Disentangling Eumelanin "black chromophore": Visible absorption changes as signatures of oxidation state- and aggregation-dependent dynamic interactions in a model water-soluble 5,6-dihydroxyindole polymer

Article abstract of DOI:10.1021/ja905162s

A fundamental unsettled issue concerning eumelanins, the functional biopolymers of human skin and hair, is why they are black. The experimental difficulty lies in the virtual insolubility of these pigments, causing marked scattering effects and hindering

Full text of DOI:10.1021/ja905162s

Published on Web 10/06/2009
Disentangling Eumelanin “Black Chromophore”: Visible
Absorption Changes As Signatures of Oxidation State- and
Aggregation-Dependent Dynamic Interactions in a Model
Water-Soluble 5,6-Dihydroxyindole Polymer
Alessandro Pezzella,* Alfonso Iadonisi, Silvia Valerio, Lucia Panzella,
Alessandra Napolitano, Matteo Adinolfi, and Marco d’Ischia
Department of Organic Chemistry and Biochemistry, UniVersity of Naples “Federico II”
Complesso UniVersitario Monte S. Angelo, Via Cintia 4, I-80126 Naples, Italy
Received June 23, 2009; E-mail: alessandro.pezzella@unina.it
Abstract: A fundamental unsettled issue concerning eumelanins, the functional biopolymers of human
skin and hair, is why they are black. The experimental difficulty lies in the virtual insolubility of these pigments,
causing marked scattering effects and hindering characterization of the intrinsic absorption properties of
the heterogeneous species produced by oxidative polymerization of 5,6-dihydroxyindole (DHI) and related
monomer precursors. The synthesis of spectrally robust, water-soluble DHI polymers is therefore an
important goal in the prospects of disentangling intrinsic absorption properties of eumelanin components
by circumventing scattering effects. Reported herein is the first water-soluble DHI polymer produced by
oxidation of ad hoc designed 5,6-dihydroxy-3-indolyl-1-thio-ꢀ-D-galactopyranoside (1). The dark brown
polymer exhibited a distinct band at 314 nm and a broad visible absorption, resembling that of natural
eumelanins. Main isolable oligomer intermediates including 2,7′- and 2,4′-biindolyls 2 and 3, attest the
close resemblance to the mode of coupling of the parent DHI. Sodium borohydride reduction caused
decoloration and a marked absorbance decrease in the visible region around 550 nm, but did not affect
the UV band at 314 nm. Measurements of absorbance variations with dilution indicated a linear response
at 314 nm, but a significant deviation from linearity in the visible region, with the largest decrease around
500 nm. It is argued that eumelanin black color is not only intrinsically defined by the overlap of π-electron
conjugated chromophores within the individual polymer components, as commonly believed, but also by
oxidation state- and aggregation-dependent interchromophoric interactions causing perturbations of the
heterogeneous ensemble of π-electron systems and overall spectral broadening.
Introduction
5,6-dihydroxyindole (DHI) and its 2-carboxylic acid (DHICA).
These are highly oxidizable catechol-containing indole deriva-
Although black is a common property in the inorganic world,
among living organisms it seems to be associated mainly with
eumelanins, the peculiar class of biopolymers of human skin,
hair, eyes, and brain.1-4 Eumelanins can rightly be regarded
among the most enigmatic pigments found in Nature because
of their high molecular heterogeneity, amorphous and particulate
character, and complete insolubility in all solvents, which have
so far hindered detailed structural elucidation.5,6 Current models
point to a hierarchical architecture of eumelanin particles7-9
consisting of π-stacked layered molecular systems building on
oligomeric motifs derived from oxidation of two key precursors,
tives which are biosynthesized in epidermal pigment cells, the
melanocytes, by the tyrosinase-catalyzed oxidation of tyrosine.10
Model chemical studies have shown that DHI undergoes
oxidative coupling mainly via 2,4′- and 2,7′-bondings to give
dimers and complex mixtures of oligomers thorough different
mechanisms.11-15 How these processes and ensuing supramo-
lecular interactions participate in eumelanin buildup remains
obscure.
An unresolved and related topic to this issue concerns the
black color of eumelanins, which is atypical of organic
(1) Prota, G. Melanins and Melanogenesis; Academic Press: San Diego,
CA, 1992.
(10) Land, E. J.; Ramsden, C. A.; Riley, P. A. Acc. Chem. Res. 2003, 36,
300–308.
