The Histochemical Journal 33: 273–281, 2001.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Autofluorescence of melanins induced by ultraviolet radiation and
near ultraviolet light. A histochemical and biochemical study
Milan Elleder1 & Jan Borovanský2
Institute of Inherited Metabolic Disorders, Ke Karlovu 2, 12808 Prague, Czech Republic
2
II Department of Medical Chemistry and Biochemistry, Charles University and Faculty Hospital, 1st Faculty of Medicine,
Prague, Czech Republic
1
Received 26 March 2001 and in revised form 28 May 2001
Summary
The induction of autofluorescence of melanins by UV radiation (330–380 nm) and near UV (400–440 nm) light (jointly called
UV light) was studied in tissue sections using three commercially available mounting media. Only Immu-Mount (Shandon)
was found suitable for this purpose. UV irradiation of melanins in sections mounted in this medium induced strong yellow
autofluorescence irrespective of the type of the polymer (eumelanin, neuromelanin, pheomelanin and ochronotic pigment).
The phenomenon of autofluorescence induction was also observed with isolated natural and in vitro prepared melanins. It
was inhibited by anhydrous conditions, sodium azide and catalase. In parallel experiments, rapid degradation of melanins
with an intermediate fluorescent stage was achieved in UV-irradiated sections mounted in media artificially enriched with
hydrogen peroxide, or directly in aqueous solutions of H2 O2 , Na2 O2 or HIO4 . Oxidations not associated with UV light led to
nonfluorogenic breakdown of melanins. These observations indicate that the common mechanism may be an oxidative attack
resulting from a concerted action of hydrogen peroxide and UV light leading, through strongly fluorescent intermediates, to
a complete bleaching and oxidative breakdown of melanin and melanin-like polymers. Reactive oxygen species (including
ozone) are considered to be important reactants in these experiments. Lipopigments differ from melanin-like pigments by
their primary autofluorescence, which mostly faded during continuous prolonged irradiation. The only regular exception was
melanosis coli pigment, the autofluorescence of which was considerably augmented by UV irradiation. Our results demonstrate
a novel type of fluorogen in autofluorescent pigment histochemistry. The implications of the results are discussed especially
in the light of the possible presence of melanin-based fluorogens in lipopigments.
Introduction
Melanins are natural ubiquitous pigments. They are opaque,
mostly insoluble, materials of an aromatic polymeric irregular structure consisting of different monomer units linked by
a variety of bonds (Prota et al. 1998). They absorb light over
a wide spectral range, can convert energy from one type to
another, have stable free radical properties, are involved in
redox reactions, can behave as semiconductors and often display ion exchange properties (Borovanský 1996). All these
properties can be exploited for maintaining cell homeostasis. Melanins are formed by the oxidation of phenolic precursors with a subsequent polymerization of the arising
quinones (Riley 1995). Authentic melanins, deposited in
subcellular organelles–melanosomes, include two classes of
pigment – eumelanins derived from l-tyrosine and pheomelanins synthesized from l-tyrosine and l-cysteine. Precursors
of neuromelanin are l-dopamine and l-cysteine. Ochronotic
pigment arising from the oxidation of homogentisic acid
shares many features with melanin (Norfray et al. 1988).
In terms of histochemistry, melanins are reputed to reduce
the ‘simple’ and complex (diamo) silver cations to metallic silver (Lillie and Fullmer 1964) and to be sensitive to
oxidative bleaching agents, which are capable of destroying the melanin polymer (Pearse 1972, Korytowski & Sarna
1990, Mosca et al. 1999). These, together with the absence
of fluorescence expected hitherto, have been pivotal methods
of pigment histochemistry and serve to differentiate melanin
and melanin-like pigments from the family of lipopigments
(Pearse 1972, Elleder 1981, Senba 1986, Lyon 1991).
We present evidence that irradiation with UV light of
samples containing melanin can, in the presence of peroxides
or periodic acid, induce a strong yellow autofluorescence due
to products of oxidative degradation. Since the induction of
fluorescence proved to be a common novel feature of all
types of melanins, it represents a new tool with a possible
impact on melanin and generally on autofluorescent pigment
histochemistry.
Material and methods
Standard paraffin sections of formaldehyde-fixed tissues
were used for most experiments. Skin samples from current biopsies of pigmented naevi of various types (n = 40),
enucleated human eye bulbi (n = 2), mouse C57BL6JJ eye
274
bulbi (n = 3), human autopsy samples of substantia nigra
with neuromelanin (n = 3) and of alkaptonuria (aortic wall,
n = 1), samples of human liver (n = 3), heart (n = 2),
brain (n = 3) with lipofuscin age pigments, and brain
from neuronal ceroid-lipofuscinosis (n = 2), spleen from
Niemann-Pick disease type B with ceroid in storage histiocytes (n = 2), liver from Dubin-Johnson syndrome (n = 5),
gut with melanosis coli (n = 10) and from HermanskyPudlak syndrome (n = 1) were studied. In selected cases,
cryostat sections from unfixed frozen tissues were also used.
