14-Hydroxylation of opiates: Catalytic direct autoxidation of codeinone to 14-hydroxycodeinone

Article abstract of DOI:10.1021/ja051682z

Codeinone (3) was efficiently and directly converted to 14-hydroxycodeinone (1) by catalytic air oxidation in aqueous solution. A number of simple manganese and copper salts were identified to be effective catalysts, including MnSO₄, KMnO₄, and CuSO₄. An appropriate reducing agent, such as sodium thiosulfate, is required in the reaction mixture presumably for the reduction of a detrimental peroxide intermediate. This discovery allows the more abundant codeine to be employed as the starting material for the synthesis of 14-hydroxylated opiate drugs without recourse to a thebaine-like intermediate. These discoveries were inspired from our study of microbial transformation of codeine to 14-hydroxycodeine by Mycobacterium neoaurum, where we found the actual 14-hydroxylation step is a chemical reaction rather than an enzymatic reaction, as previously believed. Copyright

Full text of DOI:10.1021/ja051682z

Published on Web 04/27/2005
14-Hydroxylation of Opiates: Catalytic Direct Autoxidation of Codeinone to
14-Hydroxycodeinone
Qibo Zhang, Joseph O. Rich,† Ian C. Cotterill, David P. Pantaleone,‡ and Peter C. Michels*
Albany Molecular Research, Inc., Bioscience DiVision, 21 Corporate Circle, P.O. Box 15098,
Albany, New York 12212-5098
Received March 16, 2005; E-mail: peter.michels@albmolecular.com
Scheme 1
The introduction of a 14-hydroxyl group to the morphine alkaloid
structure is known to improve its pharmacological properties.1 This
group of drugs includes oxycodone, naloxone, oxymorphone, and
naltrexone with well-established clinical uses for analgesia, anti-
tussive, and treatment of drug abuse and alcohol addiction.2 The
common starting material for the synthesis of many 14-hydroxylated
opiates is thebaine, a minor component of natural opium extracts,
which is converted by peracid oxidation to 14-hydroxycodeinone
(1), the key 14-hydroxylated intermediate.3 Because codeine (2)
and 3 are directly involved in the formation of 4, a sample of 95%
enriched [6-D]codeine was subjected to biotransformation by M.
1
neoaurum. H NMR analysis of the resulting isolated 4 indicated
and morphine are much more available than thebaine, many
a complete loss of the deuterium label during the conversion. A
approaches have been reported for their conversion to 14-hydroxy-
sample of the remaining codeine was also recovered from the same
lated opiates.4 However, each of these methods is plagued by low
1
experiment, and H NMR analysis found the majority (83%) of
yields, the formation of difficult-to-separate byproducts, or the use
the deuterium label retained, indicating the deuterium loss due to
of undesirable heavy metals. For instance, significant effort has
the equilibrium catalyzed by the dehydrogenase is slow. Thus, this
not resulted in practical success for the preparation of 14-
deuterium labeling experiment strongly supports that 3 is an
intermediate during the biotransformation of 2 to 4 by M. neoaurum.
We also observed a significantly slow turnover rate for the
hydroxycodeinone directly from codeinone (3), which can be easily
prepared from codeine.5 Indeed, it was previously concluded after
a systematic study that “the direct conversion of [codeinone] to
deuterated substrate (4 days, 21%) compared to that for unlabeled
[14-hydroxycodeinone] cannot be performed with the vast majority
codeine (3 days, ∼35%). This kinetic isotope effect suggests that
of common oxidizing agents”.6
the dehydrogenase-catalyzed reaction is slow, and the 14-hydroxy-
The development of a bioprocess for the 14-hydroxylation of
morphine alkaloids has also attracted great interest.7 A number of
microorganisms have been identified for the 14-hydroxylation of
codeine and morphine.7 The strain that has received the most
lation step must be fast. The absence of 3 from the fermentation
medium during the reaction, along with the initial accumulation
and resulting disappearance of 1, also supports the hypothesis that
the hydroxylation of 3 is much faster than the dehydrogenase-
attention in the past decade is Pseudomonas putida M10, which
catalyzed oxidation of 2 to 3.
