Translator Disclaimer
25 June 2015 Raman spectroscopy of human skin: looking for a quantitative algorithm to reliably estimate human age
Author Affiliations +
Abstract
The possibility of examining soft tissues by Raman spectroscopy is challenged in an attempt to probe human age for the changes in biochemical composition of skin that accompany aging. We present a proof-of-concept report for explicating the biophysical links between vibrational characteristics and the specific compositional and chemical changes associated with aging. The actual existence of such links is then phenomenologically proved. In an attempt to foster the basics for a quantitative use of Raman spectroscopy in assessing aging from human skin samples, a precise spectral deconvolution is performed as a function of donors’ ages on five cadaveric samples, which emphasizes the physical significance and the morphological modifications of the Raman bands. The outputs suggest the presence of spectral markers for age identification from skin samples. Some of them appeared as authentic “biological clocks” for the apparent exactness with which they are related to age. Our spectroscopic approach yields clear compositional information of protein folding and crystallization of lipid structures, which can lead to a precise identification of age from infants to adults. Once statistically validated, these parameters might be used to link vibrational aspects at the molecular scale for practical forensic purposes.

1.

Introduction

In the vibrational spectra of biological molecules, indelible “fingerprints” can be found for the occurrence (and frequency) of a number of otherwise hidden natural phenomena. The discovery and rationalization of such spectral “fingerprints” represents a socially useful and fascinating practice, but also pose formidable challenges to spectroscopists. Unlike simpler inorganic structures, an explicit description of the irreducible phonon representation for complicated organic molecules is a task of enormous cumbersomeness, which can by no means be accomplished without the help of computational programs. Even an “elementary” protein structure possesses a number of distinct vibrational modes (i.e., in the order of 104),1 which makes the task of unfolding such a structural complexity one of the hardest challenges in modern science. On the positive side, however, there is the high sensitivity of vibrational spectroscopy to structural features. This method can be exceptionally sensitive to even quite small variations of bond strength, i.e., in the order of 0.02%.2 Such variations can promptly be resolved by probing with a spectral resolution better than 5cm1, and modern spectroscopes usually possess spectral resolutions at least 1 order of magnitude better than the above threshold. In other words, Raman spectroscopy provides us with a suitably high sensitivity for monitoring subtle bond modifications and molecular distortions with high accuracy.

From an analytical point of view, the main consequence of the complexity of biomolecules has been that Raman spectra have often been treated as a mere matricial sequence of mathematical data, with little emphasis being placed on the interpretation of the physical origin of individual bands in the recorded spectra. Such a statistically driven approach could be quite useful in locating spectral differences. However, despite the huge piece of information potentially contained in the Raman spectra, such spectroscopic “cryptograms” have, most of the time, remained unfolded with only a small (qualitative) part of their contents being physically interpreted. On the other hand, several research groups have continued to intensely work on basic interpretations of the Raman spectra emitted from biological samples.39 Such basic approaches have enabled clarifying a number of structural features in organic molecules, including the interpretation of Raman bands and their spectral positions from saccharides,5 lipids3,4,10 and carotenoids.9 The presence of mono-unsaturated acyl groups has also been related to the first derivative value of scattering intensities measured at 1656cm1, which relate to the vibrational stretching of cis and trans CC bonds in lipids.10 Moreover, classification overviews of Raman spectra for biomolecules have recently become available.11 Specifically regarding model protein structures, Raman bands represent vibrational modes of both the peptide backbone and its side chains. Spectral positions, intensities and polarizations of the Raman bands result in being quite sensitive to protein secondary, tertiary and quaternary structures, in addition to side-chain orientations and local environments. In a number of favorable cases, the Raman spectrum has provided a straightforward signature of the protein three-dimensional (3-D) structure, intramolecular dynamics and intermolecular interactions.12 Specifically related to skin, its constituent organic molecules generally display in the Raman spectrum according to the corresponding model structures, but significant compositional and structural variations can be expected with progressing age. Collagen fibers are composed of collagen types I and III in a ratio that depends on their location being either in the papillary or in the reticular dermis. In young people, collagen fibers in the region of papillary dermis take the form of densely packed and irregularly arranged networks, whereas in the reticular dermis, their morphology appears coarser and with loosely arranged and intertwined bundles. Upon aging, the amount of collagen fibers increases, packing becomes denser, and the stereological arrangement is less twisted.13,14 Moreover, elastic fibers consist of an amorphous fraction (90% of the mature fiber), which is exclusively composed of elastin and a microfibrillar component consisting of nanostructured fibrils.15 Superficial microfibril bundles gradually thicken and merge with increasingly large amounts of amorphous elastin as the papillary dermis changes into reticular dermis. However, with increasing age, the concentration of elastin fibers in the papillary dermis decreases. Accordingly, fibers from the skin of older individuals lose some of their elasticity, thicken (fold) into agglutinated masses and unravel.14 While the anatomical details of structural skin modifications with aging are well studied and documented, methods for finding such features in the recorded Raman spectra are yet at their early development. Consequences of aging processes are obviously the changes in the biochemical structure of tissue, which should also be seen in its Raman spectrum. The challenge, therefore, shifts now to how to translate such qualitative notions into a quantitative spectroscopic algorithm capable of assessing human age to a degree of precision.

Building upon previous outputs of different approaches to Raman evaluations of biological samples, we challenged, in this paper, the establishment of parameters for the quantitative assessment of human age from skin samples belonging to cadaveric donors. This study was mainly motivated by the need in the field of forensic pathology to determine within an improved degree of accuracy the age of human subjects lacking specific identity information. Spectroscopic outputs were rationalized and discussed in terms of selected parameters, including the degree of protein folding and the degree of lipid crystallization. Some other parameters—such as the fractional ratio between α-helix and β-sheet, the presence of sphingomyelin in the ceramide structure, and the content of collagen versus lipids—were also noticed. It should be stated at the outset of our awareness that this study lacks statistical relevance, which has been a direct consequence of the cadaveric origin of the studied samples and the need to examine the skin sample within a narrow interval of time (i.e., <1 week) since the date of decease and the necessity of preliminary clearing up ethical issues with respect to donors. Nevertheless, specific care was taken in obtaining “standard” spectra for each donor with averaging on a large number of acquisitions on each sample. In other words, the validity of the shown concepts relies on the basic assumption that if an age representing parameter (i.e., a “natural clock”) actually exists in the vibrational behavior of skin tissue, this should be independent of individual classes. The practical possibility of retrieving age information from skin samples is a confirmation that sensitive and selective “fingerprints” of natural aging exist in the intermolecular interactions and dynamics of the constituent phases of human skin. Further studies are presently ongoing for obtaining a statistical validation of the proposed parameters and procedures.

2.

Experimental Procedures

A series of cadaveric skin samples was obtained from donors (human patients deceased at different ages spanning from a few months to 62 years old) upon preliminary clearance of ethical procedures at the Department of Forensic Medicine of the Graduate School of Medical Science of Kyoto Prefectural University of Medicine. Details of the cadaver samples available for this study are provided in Table 1. The samples were taken from abdominal locations in the body unexposed to solar irradiation. Skin samples were typically 10×10mm2 wide (cut by hand) and encompassed the full thickness of the skin structure, from the stratum corneum to the hypodermis. All the skin samples investigated in this study belonged to the same race, showing similar color (pale white skin, with some yellow tones) on the skin surface. No special medical treatments were performed on the skin samples before and after they were extracted from the human bodies. More importantly, all the donors examined in this study died in a healthy situation and their pathological records did not show specific items either related to liver or other diseases specifically impacting on protein and lipid structures. Neither the infant nor other donors received medical therapy that might have induced changes in the molecular structure of their skin. Raman experiments were conducted on the skin samples within less than 1 week after the date of the patient’s decease. The skin samples were preserved at 70°C in a freezer before Raman analysis. Immediately before the Raman experiments, skin samples were thawed and kept on ice. Prior to Raman spectroscopic characterizations, each sample was divided into smaller specimens and hematoxylin and eosin (H&E) stain histology and optical microscopy were applied on the cross sections of one specimen for each sample in order to preliminarily assess the location of different zones of epidermis and dermis from morphological features. As a spectroscopic reference, a sample of skin type I collagen was also investigated, which was purchased from Sigma-Aldrich, Co.16

Table 1

Information on the human cadaver samples investigated in this study.

SampleAgeSexCause of death
13 monthsFemaleAspiration asphyxia
215 yearsMaleHemorrhagic shock
317 yearsMaleDrowning
435 yearsMaleDrowning
561 yearsMaleCardiac rupture

All the Raman spectroscopic experiments described in this paper were carried out in backscattering optical probe configuration with using a triple monochromator (T-64000, Horiba/Jobin-Yvon, Kyoto, Japan) equipped with a liquid nitrogen-cooled charge coupled device. The excitation source in the present experiments used a 532-nm Nd:YVO4 diode-pumped solid-state laser (SOC JUNO, Showa Optronics Co. Ltd., Tokyo, Japan) operating with a power of 200 mW. An objective lens with a numerical aperture of 0.5 was used both to focus the laser beam on the sample surface and to collect the scattered Raman light. All the experiments described in this paper were conducted with a pinhole aperture of 100μm and by employing an objective lens with a magnification of 100×. All the experiments were conducted at room temperature with a relative humidity of 68%. Each spectrum at a measurement location was collected for five scans with the accumulation time of each scan being 1 min. For each studied sample, several tens of spectra were collected on the top skin surface and an average spectrum could be obtained with a statistical validity using commercial software (LabSpec Ver. 4.02, Jobin-Ivon/Horiba, Tokyo, Japan). On the cross section of each skin sample, spectral line scans were also performed, starting from the stratum corneum to 800 μm deep in the depth direction. Spectral Raman lines were analyzed using a commercially available software package (Origin 9.1, OriginLab Co., Northampton, Massachusetts, United States). Fitting was performed according to Gaussian-Lorentzian functions after subtracting the baseline.

3.

Experimental Results

3.1.

Labeling the Raman Spectrum of Human Skin

Similar to the case of other soft tissues, the Raman spectrum of skin is dominated by the vibrational bands of its structural proteins, amino acids and lipids. Figure 1 shows a Raman spectrum detected by our microprobe equipment with focusing on the stratum corneum of a sample (top-view spectrum) from a 3-month-old donor. Raman spectra from the healthy skin of infants are seldom found in the published literature. For this reason, we have considered this spectrum as a “reference” one since it was almost unaffected by environmental effects. In this study, it is used for a preliminary screening of the emitted Raman bands and to label them according to the published literature.

Fig. 1

Raman spectrum as detected with focusing on the stratum corneum of a sample (top-view spectrum) belonging to a 3-month-old infant donor. The spectrum has been arbitrarily divided into low frequency and high frequency zones of investigation. The labeled bands are assigned in Table 2.

JBO_20_6_065008_f001.png

The Raman spectrum in Fig. 1 has arbitrarily been divided into a low-frequency zone (250 to 1800cm1) and a high-frequency zone (2800 to 3200cm1). In these two spectral zones, a total of 20 bands could be distinctly observed, as labeled in Fig. 1 (bands 1 to 20). Table 2 summarizes the observed Raman bands and their physical origin in the two noticed spectral zones reported in literature.1761 Bands associated with vibrations of amide bonds in polypeptide chains were observed, with the amide I emission being dominated by CO stretching vibrations and the amide III band by CN stretching and NH bending vibrations. The former emission (band 13) at 1652cm1 (here often observed, in agreement with other authors, at a slightly higher frequency of 1657cm1)23,31,62 is typical of mammalian keratins with mainly α-helical conformation.18,19,63 On the other hand, the latter emission presented two maxima: one at 1271cm1 (band 10, assigned to nonpolar fragments with high proline content forming collagen triple helix) and the other at 1244cm1 (overlapped to band 8 already assigned to tryptophan and phenylalanine) from polar fragments of collagen characterized by a low proline content.24,31,32 Band 9, which appears as a low-frequency shoulder to the 1271cm1 band 10, is most likely an overlap of vibrational modes from the adenine and cytosine belonging to the β-sheet structure of amide III (reported at 1258cm1)33 and lipids (reported at 1255cm1).34 The strong emission detected at 1440cm1 (band 11) and the weaker but clearly detectable band at 1750cm1 (band 14) can both be assigned to vibrational modes in lipids (with a contribution from proteins, especially in the former band).1719,26,32,64 In particular, the former band arises from CH2 scissoring and CH3 bending,35 whereas the latter one mainly represents the CO stretching mode in lipids and phospholipid molecules.36,37

Table 2

Assignment of Raman bands in skin and skin collagen samples: v, stretching mode; vs, symmetric stretch; vas, asymmetric stretch; δ, bending mode; and w, wagging mode.

