Photogeneration of an o-quinone methide from pyridoxine (vitamin B6) in aqueous solution

Article abstract of DOI:10.1039/a707231g

Photolysis (254, 266 or 308 nm) of pyridoxine (vitamin B₆) in aqueous solution gives an o-quinone methide efficiently which is trapped by MeOH and ethyl vinyl ether.

Full text of DOI:10.1039/a707231g

View Article Online / Journal Homepage / Table of Contents for this issue
Photogeneration of an o-quinone methide from pyridoxine (vitamin B ) in
6
aqueous solution
Darryl Brousmiche and Peter Wan*†
Department of Chemistry, Box 3065, University of Victoria, Victoria, British Columbia, Canada V8W 3V6
Photolysis (254, 266 or 308 nm) of pyridoxine (vitamin B
6
) in
ether. Interestingly, none of the product studies show any
evidence for the formation of the m-QM, thereby indicating that
o-QM 2 is formed selectively.
aqueous solution gives an o-quinone methide efficiently
which is trapped by MeOH and ethyl vinyl ether.
2
Photolysis of 1 in 1:1 H O–MeCN with 0.26 m ethyl vinyl
Over the past several years there has been a great deal of interest
in the biological chemistry of quinone methides‡ (QMs), due
mainly to their toxicological properties against both normal and
cancerous cells, as well as their proposed intermediacy in the
formation of many biologically important polymers (melanin,
lignin and insect cuticles).1 These properties have been
attributed to the electrophilic nature of QMs which can result in
both alkylation of DNA and amino acids, and in self
polymerization. QM reactivity in biological systems has been
studied extensively with QM precursors such as butylated
hydroxytoluene (BHT), eugenol, mitomycin C and the anthra-
cyclines.1 The formation of these QMs has traditionally been
via thermal reaction, through the use of enzymes in vivo,
oxidation of phenols, or nucleophilic substitution of silyl or
quartenary ammonium groups, amongst others.1 Recently,
however, our group has developed a clean and efficient
photochemical method for the formation of o-, m- and p-QMs
via UV (254 nm) photolysis of hydroxybenzyl alcohols.6
ether (EVE) gave the Diels–Alder adduct 4¶ cleanly ( > 70%) in
a regioselective fashion (Scheme 1). Formation of 4 is only
possible through the intermediacy of a 1,4-dipolar species and
therefore provides conclusive evidence for the efficient photo-
generation of o-QM 2.
We have shown that laser flash photolysis (LFP) provides an
effective method for the direct detection of appropriately
,2
6
substituted o-, m- and p-QMs. The choice of 5 for these studies
Ph
Ph
CHOH
CHOMe
CH2OH
,2
HO
Me
CH2OH
HO
Me
N
N
–5
5
6
Ph
Ph
O
CH OH
Pyridoxine 1 (vitamin B
6
) provides an ideal system for the
2
O
2
CH OH
continued study of photochemically generating QMs as it is a
biologically relevant molecule and it contains the required
hydroxybenzyl alcohol-type system. Moreover, it allows for a
competition between the formation of an o-QM vs. a m-QM due
Me
N
H
N
+
Me
7
(lmax 370 nm)
8 (lmax 430 nm)
2
to the presence of a second CH OH group meta to the aromatic
hydroxy group. Formation of a QM from 1 via exposure to UV
light§ could lead to cell and DNA damage which to the best of
our knowledge has not been explored.
will enable the resulting QM to be more readily detectable due
to the expected longer wavelength of absorption and corre-
spondingly larger extinction coefficient. LFP of 5 (lexc = 266
or 308 nm, YAG or excimer lasers, < 20 mJ per pulse) in 100%
2
4
Photolysis of 1 and 5 (10 m; Rayonet photoreactor; 254
nm; ca. 15 °C; argon) in 1 :1 MeOH–H O gave the corre-
2
H O (oxygen purged to remove triplet states and possible
2
sponding methyl ethers 3 and 6 (f ≈ 0.2 for reaction of 5)
cleanly ( > 40%) at low conversion (Scheme 1). The location of
the methoxy group was unambiguously assigned based on NOE
radical species) yielded a species at 370 nm, which is observed
at pH 7 (t > 10 ms) and 12 (t ca. 2 ms) (Fig. 1). When LFP
experiments were carried out at pH 1, this band can be seen
6
data. Previous work by this group has shown that the quantum
yield of methyl ether formation from m-QMs is approximately
half of that from the corresponding o-QM. Thus, if the m-QM
was formed competitively with the o-QM of 1, it would be
easily discernable by formation of the corresponding methyl
0
.08
0.06
0.04
CH2OH
HO
Me
CH2OH
O
CH OH
2
i
0
.02
.00
N
Me
N
0
1
2
iii
ii
–0.02
–0.04
–0.06
EtO
CH2OMe
HO
Me
CH2OH
O
CH2OH
N
Me
N
350
400
450
500
550
600
650
700
4
3
l / nm
Scheme 1 Reagents and conditions: i, hn; ii, MeOH–H
0.26 m), MeCN–H O (1:1)
2
O (1:1); iii, EVE
2 2
Fig. 1 LFP spectra (lmax 308 nm; 100% H O; O purged) of 5 at (8) pH 1,
(/) 7 and (2) 12
(
2
Chem. Commun., 1998
491

