Study on Fracture Healing with Small-Splint-Fixation Therapy by Near-Infrared Raman Spectroscopy

In this study, near-infrared (NIR) Raman spectroscopy was explored to assess the incorporation of calcium hydroxyapatite (CHA ∼960 cm) and other biochemical substances during the recovery of rabbits with complete radial fractures treated with or without small splints. 24 rabbits were randomy divided into two groups, one treated with small-splint-�xation therapy and the other without any intervention.e rabbits were sacri�ced at 7, 15, 23, and 30 days a�er surgery, and the surface layers of the calluses in the fracture healing site from control and treated groups were routinely prepared for Raman spectroscopy. e prominent Raman bands were observed, includingminerals at 430, 590, 960, 1003, and 1071 cm, protein at 856, 876, 1246, and 1667 cm, and lipid at 1767 cm. e carbonate-to-phosphate ratio (CO3 to υυ1 PO4) and the mineral-to-matrix ratio (υυ1 PO4 to amide I) were calculated from these normalized Raman bands. Comparison of the υυ1 PO4-to-amide I ratio for the control group with that of the treated group probably indicated that the small-splint-�xation therapy could be useful for the gradual mineralization of the collagenmatrix during fracture healing.


Introduction
Bone is a special material integrating an organic matrix, principally type I collagen, and an inorganic component [1].Bone healing is an extremely complex process in human body, which has been under extensive investigation for many years [2].e �xation therapy is the important and basic method to treat the bone fracture, and it is the fundamental condition that the bone fracture canbe concrescent on time [3].Among numerous �xation therapies, elastic �xation therapy is the most reasonable and efficient technique for bone healing because, in elastic �xation therapy, the micromovement in the fracture ends of the bone is permitted, which can accelerate formation and calci�cation of callus and can be e�ective in promoting healing [4,5].As an important elastic �xation therapy, small-splint-�xation therapy �ts on the trend of the healing of the fracture [6], whose guiding idea is "dynamic and static combination, equal stress on bones and muscles, exterior and interior treatment, and collaboration between doctors and patients." Microscopic examination and imaging are the cardinal methods for the assessment of small-splint-�xation therapy in promoting bone healing of rabbits with complete radial fractures nowadays.However, the in�uence of small-splint-�xation therapy on bone healing is still vague.Although the macroscopic properties of bone have been extensively examined and are somewhat established, the early steps in its mineralization are not well understood.
Raman spectroscopy (RS) can provide �ngerprinting type information on the structure and conformation of macromolecules such as proteins, nucleic acids, and lipids [7].RS has drawn considerable attention due to its great potential for improving clinical diagnosis [8][9][10] and obtaining precise information on biochemical composition [11][12][13][14][15]. Several groups have already investigated the applications of laser Raman spectroscopy for bone and tooth chemistry studies [16][17][18].During the bone fracture healing process, immature bone is replaced by a mature type of bone, which incorporates inorganic components with calcium hydroxyapatite (CHA, Ca 10 (PO 4 ) 6 (OH) 2 ) being the most important one.Similar to CHA, the level of bone mineral can characterize the presence of mineral on the bone; however, there may be other mineral phases present in the early stages of mineralization such as amorphous calcium phosphate and tricalcium phosphate.According to some researches, in less mineralized bone, there is a larger relative amount of monohydrogen phosphate and a smaller relative amount of carbonate.And the degree of mineralization will vary associated with bone development from the initial stages of formation to full maturity [19,20].
In this study, NIR Raman spectroscopy was �rstly used to investigate the e�ects of small-splint-�xation therapy on the healing of complete radial fractures in rabbits by monitoring the level of CHA at 7, 15, 23, and 30 days aer surgery, using the P-O symmetric stretch band (960 cm −1 ) as a prominent marker band.Moreover, carbonate-to-phosphate ratio and mineral-to-matrix ratio are also calculated from the normalized Raman band areas to obtain some useful information about bone components.China)), the rabbits had their right legs shaved, and a 3 cm long incision was performed at the lower one-third part of radius with a n ∘ 15 scalpel blade.Skin and subcutaneous tissues dissected down to periostea, which were gently sectioned so as to expose the bones.A 3 mm complete radial fracture was surgically produced as fracture healing model (low-speed drill, 1200 rpm, under refrigeration) in each rabbit [2].All wounds were routinely sutured, and the rabbits received a single dose of benzylpenicillin sodium for injection (80.000UI, Huaxing, Henan, China) immediately aer surgery.

Randomization and Intervention
Methods.Aer operation, the 24 rabbits were divided into two groups randomly by SPSS (Version 17.0) [21], including a control group (  12) and a small-splint-�xation treated group (  12).ere was no internal or external �xation treatment for the control group.In the treated group, four pieces of small splints made by elastic �r tree barks are used in upper, lower, le, and right positions of the fractured leg, respectively, to �x the fracture bone well.Finally, small splints were tied up by bandage in order to prevent slippage.

