In vitro enzymatic oxidation of a fluorine-tagged sulfido substrate analogue: A19F NMR investigation

Article abstract of DOI:10.1002/mrc.1797

¹H-decoupled ¹⁹F NMR has been used to monitor the highly regioselective oxidation of a fluorine-tagged thia-fatty acid derivative by castor stearoyl-ACP Δ⁹ desaturase. The major enzymatic product, after reductive work-up, was identified as 9-fluoro-1-nonanol. This compound could be easily distinguished from substrate and a 9-sulfoxy by-product on the basis of its ¹⁹F NMR chemical shift and spiking experiments using authentic standards. Structural assignment of the cleavage product was confirmed by GC-MS analysis of the enzymatic products. Copyright

Full text of DOI:10.1002/mrc.1797

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2006; 44: 629–632
Published online 14 March 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1797
Note
In vitro enzymatic oxidation of a fluorine-tagged
sulfido substrate analogue: a 19F NMR investigation
Amy E. Tremblay,1 Peter H. Buist,1∗ Derek Hodgson,2 Brian Dawson,2 Ed Whittle3 and
John Shanklin3
1
Department of Chemistry, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario, Canada K1S 5B6
Centre for Biologics Research, Biologics and Genetic Therapies Directorate, Health Canada, Tunney’s Pasture, Ottawa, Ontario, Canada K1A 0L2
Brookhaven National Laboratory, Biology Department, Upton, NY 11973, USA
2
3
Received 2 December 2005; Revised 25 January 2006; Accepted 26 January 2006
1H-decoupled 19F NMR has been used to monitor the highly regioselective oxidation of a fluorine-tagged
thia-fatty acid derivative by castor stearoyl-ACP 19 desaturase. The major enzymatic product, after
reductive work-up, was identified as 9-fluoro-1-nonanol. This compound could be easily distinguished
from substrate and a 9-sulfoxy by-product on the basis of its 19F NMR chemical shift and spiking
experiments using authentic standards. Structural assignment of the cleavage product was confirmed by
GC-MS analysis of the enzymatic products. Copyright  2006 John Wiley & Sons, Ltd.
KEYWORDS: NMR; 19F substituent effects; oxidation; desaturase
INTRODUCTION
RESULTS AND DISCUSSION
The ACP thioester derivative of 18-fluoro-9-thiaoctadecanoic
acid (1-ACP, 160 nmol) was incubated for 40 min with freshly
prepared 9 desaturase (25 nmol) and cofactors essentially as
previously reported.7 The enzymatic reaction was quenched
by the addition of THF/NaBH4 and the resultant terminal
alcohols were extracted with dichloromethane; previous
experiments have demonstrated that 19F NMR signals of
related compounds in aqueous solutions are substantially
broadened. An aliquot of the organic extract was allowed
to evaporate and the residue so obtained examined by
1H-decoupled 19F NMR after the addition of CDCl3 (570 µl).
The 19F NMR spectrum of the product mixture was similar
to that obtained previously6 and is presented in Fig. 1(A). As
alluded to in the Introduction, previous experiments6,7 had
suggested that the major fluorinated analyte (U) observed
at ꢀ218.23 ppm might be derived from chain cleavage of
the substrate (Scheme 1). Accordingly, 9-fluorononanol was
prepared from commercially available 9-bromononanol (See
Experimental Section) and an appropriate amount of this
material (7 nmol) was added to the enzymatic product
mixture (Fig. 1(B)). A single, unbroadened signal of enhanced
intensity relative to the minor peaks was obtained at the
expected chemical shift (ꢀ218.23 ppm). The identification
of compound U as 9-fluoro-1-nonanol was confirmed by
GC-MS analysis of a portion of the enzymatic extract after
silylation (See Experimental Section for GC conditions). A
GC peak corresponding to the TMS derivative of 9-fluoro-
1-nonanol was observed at the anticipated retention time
(Fig. 2(A)). The mass spectrum of this analyte also matched
that of the reference standard (Fig. 2(B)).
19F NMR is a useful technique for monitoring the biotransfor-
mation of fluorine-tagged substrates at the trace analytical
level.1 We have recently developed a methodology that
features the use of ω-fluorinated fatty acyl thia-analogues
such as 1-ACP to probe for oxo-transfer reactions carried
out by an important class of enzymes known as desaturases
(Scheme 1).2 The latter enzymes are critical to lipid biosyn-
thesis in virtually all aerobic life forms3,4 but are relatively
poorly understood from a mechanistic point of view. Our
fluorine-based approach was tested recently using an in vivo
membrane-bound yeast 9 desaturating system5 and a puri-
fied in vitro soluble plant 9 desaturase enzyme.6 In both
cases, regioselective sulfoxidation could be easily monitored
by 19F NMR owing to long-range substituent effects on a
remote fluorine reporter group. Interestingly, in the case of
the plant 9 desaturase, we were able to detect an unknown
fluorinated product (Compound U) produced from 1-ACP
in addition to the sulfoxide product.6 We proposed that this
compound was a chain-cleavage product6,7 and in this paper
we confirm this hypothesis.
ŁCorrespondence to: Peter H. Buist, Department of Chemistry,
Carleton University, 1125 Colonel By Drive, Ottawa, Ontario,
Canada K1S 5B6. E-mail: pbuist@ccs.carleton.ca
Contract/grant sponsor: NSERC; Contract/grant number:
OGP-2528.
Contract/grant sponsor: Office of Basic Energy Sciences of the
United States Department of Energy.
Copyright  2006 John Wiley & Sons, Ltd.