Band labelBand position (cm−1)Principal assignmentReferences
Band 1510v(SS) in cysteine1718.19
Band 2745Symmetric breathing of tryptophan21,22
Band 3877v(CC) in hydroxyproline, δ(tryptophan ring)18,20,232425
Band 4919v(CC) of proline rings17,18,26,27
Band 51003v(CC) in phenylalanine17,18,26,27
Band 61081v(CO32) and v(PO43), v(CC) in lipids22,24
Band 71157Sphingomyelin28
Band 81208v(CC6H5) in tryptophan and phenylalanine29,30
Band A1210v(CC6H5) in tyrosine and phenylalanine35
Band A1218Amide III, v(CN) and δ(NH) in α-helix36,4546.47.48.49
Band 1*1242Amide III, v(CN) and w(CH2) in collagen65
Band 91255, 1258Amide III, adenine and cytosine β-sheet structure + lipids33,34
Band 101271Nonpolar fragments of proline in α-helix24,31,32
Band 3*1315Amide III, δ(CH2) in α-helix collagen65
Band B1338Amide III, v(CN) and δ(NH) in hydrated α-helix37
Band C1370Ring and v(CN) in cytosine and guanine44
Band D1395Symmetric δ(CH3) of the methyl groups of proteins50
Band 11A1428Proteins + lipids, CH2 scissoring5657.58
Band 11B1450δ(CH) in proteins + lipids5657.58
Band 11C1467δ(CH2) in proteins + lipids5657.58
Band 121526Sphingomyelin28
Band E1520v(CN) in adenine and cytosine in α-helix51,52
Band F1551Amide II, v(CN) and δ(NH)53,54
Band G1584δ(CC) in phenylalanine44
Band H1605v(CC) in phenylalanine and tyrosine41,55
Band I1638Amide I, v(CO) in α-helix + β-sheet39,56
Band 131652Amide I, v(CO) in α-helix23,31,62
Band J1681Amide I, v(CO) in disordered structure40,56
Band 141750v(CO) in lipids and phospholipids36,37
Band 152855vs(CH3) in lipids (liquid state)38
Band 162877vs(CH3) in lipids (hexagonal)60,61
Band 162881vs(CH3) in lipids (orthorhombic)56
Band L2910v(CH2) and v(CH3) in cholesterol and phospholipids59
Band 172928vs(CH3) in proteins42
Band 17+2950vas(CH3) in proteins42
Band 182956vas(CH3) in proteins42
Band 18+2980v(CHα,α)41
Band 193010vas(CH) in lipids and unsaturated acids40,43

In the high-frequency zone, a broad overlapping emission, contributed by at least five relatively strong bands (bands 15 to 19), could be detected, in addition to a rather weak and isolated band centered at higher frequencies (band 20). In this broad emission, the sub-band labeled as band 15 was centered at around 2855cm1 and represented the symmetric stretching of lipids in the liquid state.38 However, band 15 is likely to be overlapped by the CH symmetric stretching band of collagen centered at 2850cm1.39 The strongest sub-band in this spectral zone was centered at 2885cm1 (band 16). This band has been assigned to symmetric stretching in lipids.38,40 However, band 16 seemed to show a low-frequency shoulder at around 2881cm1, also of protein origin, which we did not explicitly label here. This sub-band shoulder was related to CH symmetric stretching in collagen.41 Overlapping effects between lipids and collagen will be discussed in more detail in the following sections. Bands 17 and 18 are seen as two consecutive shoulder sub-bands to the overall emission toward higher frequencies. The origin of both these bands, which are located at 2928 and 2956cm1, respectively, is probably also a composite one with components from both lipids and collagen. However, the former band corresponds to CH3 symmetric stretching (i.e., due primarily to proteins),66 whereas the latter one hits a frequency related to their CH3 asymmetric stretching.42 Band 19 is weaker than the other bands composing the overall high-frequency Raman emission, but clearly appears as a more separate sub-band. Its location is at 3010cm1, which corresponds to asymmetric stretching of CH groups in lipids, fatty and unsaturated acids.40,43 Finally, a quite weak but resolvable band was observed at 3140cm1. We labeled it as band 20, and it was tentatively assigned to NH symmetric stretching.44

It has been reported that in tissue, the spectral range from 3200 to 3600cm1 is occupied by a broad band peaking at 3250cm1, which is associated with OH stretching vibrations of tissue-bound water and NH stretching vibrations of proteins.17,18,32 Conversely, the presence of unbound water (i.e., tetrahedral water clusters) in skin is represented by a band located at 180cm1 in the Raman spectrum.67 The status of hydration of skin has been estimated through the intensity ratio between the protein stretching band at 2938cm1 (which we observed at 2928cm1) and the water stretching band at 3250cm1.67 Raman spectra from skin samples before and after sunlight exposure revealed a total fraction of water higher by 30% in the latter sample as compared to the former one.17 We also observed water-related features in our skin samples. However, although it has been recognized that the water content in skin generally increases with age,68 we ruled out a priori the possibility of using the hydration ratio as a meaningful parameter for age assessments because of the strong environmental effects on the hydration state of skin and the fact that each skin layer might show a different hydration level.69,70 We failed in observing a reliable trend in hydration levels independent of environmental effects and depending on age, probably also because the Raman probe reached different structures while in-depth penetrating samples from donors with different ages, thus giving different intensity ratios. On the other hand, an important hint in this work was that skin-aging processes involve conformational changes in structural proteins. In the Raman spectra of skin samples from older individuals and/or of skin exposed to sunlight, the maxima of the amide I and III bands were systematically detected at spectral positions shifted toward lower frequencies as compared to the spectrum of young and/or unexposed skin samples. Moreover, reduced intensities were generally found and a shift occurred of the CH2 stretching band in the aliphatic side chains of amino acids toward lower frequencies. This latter spectral perturbation was interpreted as the consequence of structural changes in protein folding.70,71

3.2.

Average Raman Spectra as a Function of Donors’ Age

Prior to Raman spectroscopic characterizations, H&E stain histology and optical microscopy were applied on the cross sections of all samples in order to preliminarily assess the location of different zones of epidermis and dermis from morphological features. Figure 2 shows a micrograph of the investigated histological section with the H&E stain of the sample from the 3-month-old infant donor. Labels show the protocol for the line-scan Raman spectroscopy characterization in various regions of the skin at increasing depths. Similar line scans were performed on all the investigated samples from donors of different age.

Fig. 2

Skin cross-sectional sample belonging to the 3-month-old donor with H&E stain. Labels illustrate the procedure of Raman microprobe line scan along the in-depth abscissa, z, with origin at the free surface of the sample.

JBO_20_6_065008_f002.png

In this section, we compare typical Raman spectra collected as a function of patient age from the stratum corneum (i.e., at depth from the sample surface, z=0μm), from a deeper zone in the epidermis zone (just below the stratum corneum, z=100μm; simply referred to as “epidermis,” henceforth) and from a zone further in depth, which was thought to preponderantly be part of the dermis in all the investigated samples (z=700μm). It should be noted at the outset that the thickness of various zones along the depth of skin significantly varies with location in the body and with age. We have minimized the former difference by sampling always from the same part of the donors’ body (abdomen). However, the latter difference is itself a part of our assessments and yet stems from our sampling. Figures 3Fig. 4Fig. 5Fig. 67 show average spectra collected on skin cross-section samples of a 3-month-old infant, 15-, 17-, 35-, and 62-year-old donors, respectively. The recorded spectra showed a significant degree of complexity and there were a large number of features coming out from a comparison of average Raman spectra collected at different locations and from donors with different ages. Table 2 lists all the Raman bands discussed in the remainder of this paper, together with their physical origin.

Fig. 3

Average Raman spectrum collected on the skin cross-section sample belonging to a 3-month-old infant. Spectra from both low frequency and high frequency zones are shown as collected at different depths. The origin of the labeled bands is assigned in Table 2.

JBO_20_6_065008_f003.png

Fig. 4

Average Raman spectrum collected on the skin cross-section sample belonging to a 15-year-old donor. Spectra from both low frequency and high frequency zones are shown as collected at different depths. The origin of the labeled bands is assigned in Table 2.

JBO_20_6_065008_f004.png

Fig. 5

Average Raman spectrum collected on the skin cross-section sample belonging to a 17-year-old donor. Spectra from both low frequency and high frequency zones are shown as collected at different depths. The origin of the labeled bands is assigned in Table 2.

JBO_20_6_065008_f005.png

Fig. 6

Average Raman spectrum collected on the skin cross-section sample belonging to a 35-year-old donor. Spectra from both low frequency and high frequency zones are shown as collected at different depths. The origin of the labeled bands is assigned in Table 2.

JBO_20_6_065008_f006.png

Fig. 7

Average Raman spectrum collected on the skin cross-section sample belonging to a 62-year-old donor. Spectra from both low frequency and high frequency zones are shown as collected at different depths. The origin of the labeled bands is assigned in Table 2.

JBO_20_6_065008_f007.png

In the following 10 points, we attempt to rationalize the main features that came to light from the recorded Raman spectra. The main findings can be summarized as follows:

  • (1) The spectrum acquired in the stratum corneum of the 3-month-old infant was quite different when recorded from the sample surface and from cross section (cf. Figs. 1 and 3). On the low-frequency foot of the 1150 to 1750cm1 spectral interval, a new band appeared at around 1210cm1 (labeled band A in Fig. 3), which was rather weak in the stratum corneum but became well resolved in both epidermis and dermis spectra. Band A was assigned to the stretching mode of CC6H5 in tyrosine and phenylalanine.35 Additional new features also appeared in spectra from epidermis and dermis. Although the intensity ratio between band 10 (assigned to nonpolar fragments with high proline content forming collagen triple helix) and band 9 (overlap of vibrational modes from the adenine and cytosine belonging to the β-sheet structure of amide III and lipids) was similar in the stratum corneum for spectral acquisitions from top surface and cross section, four additional bands appeared in the latter measurements at 1218, 1338, 1370, and 1395cm1 (labeled as bands A, B, C, and D, respectively, in Fig. 3). Band A was especially pronounced in both epidermis and dermis, quite different from the morphology of a low-frequency shoulder to band 9, as detected in the case of stratum corneum. It was assigned to amide III and, specifically, to coupling of CN stretching and NH bending (mixed with vibration of side chains in proteins).45 Note also that band 9 could be considered as being composed of two distinct sub-bands at lower and higher wavenumbers (at 1243 and 1274cm1, respectively). However, these two bands were both assigned to CN stretching in the α-helix conformation of amide III,36,4649 and we neglected their distinction in this study. Band B has also been reported to arise from amide III hydrated α-helix (NH bending and CN stretching)37 and to be partly contributed by tryptophan (CH2/CH3 wagging, twisting and bending).22 Band C was interpreted as ring and CN stretching in cytosine and guanine,44 whereas band D arose from symmetric CH3 bending of the methyl groups of proteins.50 Bands B, C, and D became extremely pronounced (comparable or even more intense than bands 9 and 10) in both epidermis and dermis spectra of the infant sample (cf. Fig. 3).

  • (2) The intensity of the sphingomyelin band (band 12) appeared relatively weak in the cross-section spectra of the stratum corneum of the 3-month-old-donor sample (i.e., comparable with that detected on the top surface; cf. Figs. 1 and 3 at z=0). This band did not appear at all in both epidermis and dermis spectra. Another striking feature in the cross-sectional spectra of both epidermis and dermis of the infant sample was a quite pronounced Raman activity in the zone 1550 to 1682cm1. While in the spectrum of stratum corneum recorded from the top surface, we could only observe one main band (band 13; related to mammalian keratins with mainly α-helical conformation) in cross section and six additional bands could be observed in this spectral zone (i.e., besides band 13) in both epidermis and dermis. Those bands were centered at 1520, 1551, 1584, 1605, 1638, and 1681cm1 (labeled as bands E, F, G, H, I, and J, respectively). Assignments were made, as follows: (1) band E, probably an overlap of (CC) stretching in carotenoid and CN stretching in adenine and cytosine in α-helix;51,52 (2) band F, amide II of proteins (NH bending and CN stretching);53,54 (3) band G, CC bending mode of phenylalanine;35 (4) band H, protein CC stretching in phenylalanine and tyrosine;41,55 (5) band I, amide I band related to both α-helix and β-sheet (CO stretching vibrations);39,56 and (6) band J, amide I band related to a disordered structure, nonhydrogen bonds, and CO stretching.40,56 Note that band E, which is absent in normal tissues, might be contributed here by carotenoids contained in the infant’s artificial milk. Bands F, G, H, and I were also observed in the stratum corneum of the cross-section sample, but with significantly lower intensity. Moreover, band F in the stratum corneum was shifted by few wavenumbers toward higher frequencies as compared to epidermis and dermis spectra.