Tissue Specimens Preparation
. ree rabbits from each group were humanely killed at 7, 15, 23, and 30 days aer the surgery with an overdose of general anesthetics, and the calluses from the fracture healing site were removed and longitudinally cut down to obtain the surface layer of the calluses as specimen by a Leica SP1600 saw microtome, which was specially designed for the cutting of extremely hard and brittle industrial materials.In order to minimize the growth of aerobic bacteria, the specimens were stored in li�uid nitrogen, and the chemical �xation was not advisable as a result of the �uorescence emissions from the �xative substances.Prior to Raman study, without decalci�ed procedure, the specimens were warmed gradually to room temperature, washed by PBS, and then for Raman measurement.measurement of Raman spectra.Brie�y, all wavelengthcalibrated spectra were corrected for the wavelength dependence of a standard 520 cm −1 vibrational band of a silicon wafer before Raman measurements and the soware package WIRE 2.0 (Renishaw) was employed for spectral acquisition and analysis.e microscope was operated under a Leica 20× objective, which focused the laser beam onto a spot on the sample surface.For each sample, Raman measurements were performed on twenty random positions and two spectra were acquired from each position by repeated measurements.erefore, in total forty Raman spectra were obtained from each sample.e signal was integrated for 10 s and measured over a spectral range of 400 to 1700 cm −1 .e raw spectra of fracture healing sample acquired from control group and small-splint-�xation-treated groups in the 400-1700 cm −1 range represented a combination of prominent auto�uorescence, weak Raman scattering signals, and noise.us, the raw spectra were preprocessed by adjacent �ve-point smoothing to reduce noise.�nd, then, an automated algorithm for auto�uorescence background removal was applied to extract pure Raman spectra.e program was kindly offered by the BC Cancer Research Centre [22].en each of the background-subtracted Raman spectra was normalized to the integrated area under the curve from 400 to 1700 cm −1 to enable a better comparison of the spectral shapes and relative peak intensities among the different samples [7].