(2) Ito, S. Pigment Cell Res 2003, 16, 230–236.
(3) Seagle, B.-L. L.; Rezai, K. A.; Kobori, Y.; Gasyna, E. M.; Rezaei,
K. A.; Norris, J. R., Jr. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 8978–
8983.
(11) d’Ischia, M.; Napolitano, A.; Pezzella, A.; Land, E. J.; Ramsden, C. A.;
Riley, P. A. AdV. Heterocycl. Chem. 2005, 89, 1–63.
(12) Pezzella, A.; Panzella, L.; Crescenzi, O.; Napolitano, A.; Navaratnam,
S.; Edge, R.; Land, E. J.; Barone, V.; d’Ischia, M. J. Am. Chem. Soc.
2006, 128, 15490–15498.
(4) Zecca, L.; et al. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17567-
17572.
(13) Panzella, L.; Pezzella, A.; Napolitano, A.; d’Ischia, M. Org. Lett. 2007,
9, 1411–1414.
(5) Simon, J. D.; Ito, S. Pigment Cell Res. 2004, 17, 422–424.
(6) Meredith, P.; Sarna, T. Pigment Cell Res. 2006, 19, 572–594.
(7) Clancy, C. M. R.; Simon, J. D. Biochemistry 2001, 40, 13353–13360.
(8) Liu, Y.; Simon, J. D. Pigment Cell Res. 2003, 16, 606–618.
(9) Liu, Y.; Simon, J. D. Pigment Cell Res. 2005, 18, 42–48.
(14) Pezzella, A.; Panzella, L.; Natangelo, A.; Arzillo, M.; Napolitano, A.;
d’Ischia, M. J. Org. Chem. 2007, 72, 9225–9230.
(15) Okuda, H.; Wakamatsu, K.; Ito, S.; Sota, T. J. Phys. Chem. A 2008,
112, 11213–11222.
9
15270 J. AM. CHEM. SOC. 2009, 131, 15270–15275
10.1021/ja905162s CCC: $40.75  2009 American Chemical Society

Disentangling Eumelanin “Black Chromophore”
A R T I C L E S
diacetoxyindole (0.587 g, 2.5 mmol), and N-bromosuccinimide (448
mg, 2.7 mmol) was dissolved in acetonitrile (60 mL) under an argon
atmosphere. The mixture was kept under stirring for 4 h at 60 °C,
cooled to rt, and diluted with dichloromethane. The mixture was
washed with water, and the aqueous phase re-extracted with
dichloromethane. The collected organic phases were dried over
anhydrous sodium sulfate and concentrated in Vacuo. The residue
was purified by silica gel flash-chromatography (eluant: petroleum
ether/ethyl acetate mixtures) to yield pure 1 acetyl derivative (1.5
g, 49% yield) High resolution ESI+MS found 618.1264 [M + Na]+,
calcd for C26H29NO13SNa 618.1257. Spectral data are listed in Table
S1, Supporting Information
chromophores and is commonly associated to a broadband
monotonic absorption throughout the entire UV/visible range.6
Due to large amounts of proteins and other impurities present
in natural pigments, studies of the optical properties of eumela-
nins have usually been carried out on commercially available
synthetic pigments made from tyrosine and partially solubilized
at pH 11.5.16-18 The lack of information about the fundamental
structural units of such synthetic pigments and the harsh alkaline
conditions favoring extensive structural degradation might,
however, affect interpretation of the experimental data and their
relevance to natural eumelanins.
Isolation of Dimers 2 and 3 (As Acetylated Derivatives). A
solution of 1-Ac (60 mg, 0.10 mmol) in MeOH (2 mL) was treated
with sodium t-butoxide under a nitrogen atmosphere for 1 min to
obtain complete deacetylation (MS evidence) and then diluted with
0.1 M phosphate buffer (pH 7.0, 10 mL) and treated with HRP (77
U/mL) and H2O2 (36 µL of a 30% solution, 0.35 mmol). After
15 s reaction time, the oxidation was halted by addition of sodium
dithionite in water up to a 50 mM final concentration and
lyophilized. The powder was acetylated and then extracted with
10% ammonium chloride. The ethyl acetate extractable fraction was
fractionated by thin layer chromatography (CHCl3 sAcOEt 65/
35) to afford 2-Ac (Rf )0.4, 5 mg, 10% yield, High resolution
ESI(+)-MS found 1211.2468 [M+Na]+, calcd for C52H56N2O26S2
Na 1211.2460) and 3-Ac (Rf 0.3, 4 mg, 8% yield, high resolution
ESI(+)-MS found 1211.2455 [M+Na]+, calcd for C52H56N2O26S2
Na 1211.2460) and. Spectral data are reported in Tables S2 and
S3, Supporting Information.