630
A. E. Tremblay et al.
Scheme 1
Two of the minor peaks in the 19F NMR spectrum of
the enzymatic product mixture were identified as those
corresponding to the residual substrate (S, ꢀ218.22 ppm)
and sulfoxide product (SO, ꢀ218.29) via spiking experiments
with authentic reference standards (data not shown). Two
minor signals (ꢀ218.19, ꢀ218.28 ppm) remain unidentified;
these were not present in the 19F NMR spectrum of the
starting material (S) (detection limit: 0.05% of major analyte),
and the GC-MS analysis of the extract did not reveal
appreciable amounts of any analyte in addition to the TMS
derivative of 9-fluorononanol (the sulfoxide by-product is
thermally labile and is not GC-observable). It should be noted
that owing to the enhanced volatility of 9-fluoro-1-nonanol
relative to longer chain compounds, the intensity of these
minor peaks in the 19F NMR spectrum has, in all likelihood,
been artificially enhanced upon evaporative concentration of
the extract.
The identification of compound U substantiates our
previous prediction that 1-ACP undergoes desaturase-
mediated oxidation primarily ˛ to sulfur owing to a strict
enzyme-imposed regiochemical imperative.6,7 Early work
using thia-analogues as mechanistic probes8 also led to
the hypothesis that when the enzyme oxidant (probably
a hypervalent diirondioxo complex)3 is not optimally
aligned with the sulfur atom of a thia-substrate, a chain-
cleavage process can occur. Our data correlate well with
the results obtained recently by Fox and coworkers who
used mass spectrometry to analyze enzymatic products from
9 desaturase-mediated oxidation of the nonfluorinated
analogue of 1-ACP.9 One possible pathway to compound
U is illustrated in Scheme 2 and involves hydroxylation at
C-10 followed by spontaneous collapse of the intermediate
hemithioketal to aldehyde and thiol. NaBH4 reductive work-
up generates 9-fluorononanol (U) from the intermediate
aldehyde.
Figure 1. 1H-decoupled 19F NMR (376.5 MHz) spectra of the
products (CDCl3 solution) obtained from 9 desaturase-
mediated oxidation of 18-fluoro-9-thiaoctadecanoyl ACP
before (A) and after (B) addition of 9-fluorononanol reference
standard. Analytes were obtained by reductive work-up of the
enzymatic reaction using NaBH4, followed by CH2Cl2
extraction. *Unidentified resonances. S D Substrate;
SO D 9-sulfoxide product; U D Previously unidentified
resonance; 9F1-OL D 9-fluorononanol.
CONCLUSIONS
A novel reaction product of a desaturase-mediated reaction
1
was detected for the first time by H-decoupled 19F NMR6
Copyright  2006 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2006; 44: 629–632

