- Email: [email protected]
Detailed Analysis of the Structural Changes of Bone Matrix During the Demineralization Process Using Raman Spectroscopy
Available online at www.sciencedirect.com
ScienceDirect Physics Procedia 73 (2015) 221 – 227
4th h International Confereence Photon nics and Info formation Optics, O PhIO O 2015, 28-330 January 2015 2
Deetailed annalysis of o the strructural changess of bonee matrix during the t deemineralization process p uusing Raaman speectroscoopy E.V. Timcheenkoa*, L.A A. Zherdev vaa, P.E. Tim mchenkoa, L.T. Volo ovab, U.V. P Ponomarev vab b
a Samara Statte Aerospace Univversity (SSAU), M Moskovskoye shossse 34, Samara 443086, Russia Experimental M Medicine And Biotechnologies Insttitute of the Samaara Medicine University, Gagarin Street 20, Samarra 443079, Russia a
Absttract The results of experimental researrch of human cortical bone tisssue depending g on demineralization time weere represented using Raman spectroscopyy. Depeending on demiineralization tim me the ratio of the t mineral (ɊɈ Ɉ43- and CO32-) and organic co omponents (ami mide I) of bone tissue, t as well as changes iin the spectral regions r responssible for the strructural integritty of the collageen fibers in bonne tissue (1200-1460 cm-1 and 2880-30000 cm-1) were innvestigated. Thee observed chaanges show a decrease in mineeral componentts: thus, the vallue of Raman band intensiity at 956 and 1069 1 cm-1 for 5 minutes deminneralization is 68.5 6 and 77.3 %, % for 20 minutees - 55.1 and 61 1.1 %, for 120 minutes - 322.8 and 37 % frrom Raman inteensity values off not demineraliized tissue objects respectivelyy. ©©2015 Authors. Published by Elsevier B.V. B.V. This is an open access article under the CC BY-NC-ND license 015The The Authorrs. Published byy Elsevier 20 (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer--review under rresponsibility of the National Research R Nucleear University MEPhI M (Moscow w Engineering PPhysics Institutte). Peer-review under responsibility of the National Research Nuclear University MEPhI (Moscow Engineering Physics Institute) Keyw words: Raman speectroscopy; corticcal bone; hydroxy yapatite; collagenn; scanning electro on microscopy.
1. In ntroduction In n surgical andd traumatologgy practice thee alteration off the bone tisssue integrity, associated w with the diseasses of musculoskeletal ssystem, with thhe occurrencee of microfracttures amid weeakening of th he bone remoddeling processs with nsequent increease of bone fragility f and porosity, p is speecific for a sig gnificant perceentage of patieents. Currentlly, for a con the replacement r oof bone defectts is accepted to use the "ggold standard"" - autogenouss bone grafts.. At early stag ges of the study s and usee of autogenouus tissues in transplantation t n process, thee main problem m with infectitious complicaations follo owing transplantation has forced speciaalists to refusse of this kin nd of grafts. Over O time, thhe developmeent of
* Corresponding C autthor. Tel.: +7-9277-711-13-87. E-mail E address: [email protected]
1875-3892 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the National Research Nuclear University MEPhI (Moscow Engineering Physics Institute) doi:10.1016/j.phpro.2015.09.161
222
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
biom medical sciennces in conjuunction with innovative i prrocessing tech hniques of au utogenous tisssue resulted in the creaation of new sstandards for the t safe use off autogenous ggrafts [Flierl M.A. M et al. (20 013)]. The T use of deemineralized lyophilized l bo one tissue of donor that allmost is not inferior to a bbone graft bassed on patiient's tissue, iss one of the suuch standards.. The importannt step in obtaaining of demiineralized bonne biomateriall is the proccessing of tisssue in acidic solution thatt led to loss oof most of the mineral com mponents [Ʉirrillova I.Ⱥ. (2 2004)] whiile maintaininng of collagenn and protein ns proportion necessary fo or tissue regen neration [Tim mchenko P.E. et al. (2011)]. It is alsoo known that the recovery speed of bonne is higher in n the transplan ntation of parttially deminerralized biom material as coompared with a completely y demineralizeed [West P.A.. et al. (2005)]. The effect oof demineraliization inco ompleteness iis explained by b strong link ks of osteoindductor (respon nsible for thee osteogenesiss) with the mineral m com mponent of boone tissue [Sluutsky L.I. et al. a (1986)]. Ass a result, corrrectly selected the deminerralization deg gree of bon ne matrix provvides the perffect combination of osteoinnductive and osteoconductive propertiess for effectivee bone regeeneration. Used U biochem mical and morpphological meethods to deteermine the dem mineralization n degree oftenn require signiificant timee and resourcce costs that directly affecct on the quaality of patien nt care. At th he moment, aamong the po ossible metthods to solvee this problem m are the optiical methods, and authors of paper [Tarrnowski C.P. et al. (2002)]] have prop posed the metthod of Ramaan spectroscop py (RS) as a ffast and efficieent method fo or investigatinng the bone mineral m com mponents. The T aim of thhis work was the t establishm ment of depenndence for min neral and orgaanic componeents of human n bone biom material in thee demineralizaation process using Raman spectroscopy. 2. Materials M and d methods As A objects of study the 36 samples s of thrree-dimensionnal cortical hu uman bone tissue of the "Liioplast" seriess (TU939 98-001-019631143-2004) weere used. The first step in ttechnological preparation process p of graafts was degreasing and d complete rem moval of bonee marrow and d blood from bbone scaffold ds by brief low w-frequency uultrasonic treaatment (24 - 40 kHz duriing 2-3 minutes), while the destruction oof bone was no ot occurring. Then, T to obtainn the deminerralized bon ne biomaterial we placed boone tissue in a 1,2 % solutioon of HCl whiile maintaining g a constant ssolution tempeerature at +4 + 0C. The neext step was lyophilization l n of samples oon the ALPH HA 2-4 LSC which w consistss in freeze-drried of tissu ue. Spectral S charaacteristics of the samples were studiedd using an ex xperimental setup s (Fig. 1)), includes a highreso olution digitall spectrometeer Shamrock sr-303i s with an integrated d cooled cameera DV420A--OE CCD (sp pectral reso olution of 0.155 nm (~ 1 cm m-1)), an RPB7 785 fiber-optiic probe comb bined with thee LuxxMasterr LML-785.0R RB-04 laseer module (7855 nm wavelenngth).
