Photobleaching of Compounds of the 5,10,15,20-Tetrakis(m-hydroxyphenyl)-porphyrin Series (m-THPP, m-THPC, and m-THPBC)

Article abstract of DOI:10.1021/ol025842c

(Matrix Presented) 5,10,15,20-Tetrakis(m-hydroxyphenyl)porphyrin (m-THPP) yielded novel quinonoid porphyrins upon irradiation in aqueous methanol. True photobleaching was observed for 5,10,15,20-tetrakis(m-hydroxyphenyl)chlorin (m-THPC) and 5,10,15,20-tet

Full text of DOI:10.1021/ol025842c

ORGANIC
LETTERS
2002
Vol. 4, No. 12
2013-2016
Photobleaching of Compounds of the
5,10,15,20-Tetrakis(m-hydroxyphenyl)-
porphyrin Series (m-THPP, m-THPC, and
m-THPBC)
Raymond Bonnett* and Gabriel Mart´ınez
Department of Chemistry, Queen Mary College, UniVersity of London, Mile End Road,
London E1 4NS, UK
r.bonnett@qmul.ac.uk
Received March 8, 2002
ABSTRACT
5,10,15,20-Tetrakis(m-hydroxyphenyl)porphyrin (m-THPP) yielded novel quinonoid porphyrins upon irradiation in aqueous methanol. True
photobleaching was observed for 5,10,15,20-tetrakis(m-hydroxyphenyl)chlorin (m-THPC) and 5,10,15,20-tetrakis(m-hydroxyphenyl)bacteriochlorin
(m-THPBC) under the same conditions; several fragmentation products (imides, methyl p-hydroxybenzoate, dipyrrin derivatives) were recognized.
Photodynamic therapy (PDT) is emerging as a treatment for
a variety of conditions, but particularly for cancer and wet
age-related macular degeneration. It requires a combination
of oxygen, visible light, and a drug (a photosensitizer), which
causes damage to living tissue.1 The 5,10,15,20-tetrakis(m-
hydroxyphenyl)porphyrin series2 (m-THPP (1), m-THPC (2),
and m-THPBC (3)) includes some of the most potent
photosensitizers discovered to date. The chlorin, m-THPC
(2) (FOSCAN, temoporfin), has recently received regulatory
approval in the EU for the treatment of head and neck
cancer.3 In photochemistry and photobiology, photobleaching
is defined as the loss of absorption or emission intensity4
caused by light. Photobleaching of porphyrin analogues most
commonly used in PDT can occur by photooxidation or
photoreduction processes, the former being more common.
Two types of irreversible photobleaching process leading to
chemical change of the chromophore are encountered:5 (i)
photomodification, where the chromophore is retained in a
modified form; (ii) true photobleaching, where chemical
changes are deep seated and result in small fragments that
no longer have appreciable absorption in the visible region.
Photobleaching of photosensitizers during PDT has been
observed both in cells in Vitro and in skin in ViVo using the
fluorescence mode of observation, both directly and after
pigment extraction.6 While photobleaching might be thought
to be a disadvantageous property in a tumor photosensitizer
(in that the source of the active species is being destroyed),
(3) Ernst & Young, press release, 28 June 2001.
(1) Bonnett, R. Chemical Aspects of Photodynamic Therapy; Gordon and
Breach Science: Amsterdam, 2000.
(2) (a) Berenbaum, M. C.; Akande, S. L.; Bonnett, R.; Kaur, H.; Ioannou,
S.; White, R. D.; Winfield, U. J. Br. J. Cancer 1986, 54, 717. (b) Bonnett,
R.; White, R. D.; Winfield, U. J.; Berenbaum, M. C. Biochem. J. 1989,
261, 277.
(4) Verhoeven, J. W. Pure Appl. Chem. 1996, 68, 2223.
(5) Bonnett, R.; Mart´ınez, G. Tetrahedron 2001, 57, 9513.
(6) (a) Moan, J.; Christiansen, T.; Jacobsen, P. B. Porphyrin Localization
and Treatment of Tumours; Dorion, D. R., Gomer, C. J., Liss, A. R., Eds.;
New York, 1984; p 419. (b) Mang, T. S.; Dougherty, T. J.; Potter, W. R.;
Boyle, D. G.; Somer, S.; Moan, J. Photochem. Photobiol. 1987, 45, 501.