converts codeine and morphine to 14-hydroxycodeine (4) and 14-
A CFE and membrane fractions of P. putida M10 have been
hydroxymorphine, respectively.8 Interestingly, this microbial trans-
reported to catalyze the 14-hydroxylation of 3 without the require-
ment of any cofactors, and the activity was found to be extremely
formation has been demonstrated to be the result of a sequence of
reactions (Scheme 1). Briefly, codeine is converted to 3 by morphine
heat-stable.7a We found that a CFE of M. neoaurum was unable to
dehydrogenase (MDH), followed by hydroxylation of 3 to form 1,
catalyze the 14-hydroxylation of 3. Conversely, excellent activity
which is then converted to 4 by the same MDH. While the MDH
was detected in the spent medium, which was prepared from a 72
h fermentation broth after removal of all cells. When 3 (2 mM)
and the spent medium (4 mL, adjusted to pH 8.0) were kept in a
has been well characterized, cloned, and expressed, the key
hydroxylation activity was barely detected.7a
During our study of opiates, we identified Mycobacterium
shaker at 29 °C and 300 rpm for 2 h, HPLC analysis indicated that
neoaurum (MTP650) to be an efficient microorganism for the
nearly all 3 was converted to 1 (∼60%) and other products. Similar
biotransformation of 2 to 4 and a few other derivatives (see
activity was also observed after the spent medium was boiled for
30 min. Size-exclusion fractionation experiments determined that
Supporting Information), similar to those reported for P. putida
M10. We report here our evidence that the 14-hydroxylation of 3
the activity was in a fraction with low molecular weight (<1 kD).
by M. neoaurum is a chemical hydroxylation, rather than an
These facts strongly support that the 14-hydroxylation of 3 is not
enzyme-catalyzed reaction. More importantly, we have discovered
catalyzed by an enzyme as previously believed. When the reaction
a simple chemical reaction for the direct conversion of 3 to 1 in
was carried out using fresh medium (pH 8.0), a complex mixture
was generated with 1 only as a minor component (<5%). Clearly,
a component(s) of the spent medium facilitates the 14-hydroxylation
reaction.
aqueous solution, providing an attractive strategy for the synthesis
of 14-hydroxylated opiates starting from codeine.
The presence of a dehydrogenase in M. neoaurum was implicated
by the identification of 1 as one of the bioconversion products and
Upon consideration of the possible reaction mechanisms and the
confirmed by the conversion (7%) of 1 to 4 with a cell-free extract
requirement of molecular oxygen for the 14-hydroxylation of 3,
(CFE) in the presence of NADPH. To verify that the dehydrogenase
we envisioned that a peroxide intermediate might be involved. The
† Current address: High Throughput Analysis Facility, Northwestern University,
2205 Tech Drive, Evanston, IL 60208.
spent medium may contain a reducing equivalent that can reduce
the peroxide intermediate to form the product. Because the fresh
medium contains no reductant, a peroxide intermediate may undergo
‡ Current address: Baxter Healthcare Corporation, Renal Division, 1620
Waukegan Road, MPGR-D1, McGaw Park, IL 60085.
9
7286
J. AM. CHEM. SOC. 2005, 127, 7286-7287
10.1021/ja051682z CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2
the oxidation state of the catalytic metal species. Thiourea was found
to be another efficient reductant capable of replacing thiosulfate,
whereas a variety of other reagents could also be used for the
reaction with somewhat decreased yield, including sodium iodide,
sodium thiocyanate, 2-ketoglutarate, methionine, N-acetylmethion-
ine, and alanine. The identification of these reductants suggests that
the reaction using spent medium is likely accomplished with the
combined reduction by a number of reducing agents, such as amino-
and sulfur-containing compounds generated during the fermentation.
