B02: RNA determinants of the antiviral innate immune response

RNA determinants of the antiviral innate immune response

Viral RNAs trigger the innate immune response, which is able not only to distinguish foreign RNA from the cell’s own diverse complement of RNA molecules, but also varies among viral pathogens. Macrophages act as the first line of defence against invading viruses and are pivotal for the innate immune response, crucially regulating the entire inflammatory process during viral infections. Single- or double-stranded RNA viruses are recognised by several classes of RNA-binding proteins, in particular, the RNA-dependent toll-like receptors (TLR), TLR-7 and TLR-8, the RIG-I-like receptors (RLR) RIG-I, MDA5, LGP2, and their homologues, and the double-strand sensing protein kinase R (PKR). While individual RNA ligands of these sensory proteins have been well studied, a comprehensive and comparative understanding of their evolutionary and structural diversity across viral groups remains lacking.

In our project, we therefore aim to determine systematically the key features of both RNA sequence and RNA structure required for binding to these sensory proteins and for the subsequent activation of the innate immune response. We aim to make this connection actionable by developing a comprehensive predictive toolkit, RNAinnate, designed to predict the tempo and mode of activation of the innate immune system from the viral RNA sequence.

RNA-sensitive receptors and subsequent inflammatory response

In detail, we will first investigate the specificity of RNA-protein binding for each purified RNA sensor in vitro, using randomised synthesised RNA libraries and employing cross-linking and immunoprecipitation sequencing methods (CLIP-seq). We will then transfect cells with distinct RNA sequences to validate our findings on the purified RNA sensors using electroporation- and DharmaFECT-based transfection. We will establish a transfection protocol using human lung epithelial cells (A549 cells and Calu-3 cells) and subsequently apply this technique to human primary macrophages. We will thoroughly examine the innate immune response, in particular the subsequent phosphorylation of downstream kinases of the specific RNA-binding proteins in the lung epithelial cells and in the macrophages. Afterwards, we will infect lung epithelial cells and human macrophages with SARS-CoV-2 to investigate virus-specific RNA-protein binding. Ultimately, we will infect cells in parallel with intact SARS-CoV-2, influenza A virus (IAV), and respiratory syncytial virus (RSV) to map the previously identified triggering elements within their natural genomic contexts. In order to determine characteristics of binding RNAs, we will evaluate enrichment and depletion of features such as secondary structure elements and local sequence motifs, as well as assess the distribution of folding energies. Moreover, we will employ unsupervised clustering techniques on these features, and combine the results of the different methods to extract descriptors of binding patterns in the form of covariance models and Bayesian descriptors similar to Dimont. Besides the determination of the RNA sequences, we will analyse the transcriptomic, metabolomic, and lipidomic changes of human primary macrophages related to the specific RNA sensors that occur following distinct RNA recognition. Finally, we will uncover evidence of selection pressure that removes or attenuates the individual RNA trigger elements for the innate immune system, providing insights into the mechanisms of RNA genome evolution. To achieve the latter goal, scalable, high-quality alignments of viral genomes are required that can combine inter-species comparison with information of strain-level variations. As no such tool exists, we fill this gap with VirAligner, using novel combinations of existing approaches in comparative sequence analysis.

We aim to understand the sequence and structure requirements of viral RNAs that are necessary to trigger the innate immune system, and to disentangle the reasons for differences in the response. This overarching biological question requires that we tackle the following research tasks:

WP 1: Identification RNA features (Jordan/Stadler)
WP 2: Identification of triggering patterns to activate innate immune response (Jordan/Stadler)
WP 3: Triggering elements in viral genomes (Jordan/Stadler)
WP 4: Evolutionary Implications (Stadler)
WP 5: Identifying antiviral treatment strategies in the event of pandemic outbreak (Jordan)

Team Members

Dr. Paul Jordan

Project Leader

Prof. Dr. Peter Stadler

Project Leader

N. N.

Doctoral Researcher

N. N.

Doctoral Researcher

Dr. Maria Waldl

Associated Postdoctoral Researcher

Eva Laderick

Associated Office Manager

Project-Specific Publications

2026

Soltaninezhad, Mohammad; Rouzbahani, Yashar; Contreras, Jhonatan; Larios, Francisco Paez; Jordan, Paul M; Werz, Oliver; Chippalkatti, Rohan; Abankwa, Daniel Kwaku; Eggeling, Christian; Bocklitz, Thomas

Lightweight CycleGAN models for cross-modality image transformation and experimental quality assessment in fluorescence microscopy Journal Article

In: Biomed Opt Express, vol. 17, no. 3, pp. 1476–1498, 2026, ISSN: 2156-7085.

Abstract | Links | BibTeX

@article{pmid41970592,

title = {Lightweight CycleGAN models for cross-modality image transformation and experimental quality assessment in fluorescence microscopy},

author = {Mohammad Soltaninezhad and Yashar Rouzbahani and Jhonatan Contreras and Francisco Paez Larios and Paul M Jordan and Oliver Werz and Rohan Chippalkatti and Daniel Kwaku Abankwa and Christian Eggeling and Thomas Bocklitz},