Fig. F 1. Experimenntal setup: 1 – obbject; 2 – Raman n probe RPB785; 3 – laser modulee LuxxMaster LM ML-785.0RB-04;; 4 – laser power supply mod dule; 5 – spectrom meter Shamrock srr-303i; 6 – CCD camera ANDOR DV-420A-OE; 7 – computer; 8 – transmitting fibeer; 9 – receiving fiber f
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
Due to the low level of autofluorescence the 785 nm wavelength of the visible wavelength range is well-proven in Raman spectroscopy for biological applications. Procedure description for work with the experimental stand earlier was presented in the article [Timchenko E.V. et al. (2014)]. 3. Results Before data analysis, the obtained Raman spectra of research objects were preprocessed in MatLab’7 for deletion of setup noise and fluorescence background using an approximation of the spectral curve by a fifth order polynomial, based on the algorithm proposed in the paper of [Zhao J. et al. (2007)]. A feature of the algorithm is a sequential and many iterative comparisons of the initial spectrum with a polynomial approximation of autofluorescence component, and the preservation of spectral contours and intensity of the initial Raman spectrum positively proved the algorithm in spectroscopy for biomedical research [Huang Z. et al. (2011)]. However, during the spectra study was found that at the large spectral range (exceeding the 1500-3000 cm-1) the approximation of spectral curve led to cuts of Raman band near more intensive neighboring band such as 1003 cm-1 (Ȟ (C-C) of phenylalanine) near 956 cm-1 (Ȟ1 (PO4)3- of hydroxyapatite). Therefore, the processing was carried out at the spectral regions 400-1150, 1200-1500, 1600-1800 and 2800-3000 cm-1 with appropriate criteria of approximation for sensitivity increase and maximum display without loss of informative bands. Informativeness of Raman spectra of bone tissue samples depends on the complexity and structure of studied biomaterial (on the number of bands in Raman spectrum), because the resulting spectrum is a superposition of several spectral bands that leading to distortion of the real values of intensity. Therefore, to extract the Raman bands against the background of spectral curve, the decomposition of obtained spectra into spectral contours using deconvolution of Gauss-Lorentz function was conducted by PeakFit software (Systat Software, Inc.). Specific Raman spectrum for non-demineralized cortical human bone is presented in Fig.2. The determination coefficients of the resulting spectrum (generated by the spectral contours) from the initial spectrum for 790-1140 cm-1, 1200-1480 cm-1, 1590-1790 cm-1 and 2840-3020 cm-1 regions were R2 = 0.9999, 0.9978, 0.9998 and 0.9993 respectively. The spectral region (Fig. 2a) contains the Raman bands responsible for the mineral structure of bone matrix, such as Ȟ1 (PO4)3- vibrational mode in the molecule of hydroxyapatite at 956 cm-1, Ȟ3 (PO4)3- at 1030 and 1045 cm-1, Ȟ1 (CO3)2- (B type substitution where (PO4)3- a substituted (CO3)2-) at 1069 cm-1 and Ȟ1 (CO3)2- vibration mode (A type substitution where a substituted OH- (CO3)2-) at 1098 cm-1 [Yamamoto T. et al. (2011), Crane N.J. et al. (2013)]. Band at 956 cm-1 is dominant by the intensity throughout all Raman spectrum of bone tissue, which determines the content of hydroxyapatite in bone matrix, and also possible to study the content of carbonate (CO3)2- in the process of demineralization. Less intense bands at 816, 852, 920, 941 and 870 cm-1 corresponding to the stretching vibration of C-C in proline and hydroxyproline [Yamamoto T. et al. (2011), Crane N.J. et al. (2013)], and the band at 1003 cm-1 responsible for C-C stretching vibration in phenylalanine [Crane N.J. et al. (2013)] represent the organic component of bone matrix. Further, Fig. 2b shows the spectral region responsible for structural integrity of bone collagen, where 1239, 1264, 1284, 1316, 1338 and 1381 cm-1 bands correspond to amide III (modes of Ȟ (CN) and į (NH)) [Yamamoto T. et al. (2011), Crane N.J. et al. (2013), Rehman I. et al. (2012), Poáomska M. et al. (2010)] and 1396, 1418 and 1450 cm-1 correspond to mode of į (CH2) and į (CH3) [Rehman I. et al. (2012), Poáomska M. et al. (2010)]. Spectral region of 1600-1700 cm-1 (Fig. 2c) in [Crane N.J. et al. (2013)] is a defining in the assessment of structural integrity of the collagen fibers. This region corresponds to amide I (stretching of C=O) with Raman bands at 1634, 1660, 1681 cm-1 (Į-helix and ȕ-sheet [Poáomska M. et al. (2010), Shen J. et al. (2010)]), and 1604 cm-1 (stretching vibration of C=C in tyrosine [Rehman I. et al. (2012)]). The most remote region of the spectrum (Fig. 2d) at 2879, 2910, 2938 and 2980 cm-1 corresponds to the stretching vibrations of CH2 and CH3 in collagen [Poáomska M. et al. (2010), Ali S.M. et al. (2013)].