  • (3) When collected from cross section, band 11 in the sample from the 3-month-old donor became clearly resolvable into three distinct sub-bands located at 1428, 1450, and 1467cm1 (labeled 11A, 11B, and 11C, respectively). We have mentioned in the previous section that band 11 can be assigned to vibrational modes in both lipids and proteins. Deconvolution into three distinct sub-bands enabled us to resolve CH2 scissoring (11A), CH bending (11B), and CH2 bending (11C) components.5658 In the high-frequency zone of the cross-section spectra from the 3-month-old-infant sample, a comparison with topside spectra showed that the basic structure of five sub-bands was still preserved, but several differences were also spotted. In the stratum corneum, one significant variation was observed in the presence of a new band centered at 2910cm1. This new band, labeled as band L in Fig. 3, was assigned by other authors to CH3 stretching vibrations in proteins.66,72 However, there is also a clear overlap with CH2 and CH3 stretching contributions from cholesterol, phospholipids and creatine.59 Significant variations in the relative intensity of various sub-bands could also be recorded. While spectra from the stratum corneum of the infant sample were found, except for the presence of the newly detected band L, to experience a quite similar balance between sub-band intensities when recorded from top surface and cross-section, a significant dropdown was found in both epidermis and dermis spectra for bands 15 and 16 (cf. Fig. 3), which could be assigned to symmetric stretching of liquid-state and crystalline lipids, respectively. This band was relatively strong only in the stratum corneum, whereas a new sub-band appeared and was labeled as band 16, at 2877cm1 in both epidermis and dermis. We justified the intensity annihilation of band 16 with the low content of lipids in infant skin. The presence of band 16 could then be interpreted according to not only a contribution from collagen (CH symmetric stretching), but also with the presence of lipids in different physical states. The packing of lipids within skin lamellae is usually referred to as the lateral lipid organization. In the order of increasing packing density, such lipid organization could either be liquid, hexagonal, or orthorhombic.60 In a healthy human stratum corneum from adult skin, the lipids mainly assemble themselves in an orthorhombic lipid organization at skin temperature, although domains with a hexagonal or liquid organization also exist along the in-depth abscissa.60,61 In infant skin, the lipid structure is expected to be strongly oriented toward a liquid and a hexagonal package. In substance, the conspicuous intensity reduction in intensity of band 15 in spectra collected upon cross-section sampling enabled resolving the actual structure of what we have detected in topside spectra (cf. Fig. 1) as a cumulative band 16 (seen at around 2881cm1). Band 15, the sub-band 16, band 16, and band L can all be assigned to symmetric stretching of CH3 units.56 However, it is hard to distinguish where such CH3 units exactly belong in the structure of skin. Bands 15 and 16 should mainly arise from CH3 units embedded in lipids with different physical states, namely liquid and crystalline (i.e., orthorhombic), respectively. Band 16 stems from CH3 units embedded in collagen, but should also be contributed by lipids in hexagonal packing (i.e., intermediate between liquid and orthorhombic structures). Band L is also a mixed contribution from proteins, cholesterol, phospholipids, and creatine. Additional discussion regarding the fractional contributions of proteins and lipids in this spectral zone will be given in the forthcoming discussion section.

  • (4) On the high-frequency side of the spectral zone at 2800 to 3200cm1, for both epidermis and dermis spectra, the sub-band related to asymmetric stretching of CH groups in lipids, fatty and unsaturated acids (band 19) completely disappeared, while a strong band newly appeared at 2980cm1 shifted by about 12cm1 toward higher frequencies as compared to band 18 (cf. Fig. 3). This new sub-band was labeled as 18+ and was assigned to CHα,α stretching.41 Moreover, band 17 at 2928cm1, which corresponds to CH3 symmetric stretching primarily in proteins, became accompanied by a new sub-band located at 2950cm1, which was labeled as band 17+. This latter band could also be assigned to CH3 asymmetric stretching in proteins.42

  • (5) As a general trend in the low-frequency region between 1150 and 1750cm1, the sample from the infant donor was the one that showed the most marked differences between spectra collected at the three investigated locations, namely at the stratum corneum, epidermis and dermis (i.e., z=0, 100, and 700μm in depth from the sample surface, respectively). The remaining samples showed less significant variations in the spectral characteristics recorded at different depths with increasing donor’s age, with the least variations apparently recorded for the oldest 62-year-old donor (compare Figs. 3 and 7). In particular, the high-frequency part of the spectra in the sample from this latter donor was almost exactly the same at z=0, 100 and 700μm in depth (cf. Fig. 7) and also exhibited a morphology profile similar to the spectrum recorded from the top-side acquisition in the stratum corneum (not shown here). Regarding the low-frequency side of the spectrum from the oldest donor in Fig. 7, the morphology of bands 11 (lipids) and 12 (sphingomyelin) was very similar for detections at different locations in cross section, although their intensity ratios were different. The intensity ratio of band 12 over band 11 (considered as the sum of sub-bands 11A, 11B, and 11C) increased significantly in both epidermis and dermis of the sample from the oldest donor, reaching values 1.3 to a value 0.87 in the stratum corneum. Note that in the spectra retrieved from the sample belonging to the youngest donor (cf. Fig. 3), the average ratio between band 12 and band 11 in the stratum corneum was very low, namely 0.06. Similarly, band 12 was also hardly resolvable in spectra collected from both epidermis (i.e., at z=200μm) and dermis (at z=700μm). Another main feature in the spectrum retrieved from the oldest donor investigated was that band B, only pronounced in the stratum corneum, apparently tended to disappear with increasing depth until becoming completely annihilated in the dermis. Band B corresponds to molecular vibrations in amide III hydrated α-helix, in particular NH bending and CN stretching. Its full annihilation might thus suggest a drastic reduction of hydrated α-helix in the amide III structure of dermis. Limited to the stratum corneum, the skin of the oldest patient showed a detectable intensity for band G at 1585cm1, while this band also tended to become very weak in the epidermis and almost disappeared in the dermis (cf. Fig. 7). Note that we noticed the inverse trend in the skin sample from the infant donor, namely a conspicuous increase in the intensity of band G in both epidermis and dermis as compared to the stratum corneum. Band G is related to the CC bending mode of phenylalanine and its decrease suggests a lack of this essential α-amino acid in the skin structure. Phenylalanine is found naturally in the breast milk of mammals, which explains its abundance in the skin sample from the infant donor. Finally, another feature appearing from the comparison between samples from the youngest and the oldest donors was the different intensity of band J in epidermis and dermis (cf. Figs. 3 and 7). This band showed a quite weaker intensity in the older donor as compared to the younger one, a characteristic that confirmed the notion of a more disordered structures for amide I in the skin of infants.

  • (6) A morphological comparison between epidermis and dermis spectra from skin samples of the 35-year-old and 62-year-old donors showed very few differences, except for a stronger Raman activity in the spectral zone between 1200 and 1400cm1 in the stratum corneum of the sample from the former donor (cf. Figs. 6 and 7). Independent of location and unlike the case of the infant donor (cf. Fig. 3), band J was quite weak in samples from both the 35- and 62-year-old donors (cf. Figs. 1 and 7). A pronounced intensity for band J in the epidermis and dermis was also found in samples from both 15- and 17-year-old patients (cf. Figs. 4 and 5). An attempt to rationalize the findings related to band J will be given in the forthcoming discussion section. In the high-frequency zone, the different behavior of band 19 was also resolvable and was more markedly present in the oldest donor. However, samples from both the 15- and 17-year-old donors showed quite different patterns regarding band 19, which seems to be quite affected by the donor habits and lifestyle. Regarding spectra from the stratum corneum and specifically the overlapping set of low frequency bands (A, 9, 10, B, C, and D), a comparison as a function of donors’ age revealed tangible morphological differences. However, spectra from the stratum corneum appeared by far the most difficult to rationalize with the same criteria. The spectrum from young donors (3 months, 15 and 17 years old, in Figs. 35, respectively) showed low intensities for bands B and C relative to the intensity of band 10. The intensity of these bands increased in the 35-year-old donor (cf. Fig. 6), but it was similar to the infant donor in the spectrum from the 62-year-old donor (Fig. 7). A similar trend was found for the intensity of band D in the stratum corneum. Some inconsistencies between the samples from 15- and 17-year-old donors could be found in the low-frequency spectral zone. Although these donors were close in age, quite different spectral morphologies could be found in the stratum corneum (cf. Figs. 4 and 5). These considerations lead us to the conclusion that low-frequency Raman spectra from the stratum corneum could be less suitable for age analyses due to the strong effect of environmental conditions and individual habits.

  • (7) The high-frequency part of the Raman spectra in samples from donors of intermediate age was also analyzed in comparison with the two extreme cases of infant and older donors. In dermis spectra, band 18+ was only present with relatively high intensity in samples from the two youngest donors (Figs. 3 and 4), while it disappeared in samples from the three older donors together with the appearance of band 18 (with intensity decreasing with increasing age) and band 19 (with intensity almost constant with age; cf. Figs. 57). Both band 18 (CH3 asymmetric stretching) and band 18+ (CHα,α stretching) are related to vibrational modes in proteins; therefore, the preferential presence of one over the other might serve to specify protein quality in the donor. However, the present data do not seem to allow qualifying whether or not such a feature specifically relates to the donors’ ages. For example, in the 35-year-old sample, band 18+ from CHα,α stretching was present in the stratum corneum and epidermis, but it disappeared in dermis (cf. Fig. 6). On the other hand, in the infant donor sample, band 18+ was detected in epidermis and dermis, but it disappeared in the stratum corneum (cf. Fig. 3). In the sample from the 15-year-old donor, band 18+ appeared at any location (cf. Fig. 4), while it disappeared at any location in samples from the 17- and 62-year-old donors in favor of a clearly resolvable band 18, present at any location along the cross section (cf. Figs. 5 and 7). A meaningful trend with increasing donor age seemed to arise in the variations of Raman intensity in the low-frequency shoulder of the cumulative spectrum at 2800 to 3200cm1, namely for the group of bands 15, 16, 16, and L. Upon considering the relative intensity of bands (16, 16, L) versus band 15, namely the fraction of bands contributed by crystalline lipids versus the one contributed by lipids in the liquid state, a consistent trend showing a tangible increase of the former state with increasing donor age could be observed. A plot and a more detailed discussion of this trend will be given in Sec. 4.2.

  • (8) Band B, which arises from NH bending and CN stretching in amide III hydrated α-helix, seemed to provide a coherently decreasing trend with age in the dermis zone of all the available samples. A decrease in content of hydrated α-helix with increasing age represents a reasonable argument from the viewpoint of ceramide containing structures and skin barrier functions.73 This trend was less clear (i.e., more scattered) in epidermis spectra and completely absent in a comparison among spectra from the stratum corneum among the investigated samples. This means that environmental factors can also strongly affect hydration of α-helix in portions of skin exposed to sunlight or externally treated. Moreover, dietary effects have also recently been reported in this context, which might alter both dryness and dehydration of the skin.73 Renugopalakrishnan et al.74 noticed the amide I emission (i.e., dominated by CO stretching vibrations and typical of α-helical conformation in mammalian keratin) for spectroscopically evaluating thermal denaturation in chick skin. The used band corresponds to band 13 in Figs. 37, which we observed as centered at a frequency of 1657cm1. Moreover, Tosato et al.75 discussed band 13 as a sensor for structural modifications of the amide group. The rationale of the discussion was based on the observation that the CO stretching mode in amide I is weakly coupled to the stretching mode of the carbon–nitrogen bond and to the in-plane deformation mode of the nitrogen–hydrogen amide bond. Accordingly, changes in the molecular geometry due to degradation of collagen triple helix chains might result in the dissociation of the triple helix into a simple string or a double string.76,77 Band 13 in spectra at any location of our samples showed significant shifts in frequency, but no morphological variations nor any obvious correlation with band B in different samples. On the other hand, again in Ref. 74, the most striking difference of thermal denaturation was observed for the amide III doublet, which we labeled as band 9 (vibrational modes from the adenine and cytosine belonging to the β-sheet structure of amide III) and band A (coupling of CN stretching and NH bonding in amide III). This doublet collapsed upon denaturation to produce a strong band at lower frequencies at around 1243cm1. We also noticed a collapse of band A in dermis spectra with increasing age, but no similar trend could be observed for band 9. Moreover, no strong new band could be detected at around 1243cm1. Following Renugopalakrishnan et al.,74 the collapse of band A could, in principle, be interpreted as a signal of incipiency for a denaturation process of protein and polypeptides in the amide III structure, which progressively occurs with increasing age.

