March 2012
Volume 53, Issue 3
Free
Biochemistry and Molecular Biology  |   March 2012
Alterations in the Biomolecular Signatures of Developing Chick Corneas as Determined by Biospectroscopy and Multivariate Analysis
Author Affiliations & Notes
  • Weiyi Pang
    From the The School of Public Health, Guangxi Medical University, Nanning, Guangxi, People's Republic of China;
    Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, United Kingdom;
  • Abdullah A. Ahmadzai
    Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, United Kingdom;
  • Imran I. Patel
    Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, United Kingdom;
  • Xiaoqiang Qiu
    From the The School of Public Health, Guangxi Medical University, Nanning, Guangxi, People's Republic of China;
    The School of Public Health, Guilin Medical University, Guilin, Guangxi, People's Republic of China; and
  • Melody Liles
    The Structural Biophysics Group, School of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales, United Kingdom.
  • Andrew J. Quantock
    The Structural Biophysics Group, School of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales, United Kingdom.
  • Francis L. Martin
    Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, United Kingdom;
  • *Each of the following is a corresponding author: Xiaoqiang Qiu, The School of Public Health, Guangxi Medical University, Nanning, Guangxi, PR China; and The School of Public Health, Guilin Medical University, Guilin, Guangxi, PR China; xqqiu9999@sina.com. Francis L. Martin, Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Bailrigg, Lancaster LA1 4YQ, UK; f.martin@lancaster.ac.uk
Investigative Ophthalmology & Visual Science March 2012, Vol.53, 1162-1168. doi:10.1167/iovs.11-9262
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Weiyi Pang, Abdullah A. Ahmadzai, Imran I. Patel, Xiaoqiang Qiu, Melody Liles, Andrew J. Quantock, Francis L. Martin; Alterations in the Biomolecular Signatures of Developing Chick Corneas as Determined by Biospectroscopy and Multivariate Analysis. Invest. Ophthalmol. Vis. Sci. 2012;53(3):1162-1168. doi: 10.1167/iovs.11-9262.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: Biospectroscopy tools are increasingly being recognized as novel approaches toward interrogating complex biological structures in a nondestructive fashion. This study was conducted to apply these tools to interrogate alterations in the molecular signatures of developing chick corneas during the onset and development of transparency.

Methods.: Embryonic chick corneas (n = 46) were obtained at 2-day intervals from embryonic day (E)10 to E18 of incubation and investigated with attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy and Raman microspectroscopy. Resultant spectra were analyzed for variance by using principal component analysis and linear discriminant analysis (PCA-LDA).

Results.: Mean spectra after ATR-FTIR spectroscopy or Raman microspectroscopy derived from corneas at each developmental stage showed some overlap; however, in PCA-LDA scores plots, a clear segregation of spectra was evident, and two-category discrimination indicated that significant molecular alterations occur during tissue morphogenesis. Notable by both techniques was the increasing intensity of DNA signal (1080 cm−1) from E10 onward. Major segregating biomarkers identified by ATR-FTIR spectroscopy between E10 and E18 were in the DNA/RNA (1126 cm−1), glycogen (1045 cm−1), protein (1470 cm−1), and amide II (1512 cm−1 and 1524 cm−1) spectral regions. Raman spectroscopy also identified major distinguishing vibrational modes that included proteins, amino acids (tyrosine, proline phenylalanine, and valine), and secondary structures of proteins (amide I and amide II).

Conclusions.: The developing chick cornea undergoes significant changes in its biomolecular composition in the E10 to E18 developmental period, with the major changes occurring in the spectral regions associated with DNA/RNA, proteins, glycogen, and secondary protein structures.

