University of Heidelberg

Assay Technology

Coordination: Dr. Holger Erfle | BioQuant, ViroQuant-CellNetworks RNAi Screening Facility | Holger.Erfle (at)


Resources and technologies for the systematic high-throughput and high-resolution functional analysis of non-coding RNAs as regulators of secretory membrane trafficking


Functional Analysis of non-coding RNAs in living Cells


The last years have seen an explosion in the number of RNA transcripts with no apparent protein coding potential. This raises the question, if non-coding RNA (ncRNA) might be as important as proteins in the regulation of vital cellular functions and what their contribution to a systematic description of live is.

We here combine the expertises of seven academic groups and two industrial partners with the aim to establish a high-throughput and high-content microscopy-based system to modulate the expression of non-coding endogenous RNAs for analysing their regulatory role in living cells in a comprehensive and predictive way. The focus will be done on the analysis of the secretory membrane traffic as a response system to changes of to levels of ncRNAs under the rationale that 1) the secretory membrane traffic evolves to be an essential part of adaptive cellular response; 2) ncRNAs are currently gaining a role of potent and comprehensive regulators of diverse cellular functions.

This project is going to be the milestone in a high-throughput characterisation of the regulatory role of ncRNA molecules as well as the complexity of their regulatory potential.It will open a new avenue towards the understanding of biological systems on a yet unprecedented level if to compare to current experimentation with coding genes, their transcripts and expressed proteins. The experimental platform combining high-throughput and ultra-high resolution microscopy, strategies for data handling, evaluation and modeling, and reagents to target selected ncRNAs specifically will produce a unique set of data enabling to analyse the secretory membrane trafficking as a complex and adaptive cellular response system for the first time.


Links between development platforms:



Project partners 


As the number of ncRNAs might reach thousands or even tens of thousands, a ultra-high throughput experimental format compatible with fluorescence microscopy - high density cell arrays - will be developed together with Benedikt Busse and Klaus Burkert. With this we will quantitatively examine changes in protein secretion and the dynamics of secretory organelles and the cytoskeleton by automated correlative microscopy. This approach as no other technique will enable to detect changes in secretory processes within the cellular context due to the power of the time-resolved high-throughput scale.

The quantitative, time-resolved data acquisition platform developed will not only provide the foundation for dynamic modeling of regulatory processes in secretory membrane trafficking (Karl Rohr / Lars Kaderali / Reinhard Schneider), but will at the same time yield an experimental platform for high throughput data acquisition of ncRNAs for systems biology in general, allowing the acquisition of quantitative, time-resolved and multiplexed phenotypic readouts after ncRNA knockdowns.


The group of Reinhard Schneider will generate a comprehensive and non-redundant database of miRNAs and their targets from already available resources will be produced. The annotations will store the origin and evidence of the information source, which will be extended by part- or fully automatic workflows where we will use an existing text-mining analysis pipeline. The framework will allow for interactive and batch sequence database searches and will provide easy access to already existing prediction and analysis tools. That will be extended to analyse other species of ncRNAs (antisense RNA, long-non-coding RNA, pseudogenes), which will be targeted in this project. To ensure a comprehensive view of the available information the database of ncRNA’s will be linked to other data sources and retrieval systems in molecularbiology. To take full advantage in the context of systems biology, the ncRNAs and their targets will be mapped in addition to other biological information like diseases links, involved pathways, Gene Ontology annotations, protein–protein or protein-chemistry interactions. In addition a range of existing and newly developed data mining technologies will be applied to disentangle the inherent networks. The goal will be the better understanding of RNA-based regulatory interactions and the prediction and identification of new targets. The data produced within the project will require a sophisticated management and analysis system. The group will set up an appropriate general-purpose framework and maintain it throughout the project. An interactive web-interface and a programmatic API via web-services will be developed and deployed in collaboration with the project partners: Dr. H. Erfle, Dr. A. Mokhir, Dr. L. Kaderali and Dr. V. Starkuviene. In addition, experimental annotation to existing database entries of ncRNA will be deposited in existing databases for public use.


In this Project Graffinity will contribute to the development of high density cell-arrays. This platform should allow about 10.000 fluorescence microscopy experiments for one cell culture incubation. In close cooperation with the project partners (Benedikt Busse, Dr. H. Erfle) an efficient layout will be designed. To manage this array setup, miniaturization technologies including photolithography and thin film evaporation will be provided. The establishment of suitable spotting technology to ensure a reliable transfer of the ncRNA interacting substance library to the high density array is also necessary. By considering bio-compatibility for the  cell-array design, a usefull means of  high throughput  and high content ncRNA analysis will be generated.


Data normalization and statistical analysis are essential procedures in the analysis of RNAi screens in order to reliably compare data from different experiments. Many algorithms for statistical analysis of DNA microarray gene expression data have been reported over the last years, however, these are not directly suitable for high-throughput microscopy data. Frequently, even functional analysis data are analyzed by excluding there resolution power of microscopy readouts and only averaging the signal over all cells within one well. We will closely interact with Dr. V. Starkuviene and the ViroQuant-CellNetworks RNAi Screening Facility in designing the experiments for optimal statistical power. Based on results of and in close collaboration with Prof. K. Rohr, the group will  develop methods for the normalization and statistical analysis of high-content, high-resolution cell arrays with microscopy-readout at the level of individual cells. We will use and develop methods based on mixture models and adapt appropriate  procedures for hit scoring.
In parallel, the group will focus on developing tools for network inference and parameter estimation, making use of multiple phenotypic readouts obtained for validated hits, images features and making use of time resolved readouts in living cells.
Finally, in an iterative procedure between experiment and modeling this will make it possible to develop a mathematical model of regulatory processes in the secretory pathway. This work will be accomplished in close co-operation with Dr. V Starkuviene and Dr. R. Schneider. Combining Bayesian statistical learning, Bayesian networks, and ordinary differential equations, this approach reaps the advantages of both statistical learning and dynamic modeling based on differential equations.



 Strategical partnership
Contact: E-Mail (Last update: 23/07/2011)