10.1021/ol025842c CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/23/2002

it is possible to envisage potential benefits, for example, in
limiting the damage to healthy tissue by working at the
appropriate dose of photosensitizer.5,6 In recent years, photo-
bleaching has received increasing attention. However, most
of the reports to date are kinetic studies and often little is
known about the structures of the photoproducts.5
Scheme 1. Photobleaching of m-THPP (1)
The first comparative kinetic study7 of the photobleaching
of the m-THPP series in Vitro using laser irradiation was
recently reported. True photobleaching was observed for
m-THPC (2) and m-THPBC (3), but the porphyrin (1) reacted
more slowly, with photomodification. Evidence was found
for the involvement of singlet oxygen in all three systems.
We now wish to report our preparative studies using broad
band UV-vis radiation.
m-THPP (1) was irradiated (two Philips MLU 300 W
lamps) for 24 h in methanol-water (60:40) in a water-cooled
photoreactor.8 The absorbance decay at band IV (λmax ) 512
nm) was 20%, band III (λmax ) 546 nm) intensity decreased
22%, and the absorbance of band II (λmax ) 587 nm) decayed
8%. The methanol of the irradiated solution was then
removed, and the resulting aqueous suspension was extracted
with ethyl acetate, leaving a colorless water layer. Prelimi-
nary analysis of the organic layer by TLC on silica showed
the presence of four major components, m-THPP (1), two
brown bands less polar than 1, and a purple base line. The
least polar brown band was found to contain four minor
components separated by preparative TLC. The second brown
band was isolated in larger amount (25%) as puce solid 4
after flash chromatography; it was followed by m-THPP (1,
32% recovery). The porphyrin nature of the five photopro-
ducts was immediately apparent from the UV-vis spectra.
The measured accurate mass and elemental analyses for
the major photoproduct 4 were consistent with the molecular
formula C44H28N4O5 (M + H 693.2084, calcd 693.2138), a
compound containing an extra oxygen when compared with
m-THPP (1). The UV-vis spectrum of 4 showed bands at
413, 510, 547, 588, and 645 nm. Qualitatively, 4 showed
little fluorescence on the TLC plate (366 nm), in contrast to
the bright red fluorescence observed for m-THPP (1). This
difference in fluorescence was confirmed when the emission
spectra (λexc ) 423 nm) of m-THPP (1) and 4 were recorded
at nearly the same concentration (c ) 1.55 × 10-6 and 1.7
× 10-6 M, respectively). The fluorescence of m-THPP (1)
at 648 nm was three times more intense than that of the
photoproduct. It has been shown that meso-benzoquinonylpor-
phyrins have their fluorescence diminished considerably,9 so
4 was tentatively regarded as a benzoquinonylporphyrin. The
quinonoid nature of this porphyrin was confirmed after
1
rapidly ruled out on the basis of H NMR evidence. To
distinguish between the 3,4-o-benzoquinone and the 2,5-p-
benzoquinone is not simple on the basis of spectroscopic
evidence alone. We then embarked in the synthesis of a series
of 5-(dihydroxyphenyl)-10,15,20-tris(3-hydroxyphenyl)por-
phyrins 11-14 (Scheme 2).10
With these porphyrins in hand, we proceeded to reduce
the photoproduct. When 4 was treated with NaBH4 (MeOH,
2.5 h), the starting material disappeared and a new material
was formed. On the TLC plate, the new porphyrin was more
polar than the photoproduct and m-THPP (1) and it showed
a strong red fluorescence as did porphyrins 11-14, in
contrast to the weak brown fluorescence shown by 4. This
new material showed a molecular ion at m/z 695 (695.2264)
which corresponded to the molecular formula C44H30N4O5
+ H (695.2294). Thus, the reduction of photoproduct 4 to
yield the corresponding dihydroxyphenylporphryrin had been
achieved. In the 1H NMR spectra of 11-14, each benzenoid
region is very different and can be used as a fingerprint.8
The reduction product from the photoproduct was only
similar in the benzenoid region to 5-(3,4-dihydroxyphenyl)-
10,15,20-tris(3-hydroxyphenyl)porphyrin (11) and the two
samples were undistinguishable on the TLC plate. Further-
more, the dehydrogenation of 11 with DDQ yielded a mixture
of mobile porphyrins: the products showed reduced fluo-
rescence on TLC under 366 nm light, and one of them was
identical with authentic material 4 on mixed TLC. However,
it was not possible to isolate a pure sample for further
analysis. Although we have attempted to grow crystals of 4,
they have been unsuitable for X-ray analysis. Our conclusion
is that the structure which best fits with these experimental
results is 5-(3,4-benzoquinonyl)-10,15,20-tetrakis(3-hydroxy-
phenyl)porphyrin (4) (Scheme 1).
