Photochemical behavior of folic acid in alkaline aqueous solutions and evolution of its photoproducts

Article abstract of DOI:10.1002/1522-2675(200208)85:8<2300::AID-HLCA2300>3.0.CO;2-B

The photolysis of folic acid (= N-(4-{[(2-amino-1,4-dihydro-4-oxopteridin-6-yl)methyl]amino}benzoyl)-L glutamic acid) in alkaline aqueous solution (pH 10.0-11.0) was carried out at 350 nm at room temperature and monitored by UV/VIS spectrophotometry, anal. and prep. thin-layer chromatography (TLC). and high-performance liquid chromatography (HPLC, HPLC/MS). The folate species underwent at least two independent photo-oxidation pathways, which were not observed when the acid form was photolyzed at pH < 7. The presence of O₂ was essential in these oxidation pathways. Evidence for the role of singlet oxygen was established. In one of the pathways, the folate underwent cleavage, yielding 6-formylpterin (=2-amino-1.4-dihydro-4-oxopteridine-6-carboxaldehyde) and (4-aminobenzoyl)glutamic acid as photoproducts. The other pathway yielded a new photostable product A of molecular mass 455, which could be isolated and stored in acidic or neutral aqueous solution. However, A was rather unstable in alkaline media undergoing a thermal reaction to a product B of lower molecular mass (427). The kinetics of this thermal reaction was analyzed with a stopped-flow spectrophotometer. A linear dependence of the first-order rate constant with the OH-concentration was observed. The corresponding bimolecular rate constant was 1.1 M^-1 S^-1. The quantum yields of substrate consumption and of photoproduct formation were determined. The here-reported photochemical behavior of folate solutions departs from results in acid media, where phototransformation proceeded via the cleavage of the acid form into 6-formylpterin and (4-aminobenzoyl)glutamic acid as the first major photoproducts, and where no thermal reactions were observed.

Full text of DOI:10.1002/1522-2675(200208)85:8<2300::AID-HLCA2300>3.0.CO;2-B

2
300
Helvetica Chimica Acta ± Vol. 85 (2002)
Photochemical Behavior of Folic Acid in Alkaline Aqueous Solutions and
Evolution of Its Photoproducts
a
a
a
a
by Andr e¬ s H. Thomas ), Gustavo Su a¬ rez ), Franco M. Cabrerizo ), Fernando S. GarcÌa Einschlag ),
b
c
c
a
Ra u¬ l Martino ), Claudio Baiocchi ), Edmondo Pramauro ), and Alberto L. Capparelli* )
a) Instituto de Investigaciones FisicoquÌmicas Te o¬ ricas y Aplicadas (INIFTA),
Universidad Nacional de la Plata, Casilla de Correo 16, Sucursal 4, 1900 La Plata
(
Tel/fax: 00542214242333; e-mail: alcappa@dalton.quimica.unlp.edu.ar)
b) Centro de Investigaci o¬ n y Desarrollo en Procesos CatalÌticos (CINDECA),
Universidad Nacional de La Plata, Casilla de Correo 16, Sucursal 4, 1900 La Plata, Argentina
c
)
Dipartimento di Chimica Analitica, Universit a¡ di Torino, Via P. Giuria 5, I-10125 Torino
The photolysis of folic acid (ˆ N-(4-{[(2-amino-1,4-dihydro-4-oxopteridin-6-yl)methyl]amino}benzoyl)-l-
glutamic acid) in alkaline aqueous solution (pH 10.0 ± 11.0) was carried out at 350 nm at room temperature and
monitored by UV/VIS spectrophotometry, anal. and prep. thin-layer chromatography (TLC), and high-
performance liquid chromatography (HPLC, HPLC/MS). The folate species underwent at least two
independent photo-oxidation pathways, which were not observed when the acid form was photolyzed at
2
pH < 7. The presence of O was essential in these oxidation pathways. Evidence for the role of singlet oxygen was
established. In one of the pathways, the folate underwent cleavage, yielding 6-formylpterin (ˆ2-amino-1,4-
dihydro-4-oxopteridine-6-carboxaldehyde) and (4-aminobenzoyl)glutamic acid as photoproducts. The other
pathway yielded a new photostable product A of molecular mass 455, which could be isolated and stored in
acidic or neutral aqueous solution. However, A was rather unstable in alkaline media undergoing a thermal
reaction to a product B of lower molecular mass (427). The kinetics of this thermal reaction was analyzed with a
À
stopped-flow spectrophotometer. A linear dependence of the first-order rate constant with the OH
concentration was observed. The corresponding bimolecular rate constant was 1.1 m
À1 À1
s . The quantum yields
of substrate consumption and of photoproduct formation were determined. The here-reported photochemical
behavior of folate solutions departs from results in acid media, where phototransformation proceeded via the
cleavage of the acid form into 6-formylpterin and (4-aminobenzoyl)glutamic acid as the first major
photoproducts, and where no thermal reactions were observed.
1. Introduction. ± Folic acid and related compounds play very important roles in
biological processes. This compound, also known as pteroylglutamic acid (ˆ N-(4-{[(2-
amino-1,4-dihydro-4-oxopteridin-6-yl)methyl]amino}benzoyl)-l-glutamic acid), can
be considered a conjugated pterin (ˆ2-aminopteridin-4-(1H)-one), and is a vitamin
that acts as a coenzyme in the metabolism of purine and pyrimidine bases [1]. Three
subunits can be recognized in the structure of folic acid [2][3]: 6-methylpterin, 4-
aminobenzoic acid, and glutamic acid (see Fig. 1). The folic acid derivative 5,10-
methenyltetrahydrofolate is present as the light-harvesting antenna in DNA photo-
lyases, which are involved in DNA-repair processes [4 ± 6] after UV-light irradiation.
The pteridine moiety present in folic acid behaves as a weak acid in aqueous
solutions. As reported by Albert [7] for several pterin derivatives, the dominant
equilibrium at pH > 5 involves an amide group (acid form) and a phenolate group (base
form). The pK of this equilibrium is ca. 8 for the different pterin derivatives studied
a
[
7 ± 10], including folic acid itself. Other functional groups of the pterin moiety (e.g., the