  • (9) The lipidic content of skin is a key factor in the permeability barrier function, including cosmetics effects and transdermal drug delivery. Abnormalities in barrier function, which are related to lipid content, have been shown to lead to atopic dermatitis and other common cutaneous diseases.78 Free fatty acids and triglycerides are affected in their compositions by ultraviolet irradiation,79 and a decrease in lipid content with age leads to an increased susceptibility to exogenous insults.80 These latter two aspects are essential to our research because they show that if the effect of environmental factors could be separated, a detailed spectroscopic knowledge of lipidic content, composition, and structure might not only help to elucidate skin function, malfunction and abnormality, but it could also be a marker of the actual age of skin. An early study by Ghadially et al.80 showed that fatty acids in senescent murine epidermal lipids are clearly different from young controls, since they contain a disproportionate increase in medium and long chain species. On the other hand, no changes were apparent in very long chain species. Moreover, in both aged human and mouse epidermis, a paucity of secreted lamellar body contents was present at the stratum granulosum/stratum corneum interface, a finding that matched the lipid biochemical observation showing that lipids in aged stratum corneum are globally reduced in quantity, without exhibiting any specific abnormality in species distribution or fatty acid composition. A former study by Gaber and Peticolas38 suggested the use of the Raman activity of lipids in the spectral interval between 2850 and 2890cm1 (our bands 15 and 16) for assessing the physical state of phospholipids. As we have discussed in point (4), we detected in this study a clear monotonic variation with donors’ ages of the Raman activity in this high-frequency region, which suggests the presence of lipids in an increasingly crystalline state (over liquid and semicrystalline states) in donors with increasing age. On the other hand, if we look at the lipid/protein activity in the low-frequency region, as represented by bands 11A, 11B, and 11C at around 1440cm1, we could not find any monotonic trend with increasing age at any investigated location. This trend is probably due to the fact that both lipids and proteins contribute to the three sub-bands. This point will be discussed in more detail in the initial subsection of the forthcoming discussion. Moreover, the balance between the intensity of various sub-bands was barely altered as a function of age and location (cf. Figs. 37). Translating the above set of information into physical arguments, our experiments prove that band 11, which “feels” local motifs encoded into the primary structure of lipids and proteins [i.e., CH2 scissoring (11A), CH bending (11B), and CH2 bending (11C)], is not strongly affected by age; but, regarding lipids, it is their physical state that monotonically changes with progressing age, namely bands 15 and 16, which reflect the tertiary structure of the lipid molecules, actually undergo a progressive crystallization process of lipid assembly in older patients.

  • (10) The Raman activity in the interval of frequencies from 1600 to 1700cm1 typically appeared as a triplet (bands I, 13, and J), with bands 13 and I being the most prominent in the stratum corneum and in epidermis or dermis, respectively. As already mentioned, band 13 has unambiguously been assigned to the amide I vibrational mode,81 which mainly involves CO stretching and, to a lesser extent, NH in-plane bending of peptide groups.82,83 The exact spectral location of this band strongly depends on the secondary structure of the polypeptide chain and could, therefore, be useful for estimating secondary structure fractions of proteins.84 In comparing epidermis and dermis spectra from the oldest and the youngest donors in this study, band 13 was found to be clearly shifted towards lower frequencies in the former donor. Regarding samples from donors with intermediate age, the sample from the 35-year-old donor displayed band 13 at nearly the same frequency as the sample from the 62–year-old donor. On the other hand, band 13 from both samples from the 15- and 17–year-old donors was very close in frequency to that of the infant donor. Despite a possible interaction with vibrational modes in lipids, a strong intensity of band 13 in the Raman spectra should testify to a preponderance of proteins with high α-helix content.82,83,85 On the other hand, band I, which comprises amide I contributions from both α-helix and β-sheet (CO stretching vibrations) is more difficult to interpret.39,56 A spectral shift toward lower frequencies of bands 13 and I could be interpreted as a mark for the presence of heavier structures in α-helix and β-sheet, and thus for a larger interaction with lipids. The spectral intensity of band J, which encompasses all the amide I contributions from disordered structures, nonhydrogen bonds, and CO stretching,40,56 to the sum of α-helix-representative band 13 and α-helix and β-sheet-related band I can be assumed as a direct measure of the degree of disorder of the protein structure in the tissue. Our data showed that in epidermis and dermis, the intensity of band J with respect to band 13 seemed to decrease with increasing age, although its absolute value apparently depended on individual donors. This point will be discussed in more detail in the forthcoming sections. In principle, there could also be another additional spectral feature that characterizes the fractional presence of β-sheet structures. This is a band appearing at around 1517cm1, which mainly involves NH in-plane bending and CN stretching of the trans peptide group of amide II vibrations.81 However, this band was hardly observed in this study. According to a study by Shao et al.,86 dealing with structural changes in heated up proteins upon interaction with lipids, the increase of β-sheet formation is due to proteins–lipids and proteins–proteins interactions during environmental exposure. In general, the behavior of Raman bands related to peptides and proteins, also including the band at 1517cm1, in the stratum corneum is not necessarily intrinsic to the skin structure but might be altered by environmental conditions. For this reason, we have neglected it in our further characterizations.

According to the spectral characterizations shown above, we monitored the following spectral features:

  • (1) the intensity ratio of band 10 on band 9, I10/I9, representative of the fractional ratio between α-helix and β-sheet;

  • (2) the intensity ratio of band J on band 13, IJ/I13, representative of the fractional ratio between disordered and ordered structures (in peptides and proteins);

  • (3) the intensity ratio of band 12 on band 11B, I12/I11B, representative of the content of sphingomyelin as compared to the overall amount of lipids in the sample;

  • (4) the intensity ratio of band 16 on band 15, I16/I15, representative of the physical state of lipids, namely liquid versus crystalline state; and

  • (5) the spectral shift of band 13, ω13, representative of changes in molecular geometry of amide I, due to the degradation of collagen triple helix chains and their dissociation into simple or double strings.

More information about the variation of these five factors along the in-depth axis on cross sections of the investigated skin samples is available in the Appendix, and the dependence of these parameters on age will be further discussed in the following section.

4.

Discussion

4.1.

Subtracting from Skin Spectra the Contribution of Standard Collagen

The Raman spectrum of dermis is dominated by collagen, which in turn constitutes 70% of the dry weight and 90% of the total protein content. Important contributions to the Raman spectrum of skin are also expected in the stratum corneum and in the epidermis zones. Of the about 20 types of collagen existing in the human body, skin consists of 80% type I and 15% type III, while the remaining 5% is predominantly type IV collagen.8789 Figure 8 shows the Raman spectrum of skin (type I) collagen purchased from Sigma-Aldrich Co, which was collected under exactly the same experimental conditions as the spectra of skin as shown in Figs. 37.

Fig. 8

Raman spectrum of skin (type I) collagen in (a) low and (b) high frequency regions, which was collected under exactly the same experimental conditions as that the spectra of skin. The labeled bands are assigned in Table 2.

JBO_20_6_065008_f008.png

On the low-frequency side of the spectrum, as shown in Fig. 8(a), most of the bands could be labeled, similar to what was detectable in the skin samples (cf. Table 2). The relatively strong intensity of the CH vibration bands (11A, 11B, and 11C) in skin collagen suggests that the contribution of collagen to their overall intensity in skin samples is not negligible as compared to the respective contributions in lipids. Only two additional bands (i.e., bands 1* and 3*) remained to be assigned; these bands were not obviously detectable in the skin samples as they appeared on the low-frequency side of the collagen spectrum. Band 1* was centered at around 1242cm1 and could be related to amide III disordered structures (CN stretching and CH2 wagging), whereas band 3*, at 1315cm1, was again related to α-helix amide III (CH2 bending mode).65

Regarding the high-frequency side of the Raman spectrum of pure collagen, we have collected six distinct bands (labeled as 15, 16, 17–19 and L) with an overall morphology comparable to that found in skin spectra. However, non-negligible shifts (toward lower or higher wavenumbers) could be found for some bands. Interestingly, band L was quite weak, thus suggesting a preponderant contribution of lipids over proteins. Moreover, band 16, observable in skin, could not be found in the collagen spectrum. The important implication in these findings is that band 15 and band 16 as observed in the skin sample, which were assigned to symmetric stretching of CH3 units in lipids embedded in phases with different physical states, also exist in collagen with the same physical origin, but appearing at shifted frequencies. In an attempt to better visualize this complex spectroscopic situation, we normalized the high-frequency side of the skin spectra to the respective intensity maxima along the cross-section scan (in Figs. 37), and from them we subtracted the contribution of the similarly normalized collagen spectrum (Fig. 8). The results of this procedure are shown in Fig. 9 for spectra belonging to the youngest and to the oldest donors in this study [spectra from the 3-month- and 62-year-old donors in (a) and (b), respectively]. In the plots, the Raman intensity difference, ΔI=IsIc, is plotted as a function of spectral locations, where Is and Ic represent the normalized intensities of skin and collagen spectra, respectively. In each figure, three plots are given which correspond to different depths along the z-axis, namely z=0 (stratum corneum), z=100μm (epidermis), and z=700μm (dermis). In the plots, positive and negative values for ΔI correspond to a preponderance of lipids or collagen contributions to the skin spectra, respectively. For better clarity, the spectral location of these bands, belonging to collagen, is also shown in insets of both figures.

Fig. 9

Raman intensity difference, ΔI=IsIc, as a function of spectral locations, as calculated after spectral normalization to the respective intensity maxima; the plots refer to the 3-month- and 62-year-old donors [in (a) and (b), respectively]. Three plots are displayed for each donor, which represent averages over the stratum corneum, epidermis and dermis.

JBO_20_6_065008_f009.png

The following conclusions could be drawn as the main outputs of the spectral subtraction procedure:

  • (1) The Raman spectrum from the dermis zone of the skin sample was completely composed of type I collagen in the case of the infant donor [cf. ΔI0 over the entire high frequency zone in Fig. 9(a)]. However, increasingly pronounced spectral contributions from lipids could be noticed at 2800 to 2930cm1 when proceeding along the cross-section line-scan toward the surface of the sample.

  • (2) In the Raman spectrum from the 62-year-old donor, from surface to z=700μm, subtraction of the collagen spectrum from the skin spectrum produced the same results, independent of location along the line scan. Lipids were preponderant in the spectral interval 2800 to 2930cm1, while the main spectral contribution at higher frequencies mainly stemmed from type I collagen structures [cf. Fig. 9(b)].

  • (3) The spectral subtraction procedure, showing that the band labeled as 16 in skin corresponded to CH symmetric stretching in collagen, has provided an important confirmation about the possibility of using bands 15 and 16 (centered at 2855 and 2885cm1, respectively) for discussing the physical state of lipids in skin. However, discussions on lipid structures using the above two bands should be limited to the stratum corneum and epidermis, since their contributions in the spectrum from dermis were, in young donors, completely contributed by collagen structures.

4.2.

Possible Spectroscopic Parameters for Evaluating Human Age

Our attempts “to decode” skin spectra in search for a possible natural “clock” in the vibrational behavior of skin have brought us several hints concerning the evolution of the chemical and physical nature of proteins and lipids with age. Figure 10 shows plots of different selected parameters as a function of donors’ ages.

Fig. 10

Plots of different parameters as a function of donors’ ages: the intensity ratio, I10/I9, (a) representative of the fractional ratio between α-helix and β-sheet; the ratio, IJ/I13, (b) representative of protein folding upon aging; the spectral shift of band 13, ω13, (c) representative of the amount of rupture of triple helix molecules within the collagen macromolecule due to degradation of collagen type IV; and (d) the ratios of I16#=I16+I16+IL to I15 and IS=I16+I16 to I15 as representative of the process of lipid crystallization upon aging.

JBO_20_6_065008_f010.png

As previously discussed, the intensity ratios I10/I9 and IJ/I13 can be considered as representative of the fractional ratio between the α-helix and β-sheet, and of the fractional ratio between disordered and ordered structures (in peptides and proteins), respectively. The plots in Figs. 10(a) and 10(b) display these (average) parameters as a function of donors’ ages. In the averaging process, we have excluded the stratum corneum in order to minimize environmental effects. The trend for I10/I9 in the former plot, when averaged on different areas of the cross section of the samples, showed a minimum for the 17-year-old donor. This clearly nonmonotonic nature was more pronounced in the epidermis than in the stratum corneum. We have similarly found nonmonotonic plots (not shown here) for other parameters [e.g., the sphingomyelin ratio, I12/I11B], as proposed above. On the other hand, the plot of IJ/I13 versus donors’ ages remarkably fitted to a high degree of precision with an exponential decrease in donors’ ages. According to our data, the algorithm relating the Raman intensity ratio IJ/I13 to age, xA, can be given as

Eq. (1)

IJ/I13=0.74×Exp[0.048×xA].

