Molecular mapping of periodontal tissues using infrared microspectroscopy
© Hynes et al; licensee BioMed Central Ltd. 2005
Received: 27 January 2005
Accepted: 12 May 2005
Published: 12 May 2005
Chronic periodontitis is an inflammatory disease of the supporting structures of the teeth. Infrared microspectroscopy has the potential to simultaneously monitor multiple disease markers, including cellular infiltration and collagen catabolism, and hence differentiate diseased and healthy tissues. Therefore, our aim was to establish an infrared microspectroscopy methodology with which to analyze and interpret molecular maps defining pathogenic processes in periodontal tissues.
Specific key cellular and connective tissue components were identified by infrared microspectroscopy and using a chemical imaging method.
Higher densities of DNA, total protein and lipid were revealed in epithelial tissue, compared to the lower percentage of these components in connective tissue. Collagen-specific tissue mapping by infrared microspectroscopy revealed much higher levels of collagen deposition in the connective tissues compared to that in the epithelium, as would be expected. Thus inflammatory events such as cellular infiltration and collagen deposition and catabolism can be identified by infrared microspectroscopy.
These results suggest that infrared microspectroscopy may represent a simple, reagent-free, multi-dimensional tool with which to examine periodontal disease etiology using entirely unprocessed tissue sections.
Periodontitis is defined by the inflammatory destruction of the supporting structures of teeth, including the periodontal ligament and alveolar bone. Periodontal diseases are generally chronic in nature and usually persist in the absence of treatment [1, 2]. These diseases are the result of exposure of the periodontium to dental plaque biofilms that accumulate on the teeth to form bacterial masses at or below the gingival margin .
IR microspectroscopy is a relatively new technique in which infrared spectra are observed through a microscope that transmits and detects infrared radiation. IR microspectroscopy has been previously utilized to monitor variations in the catabolism and anabolism of collagen and other components within cardiac tissues and oral cancer [5, 6]. Collagens exhibit a series of unique IR absorption bands between 1000 and 1300 cm-1. Specifically, the strong band at 1204 cm-1 has been identified as typical of collagen deposition [5, 6]. By integrating the intensity of this absorption band, one can readily plot molecular contour maps that clearly delineate areas of collagen deposition. Inflammation-driven collagen degradation is a hallmark of periodontitis [7, 8], with strong dissolution of collagen types I and III observed [9, 10].
Particular absorptions are assigned to various functional groups in an attempt to extract biochemical information. Absorptions between 1620–1680 cm-1 are usually attributed to amide I vibration of proteins, while absorptions at 1080 and 1240 cm-1 are attributed to PO2 - symmetric and asymmetric stretching vibrations of DNA phosphodiester groups . Using this data qualitative and semi-quantitative information can be extracted from the spectra.
Therefore, we set out to establish an IR microspectroscopic methodology that would allow the analysis and interpretation of molecular maps defining pathogenic processes in unprocessed periodontal tissues. This task is much more difficult in periodontal tissues in comparison to most other tissues due to the following factors; (1) Periodontal soft tissues are very thin (to a few mm); (2) Periodontal tissues excised at surgery are generally small (around 5 mm2); and (3) Excised periodontal tissues are particularly friable. Each of these factors present problems in embedding, sectioning, and orientating unfixed periodontal samples.
Patients referred to the Graduate Periodontics Clinic, University of Manitoba for periodontal therapy and giving written, informed consent were recruited. Inclusion criteria were a clinical diagnosis of chronic periodontitis and a requirement for periodontal surgery at a diseased site (probing pocket depth ≥ 5 mm, bleeding on probing, and clinical attachment loss ≥ 3 mm). Exclusion criteria were tobacco smoking, pregnancy, a requirement for antibiotic prophylaxis prior to periodontal probing, prolonged anti-inflammatory medications within the past 3-months (e.g. NSAIDs, steroids, antibiotics, or immunosuppressants), any systemic condition that may interfere with the study, such as inflammatory diseases, diabetes or blood dyscrasias, and lesions of the gingiva unrelated to plaque-induced periodontal disease. All subjects had received oral hygiene instruction and root planing prior to surgery.
Expired-air Carbon Monoxide measurement
Self-reported non-smoking status was validated at the time of recruitment by analysis of expired-air CO concentrations (PiCO meter, Bedfont Sci., UK and associated PiCO chart software), calibrated according to the manufacturers instructions. Non-smokers were required to exhibit expired-air CO concentrations <10 ppm.
Periodontal tissue collection and processing
On the day of surgery, prior to any anesthesia being administered, clinical measurements were obtained from the surgical site: probing depths, bleeding on probing and clinical attachment level (in that order). Twenty periodontal tissue samples were obtained from patients in good general health who were undergoing surgical treatment for chronic periodontitis. Mid-interproximal tissue was preferred, as these sites show increased clinical and histological signs of inflammation compared to other gingival sites . Periodontal tissue samples were snap frozen in isopentane supercooled in liquid nitrogen and then stored at -80°C.