Helvetica Chimica Acta ± Vol. 85 (2002)
2301
Fig. 1. Molecular structure of folic acid in alkaline aqueous solutions (pH > 8)
2
-amino group or ring N-atoms) have pK values lower than 2 [7], and the pK of the
a
a
carboxylic acid functions of the glutamic acid moiety of folic acid are lower than 5 [11].
The acid and base forms of the pterin moiety present differences in their electronic
structures and charge densities that could also be reflected in its photochemical
behavior. In fact, as reported in previous communications [12 ± 18], the photochemistry
as well as the photophysical behavior of the different acid ± base forms of several pterin
derivatives show significant differences.
There are few reports in the literature that deal with the photochemistry of folic
acid. In 1949, Lowry et al. [17] employed indirect methods to identify products of
photolysis of folic acid and proposed the formation of (4-aminobenzoyl)glutamic acid
and 6-formylpterin (ˆ2-amino-1,4-dihydro-4-oxopteridine-6-carboxaldehyde) as pho-
toproducts. No information about the reaction pathway, the effect of O , or the
2
quantum efficiency was reported. In a more-recent communication [10], we inves-
tigated the photochemistry of the acid form of this vitamin in aqueous solutions
(
pH 4.5 ± 6.0). The photolysis quantum yields were reported, the photoproducts were
identified (which were coincident with those proposed by Lowry et al. [17]), and some
hints on its kinetic and mechanistic aspects were suggested. Indirect evidence for the
role of singlet oxygen as a triggering agent for the photooxidation was shown, i.e., the
photodegradation of folic acid was inhibited by singlet-oxygen scavengers. The
influence of 6-formylpterin, one of the photoproducts detected, on the rate of
photodegradation was also described, i.e., photodegradation was enhanced either when
the concentration of 6-formylpterin increased during the photolysis or when this
photoproduct was present in the initial solutions of folic acid. Therefore, excitation of
6-formylpterin led to more-efficient generation of singlet oxygen that induced the
degradation of folic acid.
In the context of our investigations on photochemistry and photophysics of pterin
derivatives [10][13][15][16][19] as well as potential biological implications [20], we
now report the photolysis of folic acid in alkaline aqueous solutions, where only the
base form of the molecule, the folate, is present (Fig. 1). Photoproducts were
characterized, and the quantum efficiencies of the photodegradation of folate and of
the formation of photoproducts as well as the effects of O on these photochemical
2
reactions were determined. The results were compared with those obtained for the
photochemistry of folic acid in acid aqueous solution [10].

