The 5th Workshop of Structural Biology in the Helmholtz Association of the Cross Programme Activity will occur December 17 – 18, 2019 at the DKFZ in Heidelberg.
Register now and make hotel reservations in advance of the holiday season in Heidelberg.
More details on the program will follow soon.
EREC STEBBINS - Conference Host
JOHAN ZEELEN - Conference Host
“The Bio-imaging and Diffraction Beamline P11” at PETRA III in Hamburg is dedicated to structural investigations of biological samples from atomic to micrometer length scales. The beamline provides two experimental endstations: an X-ray crystallography experiment which is open to users since 2013  and a scanning X-ray microscope which allows for e.g. ptychographic imaging combined with X-ray fluorescence measurements which is currently under construction.
The flexible X-ray optics at P11 allow for tailoring the beam properties to the experimental requirements. A first mirror system located in the optics hutch is used for generating a secondary source at 65.5 m downstream from the X-ray source. With this, a parallel beam can be generated which is ideally suited for structure determinations from large unit cell systems, such as large molecular complexes . The KB system installed in the experimental hutch can be used for refocusing the secondary source in order to generate a highly intense microbeam with more than 1.3 × 10^13 ph/s in a 4 × 9 μm^2 (v × h, FWHM) focal spot at the crystallography experiment. This allows for the investigation of microcrystals and the application of novel data collection routines, such as serial crystallography using jets, the inhouse-developed tape-drive  or solid supports (e.g. Si-chips) for sample delivery [3-6].
The P11 crystallography endstation can be operated between 5.5 and 28 keV and provides full SAD/MAD capability. Energy and beam size changes can be easily realized by the users within a few minutes. The endstation is equipped with a high precision single axis goniostat. Crystals can be rapidly exchanged in less than 20 s using an automatic sample changer equipped with an in-house designed cryogenic sample gripper and a large capacity storage Dewar, providing space for 23 uni-pucks (368 samples = 3 - 4 dry shippers). Together with the Pilatus 6M detector in place, P11 allows for high-throughput crystallography and is ideally suited for industrial applications, such as e.g. fragment screening.
For 2020 an upgrade of the crystallography endstation is planned. It will include the implementation of a new Roadrunner goniometer  which will allow for conventional (rotation series) and serial crystallography on fixed targets (still images). In addition, the Pilatus 6M detector will be replaced by a faster EIGER2 16M and a new on-axis microscope for better sample visualization of micrometer-sized crystals will be installed.
 - A. Burkhardt et al., EPJ Plus 131, 56 (2016).
 - K. Schulte et al., BMC Biology 14, 14 (2016).
 - F. Stellato et al., IUCrJ 1, 204 (2014).
 - P. Roedig et al., Sci. Rep. 5, 10451 (2015).
 - K. R. Beyerlein et al., IUCrJ 4, 769 (2017).
 - A. Meents et al., Nat. Commun. 8, 1281 (2017).
X-ray crystallographic fragment screening has gained great popularity for development of tool compounds and drug candidates. At the HZB, we have established over the last years the complete workflow of a fragment-screening experiment. Protein crystals can be soaked with individual fragments (<300 Da compounds) at high concentrations and prepared for a diffraction experiment. Diffraction data can then be collected at one of our state-of-the-art beamlines for macromolecular crystallography, BL14.1 or BL14.2. Automated data processing is achieved via XDSAPP and refinement and hit detection via fspipeline.
In addition, we have recently developed in collaboration with the Drug Design Group in Marburg a novel compound library aiming for maximal chemical diversity called the F2X-Universal Library. From this over 1000 fragment large set we derived a compact, yet again maximally diverse 96 fragment subset called the F2X-Entry Screen. The libraries are dispensed via acoustic liquid handling and dried on ready-to-use 96-well plates, enabling easy experimental workflows. In the presentation the complete workflow will be presented, as well as the requirements and possibilities.