The chicken embryo cornea, which develops into a fully functioning transparent tissue in the week leading up to hatching at day 21 of incubation, is an excellent model for the study of matrix morphogenesis and corneal structure–function relationships. 1,2 Pioneering work documented fundamental anatomic and biochemical aspects of chick corneal development. 3 10 Later investigators added more detail about the roles of various collagen types that are influential in corneal development 11 17 ; the mechanisms by which cells deposit an organized fibrillar matrix 18 24 ; the tissue's changing gene expression profile 25 ; and the compositional alterations in proteoglycans, their sulfation status, and their relationship with collagen. 26 33 To further clarify the developmental events in chick corneal morphogenesis, we undertook a series of investigations using infrared (IR) spectroscopy or Raman microspectroscopy to elucidate the biomolecular changes in the cornea as it grows and becomes transparent between embryonic day (E)10 and E18. 
Attenuated total reflection Fourier-transform IR (ATR-FTIR) spectroscopy is based on the concept that biomolecules absorb in the mid-IR region (λ= 2.5–25 μm) through the vibration of chemical bonds. The vibration is detected, and derived from this is a detailed biomolecular fingerprint of the cells or tissue being interrogated in the form of an IR spectrum that relates to chemical structure. 34,35 Raman microspectroscopic techniques, on the other hand, work on the principle that chemical bonds are excited to a virtual state through excitation of photons that subsequently relax to a different vibrational state. This relaxation causes inelastic photon-scattering and a frequency shift, 36 and acquired spectra provide data with a high spatial resolution and a wide vibrational profile, with minimal impact of aqueous. 34,37 ATR-FTIR spectroscopy and Raman microspectroscopy, although fundamentally different, provide complementary information about the vibrational modes of molecules within tissues and have recently been used for the characterization of cell cycle events, disease diagnosis, and cancer diagnosis. 34 38 ATR-FTIR spectroscopy and Raman microspectroscopy experiments generate complex biochemical datasets containing hundreds and thousands of spectra, and analysis is best achieved by multivariate approaches such as principal component analysis (PCA) and/or linear discriminant analysis (LDA). This allows for data reduction to facilitate the identification of wavenumber-related biomarkers, such as glycogen content, conformational protein changes and phosphorylation, and structural alterations in DNA/RNA. 39 In the present study, we used ATR-FTIR spectroscopy and Raman microspectroscopy to obtain biomolecular signatures from developing chick corneas in the E10 to E18 developmental period when the tissue is undergoing embryonic growth and transition into a mature transparent tissue. 
Materials and Methods
Tissue Source
Fertilized white leghorn chicken eggs were obtained from a commercial hatchery (Henry Steward and Co., Lincolnshire, UK), left to settle for approximately 24 hours, and placed in an incubator at 37.5°C (Octagon-100; Brinsea Products Ltd., Sandford, UK). After the required period (i.e., at embryonic day [E]10, E12, E14, E16, and E18), the corneas (n = 48) were excised at the limbus and dried under light compression between paraffin film–covered glass microscope slides. All work was conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and all corneas were interrogated by both spectroscopic approaches. 
ATR-FTIR Spectroscopy and Raman Microspectroscopy
Dried and flattened corneas were placed on 1-cm2 low-E-reflective glass slides (Kevley Technologies, Chesterland, OH) for analysis. IR spectra for ATR-FTIR spectroscopy were obtained with an FTIR spectrometer with an ATR attachment containing (Helios) a diamond crystal (Vector 22; all from Bruker Optics Ltd., Coventry, UK). Spectra through a ≈250 μm2 aperture ATR crystal were acquired from 50 different locations across each cornea (n = 46). A new background was taken after the acquisition of every 10 spectra. 
After ATR-FTIR spectroscopic analysis, Raman spectra were acquired from each cornea with a Raman spectrometer (InVia; Renishaw PLC, Gloucestershire, UK) containing a 785-nm laser diode operating at 100-mW and a Rayleigh holographic edge filter. Corneas were placed on gold slides (which have no background fluorescent signature) for data collection. The spectrometer's entrance slit was 50-μm and a 1200-lines/mm (1-cm−1 spatial resolution) diffraction grating dispersed Raman signals onto a Pelletier cooled CCD (Master; Renishaw). A white light camera mounted on the microscope was used to obtain dark-field images to visualize locations for spectral acquisition. Spectra were acquired with an epidetecting ×50 objective lens (numerical aperture 0.75, ∼1 μm2 spatial resolution), 10 seconds of 100% laser power exposure, and three repeat acquisitions. 
Computational Analysis
PCA allows for the reduction of the number of variables in the spectral dataset, whose principal components (PCs) can capture more than 95% of the variance present in the original dataset. 40 LDA is a data separation technique that maximizes interclass variance. PCA can be applied before LDA (thus, PCA-LDA) to reduce computational complexity and increase the recognition accuracy in different classes. 40,41 Initially, using OPUS software, ATR-FTIR spectra were individually cut to the biochemical cell fingerprint region (1800–900 cm−1). Subsequently, spectra were baseline corrected and normalized to amide I (1650 cm−1). 
Raw Raman spectra were subtracted for cosmic rays or atypical peaks using the spectrometer system software (Wire 3.1; Renishaw). In particular, some residual paraffin peaks 42 caused by the specimen environment were evident; thus, the following Raman spectral regions were removed before analysis: 1470 to 1405, 1305 to 1285, 1145 to 1115, 1070 to 1050, and 760 to 690 cm−1. Data from every three spectra were then averaged and cut between the 1750 to 500-cm−1 (1336 data points) spectral range. Subtraction of background fluorescence was achieved by baseline correction using the polynomial technique followed by vector normalization (MatLab; The MathWorks, Natick, MA). Multivariate analysis (i.e., PCA-LDA) was applied to the spectrometric data with the same software (R2009a, MatLab; The MathWorks) and a graphic user interface toolkit for spectroscopy (http://biophotonics.lancs.ac.uk/software). 
Results
All corneas were interrogated by ATR-FTIR spectroscopy followed by Raman microspectroscopy, and typical representative spectra are shown (Fig 1A, ATR-FTIR spectroscopy; Fig. 1B, Raman microspectroscopy). Initial PCA on the average ATR-FTIR spectra on each embryonic day (E10 [n = 6], E12 [n = 9], E14 [n = 9], E16 [n = 10], and E18 [n = 12]) indicated a degree of overlap between the developmental days, but with a notable difference in the increasing of intensity of DNA (1080 cm−1; symmetric phosphate, νasPo 2 ) from E10 onward (data not shown). The average Raman spectra (E10 [n = 4], E12 [n = 5], E14 [n = 9], E16 [n = 10], and E18 [n = 11]) confirmed the overlap observed after ATR-FTIR spectroscopy, and again highlighted the between-class discrimination noted at 1080 cm−1
Figure 1.
 