2
302
Helvetica Chimica Acta ± Vol. 85 (2002)
2. Experimental. ± Materials. Folic acid (Merck) and the other pterin derivatives (Schircks Laboratories)
2
were used without further purification. Solns. of folic acid stored in the absence of O were stable (HPLCor
TLC). Other chemicals from Sigma-Aldrich were used as received. The pH of aq. solns. was adjusted to 10.0 ±
À3
1
1.0 by adding drops of conc. HCl or NaOH solns. with a micropipette. The ionic strength was ca. 10 m in all the
experiments.
Photolyses. The continuous photolysis of aq. folate solns. was carried out in the presence or in the absence of
air. Deaerated solns. were obtained by bubbling with O -free N for 20 min. Rayonet RPR-3500-ä lamps
Southern N.E. Ultraviolet Co.) were employed for irradiating. Photolysis products were identified or
2
2
(
characterized by analyzing the absorbance changes during irradiation, by TLC, and by high performance liquid
chromatography (HPLC).
UV/VIS Spectroscopy. UV/VIS Absorption spectra were collected with a Cary 3 (Varian) spectropho-
tometer and refined with a program for smoothing and averaging signals. Quartz cells of 1-cm optical path
length and double cells of 1 cm (for irradiation) and 0.2 cm (for absorbance measurements) were used.
Absorption spectra of the solns. were recorded at regular intervals of irradiation time. From the absorbance
changes recorded under these conditions, experimental difference (ED) and normalized experimental
difference (NED) spectra were obtained. Reference difference (RD) and normalized reference difference
(
NRD) spectra were obtained from aq. solns. of commercial standards. The analysis based on these difference
spectra is described elsewhere [10][13]. When necessary, further details are described in the Figure legends.
TLC Analyses. For anal. TLC, 100-mm DEAE-cellulose plates and aq. NH Cl solns. (0.3%(w/v) and 3%(w/
v)) as eluent were employed. Irradiated folate solns. (450 mm) and standards of pterin derivatives were detected
with 350-nm light. Under these conditions, most pterin derivatives show blue fluorescent spots. R Values
obtained for several standard solns. under these conditions were reported elsewhere [10]. In some experiments,
50-mm cellulose plates with fluorescent indicator were employed to detect compounds that cannot be revealed
4
f
2
by 350-nm light (e.g. (4-aminobenzoyl)glutamic and folic acid).
To separate some photolysis products, prep. TLCwith 250- mm DEAE-cellulose plates were performed. To
increase the amount of product, irradiated solns. of folate were concentrated at low pressure before carrying out
the chromatographic runs. The same eluents as for anal. TLCwere used. Products were extracted from the TLC
plates with aq. solns. (0.5m HCl or 3%(w/v) NH Cl soln.). The solns. thus obtained were centrifuged and filtered
4
before analysis.
HPLC Analyses. For monitoring the reaction and identifying photoproducts, a Konic-Instruments 500-B
chromatograph, a C18 Spherisorb-S5-ODS2 (250 Â 46 mm) and a Lambda-1000 Bischoff HPLC spectropho-
tometer as detection system column were employed. The eluent flow rate was 1.0 ml/min. Eluent 1: 7% MeCN/
2
H O 6 :4 (soln. a), 93% aq. soln. containing 20 mm potassium phosphate (pH 5.5) and 2.5 mm EDTA (soln. b).
Eluent 2: 5% MeOH, 95% aq. soln. containing 50 mm NaOAc and 5 mm citric acid (pH 5.2). Eluent 3: 10%
soln. a, 90% soln. b. Eluents 1 and 3 were employed for monitoring folic acid and other pterin derivatives at
3
40 nm, whereas eluent 2 was used for identification and quantification of (4-aminobenzoyl)glutamic acid at
2
80 nm.
QuantumYields . The quantum yields for folate disappearance and 6-formylpterin formation were
determined in experiments performed under different conditions. A chemical actinometer ([CoIIIBr(NH
)
]
2‡
)
3
5
was employed for the measurements of photon rates [21]. The initial concentration of reactant was adjusted so
that more than 99.99% of the incident light at 350 nm was absorbed. The evolution of the concentrations of
reactant and photoproducts during the irradiation time was followed by HPLC. Aq. solns. of commercial
standards were employed for obtaining the corresponding calibration curves.
Oxygen Consumption. The O
electrode (Orion, model 37-08-99). The experimental set-up for these measurements was described elsewhere
22].
Experiments in the Presence of Furfuryl Alcohol. Photolysis in the presence of 10 mm furfuryl alcohol
ˆ furan-2-methanol; Riedel-de Ha e» n) was carried out. This compound exhibits high reactivity towards singlet
2 2
consumption during the photolysis was measured with an O -selective
[
(
oxygen [23][24] reducing greatly its concentration when formed during the photolysis. In this group of
experiments, folate and 6-formylpterin concentrations as a function of irradiation time were determined by
HPLC. Results were compared with those obtained in the absence of the selective singlet-oxygen scavenger.
HPLC/MS Analyses. A ThermoFinnigan-AS300 auto sampler connected to a pump ThermoFinnigan,
model TP 4000, interfaced with a LCQ-ThermoFinnigan ion-trap mass spectrometer equipped with an ESI ion
source was employed for MS and MS/MS in the negative-ion mode. Separations were obtained under gradient-
elution conditions with a RP-18-C18 column (Lichrospher RP-18, 250 Â 4.6 mm, 5 mm particles (Merck,
4
Darmstadt, Germany)) and a two-component mobile phase. Phase A: MeCN, phase B: aq. 0.05 mm NH OAc