The transformation of the cryo-EM technique to a mainstream structural biology method is ongoing. After the cryo-EM “resolution revolution”, the method has delivered a plethora of critical structures in fundamental biological fields such as transcription, membrane biology and many more. As a result, the field has undergone tremendous growth in user base, but at the same time, it still requires further sustainable development of hardware and software. In addition to high-resolution structure determination, it also has unique capabilities to obtain structural information in situ, within complex environments, and to deal with heterogeneous samples. In my talk, I will review the recent activities about the cryo-EM setup at the Ernst-Ruska Centre (ERC) in Jülich and illustrate them using several protein assemblies involved in membrane trafficking events such as autophagy. The recently funded ERC 2.0 will provide the framework for future activities including further development of novel hardware to improve image quality and extend the spectrum of structural biology applications of the cryo-EM technique.
The Helmholtz PSPF is a cooperation between the HZI, MDC, HMGU and CSSB. Experts in various techniques for high quality protein expression offer their know-how and support to produce difficult to express proteins for the Helmholtz research community. An overview of the available technologies and the latest development in plasmid-based transfection will be presented.
The virus-free transient gene expression (TGE) in High Five insect cells recently evolved to an efficient protein production method. However, up to now the published TGE protocols are not standardized and remain difficult to establish and reproduce in different labs.
Here we present the factors determining the reproducibility and robustness of the method. The type of polyethylenimine, growth phase of the cells, passage number, origin of cell line isolates and cultivation media were considered to optimize and create a robust method with high protein yields.
Chair Erec Stebbins
Immunotherapy is a burgeoning field in the treatment of numerous diseases beyond infection. The use of monoclonal antibodies and vaccines for cancer and other maladies is revolutionizing medicine. However, the approaches to develop successful immunotherapies are limited and tied to established, older technologies. We present here a novel immunotherapeutic platform with great promise. The technology is based on fusing epitopes to the Variant Surface Glycoprotein of the African trypanosome, resulting in an array of ten million epitope copies presented on the surface of a highly immunogenic organism. We have established technologies to link protein, peptide, carbohydrate, and small molecule moieties to this coat. Structural considerations of the coat protein were integral in the development of the vaccine. Finally, we present a test case with a difficult immunotherapeutic problem that has eluded researchers to date: how to develop robust and effective immunity to small molecules of interest. We show that we are able to use our novel platform to solve this problem, and present an application in which we successfully develop immunity in animals against the opioid drug fentanyl.
cis-Aconitate decarboxylase (CAD, also known as ACOD1 or Irg1) converts cis-aconitate to itaconate and plays central roles in linking innate immunity with metabolism and in the biotechnological production of itaconic acid by Aspergillus terreus. We have elucidated the crystal structures of human and murine CADs and compared their enzymological properties to CAD from Aspergillus terreus. Recombinant CAD is fully active in vitro without a cofactor. Murine CAD has the highest catalytic activity, whereas Aspergillus CAD is best adapted to a more acidic pH. CAD is not homologous to any known decarboxylase and appears to have evolved from prokaryotic enzymes that bind negatively charged substrates. CADs are homodimers and the active center is located in the interface between two distinct subdomains. We identified eight active site residues critical for CAD function. Rare naturally occurring human mutations in the active site abolished CAD activity, whereas a variant that is common in African populations (Asn152Ser allele frequency 20%) increased CAD activity. Residues near the active site of human CAD were replaced with those present in the mouse counterpart and several mutations lead to increased activity. It was found that the activity of CAD was strongly decreased during human evolution by a single mutation that changed a methionine residue, which is conserved in mammals, to isoleucine. The present study opens the way for (i) assessing the potential impact of human CAD variants on disease risk at the population level, (ii) developing therapeutic interventions to modify CAD activity, and (iii) improving CAD efficiency for biotechnological production of itaconic acid.
