C.J. Kok Public Award
The Faculty of Science grants two C.J. Kok awards each year: the C.J. Kok Public Award, also known as the award for the ‘Discoverer of the Year’, and the C.J. Kok Jury Award, the award for the best PhD thesis from the past year. All institutes within the faculty are given the opportunity to nominate candidates for both awards.
Mapping the invisible universe
It is a task of truly mind boggling proportions: creating a map of the threedimensional structure of the entire universe. Not content with just that Herculean effort, the members of the Kilo-Degree Survey (KiDS) team managed to up the ante. They didn’t just map the structure of the universe we can see, but also managed to chart that which we can’t.
Of course, mapping something that’s truly invisible is impossible. The KiDS team charted the structure of dark matter: invisible stuff that does not interact with regular matter. It does, however, warp space-time just like regular matter, deforming it in a way that’s visible by how light changes its path while moving. This so-called weak lensing effect revealed the distribution of regular and dark matter in greater detail than ever before.
‘We found a difference in the statistical properties of our distribution and previous measurements’, says team member Henk Hoekstra. ‘Although our result isn’t statistically significant, it’s certainly interesting.’
The publication about this discovery was one of the three most cited in astronomy in 2017. ‘The next data set covers twice the sky surface, and should halve our uncertainty’, adds Koen Kuijken, overall leader of the KiDS team. If that data increases the significance, the team might be onto something big. ‘The chance we haven’t found anything is only a couple percent. It’s exciting.’
‘Mapping the dark matter in the universe with gravitational lensing has been talked about since the 1990’s’, says Kuijken, ‘but it is only now that we have finally developed the instruments and tools to actually do it.’
The technique is mathematically comparable to looking at coins in a bath tub. ‘When the water’s still, the coins look circular’, Hoekstra says. But when the water’s moving, the images deform. The same happens with galaxies. Deformations caused by (dark) matter changes their shapes in subtle ways that are difficult to spot in a single galaxy, but emerge when looking at many of them. ‘We average our results over tens of millions of galaxies’, says Hoekstra.
Their work leads to interesting possibilities. ‘We’re able to test some alternative theories of gravity’, says Hoekstra. And future surveys, like the Euclid ESA mission, in which both Kuijken and Hoekstra play a prominent role, will lead to even more detailed results.
Astronomers might even get to tackle the most mysterious substance in our universe: the ubiquitous dark energy, which pushes the universe to expand. While some think dark energy is the result of Einstein’s so-called cosmological constant, others disagree. ‘If we find the cosmological constant isn’t actually constant, that’ll be a big deal’, says Hoekstra. We’ll need to find new physics to explain how the universe actually works.’
The KiDS team started their research in 2012 and is an ever evolving, international organism; a collaboration between different universities in different countries. The team is led by principle investigator Koen Kuijken from Leiden Observatory.
Koen Kuijken (Kapellen, Belgium, 1963) has been working as a professor of Astrophysics at Leiden University since 2002. As principal investigator of the KiDS survey, he was responsible for everything from observations and scientific analysis to funding. He also headed the international consortium that built the 300 million pixel camera with which the observations was done. ‘The 15 year, end-to-end involvement with this was very satisfying. It was great to finally see the initial plans come to fruition,’ he says. ‘And I loved working with this great, dedicated team of young people.’
Henk Hoekstra (Heerenveen, The Netherlands, 1973) has been at Leiden University since 2008 and became a full professor in 2017. He was a coordinating member of the KiDS team and contributed to the image simulations that were used to test the algorithm that measured the lensing of galaxies. ‘We are pushing the boundaries of this field’, he says, ‘discovering subtle effects as we go along. This means that the research is varied and has a nice, smallish international team with many young researchers.’
Massimo Viola (Trieste, Italy, 1983) studied physics and astrophysics in Italy and came to Leiden as a research assistant from 2012 to 2017. He coordinated the weak lensing team of the KiDS collaboration and co-led the main cosmological analysis. ‘I loved working with a friendly and committed group of top-class scientists to improve our knowledge about the origin and evolution of our universe’, he says. ‘It was truly an exciting journey of discovery, done with enthusiastic colleagues, many of whom I now call friends.’