Average spectra of all spectral acquisitions from developing chick corneas at different embryonic days (E10 [red], E12 [blue], E14 [green], E16 [magenta], and E18 [yellow]) after (A) ATR-FTIR spectroscopy and (B) Raman microspectroscopy.
Figure 1.
 
Average spectra of all spectral acquisitions from developing chick corneas at different embryonic days (E10 [red], E12 [blue], E14 [green], E16 [magenta], and E18 [yellow]) after (A) ATR-FTIR spectroscopy and (B) Raman microspectroscopy.
Owing to the overlap of data and subtlety of the spectral variations between developmental days, exploratory multivariate analysis (i.e., PCA-LDA) was performed. In this more powerful discriminatory approach, each spectrum becomes a point at which the best segregation and associated wavenumber are identified. By this analysis, a clear cluster segregation of the chick corneas at different developmental days was observed, as seen in 2-D and 3-D PCA-LDA scores plots (Figs. 2A, 2C). In particular, the PCA-LDA data indicated that there was less overlap of spectral points between developmental stages with greater intervals (e.g., E10 vs. E18). Consistent with the ATR-FTIR spectroscopy observations, 2-D and 3-D PCA-LDA score plots of the Raman data reveal clustering and segregation of data points from corneas at different developmental days (Figs. 2B, 2D), although these clusters were less tight than those derived after ATR-FTIR spectroscopy, especially at E10 and E14. The discriminating wavenumbers responsible for the segregation ATR-FTIR and Raman spectra with progressive corneal development as revealed by PCA-LDA are shown in the derived cluster vector plots in Figure 3, and the main biomarkers in each category are listed in Table 1. These cluster vector plots indicate the discriminating variables in the corresponding scores plots proportional to peak intensity. 
Figure 2.
 
2-D or 3-D PCA-LDA scores plots performed on the dataset from the spectra derived from chick corneas at different days. (A) 2-D PCA-LDA scores plot after ATR-FTIR spectroscopy; (B) 2-D PCA-LDA scores plot after Raman microspectroscopy; (C) 3-D PCA-LDA scores plot after ATR-FTIR spectroscopy; and (D) 3-D Raman scores plot after Raman microspectroscopy. The different classes were labeled as shown in the figure.
Figure 2.
 
2-D or 3-D PCA-LDA scores plots performed on the dataset from the spectra derived from chick corneas at different days. (A) 2-D PCA-LDA scores plot after ATR-FTIR spectroscopy; (B) 2-D PCA-LDA scores plot after Raman microspectroscopy; (C) 3-D PCA-LDA scores plot after ATR-FTIR spectroscopy; and (D) 3-D Raman scores plot after Raman microspectroscopy. The different classes were labeled as shown in the figure.
Figure 3.
 
PCA-LDA cluster vector plots derived from infrared spectral datasets: (A) ATR-FTIR spectral dataset; and (B) Raman spectral dataset. The different categories are shown in the figure.
Figure 3.
 
PCA-LDA cluster vector plots derived from infrared spectral datasets: (A) ATR-FTIR spectral dataset; and (B) Raman spectral dataset. The different categories are shown in the figure.
Table 1.
 
Top Five Ranked Tentative Peaks of Every Category Observed in the Cluster Vector Plots
Table 1.
 