19F NMR probe for enzymatic sulfide oxidation 631
Figure 2. Partial GC-MS TIC chromatogram of products (TMS derivatives) obtained from 9 desaturase-mediated oxidation of
18-fluoro-9-thiaoctadecanoyl ACP (A) and TMS derivative of reference standard, 9-fluorononanol (B). Mass spectral analysis of
major enzymatic product U (C) and TMS derivative of reference standard, 9-fluorononanol (D). U D Previously unidentified enzymatic
product; 9F1-OTMS D TMS derivative of 9-fluorononanol.
NMR and GC-MS. This result, in combination with other
mechanistic information,2 suggests that the putative diiron-
dioxo oxidant initiates the parent dehydrogenation reaction
at C-10 (Scheme 1). The lack of background interferences,
high sensitivity and wide chemical shift range of 19F NMR
greatly enhance the utility of a remote fluorine substrate tag
in probing enzymatic reactions. We are currently exploring
other applications of 19F NMR in the trace analytical mode
that will contribute to the bioorganic investigation of enzyme
mechanisms.
EXPERIMENTAL
Materials
18-fluoro-9-thiaoctadecanoic acid and the corresponding
sulfoxide were available from a previous study.6 9-Fluoro-1-
nonanol was synthesized by fluorination of the correspond-
ing commercially available 9-bromo-1-nonanol following a
standard literature procedure:10 A solution of tetrabutylam-
monium fluoride TBAF (5.86 g, 22 mmol) in THF (45 ml)
was stirred over 3A molecular sieves under N2 for 1 h.
9-bromo-1-nonanol (1.02 g, 4.5 mmol) in THF (10 ml) was
transferred to the TBAF solution and stirred under N2 at
Scheme 2
r.t. for 3 h. After THF was removed in vacuo, H2O (20 ml)
was added to the resultant syrup and the aqueous layer
and in the present work, the predicted structure for this
compound has now been independently confirmed by 19F
extracted with hexanes (4 ð 20 ml). The organic layer was
washed with sat. NaCl (30 ml) and dried (Na2SO4ꢀ. The crude
Copyright  2006 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2006; 44: 629–632