Human noroviruses are a leading cause of viral acute gastroenteritis, but so far neither vaccines or antivirals are available. Expression of the norovirus capsid protein (VP1) typically results in the formation of VLPs that are morphologically similar to native virions that have a T=3 icosahedral symmetry. In T=3 particles, 180 copies of VP1 adapt three quasiequivalent subunits (A, B and C). However, we discovered that several genetic variants of GII.4 VLPs exhibited T=4 icosahedral symmetry, and 240 copies of VP1 form four subunits (A, B, C and D). In this study, we showed that the sequenced-modified VLP vaccine candidate, GII.4c, mostly forms T=4 icosahedral symmetric particles, while a smaller percentage of particles had T=3 and T=1 icosahedral symmetry. Preliminary antigenic and structural analysis between the T=1 and T=4 VLPs indicated that for several epitopes, antigenicity and accessibility likely differed, suggesting that structural differences could be important for vaccine development.
In recent decades, antibiotic resistance has emerged as a significant risk to public health. Resistance to all known classes of antibiotics, including the drugs of last resort, the carbapenems, has now emerged. Carbapenems are members of the beta-lactam class, the most commonly and clinically relevant antibiotic class used in treatment of gram-negative and gram-positive bacterial infections and are now being rendered useless by beta-lactamases synthesized by bacteria.
Our work focuses on elucidating the structure and dynamics of the active site of beta-lactamases and discovering low molecular weight inhibitors which bind and inhibit the activity of the enzyme. We screened a diverse fragment library using NMR methods to find hits. Selected molecules were assayed to characterize inhibitory activity and then co-crystallized to elucidate fragment location, binding mode and to guide possible further development. For the first time, several clinically relevant antibiotics were also crystallized with IMP-13, giving us vital information on the availability of chemical space around the active site and its use in the context of developing fragment screen hits. Protein dynamics studies using NMR and molecular dynamics simulations show that metallo-beta-lactamase IMP-13 shows high L1 loop flexibility crucial for substrate recognition and processing. This is supported by crystal structures apo-state IMP-13, demonstrating different loop conformations, as well as in complex with hydrolyzed carbapenems.
Utilizing the variety of obtained results, we hope to develop potent new molecules targeting beta-lactamase activity, to aid in the fight against antibiotic resistance.
We present a Focused Library Generator that is able to create from scratch new molecules with desired properties. After training the Generator on the ChEMBL database, transfer learning was used to switch the generator to producing new Mdmx inhibitors that are a promising class of anticancer drugs. Lilly medicinal chemistry filters, molecular docking, and a QSAR IC50 model were used to refine the output of the Generator. Pharmacophore screening and molecular dynamics (MD) simulations were then used to further select putative ligands. Finally, we identified five promising hits with equivalent or even better predicted binding free energies and IC50 values than known Mdmx inhibitors. The source code of the project is available on https://github.com/bigchem/online-chem.
Parkinson’s disease (PD) and Multiple System Atrophy (MSA) are clinically distinctive diseases that feature a common neuropathological hallmark of aggregated α-synuclein. Little is known about how differences in α-synuclein aggregate structure affect disease phenotype. Here, we amplified α-synuclein aggregates from PD and MSA brain extracts and analyzed the conformational properties using fluorescent probes, NMR spectroscopy and electron paramagnetic resonance. We also generated and analyzed several in vitro α-synuclein polymorphs. We found that brain-derived α-synuclein fibrils were structurally different to all of the in vitro polymorphs analyzed. Importantly, there was a greater structural heterogeneity among α-synuclein fibrils from the PD brain compared to those from the MSA brain, possibly reflecting on the greater variability of disease phenotypes evident in PD. Our findings have significant ramifications for the use of non-brain-derived α-synuclein fibrils in PD and MSA studies, and raise important questions regarding the one disease-one strain hypothesis in the study of α-synucleinopathies.
