Possible role of hydroxyl radicals in the oxidative degradation of folic acid

Article abstract of DOI:10.1016/j.bmcl.2004.10.038

An unprecedented ^.OH radical-induced N-dealkylation of amines including folic acid is reported, the reaction proceeding via a direct hydrogen atom abstraction mechanism as revealed by pulse radiolysis experiments. Hydroxyl radicals have been found to cause oxidative N-dealkylation of amines including folic acid via a hydrogen atom transfer mechanism.

Full text of DOI:10.1016/j.bmcl.2004.10.038

Bioorganic & Medicinal Chemistry Letters 15 (2005) 67–71
Possible role of hydroxyl radicals in the oxidative
degradation of folic acid
Birija S. Patro,a Soumyakanti Adhikari,b Tulsi Mukherjeeb and Subrata Chattopadhyaya,*
aBio-Organic Division, Bhabha Atomic Research Centre, Mumbai 400 085, India
bRadiation Chemistry and Chemical Dynamics Division, Bhabha Atomic Research Centre, Mumbai 400 085, India
Received 11 August 2004; accepted 12 October 2004
Available online 5 November 2004
Abstract—Hydroxyl radicals have been found to cause oxidative N-dealkylation of amines including folic acid via a hydrogen atom
transfer mechanism.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
vealed by the Bratton–Marshall (BM) procedure2c and
2,4-DNPH test respectively. Thus, cleavage of 1 by
ÅOH radicals was apparent. However, the above Fenton
system can generate the Fe4+–OH complex3 which is also
implicated in the hemin-catalyzed dealkylation of vari-
ous aÅmines1a including 1.2c Consequently, 1 was reacted
with OH, generated by an iron independent, radiolytic
process, which also led to the same amine 2 and aldehyde
3b. The time dependent growths of 2 and 3b were fol-
lowed spectrophotometrically over a period of 0–7h,
by recording the absorbances at 550nm for the BM
derivative of 2 and at 497nm for the DNPH derivative
of 3b (Fig. 1). The product distribution and characteriza-
tion were also carried out by paper chromatography and
HPLC by comparing with authentic samples. The time
dependent HPLC analyses showed 3b as the major pterin
product, while a linear relationship between the deple-
tion of 1 and production of 2 was also observed.
Oxidative N-dealkylation of amines, catalyzed by sev-
eral oxidative hemin-containing enzymes and hemin-
hydroperoxides combinations, is of growing interest1a–e
due to its central role in xenobiotic metabolism. One
of the important examples of this biochemical reaction
is the C9–N10 bond cleavage (Scheme 1) of folic acid
(1) leading to the formation of the amine 2 and 6-substi-
tuted pterins 3a–cÀ. ÅFor this, participation of the super-
or a ferryl–hydroxo (Fe –OH)
4+
2a,b
oxide radicals (O2
)
complex2c,d has been reported. Herein, we report the
possible role of the ubiquitous hydroxyl radicals in the
reaction, and provide insights about the reaction
mechanism.
2. Results and discussion
Å
Exposure of 1 to the hydroxyl radical generating Fenton
system (Fe2+–EDTA–H2O2) furnished 2 and 3b as re-
The possible involvement of the OH radicals in amine
dealkylation was ascertained further by carrying out
N
N
N
N
N
H2N
H2N
H
N
H2N
N
N
+
R
N
COGlu
OH
OH
1
2
COGlu
3a-c
a/b/c: R = Me/CHO/CO H
2
Scheme 1.
Keywords: Folic acid; N-Dealkylation; Pulse radiolysis; Radicals.
*
Corresponding author. Tel.: +91 22 5593703; fax: +91 22 5505151; e-mail: schatt@apsara.barc.ernet.in
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.10.038

68
B. S. Patro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 67–71
3.0
2.5
2.0
1.5
1.0
0.5
0.0
2.5
497nm
396nm
550nm
2.0
1.5
1.0
550nm
0.5
0.0
0
2
4
6
8
Irradiation time/h
0
2
4
6
8
Irradiation time/ h
Figure 2. Time dependent growth of 5 and 6 formed during oxidation
Å
Figure 1. Time dependent growth of 2 and 3b formed during oxidation
Å
of 4 by radiolytically generated OH. (–j–) DNPH derivative of 6
(absorbance at 396nm); (–d–) BM derivative of 5 (absorbance at
550nm). Values are mean SE (n = 4).
of 1 by radiolytically generated OH. (–j–) DNPH derivative of 3b
(absorbance at 497nm); (–d–) BM derivative of 2 (absorbance at
550nm). Values are mean SE (n = 4).
