As my soccer coach liked to say, “the harder you work, the luckier you get”. I think the pursuit of scientific discovery requires quite a lot of hard work, persistence and, in some cases, a little bit of luck. The development of the imaging method described in our paper just published in Nature Biomedical Engineering involved all of these factors together with the collaborative efforts of everyone involved. The impetus of this paper started with studying aortic heart-valve tissue in the context of calcific aortic-valve disease (CAVD). CAVD affects primarily older patients or younger individuals with a genetic determinants that turn their aortic valve from a normal tri-leaflet to an abnormal bi-leaflet anatomy. In patients with CAVD, the soft valve tissue stiffens to the point of hindering the nominal opening and closing of the valve. In some cases, the valve tissue can develop bone-like nodules, further impeding blood flow and potentially requiring the only viable treatment option to date: full valve replacement surgery.
In an effort to better understand the changes in the extracellular matrix (ECM) proteins, as a way to more fully describe what is happening during CAVD progression, our labs started a collaboration to look at changes in the collagen structure. For these analyses, we decided to use a well-established, nonlinear microscopy imaging method capturing the second harmonic generation (SHG) signal of the collagen fibers. During these imaging sessions, the microscope collected signals in several different channels at different wavelengths. While collecting these multi-channel images, I noticed what I thought to be random background noise in the 525nm emission channel. After consultation with Prof Georgakoudi and my co-authors, we decided that the ‘random background’ signal that kept showing up in that extra channel wasn’t just a coincidence and warranted further investigation, particularly given that the group had observed similar strong natural fluorescence signals in engineered bone tissues. Upon further testing, we determined that indeed the probable source of this fluorescence was calcification in the valve tissue.
Following a spectral analysis of different tissue types and a calcific model system, all with differing levels of calcification, we developed a simple approach using two imaging filters to separate the mineralization signal from that of collagen and to quantify the level of mineralization in several different tissue samples. This system, which embodies many of the advantages of using nonlinear microscopy, is label-free and non-destructive to the imaged samples. We were able to use our method to image a calcifying cell culture system to track nodule development over time and saw that calcification occurs at varying rates and to different degrees, sometimes disappearing in spots even as the overall calcification increases.
We believe that this technique has a wide range of applications for studying calcification in disease states as well as for the visualization of healthy mineralization, such as in bone and tooth tissue engineering. As optical probe technology improves, this system could be used in vivo to aid with CAVD diagnosis, in determining the stability of atherosclerotic plaques, or in determining breast cancer prognosis, providing high resolution images, within the context of the natural tissue environment. The mineralized fluorescence could also be paired with signal from other endogenous fluorophores, such as collagen, lipids, or elastin to better understand the presence of calcification in the context of other ECM molecules. With this manuscript and the development of our new imaging methodology, we hope to share our luck and help advance the understanding of diseases like CAVD.
Our paper: Baugh, L. M. et al. Non-destructive two-photon excited fluorescence imaging identifies early nodules in calcific aortic-valve disease, Nat. Biomed. Eng. (2017) doi:10.1038/s41551-017-0152-3.