Article pubs.acs.org/ac
Effects of Ovariectomy on Rat Mandibular Cortical Bone: A Study Using Raman Spectroscopy and Multivariate Analysis Xiaoming Fu,† Jiang Chen,*,‡ Dong Wu,‡ Zhibin Du,§ Qun Lei,† Zhiyu Cai,† and Stefan Schultze-Mosgau⊥ †
School of Stomatology, Fujian Medical University, Fuzhou, Fujian 350000, China Department of Oral Implantology, Affiliated Stomatological Hospital of Fujian Medical University, Fuzhou, Fujian 350002, China § Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland 4001, Australia ⊥ Department of Oral and Maxillofacial Surgery/Plastic Surgery, University of Jena, Jena, Thuringia 07747, Germany ‡
ABSTRACT: To investigate the effects of ovariectomy (OVX) on rat mandibular bone, the physicochemical compositions of mandibular cortical bone of ovariectomy and sham operated rats 2, 4, and 8 months after surgery were compared using Raman spectroscopy. With principal component analysis and linear discriminant analysis based on the Raman spectra, the mandibular cortical bone of the OVX group was clearly distinguished from that of the sham-operated group 8 months after surgery with no overlap. Specifically, significant reductions in the mineral-to-matrix ratio and full width at half-maximum as well as a significant increase in the carbonate-to-phosphate ratio were observed in the mandibular cortical bone of the OVX group. Results support the hypothesis that Raman spectroscopy is sensitive enough to distinguish between OVX and sham-operated mandibles with multivariate analysis by detecting the chemical composition of the mandibular cortical bone. The parameters mineral-tomatrix ratio, carbonate-to-phosphate ratio, and full width at half-maximum can appropriately characterize changes in the chemical composition of the mandibular cortical bone after OVX.
Ovariectomy (OVX) is widely used to establish an experimental model of postmenopausal osteoporosis. With a continuous bone loss after OVX, it is confirmed that OVX could induce systemic osteoporosis ultimately in female rats. However, skeletal sites have been observed to be not equally susceptible to OVX-induced bone loss. Bone mass is reduced obviously in the tibiae, femora, and vertebrae of ovariectomized rats.1−4 On the contrary, skull vaults do not develop osteoporosis at all.5 Many studies have concentrated on the effects of OVX on jawbones as well, but the extent of jawbone degradation postOVX6,7 at the ultrastructural level remains unclear. Research has shown that osteoporosis occurs in the jawbone after OVX. Tanaka et al.8 histomorphometrically analyzed trabecular structural changes in the mandibular alveolar bone of ovariectomized rats and found osteoporotic changes and a thin alveolar bone in the interradicular septum of the rat first molar. Using a standardized radiographic protocol, Yang et al.9 found that the mean mandibular cortical thickness of ovariectomized mandibles was significantly less than that of sham-operated ones. With high-resolution micro-CT (15 μm), another study reported that OVX decreased the bone volume/soft-tissue volume ratio and trabecular thickness while significantly increasing trabecular separation and the structure model index in the mandible.10 However, these findings contradict other studies. In their histomorphometric analysis of maxillae in ovariectomized rats, Ishihara et al.11 reported that the net bone area and bone structure in the maxillae of ovariectomized rats were similar to those of sham-operated rats. Miller et al.12 also found a few differences in the histomorphometric indices of © 2012 American Chemical Society
bone formation in rat mandibles more than 1 year after OVX or sham operation. Clearly, traditional histomorphometric studies of the effects of OVX on the rat jawbone have presented conflicting data. Furthermore, the rat has not haversian systems in the cortical bone. Morphological changes in the mandibular or maxillary cortical bone after OVX have not been reported as well. Identifying a novel way to assess the quality of jawbones is thus crucial to monitor the effects of OVX. Cortical geometry, cancellous architecture, and material composition are used to define bone quality.13−15 Detecting changes in material properties is more sensitive to monitoring variations in the jawbone after OVX. In recent years, more investigators have used Raman spectroscopy to detect and quantify the bone organic matrix and minerals with microlevel spatial resolution. For example, one study used Raman spectroscopy to identify differences between the chemical composition of the cortical bone of C57 Black 6 mice with that of their trabecular bone.16 Another study that focused on agerelated changes in the physicochemical properties of mineral crystals in rat cortical bone discovered that the elastic deformation capacity of aged rats is significantly impaired both at the tissue level and the organ level with increasing age.17 Other studies have demonstrated lower mineral and matrix contents, decreased mineral-to-matrix ratios, and Received: January 5, 2012 Accepted: March 7, 2012 Published: March 7, 2012 3318
