Article pubs.acs.org/Biomac
Observations of Multiscale, Stress-Induced Changes of Collagen Orientation in Tendon by Polarized Raman Spectroscopy Admir Masic,† Luca Bertinetti,† Roman Schuetz,†,‡ Leonardo Galvis,†,§ Nadya Timofeeva,†,§ John W. C. Dunlop,† Jong Seto,∥ Markus A. Hartmann,⊥ and Peter Fratzl*,† †
Department of Biomaterials, Max-Planck-Institute of Colloids and Interfaces, Research Campus Golm, 14424 Potsdam, Germany BAM Federal Institute for Materials Research and Testing, Unter den Eichen 87, 12205 Berlin, Germany § Berlin-Brandenburg School for Regenerative Therapies (BSRT), Charité Campus Virchow-Klinikum, Augustenburger Platz 1, 13353 Berlin, Germany ∥ Department of Chemistry, Universität Konstanz, POB 714, 78457 Konstanz, Germany ⊥ Institut für Physik, Montanuniversität Leoben, Franz-Josef Strasse 18, A-8700 Leoben, Austria ‡
ABSTRACT: Collagen is a versatile structural molecule in nature and is used as a building block in many highly organized tissues, such as bone, skin, and cornea. The functionality and performance of these tissues are controlled by their hierarchical organization ranging from the molecular up to macroscopic length scales. In the present study, polarized Raman microspectroscopic and imaging analyses were used to elucidate collagen fibril orientation at various levels of structure in native rat tail tendon under mechanical load. In situ humidity-controlled uniaxial tensile tests have been performed concurrently with Raman confocal microscopy to evaluate strain-induced chemical and structural changes of collagen in tendon. The methodology is based on the sensitivity of specific Raman scattering bands (associated with distinct molecular vibrations, such as the amide I) to the orientation and the polarization direction of the incident laser light. Our results, based on the changing intensity of Raman lines as a function of orientation and polarization, support a model where the crimp and gap regions of collagen hierarchical structure are straightened at the tissue and molecular level, respectively. However, the lack of measurable changes in Raman peak positions throughout the whole range of strains investigated indicates that no significant changes of the collagen backbone occurs with tensing and suggests that deformation is rather redistributed through other levels of the hierarchical structure.
■
information, contributing to an understanding of the role of molecular level features to tissue behavior. Polarized-light microscopy was one of the earliest physical techniques to be used to study the morphology of collagen-rich tissues. Recently, second harmonic generation (SHG) microscopy, a technique that exploits nonlinear optical properties of molecules for imaging,15 became one of the main techniques to display morphology and therefore orientation of collagen in tissues.16 PR has an advantage with respect to most of the other advanced optical methods to evaluate collagen orientation that lies in its reliable chemical fingerprint that is acquired in real time. Such information allows studies associated with molecular and atomic level properties that are often fundamental in material behavior. Recently, potential uses of the PR approach in determination of collagen orientation within tissue have been reported.17−20 However, conventional single point PR spectroscopy is inadequate to describe the chemical information
INTRODUCTION
The collagen molecule is a fundamental structural building block for various types of natural tissues. 1,2 Hierarchical structuring from molecular to tissue levels allows collagen to fulfill a variety of mechanical functions, particularly in vertebrates. It is a major constituent of tendons and ligaments as well as the organic matrix of bone and dentin and is also present in skin and arteries. In all the aforementioned biological materials, the collagen orientation plays a fundamental role in the overall mechanical properties of the tissue,3−5 and it may be altered in various diseases such as skin cancer,6 keratoconus of the cornea,7 or osteogenesis imperfecta.8 The mechanical properties of different types of collagencontaining tissues have been extensively investigated on various hierarchical levels of structure. 9−14 However, a lack of experimental methodologies able to assess the molecular scale in situ prevented the understanding on how this level of structure contributes to the macroscopic mechanical behavior. In this context, polarized Raman (PR) spectroscopic and imaging techniques can provide several length scales of © 2011 American Chemical Society
Received: July 21, 2011 Revised: September 22, 2011 Published: September 28, 2011 3989
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
contains the original sequence of the human collagen III (Figure 1B,C). In this study, Raman microspectroscopic and imaging analyses were used to elucidate collagen fibril orientation and behavior under uniaxial tension. A significant experimental result reported here is the possibility to image collagen fibril orientation directly in situ. By combining PR microspectroscopy with a uniaxial micromechanical stretching device, changes in orientation and interaction of RTT collagen at both the molecular and the tissue scale can be monitored during tissue tensing.
