Edit

Duke Team

Biological X-ray imaging

Our research programme focuses on new approaches in X-ray imaging of biological samples and encompasses experimental, technical and computational developments.

Edit

We will develop High-Throughput Tomography (HiTT), extending the pioneering work of Thomas Schneider and his team on beamline P14 at PETRA III in Hamburg. We will develop a correlative workflow combining our HiTT data with data acquired from the same sample using other state-of-the-art imaging modalities such as light microscopy and electron microscopy.

X-ray imaging has a notable application in the area of neuroscience and we will work with those in the field to embed HiTT into established neuroscience workflows. Extending HiTT to include imaging of other bulk soft tissues samples will open up new areas of biology as it will be possible to view internal structures in samples on a scale not previously seen.

Previous and current research

My primary area of expertise is in the development of synchrotron X-ray methods with a particular focus on their application to life science research.

After many years of developing synchrotron instrumentation and techniques for macromolecular crystallography, I moved into the area of X-ray imaging. I developed the technique of cryo soft X-ray tomography which is a powerful technique to image the internal structures of intact frozen hydrated cells. A particular interest of mine is correlative imaging. While X-ray imaging in its own right has clear benefits, answering many biological questions, additional information is often required to understand the functional significance of specific internal structures within cells. In these situations correlative imaging is the right approach. This means visualising exactly the same specimen or region with two or more complementary imaging modalities such that, when the results are combined or overlaid, the information generated is greater than the sum of its parts. Correlative imaging is by necessity collaborative, in the sense that it requires cooperation between specialists in different techniques – no single imaging modality can probe the inner workings of a section of tissue or a cell whilst simultaneously visualising the entire organ or even organism. To date, most work in this area has been dominated by the approach of Correlative Light and Electron Microscopy (CLEM). However, significant progress has also been made in Correlative Light and X-ray Microscopy (CLXM) in which cryo soft X-ray microscopy has been used to image the internal structures of intact frozen hydrated cells.

Over the past few years, advances have been made in establishing hard X-ray imaging of soft biological tissue. The use of hard X-rays allows data to be collected from intact bulk samples such as brain tissue which – when combined with advanced fluorescence techniques – creates amazing opportunities to correlate in vivo with ex vivo imaging. This enables scientists to explore structure and function and simultaneously to bridge length scales allowing a multitude of questions to be addressed using the same sample.

Future projects and goals

Our goal is to establish biological X-ray imaging as a routine method for the collection of 3D data on biological materials ranging from cells to organoids, organs and organisms using existing facilities at EMBL and elsewhere.

In particular, HiTT will open up X-ray imaging to the vast array of bulk soft tissue samples being studied by biologists around the world such as mouse brains, mouse retinas, kidneys, vascular networks, cancerous tumours and so on. Delivering HiTT on P14 at EMBL Hamburg will revolutionise biology in ways that we can currently only dream of.

In parallel with establishing data collection techniques we will develop the data analysis pipeline, ensuring that it is as streamlined and as easy to use as possible. Imaging produces a vast amount of data so the infrastructure to allow it to be handled painlessly is vital alongside the provision of user-friendly interfaces for image acquisition and powerful reconstruction algorithms.

Edit