223
224
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
b
a
d
c -1
-1
Fig. 2. Sppecific spectral coontours for region ns: (a) 790-1140 cm , (b) 1200-14 480 cm , (c) 1590-1790 cm-1, (d) 22840-3020 cm-1
In I this article,, the relative concentration c of organic coompounds in bone b matrix was w evaluatedd by the intensity of the amide III bannd (1239 cm-1), ) CH2 deform mation band (1 450 cm-1) and d amide I regio on (1660 and 1681 cm-1). In n turn, -1 man band at 9956, 1069 andd 1098 cm were w used forr investigation n changes in the t mineral co components off bone Ram mattrix. Further, F Fig. 3 shows the inntensity chang ge of Raman bbands for bone matrix samp ples at differennt demineralizzation timee relative to iintensity valuees of Raman bands for nonn-demineralizzed samples. The T intensity value at 2938 8 cm-1 werre used for intensity normallization of all spectral bandss:
In n( k )
I ( k )t I (k )0 I (2938)t
(1)
wheere I(k) t – Ram man intensity at a k wavenu umber with a t demineralizzation time of sample, I(k) 00 – Raman inttensity at a k wavenum mber without demineralizati d ion of samplee (t = 0), and I(2938) t – Raman intennsity at 2938 8 cm-1 wav venumber withh a t demineraalization time of sample.
225
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
a
b
c
d
m of: (a) I - In (956) and In (1069 ), II - In (1098) and a In (1069); (b) I - In (956) and IIn (1681), II - In (956) Fig. 3. The two-dimeensional histogram d In (1660); (c) Inn (956) and In (12239); (d) In (956) and In (1450) forr different demin neralization time: 1 - 0 min., 2 - 2 m min ., 3 - 5 min., 4 - 8 and min., 5 - 11 min., 6 - 15 min., 7 - 20 min., m 8 - 30 min., 9 - 45 min., 10 - 60 min. 11 - 90 min., m 12 - 120 minnutes
As A can be seeen from Fig. 3a, bands at 956, 1069 annd 1098 cm-11 are characteerized by a siignificant inteensity decrrease with dem mineralizationn time: at 5 minutes m of dem mineralization the values off In(956), In(11069) and In(1098) are 68.5, 6 77.3 andd 83.2 %, at 20 2 minutes - 55.1, 5 61.1 andd 76.2%, at 12 20 minutes - 32.8, 3 37 and 666.4% respectively. Fig. 3b shows the intensity channges in Ramaan bands of am mide I region, indicates the cross-links in collagen: the main d at 1660 cm-11 is characterizzed by an inteensity decreasee to 89.2% at 5 minute dem mineralization and to 83.2% at 20 band minu utes, but 1681 cm-1 is chaaracterized by y an intensityy increase to 19.4% at 5 minutes m and 337% at 20 minute m dem mineralization. Raman R intensitty at 1239 cm m-1, responsiblee for the amidde III group, during d all dem mineralization time varies by y less than n 5.6% (Fig. 33b), unlike thee band at 1450 0 cm-1 whosee intensity afteer 90 minute demineralizati d tion is decreassed to 89.1% relative too Raman intennsity of non-d demineralizedd bone matrix x sample (Fig g. 3d). Practiccally unchang ged in nsity the Ram man band of coollagen inform ms about an opptimal parameters of the deemineralized medium (HCl acid inten temp perature and cconcentration)) that does nott lead to deforrmation and destruction d of collagen. Avaailability of Raman R band ds, that corresppond to the coollagen structu ure and whosee intensity inccreases with in ncreasing dem mineralization time,
226
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
can be explained by the restruccturing into ch hemical compposition collag gen, associated with a recovvery or oxidattion of w with demineeralization tim me and reach h their CH-bond. Thus, the band inttensities at 1284 and 28779 cm-1 grow ximum of 1117.2 and 1177.1% at 20 minutes dem mineralization relatively non-demineraliized sampless with max subsequent intenssity relaxationn at 45-120 miinutes demineeralization. The T mineral composition change of investigated samples parrallel was monitored m by scanning electron miccroscopy. Miccrographs of sample s surfacce structure foor cortical bo one matrix dep pending on thhe demineraliization timee are presenteed in Fig. 4.