The notion of folding and misfolding in peptides and proteins is a well-known one in the field of biophysics.9092 In order to become functionally active, newly made (or “nascent”) protein chains must assemble into a “fold,” namely a well defined 3-D pattern. An amino acid sequence specifies the information that specifies the fold, which is a process thermodynamically driven by the hydrophobic effect. Structural rearrangement gives rise to the correct amino acid packing that corresponds to the most stable and active state in healthy subjects. This task is completed within intervals of time between milliseconds and many minutes, depending on protein size. On the other hand, band 13, in the region at around 1659cm1, is associated with the triple-stranded helix stabilized in collagen by a large number of interchain hydrogen bonds.93 Fourier transform infrared spectroscopy data by Federman et al.94 have assessed band displacements from the 1652 toward 1658cm1, which indicates the rupture of the triple helix molecule within the collagen macromolecule due to degradation of collagen type IV by the metalloproteinase trypsin. In other words, the plot in Fig. 10(c), which shows the spectral displacements of band 13 as a function of age, is another aspect of protein changes related to aging. It should be noted that in the case of collagen, newly synthesized premature collagen is imported into the lumen of endoplasmic reticulum and folded and modified during transportation through the Golgi apparatus and is then secreted as a mature form. During this protein maturation, misfolded proteins are subjected to degradation. Mature proteins are also degraded when damaged. This protein metabolism can be affected by both age and environmental stress. The protein state in the skin should thus be the result of both synthesis and degradation, which is altered in an age-dependent manner, so young skin contains a higher amount of fresh collagen than old.

In conclusion, our data in Fig. 10(b) simply show that the amount of proteins in skin, which are yet to assembly into folds, is largest in infants and gradually reduces with aging. This circumstance is thus the spectroscopic representation of the fact that the dermis of newborns contains less mature collagen than adults. Moreover, Fig. 10(c) shows that stabilization of the triple helix is maximized at intermediate ages, a behavior that is in good agreement with the maximization of α-helix contents testified by the concurrent minimum of the intensity ratio, I10/I9[Fig. 10(a)]. In other words, as far as proteins are concerned, Raman spectroscopic data have provided us with a consistent picture of structural evolution with age which is in agreement with well-established concepts in biophysics of the aging process.

Regarding lipids, we show here another spectroscopic plot which apparently contained characteristics of remarkable precision with respect to the evolution with increasing age. This plot is shown in Fig. 10(d) and represents the relationship found between the first two maxima appearing in the high-frequency zone investigated at around 2885 and 2890cm1. In epidermis and dermis, the presence of more crystallized and ordered structures (i.e., a decreasing trend for the IJ/I13 ratio) through aging corresponds to a general and well-established notion in biophysics.9597 We plotted, in Fig. 10(d), both the ratio of I16#=I16+I16+IL to I15 and the ratio of IS=I16+I16 to I15. Both plots show a similar trend of exponential increase with increasing age, although the former plot gives a more distinct separation between young and old donors. Note also that the former plot could be preferable to the latter one because it just represents the ratio between the first two maxima in correspondence of the lower foot of the high-frequency zone of the skin spectrum; thus, it does not require any spectral deconvolution procedure to be calculated. Interestingly, a plot that simply uses the intensity of the deconvoluted band 16 to band 15 leads to the opposite trend (i.e., a decreasing ratio with age; cf. Figs. 11 and 12 in Appendix). This finding, together with the observation of the remarkably low intensity of band L in collagen, suggests the preponderance of the lipid contribution (i.e., cholesterol, phospholipids and creatine) of band L in skin samples. Following the definition given by Gaber and Peticolas,38 we defined the order parameter for lateral interaction, SL, according to the following equation:

Eq. (2)

SL=(I16#/I15)0.71.5,
which refers to lipids in fully crystalline and fully liquid states when equal to 1 and 0, respectively. This parameter, given in the inset of Fig. 10(d), reflects the intermolecular structure of lipids and decreases in the order from lamellar liquid, hexagonal liquid crystal and orthorhombic crystal state. According to our data, an empirical equation that represents the Raman intensity ratio, I16#/I15, as a function of age, xA, can be drawn, as follows:

Eq. (3)

I16#/I15=1.06×Exp[0.00028×xA1.75]
or, alternatively, in terms of the order parameter, SL:

Eq. (4)

SL=0.707×Exp[0.00028×xA1.75]0.467.

4.3.

Needs for the Raman Probe in Forensic Assessments of Human Age

A reliable estimation of age at death, which is a main element in the identification of bodies of unidentified origin, is also one of the main challenges in forensic sciences. For example, official data from Kyoto Prefecture98 report an average number of 13 persistently unidentified human bodies per year in the last 12 years. Such a large number of cases for a limited geographical area could give an idea of the severity of this problem from the social side. A number of anthropological techniques have been put forward to estimate the age at death in children and adults, but they are obviously insufficient in a number of cases. These techniques include long-bone length, epiphyseal fusion, dental eruption, and lengths of diaphysis at birth for infant and juvenile remains; eruption of third molars, fusion of the spheno-occipital synchondrosis, pubic symphysis, auricular surface, cranial sutures, sternal rib ends (costal cartilage), maxillary suture closure, tooth-root translucency, and formation of osteoarthritis characteristics in adults (for a complete review of these methods, see Ref. 99). However, two main shortcomings appear in applying these methodologies: (1) an increasingly lower accuracy for adult subjects and (2) the absolute necessity of specific references depending on population. In order to overcome these deficiencies, new methodologies have been developed, which have so far included both biochemical and chemical methods. The former methods basically consist of screening the natural processes of aging, thus including different biochemical changes that lead to alterations in cells and tissues. Biochemical methods have so far relied on forensic analyses of hair, drugs in hair blood, and semen, based on infrared spectroscopy, chromatography, ultraviolet, and mass spectrophotometry.100 The chemical methods, on the other hand, involve manipulation and modifications of molecules or accumulation of selected products as, for example, modifications that take place in DNA and chromosomes. Among the chemical methods, the most accurate technique today is considered to be aspartic acid racemization in noncollagenous bone proteins or tooth enamel,101,102 although other techniques are being concurrently used depending on the forensic context under examination.103 In this paper, we have described a Raman approach to age evaluation on skin samples which basically belongs to the biochemical type of analyses since it requires no sample manipulation except for the extraction of a skin sample from the ventral part of the body. Although Raman spectroscopy could, in principle, also be applied in situ without any physical extraction of a skin sample from the body, our study has clearly shown how crucial it is to obtain cross-sectional skin samples in order to retrieve more detailed spectral information on protein and lipid structures and to avoid biases of environmental nature unavoidably contained in the stratum corneum (i.e., on the top surface) of the skin sample. Although less established as compared to radiographic analyses,104,105 Raman spectroscopy has been widely used in different branches of forensic science.106,107 However, studies using the Raman method for age assessments are few and are based on phenomenological approaches.108,109 In Ref. 109, the authors used a correlation between the variability of Raman spectra and the stages of dentinal evolution with advancing age and remarkably obtained predictions of a correct age, with a mean error of ±5 years. The main benefit of this method consisted of minimal and nondestructive sample preparation, which in turn led to an efficient age prediction for any age group. From this viewpoint, our study of skin presents a similar advantage but also necessitates more detailed and physics-driven Raman analyses of organic molecules. The very least outputs of our study are a complete labeling of Raman bands in cadaveric samples of skin (including an infant), their spectral deconvolution and their different trends with age as a function of in-depth abscissa. Despite the limited number of cadaveric samples investigated, which unavoidably limits the nature of this investigation to a proof-of-concept study, the Raman response of proteins and lipids in skin samples seemed to remarkably obey precise patterns, consistent with general biophysical concepts. Thus, Raman spectroscopy seems to provide a reliable and more straightforward path to age determination as compared to other spectroscopic techniques involving electrons, microscopic forces or neutrons.110113

5.

Conclusion

The aim of this work was to establish a correlation between age and Raman spectra retrieved from human skin. We explored both top-surface and cross-sectional Raman responses in skin samples from cadaveric donors of different ages and found the latter spectra more meaningful and richer in structural details than the former ones, regarding both proteins and lipids. Clear correlations could be found between the relative intensity of selected Raman bands and the stages of protein folding and evolution of lipid crystallization with advancing age. Protein folding seems to be a more sensitive parameter for infants and young patients, while lipid crystallization follows more sensitive variations with age for patients in their adulthood. It is probable that an algorithm combining both these two spectral parameters could improve the precision of age estimation. Moreover, there were hints for the presence of additional “biological clocks” in the Raman spectrum of skin. However, the complexity of the retrieved spectra poses considerable challenges to additional findings. Although a limited number of only five cadaveric donors that were available for our experiments necessarily confines our data into a mere proof-of-concept frame, the degrees of protein folding and lipid crystallization seemed to represent precise predictors of biological age. While we hope that our findings will stimulate other researchers to prove the newly stated concepts, additional experiments are going on in our laboratory to enlarge the number of donors every time they become available to us. The proposed Raman method should especially be suitable for those kinds of situations where traditional methods fail and a prompt and minimally invasive evaluation of age is needed.

Appendices

Appendix

Profiles along the in-depth axis on cross sections of the skin samples. Figures 11 and 12 show line-scan plots of these spectroscopic parameters as a function of distance, z, from the surface of the skin sample. Plots are displayed for samples from the 3-month- and 62 year-old donors in Fig. 11, whereas Fig. 12 compares the results obtained from samples belonging to the 15-year- and 35-year-old donors. For brevity’s sake, data from the 17-year-old donor were not explicitly shown, since they were very similar to those collected on the sample from the 15-year-old sample. As discussed in the manuscript, the intensity ratios I10/I9 and IJ/I13 [displayed in Figs. 11(a) and 11(e), and 12(a) and 12(e)] were considered as representative of the fractional ratio between α-helix and β-sheet and of the fractional ratio between disordered and ordered structures (in peptides and proteins), respectively. A comparison between the youngest and the oldest donors in this study [i.e., as displayed in Figs. 11(a) and 11(e)] revealed a similar in-depth distribution for the ratio I10/I9, while the trends for the ratio IJ/I13 were quite different. Leaving aside data from the stratum corneum (strongly affected by environmental effects), we examined data collected in the epidermis and dermis region. In these zones, the IJ/I13 ratio was almost constant at 0.7 for the infant sample, while it was 0 in the sample from the old donor. This finding was interpreted with the presence of a more crystallized and ordered structure in the older donor as compared with the infant donor. Looking at the samples from the donors of intermediate age [in Figs. 12(a) and 12(e)], the trend of a decreasing IJ/I13 ratio (i.e., of an increased crystallinity and structural order) in both epidermis and dermis with increasing age seemed to be confirmed. Moreover, the ratio I10/I9 also experienced lower values at intermediate ages as compared to both samples from the youngest and oldest donors. The ratio I12/I11B was monitored [cf. Figs. 11(b), 11(f), 12(b), and 12(f)] because it was thought to represent the content of sphingomyelin as compared to the overall amount of lipids in the sample. The amount of sphingomyelin showed a trend that was not monotonic with donors’ ages. It was highest in the deep portions of the sample from the 35-year-old donor, while the minimum value was detected in the infant sample.

Fig. 11

Line-scan plots as a function of distance, z, from the surface of the skin sample comparing the trends found for the 3-month [(a) to (d)] and the 62-year-old donor [(e) to (h)]. Plots show the intensity ratio of band 10 on band 9, I10/I9 and the intensity ratio of band J on band 13, IJ/I13 [in (a) and (e)]; the intensity ratio of band 12 on band 11B, I12/I11B[in (b) and (f)]; the intensity ratio of band 16 on band 15, I16/I15 [in (c) and (g)]; and, the spectral shift of band 13, ω13 [in (d) and (h)].

JBO_20_6_065008_f011.png

Fig. 12

Line-scan plots as a function of distance, z, from the surface of the skin sample comparing the trends found for the 15-year-old [(a) to (d)] and the 35-year-old donor [(e) to (h)]. Plots show the intensity ratio of band 10 on band 9, I10/I9 and the intensity ratio of band J on band 13, IJ/I13 [in (a) and (e)]; the intensity ratio of band 12 on band 11B, I12/I11B[in (b) and (f)]; the intensity ratio of band 16 on band 15, I16/I15 [in (c) and (g)]; and, the spectral shift of band 13, ω13 [in (d) and (h)].

JBO_20_6_065008_f012.png

Sphingomyelin (also referred to as ceramide phosphorylcholine) consists of a ceramide unit with an attachment of phosphorylcholine moiety. Although sphingolipids are the main polar lipid constituents of milk, they represent an important but minor nutrient for infants, which justifies the low level found for this sphingolipid in the sample from the infant donor. Holleran et al.114 have shown that sphingolipids, including sphingomyelin, represent 25% of the lipids located in the stratum corneum and are a major element of the epidermal permeability barrier. Moreover, alterations in epidermal barrier function lead to a rapid increase in cholesterol and fatty acid synthesis, which parallels the early stages of the repair process. Sphingolipid synthesis was also found to increase during barrier repair but with a delayed response in comparison to cholesterol and fatty acid synthesis.115 Since a deficiency of any of these species will result in abnormal membrane structures with a reduced capacity to impede trans-epidermal water flux,116 our findings here might simply show that such a protective capacity in skin is maximized in the medium age of humans, initially increasing from infant age and then again turning back to low levels for advanced ages.