■
MATERIALS AND METHODS
Rat Tail Tendon (RTT). Fascicles of approximately 20 mm in length and 200 μm in thickness were dissected from the proximal end of the tails of male Wistar rats aged between 4 and 8 months. Raman Spectroscopy. For Raman microspectroscopy, a continuous laser beam was focused down to a micrometer sized spot on the sample through a confocal Raman microscope (CRM200, WITec, Ulm, Germany) equipped with a piezo-scanner (P-500, Physik Instrumente, Karlsruhe, Germany). The diode-pumped 785 nm near-infrared (NIR) laser excitation (Toptica Photonics AG, Graefelfing, Germany) was used in combination with a water immersed 60× (Nikon, NA = 1.0) microscope objective. For in situ mechanical tests, a 20× (Nikon, NA = 0.4) objective was used. The linearly polarized laser light was rotated using a half-wave plate and scattered light was filtered introducing a further polarizer (analyzer) before the confocal microscope pinhole. The depolarization caused by the experimental setup is negligible.43 The spectra were acquired using a CCD (PI-MAX, Princeton Instruments Inc., Trenton, NJ) behind a grating (300 g mm−1) spectrograph (Acton, Princeton Instruments Inc., Trenton, NJ) with a spectral resolution of ∼6 cm−1. Thirty accumulations with integration time of 1 s were used for single spot analyses. For mapping purposes the surface was scanned with steps of 1 μm integrating the signal for 0.1 s. Microtensile Tests. In situ micro-Raman stress−strain experiments were performed using a custom-built microtensile tester with a 50 N load cell integrated into a sealed chamber allowing for controlled relative humidity (RH) and temperature. A quartz window (SPI Supplies, West Chester, PA) installed on one of the sides allowed the transmission of the light used for Raman microspectroscopy, and an inlet and an outlet port allowed the stream of moisture produced by a WETSYS humidity generator flowing in the chamber. To keep the temperature constant during the experiments, water was circulated from a constant temperature bath (∼30 °C) into the walls of the chamber. Temperature and RH were monitored throughout the experiments. A wet RTT with an effective strain length of 10 mm was clamped by pressure bars and fixed in the tensile tester chamber, preequilibrated with 95% RH to avoid sample dehydration during the measurements. To strain the sample, a displacement rate of 1 μm s−1 was used. The stresses were calculated from the load cell data using the cross-sectional area of the tendon measured with confocal microscope at the beginning of the experiment. Data Processing. The ScanCtrlSpectroscopyPlus (version 1.38, WITec, Ulm, Germany) and WitecProjectPlus (version 2.02, WITec, Ulm, Germany) were used for the experimental setup and spectral data processing, respectively. Chemical images were achieved by integration over defined Raman shift regions in the spectrum using a sum filter. The filter calculates the intensities within the chosen borders and the background is subtracted by taking the baseline from the first to the second border. The amide I intensity was obtained by integrating the total intensity of the amide I band (1600−1700 cm−1). The Raman orientation map was produced by a nonlinear least-squares fitting procedure provided by Matlab 7.5 (MathWorks Inc., Natick, MA) using built-in and locally written scripts. The equation
Figure 1. (A) Hierarchical structure of RTT collagen. (B) Molecular structure of protein-like peptide molecule 1BKV (name code in Protein Data Bank (PDB))42 visualized with PyMOL. Amino acids are represented by nitrogen (blue), carbon (green), and oxygen (red) whereas hydrogen atoms are not shown. (C) Amino acid chains of the same peptide as in (B) represented now only by CO backbone bonds (green: carbon; red: oxygen).
and orientation distribution in relation to the macroscopic scale of tissues. Gamsjäger and co-workers as well as Kazanci and coworkers demonstrated the use of PR imaging techniques in determining orientation and composition in cortical bone tissue.21−23 These results demonstrate the utility of PR imaging in studying complex biological architectures at the tissue level. Rat tail tendon (RTT) is composed entirely of type I collagen fibrils with small amounts of proteoglycans. 24 The organization from the molecular to the tissue length scales of collagen (Figure 1A) shows a high degree of alignment.9 Of specific interest is the existence of a macroscopic crimp structure which has its origins at the molecular scale. 25,26 The crimp corresponds to regions of the collagen fibrils where molecular kinking occurs. This is due to a realignment of the fibril orientation that maximizes interfibrillar interactions and releases intrafibrillar strains from the helical arrangement. 27 At the level of the tissue, a sharp change in the orientation of collagen fibrils can even be observed by optical microscopy. In a uniaxial tensile experiment, a macroscopic crimp is the first structural motif involved in dissipation of applied stress, giving rise to the “toe” region of the stress−strain curve.28 At strains beyond ∼3%, the stiffness of rat tail tendon increases considerably with deformation (heel region) as the molecular kinks in the gap are straightened.29−31 When tendon is stretched beyond the heel region of the stress/strain curve, a gliding of molecules takes place before final failure of the tendon occurs.13,31 The orientation of one molecular unit within each macromolecule can be derived from Raman band anisotropy measurements, assuming that Raman band tensors of the relevant molecular unit are already known.32−34 From the parameters provided by the Raman tensor, the arrangement or orientation of the molecular structure with respect to the biological system can be determined.35−40 Scattering of the amide I band is more intense in the direction which is perpendicular to the collagen fibril orientation41 compared to the direction parallel to the fibril because the collagen carbonyl groups are mainly oriented perpendicular to the collagen chain, as can be observed in the molecular visualization of the collagen-like peptide 1BKV (Protein Data Bank code) which
(1) 3990
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
Figure 2. CO angle distribution (abscissa, deg) of collagen-like peptides derived from PDB with respect to the molecular axis. Angular distribution of CO in peptide 3DMW (green line),52 1BKV (blue line), 1QSU (gray line),53 1CAG (cyan),54 1K6F (magenta), 1K6F_1 (yellow),55 and 1WZB (black)56 molecule. The curves are fitted by a Gaussian distribution: red thick line. The Gaussian fit takes into account cylindrical symmetry of the collagen molecule and indicates angular distribution of CO bonds centered approximately at 90° with a full width at half-maximum of 30°. where I is the amide I intensity response, a the mean intensity, b the amplitude of fitting/mean intensity, x the polarization angle of the laser (radians), and c the phase shifting (radians), was used. The fitting parameters in Figure 4 are plotted as vectors. Their lengths represent b (ratio amplitude/mean intensity), while their direction correspond to the calculated fitted angle of the collagen fibril, the phase shift c. The color code in Figure 4 corresponds to the mean intensity, a.