Spectrophotometric Studies. Oxidation of 1 in 0.1 M phosphate
buffer pH 7.0 was carried out with the substrate at 12.5 mM using
HRP (25 U/mL) and H2O2 (12.5 mM), and after 20 min the solution
was diluted 25-100 fold with 0.1 M phosphate buffer pH 7.0; the
spectrum was recorded in a thermostatted cell at 298 K with or
without filtration through a membrane with 0.47 µm pores. When
necessary, the mixture was treated with a solution of NaBH4 in
water (20 mM). DHI was also used at 12.5 mM, and after 20 min
spectra were taken on the solution after separation of the pigment.
Although in the aggregated form both scattering and pure
molecular absorption effects have been suggested to cause the
black appearance of eumelanins,19 a substantial body of evidence
suggests that the monotonically decreasing spectrum is in fact
due to “real absorption”, possibly by a multichromophoric
heterogeneous ensemble of species of different chemical struc-
tures.20
In this context, the crucial gap concerns the absorption
properties of the oligomer constituents, and how they vary with
molecular size and redox state. While it is commonly agreed
that the oxidized forms of DHI oligomers have red-shifted
HOMO-LUMO gaps, no definite evidence for the existence
of stable visible light-absorbing chromophores derived from DHI
is so far available, apart from pulse radiolysis and computational
investigations of labile transient species.12 In fact, no stable
colored DHI-based chromophore has been so far isolated or
created, an issue of the utmost relevance to the key question of
why eumelanin is black. Further progress in this area depends
therefore on the availability of reliable, structurally defined, and
water-soluble model polymers, which would not only yield
useful information as to one of the supposed crucial functions
of eumelanins, namely photoprotection, but may also inspire
solid state physicists and materials scientists pursuing soft
biocompatible functional materials.21,22 The information gained
in aqueous media may moreover effectively integrate recent data
from organic-soluble eumelanin-like pigments from DOPA23
or DHICA benzyl and octyl esters.24
Results
Synthesis of Glycated 5,6-Dihydroxyindole Monomer. Gly-
cosylation processes have been selected by Nature as a highly
effective means of solubilizing organic compounds in water to
allow for their excretion. Capitalizing on a recently reported
procedure for installing sulfur-containing functionalities onto a
5,6-dihydroxyindole ring,25 we have thus focused on the
synthesis of 5,6-dihydroxy-3-indolyl-1-thio-ꢀ-D-galactopyrano-
side (1) as a candidate model system for preparation of a water-
soluble eumelanin (Figure 1).
Herein, we report the first water-soluble eumelanin-type
polymer from DHI as a unique investigative tool to inquire into
the origin of eumelanin black color and optical properties.
Materials and Methods
5,6-Diacetoxyindol-3-yl-1-thio-ꢀ-D-galactopyranoside (1-Ac).
Synthesis of the thiogalactosyl donor S-GalDonor is reported in
detail in SI. A mixture of S-GalDonor (1.31 g, 2.5 mmol), 5,6-
Thioglycosidation proved a viable route to indole function-
alization, owing to the chemical versatility of the sulfur
derivatives which can be subjected to a variety of transforma-
tions, thus opening several pathways for further postpolymer-
ization manipulations. Additionally, thioglycosidation at the
3-position, which is negligibly involved in the polymerization
process, was expected to cause lesser perturbations of the
monomer, oligomer, and polymer chromophores.
Because of the marked facility of the 5,6-dihydroxyindole
system to oxidation, synthetic strategies were targeted to the
acetylated derivative, 5,6-diacetoxyindole. The choice of acetyl
was favored over other catechol protecting groups because these
groups can be conveniently removed in situ just prior to
oxidative polymerization experiments.12-14
(16) Riesz, J. J.; Gilmore, J. B.; McKenzie, Ross, H.; Powell, B. J.;
Pederson, M. R.; Meredith, P. Phys. ReV. E.: Stat. Nonlinear Soft
Matter Phys. 2007, 76, 21915/1–21915/10.