Arun Kannawadi (Chennai, India, 1990) finished his PhD at Carnegie Mellon University in the United States in 2016, before coming to Leiden as a postdoc. ‘I have dwelved into quantum information, biophysics and theoretical physics before I finally picked cosmology as my field of interest’, he says. For the KiDS survey he worked on so-called shear calibrations, which help make the galaxy image simulations more realistic. ‘I liked that the team was small, which brings the members closer and avoids overhead’, he says. ‘This group brings together experts I already knew through my PhD work, but I did not have a chance to meet before coming to Leiden.’
Maciej Bilicki (Łódź, Poland, 1978) studied mathematics, computer science and astronomy in Poland before coming to Leiden as a postdoc in 2015. He estimated the distances to galaxies without directly measuring their spectra, by analysing their properties with machine learning. ‘I liked being able to contribute to publicly released catalogues and working with state-of-the-art data of unprecedented quality’, he says.
Andrej Dvornik (Maribor, Slovenia, 1990) did his masters in Leiden and is currently working on his PhD. He’s looking into the gravitational lensing effects of individual galaxies, galaxy groups and clusters to study their properties and helped to reduce the enormous amount of data the KiDS study gathered. ‘I liked working with this lively and super motivated team, with a presence from Canada to Australia’, he says, ‘while at the same time learning new things about dark energy and dark matter and getting to know efficient methods to deal with large data sets.’
Christos Georgiou (Ioannina, Greece, 1992) graduated in theoretical physics at the university of Edinburgh before starting his PhD in Leiden in 2015. He studies the way galaxies tend to align with each other, and how that alignment can be used to make gravitational lensing measurements more accurate. ‘I really enjoyed the friendly atmosphere in this team’, he says. ‘Everyone was always willing to help, which made our work fluid and enjoyable.’
Maria Christina Fortuna (Assisi, Italy, 1991) did a bachelor in physics and a master in astrophysics before beginning her PhD in Leiden in 2016, where she’s developing a model for the description of the intrinsic alignment of galaxies, which means she got to work with the KiDS team. ‘I like being on the edge between modelling and data analysis’, she says. ‘I am very fascinated by the connection between galaxies and dark matter, as well as the large-scale structure of the universe.’
Margot Brouwer (Amsterdam, The Netherlands, 1989) started her PhD at Leiden Observatory in 2013. She studied how to use weak gravitational lensing to learn more about the dark matter distribution around galaxies, clusters and larger cosmic structures. ‘Our whole team is working together to create state-of-the-art weak gravitational lensing observations’, she says. ‘Having these unique data at my disposal allowed me to do ground-breaking research, such as studying the cosmic dark matter web and performing the first test of Erik Verlinde's new theory of emergent gravity.’
Ricardo Herbonnet (Rotterdam, The Netherlands, 1988) spent 11 years at Leiden university, starting with studying for his bachelors in 2006 and ending with his PhD in 2017. His most important contribution to the KiDS team was determining the accuracy of the shape measurements of the galaxies in the dataset. ‘I really loved working with people from many different countries’, he says. ‘Our team is small, which means that there’s a lot of work for everyone. But it also means you really get to know the team members. That led to a great atmosphere.’
Screaming and dancing balls lead to new physics discovery
Viral YouTube videos are not where you’d typically expect new physics to pop up. But when physicist Scott Waitukaitis saw one in which an aspiring Ukrainian rapper dropped hydrogel balls on a hot plate, he knew something was up. His careful analysis revealed a new physical effect and opened the door to possible applications in soft robotics.
1.2 million. That’s how many people have currently viewed ‘Hydrogel Beads in a Frying Pan’, the YouTube video by aspiring rapper Ivan Panchenko. The video’s popularity is easily explained: when the balls hit the hot surface, something surprising happens. They keep on bouncing, while screaming in high tones. It’s an outcome that tickles the viewer and appears to be magic, but isn’t. It’s physics. Physics that Scott Waitukaitis has since discovered.
‘We discovered a new version of the Leidenfrost effect’, Waitukaitis says. An effect which can be seen when water drops dance on a hot place, floating on a cushion of their own vapour.
In hydrogel balls – soft solids comprised of mostly water – something similar happens. When they hit the surface, vapour escapes and pushes them up, adding a spring to their bounce. But it also causes the surface to vibrate, creating pressure waves in the same way the cone of a speaker does. In short: the balls don’t just jump, they scream.