1
extensive spectroscopic studies (IR, H and 13C NMR).8
Three benzoquinone isomers of 4 are possible, namely,
two o- and one p-benzoquinone. The 2,3-o-benzoquinone was
(7) Bonnett, R.; Djelal, B. D.; Hamilton, P.; Mart´ınez, G.; Wierrani, F.
J. Photochem. Photobiol. B: Biol 1999, 53, 136.
(8) For experimental procedures and characterization of new compounds,
see Supporting Information.
The presence of additional oxygen atoms in the remaining
four photoproducts compared to the starting material 1 was
(9) (a) Chin, A.; Dalton, J.; Milgrom, L. R. J. Chem. Soc., Perkin Trans.
2 1982, 707. (b) Dalton, J.; Milgrom, L. R. J. Chem. Soc., Chem. Commun.
1979, 609. (c) D’Souza, F.; Deviprasad, G. R.; Hsieh, Y. Y. J. Chem. Soc.,
Chem. Commun. 1997, 533. (d) Deviprasad, G. R.; Keshavan, B.; D’Souza,
F. J. Chem. Soc., Perkin Trans. 1 1998, 3133.
(10) Novel porphyrins 11-14 were prepared by the cross condensation
in propionic acid of 3-methoxybenzaldehyde (3 equiv) and the appropiate
dimethoxybenzaldehyde (1 equiv), followed by deprotection of the methyl
ethers to yield the phenols using BBr3. See details in Supporting Information.
2014
Org. Lett., Vol. 4, No. 12, 2002

Scheme 2. Porphyrins 11-14 and photobleaching of m-THPC (2) and m-THPBC (3)
proved by mass spectroscopy. The mass spectra (FAB)
respectively. The same type of behavior has been reported
for a series of 2,5-benzoquinonylporphyrins.9 Photoproducts
5-8 also showed little fluorescence on the TLC plate when
irradiated with 366 nm light. Further spectroscopic studies
showed that 5 and 6 had two extra oxygen atoms. The other
two minor photoproducts 7 and 8 appeared to have three
and four extra oxygen atoms, respectively (Table 1).
1
(IR, H and 13C NMR) led to the formulation of these four
compounds as benzoquinonylporphyrins 5-8 (Scheme 1).8
Kinetic studies7 show that the photooxidation of m-THPP
(1) is slower in methanol than in methanol-water (60:40),
and the same was observed in these preparative studies. After
41 h irradiation in methanol,8 the absorbance decay of band
IV (λmax ) 512 nm) was only 16% and the recovery of
m-THPP (1) was 53%. In addition to m-THPP (1), small
amounts of 4 (5%) were obtained. A number of fractions,
more polar than 1, were also isolated and appeared to be
porphyrins according to their UV-vis spectra. Due to the
small quantities formed, their structures were not studied fur-
ther. After 140 h, methyl 3-hydroxybenzoate (9, 2.5%) and
maleimide (10, 1.1%) were identified as the components of
a mobile fraction by 1H NMR, MS(EI), and GLC (Scheme 1).8
Table 1. MS(FAB) for Benzoquinonylporphyrins 5-8
5: C44H26N4O6 6: C44H26N4O6 7: C44H24N4O7 8: C44H22N4O8
+ 5H
+ 5H
+ 7H
+ 7H
711.2270a
711.2244b
711.2198a
711.2244b
727.2170a
727.2193b
741.1960a
741.1960b
a Found. b Calculated.
However, the molecular ions reported showed extra hydrogen
atoms above those required by the expected quinonoid
formula. Reduction of quinones is often observed during the
mass spectrometric process, and this appears to be occur-
ring.11 The mass of 6 was also measured by the APCI
technique, and the quinone rings did not appear to be
reduced, since m/z 707 (M + H) was measured.