Figures 11(c), 11(g), 12(c), and 12(g) show the line-scan trends retrieved for the Raman intensity ratio I16/I15, which is representative of the physical state of lipids, namely liquid versus crystalline state. A comparison between the youngest and oldest donors [Figs. 11(c) and (g)] shows a clearly positive gradient up to high values of this intensity ratio (i.e., up to 5 to 6) as a function of in-depth abscissa, z, in the former sample, versus a conspicuously constant (and low) value 2 in the latter sample. Samples from donors from intermediate ages [Figs. 12(c), and 12(g)] also showed intermediate trends with conspicuously constant values in large portions of both the epidermis and dermis. High I16/I15 ratios represent a preponderance of crystalline lipids over liquid ones, which conceivably accompany the increase in age in humans. Interestingly, a similar trend could be obtained from retrieving the spectral shifts, ω13, for band 13 as a function of the abscissa, z [cf. Figs. 11(d), 11(h), 12(d), and 12(h)]. A gradual increase as a function of the abscissa, z, was found for the infant donor versus a lesser increase up to lower frequencies in the oldest donor. Intermediate trends appeared in samples from donors of intermediate ages. As discussed in the paper, the shift of band 13 toward higher frequencies has been considered to represent changes in the molecular geometry of amide I due to degradation of collagen triple helix chains and their dissociation into simple or double strings. Higher frequencies, such as those being observed in younger donors, thus might represent a higher fraction of unfolded structures in young skin samples, as also suggested by the higher values of the IJ/I13 ratio. On the other hand, shifts toward low wavenumbers have been interpreted as the effect of the squeezing out of water (and a consequent reinforcement of hydrogen bonds) between adjacent chains in a study of keratin fibers under strain by Paquin and Colomban.65 This interpretation leads us to consider a higher chain packing in adult skin samples, which is also a reasonable output when considering the chemistry of skin aging.

References

1. 

W. J. Netzer and F. U. Hartl, “Recombination of protein domains facilitated by co-translational folding in eukaryotes,” Nature, 388 343 –349 (1997). https://doi.org/http://dx.doi.org/10.1038/41024 NATUAS 0028-0836 Google Scholar

2. 

A. Barth and C. Zscherp, “What vibrations tell about proteins?,” Q. Rev. Biophys., 35 369 –430 (2002). https://doi.org/http://dx.doi.org/10.1017/S0033583502003815 QURBAW 0033-5835 Google Scholar

3. 

J. J. L. Lippert and W. L. Peticolas, “Laser Raman investigation of the effect of cholesterol on conformational changes in dipalmitoyl lecithin multilayers,” Biochim. Biophys. Acta, 282 8 –17 (1972). https://doi.org/http://dx.doi.org/10.1073/pnas.68.7.1572 BBACAQ 0006-3002 Google Scholar

4. 

Y. M. Weng et al., “Structural analysis of triacylglycerols and edible oils by near-infrared Fourier transform Raman spectroscopy,” Appl. Spectrosc., 57 413 –418 (2003). https://doi.org/http://dx.doi.org/10.1366/00037020360625952 APSPA4 0003-7028 Google Scholar

5. 

M. F. Mrozek and M. J. Weaver, “Detection and identification of aqueous saccharides by using surface-enhanced Raman spectroscopy,” Anal. Chem., 74 4069 –4075 (2002). https://doi.org/http://dx.doi.org/10.1021/ac020115g ANCHAM 0003-2700 Google Scholar

6. 

R. J. Weesie et al., “Semiempirical and Raman spectroscopic studies of carotenoids,” Biospectroscopy, 5 19 –33 (1999). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1520-6343 BIOSFS 1075-4261 Google Scholar

7. 

K. Maquelin et al., “Identification of medically relevant microorganisms by vibrational spectroscopy,” J. Microbiol. Methods, 51 255 –271 (2002). https://doi.org/http://dx.doi.org/10.1016/S0167-7012(02)00127-6 JMIMDQ 0167-7012 Google Scholar

8. 

K. De Gussem et al., “Raman spectroscopic study of Lactarius spores (Russulales, Fungi),” Spectrochim. Acta A, 61 2896 –2908 (2005). https://doi.org/http://dx.doi.org/10.1016/j.saa.2004.10.038 SAMCAS 1386-1425 Google Scholar

9. 

H. G. M. Edwards et al., “Non-destructive analysis of pigments and other organic compounds in lichens using Fourier-transform Raman spectroscopy: a study of Antarctic epilithic lichens,” Spectrochim. Acta A, 59 2301 –2309 (2003). https://doi.org/http://dx.doi.org/10.1016/S1386-1425(03)00073-8 SAMCAS 1386-1425 Google Scholar

10. 

V. Baeten, “Raman spectroscopy in lipid analysis,” Lipid Technol., 22 (2), 36 –38 (2010). https://doi.org/http://dx.doi.org/10.1002/lite.200900082 LITEEI Google Scholar

11. 

J. De Gelder et al., “Reference database of Raman spectra of biological molecules,” J. Raman Spectrosc., 38 1133 –1147 (2007). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

12. 

G. J ThomasJr., “Raman spectroscopy of protein and nucleic acid assemblies,” Annu. Rev. Biophys. Biomol. Struct., 28 1 –27 (1999). https://doi.org/http://dx.doi.org/10.1146/annurev.biophys.28.1.1 ABBSE4 1056-8700 Google Scholar

13. 

C. Raschke and P. Elsner, “Skin aging: a brief summary of characteristic changes,” Textbook of Aging Skin, 37 –44 Springer-Verlag, Berlin (2010). Google Scholar

14. 

F. H. Silver, Mechanosensing and Mechanochemical Transduction in Extracellular Matrix Biological Chemical Engineering and Physiological Aspects, Springer, New York (2006). Google Scholar

15. 

J. Rosenbloom, W. R. Abrams and R. Mecham, “Extracellular matrix 4: the elastic fiber,” FASEB J., 7 (13), 1208 –1218 (1993). FAJOEC 0892-6638 Google Scholar

17. 

M. Gniadecka et al., “Structure of water proteins and lipids in intact human skin hair and nail,” J. Invest. Dermatol., 110 393 –398 (1998). https://doi.org/http://dx.doi.org/10.1046/j.1523-1747.1998.00146.x JIDEAE 0022-202X Google Scholar

18. 

W. Akhtar and H. G. M. Edwards, “Fourier-transform Raman spectroscopy of mammalian and avian keratotic biopolymers,” Spectrochim. Acta A, 53 81 –90 (1997). https://doi.org/http://dx.doi.org/10.1016/S1386-1425(97)83011-9 SAMCAS 1386-1425 Google Scholar

19. 

H. G. M. Edwards, A. C. Williams and B.W. Barry, “Potential applications of FT-Raman spectroscopy for dermatological diagnostics,” J. Mol. Struct., 347 358 –379 (1995). https://doi.org/http://dx.doi.org/10.1016/0022-2860(95)08560-I JMOSB4 0022-2860 Google Scholar

20. 

R. Manoharan, Y. Wang and M. S. Feld, “Histochemical analysis of biological tissues using Raman spectroscopy,” Spectrochim. Acta, 52 215 –249 (1996). https://doi.org/http://dx.doi.org/10.1016/0584-8539(95)01573-6 SPACA5 0038-6987 Google Scholar

21. 

N. Stone et al., “Near-infrared Raman spectroscopy for the classification of epithelial pre-cancers and cancers,” J. Raman Spectrosc., 33 564 –573 (2002). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

22. 

W. T. Cheng et al., “Micro-Raman spectroscopy used to identify and grade human skin pilomatrixoma,” Microsc. Res. Tech., 68 (2), 75 –79 (2005). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0029 MRTEEO 1059-910X Google Scholar

23. 

R. Dong et al., “Temperature-dependent Raman spectra of collagen and DNA,” Spectrochim. Acta A, 60 557 –561 (2004). https://doi.org/http://dx.doi.org/10.1016/S1386-1425(03)00262-2 SAMCAS 1386-1425 Google Scholar

24. 

W. T. Cheng et al., “Micro-Raman spectroscopy used to identify and grade human skin pilomatrixoma,” Microsc. Res. Tech., 68 (2), 75 –79 (2005). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0029 MRTEEO 1059-910X Google Scholar

25. 

M. Polomska et al., “Fourier transform near infrared Raman spectroscopy in studies on connective tissue,” Acta Phys. Pol. A, 118 (1), 136 –140 (2010). ATPLB6 0587-4246 Google Scholar

26. 

M. G. Shim and B. C. Wilson, “Development of an in vivo Raman spectroscopic system for diagnostic applications,” J. Raman Spectrosc., 28 131 –142 (1997). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

27. 

H. G. M. Edwards et al., “NIR-FT Raman spectroscopy as a diagnostic probe for mummified skin and nails,” Vibr. Spectrosc., 28 3 –15 (2002). https://doi.org/http://dx.doi.org/10.1016/S0924-2031(01)00141-2 VISPEK 0924-2031 Google Scholar

28. 

S. M. Ali et al., “Raman spectroscopic analysis of human skin tissue sections ex-vivo: evaluation of the effects of tissue processing and dewaxing,” J. Biomed. Opt., 18 (6), 061202 (2013). https://doi.org/http://dx.doi.org/10.1117/1.JBO.18.6.061202 JBOPFO 1083-3668 Google Scholar

29. 

Z. Huang et al., “Near-infrared Raman spectroscopy for optical diagnosis of lung cancer,” Int. J. Cancer, 107 1047 –1052 (2003). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0215 IJCNAW 1097-0215 Google Scholar

30. 

Z. Huang et al., “Raman spectroscopy in combination with background near-infrared autofluorescence enhances the in vivo assessment of malignant tissues,” Photochem. Photobiol., 81 1219 –1226 (2005). https://doi.org/http://dx.doi.org/10.1562/2005-02-24-RA-449 PHCBAP 0031-8655 Google Scholar

31. 

F. S. Parker, Application of Infrared Raman and Resonance Raman Spectroscopy in Biochemistry, Plenum Press, New York (1983). Google Scholar

32. 

T. M. Greve, K. B. Andersen and O. F. Nielsen, “ATR-FTIR FT- NIR and near-FT-Raman spectroscopic studies of molecular composition in human skin in vivo and pig ear skin in vitro,” Spectroscopy, 22 (6), 437 –457 (2008). SPECET 0887-6703 Google Scholar

33. 

P. R. T. Jess et al., “Early detection of cervical neoplasia by Raman spectroscopy,” Int. J. Cancer, 121 2723 –2728 (2007). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0215 IJCNAW 1097-0215 Google Scholar

34. 

R. K. Dukor, “Vibrational spectroscopy in the detection of cancer,” Handbook of Vibrational Spectroscopy, 3335 –3359 John Wiley & Sons, Ltd, Chichester (2002). Google Scholar

35. 

D. P. Lau et al., “Raman spectroscopy for optical diagnosis in the larynx: preliminary findings,” Lasers Surg. Med., 37 192 –200 (2005). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1096-9101 LSMEDI 0196-8092 Google Scholar

36. 

N. Stone et al., “Raman spectroscopy for identification of epithelial cancers,” Faraday Discuss., 126 141 –157 (2004). https://doi.org/http://dx.doi.org/10.1039/b304992b FDISE6 0301-7249 Google Scholar

37. 

R. E. Kast et al., “Raman spectroscopy can differentiate malignant tumors from normal breast tissue and detect early neoplastic changes in a mouse model,” Biopolymer, 89 134 –141 (2008). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0282 BIPMAA 0006-3525 Google Scholar

38. 

B. P. Gaber and W. L. Peticolas, “On the quantitative interpretation of biomembrane structure by Raman spectroscopy,” Biochim. Biophys. Acta, 465 260 –268 (1977). https://doi.org/http://dx.doi.org/10.1016/0005-2736(77)90078-5 BBACAQ 0006-3002 Google Scholar

39. 

R. Calheiros et al., “Antioxidant phenolic esters with potential anticancer activity: a Raman spectroscopic study,” J. Raman Spectrosc., 39 95 –107 (2008). https://doi.org/http://dx.doi.org/10.1002/jrs.v39:1 JRSPAF 0377-0486 Google Scholar

40. 

G. Shetty et al., “Raman spectroscopy: evaluation of biochemical changes in carcinogenesis of oesophagus,” Br. J. Cancer, 94 1460 –1464 (2006). https://doi.org/http://dx.doi.org/10.1038/sj.bjc.6603102 BJCAAI 0007-0920 Google Scholar

41. 