Paraffin sections were dewaxed with xylene and acetone, hydrated, then washed in distilled water and mounted
into Immu-Mount (Shandon, Pittsburgh, PA, USA) containing polyvinyl alcohol resin and glycerol as the main
components and less than 1% amino alcohol and cationic
quarternary ammonium chloride. They were examined for
autofluorescence using the filter blocks specified below, at
time zero and after continuous irradiation up to 3 h. In selected
experiments, the sections were mounted in the following artificial media: glycerol, 10% polyvinyl alcohol in distilled
water (mol. wt. 30,000–70,000), distilled water, distilled
water alkalized with concentrated ammonium hydroxide
(pH 8), 1% aqueous periodic acid, 1% sodium peroxide
or a drop of 10% hydrogen peroxide, or mineral oil. The
mounted sections were covered with a standard glass coverslip and irradiated with UV light from a 100 W mercury UV
lamp in a Nikon E800 microscope (excitation/emission filter
blocks: 330–380/420, 400–440/470 and 510–560/590 nm).
Most experiments were carried out with the 400–440/470 nm
filter block. Other commercially available mounting media
tested were Vectashield D (Vector Laboratories, Burlingame,
CA, USA), and Permafluor (Immunotech, Marseille, France).
Melanin bleaching was effected either with 10% hydrogen
peroxide alkalized with several drops of concentrated ammonium hydroxide to pH 7.4–7.8 (1–3 h at room temperature)
or with 1% potassium permanganate in 1% H2 SO4 (30, 60
and 120 sec at room temperature), followed by oxalic acid to
remove MnO2 . In some experiments the sections were preextracted using 2 : 1 v/v chloroform–methanol overnight at
room temperature.
Black (human and dog) and red (human and cat) hairs were
placed directly onto glass histology slides and mounted in
Immu-Mount. Melanosomes from black poodle hair were isolated by the method of Borovanský and Hach (1986) while the
classic method of Seiji et al. (1963) was employed to obtain
melanosomes from the minipig MeLiM melanoma (Horák
et al. 1999) and the Bomirski hamster melanoma.
Melanin from the MeLiM strain of minipig melanoma
was prepared as described previously (Borovanský 1978).
Synthetic eumelanin was synthesized by the auto-oxidation
of l-DOPA (Sigma, USA) (Borovanský et al. 1986) and
extensively dialyzed to remove soluble melanin precursors.
Synthetic pheomelanin was obtained by the auto-oxidation
of l-DOPA and l-cysteine (Sigma, USA) as recommended
by Deibel and Chedekel (1982). Synthetic neuromelanin
was made by the procedure of Zecca et al. (2000)
from l-dopamine (Sigma, USA) and l-cysteine. Synthetic
ochronotic pigment was prepared from homogentisic acid
M. Elleder & J. Borovanský
(Fluka, Switzerland) in two ways: (a) by auto-oxidation with
air in bicarbonate/carbonate buffer pH 10.5 (ochro A); and
(b) by enzyme catalysis with tyrosinase (Sigma, USA) in
0.1 M phosphate buffer, pH 6.8 (ochro T).
In experiments aimed at the inhibition of melanin fluorescence development, two drops of Immu-Mount were mixed
with (a) two drops of TRIS-HCl buffer pH 7.99 (buffer T/H)
(control samples), (b) two drops of catalase solution (catalase
from bovine liver, Sigma, FRG, 17,000 units/mg solid; 1 mg
dissolved in 100 µl of the T/H buffer as in a), (c) two drops of
sodium azide (Fluka, Switzerland; 2.7 mg dissolved in 200 µl
T/H buffer).
Results
Induction of autofluorescence
Induction of autofluorescence of melanin of any type
was strongly dependent on the type of mounting medium
used. Positive results were obtained only with ImmuMount medium (irrespective of the date of expiry). Neither
of the other two commercially available mounting media
(Permafluor, Vectashield) exhibited this effect. Induction of
autofluorescence was accompanied with partial bleaching
of the pigment. A shift from the near UV light (400–440)
to UV radiation (340–380 nm) accelerated the induction of
autofluorescence in Immu-Mount medium, and the addition
of catalase to the mounting medium caused a significant
delay similar to the addition of sodium azide. No induction
was observed with the following mounting media: paraffin oil, concentrated glycerol, 10% polyvinylalcohol (mol.
wt. 30,000–70,000), distilled water, distilled water alkalized
with ammonium hydroxide (pH 8). Mounting of sections in a
drop of 10% hydrogen peroxide, 1% periodic acid or 1%
sodium peroxide led to an enormous acceleration of both
autofluorescence induction and complete bleaching by UV
within 5 min. Heavy deposits of melanin in strongly pigmented melanoma were destroyed within several minutes.