■
function. However, this approximation can differ significantly from the real distribution function and in some case might take negative values. According to Pottel et al.,49 a more realistic distribution function can be obtained from any set of Pl values maximizing the information entropy of the distribution, subjected to the constraint that f(θ) is always positive and that its integral over the range [0, π] is equal to 1 (i.e., requiring that f(θ) is a probability distribution). The function satisfying these constraints is called the most probable distribution function (f(θ)mp) that is given by
THEORETICAL BASIS
The theoretical background of orientation measurements by Raman microspectroscopy in polymer fibers has been extensively described in several papers in the past decades, 44−46 and the theory has been applied to synthetic polymers as well as to biological systems.34,39,47 Here, only the basic concepts are recalled for the convenience of the reader. For a system with fiber-type symmetry (i.e., cylindrical symmetry), such as collagen fibrils found in rat tail tendon, the orientation distribution function of the molecular moieties can be represented by a series of even Legendre polynomials:
(3)
when the series is limited to the first two polynomials. The Lagrange multipliers λ 2 and λ 4 can be calculated numerically considering that the average values of each Legendre polynomial weighted with f(θ)mp must be equal to the corresponding order parameter obtained by the PR measurements. Here, we focused on the orientation of the CO bonds, and we considered the Raman tensor of the CO vibration as cylindrically symmetric. However, the orientation parameters of the collagen molecule can also be evaluated, provided that the angular relationships between the collagen axis and the CO oscillators is known, using the addition theorem for Legendre polynomials,50,51 that is
(2)
where Pl00(cos θ) are the Legendre polynomials and ⟨Pl00⟩ are the coefficients of the series, also called order parameters (in the following referred to as ⟨Pl00⟩ = Pl, l = 2, 4), that represent average values of the Legendre polynomials over the orientation distribution of scattering centers. θ is the angle between the principal axis of the Raman tensor (α) for the considered vibrational mode and the symmetry axis of the sample (in this case the direction of the fiber). P2 and P4 values can be determined by polarized microRaman spectroscopy following the procedures described by Bower,45 taking into account the depolarization effect, as described by Turrel,48 when objectives with high numerical aperture are used. The knowledge of the P2 and P4 values leads to a Legendre series expansion truncated at the second even term (P 4) providing a first approximation of the orientation distribution
(4)
where Pl,CM, Pl,CO, and Pl,CO/CM are the order parameter of the orientation distributions of: collagen molecules with respect to the fiber axis, CO oscillators with respect to the fiber axis, and CO oscillators with respect to the collagen molecule, respectively. While Pl,CO could be determined experimentally, we performed calculations using available collagen structures from the Protein Data Bank (PDB, see Figure 2) to get P l,CO/CM . The distribution of CO angles has been investigated for different collagen-like peptides based on the 3991
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
X-ray structures taken from the PDB. CO angle distributions of various collagen-like peptides with respect to the main molecular axis are shown in Figure 2. The single Gaussian fitting curve (red thick line) shows that CO oscillators lie almost perpendicular to the collagen molecules in the range from 60° to 120°. In order to calculate the orientation distribution function using the Legendre addition theorem, we hypothesized a Gaussian CO distribution centered at 90° with a full width at half-maximum of 30°. This gives rise to a P2,CO/CM of −0.43 and P4,CO/CM of 0.23. The orientation distribution function of CO with respect to the collagen molecule was considered fixed during the experiments.
■
RESULTS AND DISCUSSION
PR of RTT. Simultaneous uniaxial tensile stretching and PR spectroscopic and imaging analysis of RTT fascicles were performed in a humidity-controlled environment. A schematic
Figure 4. In situ polarized Raman mapping of the collagen fiber orientation in unstretched fully hydrated RTT. (A) Optical microscopy image of the analyzed region where the crimp structure of collagen is visible. (B) Map obtained by fitting 13 Raman images collected with different polarization angles of the incident laser light. The direction of arrows indicates the orientation of collagen fibers, their length represents the amplitude of the fitting curve, and the color code represents the average intensity of the amide I band (parameter a). (C, D, E) Magnified regions of interest reveal specific structural characteristics at the level of tissue. Note the change in collagen fiber orientation in (D) corresponding to the crimp (at about 50 μm). (F) Frequency plot of the c parameter with respect to the rat tail fiber direction z.