Hydrogenases are the bio-catalysts that catalyze the oxidation and generation of molecular hydrogen in a highly efficient and reversible manner. While [NiFe] hydrogenases are the most common in nature, it is the [FeFe] hydrogenases that are the most active, but at the same time they are the most oxygen-sensitive. The active site of [FeFe] hydrogenases consists of the H-cluster, which is constructed from a canonical [4Fe-4S] cluster coupled to a unique [2Fe] sub-cluster containing a unique bridging azapropane dithiolate as well as carbonyl and cyanide ligands. The H-cluster has been subjected to numerous spectroscopic and functional studies over the years, but as yet there is no clear consensus concerning how the catalytic cycle operates and how spectroscopically identified intermediates can be structurally rationalized. The enzyme from the sulfate-reducing bacterium Desulfovibrio desulfuricans (DdHydAB) is one of the most active and bidirectional [FeFe] hydrogenases. An additional interesting feature of DdHydAB is that it can be purified aerobically in an oxygen-stable inactive state called Hinact. This state is thought to be over-oxidized with a Fe(II)Fe(II) configuration possessing an additional ligand at the open coordination site on the distal Fe. We have crystallized DdHydAB and solved its structure in the oxygen-stable, inactive Hinact state. The structural data is supported by spectroscopy and theoretical calculations.
The dynamin protein catalyzes the end stage of clathrin-mediated endocytosis by forming oligomeric filaments around the membrane neck and performing scission in a GTP-dependent manner. Because of its central role in this important process, significant effort has been spent in determining the structural nature of membrane-bound dynamin oligomers. Important insights have included cryo-EM studies of dynamin filaments decorating membrane tubes and X-ray crystallography studies determining the ligand-dependent conformations of the GTP-binding domain (Gdomain). Using molecular simulation and mesoscale modeling, we combine the structural knowledge with new stopped-flow kinetics and smFRET studies of the Gdomain to show that the assumed GTP-dependent powerstroke is geometrically and kinetically consistent with the available data. Overall, our study indicates that dynamin can function as a processive molecular machine.
Five decades after its initial discovery, the molecular mechanisms underlying ρ-dependent transcription termination and its regulation remain unresolved1. We have determined cryo-electron microscopy structures of ρ/NusA/NusG/rut RNA-modified transcription elongation complexes that portray four stages of an RNA polymerase-dependent termination pathway. Together with structure-informed functional analyses, the atomic models illustrate how ρ can contact RNA polymerase at multiple sites, how ρ recruitment depends on NusA, how NusA can counteract ρ on polymerase by multiple mechanisms, how ρ can take advantage of the NusG C-terminal domain to inactivate RNA polymerase by modulating DNA re-annealing and displacing template DNA in the active site, how the NusG N-terminal domain guides ρ through part of the termination pathway and how a dynamic interplay of the RNA polymerase zinc-binding domain, NusA and NusG can initiate RNA handover to the ρ primary RNA binding sites. Fundamental principles of ρ action uncovered here may apply to diverse termination factors across kingdoms.
Legionella pneumophila is a bacterial pathogen that, upon inhalation of contaminated aerosols, causes Legionnaires’ disease, a severe form of pneumonia. L. pneumophila replicates intracellularly in macrophages, epithelial cells and in environmental hosts such as amoebae. Numerous virulence factors of L. pneumophila are known, including secreted effectors and surface-associated proteins. Its virulence together with the ability of L. pneumophila to lyse membranes and trigger pore formation for vacuolar escape after intracellular replication likely is related to a number of A-type phospholipases (PLA). The PLAs of L. pneumophila count at least 15 members, clustered in three families .
PlaB, which is the first representative of a new PLA family distinguished by a conserved THSTG-motif, is an outer membrane-associated virulence factor of L. pneumophila. It possesses the highest PLA activity of the phospholipases produced by this pathogen and is also required for its hemolytic activity . Concentration-dependent tetramerization leads to inhibition of PlaB, pointing towards a mechanism for bacterial self-protection . It is known that the N-terminal domain of PlaB contains the catalytic triad Ser85/Asp203/His251, which is essential for phospholipase as well as hemolytic activity. Hemolysis, however, also depends on fifteen C-terminal residues as indicated by truncation mutants. Interestingly, a 170 amino acid long region in the C-terminal half has not been assigned to any function yet.