Their temporary wave of autofluorescence was difficult to
photograph. The induced autofluorescence always persists
for some time after complete bleaching. A similar effect was
observed after addition of several drops of 10% hydrogen
peroxide to glycerol or polyvinylalcohol. Irradiation of dry
cryostat or dry unmounted paraffin sections was without any
effect; autofluorescence induction was seen in both cryostat
and paraffin sections after prolonged lipid extraction. The
induction of melanin autofluorescence was always accompanied by the disappearance of the background nonspecific
tissue autofluorescence.
In control experiments the induction of autofluorescence
was tested via oxidation with slightly alkalized 10% hydrogen peroxide or with acidified potassium permanganate without UV irradiation. Although there was significant bleaching, there was no sign of autofluorescence induction in any
of the samples tested (epidermal melanin, neuromelanin,
ochronotic pigment). The residual pigment displayed autofluorescence after a subsequent UV irradiation.
Autofluorescence of melanins
The following results were obtained in paraffin sections
mounted into the Immu-Mount medium and irradiated with
the near UV light.
Eumelanin
Epidermal melanin
The melanin granules (normal epidermis, pigmented naevi)
became autofluorescent during several minutes of irradiation
with the maximum attained after 15–20 min (Figure 1A,B).
During this period there was a partial pigment bleaching.
Retinal pigment epithelium melanin
In the adult human retina there was a mass of primarily autofluorescent lipofuscin in the basal part of the retinal pigment
epithelium forming a continual autofluorescent line. Melanin
granules of the chorioidal melanocytes and in the apical
part of the retinal pigment epithelium displayed strong autofluorescence within 10–20 min. The intensity of lipofuscin
autofluorescence did not change or increased slightly during
UV irradiation. Fetal human eye retinal pigment epithelium, strongly melanin pigmented, was free of primary autofluorescent granules. Melanin autofluorescence was induced
within the same interval of UV irradiation (Figure 1C–F).
The C57BL6JJ mouse pigmented retinal pigment epithelium
behaved identically.
Synthetic aggregated eumelanins
Synthetic aggregated eumelanins (nondialyzed, dialyzed)
displayed strong yellow autofluorescence after 1 h UV irradiation with a slight decoloration (Figure 1G,H). An aqueous
solution of l-DOPA did not exhibit fluorescence. However,
when dissolved in the 0.1 M Na2 CO3 , it quickly turned brown
and fluorescent when irradiated. Purified melanosomes from
the Bomirski hamster (Mesocricetus auratus) melanoma,
from the minipig MeLiM melanoma and from the canine hair
behaved similarly. For each, maximum fluorescence was displayed after 3 h irradiation and the samples became slightly
decolorized.
Pheomelanin
Artificially prepared pheomelanin gave an immediate autofluorescence at neutral and alkaline pH. At acid pH the synthetic pheomelanin precipitated and failed to react. Attempt at
pheomelanin screening in skin naevi showed primary autofluorescence attributable to pheomelanin in only one of 27
different samples of pigmented benign naevi. Three dysplastic naevi, known to be rich in pheomelanin (Salopek et al.
1991) were unexpectedly negative.
Hair samples
In black (= eumelanin) human and dog hair, strong yellow
autofluorescence was induced by a 2-h period of irradiation;
in red (= pheomelanin) human and cat hair, melanin granules exhibited primary autofluorescence, which was further
augmented during the UV irradiation (up to a 2-h period).
275
Neuromelanin
Neuromelanin granules in adulthood consist of two components – neuromelanin and lipofuscin (Barden 1969, 1981).
Physical properties of the lipofuscin component of neuromelanin (primary yellow autofluorescence, sudanophilia) are
blocked in intact granules and are apparent after removal of
the melanin. Neuromelanin granules become autofluorescent
during the UV irradiation at a rate comparable with eumelanin (Figure 2A,B). Identical results were observed in both
adult and childhood nigral neurons (the latter free of lipofuscin). After a 30-min period of irradiation, the autofluorescent
neuromelanin granules were still partly pigmented.
