Table of Contents
2. The Key Concept of the Imaging Centre: “Systems Topobiochemistry”
3. Activities of the Imaging Centre
A combination of modern microscopy and digital image processing techniques has radically changed biological research. We now visualize molecular interactions and structural dynamics as digital images to study the biological system. Extraction of numerical information out of three-dimensional space within organisms became a novel and important feature with the technological advancements. Computer simulation has become a realistic approach for biologists, as we can now simulate molecular reactions and mobility as the direct extrapolation of actual digital image sequences acquired through wet lab experiments. The Centre for Molecular and Cellular Imaging (CMCI; in the following, The Imaging Centre) will be the key for enforcing top-quality science at EMBL in this direction. I will discuss the concept for this task and then will clarify the prospective activities.
2. The Key Concept: “Systems Topobiochemistry”:
During the 20th century, biochemistry has accumulated an enormous amount of information on how biological system components are functioning. Very recently, advances in fluorescence microscopy & imaging are pushing the knowledge forward by adding spatial dimension as a new parameter. I call this new area of biochemistry the “Systems Topobiochemistry” *1 for the following reason: In traditional biochemical experiments on enzyme kinetics, reactions have been studied using solutions containing even distributions of enzymes and substrates to characterize details of the enzyme function (Fig.1 A). Spatial information has been ignored in these in vitro systems partly also due to the simple lack of technology. It allowed a clear and straightforward analysis of the results by decreasing the number of parameters and unknowns. However, it has the critical draw back that in actual biological systems, spatial position and local environment are key parameters contributing to many biochemical reactions (Fig.1 B). Today, with the advancement of microscopic and imaging technologies, the visualization of temporal changes in three-dimensional space has become possible. This enables the measurement of molecular reaction kinetics and dynamics in vivo and in situ. Various fluorescence microscopic techniques and novel probes based on fluorescent proteins have evolved in last two decades. Molecular dynamics can now be analyzed in the natural context of the biological system.
Together with the advancements in light microscopy technology and development of novel probes, two further techniques become increasingly important for the understanding of biological systems in terms of Topobiochemistry: computer simulation *2 and substrate designing.
Computer simulation: Computer simulation is becoming a powerful technique to evaluate the image-based experimental results and its interpretation of biological system. Following the analysis of image sequences from a real experiment, models of the molecular reactions and dynamics can be generated and the reaction or movement process can be simulated in silico. The experimental results may suit as an initial value, from which a parameter space can be explored and compared with the actual experiment. The simulation results will be fed back into the design and refinements of further experiments.
Substrate Designing: Taken that the computer simulation is becoming an important part of the biological research, and that it is directly reflecting the actual experiments, initial and boundary conditions in the actual experiments will be a key for the preciseness of modeling and simulation. I consider geometry as the most important initial and boundary condition, which in a biological system is typically complex and irregular. One way to manage the problems arising from irregular and complex initial and boundary conditions is to perform a geometrical transformation of the image-based experimental results for the simulation. A more direct and honest way to set a proper initial- and boundary conditions is to experimentally control the geometry of the experimental system through engineered environments (Fig.1 C). By means of various nano- and micro- fabrication techniques my colleagues and I have recently succeeded to control the cell shape and thereby also to influence the shape and position of intracellular structures such as the Golgi apparatus and mitotic spindle. Substrate designing approaches will become a standard protocol for imaging in a near future, because they allow highly parallel image acquisition of many accurately positioned and identically shaped samples under identical environmental conditions. Moreover, as it has been shown that the structure of the biological system is directly related to its function, shape has to be tightly controlled throughout the experiments. Substrate design will be an essential part of the Systems Topobiochemistry.
To summarize, Systems Topobiochemistry consists of three cutting-edge technologies (Fig.1 D): (1) visualization and analysis of molecular reactions while maintaining their spatial context within the system, (2) modeling and computer simulations of observed results within multidimensional space and (3) substrate designing to control the boundary conditions of the system. Top-biochemistry will be a powerful way to study the biological systems in this century and will be the central concept of the Imaging Centre at EMBL.
