High-precision insights into bones, the brain and cell nuclei

Why do older people’s bones break so easily? Bone density is typically measured using a special X-ray scan to assess the risk of a fracture. “But that alone is a relatively poor indicator,” says Marianne Liebi, a researcher at the PSI Center for Photon Science. Bones don’t just become more porous with age. Other factors also play a role. And a method that Liebi’s team has refined for this purpose can now measure these changes in small individual samples. This fundamental research is giving the scientists insights into the structure of bones and how it changes with age. Their findings are helping to design better implants that stabilise bones optimally. 

Liebi’s research group uses the special X-ray light from the Swiss Light Source SLS, one of the five large research facilities at PSI. For more than ten years, she and her colleagues have been perfecting a method called small-angle X-ray scattering tensor tomography, or SAXS-TT for short. This allows the researchers to obtain high-resolution 3D images and, in particular, it reveals a factor that has a major influence on bone stability: the orientation of the collagen fibres that make up the bone material. “The fibres are a thousand times finer than a human hair and form an organic matrix, so to speak,” says Liebi. “In the femural neck, they are aligned more or less parallel to each other, to cushion the forces acting upon them. In some other areas of the bone, however, they criss-cross one another like the wood fibres in chipboard for maximum static stability.” 

While SAXS-TT doesn’t resolve these nanostructures individually, it provides information on their shape and orientation across different regions by analysing X-ray scattering. “This allows us to see the quality of the bones and how well they are fusing with a bone substitute,” says Liebi. In a project with the University of Bern, her team is investigating the circumstances associated with femoral neck fractures – a common problem among older people. Bone density is known to decrease with age, particularly on the upper side of the femoral neck. Now, the measurements taken by Liebi’s team show for the first time that the collagen fibres on this fracture-prone side are less ordered and that the bone platelets are a different shape. These tiny lamellae of calcium phosphate sit between the collagen fibres and stabilise them. 

“Our method also allows us to examine the orientation of the smallest structures in other parts of the body,” says Liebi. For example, PSI researchers have mapped the orientation of dentinal tubules in human teeth. “We have also used SAXS-TT to examine the myelin sheath that surrounds the nerve fibres of brain cells,” Liebi reports.

A circuit diagram of the brain 

When carrying out high-precision measurements of the brain, Liebi’s team sometimes collaborates with the group of PSI neurobiologist Adrian Wanner, who works at the PSI Center for Life Sciences. His aim is to decipher how the countless brain cells, known as neurons, are interconnected and how memories are retrieved. The ultimate goal is to create a detailed circuit diagram of the brain, not least to help pinpoint the exact causes of neurodegenerative diseases such as Alzheimer’s and Parkinson’s. 

“A single cubic millimetre of brain contains 100,000 neurons,” says Wanner. Each of these brain cells has thousands of connections with other neurons, meaning that this cubic millimetre contains several hundred million synapses and approximately four kilometres of nerve pathways. Wanner’s hypothesis is that, in order to understand the processes within this dense tangle, the synaptic connections and the organisation of neurons must be captured at high resolution within a large-scale context – ideally, the entire brain. 

The team is tackling this challenge from two angles. On the one hand, the researchers are examining dead brain tissue using an electron microscope – where they achieve a resolution of four nanometres in the tissue sections – as well as SAXS at the SLS. This allows them to obtain ultrahigh-resolution 3D images of the brain’s neural connections. On the other hand, they are observing live mice trained to navigate a virtual maze. Using a technique called calcium imaging, the researchers track which individual brain cells are activated when the mouse remembers the correct path. When the cells fire, the flow of calcium ions stimulates certain proteins to fluoresce. “So, as the mouse scurries through the corridors, its brain lights up here and there like a Christmas tree,” says Wanner. 

This activity varies slightly from mouse to mouse, of course. Nevertheless, certain neural pathways for specific tasks can be identified and thus applied to the human brain. “And that’s what we’re aiming for,” says Wanner. “We want to derive the principles by which a healthy brain operates – in humans as well as in mice.” 

Wanner’s team then repeats the experiments with genetically modified mice that have developed Alzheimer’s disease at various stages. The researchers then compare the patterns: what prevents the brain from retrieving memories? And to what extent is this related to the “plaques”, the deposits of harmful proteins in the brain to which Alzheimer’s is generally attributed? It might be possible to find an explanation as to why some people with Alzheimer’s are hardly affected despite having plaques, while others are affected long before the plaques appear. “Perhaps we’ll even discover entirely different factors – for example, that certain cell organelles are missing when synapses are defective,” says Wanner. “That could open up new avenues for treatment.”