Top Five Ranked Tentative Peaks of Every Category Observed in the Cluster Vector Plots
Embryonic Day Discriminating Wavenumber (cm−1)
ATR-FTIR Spectroscopy Raman Microspectroscopy
10 1038, 1126, 1470, 1620, 1524 853, 1035, 1679, 1684, 1675
12 1686, 1701, 1717, 1736, 1539 814, 919, 938, 1084, 1658
14 1015, 1701, 1593, 1616, 1736 814, 939, 919, 1084, 1635
16 1126, 1045, 1462, 1528, 1416 919, 938, 1250, 814, 1028
18 1701, 1736, 1531, 1601, 1123 856, 814, 1685, 1635, 919
Two-category discriminant analysis of ATR-FTIR spectra from chick corneas at different stages of embryonic development was used to determine the wavenumbers responsible for between-category spectral variance. PCA-LDA scores plots show an obvious discrimination between E10 versus E18 with little overlap (Fig. 4A). Wavenumbers that contributed to the E10 versus E18 segregation were 1126 (DNA/RNA), 1045 (glycogen), 1470 (proteins), 1524 (amide II), and 1512 (amide II) cm−1 (Fig 4B). The E12 versus E18 comparison reveals clustering and segregation of data, with distinguishing wavenumbers observed at 1543 (amide II), 1732 (lipid), 1713 (lipid), 1609 (amide I), and 1119 (DNA/RNA) cm−1 (Figs. 4C, 4D). A relatively good segregation was notable between E14 and E18, with the main distinguishing wavenumbers being 1524 (amide II), 1543 (amide II), 1709 (amide I), 1126 (DNA/RNA), and 1053 (glycogen) cm−1 (Figs. 4E, 4F). Owing to progressive corneal development, it is not surprising that the segregation between E16 and E18 was not as marked as that in the other between-category comparisons. Nevertheless, some segregation was found, and the wavenumbers most responsible were 1466 (proteins), 1512 (amide II), 1524 (amide II), 1732 (lipid), and 1612 (amide I) cm−1 (Figs. 4G, 4H). 
Figure 4.
 
Two-category discriminant analysis using PCA-LDA after ATR-FTIR spectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Figure 4.
 
Two-category discriminant analysis using PCA-LDA after ATR-FTIR spectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Scores plots of the Raman microspectroscopy data when investigated by PCA-LDA confirmed between-category discrimination based on developmental day and revealed that, consistent with progressive developmental change, the segregation was more notable the larger the interval between developmental time points. No overlap was apparent in the E10 versus E18 comparison in which the following wavenumbers were responsible for segregation: 1395 (proteins), 1384 (proteins), 855 (tyrosine and proline), 1673 (amide I), and 1679 (amide I) cm−1 (Figs. 5A, 5B). There was also a clear discrimination between E12 and E18, with 855 (tyrosine and proline), 1681 (amide I), 919 (proline and valine), 935 (proline and valine), and 1656 (amide I) cm−1 contributing to the segregation (Figs. 5C, 5D). The degree of segregation was less evident when the E14 and E18 spectral points were compared, for which the main distinguishing wavenumbers were found to be 856 (tyrosine and proline), 937 (proline and valine), 814 (proteins), 1635 (amide I), and 1002 (phenylalanine) cm−1 (Figs. 5E, 5F). Unlike the clear segregation seen in comparisons between E18 and E10/E12, notable overlap was seen for the E16 versus E18 between-category comparison, with the wavenumbers 1002 (phenylalanine), 1685 (amide I), 1688 (amide I), 1035 (phenylalanine), and 1657 (amide I) cm−1 responsible for the segregation that was observed (Figs. 5G, 5H). 
Figure 5.
 
Two-category discriminant analysis using PCA-LDA after Raman microspectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Figure 5.
 