Fig. 4. The structture of cortical boone surface at diffferent demineraliization time for biomaterial: b (a) 0 min, (b) 2 min, (cc) 5 min, (d) 8 min. The a accelerating voltag ge - 20 kV, beam m current - 1 nA, number n of points - 16
In I Fig. 4 can oobserve as thee demineralizeed samples havve a certain deegree of bone surface roughhness (Fig. 4c) with the appearance oof pores at 5 and a 8 minutes demineralizaation. These ch hanges are the result of "w washing out" apatite a o the object. That causes the bones to lose most off their hardnesss and compreessive crysstals under ann acid action on streength, and at a large demineeralization tim me leads to defformation and destruction of the trabeculaae. 4. Conclusion C As A a result, Raaman spectra of human corttical bone witth a different demineralizati d ion degree weere obtained. Shown S thatt the demineraalization process can be qu uantitatively coontrolled usin ng the ratio ɊɈ Ɉ43- and ɋɈ322-: for 956 and d 1069 cm-1 at 5 minutes demineralizzation the inteensity value iss 68.5 and 77.3%, for 20 minutes m - 55.1 and 61.1%, at 120 min nutes - 32.8 annd 37% respeectively. The spectral featuures for bone collagen depending on dem n time mineralization werre investigatedd by analyzingg spectral regiions of 1200-11460 cm-1 and d 2880-3000 cm m-1. Ack knowledgemeent This T work wass supported byy the Ministry y of Educationn and Science of the Russian n.
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
227
References Ali S.M., Bonnier F., Tfayli A., et al., 2013, Raman spectroscopic analysis of human skin tissue sections ex-vivo: evaluation of the effects of tissue processing and dewaxing, Journal of Biomedical Optics. 18(6), 061202–10. Crane N.J., Polfer E., et al., 2013, Raman spectroscopic analysis of combat-related heterotopic ossification development, Bone. 57, 335–342. Flierl M.A., Smith W.R., et al., 2013, Outcomes and complication rates of different bone grafting modalities in long bone fracture nonunions: a retrospective cohort study in 182 patients, Journal of Orthopaedic Surgery and Research. 8, 33. Huang Z., Chen X., et al., 2011, Raman spectroscopic characterization and differentiation of seminal plasma, J. Biomed. Opt. 16, 110501. Ʉirillova I.Ⱥ., 2004, Demineralized bone graft as a stimulator of bone formation: modern concepts, Spine Surgery. 3, 105–110. Poáomska M., Kubisz L., et al., 2010, Fourier Transform Near Infrared Raman Spectroscopy in Studies on Connective Tissue, Acta Phys. Pol. A. 118(1), 138. Rehman I., Movasaghi Z., et al., 2012, Vibrational Spectroscopy for Tissue Analysis, Series in Medical Physics and Biomedical Engineering, CRC Press, 271. Shen J., Fan L., Yang J., et al., 2010, A longitudinal Raman microspectroscopic study of osteoporosis induced by spinal cord injury, Osteoporos Int, 21, 81–87. Slutsky L.I., Sevastianova N.A., 1986, Podiatrist., TRAUMA., 8 (82). Tarnowski C.P., et al., 2002, Mineralization of developing mouse calvaria as revealed by Raman microspectroscopy, Journal of Bone and Mineral Research. 17, 1118–1126. Timchenko E.V., Timchenko P.E., Volova L.T., et al., 2014, Raman spectroscopy of the organic and mineral structure of bone grafts, Quantum Electronics. 44 (7), 696–699. Timchenko P.E., Timchenko E.V., Zakharov V.P., et al., 2011, Diagnostics of bone implantat and control of their process osteointegration with of a method confocal microscopy, Computer Optics. 35(2), 183-187. West P.A., Torzilli P.A., et al., 2005, Fourier transform infrared imaging spectroscopy analysis of collagenase-induced cartilage degradation, Journal of Biomedical Optics. 10 (1), 014015. Yamamoto T., Uchida K., Naruse K., et al., 2011, Quality assessment for processed and sterilized bone using Raman spectroscopy, Cell and Tissue Banking. 409–414. Zhao J., et al., 2007, Automated autofluorescence background subtraction algorithm for biomedical Raman spectroscopy, Society for applied spectroscopy. 61(11), 1225–1232.