Å
the reaction of the radiolytically generated OH radicals
125
100
75
50
25
0
with 4, a simpler analogue of 1. This also produced the
expected cleavage products 5 and 6 (Scheme 2), which
were quantified spectrophotometrically (Fig. 2) and
1
identified from their H NMR spectra.
ÀÅ
Given the reported2a,b involvement of the O2 in the
above reaction, its reactions with 1 and 4 were also stud-
ied. The degradations of 1 and 4 were significantly less
(40–50%) in these cases that was consistent with the
less reactivity of the O2ÀÅ compared to the ÅOH radicals.
Between 1 and 4, more dealkylation products were
formed with the latter (Fig. 3). Besides hydrogen atom
transfer (HAT), the O2ÀÅ radicals can react with 1 in var-
ious other pathways including reduction of its pterin
moiety. With 4, however, only a HAT is possible, result-
ing in higher amount of its oxidized metabolites.
0
1
2
3
4
5
6
7
8
Irradiation Time/h
Figure 3. Time dependent growth of 2 and 5 formed during oxidation
of 1 and 4 by O2ÀÅ. (–j–) BM derivative of 2 (for 1); (–d–) BM
derivative of 5 (for 4). Values are mean SE (n = 4).
Next, we turned our attention to the mechanism of the
ÅOH-mediated oxidation. In spite of being long known,
the exact mechanism of oxidation of 1 remains consider-
ably speculative. The reaction can proceed either
through an initial HAT or a single electron transfer
(SET) process as shown in Scheme 3. For example, the
key intermediate 1d of the reaction can be produced
either via the (i) HAT intermediate 1b, or (ii) SET inter-
mediate 1a through 1c which also involves a HAT as the
secondary process. The third process involving a proton
transfer from 1a to 1b can be excluded in the present
case, due to the poor basicity of the system.
oxidative N-dealkylation. Further, distinction between
the intermediacy of 1b and 1c could not be made as gen-
eration of both the intermediates involve a hydrogen
atom transfer. The mechanistic controversies possibly
arose because of overemphasis on indirect evidences,
which were not applicable universally, and recognition
of the minor reactions as the major ones. Very recently,
in a more direct approach, two groups have estab-
lished1d,e the operation of a HAT mechanism in the
cytochrome P450-catalyzed dealkylation of cyclopropyl
amines, by characterizing the oxidized metabolites.
However, these evidences could only be generated
because of the presence of the cyclopropyl group in
Alternate views involving HAT1a,d,e,c and SET mecha-
nisms1b,c have been proposed by different groups for
˚
PhCHNH
PhCH2NH
CO2Et
CO2Et
H2N
CO Et
+
PhCHO
2
4
6
4a
5
Scheme 2.

B. S. Patro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 67–71
69
.
-H
.
+
+
Ar NH CH2Pte
Ar NH CHPte
-
1c
-e
..
1a
Pte = Pterin
Ar NH CH2Pte
+
+H O
2
-H
.
1
..
.
.
-H
+
Ar NH CHPte
PteCHO
ArNH2
Ar NH CHPte
.
1b
OH
1d
2
Scheme 3.
the chosen substrates and may not be used for all kinds
of substrates such as 1.
constant of 9 · 109 dm3 molÀ1 sÀ1) and decay of the tran-
sients matched exactly and hence these were assigned to
the same radical species. Given that 4 is nonphenolic,
the absence of any transient absorption peak at
ꢀ430nm was anticipated. The transient absorption at
390nm decayed faster in the presence of oxygen (Fig.
4, inset), revealing it to be a carbon-centred radical.