Because the gels are somewhat fragile, they can’t keep bouncing and screaming forever. After about three to four minutes, the balls break. ‘If they were tougher, they could keep going longer’, Waitukaitis says. ‘They could theoretically go on for as long as thirty minutes.’
Waitukaitis is a self-confessed lover of the obscure physics behind everyday occurrences. And while he was mostly interested in understanding a mechanism, he quickly realised that there might be more to this elastic Leidenfrost effect, as he has christened it, than just some fun science.
‘There’s this field of research called soft robotics’, he says – robots made not of metal, but of softer materials. ‘People are struggling with ways to make these robots move.’ A current solution for this problem is to use inflatable balloons in cavities. ‘But that’s a pretty lousy way to make things move’, Waitukaitis says. ‘Using our effect might be more powerful and faster.’ Discovering the best way to implement these bouncing hydrogels in soft robots is something he’s happy to leave to the engineers working in that field. ‘But this is the only effect we know of where softness is an actual asset for motion, instead of a hindrance’, he says. ‘That makes this discovery a potentially very useful one.’
Scott Waitukaitis (Arizona, 1983) is an American physicist living in Amsterdam. He has a habit of taking on the physics of everyday things. He has previously also described why you can run on cornstarch and water, and why you might get shocked by doorknobs.
What do cryptography and string theory have in common? They both use mathematical equations called elliptic curves. String theorists use these equations to calculate the path of a particle, whereas cryptographers use them to encrypt data. David Holmes studies elliptic curves from a purely mathematical point of few, only occasionally touching on the many applications these equations have.
Contrary to what the name might suggest, elliptic curves are not ellipses. They are equations of the form y2=x3+ax+b. The name is mostly historical: the equations were used to calculate the circumference of ellipses. Nowadays, cryptographers use them to encrypt data because they require only small encryption keys compared to other cryptography methods. This reduces the encryption time. The equations also play a role in the pure mathematics of Fermat’s last theorem and the abc conjecture.
Physicists use elliptic curves in string theory to describe the path of a particle in space-time. To do this, the parameters ‘a’ and ‘b’ in the equations have to be a specific type of numbers, called complex numbers. Cryptographers and number theorists want the parameters to be a different type of numbers, for example rational numbers or integers. ‘But mathematically, the parameters can be anything you want’, says Holmes. ‘Mixing different types of numbers together in the equations is what I do. And it is a lot of fun.’
At first, Holmes studied elliptic curves in the context of number theory, a part of mathematics mainly devoted to the study of integers. In response to a paper he published on the preprint website ArXiv, he was invited by mathematician Alessandro Chiodo to give a talk in Paris. Chiodo works on string theory. Holmes: ‘He is one of the few who understands string theory from a mathematical as well as from a physics point of view.’ Holmes learned that his mathematical work has applications in string theory. Now he works not only on purely mathematical number theory, but also on the slightly more applicable mathematics of string theory.
Currently, Holmes looks at elliptic curves with varying parameters. If you have an equation for an elliptic curve and you vary one parameter, for example time, for each moment in time you will get another curve. The different curves you obtain by doing that are called a family. They can, for example, represent how a string in string theory evolves over time. ‘The situation with one varying parameter is quite well understood’, says Holmes: ‘I am trying to find out what happens if you vary two parameters instead of one. Then a family can evolve in two different ways. That makes it more difficult, but also more fun.’
David Holmes (London, 1986) grew up in the South of England and did his PhD in Warwick, United Kingdom. In 2012 he started as a postdoc in Hamburg, but after 3 months he moved to Leiden. He became an assistant professor there in 2014.
When searching on the Internet, you expect your search engine to come up with the results you are looking for. But what information is relevant for you, turns out to be a very personal matter. Data scientist Suzan Verberne develops search systems that help you find just the information you need.
Fake news never pops up in the timeline of Suzan Verberne’s Facebook. However, she regularly sees posts in which an article is being labelled as fake. These messages are written by Nieuwscheckers, an initiative in which students from the master Journalism & new media from Leiden University check whether news is true or false. ‘Wouldn’t it be fantastic if their manual analysis could be automated?’, Verberne asks rhetorically.