(17) Meredith, P.; Riesz, J. Photochem. Photobiol. 2004, 79, 211–216.
(18) Capozzi, V.; Perna, G.; Carmone, P.; Gallone, A.; Lastella, M.;
Mezzenga, E.; Quartucci, G.; Ambrico, M.; Augelli, V.; Biagi, P. F.;
Ligonzo, T.; Minafra, A.; Schiavulli, L.; Pallara, M.; Cicero, R. Thin
Solid Films 2006, 511-512, 362–366.
(19) Nofsinger, J. B.; Weinert, E. E.; Simon, J. D. Biopolymers 2002, 67,
302–305.
(20) Tran, M. L.; Powell, B. J.; Meredith, P. Biophys. J. 2006, 90, 743–
752.
(21) Bothma, J. P.; de Boor, J.; Divakar, U.; Schwenn, P. E.; Meredith, P.
AdV. Mater. 2008, 20, 3539–3542.
(22) d’Ischia, M.; Napolitano, A.; Pezzella, A.; Meredith, P.; Sarna, T.
Angew. Chem., Int. Ed. 2009, 48, 3914–3921.
(23) Hatcher, L. Q.; Simon, J. D. Photochem. Photobiol. 2008, 84, 608–
612.
(24) Lawrie, K. J.; Meredith, P.; McGeary, R. P. Photochem. Photobiol.
2008, 84, 632–638.
(25) Pezzella, A.; Palma, A.; Iadonisi, A.; Napolitano, A.; d’Ischia, M.
Tetrahedron Lett. 2007, 48, 3883–3886.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15271

A R T I C L E S
Pezzella et al.
Figure 1. Synthetic strategy for preparation of 1 as the acetyl derivative.
Figure 2. Eumelanin-like pigments produced by oxidation of 1 (left) and
DHI (right).
Unfortunately, several attempts to achieve thioglycosidation
of 5,6-diacetoxyindole at the 3-position by using a suitable soft
electrophilic agent proved unfruitful. An expedient solution to
this problem came, however, when the opportunity to generate
soft thioglycosidation agents by synthesizing selenophenyl
thioglycosides was considered.26 These derivatives could be
prepared by a straightforward sequence of three fast reactions
requiring a single chromatographic purification. Whereas direct
coupling of 5,6-diacetoxyindole with selenophenyl glycoside
was unproductive, activation with stoichiometric amounts of
N-bromosuccinimide (NBS) eventually resulted in the neat
generation of the desired conjugate 1-Ac in very satisfactory
yields (Figure 1).
Figure 3. Structures of the biindolyl intermediates isolated from the
oxidation of 1 after reduction and acetylation steps.
dynamic radius distribution centered at 100 nm suggestive of a
monodisperse high-molecular weight polymer (See Supporting
Information (SI)).
Direct mass spectrometric analysis on the soluble polymer
was little informative under a variety of conditions. However,
investigation of the mixture after reduction and acetylation (see
SI) allowed clear identification of oligomers up to six units,
thus ruling out significant side processes (e.g., oxidative
breakdown, loss of glycosyl residues, etc.) during oxidative
polymerization.
To gain insight into the mode of coupling of 1 during polymer
buildup, which is important to match the basic structural features
of the soluble and insoluble DHI-derived polymers and to
validate the soluble eumelanin model, a milligram-scale reaction
was stopped in the early stages by the addition of sodium
dithionite to reduce oxidized species, and the crude oligomer-
containing mixture was acetylated according to an established
protocol.14 Thin layer chromatography of the mixture thus
obtained allowed the isolation of two main oligomer intermedi-
ates, which were identified as the acetylated 2,7′- and 2,4′-
biindolyls 2 and 3 (Figure 3).
Preparation and Chemical Characterization of the Model
Soluble Polymer. Treating the acetylated derivative of 1 with
methanolic sodium t-butoxide allowed for acetate removal and
the safe generation of the desired 1. This latter was subjected
to oxidation with the horseradish peroxidase (HRP)/H2O2
system, the most efficient oxidizing agent for 5,6-dihydroxy-
indole polymerization.14 When visually observed, the reaction
proceeded with the rapid development of a dark-brown colora-
tion, whose intensity depended on substrate concentration.