The oxygen effect on transient was studied only with
the signal at 390nm in order to avoid the possible inter-
ference from the bleaching signal at ꢀ310nm. Moreo-
ver, compound 1 also showed the transient absorption
at 390nm only. The above results, in conjunction with
those obtained4a by us with 1, clearly revealed that the
ÅOH radical induced oxidation of 1 or 4 proceeds via
the HAT mechanism. Absence of any absorption due
to a N-centred radical excluded the possibility of any
SET mechanism in the reaction.
We envisaged that looking into the transients formed in
the oxidation reaction and following their reaction
dynamics by pulse radiolysis technique might provide
more direct evidence. Earlier, using the pulse radiolytic
studies, we have shown4a that the reaction between
ÅOH radical and 1 leads to the bleaching in the ground
state absorption around 300nm and a transient absorp-
tion maximum at 425nm at 10ls after electron pulse,
which shifted to 390nm at 800ls. The absorption peak
at 425nm, formed with a bimolecular rate constant of
1.13 · 1010 dm3 molÀ1 sÀ1 was attributed to a ꢀpseudo-
phenoxylꢁ radical. The absorption at 425nm decayed
in ls time scale, while that at 390nm grew steadily and
remained stable even up to seconds. Thus, generation
of the transient (kmax 390nm) both directly from 1 as
well as from the ꢀpseudo-phenoxylꢁ intermediate was in-
ferred. However, presence of multiple reactive centres in
1 precluded its exact identification and we proposed it to
be a molecular radical. It was envisaged that the pulse
radiolysis studies of the reaction using 4 as a simpler
congener of 1 would be better suited for the characteri-
zation of the intermediate.
The transients obtained with 1 and 4 were similar both
in terms of their kmax and formation rate constants.
Hence, these can be safely attributed to similar methy-
lenic radical intermediates, 1b (for 1) and 4a (for 4).
The hydrogen abstraction from the designated CH2
group is expected, as the radical 1b or 4a will be reso-
nance stabilized by the adjacent, pteryl/phenyl and amino
groups. Formation of a similar methylenic radical
absorbing at 490nm was earlier reported for the reac-
Å
Reaction of 4 with OH radicals furnished two transient
absorption peaks at ꢀ310 and 390nm (Fig. 4). Kinetic
Å
tion of curcumin, a b-diketone with the CH3 radical.4b
studies on the formation (bimolecular formation rate
Unlike the curcumin-derived radical, 4a is devoid of
extended conjugation, explaining the blue shift of its
absorption. The other C-centred radicals that may be
Å
produced by OH addition to the double bonds in 1 or
0.03
0.2
(a)
4 would not furnish the N-dealkylated products, without
getting transformed into intermediates like 1b or 4a.
Such a possibility of radical interconversion could be ex-
cluded as the pulse radiolysis oscilloscope traces of the
390nm transient with 4 were smooth without any shoul-
der and did not show any evolution within the small
time scale of the study. With 1 also, the intermediate
1b was generated by a direct hydrogen abstraction at
C-9 as well as via evolution from the ꢀpseudo-phenolicꢁ
intermediate, but not through any N-centred radical.
a
0.1
0.0
(b)
(c)
0.02
0.01
b
25
50
Time/µs
75 100
0.00
-0.01
3. Conclusion
300
400
Wavelength/nm
500
600
Overall, the present study revealed the participation of
freely diffusible hydroxyl radicals in the oxidative degra-
dation of secondary amines including 1. Based on direct
evidences, the involvement of a HAT mechanism in the
reaction has been shown unambiguously. Given the high
cellular concentrations of free iron and H2O2, the
Figure 4. Transient absorption spectra obtained from an N2O-
saturated aqueous solution containing 4 (100lM) at pH8.0 after the
electron pulse at (a) 5ls and (b) 40ls. Dose = 12Gy. Inset: absorbance
versus time plots for the traces recorded at 390nm obtained from the
above experiment. Different ratios of air and N2O were used. (a)
without air, (b) 1:4 air–N2O and (c) 1:1 air–N2O.

70
B. S. Patro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 67–71
Table 1. HPLC data for the oxidative degradation of folic acid (1)
biotoxic hydroxyl radicals may play a major role in oxi-
dative degradation of folic acid in the cellular systems.