The data scientist explains the possible advantages: ‘We could better recognize what news is fake and map the network of clickbait entrepreneurs. Besides, we would be able to analyse who reads fake news and who reads messages about fake news from Nieuwscheckers. Although the people behind Nieuwscheckers suspect these messages do not arrive in the network in which the original fake news message was shared’, says Verberne, who wrote her PhD about smart search engines at Radbout University Nijmegen. In Leiden, she focuses on predicting what search results will be relevant for specific users of search engines.
Relevance is a hyperpersonal matter. What is relevant for you, doesn’t have to be interesting for someone else. Why? ‘If you ask people to highlight the most important pieces of text in an online discussion, they all choose different parts. If you ask two individuals to search for Leiden, they come up with totally different results of what they think is important’, Verberne explains. Nevertheless, people seem to agree about the quality of a text; short messages and the use of a lot of emoticons seem to contribute to a lower quality perception.
Although relevance is a challenging concept, Verberne develops experimental methods to personalize people’s search for information. She recently built Rembench, a system for art historians to search for information about the painter Rembrandt van Rijn. Verberne also aims to help scientists who search for papers in Google Scholar. ‘If I search my area of expertise, the search engine also shows results of other disciplines. That makes sense, because Google doesn’t want to give users the feeling they have entered a certain filter bubble and as a consequence distrust the system. One of my biggest research challenges for the upcoming years is to personalize searching in a transparent way, so you don’t end up in a bubble. At the same time it remains complicated to know what is relevant for someone.’
Suzan Verberne (Nijmegen, 1980) started in Leiden as a data scientist in March 2017 and helps people to search for information on the internet. She lives in a filter bubble on Twitter and doesn’t follow too many people she tends to disagree with.
Nature is extremely good at synthesizing every molecule it needs. Imitating that proficiency in the lab is challenging, but nevertheless crucial to understand what goes on in a living cell. Hans Kistemaker developed synthetic methods that enable the study of DNA repair, one of life's key processes. In May 2017, he graduated with honors on his groundbreaking chemical research.
Proteins are the workhorses of the cells in our body. They take care of energy production, transport and repair of the cell's most precious possession, the DNA. Most proteins need to be activated through chemical signals. These are small molecules that bind to their target proteins. Access to such molecules is essential to study how proteins interact. But synthesizing them in the lab was out of reach, until Hans Kistemaker came along.
Kistemaker focused on a type of signalling called ADP-ribosylation. ‘Proteins involved in the repair of DNA damage, which is a crucial process for the cell to survive, become activated when a molecule or a chain of molecules called ADP-ribose binds to them’, he explains. ADP-ribose consists of ribose, a sugar molecule, coupled via two phosphate groups to adenosine, one of the building blocks of DNA. ‘To get a detailed picture of how DNA repair works, we need ADP-ribose molecules to perform experiments and measurements. Cells have no problem in introducing ADP-ribose modifications, but isolating these molecules from cells is complicated and the yields are extremely low.’
The fact that a cell can easily produce a molecule is no guarantee that a chemist in the lab can do the same. Kistemaker embraced the challenge for two reasons. ‘I wanted to work on organic synthesis and I wanted a strong link to biological processes.’ It took a lot of hard work and failures were plenty, but he persisted and succeeded in developing two methods that enable the production of ADP-ribose molecules with precisely defined structures in useful quantities. ‘Research groups in the USA, Switzerland and Germany used the ADP-ribose molecules that I produced and showed that the synthetic molecules behave in a similar way to the ones produced by the cell. That was an important confirmation of the value of my synthetic methods.’
Control over the exact structure is important, because the cell uses a variety of ADP-ribose molecules, ranging from a monomer (one molecule) to dimers (two molecules), trimers (three molecules) and longer chain lengths. ‘Collaborators in Germany and Denmark studied the effect of the length of these ADP-ribose combinations and found that the monomer and dimer did not activate any enzyme that starts DNA repair, but the trimer bound the enzyme with very high affinity. Apparently, the length of these chains makes a huge difference in the biological effect.’ Such results make his efforts worthwhile, says Kistemaker. ‘I really enjoyed working on the synthesis and it is great to see that my work is helping us to learn more about biology.’