Figure 3. (A) A schematic drawing of the experimental setup used for the PR measurements. (B) Raman spectra taken on stretched RTT oriented parallel (red) and perpendicular (blue) to the polarization of incident laser beam (no analyzer was used). (C) The same experiment as in (B) with the analyzer inserted in the beam path. Signals from cross-polarized experiments are not shown. (D) The same experiment as in (C) on thermally denatured RTT. No amide I anisotropy can be observed.
sketch of the experimental setup is presented in Figure 3A. The tensile tester chamber was placed under the Raman confocal microscope, and the sample was stretched to the desired strain level. In Figure 3B, spectra taken on stretched RTT (5% tissue strain, 95% RH) were collected with two different laser polarization orientations [parallel (z, red line) and perpendicular (x , blue line) to the fascicle axis]. The band at 1665 cm −1 is related to the amide I collagen vibration (mainly CO stretching of the protein backbone), whereas other bands at ∼1450 and ∼1250 cm−1 can be assigned to C−H bending and amide III vibrations, respectively. Prominent differences in intensities of most Raman bands (especially amide I) in the two different laser-to-fiber configurations were found. The results for the same experiment, but now filtering the scattered light by
introducing an analyzer (see Figure 3A) in the optical path is shown in Figure 3C. A similar anisotropy in the scattered intensity compared to Figure 3B was observed. The anisotropy of Raman bands has its origin in the preferential orientation of vibrating molecular fragments with respect to the fiber axis and incident light polarization orientation. The higher intensity of the amide I band in the direction perpendicular to the fiber axis (x) is in agreement with results reported in the literature.17,41 To investigate the effects of denaturation of collagen (thermal disruption of the collagen structure) on PR spectra, the RTT sample was thermally denatured by exposing it to an elevated temperature of 80 °C for 2 h. The results are shown in Figure 3D. No amide 3992
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
I anisotropy can be observed, which confirms a random orientation of CO groups in the sample. Mapping of Collagen Orientation. Conventional single point Raman spectroscopy is often inadequate to describe the chemical information and orientation distribution in relation to the macroscale. The intensity of the amide I band is observed to be sinusoidal in relation to the orientation of the incident laser polarization17 and can be fitted using eq 1. The phase shift c represents the orientation of the collagen fibrils in the analyzed spot. The average values a and amplitude b are related to the total amount of amide I band generating molecules as well as the out-of-plane orientation of the collagen fibrils. In principle, the three parameters a, b, and c could be used to extract information on the three-dimensional orientation of collagen fibrils, if the collagen concentration in the interaction volume of the beam is known. However, this is beyond the scope of the present paper, and we concentrate here on the situation within the plane of observation only, where b describes the degree of alignment within this plane and c the predominant orientation. The parameter a, finally, is larger when more of the collagen fibrils are oriented out of the plane. Raman spectroscopic measurements of the crimp section of unstretched totally hydrated RTT collagen were performed. A region of 100 × 10 μm was mapped using the scanning mode with 13 different polarizer orientations. No analyzer was placed in the scattered light path. Total measurement time was 1 h. The obtained data at each point were fitted using eq 1 to determine a, b, and c as a function of position. Figure 4B shows the map of the calculated angle c (orientation of black lines) around the crimp. The phase shift c is related to the maximum amide I intensity that is assumed to arise from the CO groups that are mainly distributed at 90° with respect to the molecular axis (see Figure 2). Therefore, orientation of the black lines in the laboratory reference frame reflects the average local orientation of the collagen fibrils in each point of the map. The amplitude variations are shown by the length of the lines, whereas the color scale indicates the average intensity of the amide I band (parameter a). The calculated collagen orientation map is in good agreement with the fiber directions seen using optical microscopy (Figure 4A). Enlargements of some regions of interest are shown in Figure 4C,D,E. The c parameter can be used to calculate the angular distribution function of the average orientation of collagen fibrils with respect to the rat tail fiber direction (Figure 4F) in the mapped area. The graph in Figure 4F clearly shows two different populations of collagen fibrils around the crimp. There is also a small distribution centered around 5° relative to the inversion of orientation around 50 μm in Figure 4B. Uniaxial Tensile Stretching of RTT. The crimp in RTT can be straightened by applying a tensile load in the direction along the main axis of the collagen fiber in tendon.26 The alignment of collagen fibers is easily monitored by optical microscopy.7 In Figure 5, representative stress and strain vs time curves (A) and corresponding stress−strain curve (B) obtained from tensile experiment are shown. Raman spectra (indicated by blue dots in Figure 5A,B) were collected after a certain time necessary for stress relaxation (clearly visible in Figure 5A). The stress−strain curve in Figure 5B is in agreement with data reported in the literature31 and clearly shows four characteristic regions, namely toe, heel, linear, and postlinear. The tangent modulus in the linear deformation regime lie
Figure 5. In situ tensile experiment. (A) Stress and strain curves versus time registered during the tensile experiment and (B) corresponding engineering stress−strain curve. Time/strain positions where Raman spectra were collected are indicated with blue dots.
Figure 6. Isotropic Raman spectra acquired in three characteristic regions of the stress−strain curve of RTT.