Helvetica Chimica Acta ± Vol. 85 (2002)
2303
was
(
pH 6.8); flow rate 0.8 ml/min. Gradient conditions: A/B 10 :90 at 0 min; 100% A at 30 min. High-purity N
2
used as nebulizer and auxiliary gas. The heated capillary was set at 2008 and the ion-optics parameters were
optimized to the following values: spray voltage 4.5 kV, capillary voltage À8.0 V, tube-lens voltage 5.0 V, first-
octapole voltage 2.25 V, inter-octapole-lens voltage 20.0 V, and second-octapole voltage 7.0 V.
Kinetic Studies. As a consequence of thermal chemical changes observed in irradiated solns. of folate,
kinetic studies were carried out with a SFA-20 stopped-flow accessory (Hi-Tech) and a Cary-3 (Varian)
spectrophotometer. Kinetic experiments were performed by mixing irradiated folate solns. or solns. of the
2 4 3 4
photolysis products, separated previously by prep. TLC, with buffer solns. of K HPO /K PO (0.025m) in the pH
range 8 ± 12. The pH was measured after mixing. Absorbance changes were recorded between 250 and 400 nm as
a function of time, and rate constants were calculated as the average of at least 10 values obtained at different
wavelengths.
3
. Results and Discussion. ± 3.1. Preamble. The behavior of the folate solutions
under continuous irradiation in absence of O is identical to that reported in the
2
photolysis of the acid form of this vitamin, i.e., folate does not undergo any chemical
modification under irradiation at 350 nm in the absence of O2.
3
.2. Evolution of the Spectrumof Folate Solutions . As shown in Fig. 2,a, the
spectrum of folate solutions changes significantly with irradiation time in the presence
of O . However, most of the reported changes observed in the present study differ
2
considerably from those observed in acid medium, where only the acid form of the
pterin moiety is present [10]. The main differences are: i) In acid medium, NED and
NRD spectra are almost indistinguishable. However, in alkaline medium, NED and
NRD spectra are very different (Fig. 2,b). These results strongly suggest that, under
these irradiation conditions, the folate anion does not yield the same photoproducts as
found in acid media (6-formylpterin and (4-aminobenzoyl)glutamic acid), or that these
compounds are not the only photoproducts.
ii) When the irradiated solutions of folate (pH 10 ± 11) are stored in the dark,
further chemical changes can be observed, as shown in Fig. 3. This reaction can be
followed for a few hours after interrupting the irradiation. These results suggest that
other photoproducts different from 6-formylpterin and (4-aminobenzoyl)glutamic acid
are also present. These new photoproducts should be responsible for the observed
thermal reactions. In addition, a strong dependence in the rate of spectral changes on
the pH was detected; i.e., the rate of spectral changes increases with the pH (see below,
Sect. 3.5).
iii) Experiments were designed to investigate the role of O in the thermal reaction.
2
An air-saturated folate solution (450 mm) at pH 10.5 was irradiated during 50 min, then
deareated by bubbling with N , and stored in the dark. The absorption spectra of the
2
solution recorded at regular intervals of time after bubbling showed the same changes
as observed in the presence of air, implying that O is not involved in the observed
2
thermal reaction.
3
.3. Characterization of the Photoproducts. TLCAnalyses showed the presence of
four fluorescent products after irradiation. Two of the observed R values correspond to
f
those of standard 6-carboxypterin (ˆ2-amino-1,4-dihydro-4-oxopteridine-6-carboxylic
acid) and 6-formylpterin, while the other two substances, A and B, could not be
identified with any of the available standard pterin derivatives. As expected from
previous studies on the photochemistry of pteridine compounds [19], 6-carboxypterin
should arise in a later stage from the photolysis of 6-formylpterin. TLCRuns
performed with plates containing fluorescent indicator showed the presence of a non-

2
304
Helvetica Chimica Acta ± Vol. 85 (2002)
Fig. 2. a) Evolution of the absorption spectrumof an irradiated solution of folate (450 mm) at pH 10.0 as a
function of time (air-saturated solutions, spectra recorded every 10 min; arrows indicate the observed changes).
b) NED Spectrum (–) obtained by subtracting the initial folate spectrumfromthe corresponding spectrumafter
4
0 min of photolysis, and NRD spectrum (¥¥¥¥¥) obtained by subtracting the spectrumof a standard solution of
folate from the spectrum of a standard solution of 6-formylpterin and (4-aminobenzoyl)glutamic acid, at the same
concentration and pH.
fluorescent product having a R value that corresponds to that of standard (4-
f
aminobenzoyl)glutamic acid. Folate solutions, irradiated and then stored for several
hours in the dark, contained no significant amounts of A, but increased amounts of B
(
by TLC). These results strongly suggest that the above-mentioned thermal reaction is
the chemical transformation A ! B.
HPLCAnalyses revealed the presence of six substances in irradiated solutions of
folate. Four of them were identified by comparing their retention times (t ) with those
R
of standard solutions, i.e., 6-formylpterin, 6-carboxypterin, and folic acid (HPLC
eluent 1), as well as (4-aminobenzoyl)glutamic acid (HPLCeluent 2). The other two
substances could not be identified by comparison with the available standard solutions,
which is in good agreement with the TLCresults.
Solutions of substances A and B were obtained by prep. TLC(3% ( w/v) NH Cl
4
solution) from irradiated (50 ± 60 min) solutions of folic acid. Substances A (R 0.63)
f