Although the physiological implication of the process
is not clear, this may partly explain the increased need
of folic acid supplementation in case of iron-overload
diseases.
Peak no.
Retention time (min)
Product
1
2
3
4
5
1.26
2.28
p-Aminobenzoic acid and 2
3c
3b
4.37
15.98
22.26
1
Unknown
4. Experimental
4.1. Materials
4.4. HPLC analysis of the reaction products from 1 and
ÅOH
Folic acid (Aldrich) was used as received. All solutions
were made with triply distilled water. High purity
N2O, from BOC India Pvt. Ltd, was used for all radio-
lytic experiments. Compounds 3b,c and 4 were synthe-
sized as reported.5a,b
The tubes containing the solution (each of 1mL) of 1
(500lM) in N2O purged H2O were irradiated with c-
ray for different periods (0–7h). The products obtained,
were analyzed by HPLC (Bruker HPLC instrument)
under the following conditions: RP-18 column (E. Merck,
Germany); eluent flow rate: 1mL/min; detection: 255nm
and injection volume: 20lL. The eluent was prepared by
mixing NaC2O4ÆH2O (35.1g), KH2PO4 (1.36g), aqueous
KOH (6.94mL, 1N) and methanol (40mL), making-up
the volume to 1L and adjusting the pH to 7.2 with aque-
ous KOH. The peak areas were calculated by using a
Bruker HPLC software. The results presented in Table
1 correlated well with those reported.6
4.2. Cleavage of 1 with Fenton (Fe2+–EDTA–H2O2)
reagent
The reaction mixture (1mL) contained 1 (2.8mM),
(NH4)2Fe(SO4)2Æ6H2O (80lM), EDTA (100lM)
[EDTA and (NH4)2Fe(SO4)2Æ6H2O were mixed prior
to the addition of 1], H2O2 (200lM), KH2PO4–KOH
(10mM) pH7.4 buffer and 1 (250lM). After incubating
the mixture at 37ꢁC for 1–7h in absence or presence
(300lM) of vitamin C, a solution of TBA in 50mM
NaOH (1mL, 1% w/v) and trichloroacetic acid (1mL,
2.8% w/v aqueous solution) was added. The reaction
mixture was heated for 15min in boiling water bath
and the amount of chromogen produced was spectro-
photometrically measured4a at 532nm.
The paper chromatography was carried out using
butanol–acetic acid–water (4:1:5) mixture as the solvent
system.
ÀÅ
4.5. Cleavage of 1 or 4 with O2 radicals
The aqueous solution (1.0mL) containing 1 (500lM)
or 4 (500lM) and formate (0.1M) was irradiated with
c-ray (60Co source, dose rate 560Gy/h) for 0–7h.
Å
4.3. Cleavage of 1 or 4 with OH radicals and spectro-
photometric analyses of products
4.6. Pulse radiolysis studies
An N2O purged aqueous solution (1.0mL) of 1 (500lM)
or 4 (500lM) was irradiated with c-ray (60Co source,
dose rate 560Gy/h) for 0–8h. At different time intervals,
a fixed volume of the solution was withdrawn, diluted
with an equal volume of water and treated with HCl
(40lL, 0.5N) followed by NaNO2 (100lL, 0.1% aque-
ous solution). After 5min, ammonium sulfamate
(100lL, 0.5% aqueous solution) was added, the mixture
shaken and kept for a further 5min at room tempera-
ture. Finally, the Bratton–Marshall reagent, NEDD
(100lL, 0.1% aqueous solution) was added and the
chromogen was read spectrophotometrically at
550nm.2c
The pulse radiolysis system using 7MeV electrons
has been described earlier.7 The dosimetry was
carried out using an air-saturated aqueous solu-
tion
containing 5 · 10À2 moldmÀ3 KSCN (Ge =
23,889dm3 molÀ1 cmÀ1 per 100eV at 500nm.8 The
kinetic spectrophotometric detection system covered
the wavelength range from 250 to 800nm. The optical
path length of the cell was 1.0cm. The width of the elec-
tron pulse was 50ns and the dose was 16Gy per pulse.
