OH), 3.69 (m, 2H, CH2OBn), 4.56 (m, 1H, CH), 4.62 (s,
2H, CH2Ph), 7.30–7.39 (m, 5H, ArH); 13C NMR
(CD2Cl2, 100 MHz) 61.3 (d, J 61.8 Hz, CH), 71.0 (CH2),
74.1 (CH2), 118.7 (CN), 128.4, 128.5 (C4%), 129.9, 137.4
(C1%); wmax/cm−1 3424, 2922, 2869; m/z (EI) 178 (M+,
12%), 148 (15), 107 (14), 91 (100).
Ethyl [1-13C,18O]-3-benzyloxy-2-hydroxypropionate 5a. A
solution of the nitrile 3 (2.549 g, 14.3 mmol) in dry
ethanol (40 mL) was cooled to 0°C. Acetyl chloride (162
mmol, 11.3 equiv.) was added dropwise over 45 min and
the solution stirred for 12 h. An additional aliquot of
acetyl chloride was added (18 mmol, 1.25 equiv.) and the
reaction stirred for a further 2 h. The volatiles were
removed in vacuo to yield a white solid which was
suspended in dry hexane. The crude imidate salt 4 (4.524
g) was obtained by removal of the hexane in vacuo and
was used immediately for the next step. {13C NMR
(DMSO-d6, 90 MHz) of the unlabeled imidate salt 4: 14.1
(CH3), 60.1 (CH2CH3), 70.0 (CH), 71.8 (CH2), 72.3
(CH2), 127.4 (C4%), 127.4, 128.1, 132.2 (C1%), 172.2
(CꢀNH2+)}. The imidate salt was mixed with 18O-water
(1.3 g, 65.0 mmol, 4.5 equiv., Goss, 97.26% atom% 18O,
0.96 atom% 17O) and dry THF (45 mL) and stirred for 15
h. Water (30 mL) was added and the ester 5a extracted
into dichloromethane (5×30 mL), dried (Na2SO4) and the
solvent evacuated in vacuo to give the crude ester 5a as a
brown oil. Silica chromatography (petrol 60–80°C/ethyl-
acetate, 2:1) gave the ester 5a as a pale yellow oil (2.975
g, 13.1 mmol, 92% over two steps). 18O enrichment is
89% by 13C NMR. Found M+ 227.1123, calculated for
C1113CH16O318O 227.1124.
1H NMR (CD2Cl2, 360 MHz) 2.70 (br s, 1H, OH), 3.10
(br s, OH), 3.31–3.45 and 3.71–3.83 (m, 2H, 13CH218OH),
3.45–3.54 (m, 2H, CH2OBn), 3.76–3.83 (m, 1H, CH), 4.52
(s, 2H, CH2Ph), 7.26–7.37 (m, 5H, ArH); 13C NMR
(CD2Cl2, 90 MHz) 64.3 (13CH128OH), 71.1 (d, J 40.2 Hz,
CH), 72.1 (CH2OBn), 73.7 (CH2Ph), 128.1, 128.1 (C4%),
128.7, 138.4 (C1%); wmax/cm−1 3380, 2918, 2865, 1664; m/z
(EI) 185 (M+, 19%), 107 (53), 91 (100).
[1-13C, 2H2]-3-Benzyloxy-2-hydroxypropanol 6b. Method
as for 6a. A solution of the ester 5b (2.737 g, 12.2 mmol)
in dry ether (20 mL) was added slowly to a solution of
lithium aluminium deuteride in ether (1 M, 23.0 mL, 23.0
mmol, 1.9 equiv.). The alcohol 6b was obtained as an
almost colourless oil (2.09 g, 11.3 mmol, 93%). 2H iso-
topic enrichment was 98% by 1H NMR. Found M+
185.1100, calculated for C913CH122H2O3 185.1102.