Hans Kistemaker (Oldenzaal, 1985), studied chemistry at the University of Groningen. He performed his bachelor and master research internships in the organic chemistry group of Nobel laureate Ben Feringa. He then moved to the bio-organic synthesis group of Hermen Overkleeft and Gijs van der Marel at the Leiden Institute of Chemistry, where he performed his PhD research under supervision of Dima Filippov. He is now senior scientist at ProQR Therapeutics, a drug development company focusing on severe genetic disorders.
Alireza Mashaghi studies the proteins that help other proteins to fold, using advanced techniques at the level of single molecules. Detailed knowledge of these chaperone proteins is needed to better understand diseases that result from mistakes in protein folding, like Alzheimer’s. Eventually, this may help scientists to design new drugs that target the chaperones, or that can act as synthetic chaperones.
Proteins play a vital role in nearly all processes in our body. Thousands of different proteins are active in every cell. Their function is not just determined by their chemical composition, but also by their 3D shape. Incorrect folding or loss of the right shape lies at the basis of diseases such as Alzheimer’s disease, Parkinson’s disease and certain cancers. Alireza Mashaghi studies proteins called ‘chaperones’ that make sure other proteins are folded in the right way.
‘Chaperones are like police officers who restore order when a situation goes wrong’, explains Mashaghi. ‘They literally grab proteins and hold them in the right shape.’ Mashaghi discovered that these chaperones guide proteins during the entire folding process, preventing mistakes in an early stage.
Mashaghi uses advanced techniques like optical tweezers, that employ laser beams to stretch single molecules at the nanometer scale. ‘I operate at the interface between physics, engineering and biomedicine’, he says. ‘Eventually we aim to design drugs that can help to treat and prevent diseases related to protein folding.’
Mashaghi discovered that a chaperone called Hsp70 can detect mistakes early in the folding process. ‘These mistakes often happen under stressful circumstances, such as inflammation’, he says. ‘Hsp70 can freeze the protein in a near-correct state until the stressful situation is resolved.’ And even under standard circumstances, this chaperone is constantly keeping proteins in the right shape. ‘Previously it was thought that chaperones can only bind to unfolded proteins’, says Mashaghi. ‘We showed that this is not true at all. They also control the shape of folded proteins, and thereby control their function.’
This is very important information, according to Mashaghi. ‘It helps to better understand diseases such as Alzheimer’s’, he says, ‘which result from clustering of proteins that have become sticky due to incorrect folding.’ So while this research may seem quite fundamental, it may very well result in clinical applications, for instance in the form of drugs that target chaperones or their production. Mashaghi: ‘We are the first lab that performs translational research with single-molecule techniques. When we know more precisely what chaperones do, we can start to manipulate them, and to guide the folding process – and perhaps, someday, fold proteins any way we want.’
Alireza Mashaghi studied medicine and physics at the same time and completed his PhD (cum laude) in physics at TU Delft in December 2012. Mashaghi came to Leiden in 2016 after being a neurology and ophthalmology fellow at Harvard Medical School. He is also a visiting scholar at Harvard University and an adjunct professor of ophthalmology at Fudan University, China and an active advocate of interdisciplinary research. ‘I am not a medical doctor who converted to science. I really have two professions.’
The soil bacterium Agrobacterium tumefaciens, a highly valued asset in biotechnology, produces a number of proteins that help it to infect and genetically modify plant cells. Xiaorong Zhang unravelled the role of one of them, VirD5. To his surprise, the protein hampers proper cell division, which is lethal. How can this toxic protein benefit the infection process?
Wounded plant roots attract the soil bacterium A. tumefaciens, which then infects and genetically modifies root cells, causing crown gall growth. In this tumorous tissue, the bacterium prospers and reproduces. During infection, A. tumefaciens pumps a number of proteins into the plant cells that enable transfer and insertion of an oncogenic piece of bacterial DNA. Xiaorong Zhang investigated the role of one of them, virulence protein D5 (VirD5).
He showed that VirD5, if present in plant cells in high amounts during a prolonged period of time, interferes with the segregation of chromosomes during cell division, increasing the risk that daughter cells will either miss a chromosome or carry an extra chromosome, which is lethal.
So, VirD5 is a toxic protein. How can it help A. tumefaciens to infect host cells? During infection, Zhang argues, the protein is delivered in a low dose during a short time. ‘Transient presence of limited amounts may just slow down cell division, creating more time for insertion of bacterial DNA into host plant DNA’, he says.