ScienceDirect Physics Procedia 73 (2015) 221 – 227
4th h International Confereence Photon nics and Info formation Optics, O PhIO O 2015, 28-330 January 2015 2
Deetailed annalysis of o the strructural changess of bonee matrix during the t deemineralization process p uusing Raaman speectroscoopy E.V. Timcheenkoa*, L.A A. Zherdev vaa, P.E. Tim mchenkoa, L.T. Volo ovab, U.V. P Ponomarev vab b
a Samara Statte Aerospace Univversity (SSAU), M Moskovskoye shossse 34, Samara 443086, Russia Experimental M Medicine And Biotechnologies Insttitute of the Samaara Medicine University, Gagarin Street 20, Samarra 443079, Russia a
Absttract The results of experimental researrch of human cortical bone tisssue depending g on demineralization time weere represented using Raman spectroscopyy. Depeending on demiineralization tim me the ratio of the t mineral (ɊɈ Ɉ43- and CO32-) and organic co omponents (ami mide I) of bone tissue, t as well as changes iin the spectral regions r responssible for the strructural integritty of the collageen fibers in bonne tissue (1200-1460 cm-1 and 2880-30000 cm-1) were innvestigated. Thee observed chaanges show a decrease in mineeral componentts: thus, the vallue of Raman band intensiity at 956 and 1069 1 cm-1 for 5 minutes deminneralization is 68.5 6 and 77.3 %, % for 20 minutees - 55.1 and 61 1.1 %, for 120 minutes - 322.8 and 37 % frrom Raman inteensity values off not demineraliized tissue objects respectivelyy. ©©2015 Authors. Published by Elsevier B.V. B.V. This is an open access article under the CC BY-NC-ND license 015The The Authorrs. Published byy Elsevier 20 (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer--review under rresponsibility of the National Research R Nucleear University MEPhI M (Moscow w Engineering PPhysics Institutte). Peer-review under responsibility of the National Research Nuclear University MEPhI (Moscow Engineering Physics Institute) Keyw words: Raman speectroscopy; corticcal bone; hydroxy yapatite; collagenn; scanning electro on microscopy.
1. In ntroduction In n surgical andd traumatologgy practice thee alteration off the bone tisssue integrity, associated w with the diseasses of musculoskeletal ssystem, with thhe occurrencee of microfracttures amid weeakening of th he bone remoddeling processs with nsequent increease of bone fragility f and porosity, p is speecific for a sig gnificant perceentage of patieents. Currentlly, for a con the replacement r oof bone defectts is accepted to use the "ggold standard"" - autogenouss bone grafts.. At early stag ges of the study s and usee of autogenouus tissues in transplantation t n process, thee main problem m with infectitious complicaations follo owing transplantation has forced speciaalists to refusse of this kin nd of grafts. Over O time, thhe developmeent of
* Corresponding C autthor. Tel.: +7-9277-711-13-87. E-mail E address: [email protected]
1875-3892 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the National Research Nuclear University MEPhI (Moscow Engineering Physics Institute) doi:10.1016/j.phpro.2015.09.161
222
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
biom medical sciennces in conjuunction with innovative i prrocessing tech hniques of au utogenous tisssue resulted in the creaation of new sstandards for the t safe use off autogenous ggrafts [Flierl M.A. M et al. (20 013)]. The T use of deemineralized lyophilized l bo one tissue of donor that allmost is not inferior to a bbone graft bassed on patiient's tissue, iss one of the suuch standards.. The importannt step in obtaaining of demiineralized bonne biomateriall is the proccessing of tisssue in acidic solution thatt led to loss oof most of the mineral com mponents [Ʉirrillova I.Ⱥ. (2 2004)] whiile maintaininng of collagenn and protein ns proportion necessary fo or tissue regen neration [Tim mchenko P.E. et al. (2011)]. It is alsoo known that the recovery speed of bonne is higher in n the transplan ntation of parttially deminerralized biom material as coompared with a completely y demineralizeed [West P.A.. et al. (2005)]. The effect oof demineraliization inco ompleteness iis explained by b strong link ks of osteoindductor (respon nsible for thee osteogenesiss) with the mineral m com mponent of boone tissue [Sluutsky L.I. et al. a (1986)]. Ass a result, corrrectly selected the deminerralization deg gree of bon ne matrix provvides the perffect combination of osteoinnductive and osteoconductive propertiess for effectivee bone regeeneration. Used U biochem mical and morpphological meethods to deteermine the dem mineralization n degree oftenn require signiificant timee and resourcce costs that directly affecct on the quaality of patien nt care. At th he moment, aamong the po ossible metthods to solvee this problem m are the optiical methods, and authors of paper [Tarrnowski C.P. et al. (2002)]] have prop posed the metthod of Ramaan spectroscop py (RS) as a ffast and efficieent method fo or investigatinng the bone mineral m com mponents. The T aim of thhis work was the t establishm ment of depenndence for min neral and orgaanic componeents of human n bone biom material in thee demineralizaation process using Raman spectroscopy. 2. Materials M and d methods As A objects of study the 36 samples s of thrree-dimensionnal cortical hu uman bone tissue of the "Liioplast" seriess (TU939 98-001-019631143-2004) weere used. The first step in ttechnological preparation process p of graafts was degreasing and d complete rem moval of bonee marrow and d blood from bbone scaffold ds by brief low w-frequency uultrasonic treaatment (24 - 40 kHz duriing 2-3 minutes), while the destruction oof bone was no ot occurring. Then, T to obtainn the deminerralized bon ne biomaterial we placed boone tissue in a 1,2 % solutioon of HCl whiile maintaining g a constant ssolution tempeerature at +4 + 0C. The neext step was lyophilization l n of samples oon the ALPH HA 2-4 LSC which w consistss in freeze-drried of tissu ue. Spectral S charaacteristics of the samples were studiedd using an ex xperimental setup s (Fig. 1)), includes a highreso olution digitall spectrometeer Shamrock sr-303i s with an integrated d cooled cameera DV420A--OE CCD (sp pectral reso olution of 0.155 nm (~ 1 cm m-1)), an RPB7 785 fiber-optiic probe comb bined with thee LuxxMasterr LML-785.0R RB-04 laseer module (7855 nm wavelenngth).