Notably, a completely soluble pigment was obtained from as
much as 12.5 mM 1 (Figure 2), where the well-known formation
of a precipitate is observed when the same treatment is
conducted on DHI.
Dynamic light-scattering measurements showed a profile
typical of a homogeneous solution devoid of scattering effects.
This was deduced from the absence of the typical oscillating
light intensity profile which appears when scattering units have
non-Brownian motion. Elaboration of data indicated a hydro-
The similarity of the isolated dimers with the main species
produced by oxidative coupling of DHI11 demonstrated that the
sulfur substituent on the 3-position of 1 did not affect the
oxidation behavior and the mode of coupling of the 5,6-
(26) Gamblin, D. P.; Garnier, P.; van Kasteren, S.; Oldham, N. J.; Fairbanks,
A. J.; David, B. G. Angew. Chem., Int. Ed. 2004, 43, 828–833.
9
15272 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Disentangling Eumelanin “Black Chromophore”
A R T I C L E S
Figure 4. Absorption spectra of the polymers produced by oxidation of
12.5 mM 1 or DHI (20 min reaction time), prior to (gray lines) and after
(black lines) reductive treatment with NaBH4. Shown is also the difference
spectrum (-×-) of the untreated and treated polymer from 1 (right axis).
dihydroxyindole system in the soluble polymer. This suggested
that the oligomeric soluble pigment constituents displayed
interunit bonding patterns that are similar to those involved in
eumelanin buildup, thus supporting the validity of the new
water-soluble model system for studies of eumelanin properties.
Absorption Properties of the Soluble Polymer. Figure 4 shows
the UV/visible spectrum of the dark solution obtained by
oxidation of 12.5 mM 1 with HRP/H2O2 in phosphate buffer,
pH 7.4. The spectrum was taken 20 min after the oxidant
addition, when no further spectrophotometric change was
apparent even after additional H2O2 treatment, indicating that
the melanization process was complete. Moreover, the absorp-
tion profile of oxidized mixtures from 1 did not show ap-
preciable changes with time over at least 24 h.
Data show that after oxidation was complete the chromophore
of 1 (λmax at 290 nm) was replaced by a broadband spectrum
with a well detectable maximum around 314 nm and an almost
featureless absorption throughout the visible region (380-780
nm) approaching the baseline at the low-energy end of the
spectrum. This absorption profile is not dissimilar from those
of the insoluble DHI polymers and of intact natural eumelanin
samples from Sepia27 and hair,19 in which however significant
scattering contributions are present. Given the complete lack
of scattering effects, the spectrum reported herein demonstrates
for the first time the generation of stable, visible-light-absorbing
oligomeric/polymeric chromophores by oxidation of a DHI
derivative in aqueous medium.
Careful reductive treatment with sodium borohydride caused
visual color discharge from dark brown to pale yellow.
Spectrophotometric analysis consistently indicated a marked
absorbance decrease in the visible region but no significant
variation of the extinction coefficient at the 314 nm UV peak
(Figure 4). For comparative purposes, the spectral changes
occurring during the oxidation of DHI at similar concentrations
under the same conditions were reported (Figure 4). The
resulting black pigment suspension exhibited a relatively poor
absorption in the UV and visible regions dominated by scattering
contributions. A weak absorption maximum around 314 nm was
apparent, resembling the spectrum of the soluble polymer.
Modest changes in the absorption intensity in the visible region
occurred upon treatment with sodium borohydride.
Figure 5. (a) Effect of dilution on the absorption properties of soluble
eumelanin. Gray line: 500 µM; black line: 30 µM. Concentrations are based
on starting monomer, plots are normalized against the absorbance at 314
nm. AU ) (normalized) absorbance units. Note that absorbance is on
logarithmic scale for the sake of clarity. (b) Normalized plots of the
absorbance ratios Ax/A314 at selected wavelengths (O 384 nm, ] 420 nm,
0 550 nm, 4 750 nm) as a function of dilution (for all wavelengths
normalization was obtained by dividing all absorbance ratios by the relevant
value at 150 µM concentration). (Inset) Plot of Ax/A314 ratio differences
(slopes) determined at 70 µΜ and 30 µM concentrations as a function of
wavelength. AU ) adimensional units.
spectrum, the extent of the absorbance variations as a function
of wavelength was determined. A plot of the absorbance
differences measured in the visible region after reductive
treatment (Figure 4) against wavelength gave a peak around
550 nm, suggesting a prevalent superposition of oxidized
chromophores in that region.