H. Schultz and M. Baranska, “Identification and qualification of valuable plant substances by IR and Raman spectroscopy,” Vib. Spectrosc., 43 13 –25 (2007). https://doi.org/http://dx.doi.org/10.1016/j.vibspec.2006.06.001 VISPEK 0924-2031 Google Scholar

42. 

R. Agarwal, P. Tandon and V. D. Gupta, “Phonon dispersion in poly(dimethylsilane),” J. Organomet. Chem., 691 2902 –2908 (2006). https://doi.org/http://dx.doi.org/10.1016/j.jorganchem.2006.02.032 JORCAI 0022-328X Google Scholar

43. 

C. Krafft et al., “Near-infrared Raman spectroscopy of human brain lipids,” Spectrochim Acta A Mol. Biomol. Spectrosc., 61 (7), 1529 –1535 (2005). SPACA5 0038-6987 Google Scholar

44. 

G. I. Dovbeshka et al., “FTIR spectroscopy studies of nucleic acid damage,” Talanta, 53 233 –246 (2000). https://doi.org/http://dx.doi.org/10.1016/S0039-9140(00)00462-8 TLNTA2 0039-9140 Google Scholar

45. 

M. Gniadecka et al., “Diagnosis of basal cell carcinoma by Raman spectroscopy,” J. Raman Spectrosc., 28 125 –129 (1997). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

46. 

R. Heckel et al., “Characteristic infrared spectroscopic patterns in the protein bands of human breast cancer tissue,” Vib. Spectrosc., 27 165 –173 (2001). https://doi.org/http://dx.doi.org/10.1016/S0924-2031(01)00134-5 VISPEK 0924-2031 Google Scholar

47. 

S. Sigurdsson et al., “Detection of skin cancer by classification of Raman spectra,” IEEE Trans. Biomed. Eng., 51 1784 –1793 (2004). https://doi.org/http://dx.doi.org/10.1109/TBME.2004.831538 IEBEAX 0018-9294 Google Scholar

48. 

A. Singha et al., “Quantitative analysis of hydrogenated diamond-like carbon films by visible Raman spectroscopy,” J. Appl. Phys., 100 1 –8 (2006). https://doi.org/http://dx.doi.org/10.1063/1.2219983 JAPIAU 0021-8979 Google Scholar

49. 

A. R. Viehoever et al., “Organotypic raft cultures as an effective in vitro tool for understanding Raman spectral analysis of tissue,” Photochem. Photobiol., 78 517 –524 (2003). https://doi.org/http://dx.doi.org/10.1562/0031-8655(2003)078<0517:ORCAAE>2.0.CO;2 PHCBAP 0031-8655 Google Scholar

50. 

H. P. Wang, H. C. Wang and Y. J. Huang, “Microscopic FTIR studies of lung cancer cells in pleural fluid,” Sci. Total Environ., 204 283 –287 (1997). https://doi.org/http://dx.doi.org/10.1016/S0048-9697(97)00180-0 STENDL 0048-9697 Google Scholar

51. 

Y. G. Hu, A. G. Shen and T. Jiang, “Classification of normal and malignant human gastric mucosa tissue with confocal Raman microspectroscopy and wavelet analysis,” Petrochem. Acta A Mol. Biomol. Spectrosc., 69 378 –382 (2008). https://doi.org/http://dx.doi.org/10.1016/j.saa.2007.04.009 Google Scholar

52. 

G. I. Dovbeshka et al., “Surface enhanced IR absorption of nucleic acids from tumor cells: FTIR reflectance study,” Biopolymer, 67 470 –486 (2002). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-0282 BIPMAA 0006-3525 Google Scholar

53. 

Y. Fukuyama et al., “A study on the differences between oral squamous cell carcinimas and normal oral mucosas measured by Fourier transform infrared spectroscopy,” Biospectroscopy, 5 117 –126 (1999). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1520-6343 BIOSFS 1075-4261 Google Scholar

54. 

Q. B. Li et al., “Diagnosis of gastric inflammation and malignancy in endoscopic biopsies based on Fourier transform infrared spectroscopy,” Clin. Chem., 51 346 –350 (2005). https://doi.org/http://dx.doi.org/10.1373/clinchem.2004.037986 CLCHAU 0009-9147 Google Scholar

55. 

J. W. Chan et al., “Micro-Raman spectroscopy detects individual neoplastic and normal hematopoietic cells,” Biophys. J., 90 648 –656 (2006). https://doi.org/http://dx.doi.org/10.1529/biophysj.105.066761 BIOJAU 0006-3495 Google Scholar

56. 

R. J. Lakshimi et al., “Tissue Raman spectroscopy for the study of radiation damage: brain irradiation of mice,” Radiat. Res., 157 175 –182 (2002). https://doi.org/http://dx.doi.org/10.1667/0033-7587(2002)157[0175:TRSFTS]2.0.CO;2 RAREAE 0033-7587 Google Scholar

57. 

T. Richter et al., “Identification of tumor tissue by FTIR spectroscopy in combination with position emission tomography,” Vib. Spectrosc., 28 103 –110 (2002). https://doi.org/http://dx.doi.org/10.1016/S0924-2031(01)00149-7 VISPEK 0924-2031 Google Scholar

58. 

E. O. Faolain et al., “A study examining the effects of tissue processing on human tissue sections using vibrational spectroscopy,” Vib. Spectrosc., 38 121 –127 (2005). https://doi.org/http://dx.doi.org/10.1016/j.vibspec.2005.02.013 VISPEK 0924-2031 Google Scholar

59. 

M. Huleihel et al., “Novel optical method for study of viral carcinogenesis in vitro,” J. Biochem. Biophys. Methods, 50 111 –121 (2002). https://doi.org/http://dx.doi.org/10.1016/S0165-022X(01)00177-4 JBBMDG 0165-022X Google Scholar

60. 

J. A. Bouwstra et al., “Phase behavior of lipid mixtures based on human ceramides: coexistence of crystalline and liquid phases,” J. Lipid Res., 42 (11), 1759 –1770 (2001). JLPRAW 0022-2275 Google Scholar

61. 

F. Damien and M. Boncheva, “The extent of orthorhombic lipid phases in the stratum corneum determines the barrier efficiency of human skin in vivo,” J. Invest. Dermatol., 130 611 –614 (2010). https://doi.org/http://dx.doi.org/10.1038/jid.2009.272 JIDEAE 0022-202X Google Scholar

62. 

F. M. Lyng et al., “Vibrational spectroscopy for cervical cancer pathology from biochemical analysis to diagnostic tool,” Exp. Mol. Pathol., 82 121 –129 (2007). https://doi.org/http://dx.doi.org/10.1016/j.yexmp.2007.01.001 EXMPA6 0014-4800 Google Scholar

63. 

A. N. C. Anigbogu et al., “Fourier transform Raman spectroscopy of interactions between the penetration enhancer dimethyl sulfoxide and human stratum corneum,” Int. J. Pharm., 125 265 –282 (1995). https://doi.org/http://dx.doi.org/10.1016/0378-5173(95)00141-5 IJPHDE 0378-5173 Google Scholar

64. 

L. Knudsen et al., “Natural variations and reproducibility of in vivo near-infrared Fourier transform Raman spectroscopy of normal human skin,” J. Raman Spectrosc., 33 574 –579 (2002). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

65. 

R. Paquin and P. Colomban, “Nanomechanics of single keratin fibres: a Raman study of the α-helixβ-sheet transition and water effect,” J. Raman Spectrosc., 38 504 –514 (2007). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

66. 

N. J. Kline and P. J. Treado, “Raman chemical imaging of breast tissue,” J. Raman Spectrosc., 28 119 –124 (1997). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1097-4555 JRSPAF 0377-0486 Google Scholar

67. 

M. Gniadecka, O. F. Nielsen and H. C. Wulf, “Water content and structure in malignant and benign skin tumours,” J. Mol. Struct., 661–662 405 –410 (2003). https://doi.org/http://dx.doi.org/10.1016/j.molstruc.2003.08.030 JMOSB4 0022-2860 Google Scholar

68. 

N. Nakagawa and M. Matsumoto, and Sakai, “In vivo measurement of the water content in the dermis by confocal Raman spectroscopy,” Skin Res. Technol., 16 137 –141 (2010). https://doi.org/http://dx.doi.org/10.1111/srt.2010.16.issue-2 0909-752X Google Scholar

69. 

S. Fendel and B. Schrader, “Investigation of skin and skin lesions by NIR-FT-Raman spectroscopy,” Fresenius J. Anal. Chem., 360 609 –613 (1998). https://doi.org/http://dx.doi.org/10.1007/s002160050767 FJACES 0937-0633 Google Scholar

70. 

M. Egawa, T. Hirao and M. Takahashi, “In vivo estimation of stratum corneum thickness from water concentration profiles obtained with Raman spectroscopy,” Acta Dermatol. Venereol., 87 (1), 4 –8 (2007). https://doi.org/http://dx.doi.org/10.2340/00015555-0183 ADVEA4 0001-5555 Google Scholar

71. 

S. Olsztynska-Janus et al., “Spectroscopic techniques in the study of human tissues and their components. Part II: Raman spectroscopy,” Acta Bioeng. Biomech., 14 (4), 121 –133 (2012). 1509-409X Google Scholar

72. 

P. J. Caspers et al., “In vivo confocal Raman microspectroscopy of the skin: Non-invasive determination of molecular concentration profiles,” J. Invest. Dermatol., 116 434 –442 (2001). https://doi.org/http://dx.doi.org/10.1046/j.1523-1747.2001.01258.x JIDEAE 0022-202X Google Scholar

73. 

T. Oda et al., “Effect of oral intake of ceramide-containing acetic acid bacteria on skin barrier function,” Antiaging Med., 7 (5), 50 –54 (2010). JAMEF8 1094-5458 Google Scholar

74. 

V. Renugopalakrishnan et al., “Non-uniform triple helical structure in chick skin type I collagen on thermal denaturation: Raman spectroscopic study,” Z. Naturforsch., C53 (5–6), 383 –388 (1998). ZNTFA2 0372-9516 Google Scholar

75. 

M. G. Tosato et al., “Raman spectroscopic investigation of the effects of cosmetic formulations on the constituents and properties of human skin,” Photomed. Laser Surg., 30 85 –91 (2012). https://doi.org/http://dx.doi.org/10.1089/pho.2011.3059 PLDHA8 1549-5418 Google Scholar

76. 

J.-H. Choi and M. Cho, “Calculations of intermode coupling constants and simulations of amide I, II, and III vibrational spectra of dipeptides,” Chem. Phys., 361 168 –175 (2009). https://doi.org/http://dx.doi.org/10.1016/j.chemphys.2009.05.016 CMPHC2 0301-0104 Google Scholar

77. 

M. Gniadecka et al., “Water and protein structure in photoaged and chronically aged skin,” J. Invest. Dermatol., 111 1129 –1132 (1998). https://doi.org/http://dx.doi.org/10.1046/j.1523-1747.1998.00430.x JIDEAE 0022-202X Google Scholar

78. 

K. R. Feingold, “The importance of lipids in cutaneous function,” J. Lipid Res., 48 2529 –2530 (2007). https://doi.org/http://dx.doi.org/10.1194/jlr.E700004-JLR200 JLPRAW 0022-2275 Google Scholar

79. 

E. J. Kim et al., “UV decreases the synthesis of free fatty acids and triglycerides in the epidermis of human skin in vivo, contributing to development of skin photoaging,” J. Dermatol. Sci., 57 19 –26 (2010). https://doi.org/http://dx.doi.org/10.1016/j.jdermsci.2009.10.008 JDSCEI 0923-1811 Google Scholar

80. 

R. Ghadially et al., “The aged epidermal permeability barrier: Structural, functional, and lipid biochemical abnormalities in humans and a senescent murine model,” J. Clin. Invest., 95 2281 –2290 (1995). https://doi.org/http://dx.doi.org/10.1172/JCI117919 JCINAO 0021-9738 Google Scholar

81. 

S. Krimm and J. Bandekar, “Vibrational spectroscopy and conformation of peptides, polypeptides, and proteins,” Adv. Protein Chem., 38 181 –365 (1986). https://doi.org/http://dx.doi.org/10.1016/S0065-3233(08)60528-8 APCHA2 0065-3233 Google Scholar

82. 

A. M. Herrero, “Raman spectroscopy a promising technique for quality assessment of meat and fish: a review,” Food Chem., 107 1642 –1651 (2008). https://doi.org/http://dx.doi.org/10.1016/j.foodchem.2007.10.014 FOCHDJ 0308-8146 Google Scholar

83. 

E. C. Y. Li-Chan, “The applications of Raman spectroscopy in food science,” Trends Food Sci.Technol., 7 361 –370 (1996). https://doi.org/http://dx.doi.org/10.1016/S0924-2244(96)10037-6 TFTEEH 0924-2244 Google Scholar

84. 