In summary, we elucidated the key 14-hydroxylation step during
the microbial transformation of 2 to 4. Our results clearly suggest
that the 14-hydroxylation of 3 is a chemical reaction, rather than
an enzymatic reaction, as previously believed. Most importantly,
we have developed an efficient method for the direct 14-hydroxy-
lation of 3, thereby allowing the more abundant codeine to be
employed as the starting material for the synthesis of 14-hydroxy-
lated opiate drugs. Because of the use of cheap inorganic reductant,
free oxidant, and very low amounts of more environmentally benign
metal catalysts in an aqueous solution, the catalytic hydroxylation
presents a great opportunity to significantly improve process
economics, especially through decreased waste treatment/disposal
costs.
further complex chemistry. To test this hypothesis, a number of
simple reductants were individually included in the reaction mixture.
When sodium thiosulfate (5 mM) was included in a reaction mixture
of 3 (2 mM) in fresh medium (4 mL, pH 8.0), 1 was found to be
the main product (∼85%) after the mixture was kept at 29 °C and
300 rpm for 2 h. A control experiment with sodium thiosulfate in
phosphate buffer (100 mM, pH 8.0) resulted in unchanged 3.
Obviously, the reaction is catalyzed by a component(s) of the fresh
medium. Because the fresh medium (Mineral Salts Broth) is
chemically defined, deconvolution of medium components identified
MnSO4 and CuSO4 as the most active catalysts for the hydroxylation
of 3 in the presence of molecular oxygen and thiosulfate. The
concentrations of MnSO4 and CuSO4 in the medium are only 7.4
and 1.3 µM, respectively, indicating they are very effective catalysts.
A typical reaction mixture (5 mL) consisted of 2 mM codeinone,
2.5 mM thiosulfate, and 7 µM MnSO4 in phosphate buffer (75 mM,
pH 8.0). The reaction was completed in about 3 h at 29 °C and
300 rpm in a rotary shaker, provided that sufficient O2 was included
in the reaction vessel. Manganese reagents with other valencies
were also found to be effective catalysts for the reaction, including
Mn(OAc)3, MnO2, KMnO4, and even Mn2(CO)10. In the presence
of O2 and thiosulfate, all these manganese reagents may be
converted to the same catalytic species, such as Mn(III). The
reaction could be explained by the initial coordination of Mn(III)
with codeinone dienol (5) at slightly basic conditions (Scheme 2).
The dienolate (6) transfers one electron to manganese to generate
a stabilized tertiary radical species (7), followed by the incorporation
of O2 to form a peroxy radical species (8), which may obtain one
electron back from the manganese to yield a peroxy anion and to
regenerate the catalytic species. The peroxy anion then picks up a
proton to form 14-hydroperoxycodeinone (9), which is reduced to
1 by thiosulfate. This mechanism is supported by the positive
detection of peroxide species in the reaction mixture using a
peroxide paper when the reaction was carried out in the absence
of a reductant. Such a radical mechanism is also consistent with
those of other autoxidation reactions,9 including the Mn(III)-
catalyzed R-hydroperoxidation of 1,3-dicarbonyl compounds10 and
the γ-hydroxylation of R,â-unsaturated ketones.11 It was intriguing
to find that there is a lag phase of about 1 h for the reaction with
MnSO4 (see Supporting Information), whereas no obvious lag phase
was observed for the copper-catalyzed reaction. One possible
explanation is that the Mn(II) species needs to be oxidized to the
Mn(III) species by 8 initially generated from spontaneous autoxi-
dation of 3, and this is known to be very slow.7a The reaction with
KMnO4 completes in about 1 h and does not have a lag phase,
perhaps because the Mn(III) species is formed directly by the
thiosulfate reduction of permanganate.
Acknowledgment. We thank Dr. Yuri Khmelnitsky for helpful
discussions, and Matthew Chase, Amanda Madjid-Yunus, Danilo
Sumague, Hemant Patel, and Jennifer Ton for screening microbial
catalysts and for their assistance in fermentation experiments.