Total oxidative bleaching of neuromelanin (with 10%
hydrogen peroxide, slightly alkalized to pH 8 with
concentrated NH4 OH, for 1 h at room temperature) demasked
lipofuscin autofluorescence and did not change it in the surrounding nigral neurons containing only lipofuscin. Synthetic
neuromelanin suspensions placed on histological slides and
mounted in Immu-Mount displayed strong autofluorescence
after a 20–30-min period of illumination.
Ochronotic pigment
The pigment in the aortic wall developed autofluorescence in
paraffin sections within a substantially shorter period of UV
irradiation. Strong autofluorescence of the pigment deposits,
which can be granular or dusty, was induced after only 10 min
of UV irradiation (Figure 2C,D). The initial emission was
detectable within the first minute. In cryostat sections there
was a slight but definite primary autofluorescence of the pigment deposits, which was further strongly augmented as in
the paraffin sections. The pigment in the paraffin sections of
ochronotic aorta was significantly bleached by 1 h treatment
with 10% hydrogen peroxide (pH 7.2–7.4) without inducing
fluorescence. This was, however, induced by subsequent UV
irradiation.
Synthetic ochronotic pigment had various properties
depending on the origin of the pigment. Synthetic ochronotic
pigment prepared by autoxidation (ochroA) was autofluorescent, whereas the ochronotic pigment synthesized via
tyrosinase catalysis (ochroT) failed to emit light even after
the addition of 10% hydrogen peroxide.
Lipopigments
Lipofuscins (myocardial, neuronal, hepatic), the so called
ceroid-lipofuscins in neuronal ceroid-lipofuscinoses, and
ceroids in Niemann-Pick diseases types B, and C did not
exhibit an increment with respect to the strong primary fluorescence within the 10 min interval of continuous irradiation,
but in fact faded with continuous exposure to the UV light.
The only exceptions were the liver lipofuscins, which repeatedly displayed an initial increase in autofluorescence.
The liver Dubin-Johnson pigments displayed medium
strong to strong primary autofluorescence, which was in some
cases further slightly increased during the 10 min irradiation period. The melanosis coli pigment was primarily of
276
M. Elleder & J. Borovanský
Autofluorescence of melanins
low intensity autofluorescence of yellow–brown or brown
colour. In all cases the autofluorescence was considerably
increased after 10 min irradiation (Figure 2E,F). In addition, there was an easily recognizable autofluorescence of
the peripheral endocrine cells (Kulschticky cells), which
was increased during the irradiation. There was no increase
of autofluorescence using 10% alkalized H2 O2 . The autofluorescence of the lipopigment in Hermanský-Pudlák syndrome varied according to localization. In the large bowel
it was of low intensity and brown, of intermediate intensity
in renal deposits, and strongest in spleen macrophage clusters. A 10 min UV irradiation was of equal intensity to the
autofluorescence of the spleen macrophage lipopigment.
277
unambiguous evidence that the emission-responsible fluorophores are inherent to the pigment and excludes other compounds, e.g. pteridines which may physiologically bind to
melanosomes (Obika 1976).
As for ochronotic pigment, our observation partially conflicts with the prevailing view that the pigment fluorescence is primarily autofluorescence (Pearse 1972, Lillie &
Fullmer 1976). Primary fluorescence is supported by the
notion that the fluorescence of homogentisic acid increases
during its oxidative polymerization to so-called plasma soluble melanins-lipofuscins (Hegedus & Nayak 1993). Tissue
processing during paraffin embedding might explain the different results observed in cryostat sections (see Results).
Mechanism of autofluorescence induction
Discussion
In terms of current histochemistry, melanin represents a nonfluorescent pigment which differs from the family of autofluorescent lipopigments (Senba 1986, Pearse 1972, Lyon
1991). Only Boulton et al. (1990) have reported that untreated
melanosomes in situ in bovine retinal pigment epithelium
as well as separated ones are weakly fluorescent. Indirect
support of their observation was provided by Dintelmann
et al. (1999). Evaluation of the primary autofluorescence
of pheomelanin requires further study (see below). Fluorescence induced in melanocytes by formaldehyde has long
been attributed to the presence of dopamine (Falck et al.
1965). Massive induction of autofluorescence of melanins
by UV under specific conditions in situ was, therefore,
a surprising finding. The discrepancy between the generally accepted nonfluorescent property of melanin and our
finding can be explained by the difference in the mounting media used. Usually fluorescent pigment properties are
examined either in hydrophobic media (mineral oil) or in
glycerol, both being inhibitory. There is also a fundamental
difference between commercially available mounting media,
Immu-Mount medium is the only one among the three tested
with which the fluorescence induction was observed. The
time factor may also play a role as fluorescence only begins to
be observed after about ten minutes. This may be longer than
is usually expended on examining sections in a fluorescence
microscope.