Two-category discriminant analysis using PCA-LDA after Raman microspectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Discussion
In this study, ATR-FTIR spectroscopy and Raman microspectroscopy with multivariate analysis were used to characterize the changing biomolecular fingerprint 43 45 of the embryonic chick cornea as it develops from E10 to E18 and becomes transparent. A major finding of both techniques was an enhanced peak at 1080 cm−1, specific to E10. This wavenumber corresponds to νasPo 2 , suggesting a major conformational difference in the genomic material at E10 compared with E12, E14, E16, and E18. Although the biological significance of this difference remains to be fully determined, it has been suggested that it may be associated with an altered methylation state. 46,47 Previous measurements of the relative levels of mRNA in the developing chick cornea have shown an increase in the message for the keratan sulfate proteoglycan core proteins lumican, keratocan, and mimecan (osteoglycin) between E9 and E12, with a steady decrease thereafter. 25 It is unlikely, therefore, that changes in keratan sulfate proteoglycan core proteins underlie the change in DNA conformation subsequent to E10, with the alterations in the genetic mechanisms underlying the biosynthesis of collagen, the major component of the cornea, likely to be important. In relation to the changes in DNA/RNA after E10, it is also noteworthy that recent investigations of developing chick corneas at E14 have indicated that the relative cellular volume of the stroma ascertained by volume scanning electron microscopy is in the region of 25%, 24 higher than previous estimates. Parallel analyses of chick corneas at E10 and E18 have provided a similar overall volume estimate for the keratocytes in the stromal matrix (Young RD, and Quantock AJ, unpublished data, 2011). although the cells tend to have flatter profiles. It is likely, therefore, based on the relatively high signal from νasPo 2 at E10, that a greater proportion of the keratocyte cell body is occupied by the nucleus at this stage of embryogenesis compared with later developmental stages. As stated, it could be that methylation of the genome plays a significant role, with epigenetic mechanisms allowing for subsequent development stages. 46  
More subtle changes in the complex datasets of ATR-FTIR and Raman spectra may be extracted only by a multivariate approach, such as PCA-LDA. This highlighted good between-category segregation (i.e., developmental days) in scores plots. Moreover, scores plots derived from two-category discrimination between E18 and earlier developmental days revealed that comparisons over larger intervals result in less overlap. Nearness in multivariate distance implies pattern recognition, whereas the separation of sample clusters in the plots signifies structurally dissimilar groups. Both ATR-FTIR spectroscopy and Raman microspectroscopy disclosed the greatest degree of separation between E10 and E18 and the most similarity at E16 and E18. Thus, biospectroscopy can be indicative of alterations in the biomolecular signatures of developing chick corneas that are progressive and cumulative. Biospectroscopy tools appear to hold great promise toward lending new insights into molecular questions. 47  
As the secondary chick cornea grows in ovo from E10 to E18, the presumptive keratocytes synthesize and deposit new and progressively more well-organized extracellular matrix. ATR-FTIR spectroscopy reveals how biomarkers at E10 are clearly different from those at later developmental days and, as well as changes in DNA/RNA previously discussed, these can be attributed, predominantly, to changes in glycogen, protein, and secondary structures of protein (amide I and II), consistent with new matrix deposition. Raman spectroscopy also indicates that the major distinguishing vibrational modes that differ between E10 and later stages include proteins, constituent amino acids (tyrosine, proline phenylalanine, valine), and secondary structures of proteins (amide I and II). Collagen is undoubtedly the major component of the newly deposited corneal matrix, and, as comprehensively reviewed by Linsenmayer et al., 1 a multitude of alterations in the production and assembly of the various fibril- and non–fibril-forming collagens, their subtypes and potential interactions, occur throughout the development of the chick stroma. 10 19 Previous studies of corneal morphogenesis have indicated a progressive doubling in the amount of hydroxyproline in the embryonic corneas between E12 and E14, between E14 and E16, and between E16 and E18, 33 suggestive of a fairly linear rate of growth and collagen deposition over these 2-day intervals (E10 corneas were not included in the previous analysis). The new spectroscopy data presented herein point to a post-E10 accelerated matrix deposition, with a presumed steady rate of synthesis thereafter. Collagen is the most abundant protein of animal origin and FTIR spectroscopy has been applied to investigate its molecular structure 48 ; however, it was not possible in this study to assign specific spectral regions such as the amide I peak solely to collagen structure, given that entire corneas were analyzed. Future studies could shed new insights of collagen-specific alterations in corneal development by observing analysis-isolated constituents. 