Supporting Information Available: Experimental procedures and
profiles of biotransformation and lag phase of MnSO4. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Kobylecki, R. J.; Carling, R. W.; Lord, J. A. H.; Smith, C. F. C.; Lane,
A. C. J. Med. Chem. 1982, 25, 116-120.
(2) (a) Bryant, R. J. Chem. Ind. 1988, 146-153. (b) Mello, N. K.; Mendelson,
J. H.; Kuehnle, J. C.; Sellers, M. S. J. Pharmacol. Exp. Ther. 1981, 216,
45-54. (c) Martin, W. R. Ann. Intern. Med. 1976, 85, 765-768. (d)
Altshuler, H. L. Drug Alcohol Depend. 1979, 4, 333-346.
(3) Lo´pez, D.; Quio`oa´, E.; Riguera, R. Tetrahedron Lett. 1994, 31, 5727-
5730.
(4) (a) Seki, I. Chem. Pharm. Bull. 1970, 18, 671-676. (b) Schwart, M. A.;
Wallace, R. A. J. Med. Chem. 1981, 24, 1525-1528. (c) Wallace, A.
U.S. Patent 5,112,975, 1992. (d) Huang, B.-S.; Lu, Y.; Ji, B. Y.;
Christodoulou, A. P. U.S. Patent 6,008,355, 1999. (e) Chiu, F.-T.; Lo, Y.
S. U.S. Patent 6,177,567, 2001. (f) Ninan, A.; Sainsbury, M. Tetrahedron
1992, 32, 6709-6716. (g) Barber, R. B.; Rapoport, H. J. Med. Chem.
1975, 18, 1074-1077. (h) Schwartz, M. A.; Mimi, I. S. J. Am. Chem.
Soc. 1975, 97, 1239-1240. (i) Rapoport, H.; Lovell, C. H.; Reist, H. R.;
Warren, M. E., Jr. J. Am. Chem. Soc. 1967, 89, 1942-1947. (j) Brown,
I.; Martin-Smith, M. J. Chem. Soc. 1960, 82, 4139-4140. (k) Walker, A.
J.; Bruce, N. C. Tetrahedron 2004, 60, 561-568. (l) Linders, J. T. M.;
Vrijhof, P. WO PCT 03/018588, 2003.
(5) Coop, A.; Rice, K. C. WO PCT 00/66588, 2000.
(6) Coop, A.; Rice, K. C. Tetrahedron 1999, 55, 11429-11436.
(7) (a) Lister, D. L.; Kanungo, G.; Rathbone, D. A.; Bruce, N. C. FEMS
Microbiol. Lett. 1999, 181, 137-144. (b) Liras, P.; Umbreit, W. W. Appl.
Microbiol. 1975, 30, 262-266. (c) Liras, P.; Kasparian, S. S.; Umbreit,
W. W. Appl. Microbiol. 1975, 30, 650-656. (d) Madyastha, K. M.; Reddy,
G. V. B.; Sridhar, G. R. Indian J. Chem. B 1998, 37, 749-753. (e) Kunz,
D. A.; Reddy, G. S.; Vatvars, A. Appl. EnViron. Microb. 1985, 50, 831-
836.
(8) (a) Boonstra, B.; Rathbone, D. A.; Bruce, N. C. Biomol. Eng. 2001, 18,
41-47. (b) Rathbone, D. A.; Bruce, N. C. In Industrial and EnVironmental
Biotechnology; Ahmed, N., Qureshi, F. M., Khan, O. Y., Eds.; Horizon
Scientific Press: Wymondham, U.K., 2001; Chapter 9 and references
therein.
(9) Kerdesky, F. A. J.; Ardecky, R. J.; Lakshmikantham, M. V.; Cava, M. P.
J. Am. Chem. Soc. 1981, 103, 1992-1996.
(10) Rahman, M. T.; Nishino, H. Org. Lett. 2003, 5, 2887-2890.
(11) Sabol, M. R.; Wiglesworth, C.; Watt, D. S. Synth. Commun. 1988, 18,
1-12.
The addition of an appropriate reductant is critical for the
hydroxylation reaction because it may facilitate the reduction of
the peroxy intermediate generated during the reaction and affect
JA051682Z
9
J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005 7287