A laborious structural determination finally confirmed the organization of PlaB into two domains. The unassigned C-terminal domain was discovered to consist of a bilobed beta-sandwich, decorated with a set of cation-pi interactions on the surface that likely are necessary for protein-lipid interaction. In addition, the importance of the fifteen C-terminal residues for dimerization and enzymatic activity was clarified. Remarkable structural features were found in both domains, including the mentioned cation-pi interactions but also secondary structure elements with exceptional composition and relative positioning in the protein. Those features were in the following addressed by a set of mutations. For certain mutants, large differences in enzymatic activity and cell localization of PlaB were observed.
Bringing together structure elucidation, studies of oligomerization and protein localization as well as biochemical characterization of PlaB mutants, we can now draw a better picture of the structure-function relationship of this potent virulence factor from L. pneumophila.
 S. Banerji, P. Aurass, and A. Flieger, “The manifold phospholipases A of Legionella pneumophila – Identification , export , regulation , and their link to bacterial virulence,” Int. J. Med. Microbiol., vol. 298, pp. 169–181, 2008.
 J. Bender, K. Rydzewski, M. Broich, E. Schunder, K. Heuner, and A. Flieger, “Phospholipase PlaB of Legionella pneumophila represents a novel lipase family. Protein residues essential for lipolytic activity, substrate specificity and hemolysis,” J. Biol. Chem., vol. 284, no. 40, pp. 27185–27194, 2009.
 K. Kuhle et al., “Oligomerization inhibits legionella pneumophila PlaB phospholipase A activity,” J. Biol. Chem., vol. 289, no. 27, pp. 18657–18666, 2014.
Microtubules (MTs) are dynamic polymers of αβ-tubulin and play critical roles in cell signaling, cell migration, intracellular transport processes and chromosome segregation. They assemble de novo from α/β-tubulin dimers in an essential process termed MT nucleation. Complexes containing the protein γ-tubulin serve as structural templates for the MT nucleation reaction. In vertebrates, MTs are nucleated by the 2.2 MDa γ-tubulin ring complex (γ-TuRC) composed of γ-tubulin, five related γ-tubulin complex proteins (GCP2-6) and additional factors. High-resolution structural information on the γ-TuRC is not available, strongly limiting our understanding of MT formation in cells and tissue. Here, we present the cryo-EM structure of γ-TuRC from Xenopus laevis at 4.8 Å global resolution, revealing a 14-spoked arrangement of GCPs and γ-tubulins in a partially flexible open left-handed spiral with a uniform sequence of GCP variants and unexpectedly one molecule of Actin. The γ-TuRC spiral geometry is suboptimal for MT nucleation and a controlled conformational rearrangement of the γ-TuRC is required for its activation. Collectively, our cryo-EM reconstruction provides unprecedented insights into the molecular organization, the assembly and the activation mechanism of vertebrate γ-TuRC and will serve as an important framework for the mechanistic understanding of fundamental biological processes associated with MT nucleation, e.g. meiotic and mitotic spindle formation and centriole biogensis.
Molecular machines that convert chemical energy into motions are exciting targets for NMR spectroscopists. Their functionality combines two areas in which NMR allows for unrivalled molecular insight: monitoring chemical turnover of small compounds and protein conformational transitions.
In this distribution we ask how one can dissect the underlying enzymatic reaction steps at the atomic scale. We demonstrate that the integration of solution- and solid-state NMR approaches allows us to follow the events from the perspectives of both the protein and the substrate. We can thus derive a more complete picture of the chemical events unfolding and develop a more detailed mechanistic model of the operating mode of molecular motors.