View Article Online
below 350 nm (t < 50 ms). In all three cases, however, there is
rapid drop-off signal (bleaching) as the wavelength of observa-
tion approaches the region where the ground state material
absorbs,∑ thereby masking the true extent of this band. Another
species with lmax at 430 nm is apparent in the pH 7 (t > 10 ms)
and 12 (t ca. 2 ms) spectra (Fig. 1). The relative intensities of
the bands change with pH (there is no 430 nm band at pH 1),
with an ‘inflection’ point in the pH 4–7 region. We have
assigned the 370 nm band to o-QM 7 which is protonated at
nitrogen. This is clearly the most basic site of the o-QM;
protonation at oxygen (the other possible basic site) is unlikely
at this pH as this would generate a highly reactive diarylmethyl
carbocation. The 430 nm band is thus assigned to neutral o-QM
that 2 can be formed inside biological systems, leading to cell
and DNA damage. We believe this to be the first example of the
photogeneration of a quinomethane from an important bio-
molecule (i.e. 1). Since pyridoxal is the active form of
pyridoxine, and is known to be extensively hydrated in aqueous
solution, we are investigating the possibility of analogous
photochemistry for this compound.
We acknowledge support of this research by the Natural
Sciences and Engineering Research Council (NSERC) of
Canada. D. B. thanks the University of Victoria for a graduate
fellowship.
Notes and References
8.
†
‡
§
E-mail: pwan@uvic.ca
IUPAC name: quinomethanes.
o-QM 2 has been generated thermally from 1 at 130–190 °C and trapped
The lifetime of simple QMs is expected to be lower in basic
+
2
and acidic media than at pH 7, due to attack by either H or OH
at the appropriate sites of the QM (carbonyl oxygen and
exocyclic vinyl carbon, respectively). The transient lifetimes at
pH 7 for both 7 and 8 are approximately five times longer than
at pH 12, while the lifetime of 7 at pH 1 is approximately two
hundred-fold shorter than in neutral solution, consistent with
QM reactivity. The heteroatom present in the pyridine ring will
also have an influence on the QM lifetime at low pH, where it
is fully protonated: it should act a powerful electron-with-
drawing group, making the QM more reactive, and this is
consistent with the much shorter lifetime observed in pH 1.
It has been shown via LFP,6 in the case of the simple
hydroxybenzhydrol systems, that at elevated pH ( > 10) QMs
are formed more efficiently (higher quantum yields) than at
neutral pH, as the phenolate is already present. This appears to
be verified in our system as much stronger signals are observed
when LFP experiments are carried out at pH 12. Moreover,
with various nucleophiles (ref. 7).
¶ Selected data for 4: d (300 MHz, CDCl
1.96 (m, 2 H, Ar CH CH ), 2.25 (s, 3 H, ArCH
ArCH CH ), 3.65, 3.78 [two sets of dq (diastereotopic Hs), J 7.4, 10.3, 2 H,
CH CH O), 4.50 (s, 2 H, ArCH OH), 5.33 (t, J 3.3, 1 H, CH CH OCH),
.78 (s, 1 H, ArH).
H
3
) 1.07 (t, J 7.4, 3 H, CH
3 2
CH O),
2
2
3
), 2.70 (t, J 7.4, 2 H,
2
2
3
2
2
3
2
7
∑
According to UV–VIS data, 5 is fully protonated (at the nitrogen) at pH 1
(
l
max 290 nm), is in its free base form at pH 7 (lmax 325 nm) and is in its
2
ArO form at pH 12 (lmax 310 nm), in accordance with literature data for
the parent 1 (ref. 8).
1 D. C. Thompson, J. A. Thompson, M. Sugumaran and P. Moldeus, Chem.
Biol. Interact., 1992, 86, 129.
2
J. L. Bolton, H. Sevestre, B. O. Ibe and J. A. Thompson, Chem. Res.
Toxicol., 1990, 3, 65; J. L. Bolton, L. G. Valerio Jr. and J. A. Thompson,
Chem. Res. Toxicol., 1992, 5, 816; J. L. Bolton, N. M. Acay and V.
Vukomanovic, Chem. Res. Toxicol., 1994, 7, 443; J. L. Bolton, G. R. J.
Thatcher and P. G. McCracken, J. Org. Chem., 1997, 62, 1820.
K. Karabelas and H. W. Moore, J. Am. Chem. Soc., 1990, 112, 5372.
P. D. Gardner, H. Rafsanjani and L. Rand, J. Am. Chem. Soc., 1959, 81,
3364.
2
product studies on the formation of 3 from 1 (1:1 MeOH–H O)
3
4
at pH 7 and 12 gave yields of 9 and 15% (performed under low
conversion conditions and in which samples received the same
UV dose), respectively, consistent with the notion that the QM
is more efficiently formed at high pH.
5 M. S. Chauhan, F. M. Dean, D. Matkin and M. L. Robinson, J. Chem.
Soc., Perkin Trans. 1, 1973, 120.
6
P. Wan, L. Diao and C. Yang, J. Am. Chem. Soc., 1995, 117, 5369; P.
Wan, B. Barker, L. Diao, M. Fischer, Y. Shi and C. Yang, Can. J. Chem.,
In summary, we have shown that the corresponding o-QMs of
and 5 can be formed readily in aqueous solution via irradiation
1
1
996, 74, 465.
M. Frater-Schr o¨ der and M. Mahrer-Busato, Bioorg. Chem., 1975, 4,
32.
S. A. Harris, T. J. Webb and K. Folkers, J. Am. Chem. Soc., 1940, 62,
198.
with UV light. The o-QM is formed selectively in all cases,
although m-QM formation is a possibility. The QM can exist as
the free base or in the iminium ion form, which differ in
reactivity. The pH of the solvent leads to significant differences
in both the lifetime and amount of QM formed, with the QM
being longest-lived at pH 7. These results suggest the possibility
7
8
3
3
Received in Corvallis, OR, USA, 6th October 1997; 7/07231G
492
Chem. Commun., 1998