Several features of the UV-vis spectra of 4-8 in methanol
are of interest: the Q-bands were broadened; the original
band III (λmax ) 546 nm) of m-THPP (1) decreased when
one benzoquinone ring (in 4) was present and disappeared
for 5-8; band IV was shifted to the blue (∆λ ) 2-9 nm);
and bands II and I moved to the red ∆λ ) 3 and 9 nm,
Solutions of m-THPC (2) in methanol or methanol-water
(60:40) were irradiated for 23 and 15 h, respectively.8 The
decay of absorbance was observed throughout the whole
region (450-800 nm). As observed during the kinetic
studies,7 the reaction occurred much more slowly in methanol
than in methanol-water (60:40). After 23 h of irradiation
in methanol, preparative TLC of the crude yielded m-THPC
(2, 46%) as the major fraction. A green base line and several
minor bands were also observed, but they accounted for less
than 10% of recovered material. However, a much more
extensive photodegradation of m-THPC (2) was effected
under the same experimental conditions in methanol-water
(60:40). In only 15 h, the absorbance decay of band I
(λmax ) 749 nm) was 78%. m-THPC (2, 10%) was recovered
(11) (a) Zeller, K. P. In The Chemistry of the Quininoid Compounds,
Part 1; Patai, S., Ed.; John Wiley and Sons: New York, 1974; Chapter 5,
p 231. (b) Zeller, K. P.; Muller, R. In The Chemistry of the Quininoid
Compounds, Volume 2, Part 1; Patai, S., Rappoport, Z., Eds.; John Wiley
and Sons: New York, 1988; Chapter 3, p 87. (c) Clayton, E.; Wakefield,
A. J. C. J. Chem. Soc., Chem. Commun. 1984, 969.
Org. Lett., Vol. 4, No. 12, 2002
2015

together with eight other fractions after preparative TLC on
silica (Scheme 2).
Methyl 3-hydroxybenzoate (9) and succinimide (17) were
identified as the components of the mixture. The total
recovered yield never exceeded 5%. Maleimide (10) was not
detected. The second fraction was an orange red pigment
which was less mobile than 9 and 17 but less polar than
m-THPBC (3). The color suggested that this pigment was a
linear pyrrolic pigment and it was formulated as 7,8-dihydro-
1,9-bis(3-hydroxybenzoyl)-5-(3-hydroxyphenyl)pyrro-
methene (19). The MS(FAB) spectrum showed a major peak
at 479 (found 479.1598) which corresponded to the molecular
formula C28H22N2O5 + H (calcd 479.1607). The UV-vis
spectrum showed broad bands at 332 and 429 nm, which
are characteristic of 5-phenylpyrromethenes.12b The 1H NMR
and COSY spectra also provided support for structure 19.8
The two most mobile fractions were white solids. The
major one was a mixture of two small fragments which were
identified by 1H NMR, MS(EI), and GLC as methyl
3-hydroxybenzoate (9, 1.5%) and maleimide (10, 1.2%). The
minor white solid was identified by GLC (comparison with
authentic sample) as succinimide (17, 0.6%). Two yellow
fragments, which were less polar than m-THPC (2), were
also isolated. The major one was obtained as a yellow-orange
solid. One peak in the MS(FAB) was found at m/z 373, which
can be ascribed to the molecular formula C22H15N2O4 + 2H.
A second more mobile yellow material was also isolated in
still smaller amounts. The UV-vis spectra of both substances
showed two main bands at 314 and 400 nm which suggested
that these materials could be dipyrrolic compounds. Pyrro-
methenones with no substituents in the carbon bridge show
bands at 264 and at 400 nm.12a 5-Phenyl-4,6-dipyrrin also
shows absorption bands in this region, at 354 and 466 nm.12b
Some assignments were made from the 1H NMR of the first
fraction and tentatively led us to formulate this material as 15.
Two more major fractions were isolated. Both materials
were colored, and their UV-vis spectra showed that they
were chlorins.8 The least polar chlorin was isolated as a dark
brown solid: the pigmented spot showed little fluorescence
on the TLC plate under 366 nm light. The MS(FAB)
spectrum showed a major peak at m/z 744 which could
correspond to a molecular formula such as C44H32N4O8 (four
more oxygens than m-THPC (2)). Since this chlorin was
more mobile than 2, it was thought that it could have had
the four phenol rings transformed into benzoquinones. The
formation of an unidentified species which showed an ion
at m/z 744 in the mass spectrum has also been reported by
Bezdetnaya et al.13 from the photooxidation of 2 in ethanol.