632
A. E. Tremblay et al.
product (456 mg) was purified using flash chromatography
(Hex/EtOAc 1 : 20) to remove residual starting material. A
sample of 9-fluoro-1-nonanol was obtained as a colorless
liquid11 (55 mg, >99% pure by GC-MS). The analytical data
NMR measurements
19F NMR spectra were recorded at 300 K on a Bruker
AVANCE 400 (9.4T) spectrometer operating at 376.50 MHz
with a dedicated 5 mm 19F/1H probe and a 19F-specific
preamplifier. A Bruker 1H band-pass/19F band-stop filter
was used in the proton channel and a 19F band-pass/1H
band-stop filter was used in the fluorine observe channel
for all acquisitions. WALTZ-16 was used for proton decou-
pling. Standard microprograms from Bruker software were
employed. All spectra were run using 128 K data points
with a spectral width of 37 650 Hz, which gave a final
spectral resolution of 0.287 Hz. Exponential multiplication
with a line broadening of 0.28 Hz was applied. The spec-
tra were acquired using the 90 degree pulse, a delay time
of 1.0 s, an acquisition time of 1.74 s, and 20 000 scans. All
chemical shifts are referenced to neat external trichlorofluo-
romethane (CFCl3ꢀ at 0.00 ppm. Samples were dissolved in
0.57 ml CDCl3, which had been previously filtered through
neutral alumina.
1
(MS, H, 13C, 19F NMR) of this material was in accord with
expectations: TLC (SiO2 hexane/EtOAc 70 : 30): Rf 0.21. IR
(film): ꢁmax 3336, 2929, 2856, 1465, 1391, 1058, 723 cmꢀ1
.
2
1H-NMR (400 MHz, CDCl3ꢀ: 4.44 (dt, JHF D 47.4, 2 H);
3
3
3.64 (t, JHH D 6.6, 2 H); 1.69 (dm, JHF D 25.0, 2 H); 1.57
3
(tt, JHH D 7.2, 2 H); 1.45 (bd s, 1 H); 1.28–1.45 (m, 10 H).
13C-NMR (100 MHz, CDCl3ꢀ: 84.23 (d, JCF D 164.0); 63.02;
1
2
32.76; 30.39 (d, JCF D 19.4); 29.45; 29.31; 29.16; 25.70; 25.13
(d, 3JCF D 5.5). 19F NMR (376.5 MHz, CDCl3ꢀ: ꢀ218.23. EI-MS:
m/z 144 (0.4, [M – 18]Cꢀ, 116 (19), 88 (32), 69 (66), 55 (100),
41 (79). EI-MS (TMS derivative): m/z 233 (<1, [M – H]Cꢀ; 214
(<1, [M – HF]Cꢀ; 199 (<1, [M – HF, CH3]Cꢀ, 107 (52), 103(49),
83(65), 69(100), 73(55), 55(88).
Incubations
The purification of castor stearoyl-ACP 9 desaturase,
required cofactors and the synthesis of substrate ACP
derivatives has been previously described.12,13 Incubations
of acyl-ACP derivatives (10 ð 16 nmol batches) with freshly
prepared 9 desaturase (2.5 nmol) were carried out at room
temperature. The reaction was initiated by the addition of
NADPH (1.0 mg, 1.2 µmol) in 100 mM Tricine, pH 8.0 buffer
(50 µl) and allowed to continue for 30 min. The reaction
was terminated with the addition of THF (100 µl), and
the thioester linkage of the ACP derivatives was reduced
Acknowledgements
This work was supported by NSERC (USRA to A.T.; Grant OGP-
2528 to P. H. B.) and by the Office of Basic Energy Sciences of the
United States Department of Energy (J.S.).
REFERENCES
1. Marino R, Gilard V, Desmoulin F, Malet-Martino M. J. Pharm.
Biomed. Anal. 2005; 38: 871.
2. Buist PH. Nat. Prod. Rep. 2004; 21: 249.
3. Shanklin J, Cahoon EB. Annu. Rev. Plant Physiol. Plant Mol. Biol.
1998; 49: 611.
4. Sperling P, Ternes P, Zank TK, Heinz E. Prostag. Leukotr. Ess.
2003; 68: 73.
5. Hodgson DJ, Lao KYY, Dawson B, Buist PH. Helv. Chim. Acta
2003; 86: 3688.
6. Behrouzian B, Hodgson D, Savile CK, Dawson B, Buist PH,
Shanklin J. Magn. Res. Chem. 2002; 40: 524.
7. Behrouzian B, Buist PH, Shanklin J. J. Chem. Soc. Chem. Commun.
2001; 401.
8. Buist PH, Marecak D. J. Am. Chem. Soc. 1992; 114: 5073.
9. White RD, Fox BG. Biochemistry 2003; 42: 7828.
10. Cox DP, Terpinksi J, Lawrynowicz W. J. Org. Chem. 1984; 49:
3216.
11. Pattison FLM, Howell WC, McNamara AJ, Schneider JC,
Walker JF. J. Am. Chem. Soc. 1956; 21: 739.
12. Studts JM, Fox BG. Protein Expr. Purif. 1999; 16: 109.
13. Shanklin J. Protein Expr. Purif. 2000; 19: 319.
°
to the corresponding alcohol with NaBH4 at 37 C for
15 min. The residue was diluted with H2O (1 ml) and
extracted with CH2Cl2 (2 ð 2 ml). The phases were separated
by centrifugation; the organic layer was collected and
°
concentrated by passive evaporation for 15 h at 20 C. Owing
to the volatility of the major analyte (9-fluorononanol),
evaporation of the extract under a N2 stream was avoided.
The product mixture was analyzed directly by 19F NMR
and by GC/MS (60 m Supelco 2340 poly(biscyanopropyl
siloxane) 0.25 mm ID capillary column with 1.5 ml/min
flow rate operated in the splitless mode with an inlet
°
°
temperature of 250 C and a gradient of 100–140 C at
°
5 C/min) after silylation. Using the latter technique, it could
be demonstrated that the enzymatic reaction had consumed
all but traces of the substrate. This observation was confirmed
by subsequent 19F NMR analysis.
Copyright  2006 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2006; 44: 629–632