Spectroscopic Simulation from Recent Structural Models of Eumelanin
Researchers at Innovene, the University of San Antonio, the University of Houston, and Accelrys have used the DMol3 and VAMP codes to generate energy-minimized structures of eumelanin molecular models that provide the electronic structure and subsequent optical absorption spectra.
Agreement with experimental results offers additional support to recent eumelanin models and also provides new predictions on which to base further experimentation and understanding of the melanin structure.
The melanins are a widely occurring class of biological pigments that provide coloration to animals and plants. Eumelanins, black and dark-brown nitrogen containing pigments, form a sub-class of the melanins and are also considered to be photoprotective, as evidenced by an optical absorption spectrum which matches the action spectrum for tissue damage.
Unfortunately, an adequate explanation of the eumelanin absorption spectrum has not been available - largely because the eumelanin structure has been essentially unknown. Recently, however, a new representation for a structural model has been proposed that is based upon aggregation of oligomeric units, which promises to adequately describe the optical absorption spectrum of eumelanins.1-6
In this work,7,8 scientists at Innovene, the Universities of San Antonio and Houston, and Accelrys, used Accelrys' Density Functional Theory (DFT) code DMol3 to provide optimized structures of these sub-units of eumelanin. From the optimized structures, absorption spectra were calculated using Accelrys' semi-empirical program VAMP. Combined theoretical absorption spectra were then compared with experimental results.
Three different monomers, which are believed to form subunits of the eumelanin biopolymer, were optimized, namely indolequinone (IQ), semiquinone (SQ), and hydroquinone (HQ). These molecules are shown in Fig. 1.
Fig. 1 From left to right: Structures of 5,6-indolquinone (IQ), hydroquinone (HQ), and semiquinone (SQ).
Carbon atoms are displayed in gray, oxygen atoms in red, nitrogen atoms in blue, and hydrogen atoms in white.
The DFT results agree very well with earlier ab initio calculations,9 albeit at a fractional computation time. Electronic transitions were obtained by the PECI method in VAMP applied to the DFT-optimized structures. The λmax values for the IQ, HQ, and SQ monomers, shown in Table 1, compare well with the ones obtained by Bolivar-Martinez et al.9
Table 1 Computed λmax values (nm) for the three monomers shown in Fig. 1. BGC values from ref. 9.
Among the five dimer configurations examined the entirely planar structure (Fig. 2) was found to have the lowest energy, and therefore the most stable. Physically, this is in accord with the stabilization afforded by the hydrogen bond that is formed between the hydrogen associated with the indole moiety and the adjacent carbonyl.
Fig. 2 Most stable dimer structure. The structure is stabilized by a hydrogen bond formed between adjacent O and H atoms of fused rings.
This dimer shows a strong bathochromic shift in absorption (λmax = 320 nm). The relatively planar structure associated with this particular dimer corresponds to more extensive delocalization, which, in turn, produces lower energy transitions. For all dimers under consideration, correlations between HOMO-LUMO gaps and torsion angles across the dimer bond as well as the maximum absorption wavelength have been found. That means, the more stable the dimer is, the smaller is its gap, the more planar is the structure and the higher is the red shift in its spectrum.
If one extends this work to higher oligomers8 (tetramers-hexamers), the absorption spectra are further red-shifted into the visible region of the spectrum. Broadenend absorption spectra of the oligomers together with the optimized structure of one of these oligomers (pentamer) are shown in Fig. 3.
It is important to note that for each of the simulated spectra of Fig. 3, the optical peaks predicted a) span the UV, visible, and near IR; and b) the peaks increase progressively in intensity as the wavelengths decrease. These are exactly the features the experimental absorption spectrum for melanin exhibits. The calculation for the oligomer spectra yield relatively discrete optical bands. This is in contrast to the observed smooth, structureless optical absorption for melanins (Fig. 4, left). However, the observed optical absorption for melanin represents a very large number of polymeric units and there is also much experimental evidence that melanin is very heterogeneous in its polymerization. Indeed, the sum of the calculated spectra for the tetramer, pentamer, and hexamer (Fig. 4, right) demonstrates a smoothing in the spectra relative to the individual non-superposed spectra and resembles the experimental spectrum very closely.
Fig. 4 Experimental Melanin absorption spectrum (left) and calculated spectrum (right) obtained by linear combination of the oligomer spectra of Fig. 3.
The encouraging results suggest work that should further test the molecular models. More detailed MO calculations should be done that include the amine and carboxyl derivatives that are known to further specify the general quinone-like models used in the present work. Further, various superpositions of spectra corresponding to standard molecular weight distributions for free radical polymerizations should be modeled; and even direct experimental input derived from STM and Mass Spectroscopy analysis of melanins would be useful. Finally, attempts should be made to simulate hydrogen-peroxide bleached melanins based upon appropriately modified chemical structure models.
In fact, the results for single melanin sheets have not yet involved spectra associated with stacking and may be more appropriate for bleached or pheomelanin. Such stacking will lead to further broadening and lower energy transitions, which will be reflected in greater absorption in the near IR. Modeling of such spectra may be of value in light of a recent proposal that skin cancer may be related to photo-induced reactive oxygen species produced preferentially by un-stacked oligomeric sheets.
(1) Cheng, J.; Moss, S. C.; Eisner, M.; Zschack, P. Pigment Cell Res. - Part I, 1994, 7, 255-262.
(2) Cheng, J., Moss, S.C. Moss; Eisner, M. Pigment Cell Res. - Part II, 1994, 7:263-273.
(3) Zajac, G. W.; Gallas, J. M.; Cheng, J.; Eisner, M.; Moss, S. C.; Alvarado-Swaisgood, A. E. (1994) Biochim. Biophys. Acta 1994, 1199, 271.
(4) Gallas, J. M.; Zajac, G. W.; Sarna, T.; Stotter, P. L. Pigment Cell Res. 2000, 13, 99.
(5) Zajac, G. W.; Gallas, J. M.; Alvarado-Swaisgood, A. E. J. Vacuum Sci. Technol. B 1994, 12, 1512.
(6) Gallas, J. M.; Littrell, K. C.; Seifert, S.; Zajac, G. W.; Thiyagarajan, P. Biophys J. 1999, 77, 1135.
(7) Stark, K. B.; Gallas, J.M., Zajac, G. W. ; Eisner, M.; Golab, T. J.; J. Chem. Phys. B 2003, 107(13), 3061.
(8) Stark, K. B.; Gallas, J.M., Zajac, G. W. ; Eisner, M.; Golab, T. J.; J. Chem. Phys. B 2003, 107(41), 11558.
(9) Bolivar-Martinez, L. E.; Galvao, D. S.;Caldas, M. J., J. Phys. Chem. B 1999, 103, 2993.