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
Analytical Chemistry
Article
representative of each mandibular cortical bone. In addition, the average spectrum of each group was calculated by adding all the spectra and dividing the sum by the number of spectra. Comparisons of mean Raman spectra between the SHAM and OVX groups 2, 4, and 8 months postoperation showed a difference spectrum generated by subtracting the intensity of the mandibular cortical bone spectrum in the OVX group from that in the SHAM group at each wave. Peak Parameters. Much research has shown typical spectra from bone. Phosphate ν1 (959−960 cm−1), phosphate ν2 (438 cm−1), phosphate ν4 (589 cm−1), monohydrogen phosphate (1003 cm−1), carbonate β (1070 cm−1), amide III (1250 cm−1), CH2 (1451 cm−1), and amide I (1668 cm−1) have been reported in both mineral and organic molecules.24,25 Ratios of phosphate ν1/amide I, carbonate/phosphate, and hydrogen phosphate/phosphate were measured in this study. PCA−LDA Analysis. PCA is a multivariate technique used in Raman spectroscopy in which Raman spectral data are represented by several intercorrelated quantitative dependent variables called principal components (PCs). In this work, PCA in combination with t test was performed on the spectral data to distinguish between the mandibular cortical bone of the rats in the OVX group and that of the rats in the SHAM group at three time points. The scores of the most significant PCs were incorporated into an LDA model26,27 to perform cluster discriminant analysis based on posteriori probability and determine the sensitivity and specificity of distinction between OVX and SHAM rats at different time points. A receiver operating characteristic (ROC) curve was generated to further evaluate the performance of the PCA−LDA-based diagnostic algorithm for the OVX and SHAM groups.
increased carbonate-to-phosphate ratios in osteoporotic rats.18,19 In this study, we evaluated the mandibular cortical bone of ovariectomized and sham-operated rats (OVX and SHAM groups, respectively) to determine the effects of OVX on the mandible using Raman spectroscopy. Changes in various mineral and organic contents were calculated as the peak parameters. On the basis of the Raman spectra subjected to principal component analysis (PCA) and linear discriminant analysis (LDA), we distinguished mandibles of the OVX group from those of the SHAM group.
■
MATERIALS AND METHODS Animals. Forty-eight female Sprague−Dawley rats (3 months old, 160−180 g) were obtained from SLAC Laboratory Animal Co. Ltd. (Shanghai, China). The animals were housed under a 12 h light/12 h dark cycle with controlled temperature (22 ± 1 °C) and humidity (50%−60%), and they were given free access to water and food, which was standard rat chow without phytoestrogens. After 1 week of adaptive feeding, the rats were randomly divided into the OVX group and the SHAM group according to their weight. Both ovaries of the rats in the OVX group and equal volumes of fat tissue next to each ovary of the rats in the SHAM group were excised with the animals under general anesthesia, as previously described by Du.20 Sixteen rats from each group were humanely killed 2, 4, and 8 months after surgery with an overdose of general anesthetics. All experimental procedures were performed in compliance with Institutional Animal Care and Use Committee requirements for animal care and use and approved by the Animal Care and Use Committee of Fujian Medical University. Mandibles. The bilateral mandibles of each animal were harvested with the surrounding tissue carefully removed, after which the mandibles were stored at −80 °C.19,20 After all samples had been collected, sections of approximately 1 mm thickness were transversely cut at the first molar site using a mineralogical saw (Accutom 2; Struers, Glasgow) and fitted with a diamond cutoff wheel under constant irrigation with deionized water. Slices were treated by ultrasound until no particles adhered to the surface of specimens under a stereomicroscope. Confocal Raman Microscopy. A confocal Raman microspectrometer (Renishaw, U.K.) was used to record the chemical composition of the mandibular cortical bone in the range 400− 1800 cm−1 under a 785 nm diode laser excitation. The light was linearly depolarized, and the spectra were collected using a Leica microscope with 20× objective. A Peltier cooling chargecoupled device camera was used to detect the Raman signal of the mandibular cortical bone at a spectral resolution of 2 cm−1. Each Raman spectrum was the result of two accumulations, each with a 10 s exposure time at room temperature (24 ± 1 °C).21 Alignment with the optical axis of the microscope was checked at the beginning of Raman spectral data accumulation to ascertain the comparability of spectral intensities, and silicon band intensities were obtained to ensure that the value was the same each time as an internal standard.22 All measurements (n = 12) per mandible (n = 16 each group) were performed in a randomized manner on the buccal cortical bone, and the data were stored in a microcomputer. Difference Raman Spectra. The sample background fluorescence and noise were subtracted using Vancouver Raman Algorithm.23 The average spectrum from each sample, which was calculated with Microcal Origin 8.0, was
■
RESULTS Difference Spectra. A section of the mandibular bone was shown in Figure 1A. The buccal side of the cortical bone was detected using Raman microspectroscopy at selected areas of each Raman bright-field microscope image (Figure 1B). All 12 spectra per sample were highly overlapping (Figure 1C). An average spectrum was calculated for each group. The main Raman spectra of the mandibular cortical bone of the rats in the OVX group and that of the rats in the SHAM group had similar peak positions, but their intensities (peak heights and peak areas) at certain Raman shifts differed. Peak heights and peak areas were calculated, with the results showing no difference, and peak heights were selected to represent Raman intensity in this study. Plots of the average difference spectra from the SHAM and OVX groups 2, 4, and 8 months after surgery are shown in Figure 2. The intensity of the minerals from the mandibular cortical bone of the OVX group was larger than that of the SHAM group 2 months after surgery, but the intensity of the matrix of the OVX group was smaller (Figure 2A). The intensities of both the minerals and organics were smaller in the OVX group compared with those in the SHAM group 4 months after surgery (Figure 2B). Furthermore, 8 months after surgery, all the main compositions, including minerals and organics, were significantly smaller in the OVX group (Figure 2C). These data showed that the amount of minerals was greater in the OVX group for the first 2 months postoperation but decreased from 4 months after surgery onward. Meanwhile, the composition of the matrix in the OVX group was continuously smaller than that in the SHAM group at each time point.
3319
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
Analytical Chemistry
Article
Figure 3. Peak parameters of mandibular cortical bone 2, 4, and 8 months after surgery: (A) mineral-to-matrix ratio; (B) carbonate-tophosphate ratio; (C) hydrogen phosphate-to-phosphate ratio; and (D) fwhm. *P < 0.05, OVX group versus SHAM group 8 months after surgery.
average Raman spectrum of each group. The sum of PC1, PC2, and PC3 was greater than 85%. Independent-samples t test on the PC1, PC2, and PC3 scores comparing the two groups showed that PC1, PC2, and PC3 were diagnostically significant (P < 0.05). The results of PCA based on the Raman spectra are shown in Figure 4. Until the fourth month postoperation, the
Figure 1. (A) Section of mandible bone under a stereomicroscope (original magnification ×100). (B) Raman bright-field microscope image of mandibular cortical bone marked by a square in part A (original magnification ×200). (C) Twelve overlapping Raman spectra of each sample tested as shown in part B.
Figure 4. PCA scatter plots comparing the Raman spectra of mandibular cortical bones between the OVX group and the SHAM group: (A) 2 months after surgery; (B) 4 months after surgery; and (C) 8 months after surgery. (A, B) Overlapping clusters of OVX and SHAM mandibles 2 months (A) and 4 months (B) postoperation. (C) Distinction between the OVX and SHAM groups 8 months after surgery (red circle, mandibular cortical bone of the OVX group; black star, mandibular cortical bone of the SHAM group).
Figure 2. Average difference spectra of mandibular cortical bones from the SHAM and OVX groups: (A) 2 months postoperation; (B) 4 months postoperation; and (C) 8 months postoperation.