Our finding allows us to conclude that the induction of the
autofluorescence seems to be a common feature of all natural
and synthetic melanins and melanin-like pigments because
it could be induced in eumelanin, pheomelanin, neuromelanin and also in ochronotic pigment samples both in vivo
and in in vitro conditions with synthetic samples. This gives
The results of our experiments allow us to propose that the
UV-induced fluorescence of melanins in sections mounted
in Immu-Mount medium is associated with photochemical
generation of hydrogen peroxide. The addition of catalase to
the medium inhibited the induction of fluorescence. Sodium
azide, a scavenger of free radicals (Slawinska & Slawinski
1987), also exerted a suppressive effect on the fluorescence
induction. However, details of the mechanism of the fluorescence promoting property of the Immu-Mount medium
remains to be elucidated.
Our view is corroborated by the following data. Felix et al.
(1978) demonstrated that the irradiation of melanins in the
range of 320–600 nm in aerated media produced superoxide radicals and hydrogen peroxide. Korytowski et al. (1985)
showed that dopamelanin behaved as a pseudosuperoxide dismutase producing H2 O2 with a dramatic increase of its formation at an alkaline pH. Slawinska and Slawinski (1987)
and Korzhova et al. (1989b) documented that the degradative
oxidation of synthetic melanins by reactive oxygen species
and hydrogen peroxide is accompanied by photon emission.
In contrast to eumelanin, pheomelanin has been shown to
be exceptionally photolabile. Irradiation of aerated samples
of pheomelanin, either in the solid state or in aqueous solution, produces at least 100 times the amount of superoxide
as does eumelanin (Chedekel 1982). This could explain the
immediate fluorescence of pheomelanin and pheomelaninrich structures (see also below).
However, in parallel experiments, partial bleaching of
melanins with hydrogen peroxide or with acidified permanganate failed to induce autofluorescence. This also speaks
against a possibility of the presence of masked primary fluorogens (see below). We, therefore, do not consider hydrogen
peroxide under these standard histochemical conditions to
Figure 1. (A–H) UV-induced autofluorescence in eumelanin samples in the skin (A,B), in the fetal (C,D), and adult retinal pigment epithelium (E,F)
and in artificially-prepared eumelanin samples, dialyzed (G) and nondialyzed (H). The same section is presented in two versions. The left row represents
in each sample the starting situation with the absence of melanin autofluorescence. The only exception is the strong autofluorescence of lipofuscin
at the basis of the retinal pigment epithelium (E, between arrowheads). In G, the border of the UV-irradiated area is situated between the dark and
fluorescent melanin deposits. The right row shows strong induction of melanin autofluorescence in the samples after 10 min irradiation with near UV.
Unstained paraffin sections and melanin samples mounted in Immu-Mount. A–G ×260, H ×100.
278
M. Elleder & J. Borovanský
Autofluorescence of melanins
be the exclusive factor responsible for the induction of autofluorescence. The crucial finding was that hydrogen peroxide
in combination with UV strongly augmented in the melanins
both fluorescence induction and destruction, either alone
or when added to the experimental test media. Oxidationinduced fluorescence of melanin in the retinal pigment epithelium (Kayatz et al. 2001) must also have been due, according
to the protocol used, to the combined action of hydrogen
peroxide and UV irradiation. We found it interesting that
similar enhancement of fluorogenic destruction of melanin
was observed using UV light combined with oxygen rich
molecules, such as sodium peroxide and periodic acid, the
latter known to liberate reactive oxygen species under these
conditions (Head 1950, Head & Standing 1952).
These findings are corroborated by Korytowski and Sarna
(1990) who found that bleaching of melanin with hydrogen
peroxide is significantly accelerated by UV irradiation. Other
sparse data from the literature dealing with autofluorescence
of melanin are difficult to comment on as the experimental
in vitro conditions differed significantly. Induction of fluorescence of DOPA melanin has been described after prolonged illumination with UV light (Korzhova et al. 1989a)
and after boiling with hydrogen peroxide (Rosenthal et al.
1973, Korzhova et al. 1989b).
The nature of the induced melanin fluorogen(s) is
unknown. It has been suggested that they are induced by partial oxidative breakdown of melanin and may thus depend
on structural defects of melanin polymers (Gallas & Eisner
1987, Korzhova et al. 1989b, Rosenthal et al. 1973). Our
observation of strong melanin fluorescence when using all
three different filter blocks is in accord with the data in the
literature (Gallas & Eisner 1987, Kozikowski et al. 1984,
Boulton et al. 1990, Forest et al. 2000).