To find out what effect the VirD5 protein has on host cells, Zhang conducted a vast series of experiments, on which he reported in PNAS. His first results were embarrassing, he tells: ‘I designed a construct carrying the virD5 gene coupled to a molecular switch, brought it into cells of the model plant Arabidopsis, and cultured modified plants. When we turned on the switch in seedlings that carried the construct, they died. I had no experience in this field of research, so I thought I had done something wrong.’
But he had not, it turned out. Zhang also transformed yeast cells with the virD5 gene plus switch, and also in yeast the VirD5 protein induced death. ‘So, it must target an essential process that was conserved during evolution’, he reasoned. After testing 5,000 yeast strains, in each of which one gene was deleted, he discovered that the presence of the protein Spt4, product of gene SPT4, was required for VirD5 to exert its lethal effect. Spt4 is localized to the kinetochores of chromosomes, protein complexes that enable separation of replicated chromosomes during cell division.
He went on to show that VirD5 binds to Spt4 and impairs the process of chromosome segregation. As a consequence, the chromosomes are not accurately assigned to daughter cells, resulting in the appearance of cells with abnormal chromosome numbers.
After finishing his study at the Soochow University in Suzhou, China, Xiaorong Zhang (1986) aimed for a PhD position abroad. When he contacted Paul Hooykaas, the answer he received was so enthusiastic that he came to Leiden. He currently holds a post doc position at the Hubrecht Institute in Utrecht.
More and more nanoparticles appear in everyday life. From sweat-free socks to effective cancer medicines: the minute particles have all kinds of applications, and developments are going fast. But are all these new particles safe? We are still in time to find out, says Martina Vijver.
‘Our study on hazards and risks of nanoparticles in the environment runs parallel to technological developments’, Vijver says. ‘While much harm has already been done by insecticides, there are hardly any emissions of nanoparticles yet. We are now able to inform society and industry about possible environmental repercussions of new nanotechnology products before they become available.’
The risks that organisms run when exposed to nanoparticles are largely unknown yet. ‘These particles do not form a homogeneous solution in water, but a suspension. They may sink to the bottom or float on the surface and aggregate. This behaviour determines how organisms are exposed and how they may take up the particles.’
Vijver’s group already showed in the lab that gut cells of zebrafish embryos take up particles smaller than 50 nanometres after ingestion. She also showed that water flea embryos in the brood pouch of their mother take up nanoparticles. Now, Vijver is planning to investigate how the organisms are exposed to and take up the particles under natural conditions.
Vijver’s group already showed in the lab that gut cells of zebrafish embryos take up particles smaller than 50 nanometres after ingestion, whereupon the particles are distributed through the body and accumulate in tissues and organs, even in the eyes.
Embryos of the model crustacean Daphnia magna (water flea) take up the particles in a different way. They develop in their mother’s open brood pouch that is continually flushed to provide the embryos with fresh water. When nanoparticles reach the embryos via this water flow, they invade through the enveloping membrane and accumulate in or on lipophilic cells. Now, Vijver is planning to investigate how the organisms are exposed to and take up the particles under natural conditions.
Martina Vijver was actually nominated for two reasons. Besides her nanotechnology safety research, she also set up and manages the Living Lab, together with her team. This outdoor facility enables toxicology research on organisms under controlled conditions in their natural environment. It consists of ditches, tubs and two cabins that accommodate a field lab, and is unique in the Netherlands. Vijver’s colleagues, but also other scientists and school classes, are welcome to do research.
Vijver started The Living Lab for her toxicology research on fresh water ecosystems. When nutrients and toxic substances are introduced into ecosystems, their effects on single species will propagate through the food chain, resulting in consequences that are hardly predictable, she explains. ‘To study effects on an ecosystem as a whole, you either have to move all species to your lab, or to work in a natural environment.’ The Living Lab provides such a natural environment.
In November 2016, 38 ditches were dug next to and perpendicular to a lake, and organisms living in the lake were allowed to colonise the ditches until March 2017, when access was closed and Vijver and colleagues started their research. They added nutrients and insecticides to 27 ditches (3 treatments with 9 replicas each), keeping the remaining ditches as untreated controls. ‘We are currently analysing the results, and they look promising.’
Next to the ditches, 48 tubs are set up as outdoor mesocosms. Vijver is also planning to do nano- ecotoxicological research in these tubs next year.