Fig. F 1. Experimenntal setup: 1 – obbject; 2 – Raman n probe RPB785; 3 – laser modulee LuxxMaster LM ML-785.0RB-04;; 4 – laser power supply mod dule; 5 – spectrom meter Shamrock srr-303i; 6 – CCD camera ANDOR DV-420A-OE; 7 – computer; 8 – transmitting fibeer; 9 – receiving fiber f
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
Due to the low level of autofluorescence the 785 nm wavelength of the visible wavelength range is well-proven in Raman spectroscopy for biological applications. Procedure description for work with the experimental stand earlier was presented in the article [Timchenko E.V. et al. (2014)]. 3. Results Before data analysis, the obtained Raman spectra of research objects were preprocessed in MatLab’7 for deletion of setup noise and fluorescence background using an approximation of the spectral curve by a fifth order polynomial, based on the algorithm proposed in the paper of [Zhao J. et al. (2007)]. A feature of the algorithm is a sequential and many iterative comparisons of the initial spectrum with a polynomial approximation of autofluorescence component, and the preservation of spectral contours and intensity of the initial Raman spectrum positively proved the algorithm in spectroscopy for biomedical research [Huang Z. et al. (2011)]. However, during the spectra study was found that at the large spectral range (exceeding the 1500-3000 cm-1) the approximation of spectral curve led to cuts of Raman band near more intensive neighboring band such as 1003 cm-1 (Ȟ (C-C) of phenylalanine) near 956 cm-1 (Ȟ1 (PO4)3- of hydroxyapatite). Therefore, the processing was carried out at the spectral regions 400-1150, 1200-1500, 1600-1800 and 2800-3000 cm-1 with appropriate criteria of approximation for sensitivity increase and maximum display without loss of informative bands. Informativeness of Raman spectra of bone tissue samples depends on the complexity and structure of studied biomaterial (on the number of bands in Raman spectrum), because the resulting spectrum is a superposition of several spectral bands that leading to distortion of the real values of intensity. Therefore, to extract the Raman bands against the background of spectral curve, the decomposition of obtained spectra into spectral contours using deconvolution of Gauss-Lorentz function was conducted by PeakFit software (Systat Software, Inc.). Specific Raman spectrum for non-demineralized cortical human bone is presented in Fig.2. The determination coefficients of the resulting spectrum (generated by the spectral contours) from the initial spectrum for 790-1140 cm-1, 1200-1480 cm-1, 1590-1790 cm-1 and 2840-3020 cm-1 regions were R2 = 0.9999, 0.9978, 0.9998 and 0.9993 respectively. The spectral region (Fig. 2a) contains the Raman bands responsible for the mineral structure of bone matrix, such as Ȟ1 (PO4)3- vibrational mode in the molecule of hydroxyapatite at 956 cm-1, Ȟ3 (PO4)3- at 1030 and 1045 cm-1, Ȟ1 (CO3)2- (B type substitution where (PO4)3- a substituted (CO3)2-) at 1069 cm-1 and Ȟ1 (CO3)2- vibration mode (A type substitution where a substituted OH- (CO3)2-) at 1098 cm-1 [Yamamoto T. et al. (2011), Crane N.J. et al. (2013)]. Band at 956 cm-1 is dominant by the intensity throughout all Raman spectrum of bone tissue, which determines the content of hydroxyapatite in bone matrix, and also possible to study the content of carbonate (CO3)2- in the process of demineralization. Less intense bands at 816, 852, 920, 941 and 870 cm-1 corresponding to the stretching vibration of C-C in proline and hydroxyproline [Yamamoto T. et al. (2011), Crane N.J. et al. (2013)], and the band at 1003 cm-1 responsible for C-C stretching vibration in phenylalanine [Crane N.J. et al. (2013)] represent the organic component of bone matrix. Further, Fig. 2b shows the spectral region responsible for structural integrity of bone collagen, where 1239, 1264, 1284, 1316, 1338 and 1381 cm-1 bands correspond to amide III (modes of Ȟ (CN) and į (NH)) [Yamamoto T. et al. (2011), Crane N.J. et al. (2013), Rehman I. et al. (2012), Poáomska M. et al. (2010)] and 1396, 1418 and 1450 cm-1 correspond to mode of į (CH2) and į (CH3) [Rehman I. et al. (2012), Poáomska M. et al. (2010)]. Spectral region of 1600-1700 cm-1 (Fig. 2c) in [Crane N.J. et al. (2013)] is a defining in the assessment of structural integrity of the collagen fibers. This region corresponds to amide I (stretching of C=O) with Raman bands at 1634, 1660, 1681 cm-1 (Į-helix and ȕ-sheet [Poáomska M. et al. (2010), Shen J. et al. (2010)]), and 1604 cm-1 (stretching vibration of C=C in tyrosine [Rehman I. et al. (2012)]). The most remote region of the spectrum (Fig. 2d) at 2879, 2910, 2938 and 2980 cm-1 corresponds to the stretching vibrations of CH2 and CH3 in collagen [Poáomska M. et al. (2010), Ali S.M. et al. (2013)].