Table 1. Order Parameters Derived from in Situ PR Tensile Testing of RTTa sample gelatin RTT toe RTT heel RTT linear RTT postlinear
description boiled RTC 0% 5% 8% 11.5%
P2 (amide I)
P4 (amide I)
−0.06
0.067
−0.35 −0.37 −0.37 −0.36
0.11 0.13 0.14 0.15
P2 (collagen)
P4 (collagen)
0.77 0.81 0.86 0.83
0.45 0.50 0.61 0.65
a
From these order parameters the shape of the associated most probable distribution function (f mp) can be deduced. We calculated f mp for three characteristic regions of the stress strain curve of RTT and found f mp to be identical for the toe and heel regions (Figure 7, inset A). 3993
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
Figure 7. Schematic plot of a characteristic stress−strain curve relative to the uniaxial tensile deformation of RTT. Inset (A) shows molecular level reorientation of collagen molecules that occurs only in the linear region of the stress−strain curve. Inset (B) shows tissue level changes in orientation of collagen derived from PR imaging that are mainly related to the straightening of the crimp.
between 1 and 2 GPa, in agreement with the values reported in the literature.57,58 In the heel region, the measured values (0.5− 1 GPa) also match those obtained by atomistic modeling for single wet microfibrils.59 They are both significantly lower than what estimated for collagen molecules from atomistic models and Brillouin light scattering (between 4 and 9 GPa), suggesting that the deformation of collagen fibers is distributed to the different hierarchical levels and cannot result only from the stretching of the single molecules. To confirm this hypothesis, we monitored the changes of the vibrational modes of the collagen molecule through the entire tensile experiment. Raman isotropic spectra registered within toe, heel, and linear region of the stress−strain curve (blue dots in Figure 5B) are shown in Figure 6. Isotropic spectra were calculated following the procedure reported elsewhere.60 These spectra are characterized by identical peak position as well as peak intensities, suggesting that the molecular vibrational units of the backbone of collagen are not perturbed in the entire range of strains applied. For these strain ranges significant spectral changes are detected for other macromolecular constituents of natural materials, as for example cellulose in wood61 and silk.62 In cellulose, the peak due to the C−O−C bonding is observed to shift when the tissue is strained, indicating that the glycosidic backbone is loaded. In silk several bands associated with the peptide bonds in β-sheets undergo changes when the fiber is stretched,
indicating an alteration in force constants and/or dihedral changes of the bonds involved in the polypeptide chains. The fact that the peaks of the collagen molecules are not shifting during the tensile test may be due to the low stress level attained for the collagen fibers in this case (1 order of magnitude lower than in the aforementioned cases), but it also suggests that bonding environment in the triple helix remains unchanged. From the same data collection used to produce isotropic spectra, we calculated P2 and P4 order parameters related to the amide I band (Table 1) angular distribution. Negative values of P2 indicate the transverse orientation of amide I vibrational unit (mainly stretching of CO) with respect to the main axis of the collagen molecule and fiber. Applying the Legendre addition theorem for nested distributions, these values can be used to calculate P2 and P4 order parameters of the distribution of collagen molecules with respect to the fiber axis in the analyzed spot (Table 1). However, when the strain reached the linear region of the curve, a sharpening of the distribution f mp was observed, due to the alignment of the collagen molecules compared to toe and heel region. Results from PR are in good agreement with X-ray scattering results and atomistic simulations indicating that, in the heel strain regime, the collagen molecules are straightened, starting from the gap region and then within the entire fibril.31,59 3994
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
(2) Jeronimidis, G. In Structural Biological Materials: Design and Structure Property Relationships; Elices, M., Ed.; Pergamon: Amsterdam, 2000; Chapters 1 and 2. (3) Weiner, S.; Wagner, H. D. The material bone: Structure mechanical function relations. Annu. Rev. Mater. Sci. 1998, 28, 271− 298. (4) Fratzl, P.; Weinkamer, R. Nature’s hierarchical materials. Prog. Mater. Sci. 2007, 52 (8), 1263−1334. (5) Seto, J.; Gupta, H. S.; Zaslansky, P.; Wagner, H. D.; Fratzl, P. Tough lessons from bone: Extreme mechanical anisotropy at the mesoscale. Adv. Funct. Mater. 2008, 18 (13), 1905−1911. (6) Diepgen, T. L.; Mahler, V. The epidemiology of skin cancer. Br. J. Dermatol. 