Peak Parameters. The mineral-to-matrix ratio (phosphate ν1/amide I) of the mandibular cortical bone of the SHAM group (Figure 3A) increased with time, but the difference was not statistically significant. Compared with that in the SHAM group, the mineral-to-matrix ratio in the OVX group (Figure 3A) was greater during the first 2 months after surgery but became smaller from 4 months onward. On the eighth month postoperation (Figure 3A), the difference was significant (Tukey’s honestly significant difference, P < 0.05). The carbonate-to-phosphate ratio (CO32−/ PO43−) was greater in the OVX group at all time points, but the difference became significant only on the eighth month after surgery (Figure 3B) (Tukey’s honestly significant difference, P < 0.05). Moreover, the hydrogen phosphate-to-phosphate ratio (HPO42−/PO43−) was smaller (Figure 3C) in the OVX group, with the difference not being significant as well. The full width at half-maximum (fwhm) of the phosphate ν1 peaks (Figure 3D) was narrower in the mandibular cortical bone of the rats in the OVX group. On the eighth month after surgery, the difference in fwhm between the two groups became significant (t test, P < 0.05). PCA−LDA Analysis. The first PC (PC1) accounted for more than 60% of the variance within the Raman spectral data of the OVX and SHAM groups. It corresponded with the
mandibular cortical bone of the rats in the SHAM group and that of the rats in the OVX group overlapped for some cases and could not be distinguished by any clear delineation between the groups, indicating that the difference between groups was inconspicuous. Nonetheless, 8 months after surgery, the mandibular cortical bones of the OVX and SHAM groups were distinguished at different regions with high sensitivity (100%) and specificity (100%). In consistency with the clustering trend, LDA confirmed that the mandibular cortical bone of the OVX group was clearly distinguishable from that of the SHAM group, as shown in Figure 5. The area under the ROC curve increased with time. On the eighth month postoperation, it reached 1.000 (Figure 6).
■
DISCUSSION We investigated the effects of OVX on rat mandibular cortical bone using Raman spectroscopy. To our knowledge, this is the first study to have explored the composition of the mandible using Raman spectroscopy. The amount of minerals was greater in the OVX group for the first 2 months postoperation but 3320
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
Analytical Chemistry
Article
According to previous studies in tibiae, femora, and vertebrae, OVX significantly decreases the relative mineral content with time. The results we obtained during the first 2 months after surgery contradicted those of previous studies. The amount of minerals was greater in the OVX group than in the SHAM group for the first 2 months postoperation and then decreased. This difference may be attributed to the specialty of the jawbone. Ishihara32 predicted that prominent bone loss occurs mainly in areas formed by endochondral ossification, such as distal femora, and that those areas formed by intramembranous ossification, such as mandibles and maxillae, are less affected by OVX. Another study reported that greater periosteal expansion requires more resistance applied to loads.33 These findings may be a result of intramembranous ossification not being sensitive to estrogen deficiency compared with endochondral ossification. Elsubeihi34 also found that intermittent functional loading related to biting force prevents bone loss in the dentate mandible, which differs from the case of long bones and vertebrae, which suffered sustained and much larger load. Moreover, OVX has been shown to increase osteogenic ability. All above explained why in the present study mineral content did not decrease immediately after OVX, as previously reported for vertebrae, tibiae, and femora, but increased slightly during the first 2 months. This certainly does not imply that OVX had no effect whatsoever, as the mineral content significantly decreased 8 months after surgery. With increased expressions of interleukin-1, interleukin-6, transforming growth factor, and other cytokines, OVX was found to stimulate the formation, recruitment, and activity of osteoclasts.35 On the other hand, OVX induced the downregulation of insulin-like growth factors, transforming growth factor β, and osteoprotegerin. In this way, OVX restrained the activity of osteoclasts and the survival of osteocytes,36 resulting in a disrupted balance of bone resorption and bone formation, ultimately leading to low bone mass in ovariectomized rats. However, markers of bone metabolism could not directly predict the quality of bone. Bone mineral density (BMD) was applied conventionally to quantify the variations in bone mass and to thus reflect changes in bones. Dual energy X-ray absorptiometry (DEXA) is considered the gold standard for quantifying bone mass in the spine and tibia, but DEXAmeasured BMD was found to account for only 60%−70% of the variation in bone strength.37 Recently, some investigators have indicated that a large overlap in the BMD values of patients with and those without a fracture exists, rendering bone mass alone insufficient to predict the risk of fractures.38−41 Magnetic resonance imaging and micro-CT analysis are generally reliable, but they provide architectural information only, not compositional information, which is also vital to bone quality. Use of Raman spectroscopy to obtain new insights into bone quality in the study of mandibular cortical bone after OVX could not only reflect the result of bone metabolism but also indicate the physicochemical properties of bone, regardless of the morphological and radiological changes in the bone. The mineral-to-matrix ratio was calculated by dividing the intensity of the phosphate symmetric stretch band (∼960 cm−1) by amide I (∼1670 cm−1). The intensity of the phosphate ν1 peak has been reported to be the most appropriate band for evaluating the phosphate level, which was influenced to a minor extent by environmental factors among the four vibrational modes of PO43−.19 In contrast, the amide I band has been used as a measure of collagen content.42
Figure 5. LDA scatter plots comparing the mandibular cortical bone of the rats in the OVX group with that of the rats in the SHAM group based on posteriori probability, which measures the likelihood that an event will occur given that a related event has already occurred, at three time points: (A) 2 months after surgery; (B) 4 months after surgery; and (C) 8 months after surgery (red triangle, mandibular cortical bone of the OVX group; black triangle, mandibular cortical bone of the SHAM group).