To conclude, we have provided evidence of a novel, highly
efficient and reproducible oxidative breakdown of melanin
and related polyphenols by the simultaneous action of hydrogen peroxide and UV light, giving rise to one or more strongly
fluorescent intermediates. In the light of our results, we
suggest that future studies of the fluorogenic nature of the
melanin oxidative degradation pathway should be performed
using appropriately designed experimental protocols defining
the main reactive oxygen species.
It is worth mentioning the existence of potential primary
melanin fluorogens, for instance melanin precursors and/or
metabolites, e.g. indoles (Pavel 1989), or benzothiazines and
their photolytic derivatives (Liu & Chedekel 1982). In opiomelanins, the fluorescence contains contributions from both
oligomeric units and high molecular weight polymers (Mosca
et al. 1999). Oxidative polymerization of various aromatic
279
amines, aminophenols or phenols generates autofluorescent,
low molecular weight, soluble melanins/rheomelanins –
soluble lipofuscins (Hegedus 2000). The 5,6-indolquinone
melanin units (Simon 2000), especially when altered
(Slawinska & Slawinski 1982, Kozikowski et al. 1984), or
5,6-dihydroxyindole units chemically and physically modified by the polymerization process (Gallas & Eisner 1987)
were considered to be involved in the autofluorescence of
eumelanins. There are observations suggesting the existence of primary fluorescence of melanin effectively masked
by absorption of the emitted light by its chromophores
(Gallas & Eisner 1987, Kozikowski et al. 1984, Simon
2000). This masking effect would explain the scarce attention
which melanin fluorescence has received until recently in the
chemical literature.
It is not known whether melanins and melanosomes are
physiologically degraded (rev. by Borovansky et al. 1999).
Considering the structure of melanins, redox mechanisms
seem to be more probably involved in pigment degradation than hydrolytic reactions (see Korzhova et al. 1989b,
Slawinska & Slawinski 1982, Korytowski & Sarna 1990).
A fundamental question remains whether the possible in vivo
degradative pathway might go via fluorescent intermediates which would allow speculation about the possible
relationship between melanin and lipopigment at least in
several special instances, where the two pigments are intimately associated. This age-dependent association between
neuromelanin and lipofuscin in nigral brain neurons is well
established (Barden 1969). An extremenly intimate association between melanin and lipofuscin exists in retinal
pigment epithelium melanosomes, which acquire an autofluorescent rim during ageing (Feeney 1978, Boulton et al.
1990). It is worth speculating if this could be a consequence of chronic exposure of retinal pigment epithelium
melanosomes to UV and near-UV light. This hypothesis
distinguishes between the independent aposition of lipofuscin and melanin in one compartment (Schreaermayer &
Stieve 1994) and should provide evidence for a different
fluorogen from that described for lipofuscin in retinal pigment epithelium (Eldred & Lasky 1993). Also normal dermal perivascular macrophages (melanophages) are constantly
harbouring melanin together with autofluorescent lipopigment (M.E. personal observations). The melanosis coli pigment that has been suggested to be linked to the destruction of
the colonocytes by the anthraquinone type laxatives (Walker
et al. 1988) still lacks satisfactory explanation. On the basis
of its autofluorescence, particularly its outstanding increase
on UV irradiation, the melanosis pigment strongly resembles
the behaviour of melanins. Its argentaffin pattern, however,
Figure 2. (A–F) UV-induced autofluorescence in neuromelanin in the nigral nucleus neurons (A,B), ochronotic pigment deposited in the aortal media
(C,D) and in the melanosis coli pigment (E,F). Note absence of neuromelanin fluorescence (A) at the beginning of the irradiation (neuron 1 and 2)
contrasting with the dispersed glial and clustered neuronal lipopigment (arrowhead) (A). After 10 min of irradiation with near UV light (B), the
neuromelanin attained the same autofluorescence as lipopigment fluorescence. The ochronotic pigment acquired a granular and dust-like pattern of
autofluorescence on UV irradiation (D). Note the remarkable augmentation (F) of the initially faint brownish autofluorescence (E) of the melanosis
coli pigment. Unstained paraffin sections mounted in Immu-Mount. ×260.
280
differs fundamentally from that of melanin (Lillie & Fullmer
1976).
The existence of polyphenol-related fluorogens (precursors or degradation products) should be considered in the
differential analysis of autofluorescent pigments, not only
melanins (see pheomelanin and ochronotic pigment), but also
in the broad and still undefined field of lipofuscin.
Acknowledgements
This study was supported by Charles University Grant
Agency (grant 47/1999) and partly by project VS 96127 of
the Ministry of Education and Youth of the Czech Republic.