The pH of the solution was adjusted by adding NaOH
Å
or HClO4. The OH radicals were generated by pulse
radiolysis of N2O saturated aqueous solution as
described earlier.4a
For detecting the aldehyde 3b, 2,4-DNPH (50lL,
25mM in 2N HCl) was added to the irradiated solution
of 1 (250lL) followed by NaOH (200lL, 1N) after
15min, and the absorbance of its hydrazone derivative
was monitored at 497nm.
The bimolecular rate constants were calculated by plot-
ting the pseudo-first order rate of formation of the tran-
sient against the solute concentration. The uncertainty
in the measurement in bimolecular rate constant was
10%.
For assaying 6, 2,4-DNPH (50lL, 25mM in 2N HCl)
was added to the irradiated solution of 4 (250lL), the
mixture kept for 20min and the precipitate formed
was centrifuged at 13,000rpm for 20min. The pellet
was washed twice with 2N HCl, dissolved in DMSO
(500lL) and the absorbance at 396nm was read.
References and notes
1. (a) Miwa, G. T.; Walsh, J. S.; Kedderis, G. L.; Hollenberg,
P. F J. Biol. Chem. 1983, 258, 1445–1449; (b) Guengerich,
F. P. Chem. Res. Toxicol. 2001, 14, 611–650, and references

B. S. Patro et al. / Bioorg. Med. Chem. Lett. 15 (2005) 67–71
71
cited therein; (c) Mitchell, D. J.; Nikolic, D.; Ribera, E.;
Sablin, S. O.; Choi, S.; van Breeman, R. B.; Singer, T. P.;
Silverman, R. B.. Biochemistry 2001, 40, 5447–5456; (d)
Bhakta, M. N.; Wimalasena, K. J. Am. Chem. Soc. 2002,
124, 1844–1845; (e) Shaffer, C. L.; Harriman, S.; Koen, Y.
M.; Hanzlik, R. P. J. Am. Chem. Soc. 2002, 124, 8268–8274.
2. (a) Shaw, S.; Jayatilleke, E.; Herbert, V.; Colman, N.
Biochemistry 1989, 257, 277–280; (b) Vasquez-Vivar, J.;
Whitsett, J.; Martasek, P.; Hogg, N.; Kalyanaraman, B.
Free Rad. Biol. Med. 2001, 31, 975–985; (c) Taher, M. M.;
Lakshmaiah, N. Arch. Biochem. Biophys. 1987, 257, 100–
106; (d) Innocentini, L. H.; Duran, N. Braz. J. Med. Biol.
Res. 1982, 15, 11–16.
(b) Jovanovic, S. V.; Boone, C. W.; Steenken, S.; Trinoga,
M.; Kaskey, R. B. J. Am. Chem. Soc. 2001, 123, 3064–
3068.
5. (a) Waller, C. W.; Goldman, A. A.; Angier, R. B.; Boothe,
J. H.; Hutchings, B. L.; Mowat, J. H.; Semb, J. J. Am.
Chem. Soc. 1950, 72, 4630–4633; (b) Mowat, J. H.; Boothe,
J. H.; Hutchings, B. L.; Stokstad, E. L. R.; Waller, C. W.;
Angier, R. B.; Semb, J.; Cosulich, D. B.; Subbarao, Y. J.
Am. Chem. Soc. 1948, 70, 14–18.
6. Reif, V. D.; Reamer, J. T.; Grady, L. T. J. Pharm. Sci.
1977, 66, 1112–1116.
7. Mukherjee, T. Some Recent Studies of Molecular Dynam-
ics at BARC. In Atomic, Molecular and Cluster Physics;
Ahmad, S. A., Ed.; Narosa: New Delhi, 1997; pp 299–
316.
3. Gutteridge, J. M. C.; Bannister, J. V. Biochem. J. 1986, 234,
225–228.
4. (a) Joshi, R.; Adhikari, S.; Patro, B. S.; Chattopadhyay, S.;
Mukherjee, T. Free Rad. Biol. Med. 2001, 30, 1390–1399;
8. Buxton, G. V.; Stuart, C. R. J. Chem. Soc. Faraday Trans.
1995, 91, 279–281.