Serial Crystallography at X-ray Free Electron Laser (XFEL) and modern Synchrotron sources enables structure determination from micron to sub-micron sized crystals of biological macromolecules. Because of their size and thus possible short diffusion times of substrates or ligands into them, those crystals can be used for mix-and-diffuse serial crystallography experiments to unravel structural dynamics and/or enzymatic reaction pathways at atomic spatial- and µs to ms time resolution. Moreover, it should enable faster and more accurate determination of ligand- or fragment binding to target proteins in drug discovery processes. Both is, in addition, simplified by the fact that Serial Crystallography comes with the benefit that cryo-cooling of the crystals is not required and thus all measurements can be carried out at room temperature or at any other temperature that the crystals can withstand, i.e. physiological temperatures. The ultimate goal would be to produce molecular movies from such experiments and complementary techniques (e.g. spectroscopy, ultra-microscopy) to understand the dynamics of biological systems on length-scales from atoms to cells and organisms. Here I will present current developments for Serial Crystallography at both Synchrotrons and XFELs [1,2], with special emphasis on experiments at the European XFEL [2,3] and latest progress on the way towards optimized experiments.
The free energy landscape of proteins and hence their conformational space is defined by multiple variables, such as protein concentration, ligands and solvents, but also thermodynamic parameters like temperature and pressure. Pressure as physical variable is known to have a strong influence on the protein fold and especially on enzymatic activity. High-pressure experiments are a valuable tool to reveal rare conformational states of proteins, thereby gaining insights into protein function such as enzymatic reaction cycles or identifying new allosteric binding sites of pharmaceutically relevant target proteins. As a result, the design of pressure-adapted enzymes for future high-pressure applications in biotechnology or the development of new lead compounds can be facilitated. High-pressure crystallography of macromolecules provides information on the influence of pressure on the protein structure at atomic detail. However, due to its technical complexity it still is a niche method, not accessible to a large part of the scientific community. We developed a new system for high-pressure protein crystallography including a high-pressure cell with minimized total sample volume and easier handling compared to the diamond anvil cells conventionally used for theses high-pressure experiments. At the same time, our cell provides space for numerous crystals sufficient to collect complete datasets at several adjustable pressure points with only one loading process of the cell. We plan to make this device available to the general user community at beamline P11 of the PETRA III storage ring at DESY. Here, I will present some preliminary results of the first experiments.
Today’s X-ray sources offer continuously brighter and more intense X-rays for the analysis of macromolecular crystals and other materials. Beamlines provide micrometer sized X-ray foci enabling data collection from crystals that are in the low micrometer range. However, radiation damage caused by the high-intense X-rays and background scattering through interaction of the direct beam with air limit the amount of data that can be collected from such crystals. The life time of typical microcrystals in these high-intensity X-ray beams is in the sub-second range. Radiation damage occurs when photons are absorbed by the crystals and electrons are released. Free radicals modify the macromolecule, ultimately resulting in the loss of high-resolution Bragg reflections due to increased disorder in the crystals. The energy absorption is dependent on the energy of the absorbed X-rays and it has been suggested that data collection at higher energies (20-30 keV) than typically used at macromolecular beamlines reduces radiation damage. Here we show that data collection at 26 keV is indeed beneficial for the quality of the acquired data. At the PETRAIII beamline P11 we temporarily installed a setup optimized for high-energy data collection and also low background scattering, ensuring optimal data quality. We collected data from 20 µm sized crystals of the protein kinase DRAK2. Comparing data collected at the standard energy of 12 keV (1.033 A) with 26 keV (0.479 Å), the resolution limit was 0.3 Å higher at the higher energy (2.5 vs. 2.2 Å). Furthermore, we analyzed microcrystals (5 µm) of polyhedra from the cytoplasmic polyhedral virus 18. With our high-energy, low background setup, we were able to collect data to 1.13 Å at cryo- and 1.5 Å at room-temperature, which is 0.37 Å, resp, 1 Å higher than previously published for these crystals.