In further experiments the effects of dilution on the absorption
properties of the soluble pigment were investigated. For this
purpose, polymer concentrations were assumed as the starting
monomer concentration under conditions of complete oxidation.
Figure 5a compares the absorption spectra of polymer solutions
taken at 500 µM and 30 µM concentrations after proper dilution.
The graphs clearly show a decrease of the visible absorption
contribution relative to the 314 nm band at the lower concentra-
tion. Consistent with this observation, a separate plot of the
absorbance values at 314 nm (A314) against dilution over the
500 µM f30 µM concentration range indicated a linear
response, whereas a significant deviation from linearity was
apparent for absorbance values (Ax) measured in the visible
range.
This effect was better appreciated when absorbance intensity
ratios Ax/A314 at selected wavelengths were plotted as a function
of dilution (Figure 5b). Inspection of the curves revealed
significant wavelength-dependent changes that are much more
pronounced in the visible range. Noteworthy, comparison of
the Ax/A314 ratio changes after dilution revealed a maximum
variation around 500 nm (Figure 5b, inset). These results can
be taken to suggest the occurrence of reversible aggregation
To assess whether oxidized quinonoid substructures contrib-
uted homogeneously to the various regions of the visible
(27) Pezzella, A.; d’Ischia, M.; Napolitano, A.; Palumbo, A.; Prota, G.
Tetrahedron 1997, 53, 8281–8286.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15273

A R T I C L E S
Pezzella et al.
processes in solution, reinforcing specifically the absorption of
light above 500 nm.
It is difficult, based on present data, to draw any conclusion
about the average molecular weight of the polymer. Since each
molecular entity has apparently a globular shape in solution with
a diameter of ∼100 nm, and considering that the average
dimension of a single indole unit may be approximated to about
1 nm, it can be speculated that the polymer consists of chains
of at least 100 units. Indeed, mass spectrometry was not
conclusive in this regard, and thus, it is possible that the 100-
nm species are in fact molecular aggregates. Verification of this
latter point by dilution experiments and dynamic light-scattering
measurements proved to be difficult because of the high
extinction coefficient of the pigment preventing meaningful
scattering determination in a sufficiently broad concentration
range. However, the low monomer concentration used for
polymer analysis would entail an unusually high aggregation
constant for a disordered heterogeneous polymer, whereby
formation of 100 nm aggregates seems unlikely.
Several important observations regarding the chromophoric
features of the soluble polymer deserve a comment. A first
critical point is that even when melanization is complete, soluble
DHI oligomers/polymers show a well detectable and stable
maximum at 314 nm, resembling that of reduced DHI
oligomers.30,33 This spectral feature can be reasonably attributed
to oligomeric structural components in the reduced state, based
also on its insensitivity to reductive treatments. On the other
hand, the lack of effect of additional oxidant equivalents on
the absorption profile provides unambiguous demonstration that
reduced (catecholic) and oxidized (quinonoid) substructures
coexist in the soluble pigment polymer and that oxidation cannot
be forced beyond a certain threshold. This is a most important
conclusion that would confute models of eumelanin structure
based on hypothetical fully oxidized polyquinonoid architec-
tures.34
A second, most relevant observation, stems from the NaBH4
reduction experiments. These demonstrated that the visible light
absorption and, hence, the dark color of DHI polymers depend
on the presence of oxidized subunits. It was recently shown
that no visible chromophore can be generated from reduced
(catecholic) oligomers however extended the molecular size may
be.30 The reason lies in the peculiar mode of coupling of the
indole units which is often incompatible with linear or rod-
shaped oligomeric structures allowing fully extended π-electron
delocalization. Thus, significantly red-shifted HOMO-LUMO
gaps and visible light absorption can be achieved only by
oxidation. This concept was addressed in a number of previous
papers6,34-36 but could never be unambiguously demonstrated
in aqueous medium, since oxidized quinonoid oligomers are
too elusive to be identified as stable entities in solution. Our
data would provide the first direct experimental verification of
the existence in solution of stable, visible-light-absorbing
chromophores derived from DHI-related units.