2002, 146 (Suppl 61), 1−6. (7) Daxer, A.; Fratzl, P. Collagen fibril orientation in the human corneal stroma and its implicaton to the pathology of keratoconus. Invest. Ophth. Vis. Sci. 1997, 38, 121−129. (8) Makareeva, E.; Mertz, E. L.; Kuznetsova, N. V.; Sutter, M. B.; DeRidder, A. M.; Cabral, W. A.; Barnes, A. M.; McBride, D. J.; Marini, J. C.; Leikin, S. Structural heterogeneity of type I collagen triple helix and its role in osteogenesis imperfecta. J. Biol. Chem. 2008, 283 (8), 4787−4798. (9) Wess, T. J. Collagen Fibrillar Structure and Hierarchies. In Collagen: Structure and Mechanics; Fratzl, P., Ed.; Springer Science + Business Media, LLC: New York, 2008; pp 49−80. (10) Avery, N. C.; Bailey, A. J. Restraining Cross-Links Responsible for Mechanical Properties of Collagen Fibers: Natural and Artificial. In Collagen: Structure and Mechanics; Fratzl, P., Ed.; Springer Science + Business Media, LLC: New York, 2008; pp 81−110. (11) Silver, F. H.; Landis, W. J. Viscoelasticity, Energy Storage and Transmission and Dissipation by Extracellular Matrices in Vertebrates. In Collagen: Structure and Mechanics; Fratzl, P., Ed.; Springer Science + Business Media, LLC: New York, 2008; pp 133−154. (12) Buehler, M. J. Nature designs tough collagen: Explaining the nanostructure of collagen fibrils. Proc. Natl. Acad. Sci. U. S. A. 2006, 103 (33), 12285−12290. (13) Gupta, H. S.; Seto, J.; Krauss, S.; Boesecke, P.; Screen, H. R. C. In situ multi-level analysis of viscoelastic deformation mechanisms in tendon collagen. J. Struct. Biol. 2009, 169 (2), 183−191. (14) Buehler, M. J.; Yung, Y. C. Deformation and failure of protein materials in physiologically extreme conditions and disease. Nature Mater. 2009, 8 (3), 175−188. (15) Freund, I.; Deutsch, M. Second-harmonic microscopy of biological tissue. Opt. Lett. 1986, 11, 94−96. (16) Williams, R. M.; Zipfel, W. R.; Webb, W. W. Interpreting Second-Harmonic Generation Images of Collagen I Fibrils. Biophys. J. 2005, 88, 1377−1386. (17) Janko, M.; Davydovskaya, P.; Bauer, M.; Zink, A.; Stark, R. W. Anisotropic Raman scattering in collagen bundles. Opt. Lett. 2010, 35 (16), 2765−2767. (18) Raghavan, M.; Sahar, N. D.; Wilson, R. H.; Mycek, M.-A.; Pleshko, N.; Kohn, D. H.; Morris, M. D. Quantitative polarized Raman spectroscopy in highly turbid bone tissue. J. Biomed. Opt. 2010, 15 (3), 037001. (19) Falgayrac, G.; Facq, S.; Leroy, G.; Cortet, B.; Penel, G. New Method for Raman Investigation of the Orientation of Collagen Fibrils and Crystallites in the Haversian System of Bone. Appl. Spectrosc. 2010, 64 (7), 775−780. (20) Bonifacio, A.; Sergo, V. Effects of sample orientation in Raman microspectroscopy of collagen fibers and their impact on the interpretation of the amide III band. Vib. Spectrosc. 2010, 53 (2), 314−317. (21) Kazanci, M.; Wagner, H. D.; Manjubala, N. I.; Gupta, H. S.; Paschalis, E.; Roschger, P.; Fratzl, P. Raman imaging of two orthogonal planes within cortical bone. Bone 2007, 41 (3), 456−461. (22) Kazanci, M.; Roschger, P.; Paschalis, E. P.; Klaushofer, K.; Fratzl, P. Raman spectral mapping of bone osteonal tissues. Bone 2006, 39 (5), S16−S16. (23) Gamsjaeger, S.; Masic, A.; Roschger, P.; Kazanci, M.; Dunlop, J. W. C.; Klaushofer, K.; Paschalis, E. P.; Fratzl, P. Cortical bone
Figure 7 shows the versatility of PR spectroscopy in monitoring collagen multiscale orientation changes during a uniaxial tensile experiment on RTT. Reorganization on the molecular level (Figure 7, inset A) is evaluated through the distribution function f mp whereas the tissue level (Figure 7, inset B) is accessed via the frequency plot of parameter c from the imaging experiment. At the tissue level straightening of the crimp can be observed when passing from the toe to the heel region as indicated by the frequency plot of the c parameter. The orientation distribution of the collagen molecules was found to be narrower and centered at 0° already in the heel region of the stress−strain curve, indicating a straightening of the crimp. Molecular level reorientation occurs only in the linear part of the curve, indicating further straightening of collagen molecules with respect to the direction of applied stress. In summary, the results presented here are in agreement with common models of the hierarchical deformation of RTT. Applied strain does not affect the molecular vibrational levels of the collagen structure, but it is rather distributed via higher levels of hierarchy through reorientations and lateral sliding of collagen molecules.
■
CONCLUSIONS In this work, we exploited both molecular level PR spectroscopic information in combination with tissue scale PR imaging to study the mechanical behavior of RTT under uniaxial tensile deformation. The results demonstrate the versatility of PR analytical techniques to obtain multiscale orientation information at the molecular and at the fibril level of collagen. The method can be applied to map collagen within other collagen rich tissues, and in principle, it is possible to concurrently map other chemical components associated with collagen. The in situ mechanical tests show that the orientation distribution function of collagen molecules changes its shape passing from the heel to the linear region of the stress−strain curve, indicating a significant molecular alignment. An external applied stress, however, is shown to lead only to reorientation of the collagen molecule and does not affect any of its vibrational molecular properties.
■
AUTHOR INFORMATION
Corresponding Author *Tel: +49 331 567 9400. Fax: +49 331 567 9402. E-mail:
[email protected].
■
ACKNOWLEDGMENTS The authors thank Julius Wolff Institute for Biomechanics and Musculoskeletal Regeneration for providing fresh rat tails. We also thank Drs. Notburga Gierlinger, Richard Weinkamer, Sonja Gamsjäger, Murat Kazanci, and Himadri S. Gupta for fruitful discussions. L. G. and N. T. gratefully acknowledge the BerlinBrandenburg School for Regenerative Therapies for funding. P.F. and A.M. are grateful for support by the Alexander von Humboldt Foundation and the Max Planck Society in the framework of the Max Planck Research Award funded by the Federal Ministry of Education and Research.