Results and Discussions
e fracture healing and bone strength are mainly affected by the organization of its mineralized collagen �bers and material composition.Important aspects of bone tissue quality include the physicochemical properties of its main constituents, the organic matrix, and the mineral crystals.Raman bands are sensitive to the distinction between structural and chemical changes in the compositional contribution of bone.
Figure 1 shows the mean Raman spectrum of normal bone tissue.Prominent Raman bands are observed in the following peak positions with tentative biomolecular assignments: 430 cm −1 ( 2 vibration of phosphate), 590 cm −1 ( 4 vibration of phosphate), 960 cm −1 (P-O symmetric stretch vibration of hydroxyapatite), 1003 cm −1 (phenylalanine ring breathing), 1071 cm −1 (C-O in plane stretch vibration of carbonate hydroxyapatite), 1246 cm −1 (C-N-H stretch vibration of Amide III), 1450 cm −1 (C-H bending mode of protein and lipid), 1667 cm −1 (C-C-N stretch vibration of amide I), and so on.Table 1 lists tentative assignments for the observed Raman bands [1,2,17].From Table 1, we could obtain that the molecular species of bone samples can be grouped into two principal components, the organic phase of the bone (proteins and lipid) and the inorganic (mineral) phase of the bone (phosphate, monohydrogen phosphate, and carbonate).
It was previously demonstrated that Raman band intensities depend strongly on both the chemical composition of the tissue and the �ber orientation with respect to the linear polarization direction of the exciting laser beam [23][24][25].According to some literature data, biochemical composition and �ber orientation were determined from the following Raman bands: the phosphate (PO 4 −3 ), including  1 PO 4 (integration area from 930 cm −1 to 980 cm −1 ), the  2 PO 4 (integration area from 410 cm −1 to 460 cm −1 ), and  4 PO 4 bending vibrations at 590 cm −1 , the amide I (integration area from 1620 cm −1 to 1700 cm −1 ) and amide III (integration area from 1215 cm −1 to 1300 cm −1 ), and the carbonate (CO 3 −2 , integration area from 1050 cm −1 to 1100 cm −1 ) [25].e hydroxyapatite in bone mineral is characterized principally by the phosphate stretch ( 1 PO 4 ) spectral position, whose Raman band is at 960 cm −1 .Figure 2 shows the mean normalized Raman spectra of the surface layer of the calluses in the fracture healing site from control and treated groups at 7, 15, 23, and 30 days, respectively.As shown in Figures 2(a)-2(d), there was no signi�cant shi� in any band position throughout the data set; however, the constantly increasing intensity of 960 cm −1 peak was observed both in control and treated groups.As the 960 cm −1 band (phosphate  1 PO 4 ) is a prominent marker for mineral content in bone, this result may indicate gradual mineralization during the fracture healing.Besides, the 960 cm −1 peak ( 1 PO 4 ) from the treated group shows higher percentage signals than that from the control groups in the same period of fracture recovering (7, 15, 23, and 30 days), probably indicating an increase in the percentage of hydroxyapatite content relative to the total Raman-active components in the treated group [1,26,27].
Although the Raman bands, such as  1 PO 4 and amide I, are quite sensitive to the orientation and the polarization direction of the incident light, Raman bands such as amide III,  2 PO 4 , and  4 PO 4 are less susceptible to the orientation effects [24,25].erefore, Raman band areas obtained from the normalized curves were used to calculate several useful ratios of bone components.Carbonate is the most common substitution in the bone mineral lattice.e predominant substitution is type B carbonation (CO 3 −2 replacing a PO 4 −3 ), containing a small amount of type A carbonation (CO 3 −2 in place of OH − ) [1]. ere is a strong band at 1071 cm −1 indicating B-type carbonate substitution in the bone specimen.Figure 3(a) shows the ratio of the carbonate band at 1071 cm −1 to the phosphate band ( 1 PO 4 , 960 cm −1 ) between control group and treated group as a function of the time at 7, 15, 23, and 30 days aer surgery.is ratio was calculated by dividing the integrated areas of the carbonate band (CO 3 −2 , integration area from 1050 to 1100 cm −1 ) and the phosphate band ( 1 PO 4 , integration area from 930 cm −1 to 980 cm −1 ), a recognized measure of mineral maturity [17].As shown in Figure 3(a), a slight increase in the carbonate-to-phosphate ratio (CO 3 −2 /PO 4 −3 ) was observed both in control and treated groups during the experiment.In addition, Figure 3(a) also shows that the carbonate-to-phosphate ratio of the treated group is higher than that of the control group at the same healing time.e maturation of the mineral in bone is indicated by the changes in the phosphate and carbonate modes.is change of carbonate-to-phosphate ratio is consistent with some studies such as bone tissue maturation [28] and mineralization of developing bone [29].Meanwhile, this result indicated that an increase in carbonate occurs with increasing mineralization in bone tissue [1].
Figure 3(b) shows the variation of the mineral-to-matrix ratio between control group and treated group during the fracture healing.e mineral-to-matrix ratio, one of the most useful measurements for Raman analysis of bone, has been shown to be a marker of bone mineralization [17].e mineral-to-matrix ratio was obtained from the integrated areas of any of the phosphate and amide peaks, which depend on both tissue organization and composition [24,25].is ratio was calculated by dividing the area of the phosphate band ( 1 PO 4 , integration area from 930 cm −1 to 980 cm −1 ) and the amide I band (integration area from 1620 cm −1 to 1700 cm −1 ) [17].As shown in Figure 3(b), the increase of the  1 PO 4 /amide I ratio occurs both in control group and treated group during the progress of fracture healing, which indicates the degree of mineralization, and the change in mineral-tocollagen ratio [23].Figure 3(b) also shows that the treated group retained higher mineral-to-matrix ratio compared with the control group.is result probably indicated that the small-splint-�xation therapy could be useful for the gradual mineralization of the collagen matrix during fracture healing.

Conclusions
High-quality Raman spectra in the 400-1700 cm −1 range from rabbit bone tissue specimens in vitro were obtained by NIR Raman measurements.e prominent Raman peaks, such as 430, 590, 856, 876, 960, 1003, 1071, 1246, 1450, 1667, and 1767 cm −1 corresponding to biomolecules in bone tissue, were detected.ese bands represented two principal components of bone tissue, including the organic components and the minerals.e promotion of fracture healing interfered by small-splint-�xation therapy was accessed by the analysis of phosphate band ( 1 PO 4 , 960 cm −1 ) during the experiment, which was a prominent marker for mineral content in bone.Other phosphate and carbonate vibrations could show the subtle changes in mineral composition, and the amide I and amide III could be used as typical markers in organic composition.e mineral-to-matrix ratio compared between the control group and the treated group probably indicated that the small-splint-�xation therapy could be useful for the gradual mineralization of the collagen matrix during fracture healing.

4 F 1 :
e mean Raman spectrum of the normal bone tissue sample.e shaded areas represent the standard deviations of the means.

F 2 :
Processing.A confocal Raman microspectrometer (Renishaw, UK) with a 200 mW, 785 nm diode laser excitation was used for the Normalized raman intensity (a.u.) Mean normalized Raman spectra of the bones from control group and treated groups measured at (a) 7 days, (b) 15 days, (c) 23 days, and (d) 30 days aer surgery.