These developments will be highly beneficial for projects where only microcrystals are available. Moreover, also ligand-binding studies and fragment screening campaigns will benefit from using microcrystals. With the smaller crystals, it will be possible to use ligand solutions with much lower DMSO and ligand concentrations as the diffusion into the crystals will be faster and is expected to create less stress for the crystals, preserving the initial diffraction properties of the crystals.
In summary, we present an optimized data-collection setup to get the best data out of microcrystals.
In general, structure and function of biomolecules can be strongly influenced by their environment. This is in particular true for membrane (associated) proteins and soluble proteins inside a densely-packed cell. Modern structural biology therefore requires advanced biochemical tools to generate adequate complex environments as well as techniques that can be used in these environments and still report on structural details with high resolution.
In the last years our research has contributed to both aspects and recent advancements will be addressed, in particular:
The use of adequate environments including the potential of the lipid bilayer nanodiscs system for NMR-based structural studies of membrane proteins [1-4] and protein-membrane interactions as modulators in signaling or protein aggregation [5, 6].
The usage of most suitable spectroscopic techniques including the effective use of the available magnetization in challenging systems  and the possibility to selectively hyperpolarize a target protein in a cellular context .
 M. Etzkorn, et al. (2013) Structure 21, 394-401.
 F. Hagn, et al. (2013) J Am Chem Soc 135, 1919-1925.
 A. Viegas, et al. (2016) Biol Chem 397, 1335-1354.
 T. Viennet, et al. (2019) Front Mol Biosci 6, 13.
 M. Falke, et al. (2019) Chem Phys Lipids 220, 57-65.
 T. Viennet, et al. (2018) Commun Biol 1, 44.
 A. Viegas, et al. (2016) J Biomol NMR 64, 9-15.
 T. Viennet, et al. (2016) Angew Chem Int Ed Engl 55, 10746-10750.
Activation of CD4+ T-cells by concerted ligation of the T-cell receptor (TCR) and the CD28 co-receptor drives adaptive immune responses. The MALT1 paracaspase encoded by the MALT1 gene plays a key role in the cellular signaling pathways that promote T-cell activation. The MALT1 pre-mRNA was recently shown to express two isoforms, MALT1A and MALT1B, which include or exclude exon 7, respectively. TCR engagement induces alternative splicing and thereby an increase of MALT1A expression in activated CD4+ T-cells, which ultimately augments optimal T-cell activation. The splicing factors hnRNP U and hnRNP L were identified to exert antagonistic roles in MALT1 alternative splicing. Whereas hnRNP U suppresses inclusion of exon 7, hnRNP L promotes inclusion.
Here, we study structural and functional mechanisms of alternative splicing regulation of MALT1 exon7 by hnRNP U and hnRNP L. By correlating in vitro binding studies using gel shift assays, ITC and NMR with cellular splicing data using an exon7 minigene, we have mapped several intronic hnRNP U binding sites 3’ and 5’ to MALT1 exon 7. Furthermore, competition-binding assays with hnRNP U and hnRNP L reveal that these proteins compete for MALT1 pre-mRNA, consistent with their antagonistic roles in T-cell activation. Using SHAPE chemical probing, we determined the secondary structure of a functional MALT1 minigene and mapped hnRNP U (and hnRNP L) binding sites. We have identified secondary structured motifs in the RNA that have physiological relevance in T-cells, as confirmed by mutational analysis. Moreover, we have mapped regions in hnRNP U that bind to cis-regulatory motifs in MALT1 pre-mRNA and promote exon7 skipping and are currently studying the underlying structural mechanisms using NMR and X-ray crystallography. Our data suggest a model where hnRNP U binds to multiple regions of MALT1 flanking exon 7 in the pre-mRNA, thereby exposing exon 7 for splicing, while hnRNP L has an opposite effect.