The authors wish also to thank Miss J.A. Cartwright, Headteacher of Lady Smith School, Exeter, U.K. for correcting the
English of this paper.
The results reported here were presented at the VIII Meeting of the European Society for Pigment Cell Research,
Prague, September 1998 (Pigment Cell Res. 11, S234, 1998).
References
Barden H (1969) The histochemical relationship of neuromelanin and
lipofuscin. J Neuropathol Exp Neurol 28: 419–441.
Barden H (1981) The biology and chemistry of neuromelanin.
In: Sohal RS, ed. Age Pigments. Elsevier/North Holland: Biomedical
Press, pp. 204–241.
Borovanský J (1978) Quantitative determination of melanin. Mikrochimica
Acta (Wien) II: 423–429.
Borovanský J (1996) Free radical activity of melanins and related substances: biochemical and pathobiochemical aspects. Sbornı́k lékařský
97: 49–70.
Borovanský J, Hach P (1986) Isolation of melanosomes from keratinous structures: Current state of the art. Arch Dermatol Res 279:
54–58.
Borovanský J, Hach P, Smetana Jr K, Elleder M, Matouš-Malbohan I
(1999) Attempts to induce melanosome degradation in vivo. Folia
Biologica (Prague) 45: 47–52.
Borovanský J, Melezı́nek I, Buděšı́nská A (1986) Interference of melanin
in protein determination. Anal Biochem 159: 249–252.
Boulton M, Docchio F, Dayhaw-Barker, Ramponi R, Cubeddu R (1990)
Age related changes in the morphology, absorption and fluorescence
of melanosomes and lipofuscin granules of the retinal pigment epithelium. Vision Res 30: 1291–1303.
Chedekel MR (1982) Photochemistry and photobiology of epidermal
melanins. Photochem Photobiol 35: 881–885.
Deibel RMB, Chedekel M (1982) Biosynthesis and structural studies on
pheomelanin. J Am Chem Soc 104: 7306–7309.
Dintelmann TS, Heimann K, Kayatz P, Schrayermeyer U (1999) Comparative study of ROS degradation by IPE and RPE cells in vitro. Graefe’s
Arch Klin Exp Ophthalmol 237: 830–839.
Eldred GE, Lasky MR (1993) Retinal age pigments generated by selfassembling lysosomotropic detergents. Nature 361: 724–726.
Elleder M (1981) Chemical characterization of age pigments. In: Sohal
RS, ed. Age Pigments. Elsevier/North Holland: Biomedical Press, pp.
204–241.
Falck B, Jacobson S, Olivecrona H, Rorsman H (1965) Pigmented nevi
and malignant melanomas as studied with a specific fluorescence
method. Science 149: 439–440.
Feeney L (1978) Lipofuscin and melanin of human retinal pigment
epithelium. Fluorescence, enzyme cytochemical, and ultrastructural
studies. Invest Opthalmol Vis Sci 17: 583–600.
M. Elleder & J. Borovanský
Felix CC, Hyde JS, Sarna T, Sealy RC (1978) Melanin photoreactions
in aerated media. Electron spin rezonance evidence for production of
superoxide and hydrogen peroxide. BBRC 84: 335–341.
Forest SE, Lam WC, Millar DP, Nofsinger JB, Simon JD (2000) A model
for the activated energy transfer within eumelanin aggregates. J Phys
Chem B 104: 811–814.
Gallas JM, Eisner M (1987) Fluorescence of melanin – dependence upon
excitation wavelength and concentration. Photochem Photobiol 45:
595–600.
Head FSH (1950) Effect of light on the reaction between periodate and
α-glycols. Nature 165: 236–237.
Head FSH, Standing HA (1952) Reaction in aqueous solutions of sodium
metaperiodate exposed to artificial light. J Chem Soc 1457–1460.
Hegedus ZL (2000) The probable involvement of soluble and deposited
melanins, their intermediates and the reactive oxygen side-products in
human diseases and aging. Toxicology 145: 85–101.
Hegedus ZL, Nayak U (1993) Homogentisic acid and structurally related
compounds as intermediates in plasma soluble melanin formation
and in tissue toxicities. Arch Internat Physiol Biochim Biophys 102:
175–181.
Horák V, Fortýn K, Hruban V, Klaudy J (1999) Hereditary melanoblastoma in miniature pigs and its successful therapy by devitalization
technique. Mol Cell Biol 45: 1119–1129.
Kayatz P, Thumann G, Luther TT, Jordan JF, Bartz-Schmidt KU, Esser PJ,
Schraermeyer U (2001) Oxidation causes melanin fluorescence. Invest
Opthalmol Vis Sci 42: 241–246.