223
224
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
b
a
d
c -1
-1
Fig. 2. Sppecific spectral coontours for region ns: (a) 790-1140 cm , (b) 1200-14 480 cm , (c) 1590-1790 cm-1, (d) 22840-3020 cm-1
In I this article,, the relative concentration c of organic coompounds in bone b matrix was w evaluatedd by the intensity of the amide III bannd (1239 cm-1), ) CH2 deform mation band (1 450 cm-1) and d amide I regio on (1660 and 1681 cm-1). In n turn, -1 man band at 9956, 1069 andd 1098 cm were w used forr investigation n changes in the t mineral co components off bone Ram mattrix. Further, F Fig. 3 shows the inntensity chang ge of Raman bbands for bone matrix samp ples at differennt demineralizzation timee relative to iintensity valuees of Raman bands for nonn-demineralizzed samples. The T intensity value at 2938 8 cm-1 werre used for intensity normallization of all spectral bandss:
In n( k )
I ( k )t I (k )0 I (2938)t
(1)
wheere I(k) t – Ram man intensity at a k wavenu umber with a t demineralizzation time of sample, I(k) 00 – Raman inttensity at a k wavenum mber without demineralizati d ion of samplee (t = 0), and I(2938) t – Raman intennsity at 2938 8 cm-1 wav venumber withh a t demineraalization time of sample.
225
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
a
b
c
d
m of: (a) I - In (956) and In (1069 ), II - In (1098) and a In (1069); (b) I - In (956) and IIn (1681), II - In (956) Fig. 3. The two-dimeensional histogram d In (1660); (c) Inn (956) and In (12239); (d) In (956) and In (1450) forr different demin neralization time: 1 - 0 min., 2 - 2 m min ., 3 - 5 min., 4 - 8 and min., 5 - 11 min., 6 - 15 min., 7 - 20 min., m 8 - 30 min., 9 - 45 min., 10 - 60 min. 11 - 90 min., m 12 - 120 minnutes
As A can be seeen from Fig. 3a, bands at 956, 1069 annd 1098 cm-11 are characteerized by a siignificant inteensity decrrease with dem mineralizationn time: at 5 minutes m of dem mineralization the values off In(956), In(11069) and In(1098) are 68.5, 6 77.3 andd 83.2 %, at 20 2 minutes - 55.1, 5 61.1 andd 76.2%, at 12 20 minutes - 32.8, 3 37 and 666.4% respectively. Fig. 3b shows the intensity channges in Ramaan bands of am mide I region, indicates the cross-links in collagen: the main d at 1660 cm-11 is characterizzed by an inteensity decreasee to 89.2% at 5 minute dem mineralization and to 83.2% at 20 band minu utes, but 1681 cm-1 is chaaracterized by y an intensityy increase to 19.4% at 5 minutes m and 337% at 20 minute m dem mineralization. Raman R intensitty at 1239 cm m-1, responsiblee for the amidde III group, during d all dem mineralization time varies by y less than n 5.6% (Fig. 33b), unlike thee band at 1450 0 cm-1 whosee intensity afteer 90 minute demineralizati d tion is decreassed to 89.1% relative too Raman intennsity of non-d demineralizedd bone matrix x sample (Fig g. 3d). Practiccally unchang ged in nsity the Ram man band of coollagen inform ms about an opptimal parameters of the deemineralized medium (HCl acid inten temp perature and cconcentration)) that does nott lead to deforrmation and destruction d of collagen. Avaailability of Raman R band ds, that corresppond to the coollagen structu ure and whosee intensity inccreases with in ncreasing dem mineralization time,
226
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
can be explained by the restruccturing into ch hemical compposition collag gen, associated with a recovvery or oxidattion of w with demineeralization tim me and reach h their CH-bond. Thus, the band inttensities at 1284 and 28779 cm-1 grow ximum of 1117.2 and 1177.1% at 20 minutes dem mineralization relatively non-demineraliized sampless with max subsequent intenssity relaxationn at 45-120 miinutes demineeralization. The T mineral composition change of investigated samples parrallel was monitored m by scanning electron miccroscopy. Miccrographs of sample s surfacce structure foor cortical bo one matrix dep pending on thhe demineraliization timee are presenteed in Fig. 4.
