The design and synthesis of a 5-HETE affinity chromatography ligand for 5-hydroxyeicosanoid dehydrogenase

Article abstract of DOI:10.1016/S0040-4039(00)00971-0

The first total synthesis of an ω-amino 5-HETE derivative 27 has been accomplished by a new counter-clockwise strategy, in which C-1 is constructed first and C-20 last. The ω-amino 5-HETE derivative was transformed to an affinity chromatography ligand, the biotinylated 5-HETE 30. This affinity chromatography ligand is aimed at purifying the 5-hydroxyeicosanoid dehydrogenase enzyme, which is responsible for the conversion of 5-HETE to 5- oxo-ETE, a potent eosinophil chemotactic factor. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00971-0

Tetrahedron Letters 41 (2000) 5807±5811
The design and synthesis of a 5-HETE anity
chromatography ligand for 5-hydroxyeicosanoid
dehydrogenase
Subhash P. Khanapure,a,y Goutam Saha,a William S. Powellb and Joshua Rokacha,*
aClaude Pepper Institute and Department of Chemistry, Florida Institute of Technology, 150 W. University Blvd.,
Melbourne, FL 32901, USA
bMeakins-Christie Laboratories and Department of Medicine, McGill University, 3626 St-Urbain St., Montreal,
QC H2X 2P2, Canada
Received 2 May 2000; revised 18 May 2000; accepted 19 May 2000
Abstract
The ®rst total synthesis of an o-amino 5-HETE derivative 27 has been accomplished by a new counter-
clockwise strategy, in which C-1 is constructed ®rst and C-20 last. The o-amino 5-HETE derivative
was transformed to an anity chromatography ligand, the biotinylated 5-HETE 30. This anity
chromatography ligand is aimed at purifying the 5-hydroxyeicosanoid dehydrogenase enzyme, which is
responsible for the conversion of 5-HETE to 5-oxo-ETE, a potent eosinophil chemotactic factor. # 2000
Elsevier Science Ltd. All rights reserved.
Keywords: biotin; 5-HETE; 5-hydroxyeicosanoid dehydrogenase; anity chromatography.
Our interest has recently centered on oxo derivatives of eicosanoids, many of which have
potent biological activities or are key intermediates in the synthesis of other potent mediators,
such as the 12-oxo-10,11-dihydro metabolite of 12-HETE, which is converted to the potent
angiogenic factor 12-HETrE.
We have recently reported on the synthesis,1 9 biosynthesis2 and biology3,4 of several oxo-
eicosanoids, such as 12-oxo-ETE,1 10,11-dihydro-12-oxo-ETE,5 12-oxo-LTB4,6 10,11-dihydro-12-
7
oxo-LTB4 and 5-oxo-12-HETE,8 and ®nally 5-oxo-ETE2 and its tetradeutero and tetratritiated
derivatives,9 which were essential for the development of LC/MS methodology for measuring
5-oxo-ETE in biological ¯uids and for the radioactive study of the biotransformation of 5-oxo-
ETE and receptor binding studies.
* Corresponding author. Tel: 407 674 7329; fax: 407 952 1818; e-mail: jrokach@®t.edu
y
Present address: NitroMed, Inc., 12 Oak Park Drive, Bedford, MA 01730, USA
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00971-0

5808
The enzymatic machinery for the formation and metabolism of 5-oxo-ETE is only partially
understood.10 One of our goals is to purify, characterize and identify the 5-hydroxyeicosanoid
dehydrogenase (5h-dh) responsible for the transformation of 5-HETE 2 to 5-oxo-ETE 3 (Scheme
1). To accomplish this we propose to synthesize radioactively labeled photoanity probes and
ligands for anity chromatography based on the structure of the substrate 5-HETE. A similar
approach can be used to characterize enzymes metabolizing 5-oxo-ETE as well as the 5-oxo-ETE
receptor.
Scheme 1. Four proteins involved in the biosyntheses, metabolism and biological activity of 5-oxo-ETE
We wish to report here the total synthesis of a prototype for such reagents, an o-biotinylated
derivative of 5-HETE 5, as shown in Scheme 2. This is the ®rst report of the synthesis of an
o-substituted derivative of either 5-HETE or 5-oxo-ETE.
Scheme 2.
Biotinylated 5-HETE can be used to purify 5h-dh by adsorbing it on a streptavidin±agarose
column (Scheme 2). Streptavidin is a protein that has four high anity biotin binding sites
(Kd=10^15 M)11 and would thus strongly bind 5. The simplest approach to prepare an anity
ligand based on 5-HETE would have been to attach a biotin residue via the carboxyl group.
However, it seemed likely that this group would be important for the interaction of 5-HETE with
the active site of 5h-dh, and this was con®rmed by our ®nding that the methyl ester of 5-HETE is
a very poor substrate for this enzyme. On the other hand, modi®cation of the o-end of the
molecule (e.g. 5,15-diHETE) has little e€ect on the rate of metabolism. For this reason, we
decided to attach the biotin molecule to the o-end of 5-HETE, leaving the rest of the molecule
una€ected. To couple 5-HETE to biotin in this way requires introduction of an NH2 or OH
linking group, as shown in Scheme 2. It should be noted that in this initial approach we have
chosen to synthesize an anity ligand based on the 14,15-dihydro derivative of 5-HETE based on
previous experience (unpublished results) in which a 14,15-dihydro derivative performed as well
as the unsaturated derivative.
We adopted a novel approach for the synthesis of 5, which will provide us with considerable
¯exibility for the future syntheses of a variety of other related o-derivatives of 5-HETE and 5-
oxo-ETE. In the past, most, if not all, of the eicosanoid syntheses that we and others have carried
out have been performed in a clockwise manner, i.e. building the bottom part of the structure