*1 The word “Topobiochemistry” is a branch-out of the “Topobiology” concept proposed by Gerald Edelman as a title of his book in 1988. Systems Topobiochemistry is a fusion of Systems biology and topobiochemistry.
*2 See also a recent review article: Bork and Serrano (2005), “Towards Cellular Systems in 4D” Cell Vol.121 p507-509.
System topo-biochemistry at other scales
The essential concept of topobiochemistry is applicable at different scales within a biological system. Its central dogma that spatial information is directly related to function holds also for other systems/environments such as super-molecular complexes, organelles, cells and multicellular structures. I have been recently involved in studies of intracellular organelle genesis and transport dynamics as well as analysis of cell movements within tissues in developmental biology. At both these scales, digital image processing enables the analysis of spatial parameters which together with computer simulations and control of the structural environment has yielded high-level analytical interpretations.
3. Activities of the Imaging Centre:
The major task of the centre will be the development and/or refinements of analytical techniques and tools for actual research projects. At the same time, spreading some of the basic knowledge to the beginners in imaging is essential for the success of topobiochemistry. For this purpose, the activity of the Imaging Centre will employ five different types of activity. (1) Training and Consulting on Image Processing, (2) Collaborative Research, (3) Software Integration and Development, (4) a Core Project and (5) Generation of links between Light Microscopy, Electron Microscopy and Structural Biology:
(1) Training and Consulting on Image Processing:
As a result of the rapid increase in applications of image based technologies, today’s biologists require some basic knowledge on image processing. This is not limited to the researchers who are involved in the advanced image acquisition and analysis, but also for those who do basic processing of simple single digital image files. Furthermore, “Imaging literacy” has rapidly become important in the interpretation and evaluation of scientific papers and presentations, though it is rarely taught in the course of biomedical or biophysical studies. Despite it has been realized some years ago by EMBL’s scientific community that there is a strong demand for training on image processing, up to now support has been rather un-organized through various groups and people. To fill this gap, Advanced Light Microscopy Facility (ALMF) has been functioning as a public resource for information on microscopy related imaging and education. However, digital image processing and analysis is a broad academic field involving physics, mathematics and computer science knowledge and demands intensive care to be kept up to state. The Imaging Centre will be responsible for the task of education on digital image processing and at the same time build and maintain a knowledge base in this complex area.
The aim of educational programs should be to enable individual researchers to complete a major part of their image processing jobs independently after an initial consultation on the project. For difficult problems that require in-depth involvement of Image Centre members and customization of the imaging software, in-house collaboration will be considered.
(2) Collaborative Research - Customized analysis - :
In most of the research projects that involve image processing and especially analysis as an important part of the study, customization of the imaging processing will be necessary. Functions native to the conventional software in many cases do not give satisfactory results in pioneering research projects, since the originality of the analytical method is directly related to the originality of the research outcome. For these reasons, collaborative researches for in-depth image processing and analyses will be an important activity of the Image Centre.
(3) Software Integration and Development:
Many researchers are currently using digital imaging microscopy to study molecular interactions, mobility and dynamics. Different non-commercial and commercial imaging software is available for these purposes, but none of them can be always satisfactory for individual research projects. The weak point of the publicly-available software is that they are too general to meet the originality of the individual research project. For example, tracking of organelles and movement velocity can be calculated with the commercial software package MetaMorph. However the success will be limited to those few cases when ideal shape and a high signal quality of the target organelles can be provided. In addition, in-depth analysis of the dynamics, which may provide rich information on the molecular interaction details, is typically hardly available. However, it may be possible to extract this information through both, consulting on the optimal imaging conditions, such as on the spatio-temporal resolutions amenable to the analysis algorithm used, and customization of the analysis program.