Korytowski W, Hintz P, Sealy RC, Kalyanaraman B (1985) Mechanism of
dismutation of superoxide produced during autooxidation of melanin
pigment. BBRC 131: 659–665.
Korytowski W, Sarna T (1990) Bleaching of melanin pigments. J Biol
Chem 265: 12410–12416.
Korzhova LP, Frolova EV, Romakov YA, Kuznetsova NA (1989a) The
photoinduced destruction of dopa melanin. Biokhimija 54: 992–998
(in Russian).
Korzhova LP, Frolova EV, Romakov YA (1989b) Spectrofluorimetric procedure for registration of products developed after oxidative destruction of eumelanins. Vopr Med Khim 35: 139–143 (in Russian).
Kozikowski SD, Wolfram LJ, Alfano RR (1984) Fluorescence spectroscopy of eumelanins. IEEEJ Quantum Electronics 20: 1379–1382.
Lillie RD, Fullmer HM (1976) Histopathological techniques and
practical histochemistry. 4th edn, New York: McGraw-Hill.
Liu CT, Chedekel MR (1982) Pheomelanin photochemistry: Photolysis
of model compounds. Photochem Photobiol 36: 52–54.
Lyon H (1991) Theory and strategy in histochemistry. Heidelberg:
Springer, pp. 237–250.
Mosca L, De Marco C, Fontana M, Rosei MA (1999) Fluorescence properties of melanins from opioid peptides. Arch Biochem Biophys 371:
63–69.
Norfray JF, Klingler WC, Menon IA, Persad S, Berger PA, Ainscough M
(1988) A paramagnetic agent causing ochronotic arthropathy. Invest
Radiol 23: 609–615.
Obika M (1976) Reversible binding of riboflavin and pteridines to
melanin in vitro. Comp Biochem Physiol 53B: 511–523.
Pavel S (1989) Analysis of eumelanin-related indolic compounds. Trends
in Anal Chemistry 8: 177–183.
Pearse AGE (1972) Histochemistry – Theoretical and Applied. London
and Edinburgh: Churchill-Livingstone.
Prota G, D’Ischia M, Napolitano A (1998) The chemistry of melanins
and melanogenesis. In: Nordlund JJ, Boissy RE, Hearing VJ,
King RA, Ortone JP eds. The Pigmentary System: Physiology and
Pathophysiology, Oxford University Press.
Riley PA (1995) The evolution of melanogenesis. In: Zeise L,
Chedekel MR, Fitzpatrick TB, eds. Melanin: its Role in Human
Photoprotection. Overland Park, Kansas: Valdemar Publishing
Company, pp. 1–10.
Rosenthal MH, Kreider JW, Shiman R (1973) Quantitative assay of
melanin in melanoma cells in culture and in tumors. Anal Biochem
56: 91–99.
Autofluorescence of melanins
Salopek TG, Yamada K, Ito S, Jimbow K (1991) Dysplastic melanocytic
nevi contain high-levels of pheomelanin. Quantitative comparison of
pheomelanin/eumelanin levels between normal skin, common nevi,
and dysplastic nevi. Pigment Cell Res 4: 172–179.
Sarna T, Korytowski W, Sealy RC (1985) Nitroxides as redox probes of
melanins: Dark-induced and photoinduced changes in redox equilibria.
Arch Biochem Biophys 239: 226–233.
Schraermeyer UZ, Stieve H (1994) A newly discovered pathway of
melanin formation in cultured retinal pigment epithelium of cattle.
Cell Tissue Res 276: 273–279.
Seiji M, Shimao K, Birbeck MSC, Fitzpatrick KB (1963) Subcellular localization of melanin biosynthesis. Ann NY Acad Sci 100:
497–533.
Senba M (1986) Staining properties of melanin and lipofuscin pigments.
Am J Clin Pathol 86: 556.
281
Simon JD (2000) Spectroscopic and dynamic studies of the epidermal
chromophores trans-urocanic acid and eumelanin. Acc Chem Res 33:
307–313.
Slawinska D, Slawinski J (1982) Electronically excited molecules in
the function and degradation of melanins. Physiol Chem Phys 14:
363–372.
Slawinska D, Slawinski J (1987) Ultraweak photon emission in model
reactions of the in vitro formation of eumelanins and pheomelanins.
Pigment Cell Res 1: 171–175.
Walker NI, Bennett RE, Axelsen RA (1988) Melanosis coli. A consequence of anthraquinone induced apoptosis of colonic epithelial cells.
Am J Pathol 131: 465–476.
Zecca L, Costi P, Meccaci C, Ito S, Terroni M, Sonnino S (2000) Interaction of human substantia nigra with lipids and peptides. J Neurochem
74: 1758–1765.