Within reason, one would surmise that these bioanalytical methods are relatively nondestructive in comparison with other conventional methodologies. ATR-FTIR spectroscopy requires the diamond to come into contact with the biological material, 35,40,41 so some pressure effects will be exerted on the corneal samples. However, the sample is still retained if, for instance, one wishes to extract DNA/RNA from it or stain it. The 785-nm laser diode appears to be minimally destructive of biological samples. 49,50 A promising observation was that these two biospectroscopy methodologies exhibited a good correlation between their respective findings. 
In summary, this study demonstrates that ATR-FTIR spectroscopy and Raman microspectroscopy can be used to reveal significant spectral discrimination and changing biomolecular signatures in the developing chick cornea and that, by the use of multivariate analysis, the biomarkers responsible for the segregation of corneas in the E10 to E18 developmental period can be discriminated. 
Footnotes
 Supported by Project Grant EP/F034970 from the U.K. Engineering and Physical Sciences Research Council (AJQ) and the China Scholarship Council (WP).
Footnotes
 Disclosure: W. Pang, None; A.A. Ahmadzai, None; I.I. Patel, None; X. Qiu, None; M. Liles, None; A.J. Quantock, None; F.L. Martin, None
References
Linsenmayer TF Fitch JM Gordon MK . Development and roles of collagenous matrices in the embryonic avian cornea. Prog Ret Eye Res. 1998;17:231–265. [CrossRef]
Quantock AJ Young RD . Development of the corneal stroma, and the collage-proteoglycan associations that help define its structure and function. Dev Dyn. 2008;237:2607–2621. [CrossRef] [PubMed]
Coulombre AJ Coulombre JL . The development of the structural and optical properties of the cornea. In: Smelser GK , ed. The Structure of the Eye. New York, Academic Press; 1961:405–420.
Anseth A . Glycosaminoglycans in the developing corneal stroma. Exp Eye Res. 1961;1:116–121. [CrossRef] [PubMed]
Hay ED Revel JP . Fine structure of the developing avian cornea. In Wolsky A Chen PS , eds. Monographs in Developmental Biology. Vol. 1. Basel: S. Karger; 1969:1–144.
Conrad GW . Collagen and mucopolysaccharide biosynthesis in the developing chick cornea. Dev Biol. 1970;21:292–317. [CrossRef] [PubMed]
Toole BP Trelstad RL . Hyaluronate production and removal during corneal development in chick. Dev Biol. 1971;26:28–35. [CrossRef] [PubMed]
Trelstad RL Coulombre AJ . Morphogenesis of the collagenous stroma in the chick cornea. J Cell Biol. 1971;50:840–858. [CrossRef] [PubMed]
Hart GW . Biosynthesis of glycosaminoglycans during corneal development. J Biol Chem. 1976;251:6513–6521. [PubMed]
Linsenmayer TF Smith GN Hay ED . Synthesis of 2 collagen types by embryonic chick corneal epithelium in vitro. Proc Natl Acad Sci USA. 1977;74:39–43. [CrossRef] [PubMed]
Linsenmayer TF Fitch JM Mayne R . Extracellular matrices in the developing avian eye: type V collagen in corneal and noncorneal tissues. Invest Ophthalmol Vis Sci. 1984;25:41–47. [PubMed]
Linsenmayer TF Bruns RR Mentzer A Mayne R . Type VI collagen immunohistochemical identification as a filamentous component of the extracellular matrix of the developing avian corneal stroma. Dev Biol. 1986;118:425–431. [CrossRef] [PubMed]
Hayashi M Ninomiya Y Hayashi K Linsenmayer TF Olsen BR Trelstad RL . Secretion of collagen types I and II by epithelial and endothelial cells in the developing chick cornea demonstrated by in situ hybridization and immunohistochemistry. Development 1988;103:27–36. [PubMed]
Svoboda KK Nishimura I Sugrue SP Ninomiya Y Olsen BR . Embryonic chicken cornea and cartilage synthesize type IX collagen molecules with different amino-terminal domains. Proc Natl Acad Sci USA. 1988;85:7496–7500. [CrossRef] [PubMed]
Fitch JM Birk DE Linsenmayer C Linsenmayer TF . Stromal assemblies containing collagen types IV and VI and fibronectin in the developing embryonic avian cornea. Dev Biol. 1991;144:379–391. [CrossRef] [PubMed]
Gordon MK Foley JW Linsenmayer TF Fitch JM . Temporal expression of types XII and XIV collagen mRNA and protein during avian corneal development. Dev Dyn. 1996;206:49–58. [CrossRef] [PubMed]
Young BB Zhang G Koch M Birk DE . The roles of types XII and XIV collagen in fibrillogenesis and matrix assembly in the developing cornea. J Cell Biochem. 2002;87:208–220. [CrossRef] [PubMed]
Birk DE Trelstad RL . Extracellular compartments in matrix morphogenesis: collagen fibril, bundle, and lamellar formation by corneal fibroblasts. J Cell Biol. 1984;99:2024–2033. [CrossRef] [PubMed]
Fitch JM Linsenmayer CM Linsenmayer TF . Collagen fibril assembly in the developing avian primary corneal stroma. Invest Ophthalmol Vis Sci. 1994;35:862–869. [PubMed]
Quantock AJ Kinoshita S Capel MS Schanzlin DJ . A synchrotron x-ray diffraction study of developing chick corneas. Biophys J. 1998;74:995–998. [CrossRef] [PubMed]
Siegler V Quantock AJ . Two-stage compaction of the secondary avian cornea during development. Exp Eye Res. 2002;74:427–431. [CrossRef] [PubMed]
Connon CJ Meek KM Kinoshita S Quantock AJ . Spatial and temporal alterations in the collagen fibrillar array during the onset of transparency in the avian cornea. Exp Eye Res. 2004;78:909–915. [CrossRef] [PubMed]
Gealy EC Hayes AJ Buckwell B . Actin and procollagen distribution in the developing chick cornea: association with oriented matrix deposition. Invest Ophthalmol Vis Sci. 2009;50:1653–1658. [CrossRef] [PubMed]