Discussion
The origin of eumelanin black color has been the subject of
a long-lasting debate. While early views28 favored the presence
of low-lying conduction bands and/or charge-transfer interac-
tions resembling those in quinhydrones, subsequent concepts
were based on an amorphous semiconductor model29 or
emphasized the importance of scattering effects. These latter
comprise both wavelength-dependent Rayleigh scattering, caus-
ing the steep rise in the spectrum below 300 nm,30 and
wavelength-independent Mie scattering, superimposing a broad
background on the absorption spectrum.6
The design and realization of the first water-soluble DHI-
based eumelanin-like polymer was aimed to provide a means
of obviating the notorious impossibility to solubilize oligomer/
polymer eumelanin components without affecting basic struc-
tural units, and thus to gain a direct insight into the visible
chromophore generated during DHI polymerization. The results
reported herein indicate that these expectations have been met.
A first relevant point for discussion concerns the impact of
the bulky (and water-soluble) galactosyl moiety and the thio-
ether bridge at the 3-position on the mode and degree of
polymerization of the 5,6-dihydroxyindole unit. Although this
was in principle a possible limitation of this study, chemical
analysis showed in fact that 1 retains the typical mode of
polymerization of DHI via 2,4′- and 2,7′-bonds, indicating that
the sulfur substituent at C-3 exerts little or no chemical influence
on the adjacent reactive 2-position. As an additional remark,
although studies of DHI polymerization in vitro have shown
the involvement of the 3-position during the oxidative coupling
of dimers to tetramers,13 this is expected to be only one of the
many possible modes of chain elongation in the complex process
of polymer buildup, so that the use of a 3-substituted monomer
would not detract too much from the value of the soluble
polymer from 1 as a tool for structural investigations. It is
relevant to note, in this connection, that all known trimers from
DHI feature only 2,4′- and 2,7′-bondings,11 suggesting that chain
elongation pathways via sequential addition of the monomer to
dimers and higher oligomers should not involve the 3-position
to any significant extent. Whether 1-related dimers and oligo-
mers can give rise to soluble dark polymers that can be used as
model eumelanins is yet an important goal for future work and
requires further synthetic efforts aimed at preparing S-thioga-
lactosyl dimers in sufficient amounts for chemical polymeriza-
tion studies.
The above arguments would also imply that the marked
difference between the polymers from DHI and 1 is specifically
due to the sugar group ensuring solubilization of the growing
polymer and is not simply the consequence of inhibited reactivity
at the 3-position. This is also supported by literature reports
showing that 3-alkyl-5,6-dihydroxyindoles behave on oxidation
like other 5,6-dihydroxyindoles giving rise to 2-linked oligomers
and eventually to dark, insoluble eumelanin-like pigments.31,32
The third and perhaps most significant outcome of this study
relates to the nonlinear decrease of the visible components of
the chromophore with dilution. This points to the occurrence
of reversible aggregation phenomena in solution that may
(28) Blois, M. S., Jr. Physical studies of the melanins. In Biology of Normal
and Abnormal Melanocytes; Kawamura, T., Fitzpatrick, T. B., Seiji,
M., Eds.; University of Tokyo Press: Tokyo, 1971; pp 125-139.
(29) McGinness, J. E.; Corry, P.; Proctor, P. Science 1974, 183, 853–855.
(30) Riesz, J.; Gilmore, J.; Meredith, P. Biophys. J. 2006, 90, 4137–4144.
(31) Singh, S.; Jen, J. F.; Dryhurst, G. J. Org. Chem. 1990, 55, 1484–
1489.
(33) d’Ischia, M.; Crescenzi, O.; Pezzella, A.; Arzillo, M.; Panzella, L.;
Napolitano, A.; Barone, V. Photochem. Photobiol. 2008, 84, 600–
607.
(34) Stark, K. B.; Gallas, J. M.; Zajac, G. W.; Golab, J. T.; Gidanian, S.;
McIntire, T.; Farmer, P. J. J. Phys. Chem. B 2005, 109, 1970–1977.
(35) Stark, K. B.; Gallas, J. M.; Zajac, G. W.; Eisner, M.; Golab, J. T. J.
Phys. Chem. B 2003, 107, 3061–3067.
(32) Beer, R. J. S.; Broadhurst, T.; Robertson, A. J. Chem. Soc. 1954, 1947–
1953.
(36) Powell, B. J.; Baruah, T.; Bernstein, N.; Brake, K.; McKenzie, R. H.;
Meredith, P.; Pederson, M. R. J. Chem. Phys. 2004, 120, 8608–8615.
