ERC FUNDED PROJECTS

The goal of the project is to automatically analyse human activities observed in videos. Any solution to this problem will allow the development of novel applications. It could be used to create short videos that summarize daily activities to support patients suffering from Alzheimer's disease. It could also be used for education, e.g., by providing a video analysis for a trainee in the hospital that shows if the tasks have been correctly executed. The analysis of complex activities in videos, however, is very challenging since activities vary in temporal duration between minutes and hours, involve interactions with several objects that change their appearance and shape, e.g., food during cooking, and are composed of many sub-activities, which can happen at the same time or in various orders. While the majority of recent works in action recognition focuses on developing better feature encoding techniques for classifying sub-activities in short video clips of a few seconds, this project moves forward and aims to develop a higher level representation of complex activities to overcome the limitations of current approaches. This includes the handling of large time variations and the ability to recognize and locate complex activities in videos. To this end, we aim to develop a unified model that provides detailed information about the activities and sub-activities in terms of time and spatial location, as well as involved pose motion, objects and their transformations. Another aspect of the project is to learn a representation from videos that is not tied to a specific source of videos or limited to a specific application. Instead we aim to learn a representation that is invariant to a perspective change, e.g., from a third-person perspective to an egocentric perspective, and can be applied to various modalities like videos or depth data without the need of collecting massive training data for all modalities. In other words, we aim to learn the essence of activities.

1 499 875 €

Start date: 2016-06-01, End date: 2021-05-31

The formation of a functional patterned vascular network is essential for development, tissue growth and organ physiology. Several human vascular disorders arise from the mis-patterning of blood vessels, such as arteriovenous malformations, aneurysms and diabetic retinopathy. Although blood flow is recognised as a stimulus for vascular patterning, very little is known about the molecular mechanisms that regulate endothelial cell behaviour in response to flow and promote vascular patterning. Recently, we uncovered that endothelial cells migrate extensively in the immature vascular network, and that endothelial cells polarise against the blood flow direction. Here, we put forward the hypothesis that vascular patterning is dependent on the polarisation and migration of endothelial cells against the flow direction, in a continuous flux of cells going from low-shear stress to high-shear stress regions. We will establish new reporter mouse lines to observe and manipulate endothelial polarity in vivo in order to investigate how polarisation and coordination of endothelial cells movements are orchestrated to generate vascular patterning. We will manipulate cell polarity using mouse models to understand the importance of cell polarisation in vascular patterning. Also, using a unique zebrafish line allowing analysis of endothelial cell polarity, we will perform a screen to identify novel regulators of vascular patterning. Finally, we will explore the hypothesis that defective flow-dependent endothelial polarisation underlies arteriovenous malformations using two genetic models. This integrative approach, based on high-resolution imaging and unique experimental models, will provide a unifying model defining the cellular and molecular principles involved in vascular patterning. Given the physiological relevance of vascular patterning in health and disease, this research plan will set the basis for the development of novel clinical therapies targeting vascular disorders.

1 618 750 €

Start date: 2016-09-01, End date: 2022-02-28

Dietary fibres are recognized for their health promoting properties; nevertheless, many of the physicochemical mechanisms behind these effects remain poorly understood. While it is understood that dietary fibres can associate with small molecules influencing, both positively or negatively their absorption, the molecular mechanism, by which these associations take place, have yet to be elucidated We propose a study of the binding in soluble dietary fibres at a molecular level to establish binding constants for various fibres and nutritionally relevant ligands. The interactions between fibres and target compounds may be quite weak, but still have a major impact on the bioavailability. To gain insight to the binding mechanisms at a level of detail that has not earlier been achieved, we will apply novel combinations of analytical techniques (MS, NMR, EPR) and both natural as well as synthetic probes to elucidate the associations in these complexes from macromolecular to atomic level. Glucans, xyloglucans and galactomannans will serve as model soluble fibres, representative of real food systems, allowing us to determine their binding constants with nutritionally relevant micronutrients, such as monosaccharides, bile acids, and metals. Furthermore, we will examine supramolecular interactions between fibre strands to evaluate possible contribution of several fibre strands to the micronutrient associations. At the atomic level, we will use complementary spectroscopies to identify the functional groups and atoms involved in the bonds between fibres and the ligands. The proposal describes a unique approach to quantify binding of small molecules by dietary fibres, which can be translated to polysaccharide interactions with ligands in a broad range of biological systems and disciplines. The findings from this study may further allow us to predictably utilize fibres in functional foods, which can have far-reaching consequences in human nutrition, and thereby also public health.

1 500 000 €

Start date: 2016-04-01, End date: 2022-03-31

Eukaryotic cells constantly convert signals between biochemical energy and mechanical work to timely accomplish many key functions such as migration, division or development. Filopodia are essential finger-like structures that emerge at the cell front to orient the cell in response to its chemical and mechanical environment. Yet, the molecular interactions that make the filopodia mechanosensitive are not known. To tackle this challenge we propose unique biophysical in vitro and in vivo experiments of increasing complexity. Here we will focus on how the underlying actin filament bundle regulates filopodium growth and retraction cycles at the micrometer and seconds scales. These parallel actin filaments are mainly elongated at their barbed-end by formins and cross-linked by bundling proteins such as fascins. We aim to: 1) Elucidate how formin and fascin functions are regulated by mechanics at the single filament level. We will investigate how formin partners and competitors present in filopodia affect formin processivity; how fascin affinity for the side of filaments is modified by filament tension and formin presence at the barbed-end. 2) Reconstitute filopodium-like actin bundles in vitro to understand how actin bundle size and fate are regulated down to the molecular scale. Using a unique experimental setup that combines microfluidics and optical tweezers, we will uncover for the first time actin bundles mechanosensitive capabilities, both in tension and compression. 3) Decipher in vivo the mechanics of actin bundles in filopodia, using fascins and formins with integrated fluorescent tension sensors. This framework spanning from in vitro single filament to in vivo meso-scale actin networks will bring unprecedented insights into the role of actin bundles in filopodia mechanosensitivity.

1 499 190 €

Start date: 2016-03-01, End date: 2021-08-31