Therefore the problem in imaging software development is the contradiction between the generality of public software and the originality of each research project. This contradiction can never be solved completely. However there are ways to circumvent and ease the contradiction for efficient support of innovative scientific projects. The solution could be provided in two ways: software survey and software developments.
Software survey: Non-commercial and commercial imaging software will be evaluated for available functionality. These functions will be compared for their ability and missing functions in different software packages will be listed. An outcome of this survey will be recommendation lists of software for each specific research purpose. Another outcome will be information on types of software, including those already developed by EMBL staff and used at EMBL. To ease the communication between users of common software, this information will be presented on the web.
Software Development: The key concept for software development will again consist in the “Topobiochemistry”, the concept proposed above; though the term is new, this concept has been practiced by biological researchers before. As an example, the following set of image analysis techniques is regularly used by molecular, cell and developmental biologists at EMBL at different object scales.
Single Particle Tracking: nm
FRAP, Optical Flow Estimation µm
Single Organelle / Cell tracking µm to mm
All of these algorithms are currently available in different software modules programmed in different ways, some of them are custom made, and others are publicly available. What is missing is the integration of knowledge and techniques at EMBL. Even for the same technique, we see different groups using different software, mostly because of historical but not scientific reasons. Moreover, although the scale is different, the techniques listed above share their aim in analyzing the spatial information through digital imaging. All these different techniques can be integrated in the concept of “Topobiochemistry”: measurement of spatial dynamics under defined initial and boundary conditions and computer simulation directly related to the measurements. Accordingly, software will be developed as a part of a topobiochemistry package.
Having been involved in a variety of
projects during the recent years, I realized that in most of the cases
native functions of imaging software were modified or customized
macros had to be created to achieve completely satisfying analysis
From this experience I propose that one of the major goals during the
three years of the Imaging Centre shall be to transform algorithms and
available at EMBL into a general software package. In this way, a
toolbox for biological researchers in EMBL should be developed; similar
already exist in other science institutions, for example the software
Development of novel imaging plug-ins and software will be done mainly in two steps. Primordial algorithms will mainly be designed by scientific members in close collaboration with the researchers. At this stage advice form software engineers will be helpful to identify known algorithms and increase the efficiency of calculation processes. Afterwards these primordial pilot programs will be handed to the software engineers to be transformed into user-friendly software or plug-ins for a wider public use. Two step development procedures are required since innovative ideas appear at the interface of biology experiments and image processing. These innovative ideas must be first tested and confined through actual research. Problems in the algorithms and required details for the program become also clear during this first phase. Serious developments by programmers will then be the second step, with intensive discussion and information exchange between scientific members and engineering members.
(4) Core projects:
An independent core project will be carried out in the Imaging Centre. The core project will aim at the exploitation of pioneering approach in the digital image processing in biological research, to extend the topobiochemistry by the development of cutting-edge methodology, algorithms and technology. In addition to the research outcome, an independent core research project will motivate, stimulate and maintain the pioneering atmosphere of the team. In addition, the core project could be helpful in developing the future carrier of the members becoming top scientists and engineers.
(5) Imaging Links between Light Microscopy, Electron Microscopy and Structural Biology:
In deep relation to the software integration and developments, knowledge links between light microscopy, electron microscopy and structural biology will be a valuable resource for advancing the imaging technology in EMBL. In all these areas digital imaging has become essential for displaying the results. Pioneering trials are being made not only to display, but also to use more advanced imaging methods to quantify the localization of molecular complexes in a resolution higher than that of the cell biological approach. To share such information, efforts to exchange imaging knowledge in different fields will be made by close communication with these fields, such as by regular seminars centered on imaging technologies and by inviting internal/external speakers. Inside EMBL, the Image Centre will be in a close contact with other Imaging related groups in each programme and especially with the ALMF to facilitate the exchange of the hardware information and image processing knowledge.