Figure 6. ROC curve of the discrimination results for the mandibular cortical bones of the OVX and SHAM groups: (A) 2 months after surgery (area under the ROC curve, 0.825); (B) 4 months after surgery (area under the ROC curve, 0.918); and (C) 8 months after surgery (area under the ROC curve, 1.000).
eventually decreased compared with that in the SHAM group. Meanwhile, the amount of organic matrix in the OVX group was always lower than that in the SHAM group at the same time point. In accordance with these changes, on the basis of the acquired Raman spectra, PCA−LDA completely classified the mandibular cortical bones of the OVX and SHAM groups into two categories with high sensitivity and specificity on the eighth month after surgery. This allowed us to clearly distinguish the mandibular cortical bone of the rats in the OVX group from that of the rats in the SHAM group based on the data we obtained about the chemical composition of the mandibular cortical bone with regard to the following characteristic parameters: mineral-to-matrix ratio, carbonateto-phosphate ratio, carbonate-to-amide I ratio, hydrogen phosphate-to-phosphate ratio, and fwhm. Bone tissue is a synthesis of a set of known minerals and bone matrix. Raman vibrational spectroscopy can provide sitespecific information on bone mineral and matrix contents, collagen maturity, degree of mineral crystallinity, and composition.28 In this study, all measurements per mandible (n = 12) were performed in a randomized manner on the buccal cortical bone. We found that all 12 spectra of each sample were highly overlapping, which may be the result of the homogeneous structure of the cortical bone. In addition, a depolarized laser beam was applied to eliminate variations due to tissue orientation. Compared with those for the variable trabecular bone,17 the results for the cortical bone would be more stable and reliable. Furthermore, several studies reported a novel method of assessing bone using transcutaneous Raman spectroscopy that can directly and noninvasively measure the bone through the superficial skin.29−31 Therefore, as the first step, we chose to detect the cortical bone, not the deeper trabecular bone, to determine the effects of OVX on the rat mandible. 3321
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
Analytical Chemistry
Article
mandibles. PCA assessed intensities at all wave numbers of the Raman spectra with a few important PCs. Figure 4 illustrates that the PC1, PC2, and PC3 scores for the mandibular bone of both groups formed distinct clusters. Panels A and B of Figure 4 specifically show that clusters of OVX and SHAM mandibles overlapped and could not be separated by any straight line 2−4 months after surgery. On the eighth month postoperation (Figure 4C), the mandibular cortical bone of the rats in the OVX group can be distinguished from that of the rats in the SHAM group with a high resolution ratio (100%). LDA used the PC scores to create diagnostic algorithms based on posteriori probability and ultimately to incorporate all significant Raman spectral bands. Figure 5 shows that variance increased with time between the OVX and SHAM groups, whereas the variance within the OVX group became smaller. Eight months after surgery, the mandible of the rats in the OVX group could be easily distinguished from that of the rats in the SHAM group without overlapping, on the basis of the Raman spectra. This confirms once more that using Raman spectroscopy with LDA analysis completely distinguished between OVX mandibles and SHAM mandibles on the eighth month after surgery. An ROC curve (Figure 6) was generated to further verify the PCA−LDA-based diagnostic algorithm for the ovariectomized mandible. The area under the ROC curve increased with time, being 1.000 on the eighth month postoperation. These results demonstrated that the mandibular cortical bone of both groups can be distinguished with high sensitivity (100%) and specificity (100%) 8 months postoperation.