A small amount of optimal cutting tool embedding media (Miles Inc., Elkhart, IN.) was applied to the tissue samples in order to facilitate attachment to the cryotome. A series of 10 μm sections were then cut, mounted onto an IR transparent barium fluoride window, and air-dried. Adjacent 10 μm samples were mounted and processed for conventional histology including visualization of tissue integrity and inflammatory foci by H & E staining.
Infrared microspectroscopy was performed using a Bruker FTIR spectrometer and IR microscope system equipped with a liquid nitrogen cooled mercury cadmium telluride detector. For IR spectral acquisition, the microscope aperture was closed to allow the IR beam to illuminate an area of tissue measuring 50 μm × 50 μm, thereby masking all other regions of the tissue section. For each 50 μm × 50 μm demarcation, 64 interferograms were collected. The signal was averaged against a blank area as background and Fourier-transformed to generate IR spectra with a nominal resolution of 4 cm-1. After each acquisition, the stage was stepped 50 μm under computer control and the next spectrum acquired. This process was repeated until the complete area of interest was mapped. General biochemical mappings regarding the specific components of proteins, lipid, and DNA were obtained using a chemical imaging method. Specifically, the bands used for protein mapping were the amide I band (1654 cm-1); lipid mapping used the lipid CH stretching vibrations at 2800–3000 cm-1; and DNA mapping employed the PO2 - symmetric stretching vibration of DNA phosphodiester groups at 1080 cm-1. To highlight collagen deposition, a unique collagen IR band at 1204 cm-1 was used to generate tissue maps, as we have described previously in cardiac tissues . A digital CCD video camera was coordinated to the IR microscope to record each mapped area for future band and area correlation. Microphotography of the identical mapping position in the adjacent H and E-stained tissue section allowed the comparison of general histological features with the IR mapping data.
Data processing for IR microspectroscopy
Because most IR bands are broad and are composed of overlapping components, it is necessary to pre-process the original spectra by applying a band-narrowing algorithm that separates the individual bands. All data processing was performed using the Cytospec V software package http://www.cytospec.com. To permit a useful comparison of the cluster analysis, uniformly pre-treated data was used. All the original IR spectra were converted into second derivative spectra using the Savitzky/Golay algorithm with a 9-point window for the multivariate statistical analysis. Derivative spectra were scaled before the cluster analysis where the sum squared deviation over the indicated wavelengths equals unity (vector normalization) . The unsupervised cluster analysis used Ward's minimum variance algorithm and Euclidean distances as distance measure. The Euclidean distance between spectra is calculated and the pair of spectra with the least distance is grouped to create a cluster. Then the separation between this cluster and all other spectra is calculated and another two closest spectra/clusters are joined to form a new cluster. This procedure continues until all spectra/clusters are combined. In this unsupervised classification no information about the disease state of the samples is needed, only the similarity or dissimilarity of their infrared spectra is used for this classification.
In the final step of cluster analysis, all spectra in the same family are assigned a color code, and small, colored squares are drawn at the pixel coordinates of all spectra belonging to the same family to produce our false color maps. As shown in Figure 7B, the blue color (line II) indicates that the spectra from this group were in the same family, originating from the epithelium, like the IR protein mapping of the same tissue shown in Figure 7C. The yellow color in Figure 7B (line I) represents those similar spectra generated from connective tissue. Panel A in Figure 7 displays the average spectrum produced by the cluster analysis with color codes. In other words, the average spectrum from epithelium and connective tissues are in blue and yellow, marked II and I, respectively, and also correspond to the spectra analyzed in full detail in Figure 4.
IR microspectroscopy possesses numerous advantages over traditional approaches to pathology. Fixation and staining of tissues are not required before histological viewing, little or no sample preparation is necessary and only minimal technical expertise is required by the operator. The method lends itself readily to rapid, high-volume repetitive measurements. IR microspectroscopy is non-destructive, meaning that the sample may be saved and passed on for further measurements if required. Furthermore, IR microspectroscopy provides information concerning the molecular structure of the tissue and multiple analytes may be measured simultaneously from a single spectrum. This combination of features is simply unavailable from visual microscopy.
Our initial data suggest that infrared microspectroscopy represents a suitable tool with which to simultaneously monitor multiple disease markers in periodontal biopsies, including cellular infiltration, collagen catabolism, and other differences in the molecular profile of diseased tissues. We have established a methodology by which IR microspectroscopy is capable of revealing several major biochemical components and specific features, including collagen content, in the studied tissue using "digital staining", without the need for any chemical reagents or probes. IR maps of inflammation-driven collagen degradation in periodontal tissue sections can therefore be constructed and analyzed. The promising preliminary results, obtained in establishing this IR microspectroscopic methodology, suggest a potential role for infrared microspectroscopy in understanding the inflammatory processes underlying the progression of periodontitis.