5809
®rst, starting at the o-end of the molecule (Scheme 3). Such was the case in our synthesis of LTB4
7 and its o-OH and o-COOH derivatives, as well as the 14,15-acetylene derivative. However, in
the case of the o-amino-5-HETE derivative 6 and its biotinylated form 5, we elected to use a
counterclockwise approach, starting at the carboxyl end of the molecule. The advantages of such
a strategy in the present case are obvious, as it allows us to modify the o-end of the molecule as
required without repeating the whole synthetic sequence. The objective in the present study is to
keep the top part of the molecule intact, since we know from preliminary structure±activity
experiments that this part is necessary for recognition by the catalytic site of 5h-dh.
Scheme 3.
Scheme 4 illustrates the strategy used for the construction of the o-amino derivative 27 and the
biotinylated derivative 5, which will be used for attachment to a streptavidin±agarose support.
The required synthon 16 containing the 5(S)-OH group, which we have used previously, was
prepared by the modi®ed and improved synthesis shown in Scheme 4. Of particular note, the
dithio ether deprotection was performed by our recently published method12 using periodic acid
in ether/THF. At the end of the reaction (15 min) the ether THF solvent containing a white
precipitate was ®ltered through a small amount of Celite/Florisil. The solution was extracted with
KH2PO4 bu€er followed by extraction with Na2S2O3 solution. The organic layer was dried and
evaporated to a€ord a quantitative yield of the desired 11.
Another improvement in the synthesis of the aldehyde 16 follows. Previously, we used to
perform the transformation of 14 to the aldehyde 16 in two steps, the ®rst being the deprotection
of the acetonide to a€ord the diol 15, and the second the lead tetraacetate oxidation to 16. In the
present case, we performed the transformation of 14 to 16 in one step8,13 using excess periodic
acid for 14 h (92% yield). Whereas the two-step procedure used by us before6,14 was very ecient
(two-step yield 66%), the one-step periodic acid procedure is shorter, more convenient and occurs
in excellent yield.
Synthon 24 is one major objective of this synthetic approach. As can be seen, a Wittig reaction
with 24 a€orded the desired o-azido compound 26 and eventually the amino derivative 27.
Synthon 24 can obviously be used as a pivotal intermediate for the synthesis of other 5-HETE
derivatives containing di€erent o substitutions. The OTHP deprotection (step j in Scheme 4) and
the Wittig coupling of 24 with 19 (step m) occur in modest yields. We thought that it was worth
paying this price in order to bene®t from the anticlockwise strategy of this synthetic approach.
Our ®rst preparation of this synthon was one in which the oxygen was substituted by a benzoate
group 25. A reason for initially selecting 25 lies in the fact that we have used this synthon in the
synthesis of LTB4 with excellent yield (Scheme 5).
When 25 was used in the Wittig reaction with 19, a complex mixture of products was obtained.
We attribute this unsuccessful reaction to the possibility that the Wittig anion caused the
deprotection of the benzoate group followed by secondary reactions, e.g. lactonization, etc.