Using the cheese-pasta principle to combat Alzheimer’s and Parkinson’s 

Meanwhile, a group led by biophysicist Jinghui Luo, who also works at the PSI Center for Life Sciences, is pursuing a different approach. The researchers want to better understand the disordered protein clumps associated with Alzheimer’s and Parkinson’s disease. 

Luo’s team has conducted biophysical in vitro and cell culture studies as well as experiments on the nematode Caenorhabditis elegans, which is frequently used as a model organism in ageing research. Spermine, a molecule naturally produced by our body, can protect neurons and mitigate age-related memory loss. Luo wanted to find out exactly why this happens. To this end, like Liebi and Wanner, he performed SAXS measurements at SLS. This allowed him to unravel spermine’s mechanism of action. In a degradation process called autophagy, our body is constantly working to break down harmful or unnecessary proteins, thus preventing their clumping and deposition, among other things. The molecule spermine supports this process in a way that is as simple as it is effective. “Spermine acts like cheese on spaghetti: it binds individual protein strands together to form a flexible network,” Luo explains. “It is easier for the body to break down this type of network than if the proteins form large, rigid clumps.”

Luo hopes this finding could lead to new drugs against Alzheimer’s and Parkinson’s. Perhaps even more, because spermine also has an impact on other diseases such as cancer. In addition to spermine, many other molecules from the class known as polyamines also perform important functions in the body and are therefore of medical interest. Research in this area has considerable untapped potential. “If we better understand the underlying processes,” says Luo, “we could find even better recipes for our cheese sauce, so to speak.” In this search, Luo’s team is also using a special artificial intelligence tool that can draw on all the available data to calculate promising combinations of “ingredients for the sauce” much more quickly. 

Detecting age-related diseases early on with AI 

At the PSI Center for Life Sciences, another expert is using artificial intelligence for biomolecular research. G. V. Shivashankar and his team have developed an AI-based method for the early detection of various age-related diseases. “Most organs and tissues deteriorate with age because cells divide less frequently and therefore don’t regenerate as effectively,” says Shivashankar. “It is crucial to detect this as early as possible so we can intervene and age more healthily.” 

His team focuses on blood cells: “In almost every disease, characteristic metabolic products or fragments of genetic material from the diseased tissue end up in the blood,” says Shivashankar. “The blood cells, as key components of the immune system, respond to this by initiating countermeasures.” The idea, therefore, is to create high-resolution images of a special molecular tangle inside the cell nuclei: chromatin. Chromatin is, so to speak, the packaged form of our genetic material, DNA. It changes every time the cell is activated, as the genes are unpacked to be read. Shivashankar’s team has determined that blood cells are activated differently in each disease, leading to disease-specific changes in the chromatin. 

“However, to detect these subtle changes, hundreds of characteristics have to be compared –such as shape, texture, and light spectrum,” Shivashankar acknowledges. To do this, Shivashankar’s team has developed an artificial intelligence tool in collaboration with researchers at the Massachusetts Institute of Technology in the US. “It is much faster and more reliable than any human at comparing such patterns.” In tests, the AI tool has already achieved an accuracy rate of more than 85 percent in distinguishing the chromatin of cancer patients from that of healthy individuals. “Through further improvements, we aim to increase this rate even further and achieve the accuracy required for approval by Swissmedic,” says Shivashankar. 

To get there, the researchers are first creating a database of healthy chromatin patterns – because their appearance is affected by age, sex and ethnic background. The goal is to collect more than ten thousand profiles. “First we need to define which chromatin patterns should be considered healthy, so that the system has a sound reference,” says Shivashankar. Once this atlas has been created, clinical trials could begin. 

One day, the early detection of a wide variety of diseases could become a straightforward routine test in clinics and doctor’s practices. “We have already developed a device that only needs a few drops of blood for the analysis and a microscope that simply connects to a smartphone. The detection program then runs on that,” reports Shivashankar. “The technology will therefore be very fast, inexpensive and easy to use – ideal also for regions with limited medical infrastructure.” 

Ultimately, it is about understanding diseases more precisely and detecting them earlier. “PSI is a special place where wide-ranging, high-quality research into healthy ageing converges,” says Shivashankar. And sometimes the path from fundamental research to practical applications can be surprisingly short: industry partners such as Roche are already interested in Shivashankar’s method.