Helvetica Chimica Acta ± Vol. 85 (2002)
2305
Fig. 3. Evolution in the dark of the absorption spectrumof a folate solution (450 mm) at pH 11, irradiated for
0 min. Experiment performed in the presence of O . After stopping the irradiation, the spectra were recorded
at the following times: 0, 12, 30, 60, 105, and 150 min. Arrows indicate the observed changes.
5
2
and B (R ca. 0.02) were extracted from the plates with aqueous solutions (3%(w/v)
f
NH Cl solution and 0.5m HCl, resp.), and normalized spectra of A and B in alkaline
4
media were recorded (Fig. 4,a).
Substance A is stable in the absence of light at the pH of the NH Cl solution
4
(
pH 5.5), e.g., no changes in the spectrum were detected after storing the solution for
several days in the dark at room temperature. A similar spectral analysis of solutions of
substance B revealed that this compound is stable in the dark in both acid and alkaline
media.
However, substance A is not stable in the dark at pH > 8 since important changes in
the spectrum were observed under these pH conditions. The rates of these changes
showed strong dependence on the pH, being faster at higher pH (see Sect. 3.5).
Alkaline solutions of A stored in the dark until completion of the thermal trans-
formation furnished a product with a spectrum similar to that of substance B obtained
by TLC. Moreover, also on TLC, the R value of this product was identical to that of B,
f
and the NED spectrum obtained in this case (see Fig. 4,b) was similar to that obtained
with irradiated solutions of folate (e.g., experiment shown in Fig. 3). These results
confirmed that A is transformed into B in alkaline media, and that this reaction is
responsible for the spectral changes observed in irradiated solutions of folate after
interrupting the irradiation.
Solutions of substances A and B were irradiated in alkaline media. No changes in
the spectrum of B were detected after 2 h of irradiation. The changes in the spectrum of
A were the same, irrespective of the irradiation. These results suggest that neither
substance is photodegraded by UV-A light.

2
306
Helvetica Chimica Acta ± Vol. 85 (2002)
Fig. 4. a) Normalized spectra of substances A and B in aqueous alkaline media. (solutions obtained from prep.
TLC). b) NED Spectrum (–) obtained by subtracting the initial spectrumof an alkaline solution of A (recorded
immediately after alkalinizing a solution obtained by prep. TLC) fromthe corresponding spectrumof the same
solution, stored in the dark for several hours (after completing the thermal transformation) and NED spectrum
(
¥¥¥ ) obtained by subtracting the initial spectrumof an irradiated solution of folate (450 mm) at pH 11 (recorded
immediately after interrupting the irradiation) fromthe corresponding spectrumof the same solution stored in
the dark for several hours.
Solutions of both products A and B and of folic acid were analyzed by HPLC/MS
furnishing the molecular masses M 455 (m/z 454), 427 (m/z 426), and 441 (m/z 440),
r
respectively. Oxidation of folic acid without cleavage could explain the increase in the
molecular mass of A, which could be an epoxy derivative of folic acid. Substance A
could suffer then a rearrangement with loss of the CO group of the pterin moiety ( !
B). A more-detailed characterization of the chemical structure of substances A and B
is the subject of future studies.
The MS/MS of A and B (Fig. 5) revealed that A loses H O (m/z 436) and, more
2
importantly, a fragment of mass 28 (m/z 426) with a successive fragmentation pattern
identical to that of B. There is agreement between this mass fragmentation and the
chemical degradation pathway. Moreover, fragments at m/z 408 and 382 are formed,
corresponding to losses of H O and CO from m/z 426, the latter probably arising from
2
2
one of the carboxylic groups of glutamic acid.
.4. Analysis of Concentration Profiles. Foliate, 6-formylpterin, and 6-carboxypterin
3
concentrations were measured as a function of irradiation time by means of two sets of
HPLCanalyses (eluent 1): first, concentrated solutions of folate (> 500 mm) were

Helvetica Chimica Acta ± Vol. 85 (2002)
2307
Fig. 5. MS/MS of a) substance A (m/z 454) and b) substance B (m/z 426). The corresponding solutions were
obtained from prep. TLC.