Bushby AJ P'ng KMY Young RD Pinali C Knupp C Quantock AJ . Imaging three-dimensional tissue architectures by focused ion beam scanning electron microscopy. Nat Protoc. 2011;6:845–858. [CrossRef] [PubMed]
Dunlevy JR Beales MP Berryhill BL Cornuet PK Hassell JR . Expression of the keratan sulfate proteoglycans lumican, keratocan and osteoglycin/mimecan during chick corneal development. Exp Eye Res. 2000;70:349–362. [CrossRef] [PubMed]
Funderburgh JL Caterson B Conrad GW . Keratan sulfate proteoglycan during embryonic development of the chicken cornea. Dev Biol. 1986;116:267–277. [CrossRef] [PubMed]
Midura RJ Hascall VC . Analysis of the proteoglycans synthesized by corneal explants from embryonic chicken, II: structural characterization of the keratan sulfate and dermatan sulfate proteoglycans from corneal stroma. J Biol Chem. 1989;264:1423–1430. [PubMed]
Cornuet PK Blochberger TC Hassell JR . Molecular polymorphism of lumican during corneal development. Invest Ophthalmol Vis Sci. 1994;35:870–877. [PubMed]
Hirsch M Noske W Prenant G Renard G . Fine structure of the developing avian corneal stroma as revealed by quick-freeze, deep-etch electron microscopy. Exp Eye Res. 1999;69:267–277. [CrossRef] [PubMed]
Zhang Y Conrad AH Tasheva ES . Detection and quantification of sulfated disaccharides from keratan sulfate and chondroitin/dermatan sulfate during chick corneal development by ESI-MS/MS. Invest Ophthalmol Vis Sci. 2005;46:1604–1614. [CrossRef] [PubMed]
Young RD Gealy EC Liles M Caterson B Ralphs JR Quantock AJ . Keratan sulfate glycosaminoglycan and the association with collagen fibrils in rudimentary lamellae in the developing avian cornea. Invest Ophthalmol Vis Sci. 2007;48:3083–3088. [CrossRef] [PubMed]
Gealy C Kerr B Young RD . Differential expression of the keratan sulphate proteoglycan, keratocan, during chick corneal embryogenesis. Histochem Cell Biol. 2007;128:551–555. [CrossRef] [PubMed]
Liles M Palka BP Harris A . Differential relative sulfation of keratan sulfate glycosaminoglycan in the chick cornea during embryonic development. Invest Ophthalmol Vis Sci. 2010;51:1365–1372. [CrossRef] [PubMed]
Petrich W . Mid-infrared and Raman spectroscopy for medical diagnostics. Appl Spectrosc Rev. 2001;36:181–237. [CrossRef]
Walsh MJ German MJ Singh M . IR microspectroscopy: potential applications in cervical cancer screening. Cancer Lett. 2007;246:1–11. [CrossRef] [PubMed]
Lin-Vien D Colthup NB Fateley WG Graselli JG . The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules. New York: Academic Press; 1991:1–7.
Stone N Kendall C Smith J Crow P Barr H . Raman spectroscopy for identification of epithelial cancers. Faraday Discuss. 2004;126:141–157. [CrossRef] [PubMed]
Patel II Martin FL . Discrimination of zone-specific spectral signatures in normal human prostate using Raman spectroscopy. Analyst. 2010;135:3060–3069. [CrossRef] [PubMed]
Martin FL German MJ Wit E Fearn T Ragavan N Pollock HM . Identifying variables responsible for clustering in discriminant analysis of data from infrared microspectroscopy of a biological sample. J Comput Biol. 2007;14:1176–1184. [CrossRef] [PubMed]
Llabjani V Trevisan J Jones KC Shore RF Martin FL . Binary mixture effects by PBDE congeners (47, 153, 183, or 209) and PCB congeners (126 or 153) in MCF-7 cells: biochemical alterations assessed by IR spectroscopy and multivariate analysis. Environ Sci Technol. 2010;44:3992–3998. [CrossRef] [PubMed]
Llabjani V Jones KC Thomas GO Walker LA Shore RF Martin FL . Polybrominated diphenyl ether-associated alterations in cell biochemistry as determined by attenuated total reflection Fourier-transform infrared spectroscopy: a comparison with DNA-Reactive and/or endocrine-disrupting agents. Environ Sci Technol. 2009;43:3356–3364. [CrossRef] [PubMed]
Faoláin EO Hunter MB Byrne JM . Raman spectroscopic evaluation of efficacy of current paraffin wax section dewaxing agents. J Histochem Cytochem. 2005;53:121–129. [CrossRef] [PubMed]
Trevisan J Angelov PP Patel II . Syrian hamster embryo (SHE) assay (pH 6.7) coupled with infrared spectroscopy and chemometrics towards toxicologic assessment. Analyst. 2010;135:3266–3272. [CrossRef] [PubMed]
Bonifacio A Beleites C Vittur F . Chemical imaging of articular cartilage sections with Raman mapping, employing uni- and multi-variate methods for data analysis. Analyst. 2010;135:3193–3204. [CrossRef] [PubMed]
Movasaghi Z Rehman S Rehman IU . Raman spectroscopy of biological tissues. Appl Spectrosc Rev. 2007;42:493–541. [CrossRef]
Walsh MJ Fellous TG Hammiche A . Fourier transform infrared microspectroscopy identifies symmetric PO2 modifications as a marker of the putative stem cell region of human intestinal crypts. Stem Cells. 2008;26:108–118. [CrossRef] [PubMed]
Martin FL Kelly JG Llabjani V . Distinguishing cell types or populations based on the computational analysis of their infrared spectra. Nat Protoc. 2010;5:1748–1760. [CrossRef] [PubMed]
Muyonga JH Cole CGB Duodu KG . Characterisation of acid soluble collagen from skins of young and adult Nile perch (Lates niloticus). Food Chem. 2004;85:81–89. [CrossRef]
Hodges MD Kelly JG Bentley AJ . Combining immunolabeling and surface-enhanced Raman spectroscopy on cell membranes. ACS Nano. 2011;5:9535–9541. [CrossRef] [PubMed]
Patel II Trevisan J Singh PB . Segregation of human prostate tissues classified high-risk (UK) versus low-risk (India) for adenocarcinoma using Fourier-transform infrared or Raman microspectroscopy coupled with discriminant analysis. Anal Bioanal Chem. 2011;401:969–982. [CrossRef] [PubMed]
Figure 1.
 