M. Bouraoui, S. Nakai and E. C. Y. Li-Chan, “In situ investigation of protein structure in Pacific whiting surimi and gels using Raman spectroscopy,” Food Res. Int., 30 65 –72 (1997). https://doi.org/http://dx.doi.org/10.1016/S0963-9969(97)00020-3 FORIEU 0963-9969 Google Scholar

85. 

A. J. P. Alix, G. Pedanou and M. Berjot, “Determination of the quantitative secondary structure of proteins by using some parameters of the Raman amide I band,” J. Mol. Struct., 174 159 –164 (1988). https://doi.org/http://dx.doi.org/10.1016/0022-2860(88)80151-0 JMOSB4 0022-2860 Google Scholar

86. 

J.-H. Shao et al., “Evaluation of structural changes in raw and heated meat batters prepared with different lipids using Raman spectroscopy,” Food Res. Int., 44 2955 –2961 (2011). https://doi.org/http://dx.doi.org/10.1016/j.foodres.2011.07.003 FORIEU 0963-9969 Google Scholar

87. 

P. J. Caspers et al., “In vitro and in vivo spectroscopy of human skin,” Biospectroscopy, 4 S31 –S39 (1998). https://doi.org/http://dx.doi.org/10.1002/(ISSN)1520-6343 BIOSFS 1075-4261 Google Scholar

88. 

A. Tfayli et al., “Discriminating nevus and melanoma on paraffin-embedded skin biopsies using FTIR microspectroscopy,” Biochim. Biophys. Acta, 1724 262 –269 (2005). https://doi.org/http://dx.doi.org/10.1016/j.bbagen.2005.04.020 BBACAQ 0006-3002 Google Scholar

89. 

F. H. Silver, J. W. Freeman and D. DeVore, “Viscoelastic properties of human skin and processed dermis,” Skin Res. Technol., 7 18 –23 (2001). https://doi.org/http://dx.doi.org/10.1034/j.1600-0846.2001.007001018.x 0909-752X Google Scholar

90. 

V. N. Uversky, C. J. Oldfield and A. K. Dunker, “Intrinsically disordered proteins in human diseases: introducing the D2 concept,” Annu. Rev. Biophys., 37 215 –246 (2008). https://doi.org/http://dx.doi.org/10.1146/annurev.biophys.37.032807.125924 ARBNCV 1936-122X Google Scholar

91. 

A. L. Fink, “Protein aggregation: folding aggregates, inclusion bodies and amyloid,” Folding Des., 3 (1), R9 –R23 (1998). https://doi.org/http://dx.doi.org/10.1016/S1359-0278(98)00002-9 FODEFH 1359-0278 Google Scholar

92. 

P. Leandro and C. M. Gomes, “Protein misfolding in conformational disorders: rescue of folding defects and chemical chaperoning,” Mini Rev. Med. Chem., 8 901 –911 (2008). https://doi.org/http://dx.doi.org/10.2174/138955708785132783 1389-5575 Google Scholar

93. 

E. Ly et al., “Polarized Raman microspectroscopy can reveal structural changes of peritumoral dermis in basal cell carcinoma,” Appl. Spectrosc., 62 1088 –1094 (2008). https://doi.org/http://dx.doi.org/10.1366/000370208786049187 APSPA4 0003-7028 Google Scholar

94. 

S. Federman, L. M. Miller and I. Sagi, “Following matrix metalloproteinases activity near the cell boundary by infrared micro-spectroscopy,” Matrix Biol., 21 567 –577 (2002). https://doi.org/http://dx.doi.org/10.1016/S0945-053X(02)00089-6 MTBOEC 0945-053X Google Scholar

95. 

B. D. McKersie and J. E. Thompson, “Lipid crystallization in senescent membranes from cotyledons,” Plant Physiol., 59 803 –807 (1977). https://doi.org/http://dx.doi.org/10.1104/pp.59.5.803 PLPHAY 0032-0889 Google Scholar

96. 

D. M. Small, “Progression and regression of atherosclerotic lesions. Insights from lipid physical biochemistry,” Arterioscler. Thromb. Vasc. Biol., 8 103 –129 (1988). https://doi.org/http://dx.doi.org/10.1161/01.ATV.8.2.103 ATVBFA 1079-5642 Google Scholar

97. 

K. Matsuzaki, “Physicochemical interactions of amyloid β-peptide with lipid bilayers,” Biochem. Biophys. Acta Biomembr., 1768 (8), 1935 –1942 (2007). https://doi.org/http://dx.doi.org/10.1016/j.bbamem.2007.02.009 0005-2736 Google Scholar

99. 

H. M. Garvin et al., “Developments in forensic anthropology: age-at-death estimation,” 10 in A Companion to Forensic Anthropology, 202 –223 1st ed.Wiley-Blackwell Publishing Ltd., Chichester, West Sussex, United Kingdom (2012). Google Scholar

100. 

P. H. Whitehead, “Biochemical techniques in forensic science,” Trends Biochem. Sci., 10 (8), 299 –302 (1985). https://doi.org/http://dx.doi.org/10.1016/0968-0004(85)90167-7 TBSCDB 0167-7640 Google Scholar

101. 

S. Ritz, A. Turzynski and H. W. Schutz, “Estimation of age at death based on aspartic acid racemization in noncolagenous bone proteins,” Forensic Sci. Int., 69 (2), 149 –159 (1994). https://doi.org/http://dx.doi.org/10.1016/0379-0738(94)90251-8 FSINDR 0379-0738 Google Scholar

102. 

P. M. Helfman and J. L. Bada, “Aspartic acid racemization in tooth enamel from living humans,” Proc. Nat. Acad. Sci. U. S. A., 72 (8), 2891 –2894 (1975). https://doi.org/http://dx.doi.org/10.1073/pnas.72.8.2891 1091-6490 Google Scholar

103. 

E. Baccino and A. Schmitt, “Determination of adult age at death in the forensic context,” Forensic Anthropology and Medicine: Complementary Sciences from Recovery to Cause of Death, 259 –280 Humana Press Inc., Totowa, New Jersey, United States (2006). Google Scholar

104. 

S. I. Kvaal et al., “Age estimation of adults from dental radiographs,” Forensic Sci. Int., 74 175 –185 (1995). https://doi.org/http://dx.doi.org/10.1016/0379-0738(95)01760-G FSINDR 0379-0738 Google Scholar

105. 

T. H. Hsiao, H. P. Chang and K. M. Liu, “Sex determination by discriminant function analysis of lateral radiographic cephalometry,” J. Forensic Sci., 41 (5), 792 –795 (1996). JFSCAS 0022-1198 Google Scholar

106. 

A. H. Kuptsov, “Applications of Fourier transform Raman spectroscopy in forensic science,” J. Forensic Sci., 39 (2), 305 –318 (1994). JFSCAS 0022-1198 Google Scholar

107. 

M. Clayboum and M. Ansell, “Using Raman spectroscopy to solve crime: inks, questioned documents and fraud,” Sci. Justice, 40 (4), 261 –271 (2000). https://doi.org/http://dx.doi.org/10.1016/S1355-0306(00)71996-4 SJUSFE 1355-0306 Google Scholar

108. 

A. Osmani, O. Gamulin and M. Vodanovic, “Age estimation of teeth with Raman spectrometry: preliminary study,” Bull. Int. Assoc. Paleodontol., 8 (1), 137 –143 (2014). 1846-6273 Google Scholar

109. 

P. Tramini et al., “A method of age estimation using Raman microspectrometry imaging of the human dentin,” Forensic Sci. Int., 118 (1), 1 –9 (2001). https://doi.org/http://dx.doi.org/10.1016/S0379-0738(00)00352-2 FSINDR 0379-0738 Google Scholar

110. 

F. Kosa, A. Antal and I. Farkas, “Electron probe microanalysis of human teeth for the determination of individual age,” Med. Sci. Law, 30 (2), 109 –114 (1990). MDSLA6 0025-8024 Google Scholar

111. 

S. Strasser et al., “Age determination of blood spots in forensic medicine by force spectroscopy,” Forensic Sci. Int., 170 8 –14 (2007). https://doi.org/http://dx.doi.org/10.1016/j.forsciint.2006.08.023 FSINDR 0379-0738 Google Scholar

112. 

K. L. Spalding et al., “Forensics: age written in teeth by nuclear tests,” Nature, 437 333 –334 (2005). https://doi.org/http://dx.doi.org/10.1038/437333a NATUAS 0028-0836 Google Scholar

113. 

B. A. Buchholz and K. L. Spalding, “Year of birth determination using radiocarbon dating of dental enamel,” Surf. Interfaces Anal., 42 398 –401 (2010). https://doi.org/http://dx.doi.org/10.1002/sia.v42:5 SIANDQ 0142-2421 Google Scholar

114. 

W. M. Holleran et al., “Serine-palmitoyl transferase activity in cultured human keratinocytes,” J. Lipid Res., 31 (9), 1655 –1661 (1990). JLPRAW 0022-2275 Google Scholar

115. 

W. M. Holleran et al., “Regulation of epidermal sphingolipid synthesis by permeability barrier function,” J. Lipid Res., 32 (7), 1151 –1158 (1991). JLPRAW 0022-2275 Google Scholar

116. 

W. M. Holleran et al., “Sphingolipids are required for mammalian epidermal barrier function: inhibition of sphingolipid synthesis delays barrier recovery after acute penetration,” J. Clin. Invest., 88 1338 –1345 (1991). https://doi.org/http://dx.doi.org/10.1172/JCI115439 JCINAO 0021-9738 Google Scholar

Biography

Giuseppe Pezzotti is full tenured professor and leader of the Ceramic Physics Laboratory at the Kyoto Institute of Technology, Japan. He graduated summa cum laude in mechanical engineering from Rome University “La Sapienza,” Italy, he holds three doctoral degrees in materials engineering, solid state physics, and medical sciences, and is author/coauthor of 570 scientific papers, 1 book, 13 book chapters, and 8 patents, including a world patent on nanoscale stress microscopy in the scanning electron microscope.

Marco Boffelli is a third year PhD student in material science at Kyoto Institute of Technology. He joined professor Pezzotti’s research group in 2012 after receiving his diploma in material science from the Ca' Foscari University in Venice in the same year. His research interest includes ceramic biomaterials synthesis and characterization mainly with Raman spectroscopy, scanning electron microscopy, and cathodoluminescence.

Daisuke Miyamori is a lecturer in the Department of Forensic Medicine, Kyoto Prefectural University of Medicine (KPUM). He graduated from Kyushu University. He was working at the emergency department of Shonankamakura Tokushukai Hospital before he got his position at KPUM. He is a licensed medical doctor and specialized in emergency medicine.

Takeshi Uemura is a lecturer in the Department of Forensic Medicine, Kyoto Prefectural University of Medicine (KPUM). He graduated from Toho University and the Graduate School of Pharmaceutical Science, Chiba University. He was working at Arizona University before he got his position at KPUM. He has a pharmacist license. He has been doing his main research about polyamine. Here, he is studying the relationship between polyamine and age.

Yoshinori Marunaka, MD, PhD, is professor and chairperson in the Departments of Molecular Cell Physiology and Bio-Ionomics, Kyoto Prefectural University of Medicine, Japan. He is director and professor at the Japan Institute for Food Education and Health and St. Agnes’ University. He is the president-elect of the Physiological Society of Japan, and Editor-in-Chief of the Journal of Physiological Sciences. His current research focuses on roles of H+ and Cl- in cell function, including cell growth, neurite elongation, diabetes mellitus etc.

Wenliang Zhu got his PhD in material sciences in Kyoto Institute of Technology in 2005 and is now a lecturer in the Medical School of Osaka University. His main research field is on the development and application of high resolution photo-/electrostimulated spectroscopies for quantitative analysis of structural, mechanical, and chemical properties in single crystalline and polycrystalline ceramics.

Hiroshi Ikegaya is a professor and chairman of the Department of Forensic Medicine, Kyoto Prefectural University of Medicine (KPUM). He graduated from Miyazaki Medical College and Graduate School of Medicine, The University of Tokyo. He was working at National Research Institute of Police Science before he got his position at KPUM. He has a medical doctorate and architect license. He has been doing his main research about development of individual identification methods.

© 2015 Society of Photo-Optical Instrumentation Engineers (SPIE) 1083-3668/2015/$25.00 © 2015 SPIE
Giuseppe Pezzotti, Marco Boffelli, Daisuke Miyamori, Takeshi Uemura, Yoshinori Marunaka, Wenliang Zhu, and Hiroshi Ikegaya "Raman spectroscopy of human skin: looking for a quantitative algorithm to reliably estimate human age," Journal of Biomedical Optics 20(6), 065008 (25 June 2015). https://doi.org/10.1117/1.JBO.20.6.065008
Published: 25 June 2015
JOURNAL ARTICLE
18 PAGES


SHARE
Advertisement
Advertisement
Back to Top