2
308
Helvetica Chimica Acta ± Vol. 85 (2002)
irradiated and analyzed, and, second, folate/6-formylpterin mixtures were irradiated
and analyzed.
The first group of analyses showed that, in the first 20 min of irradiation, the
evolution of the concentrations of folate and 6-formylpterin follow a zero-order rate
law (Fig. 6). In this first stage, no significant increase of the 6-carboxypterin
concentration was observed, and it can be assumed that the whole incident radiation
was absorbed by the folate. The quantum yield associated with the disappearance of
À3
À3
folate was 5.1 ¥ 10 , and that of 6-formylpterin formation was 2.9 ¥ 10 . Therefore,
consumption of folate to yield 6-formylpterin was not the only process taking place in
the time range analyzed. After 20 min of irradiation, significant increases in the rate of
photodegradation of folate and in the rate of formation of 6-formylpterin were
observed (Fig. 7), coincident with a temporary increase in 6-carboxypterin concen-
tration. As mentioned above, 6-carboxypterin is a photoproduct generated from
6-formylpterin [19]. Therefore, after ca. 20 ± 30 min, a significant proportion of
incident radiation was absorbed by 6-formylpterin, and the chemical transformation
from folate to 6-formylpterin seems to be enhanced by the participation of excited 6-
formylpterin. As can be seen in Fig. 7, during the first 60 min of irradiation under these
conditions, the sum of folic acid, 6-formylpterin, and 6-carboxypterin concentrations
was not constant, thus indicating that there is another pathway for folate photo-
degradation.
Fig. 6. Changes in folate and 6-formylpterin concentrations in irradiated solution of folate (660 mm) at pH 11.0 as
a function of time. HPLCanalysis of the first 20 min of photolysis.

Helvetica Chimica Acta ± Vol. 85 (2002)
2309
Fig. 7. Changes in folate, 6-formylpterin, and 6-carboxypterin concentrations, determined by HPLC, in irradiated
solution of 560 mm folate at pH 11.0 as a function of time. The algebraic sum of folate, 6-formylpterin, and 6-
carboxypterin concentrations, for each time, is also shown (^).
The second group of analyses were performed in such a way that only a low
proportion of the incident radiation was absorbed by folate, whereas a higher
proportion was absorbed by 6-formylpterin, i.e., a mixture of folate (65 mm) and 6-
formylpterin (150 mm) was irradiated (Fig. 8). An initial rate of consumption of 0.5 mm/
min was estimated under these conditions, taking the rate of photons absorbed by folate
as 23% of the incident photonic rate and the quantum yield of folate photodegradation
À3
reported above (5.1 ¥ 10 ). However, a rate of 5.9 mm/min was obtained for the first
10 min of photolysis. These results confirm that absorption of radiation by 6-
formylpterin induces the photodegradation of folate.
The concentration of (4-aminobenzoyl)glutamic acid was also measured by HPLC
(eluent 2) as a function of irradiation time. As expected, during the first 20 min of
irradiation, the concentration follows a zero-order rate law, and, after 20 min, an
increase of the rate of formation of (4-aminobenzoyl)glutamic acid was observed.
The relative concentration profiles of the two unidentified products A and B on
irradiating solutions of folate (450 mm at pH 11.0) were determined by HPLC(eluent 3)
analyzing at a wavelength of 340 nm (t 3.2 (A) and 6.7 min (B)). The concentrations
R
of both substances increased during the first 60 min of irradiation. Changes in the
concentrations of A and B and folate were obtained as a function of time, after
interrupting the irradiation (50 min) of a solution of folate (450 mm) at pH 11.0 (Fig. 9).
In the analyzed time range, the concentration of A decreased while the concentration

2
310
Helvetica Chimica Acta ± Vol. 85 (2002)
Fig. 8. Changes in folate, 6-formylpterin, and 6-carboxypterin concentrations, determined by HPLC, in a solution
of folate (65 mm) and 6-formylpterin (150 mm) at pH 11.0 as a function of the irradiation time
of B increased, and no variation in folate concentration was observed. Other similar
HPLCanalyses (eluents 1 and 2) showed that the concentrations of 6-formylpterin, 6-
carboxypterin, and (4-aminobenzoyl)glutamic acid did not change after interrupting
the irradiation. These results indicate that these substances, as well as folate, are not
involved in the thermal reaction that occurs in irradiated solutions of folate.
3
.5. The Role of Oxygen and Kinetics. A significant decrease in O concentration
2
during the photolysis was observed, indicating that photo-oxidation of folate is taking
place. As shown in Fig. 10, after ca. 80 min of irradiation, an increase in the O2
consumption rate was observed. The degradation of folate induced by 6-formylpterin or
another product and the photo-oxidation of 6-formylpterin itself [19] may explain this
observation.
In a previous work [10], we discussed the importance of singlet oxygen in the
photochemistry of the acid form of folic acid. This reactive form of O is formed in the
2
presence of pterin derivatives, as previously reported [15][25]. Photolysis in the
presence of furfuryl alcohol, a selective scavenger [23][24], was carried out to
investigate whether singlet oxygen is involved in the mechanism of photodegradation
of folate. The results obtained (not shown) indicate that both photodegradation
1
pathways are strongly inhibited by furfuryl alcohol, suggesting that O is involved as
2
the active intermediate in photolysis of folate.
Kinetic experiments were performed by mixing irradiated solutions of folate (in
most experiments, pH 9.5 ± 10.0) or solutions of substance A (obtained by prep. TLC)