The last major fraction was another chlorin which was more
polar than m-THPC (2). The MS(FAB) spectrum showed a
major peak at m/z 697 (found 697.2493) which corresponded
to the molecular formula C44H32N4O5 + H (calcd 697.2451)
The third product isolated was m-THPC (2). The chlorin
2 was already in the starting material as an impurity, the
content being ca. 5-10% (determined by 1H NMR). During
the short period used for the photobleaching of the bacte-
riochlorin 3, m-THPC (2) is expected to suffer little
photobleaching: it was obtained in 4-8% yields. In this
preparative study, it was not possible to establish by isolation
of material if photooxidation of bacteriochlorin to chlorin
was occurring (initial content ca. 5-10%; isolated 4-8%).
To prove the formation of m-THPC (2), the photobleaching
of m-THPBC (3) was followed by 1H NMR spectroscopy.15
The peaks corresponding to m-THPC (2) gradually increased
in intensity (e.g., the area of the peak at δ 8.40 ppm of 2
increased ca. 50%). Thus, the formation of m-THPC (2) upon
irradiation of aerated solutions was confirmed. When the
1
experiment was carried out excluding the air, the H NMR
spectra were unchanged, confirming that oxygen was es-
sential for the photobleaching to occur.
In conclusion, the first example of the photooxidation of
phenolic porphyrin 1 to give benzoquinonylporphyrins is
reported.16 The corresponding chlorin 2 and bacteriochlorin
3 are photooxidized more readily, and cleavage products
include orange compounds regarded as dipyrrin derivatives.
Small amounts of colorless fragmentation products (imides,
methyl m-hydroxybenzoate) are detected in each case, and
3 gives a small but detectable yield of 2.
1
(one more oxygen than m-THPC (2)). The H NMR was
very complex, suggesting the presence of several isomers
of 16. Such a photoproduct was also obtained by Jones et
al.14 during irradiation of 2 in methanol.
Acknowledgment. G.M. is indebted to the University of
London Postgraduate trust for the award of a Laura de
Saliceto Cancer Research Studentship (1996-1998). Spec-
troscopic support from ULIRS (School of Pharmacy, Queen
Mary College). EPRSC (Swansea) is acknolwedged.
To complete the series, the photobleaching of m-THPBC
(3) in methanol was studied.8 Decay of absorbance through-
out the whole spectrum was observed in just 4.5 h, indicating
true photobleaching.7 The photobleaching of bacteriochlorin
3 was by far the most rapid process studied in this series.
After irradiation, the methanol was removed and the crude
solid was directly purified by preparative TLC. There were
only three major bands in the plate (Scheme 2).
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
ThismaterialisfreeofchargeviatheInternetathttp://pubs.acs.org.
OL025842C
The most mobile fraction proved to be a mixture. It was
isolated as a white solid and analyzed by GLC and MS(EI).
(15) Using two Philips MLU 300 W lamps, 17 mW/cm2, m-THPBC (3)
(c ) 8.8 × 10-3 M), was irradiated in air-saturated CD3OD (0.7 mL). The
NMR tube was placed in a water bath (constant flow of water) and irradiated.
At intervals, the tube was removed from the bath and the 1H NMR spectrum
was recorded.
(12) (a) Bonnett, R.; Hamzetash, D.; Valles, M. A. J. Chem. Soc., Perkin
Trans. 1 1987, 1383. (b) Bruckner, C.; Retting, S. J.; Dolphin, D. Can. J.
Chem. 1996, 74, 2182.
(13) Belitchenko, I.; Melnikova, V.; Bezdetnaya, L.; Rezzoug, H.; Merlin,
J. L.; Potapenko, A.; Guillemin, F. Photochem. Photobiol. 1998, 67, 584.
(14) Jones, R. M.; Wang, Q.; Lamb, J. H.; Djelal, B. D.; Bonnett, R.;
Lim, C. K. J. Chromatog. A 1996, 722, 257.
(16) The formation of p-benzoquinones from phenols mediated by singlet
oxygen has been previously reported, see: Saito, I.; Matsuura, T. In Singlet
Oxygen, Organic Chemistry, Vol. 40; Wasserman, H. H., Murray, R. W.,
Eds.; Academic Press: New York 1979; p 511.
2016
Org. Lett., Vol. 4, No. 12, 2002