1H NMR (CD2Cl2, 360 MHz) 3.08 (br s, OH), 3.47 (m,
2H, CH2OBn), 3.81 (m, 1H, CH), 4.50 (s, 2H, CH2Ph),
7.26–7.37 (m, 5H, ArH); 2H NMR (CH2Cl2, 61 MHz)
3.51 (d, JCD 22.5 Hz, CDD), 3.58 (d, JCD 21.8 Hz, CDD);
13C NMR (CD2Cl2, 90 MHz) 63.6 (quin., JCD 21.6 Hz,
13CD2OH), [63.8 (t, JCD 22.5 Hz, 13CDHOH), 63.9 (t,
JCD 21.6 Hz, 13CHDOH), 64.2 (s, 13CH2OH)], 71.1 (d, J
40.2 Hz, CH), 72.1 (d, J 2.0 Hz, CH2OBn), 73.7 (CH2Ph),
128.1, 128.1 (C4%), 128.7, 138.4 (C1%); wmax/cm−1 3379,
2864; m/z (EI+) 185 (M+, 26%), 107 (58), 91 (100).
[1-13C,18O]-Glycerol 1a. A mixture of benzyl ether 6a
(1.85 g, 10.0 mmol) and 10% palladium on carbon (718
mg) in ethanol (50 mL) was stirred vigorously at room
temperature under a hydrogen atmosphere of initial pres-
sure 6 atm. The reaction was completed after 16 h when
one equivalent of hydrogen had been consumed. The
reaction mixture was filtered through Celite, washed with
1H NMR (CD2Cl2, 360 MHz) 1.25 (t, 3H, J 7.2 Hz,
CH3), 3.12 (br s, 1H, OH), 3.73 (m, 2H, CH2), 4.21 (q,
2H, J 7.2 Hz, CH2CH3), 4.27 (m, 1H, CH), 4.53 (s, 2H,
CH2Ph), 7.28–7.37 (m, 5H, ArH); 13C NMR (CD2Cl2, 90
3
2
MHz) 14.3 (d, CH3, J 2.0 Hz), 62.1 (d, CH2CH3, J 2.7
Hz), 71.1 (d, J 59.9 Hz, CH), 72.0 (CH2), 73.7 (CH2),
128.0, 128.0 (C4%), 128.7, 138.3 (C1%), 173.02 (13C=18O,
89%), 173.06 (13C=16O, 11%); vmax/cm-1 3452, 2980,
2869, 1664; m/z (EI) 227 (M+, 20%), 179 (24), 166 (11),
121 (34), 91 (84), 84 (100).
Ethyl
[1-13C]-3-benzyloxy-2-hydroxypropionate
5b.
Method as for 5a. Nitrile 3 (2.468 g, 13.9 mmol) in dry
ethanol (40 mL) and acetyl chloride (total of 12.2 equiv.)

792
A. P. Siskos, A. M. Hill / Tetrahedron Letters 44 (2003) 789–792
ethanol and the filtrate evaporated in vacuo. The crude
oil was redissolved in ethanol and re-evaporated (bath
temperature 560°C) to remove any remaining toluene.
The title compound was obtained as an almost colourless
oil (908 mg, 95%). 18O enrichment is 89% by 13C NMR.
Found [M+1]+ 96.0622, calculated for C213CH9O128O
96.0628.
benzyl ether 6b (2.04 g, 11.0 mmol) and 10% palladium
on carbon (730 mg) in ethanol (55 mL). The title com-
pound was obtained as an almost colourless oil (1.03 g,
98%). Found [M+1]+ 96.0661, calculated for
C213CH9O218O 96.0709.
1H NMR (D2O, 360 MHz) 3.48–3.53 (m, 1H, CHHOH),
2
3.58–3.64 (m, 1H, CHHOH), 3.71–3.75 (m, 1H, CH); H
1H NMR (D2O, 360 MHz) 3.29–3.82 (m, 5H, 2CH2,
CH); 13C NMR (D2O, 90 MHz) 63.28 (13CH218OH, 89%),
63.30 (CH2OH); 72.9 (d, J 41.2 Hz); wmax/cm−1 3332; m/z
(EI+) 96 ([M+1]+, 25%), 77 (46), 61 (100).
NMR (H2O, 61 MHz) 3.49 (d, JCD 21.8 Hz, CDD), 3.58
(d, JCD 21.7 Hz, CDD); 13C NMR (D2O, 90 MHz) 62.6
(quin., JCD 21.7 Hz, 13CD2OH), 63.3 (CH2OH); 72.7 (d, J
41.1 Hz); wmax/cm−1 3370; m/z (EI+) 96 ([M+1]+, 12%), 77
(61), 64 (100).
[1-13C,2H2]-Glycerol 1b. Method as for 1a. A mixture of