Helvetica Chimica Acta ± Vol. 85 (2002)
2311
Fig. 9. Changes in normalized concentrations of folate and substances A (t
R
3.2 min) and B (t
R
8.4 min) as a
function of time, after interrupting the irradiation. The folate solution (450 mm) at pH 11.0 was irradiated for
0 min and then analyzed by HPLC.
5
with buffer solutions of different pH values. Similar results were obtained in both
groups of experiments. In all cases, only one process was observed in the dark. As
shown in Fig. 11, this process follows a first-order rate law. In all experiments, the
apparent rate constant (kapp) was found to be independent of wavelength. On the other
À
hand, kapp showed a strong dependence on the pH, depending linearly on the OH
concentration (Fig. 12).
Taking these results into account, the following rate law can be written:
À
À d[A]/dt ˆ k ¥ [OH ] ¥ [A], where A is the reactant of the thermal reaction and k is
À1 À1
the second-order rate constant. A value of 1.1 m
of Fig. 12.
s
was obtained for k from the data
4
. Conclusions. ± The results obtained are summarized in the Scheme. Some aspects
of the photochemistry of the base form of folic acid, i.e., of folate, are similar to those
observed for its acid form, i.e., in the absence of O , folate is photostable, whereas, in
2
the presence of O , folate undergoes cleavage, yielding 6-formylpterin and (4-
2
aminobenzoyl)glutamate as photoproducts [10]. However, the most important differ-
ences are associated with the presence of a new reaction pathway, leading to at least one
new compound A having an M (455) higher than that of folic acid (441). This increase
r
in the molecular mass may be due to the incorporation of an O-atom into the folate
anion and the loss of two H-atoms. This reaction does not involve cleavage of the folate

2
312
Helvetica Chimica Acta ± Vol. 85 (2002)
Fig. 10. Changes in the O
2
concentration in an irradiated solution of folate (450 mm) at pH 11.0 as a function of
time
anion. Further studies aimed at complete identification of the structures of substances
A and B are in progress. The folate behavior shows that this kind of molecule presents
different reactivities in the base and acid forms as a consequence of the different charge
distribution in the two forms.
Substance A is photostable, but, in alkaline media, it undergoes a dark thermal
reaction yielding product B, which has lower molecular mass (427), whereby O is not
2
involved. The difference of M of A and B suggests a complex mechanism that may
r
involve an internal rearrangement with loss of CO. This process follows a first-order
rate law and the corresponding apparent rate constant (kapp) shows a linear dependence
À
on the OH concentration.
In both acidic and basic media, singlet oxygen, very likely produced by energy-
transfer from the electronically excited states of folic acid, is the oxidizing agent that
starts the degradation reactions.
The presence of 6-formylpterin, either photochemically generated or added into
the folic acid solution before irradiation, accelerates both photo-oxidation pathways.
These processes require photoexcitation of 6-formylpterin, do not require the
absorption of light by the folate anion, and are inhibited by a singlet oxygen scavenger.
Therefore, excitation of 6-formylpterin leads to more efficient generation of singlet
oxygen that induces the photolysis of folic acid by means of the two reactions described.

Helvetica Chimica Acta ± Vol. 85 (2002)
2313
Fig. 11. Experiment performed by mixing a folate solution (pH 10.0) irradiated for 50 min with phosphate buffer
(
pH 10.5 after mixing): a) Absorbance changes recorded at 374 nmas a function of time and b) first-order
behavior of difference absorbance at 374 nm
À
Fig. 12. Linear dependence of kapp as a function of OH concentration