9
15274 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Disentangling Eumelanin “Black Chromophore”
A R T I C L E S
crucially affect eumelanin absorption features. Charge-transfer
or redox processes would be responsible for the marked
enhancement and broadening of the chromophores around and
above 500 nm with increasing polymer concentration. Aggrega-
tion has been reported to affect the optical properties of natural
and synthetic eumelanins by extending the spectrum to lower
energies.37 The spectral changes reported in this paper are in
agreement with this conclusion and would yield for the first
time some information on the basic features of the process in
solution under conditions independent of scattering effects.
In light of the new information, it is now possible to
address the key question of why eumelanin is black on the
basis of the chemical disorder model proposed by Meredith
and co-workers.20,38 This model envisages the broadband
monotonic absorption of eumelanin-type polymers as the
consequence of the superposition of a large number of inho-
mogeneously broadened Gaussian transitions associated with
each of the components of the eumelanin ensemble. There has
been much debate in the past about the effective length of
eumelanin oligomer/polymeric chains and hence about the
spatial distance over which electronic delocalization may occur.
Pulse radiolysis investigation of quinonoid species from DHI-
derived biindolyls12 showed that intense visible chromophores
can be generated even within the small boundaries of dimeric
scaffolds. Thus, overlapping individual chromophores may in
principle account for coverage of the entire UV-visible range
even at a low degree of polymerization. In this connection, it
has been calculated that as few as 11 individual species are
needed to cover the UV, visible, and near-infrared with realistic
inhomogeneous line broadening.6
irregular intervals, e.g. by hindered rotation around interunit
bonds forcing deviation from coplanarity,39 to form more or
less discrete chromophoric units. Furthermore, within each
oligomer/polymer chain the chromophoric entities may be
dynamically equilibrated, e.g. by tautomerism, thus giving rise
to chromophore mixing.
Conclusions
We have synthesized the first water-soluble 5,6-dihydroxy-
indole polymer to address the origin of the peculiar black color
of eumelanin-type polymers. The data support the notion that
eumelanin chromophores cannot be defined exclusively on an
intrinsic basis by the conjugation length of the isolated absorbing
species but can be defined also by the external perturbations of
conjugated chromophoric units caused by interactions between
the oxidized and reduced polymer chain domains. This would
entail a dynamic composition of intrinsically and extrinsically
defined light-absorbing species rather than a mixture of static
entities with absorption features intrinsically defined by the
degree of conjugation.
Although the relevance of these model studies to the actual
optical properties of natural eumelanins remains to be assessed,
it is likely that aggregation phenomena akin to those reported
in solution occur as the early events in the multistep process of
eumelanin particle buildup. When extrapolated to the solid state,
this concept would lead to an extreme scenario of structure-
property relationships dominated by emerging chromophoric
components largely extrinsic in nature.
Acknowledgment. This work was supported by a grant from
Italian Ministry of University (PRIN 2006 project). L.P. thanks
“L’Oreal Italia Per le Donne e la Scienza” for a research fellowship.
On the basis of the results of this paper, the chemical disorder
paradigm can now be integrated and elaborated into an improved
dynamic model emphasizing the importance of interchro-
mophoric interactions. The basic underpinning of the improved
model is that it is the coexistence of oxidized and reduced
substructures that is essential for the broadband visible-light
absorption of eumelanin-like polymers. In particular, our data
suggest that in solution eumelanin components can be thought
of as chains of extended π-electron systems interrupted at
Supporting Information Available: Material and methods and
experimental procedures; mono- and bidimensional NMR
spectra for compounds 1-Ac, 2-Ac, 3-Ac; MALDI MS spectrum
of the oxidation mixture of 1 (after reduction and acetylation
treatment); complete ref 4. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA905162S
(37) Nofsinger, J. B.; Forest, S. E.; Eibest, L. M.; Gold, K. A.; Simon,
J. D. Pigment Cell Res 2000, 13, 179–184.
(39) Pezzella, A.; Panzella, L.; Crescenzi, O.; Napolitano, A.; Navaratnam,
S.; Edge, R.; Land, E. J.; Barone, V.; d’Ischia, M. J. Org. Chem. 2009,
74, 3727–3734.
(38) Meredith, P.; Powell, B. J.; Riesz, J.; Nighswander-Rempel, S. P.;
Pederson, M. R.; Moore, E. G. Soft Matter 2006, 2, 37–44.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15275