■
REFERENCES
(1) Fratzl, P. Collagen: Structure and Mechanics, an Introduction. In Collagen, Fratzl, P., Ed.; Springer: New York, 2008; pp 1−13. 3995
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
Biomacromolecules
Article
(44) Labarthet, F.; Buffeteau, T.; Sourisseau, C. Orientation distribution functions in uniaxial systems centered perpendicularly to a constraint direction. Appl. Spectrosc. 2000, 54, 699−705. (45) Bower, D. Investigation of Molecular Orientation Distributions by Polarized Raman-Scattering and Polarized Fluorescence. J. Polym. Sci., Polym. Phys. 1972, 10, 2135−2153. (46) Sourisseau, C. Polarization measurements in macro- and microRaman spectroscopies: Molecular orientations in thin films and azodye containing polymer systems. Chem. Rev. 2004, 104, 3851−3891. (47) Rousseau, M. E.; Beaulieu, L.; Lefevre, T.; Paradis, J.; Asakura, T.; Pezolet, M. Characterization by Raman microspectroscopy of the strain-induced conformational transition in fibroin fibers from the silkworm Samia cynthia ricini. Biomacromolecules 2006, 7 (9), 2512− 2521. (48) Turrell, G. Analysis of Polarization Measurements in Raman Microspectroscopy. J. Raman Spectrosc. 1984, 15 (2), 103−108. (49) Pottel, H.; Herrman, W.; Vandermeer, B.; Ameloot, M. On the Significance of the 4th-rank Orientational Order Parameter of Fluorophores in Membranes. Chem. Phys. 1986, 102, 37−44. (50) Tanaka, M.; Young, R. J. Polarised Raman spectroscopy for the study of molecular orientation distributions in polymers. J. Mater. Sci. 2006, 41 (3), 963−991. (51) Lafrance, C.; Nabet, A.; Prudhomme, R.; Pezolet, M. On the Relationship Between the Order-Parameter [P-2(Cos Theta)] and the Shape of Orientation Distributions. Can. J. Chem. 1995, 73, 1497− 1505. (52) Boudko, S. P.; Engel, J.; Okuyama, K.; Mizuno, K.; Bachinger, H. P.; Schumacher, M. A. Crystal structure of human type III collagen Gly991-Cly1032 cystine knot-containing peptide shows both 7/2 and 1-/3 triple helical symmetries. J. Biol. Chem. 2008, 283, 32580−32589. (53) Kramer, R. Z.; Venugopal, M. G.; Bella, J.; Mayville, P.; Brodsky, B.; Berman, H. M. Staggered molecular packing in crystals of a collagen - like peptide with a single charged pair. J. Mol. Biol. 2000, 301, 1191−1205. (54) Bella, J.; Eaton, M.; Brodsky, B.; Berman, H. M. Crystal and molecular structure of a collagen - like peptide at 1.9 angstrom resolution. Science 1994, 266, 75−81. (55) Berisio, R.; Vitagliano, L.; Mazzarella, L.; Zagari, A. Crystal structure of the collagen triple helix model [(Pro-Pro-Gly)(10)](3). Protein Sci. 2002, 11, 262−270. (56) Kawahara, K.; Nishi, Y.; Uchiyama, S.; Nishiuchi, Y.; Nakazawa, T.; Ohkubo, T.; Kobayashi, Y. Effect of hydration on the stability of the collagen-like triple-helical structure of [4(R)-hydroxyprolyl-4(R)hydroxyprolylglycine]10. Biochemistry 2005, 44, 15812−15822. (57) Rigby, B. J.; Hirai, N.; Spikes, J. D.; Eyring, H. The Mechanical Properties of Rat Tail Tendon. J. Gen. Physiol. 1959, 43 (2), 265−283. (58) Dee, K. C.; Gentleman, E.; Lay, A. N.; Dickerson, D. A.; Nauman, E. A.; Livesay, G. A. Mechanical characterization of collagen fibers and scaffolds for tissue engineering. Biomaterials 2003, 24 (21), 3805−3813. (59) Buehler, M. J.; Gautieri, A.; Vesentini, S.; Redaelli, A. Hierarchical Structure and Nanomechanics of Collagen Microfibrils from the Atomistic Scale Up. Nano Lett. 2011, 11 (2), 757−766. (60) Lefevre, T.; Rousseau, M. E.; Pezolet, M. Orientation-insensitive spectra for Raman microspectroscopy. Appl. Spectrosc. 2006, 60 (8), 841−846. (61) Gierlinger, N.; Schwanninger, M.; Reinecke, A.; Burgert, I. Molecular changes during tensile deformation of single wood fibers followed by Raman microscopy. Biomacromolecules 2006, 7 (7), 2077− 2081. (62) Pezolet, M.; Lefevre, T.; Paquet-Mercier, F.; Lesage, S.; Rousseau, M. E.; Bedard, S. Study by Raman spectromicroscopy of the effect of tensile deformation on the molecular structure of Bombyx mori silk. Vib. Spectrosc. 2009, 51 (1), 136−141.