2
314
Helvetica Chimica Acta ± Vol. 85 (2002)
Scheme 1
This work was supported in part by a project grant from the Agencia de Promoci o¬ n CientÌfica y Tecnol o¬ gica
de Argentina (ANPCyT Grant N8 PICT 06-03531), Consejo Nacional de Investigaciones Cientificas y
Tecnol o¬ gicas (CONICET-PIP 4354), and Universidad Nacional de La Plata. A. H. T. wishes to thank the
Consejo Nacional de Investigaciones Cientificas y Tecnol o¬ gicas de Argentina (CONICET) for a research
graduate grant. C. B. and E. P. wish to thank MURST (Rome).
REFERENCES
[
[
1] −The Biochemistry of Folic Acid and Related Pteridines×, North-Holland Publishing Co., Amsterdam, 1969.
2] R. B. Angier, J. H. Booth, B. L. Hutchings, J. H. Mowat, J. Semb, E. L. R. Stokstad, Y. SubbaRow, C. W.
Waller, E. H. Northey, D. R. Seeger, J. P. Sickel, J. M. Smith, Science (Washington, D.C.) 1946, 103, 667.
3] J. J. Pfiffner, D. G. Calkins, E. S. Bloom, B. L. O×Dell, J. Am. Chem. Soc. 1946, 68, 1392.
4] A. Sancar, G. B. Sancar, Annu. Rev. Biochem. 1988, 57, 29.
[
[
[
5] J. L. Johnson, S. Hamm-Alvarez, G. Payne, G. B. Sancar, K. V. Rajagopalan, A. Sancar, Proc. Natl. Acad.
Sci. U.S.A. 1988, 85, 2046.
[
[
[
[
6] J. E. Hearst, Science (Washington, D.C.) 1995, 268, 1858.
7] A. Albert, Biochem. J. 1953, 54, 646.
8] V. D. Mon o¬ poli, A. H. Thomas, A. L. Capparelli, Int. J. Chem. Kinet. 2000, 32, 231.
9] A. H. Thomas, M. R. F e¬ liz, A. L. Capparelli, Transition Met. Chem. 1996, 21, 317.
[
10] A. H. Thomas, G. Su a¬ rez, F. M. Cabrerizo, R. Martino, A. L. Capparelli, J. Photochem. Photobiol. A:
Chem. 2000, 135, 147.
[
[
11] A. Streitwieser, C. H. Heathcock, −QuÌmica Org a¬ nica×, McGraw-Hill, M e¬ xico, 1989.
12] W. Pfleiderer, M. Kappel, R. Baur, in −Biochemical and Clinical Aspects of Pteridines×, Eds. W. Pfleiderer,
H. Wachter, and H. C. Curtius, Walter de Gruyter & Co., Vol. 3, Berlin-New York, 1984.
13] G. Su a¬ rez, F. M. Cabrerizo, C. Lorente, A. H. Thomas, A. L. Capparelli, J. Photochem. Photobiol. A:
Chem. 2000, 132, 53.
[
[
[
14] C. Chahidi, M. Aubailly, A. Momzikoff, M. Bazin, R. Santus, Photochem. Photobiol. 1981, 33, 641.
15] A. H. Thomas, C. Lorente, L. S. Villata, A. L. Capparelli, M. Mesaros, G. M. Bilmes, C. G. MartÌnez, M. R.
Pockhrel, A. M. Braun, E. Oliveros, −XVIII IUPACSymposium on Photochemistry×, Dresden, Germany,
2000, p. 388.
[16] A. H. Thomas, C. Lorente, A. L. Capparelli, M. R. Pockrel, E. Oliveros, A. M. Braun, −XII Inter-American
Photochemical Society Conference×, Ascochinga, Argentina, 2001, p. 55.
[
[
17] O. H. Lowry, O. A. Bessey, E. J. Crawford, J. Biol. Chem. 1949, 180, 389.
18] R. Baur, M. Kappel, R. Mengel, W. Pfleiderer, in −Chemistry and Biology of Pteridines×, Eds. R. L. Kisliuk
and G. M. Brown, Elsevier/North Holland, New York, 1979.
[
[
19] A. H. Thomas, G. Su a¬ rez, F. M. Cabrerizo, A. L. Capparelli, Helv. Chim. Acta 2001, 84, 3849.
20] C. Lorente, A. H. Thomas, L. S. Villata, D. Hozbor, A. Lagares, A. L. Capparelli, Pteridines 2000, 11, 100;
see also K. Ito, S. Kawanishi, Biochemistry 1997, 36, 1774.

Helvetica Chimica Acta ± Vol. 85 (2002)
2315
[
[
[
21] J. F. Endicott, G. Ferraudi, J. R. Barber, J. Phys. Chem. 1975, 79, 630.
22] F. S. GarcÌa Einschlag, M. R. F e¬ liz, L. Capparelli, J. Photochem. Photobiol. A: Chem. 1997, 110, 235.
23] R. Atkinson, in −Active Oxygen in Chemistry×, Eds. C. S. Foote, J. S. Valentine, A. Greenberg, and J. F.
Liebman, Blackie Academic & Professional (Chapman & Hall Imprint), London, 1995, Vol. 2, p. 289.
24] W. R. Haag, J. Hoign e¬ , E. Gassman, A. M. Braun, Chemosphere 1984, 13, 631.
[
[
25] K. V. Neverov, E. A. Mironov, T. A. Lyudnikova, A. A. Krasnovsky, M. S. Kritsky, Biokhim. 1996, 61, 1627.
Received February 20, 2002