Average spectra of all spectral acquisitions from developing chick corneas at different embryonic days (E10 [red], E12 [blue], E14 [green], E16 [magenta], and E18 [yellow]) after (A) ATR-FTIR spectroscopy and (B) Raman microspectroscopy.
Figure 1.
 
Average spectra of all spectral acquisitions from developing chick corneas at different embryonic days (E10 [red], E12 [blue], E14 [green], E16 [magenta], and E18 [yellow]) after (A) ATR-FTIR spectroscopy and (B) Raman microspectroscopy.
Figure 2.
 
2-D or 3-D PCA-LDA scores plots performed on the dataset from the spectra derived from chick corneas at different days. (A) 2-D PCA-LDA scores plot after ATR-FTIR spectroscopy; (B) 2-D PCA-LDA scores plot after Raman microspectroscopy; (C) 3-D PCA-LDA scores plot after ATR-FTIR spectroscopy; and (D) 3-D Raman scores plot after Raman microspectroscopy. The different classes were labeled as shown in the figure.
Figure 2.
 
2-D or 3-D PCA-LDA scores plots performed on the dataset from the spectra derived from chick corneas at different days. (A) 2-D PCA-LDA scores plot after ATR-FTIR spectroscopy; (B) 2-D PCA-LDA scores plot after Raman microspectroscopy; (C) 3-D PCA-LDA scores plot after ATR-FTIR spectroscopy; and (D) 3-D Raman scores plot after Raman microspectroscopy. The different classes were labeled as shown in the figure.
Figure 3.
 
PCA-LDA cluster vector plots derived from infrared spectral datasets: (A) ATR-FTIR spectral dataset; and (B) Raman spectral dataset. The different categories are shown in the figure.
Figure 3.
 
PCA-LDA cluster vector plots derived from infrared spectral datasets: (A) ATR-FTIR spectral dataset; and (B) Raman spectral dataset. The different categories are shown in the figure.
Figure 4.
 
Two-category discriminant analysis using PCA-LDA after ATR-FTIR spectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Figure 4.
 
Two-category discriminant analysis using PCA-LDA after ATR-FTIR spectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Figure 5.
 
Two-category discriminant analysis using PCA-LDA after Raman microspectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Figure 5.
 
Two-category discriminant analysis using PCA-LDA after Raman microspectroscopy. (A) 1-D scores plot (E10 vs. E18); (B) cluster vector plot (E10 vs. E18); (C) 1-D scores plot (E12 vs. E18); (D) cluster vector plot (E12 vs. E18); (E) 1-D scores plot (E14 vs. E18); (F) cluster vector plot (E14 vs. E18); (G) 1-D scores plot (E16 vs. E18); and (H) cluster vector plot (E16 vs. E18).
Table 1.
 
Top Five Ranked Tentative Peaks of Every Category Observed in the Cluster Vector Plots
Table 1.
 
Top Five Ranked Tentative Peaks of Every Category Observed in the Cluster Vector Plots
Embryonic Day Discriminating Wavenumber (cm−1)
ATR-FTIR Spectroscopy Raman Microspectroscopy
10 1038, 1126, 1470, 1620, 1524 853, 1035, 1679, 1684, 1675
12 1686, 1701, 1717, 1736, 1539 814, 919, 938, 1084, 1658
14 1015, 1701, 1593, 1616, 1736 814, 939, 919, 1084, 1635
16 1126, 1045, 1462, 1528, 1416 919, 938, 1250, 814, 1028
18 1701, 1736, 1531, 1601, 1123 856, 814, 1685, 1635, 919
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×