Purchase this article with an account.
Hannah E Bowrey, E Ellen Jones, Mark Fields, Rosalie K Crouch, Lucian V Del Priore, Zsolt Ablonczy; Multimodal molecular imaging of Bruch’s membrane can predict age and macular degeneration. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):3992. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
During the course of aging the physiological functions of Bruch’s membrane (BM) gradually deteriorate, in part due to drusen accumulation between the retinal pigment epithelium (RPE) and BM. A profound accumulation of large soft drusen in the posterior pole is a key early clinical feature of age-related macular degeneration (AMD). Identification of the molecular composition of BM, including drusen deposits, is necessary for understanding AMD pathogenesis. Therefore, we sought to determine molecules in BMs that underlie its fluorescent lesions from a spectrum of ages, with- or without AMD.
Matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry (IMS) allows molecule-specific imaging of biological surfaces and can determine the spatial localization of thousands of molecules in a single experiment. We adapted an established protocol developed for imaging RPE by MALDI-IMS (Ablonczy et al. Proteomics 14: 936-944, 2014), used here on ex vivo human BMs (n = 15; mass range: m/z 200-1500; spatial resolution: 350 µm; positive mode). Individual autofluorescence images were co-registered, and correlated with each individual image in the MALDI imaging dataset from the same tissue. The total abundance of each molecular species was calculated and correlated with patient age. Finally, principal component analysis (PCA) was carried out to identify specific molecular patterns in BMs.
Within each patient sample, at least 500 molecules exhibited significant spatial correlation with autofluorescence images. In all, 311 molecules were shown to correlate with patient age. Of these, three molecular species - 268 m/z, 410 m/z and 426 m/z - increased with age (R = 0.77, 0.78 and 0.75, respectively; p < 0.05, all cases), whilst all others decreased with patient age. Bisretinoids were not found in high abundance and did not correlate with age. PCA analysis could easily distinguish BMs of different ages and disease states.
Imaging of BM by MALDI-IMS can be utilized as an important tool for the identification of molecular constituents involved in age- and disease-related changes of BM, as these methods can reliably predict the clinical diagnosis of the tissue. Our results show that age-related changes are associated with the loss of many small molecules. However, only three molecules exhibited an increase with age, suggesting their potential importance in the pathobiology of AMD.
This PDF is available to Subscribers Only