The major problem in adapting IR technology to the study of inflammatory processes in gingival tissues was the small size and fragility of the periodontal biopsies themselves. However, we are able to process these small tissue samples and can clearly differentiate predominantly cellular and predominantly acellular areas of tissue, and visualize areas of collagen deposition and degradation. The next steps will be to compare IR microscopic maps of diseased and healthy tissues; to correlate NIR-based definitions of pathological tissue changes with disease severity; and to correlate IR maps with classical clinical signs of periodontal diseases, such as edema, gingival bleeding, and periodontal pocket depths on probing, and clinical attachment loss.
As inflammatory cell infiltration and subsequent collagen degradation are hallmarks of periodontitis [7–10], and because IR spectral analyses can determine such tissue events, in addition to multiple other tissue changes at the molecular and sub-molecular level, then this IR microspectroscopic methodology can now be applied to hypothesis-driven research that aims to identify disease-related pathological changes in periodontal tissues.
List of abbreviations
This study was supported by funding from the National Research Council, Canada and by the Health Sciences Research Foundation, Canada.
- Page RC, Shroeder HE: Pathogenesis of inflammatory periodontal disease. A summary of current work. Lab Invest. 1976, 34: 235-265.PubMedGoogle Scholar
- Page RC: The role of inflammatory mediators in the pathogenesis of periodontal disease. J Periodont Res. 1991, 26: 230-242.View ArticlePubMedGoogle Scholar
- Loesche WJ: The antimicrobial treatment of periodontal disease: changing the treatment paradigm. Crit Rev Oral Biol Med. 1999, 10: 245-75.View ArticlePubMedGoogle Scholar
- Jackson M, Sowa MG, Mantsch HH: Infrared spectroscopy: a new frontier in medicine. Biophys Chem. 1997, 68: 109-25. 10.1016/S0301-4622(97)80555-8.View ArticlePubMedGoogle Scholar
- Liu KZ, Dixon IM, Mantsch HH: Distribution of collagen deposition in cardiomyopathic hamster hearts determined by in microscopy. Cardiovasc Pathol. 1999, 8: 41-47. 10.1016/S1054-8807(98)00024-6.View ArticlePubMedGoogle Scholar
- Schultz CP, Liu KZ, Kerr PD, Mantsch HH: In-situ infrared histopathology of keratinization in human oral/or pharyngeal aqueous cell carcinoma. Oncol Res. 1998, 10: 277-286.PubMedGoogle Scholar
- Ejeil AL, Gaultier F, Igondjo-Tchen S, Senni K, Pellat B, Godeau G, Gogly B: Are cytokines linked to collagen breakdown during periodontal disease progression?. J Periodontol. 2003, 74: 196-201. 10.1902/jop.2003.74.2.196.View ArticlePubMedGoogle Scholar
- van der Zee E, Everts V, Beertsen W: Cytokines modulate routes of collagen breakdown. J Clin Periodontol. 1997, 24: 297-305.View ArticlePubMedGoogle Scholar
- Hillmann G, Krause S, Ozdemir A, Dogan S, Geurtsen W: Immunohistological and morphometric analysis of inflammatory cells in rapidly progressive periodontitis and adult periodontitis. Clin Oral lnvestig. 2001, 5: 227-35. 10.1007/s00784-001-0134-3.View ArticleGoogle Scholar
- Feldner BD, Reinhardt RA, Garbin CP, Seymour GJ, Casey JH: Histological evaluation of interleukin-1 beta and collagen in gingival tissue from untreated adult periodontitis. J Periodontal Res. 1994, 29: 54-61.View ArticlePubMedGoogle Scholar
- Abrams K, Caton J, Polson AM: Histologic comparison of interproximal gingival tissues related to the presence or absence of bleeding. J Periodontol. 1984, 55: 629-632.View ArticlePubMedGoogle Scholar
- Lasch P, Haensch W, Naumann D, Diem M: Imaging of colorectal adenocarcinoma using FT-IR microspectroscopy and cluster analysis. Biochim Biophys Acta. 2004, 176-186.Google Scholar
- Liu KZ, Jia J, Newland AC, Kelsey SM, Mantsch HH: Quantitative determination of apoptosis on leukemic cells by infrared spectroscopy. Apoptosis. 2001, 6: 269-278. 10.1023/A:1011383408381.View ArticlePubMedGoogle Scholar
- Diem D, Chiriboga L, Yee H: Infrared spectroscopy of human cells and tissue. VIII. Strategies for analysis of infrared tissue mapping data and applications to liver tissue. Biopolymers. 2000, 57: 82-90. 10.1002/1097-0282(2000)57:5<282::AID-BIP50>3.0.CO;2-R.View ArticleGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2342/5/2/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.