composition and orientation as a function of animal and tissue age in mice by Raman spectroscopy. Bone 2010, 47 (2), 392−399. (24) Hulmes, D. J. S. Collagen Diversity, Synthesis and Assembly. In Collagen: Structure and Mechanics; Fratzl, P., Ed.; Springer: New York, 2008; pp 15−48. (25) Cribb, A. M.; Scott, J. E. Tendon Response to Tensile-stress an Ultrastructural Investigation of Collagen - Proteoglycan Interaction in Stressed Tendon. J. Anat. 1995, 187, 423−428. (26) Morgan, M.; Kostyuk, O.; Brown, R. A.; Mudera, V. In situ monitoring of tendon structural changes by elastic scattering spectroscopy: Correlation with changes in collagen fibril diameter and crimp. Tissue Eng. 2006, 12 (7), 1821−1831. (27) Orgel, J.; Irving, T. C.; Miller, A.; Wess, T. J. Microfibrillar structure of type I collagen in situ. P. Natl. Acad. Sci. U. S. A. 2006, 103 (24), 9001−9005. (28) Diamant, J.; Keller, A.; Baer, E.; Litt, M.; Arridge, R. G. C. Collagen: Ultrastructure and its relation to mechanical properties as a function of aging. Proc. R. Soc. London, B 1972, 180, 293−315. (29) Hulmes, D. J. S.; Wess, T. J.; Prockop, D. J.; Fratzl, P. Radial packing, order and disorder in collagen fibrils. Biophys. J. 1995, 68, 1661−1670. (30) Misof, K.; Rapp, G.; Fratzl, P. A new molecular model for collagen elasticity based on synchrotron x-ray scattering evidence. Biophys. J. 1997, 72 (3), 1376−1381. (31) Fratzl, P.; Misof, K.; Zizak, I.; Rapp, G.; Amenitsch, H.; Bernstorff, S. Fibrillar structure and mechanical properties of collagen. J. Struct. Biol. 1997, 122, 119−122. (32) Tsuboi, M.; Nakamura, K.; Aida, M.; Kubo, Y.; Benevides, J. M.; Thomas, G. J. Raman tensor of the carboxylate anion: A basis for determining aspartate and glutamate side-chain orientations in proteins by polarized Raman spectroscopy. Biophys. J. 2002, 82 (1), 457A−458A. (33) Tsuboi, M.; Kubo, Y.; Akahane, K.; Pezolet, M.; Lefevre, T.; Thomas, G. J. Determination of the amide I Raman tensor for the antiparallel chain beta-pleated sheet. Biophys. J. 2005, 88 (1), 559A− 559A. (34) Lefevre, T.; Rousseau, M. E.; Pezolet, M. Determination of molecular orientation in protein films and fibres by Raman microspectroscopy. Can. J. Anal. Sci. Spectrosc. 2005, 50 (1), 41−48. (35) Tsuboi, M. Raman scattering anisotropy of biological systems. J. Biomed. Opt. 2002, 7 (3), 435−441. (36) Tsuboi, M.; Kubo, Y.; Akahane, K.; Benevides, J. M.; Thomas, G. J. Determination of the amide I Raman tensor for the antiparallel beta-sheet: application to silkworm and spider silks. J. Raman Spectrosc. 2006, 37 (1−3), 240−247. (37) Overman, S. A.; Tsuboi, M.; Thomas, G. J. Orientations of tyrosine residues (Y21 and Y24) in coat protein subunits of the filamentous virus Ff. Biophys. J. 1998, 74 (2), A73−A73. (38) Rintoul, L.; Carter, E. A.; Stewart, S. D.; Fredericks, P. M. Keratin orientation in wool and feathers by polarized Raman spectroscopy. Biopolymers 2000, 57 (1), 19−28. (39) Rousseau, M. E.; Lefevre, T.; Beaulieu, L.; Asakura, T.; Pezolet, M. Study of protein conformation and orientation in silkworm and spider silk fibers using Raman microspectroscopy. Biomacromolecules 2004, 5 (6), 2247−2257. (40) Lefevre, T.; Rousseau, M. E.; Pezolet, M. Protein secondary structure and orientation in silk as revealed by Raman spectromicroscopy. Biophys. J. 2007, 92 (8), 2885−2895. (41) Bi, X.; Li, G.; Doty, S. B.; Camacho, N. P. A novel method for determination of collagen orientation in cartilage by Fourier transform infrared imaging spectroscopy (FT-IRIS). Osteoarthr. Cartilage 2005, 13 (12), 1050−1058. (42) Kramer, R. Z.; Bella, J.; Mayville, P.; Brodsky, B.; Berman, H. M. Sequence dependent conformational variations of collagen triplehelical structure. Nat. Struct. Biol. 1999, 6, 454−457. (43) Bremard, C.; Dhamelincourt, P.; Laureyns, J.; G., T. The Effect of High-Numerical-Aperture Objectives on Polarization Measurements in Micro-Raman Spectrometry. Appl. Spectrosc. 1985, 39 (6), 1036− 1039. 3996
dx.doi.org/10.1021/bm201008b | Biomacromolecules 2011, 12, 3989−3996