The mineral-to-matrix ratio significantly decreased in the mandibular cortical bone of the rats in the OVX group 8 months after surgery. Studies of the mineral-to-matrix ratio in the femoral neck, femoral midshaft, proximal tibial metaphysis, tibial midshaft, and lumbar vertebral body of ovariectomized rats1,2,43 also found that it decreased in the OVX group compared with the SHAM group. These data suggest that all bones (various sites and types), including the mandible, changed after OVX as the mineral-to-matrix ratio decreased and that the reduction in the amount of minerals was relatively greater than that of organic matrix. Carbonated apatite as a natural part of bone minerals as well as a carbonate substitute for hydroxide (A type) and phosphate (B type) can be detected by Raman spectroscopy. The A type is very weak at ∼1104 cm−1, whereas the B type is detectable at ∼1070 cm−1. The ratio of carbonate (∼1070 cm−1) to phosphate (∼960 cm−1) was calculated in this work and found to be greater in the OVX group, in accordance with previous studies showing that the chemistry of the mineral changed with increased carbonate-to-phosphate ratio in ovariectomized subjects. The ratio of hydrogen phosphate (∼1070 cm−1) to phosphate (∼960 cm−1) was smaller in the mandibular cortical bone of the OVX group, but the difference was not significant. The carbonate-to-phosphate ratio increased but the hydrogen phosphate-to-phosphate ratio decreased after OVX compared with sham operation. We easily determined that these changes in ovariectomized rats were somehow in accordance with age-related alterations in bone, except for the mineral-to-matrix ratio. The carbonate-to-phosphate ratio and HPO42−-to-phosphate ratio give an indication of age-related changes in bone.3,44,45 Furthermore, the ovariectomized bone and aging bone are known to be characterized by decreased organic content and increased fragility. These data suggest that, to a certain extent, the changes in such bones may be a result of a common pathological mechanism, which explains why men suffer from osteoporosis as well. The fwhm of the phosphate ν1 peak in the SHAM group was becoming narrower with aging, although significant changes in fwhm were not observed. At the same time point, compared with that in the SHAM group, the fwhm was narrower in the OVX group. On the eighth month after surgery, the change in fwhm in the OVX group was significant. This narrower fwhm meant that the degree of crystallinity of the mineral increased16 because of aging17 and OVX. Moreover, the present study found the OVX-induced changes in crystallinity to be more apparent than the age-related ones. As fully crystallized alloys that behave similarly to very brittle ceramics, the mandibular cortical bone of the OVX group would be more friable and on occasion easier to crack and fracture given its increased crystallinity. The above-described variables suggest that the mineral-tomatrix ratio, carbonate-to-phosphate ratio, and fwhm are the most appropriate parameters for characterizing changes in the chemical composition of the mandibular cortical bone after OVX. The degradation of the physicochemical property of the mandibular cortical bone in the OVX group could lead to many clinical phenomena: greater absorption of the alveolar bone, thinner mandibular cortical bone, diminished loading capacity of the alveolar bone, greater tooth loss, lower survival rate of implanted teeth, and greater susceptibility to mandibular fracture. This study used PCA−LDA multivariate analysis to distinguish between OVX group mandibles and SHAM group