Fig. 4. The structture of cortical boone surface at diffferent demineraliization time for biomaterial: b (a) 0 min, (b) 2 min, (cc) 5 min, (d) 8 min. The a accelerating voltag ge - 20 kV, beam m current - 1 nA, number n of points - 16
In I Fig. 4 can oobserve as thee demineralizeed samples havve a certain deegree of bone surface roughhness (Fig. 4c) with the appearance oof pores at 5 and a 8 minutes demineralizaation. These ch hanges are the result of "w washing out" apatite a o the object. That causes the bones to lose most off their hardnesss and compreessive crysstals under ann acid action on streength, and at a large demineeralization tim me leads to defformation and destruction of the trabeculaae. 4. Conclusion C As A a result, Raaman spectra of human corttical bone witth a different demineralizati d ion degree weere obtained. Shown S thatt the demineraalization process can be qu uantitatively coontrolled usin ng the ratio ɊɈ Ɉ43- and ɋɈ322-: for 956 and d 1069 cm-1 at 5 minutes demineralizzation the inteensity value iss 68.5 and 77.3%, for 20 minutes m - 55.1 and 61.1%, at 120 min nutes - 32.8 annd 37% respeectively. The spectral featuures for bone collagen depending on dem n time mineralization werre investigatedd by analyzingg spectral regiions of 1200-11460 cm-1 and d 2880-3000 cm m-1. Ack knowledgemeent This T work wass supported byy the Ministry y of Educationn and Science of the Russian n.
E.V. Timchenko et al. / Physics Procedia 73 (2015) 221 – 227
227
References Ali S.M., Bonnier F., Tfayli A., et al., 2013, Raman spectroscopic analysis of human skin tissue sections ex-vivo: evaluation of the effects of tissue processing and dewaxing, Journal of Biomedical Optics. 18(6), 061202–10. Crane N.J., Polfer E., et al., 2013, Raman spectroscopic analysis of combat-related heterotopic ossification development, Bone. 57, 335–342. Flierl M.A., Smith W.R., et al., 2013, Outcomes and complication rates of different bone grafting modalities in long bone fracture nonunions: a retrospective cohort study in 182 patients, Journal of Orthopaedic Surgery and Research. 8, 33. Huang Z., Chen X., et al., 2011, Raman spectroscopic characterization and differentiation of seminal plasma, J. Biomed. Opt. 16, 110501. Ʉirillova I.Ⱥ., 2004, Demineralized bone graft as a stimulator of bone formation: modern concepts, Spine Surgery. 3, 105–110. Poáomska M., Kubisz L., et al., 2010, Fourier Transform Near Infrared Raman Spectroscopy in Studies on Connective Tissue, Acta Phys. Pol. A. 118(1), 138. Rehman I., Movasaghi Z., et al., 2012, Vibrational Spectroscopy for Tissue Analysis, Series in Medical Physics and Biomedical Engineering, CRC Press, 271. Shen J., Fan L., Yang J., et al., 2010, A longitudinal Raman microspectroscopic study of osteoporosis induced by spinal cord injury, Osteoporos Int, 21, 81–87. Slutsky L.I., Sevastianova N.A., 1986, Podiatrist., TRAUMA., 8 (82). Tarnowski C.P., et al., 2002, Mineralization of developing mouse calvaria as revealed by Raman microspectroscopy, Journal of Bone and Mineral Research. 17, 1118–1126. Timchenko E.V., Timchenko P.E., Volova L.T., et al., 2014, Raman spectroscopy of the organic and mineral structure of bone grafts, Quantum Electronics. 44 (7), 696–699. Timchenko P.E., Timchenko E.V., Zakharov V.P., et al., 2011, Diagnostics of bone implantat and control of their process osteointegration with of a method confocal microscopy, Computer Optics. 35(2), 183-187. West P.A., Torzilli P.A., et al., 2005, Fourier transform infrared imaging spectroscopy analysis of collagenase-induced cartilage degradation, Journal of Biomedical Optics. 10 (1), 014015. Yamamoto T., Uchida K., Naruse K., et al., 2011, Quality assessment for processed and sterilized bone using Raman spectroscopy, Cell and Tissue Banking. 409–414. Zhao J., et al., 2007, Automated autofluorescence background subtraction algorithm for biomedical Raman spectroscopy, Society for applied spectroscopy. 61(11), 1225–1232.