5810
Scheme 4. Reagents and conditions: (a) tert-butylchlorodiphenylsilane, imidazole, CH2Cl2, 0ꢀC to RT, 97%; (b)
H5IO6, ether, THF, 0ꢀC, 15 min, quant.; (c) benzene, re¯ux, 4 h, 85%; (d) Pd±C/H2, quant.; (e) H5IO6, ether/THF, RT,
18 h, 92%; (f) NaN3, DMSO, 90ꢀC, 99%; (g) PCC, CH2Cl2, 0ꢀC to RT, 77%; (h) (triphenylphosphoranylidene)-
acetaldehyde, benzene, re¯ux, 78%; (i) LiHMDS, 20, ^78ꢀC to RT, 90%; (j) 1N HCl, THF:H2O (9:1), RT, 40 h, 48%;
(k) p-TsCl, Py, 0ꢀC to RT, 82%; (l) NaI, acetone, 65ꢀC, 99%; (m) Ph3P, CH3CN, 65ꢀC, 99%; (n) LiHMDS, 19, ^78ꢀC
to RT, 36%; (o) Ph3P, THF/H2O, RT, 88%; (p) (+)-biotin, EDCl, HOBT, NaHCO3, DMF, RT, 76%; (q) 1N NaOH
(aq), MeOH, 92%; (r) THF, TBAF:HOAc (1:1), RT, 80 h, 74%
In the original LTB4 synthesis,14 shown in Scheme 5, the reaction between 31 and 32 worked
well, presumably because of the stabilized nature of the phosphonium anion, which was not
suciently basic or nucleophilic to deprotect the benzoate.
Scheme 5.
Preliminary studies show that the anity chromatography ligand 3015,16 is recognized by the
5-hydroxy dehydrogenase from human PMN and is a good substrate for this enzyme. We are

5811
currently attempting to optimize the anity of the ligand for the enzyme by testing the e€ects of
structural modi®cations of 5 on enzyme activity. We will then be in a position to use the
biotinylated 5-HETE derivative to purify 5h-dh. In addition, the o-amino intermediate 27 can be
used to synthesize a radiolabeled azidophenyl derivative of 5-HETE that can be used to label
5h-dh, which will make it possible to detect the enzyme on electrophoretic gels.
Acknowledgements
We wish to acknowledge NIH support under Grants DK-44730 (JR), NSF for an AMX-360
NMR instrument (Grant CHE-90-13145) and the MRC of Canada under Grant MT-6254
(WSP).
References
1. Wang, S. S.; Rokach, J.; Powell, W. S.; Dekle, C.; Feinmark, S. J. Tetrahedron Lett. 1994, 35, 4051±4054.
2. Khanapure, S. P.; Shi, X. X.; Powell, W. S.; Rokach, J. J. Org. Chem. 1998, 63, 337±342.
3. Berhane, K.; Ray, A. A.; Khanapure, S. P.; Rokach, J.; Powell, W. S. J. Biol. Chem. 1998, 273, 20951±20959.
4. Stamatiou, P.; Hamid, Q.; Taha, R.; Yu, W.; Issekutz, T. B.; Rokach, J.; Khanapure, S. P.; Powell, W. S. J. Clin.
Invest. 1998, 102, 2165±2172.
5. Wang, S. S.; Shi, X.-X.; Powell, W. S.; Tieman, T.; Feinmark, S. J.; Rokach, J. Tetrahedron Lett. 1995, 36, 513±
516.
6. Khanapure, S. P.; Manna, S.; Rokach, J.; Murphy, R. C.; Wheelan, P.; Powell, W. S. J. Org. Chem. 1995, 60,
1806±1813.
7. Khanapure, S. P.; Wang, S. S.; Powell, W. S.; Rokach, J. J. Org. Chem. 1997, 62, 325±330.
8. Khanapure, S. P.; Powell, W. S.; Rokach, J. J. Org. Chem. 1998, 63, 8976±8982.
9. Khanapure, S. P.; Shi, X. X.; Powell, W. S.; Rokach, J. J. Org. Chem. 1998, 63, 4098±4102.
10. Powell, W. S.; Gravelle, F.; Gravel, S. J. Biol. Chem. 1992, 267, 19233±19241.
11. Green, N. M. Avidin. In Advances in Protein Chemistry; Academic Press: New York, 1975; Vol. 29, pp. 85±133.
12. Shi, X.-X.; Khanapure, S. P.; Rokach, J. Tetrahedron Lett. 1996, 37, 4331±4334.
13. Wu, W.-L.; Wu, Y.-L. J. Org. Chem. 1993, 58, 3586±3588.
14. Zamboni, R.; Rokach, J. Tetrahedron Lett. 1982, 23, 2631±2634.
15. NMR data of compound 30: 1H NMR (360 MHz, CDCl3) ꢀ 6.55 (dd, J=15.1 and 11.1 Hz, 1H), 6.1 (m, 1H), 5.99
(t, J=10.8 Hz, 1H), 5.40 (m, 3H), 5.24 (m, 1H), 4.54 (m, 1H), 4.38 (m, 1H), 4.21 (m, 1H), 3.24 (m, 3H), 2.94 (m,
3H), 2.74 (d, 13 Hz, 1H), 2.38 (m, 2H), 2.22 (m, 2H), 2.05 (m, 2H), 1.53±1.87 (m, 11H), 1.47 (m, 2H), 1.2±1.38 (m,
13H).
16. Compound 30 has been found to be a substrate for 5h-dh. Incubation of 30 with human neutrophil microsome
produces the expected 5-oxo product. Further measurement, quantitation and Km measurement are being
determined and will be published separately.