■
CONCLUSIONS This study evaluated the effects of OVX on rat mandibular cortical bone with progressive physicochemical degradation. Raman spectroscopy was sensitive enough to monitor the composition of the mandibular cortical bone. PCA−LDA multivariate analysis based on Raman spectra allowed the mandibular cortical bone of the OVX group to be clearly distinguished from that of the SHAM group and consequently for the mandible of the OVX group to be diagnosed as “ovariectomized”, which correlated with bone fracture healing, prognosis of periodontitis, the success rate of oral implants, and other oral issues. In a further study, we will apply transcutaneous Raman spectroscopy to noninvasively detect various site-specific bones through the superficial skin with the aim of developing a Raman spectral database of OVX and SHAM groups. Moreover, we will supply the parameters of Raman spectroscopy within a particular range in the early diagnosis of osteoporosis by analyzing their correlation with other bone properties, such as BMD, mechanical properties, and clinical features.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: +86-591-8375488. Fax: +86-591-83700838. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Science and Technology Project of Fujian Province (2010H6009) and the Education and Technology Project of Fujian Province (JK2011019). We 3322
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
Analytical Chemistry
Article
(31) Schulmerich, M. V.; Cole, J. H.; Kreider, J. M.; Esmonde-White, F.; Dooley, K. A.; Goldstein, S. A.; Morris, M. D. Appl. Spectrosc. 2009, 63 (3), 286−295. (32) Ishihara, A.; Sasaki, T.; Debari, K.; Furuya, R.; Kawawa, T.; Ramamurthy, N. S.; Golub, L. M. J. Electron Microsc. (Tokyo) 1999, 48 (4), 465−469. (33) Van der Meulen, M. C.; Ashford, M. W. Jr.; Kiratli, B. J.; Bachrach, L. K.; Carter, D. R. J. Orthop. Res. 1996, 14, 22−29. (34) Elsubeihi, E. S.; Heersche, J. N. Eur. J. Prosthodont. Restor. Dent. 2009, 17 (1), 9−21. (35) Gallagher, J. C. Rheum. Dis. Clin. North Am. 2001, 27 (1), 143− 162. (36) Stepan, J. J.; Alenfeld, F.; Boivin, G.; Feyen, J. H.; Lakatos, P. Endocr. Regul. 2003, 37 (4), 225−238. (37) Ammann, P.; Rizzli, R. Osteoporosis Int. 2003, 14 (3), 13−18. (38) Caligiuri, P.; Giger, M. L.; Favus, M. Med. Phys. 1994, 21 (4), 503−508. (39) Hellekson, K. L. NIH Consens. Statement 2000, 17 (1), 1−45. (40) Lee, C.; Almagor, O.; Dunlop, D. D.; Manzi, S.; Spies, S.; Ramsey-Goldman, R. J. Rheumatol. 2007, 34 (10), 2018−2023. (41) Bolotin, H. H. Bone 2007, 41 (1), 138−154. (42) Gamsjaeger, S.; Masic, A.; Roschger, P.; Kazanci, M.; Dunlop, J. W.; Klaushofer, K.; Paschalis, E. P.; Fratzl, P. Bone 2010, 47 (2), 392− 399. (43) Renato, A. D. S.; Diego, P. J.; Hélio, A. G.; Murilo, X.; Marco, T. D. S.; Humberto, M. Open Bone J. 2010, 2, 24−31. (44) Akkus, O.; Polyakova-Akkus, A.; Adar, F.; Schaffler, M. B. J. Bone Miner. Res. 2003, 18 (6), 1012−1019. (45) Tarnowski, C. P.; Ignelzi, M. A. Jr.; Morris, M. D. J. Bone Miner. Res. 2002, 17 (6), 1118−1126.
thank the staff of the Key Laboratory of Optoelectronic Science and Technology for Medicine of Fujian Normal University for their excellent technical assistance. We also thank Dr. Zhufang Huang for conducting the multivariate analysis.
