We are developing new approaches to quantitative, label-free histological examination of tissues by infrared (IR) micro-spectroscopy. In this technique, an infrared spectrometer with a 2D detector array is attached to a microscope. It simultaneously measures IR absorption spectra at 16,384 micron-size spots in a tissue section. Chemical composition, orientation and interactions of chemical groups are determined within each spot from unique spectral fingerprints of chemical compounds and plotted as 2D-images. To date, applications of this technique to research and diagnostics have been limited to dehydrated tissues because water strongly absorbs IR light, resulting in optical interference artifacts. However, dehydration distorts biomolecular and tissue structure, smears out spectroscopic fingerprints, and degrades chemical and spectral resolution. To overcome this limitation, we designed and constructed an IR chamber with thermo-mechanical stabilization, which allows keeping tissues in solution at desired temperature. By reducing the interference artifacts, we increased spectral reproducibility and chemical resolution by two orders of magnitude compared to commercial designs. Versatile solvent control and increased spectral accuracy of the new chamber allow qualitatively new experimental approaches. For example, with this technique we distinguish collagen from other proteins, resolve different glycosaminoglycans (GAG) and even quantify the extent of GAG sulfation in cartilage. We collected a spectral library of well-purified and characterized components of connective tissues, which we measured with significantly improved spectro-chemical resolution. We also developed a new approach to quantitative mapping of collagen orientation in tissue sections by polarized IR hyperspectral imaging. To correlate tissue composition with biochemical processes, we bridged this high-definition (HD) microspectroscopic imaging with analytical autoradiographic imaging with micrometer spatial resolution. We also designed a new thermo-mechanically stabilized, flow-through chamber for HD Raman microspectroscopy, allowing a simultaneous additional characterization of bone specimens with Raman, polarized and fluorescence microscopies. We adapted in vivo dynamic labeling of bone formation surfaces with fluorescence dyes which allows to demarcate bone regions formed at given time points and to perform HD Raman microspectroscopy on the same samples. We use these techniques to characterize collagen matrix organization in osteo- and chondro-dysplasias, bone tumors and other connective tissues pathologies. Specifically, we studied a knock-in mouse model of Diastrophic Dysplasia (DTD) caused by mutations in SLC26A2 sulfate/chloride antiporter. These mutations result in deficient sulfate uptake by chondrocytes, leading to undersulfation of proteoglycan GAG chains crucial for cartilage development and integrity. DTD has delayed skeletal development and exhibits an unusual progression: The undersulfation is normalized with age, but the articular cartilage degrades with age. To understand underlying mechanism, we collected 6-micron-resolution, quantitative images of distributions of major extra-cellular matrix components across femur head cartilage and growth plate in newborn mice. We showed that in DTD mice, GAG sulfation was low compared to wild type (WT) in the articular and prehypertrophic zones but almost normal in the resting zone. In DTD mice, polarized IR hyperspectral imaging revealed disruption of a dense layer of tangentially oriented collagen fibrils at the articular surface. Disruption of this protective layer may cause articular proteoglycan depletion, a hallmark of early osteoarthritis, which we observed at birth and which further progresses with age despite the normalization of GAG sulfation. Collagen orientation in DTD mice was also disrupted throughout the femur head and growth plate. The disruption severity correlated with GAG undersulfation, suggesting that GAG sulfation may be crucial for deposition of the oriented matrix. Using quantitative microradiography of 35S-sulfate incorporation into cartilage explants, we showed that variability of undersulfation across different cartilage regions in DTD was associated with faster chondroitin synthesis rate in the articular and prehypertrophic zones. This observation explained the undersulfation normalization with age. We assisted NIBIB scientists using our technology to demonstrate penetration of functionalized carbon nanotubes inside cancer cells that overexpress hylauronate receptors, validating this approach to intracellular delivery of anticancer agents. We investigated effects of stem cell transplantation on bone quality in the Brittle mouse model of Osteogenesis Imperfecta (OI). Using fluorescence and polarized-light microscopy, we distinguished matrix produced by host cells and green-fluorescent-protein-labeled donor cells within different types of bone (woven and lamellar) in femoral cortex. Using Raman microspectroscopy, we found that matrix mineralization heterogeneity near donor cells was lower within each material type, suggesting that better organization of matrix made by the donor cells may contribute to the increased bone strength in the treated mice. We studied endocrine bone tumors caused by cyclic AMP signaling disruption in mice with different combinations of deletions of Prkar1a+/- with Prkaca+/-, Prkar2a+/-, or Prkar2b+/- in protein kinase A. Using polarized-light microscopy, Raman microspectroscopy, and dynamic bone labeling, we found that tumor formation in adult Prkar1a+/-Prkaca+/- mice causes periosteal deposition of immature cortical bone, in which collagen and mineral organizations are intermediate between those of woven and lamellar bone. We found partial compensation of Prkar1a+/- deletion effects on local maturation of bone material, matrix mineralization and collagen organization by additional Prkar2a+/- or Prkar2b+/- deletions. We showed that osteoblast differentiation and tumor growth in R1a+/-Ca+/- mice was affected by Celecoxib treatment and that Celecoxib slowed growth of tail vertebrae tumors and improved organization and mineralization of cortical bones covering the tumors. We developed non-destructive quantification of collagen content in cell cultures using HD Raman microspectroscopy. We applied this method to human fibroblast cultures from recessive osteogenesis imperfecta caused by FKBP10-null mutations. We found reduced deposition of collagen in the extracellular matrix despite synthesis of normal collagen quantities by these cells and by ppib-/- mouse osteoblasts. However, the deposition did not correlate with the disease severity and phenotype for other FKBP10 mutations causing Kuskokwim syndrome, suggesting complexity of FKBP10 functions. We used a novel G610C mouse model of OI occurring in Old Order Amish community. We showed that unexpectedly high brittleness of G610C mouse femurs is partly due to abolished anterior-posterior drift preventing replacement of woven bone laid at young age by lamellar bone. Furthermore, mature lamellar bone, whose hypermineralization contributes to bone fragility in human OI, was hypermineralized in G610C mice as well. We developed and used an HDIR method to show that cultured G610C osteoblasts deposited thinner and less dense collagen matrix, explaining hyperminerlaization and brittleness of OI bone material. Our treatment with low-protein diet activating autophagy, rescued the hypermineralization defect, pointing at autophagy as OI treatment target. During the last year, we initiated development of HDIR and Raman analyses to improve understanding of placenta function and pathology. Initial measurements resulted in surprising findings such as previously unreported accumulation of cholesteryl ester aggregates.
Showing the most recent 10 out of 13 publications