■
REFERENCES
(1) Banu, J.; Bhattacharya, A.; Rahman, M.; Kang, J. X.; Fernandes, G. J. Bone Miner. Metab. 2010, 28 (6), 617−626. (2) Pytlik, M.; Folwarczna, J.; Janiec, W. Calcif. Tissue Int. 2004, 75 (3), 225−230. (3) Fox, J.; Miller, M. A.; Recker, R. R.; Turner, C. H.; Smith, S. Y. Calcif. Tissue. Int. 2007, 81 (1), 53−63. (4) Wronski, T. J.; Pun, S.; Liang, H. Bone 1999, 25 (4), 465−468. (5) Rawlinson, S. C.; Mosley, J. R.; Suswillo, R. F.; Pitsillides, A. A.; Lanyon, L. E. J. Bone Miner. Res. 1995, 10 (8), 1225−1232. (6) Rawlinson, S. C.; Boyde, A.; Davis, G. R.; Howell, P. G.; Hughes, F. J.; Kingsmill, V. J. J. Dent. Res. 2009, 88 (7), 615−620. (7) Elsubeihi, E. S.; Heersche, J. N. Eur. J. Prosthodont. Restor. Dent. 2009, 17 (1), 9−21. (8) Tanaka, M.; Ejiri, S.; Toyooka, E.; Kohno, S.; Ozawa, H. J. Periodontal Res. 2002, 37 (2), 161−165. (9) Yang, J.; Farnell, D.; Devlin, H.; Horner, K.; Graham, J. J. Dent. 2005, 33 (2), 123−129. (10) Yang, J.; Pham, S. M.; Crabbe, D. L. Dentomaxillofac. Radiol. 2003, 32 (4), 247−251. (11) Ishihara, A.; Sasaki, T.; Debari, K.; Furuya, R.; Kawawa, T.; Ramamurthy, N. S.; Golub, L. M. J. Electron Microsc. (Tokyo) 1999, 48 (4), 465−469. (12) Miller, S. C.; Hunziker, J.; Mecham, M.; Wronski, T. J. J. Dent. Res. 1997, 76 (8), 1471−1476. (13) Judex, S.; Boyd, S.; Qin, Y. X.; Miller, L.; Muller, R.; Rubin, C. Curr. Osteoporosis Rep. 2003, 1 (1), 11−19. (14) Seeman, E.; Delmas, P. D. N. Engl. J. Med. 2006, 354 (21), 2250−2261. (15) Jacqueline, H.; Marjolein, C. H. Clin. Orthop. Relat. Res. 2011, 469, 2139−2149. (16) Goodyear, S. R.; Gibson, I. R.; Skakle, J. M.; Wells, R. P. Aspden RM. Bone 2009, 44 (5), 899−907. (17) Akkus, O.; Adar, F.; Schaffler, M. B. Bone 2004, 34 (3), 443− 453. (18) Van, A. A. A.; de Boer, J.; van, S. H.; Hoeijmakers, J. H.; Otto, C.; van, B. C. A. J. Gerontol. A Biol. Sci. Med. Sci. 2007, 62 (1), 34−40. (19) Shen, J.; Fan, L.; Yang, J.; Shen, A. G.; Hu, J. M. Osteoporosis Int. 2010, 21 (1), 81−87. (20) Du, Z.; Chen, J.; Yan, F.; Xiao, Y. Clin. Oral Implants Res. 2009, 20 (2), 145−150. (21) Feng, S.; Chen, R.; Lin, J.; Pan, J.; Chen, G.; Li, Y. Biosens. Bioelectron. 2010, 25 (11), 2414−2419. (22) Kozielski, M.; Buchwald, T.; Szybowicz, M.; Blaszczak, Z.; Piotrowski, A.; Ciesielczyk, B. J. Mater. Sci.: Mater. Med. 2011, 22 (7), 1653−1661. (23) Hequn, W.; Naiyan, H.; Jianhua, Z.; Harvey, L.; Mladen, K.; Haishan, Z. J. Raman Spectrosc. 2011, 42 (2), 160−166. (24) Kazanci, M.; Wagner, H. D.; Manjubala, N. I.; Gupta, H. S.; Paschalis, E.; Roschger, P.; Fratzl, P. Bone 2007, 41 (3), 456−461. (25) Cukrowski, I.; Popovic, L.; Barnard, W.; Paul, S. O.; van, R. P. H.; Liles, D. C. Bone 2007, 41 (4), 668−678. (26) Baur, H.; Leuenberger, C. Syst. Biol. 2011, 60 (6), 813−825. (27) Bergholt, M. S.; Zheng, W.; Lin, K.; Ho, K. Y.; Teh, M.; Yeoh, K. G.; So, J. B.; Huang, Z. Biosens Bioelectron. 2011, 26 (10), 4104− 4110. (28) Boskey, A. L. Curr. Osteoporosis Rep. 2006, 4 (2), 71−75. (29) Draper, E. R.; Morris, M. D.; Camacho, N. P. J. Bone Miner. Res. 2005, 20 (11), 1968−1972. (30) Schulmerich, M. V.; Dooley, K. A.; Morris, M. D.; Vanasse, T. M.; Goldstein, S. A. J. Biomed. Opt. 2006, 11 (6), 060502. 3323
dx.doi.org/10.1021/ac300046x | Anal. Chem. 2012, 84, 3318−3323