doi = {10.1364/BOE.578297},

issn = {2156-7085},

year  = {2026},

date = {2026-03-01},

urldate = {2026-03-01},

journal = {Biomed Opt Express},

volume = {17},

number = {3},

pages = {1476--1498},

abstract = {With the growing integration of artificial intelligence in scientific and medical applications, lightweight deep learning models have become increasingly important. These models offer substantial reductions in memory usage and computational time. Given that GPU-based model training and inference contribute significantly to carbon emissions, lightweight architectures with comparable performance to parameter-rich models present a more environmentally friendly alternative. Specifically, we build upon CycleGAN with a fixed-channel lightweight U-Net generator for modality transfer from standard confocal to super-resolution STED and deconvolved STED images, and systematically compare it against Pix2Pix and standard CycleGAN baselines. Obtaining paired datasets in medical imaging and super-resolution microscopy is often infeasible due to the need for additional experiments and the intrinsic complexity of biological sample preparation. To address this, we investigate the performance of lightweight CycleGAN models, demonstrating their ability to achieve high-fidelity modality transfer despite reduced model complexity. We introduce a fixed channel strategy within the U-Net-based generator, in contrast to the traditional channel-doubling approach. This modification significantly reduces the number of trainable parameters from 41.8 million to approximately 9 thousand, while achieving comparable or slightly improved performance. We explore the utility of GAN models as a qualitative marker for assessing experimental and labeling quality. When trained on high-quality microscopy images, the GAN implicitly learns the characteristics of optimal imaging. Deviations between GAN-generated outputs trained on high-quality data and low-quality experimental images can highlight potential issues such as photobleaching, experimental artifacts, or inaccurate labeling. In this way, the model can support qualitative assessment of experimental consistency and image fidelity in fluorescence microscopy workflows.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

With the growing integration of artificial intelligence in scientific and medical applications, lightweight deep learning models have become increasingly important. These models offer substantial reductions in memory usage and computational time. Given that GPU-based model training and inference contribute significantly to carbon emissions, lightweight architectures with comparable performance to parameter-rich models present a more environmentally friendly alternative. Specifically, we build upon CycleGAN with a fixed-channel lightweight U-Net generator for modality transfer from standard confocal to super-resolution STED and deconvolved STED images, and systematically compare it against Pix2Pix and standard CycleGAN baselines. Obtaining paired datasets in medical imaging and super-resolution microscopy is often infeasible due to the need for additional experiments and the intrinsic complexity of biological sample preparation. To address this, we investigate the performance of lightweight CycleGAN models, demonstrating their ability to achieve high-fidelity modality transfer despite reduced model complexity. We introduce a fixed channel strategy within the U-Net-based generator, in contrast to the traditional channel-doubling approach. This modification significantly reduces the number of trainable parameters from 41.8 million to approximately 9 thousand, while achieving comparable or slightly improved performance. We explore the utility of GAN models as a qualitative marker for assessing experimental and labeling quality. When trained on high-quality microscopy images, the GAN implicitly learns the characteristics of optimal imaging. Deviations between GAN-generated outputs trained on high-quality data and low-quality experimental images can highlight potential issues such as photobleaching, experimental artifacts, or inaccurate labeling. In this way, the model can support qualitative assessment of experimental consistency and image fidelity in fluorescence microscopy workflows.

Schultz, Constanze; Jordan, Paul Mike; Dahlke, Philipp; Maz, Zehra Tuğçe Gür; Banoğlu, Erden; Meyer-Zedler, Tobias; Schmitt, Michael; Werz, Oliver; Popp, Juergen

Unveiling the molecular dynamics of a nitrile-containing 5-lipoxygenase-activating protein antagonist in primary macrophages through Raman spectroscopy Journal Article

In: Chem Sci, vol. 17, no. 3, pp. 1613–1623, 2026, ISSN: 2041-6520.

Abstract | Links | BibTeX

2025

Klepsch, Lea C; Dahlke, Philipp; Behnke, Mira; Tsarenko, Ekaterina; Göppert, Natalie E; Klemm, Paul; Meyer, Jakob; George, Alan; Chi, Mingzhe; Czaplewska, Justyna A; Vollrath, Antje; Weber, Christine; Jordan, Paul M; Schubert, Stephanie; Hoeppener, Stephanie; Nischang, Ivo; Sierka, Marek; Werz, Oliver; Schubert, Ulrich S

Matching drug and polymer for efficient delivery of anti-inflammatory drugs: PLGA, polyesteramides, and acetalated dextran Journal Article

In: J Mater Chem B, vol. 13, no. 48, pp. 15516–15529, 2025, ISSN: 2050-7518.

Abstract | Links | BibTeX

@article{pmid41251545,

title = {Matching drug and polymer for efficient delivery of anti-inflammatory drugs: PLGA, polyesteramides, and acetalated dextran},

author = {Lea C Klepsch and Philipp Dahlke and Mira Behnke and Ekaterina Tsarenko and Natalie E Göppert and Paul Klemm and Jakob Meyer and Alan George and Mingzhe Chi and Justyna A Czaplewska and Antje Vollrath and Christine Weber and Paul M Jordan and Stephanie Schubert and Stephanie Hoeppener and Ivo Nischang and Marek Sierka and Oliver Werz and Ulrich S Schubert},

doi = {10.1039/d5tb01949d},

issn = {2050-7518},

year  = {2025},

date = {2025-12-01},

urldate = {2025-12-01},

journal = {J Mater Chem B},

volume = {13},

number = {48},

pages = {15516--15529},

abstract = {The hydrochalcone derivative MF-15 and the synthetically derived BRP-201 are potent anti-inflammatory active pharmaceutical ingredients (APIs) that suffer from poor bioavailability. This necessitates their incorporation into drug delivery systems. To address this limitation, we investigated four polymeric carrier materials. The poly(ester amide)s poly(3-benzylmorpholine-2,5-dione) (PPheG) and poly(3-isopropyl-morpholine-2,5-dione) (PValG), the benchmark poly(lactic--glycolic acid) (PLGA), and the polysaccharide acetalated dextran (Ac-Dex) were used to formulate nanoparticles  nanoprecipitation. The nanoparticles had sizes of around 110 to 190 nm with negative zeta potentials. Although atomistic molecular dynamics (MD) simulations predicted enhanced miscibility of PPheG and PValG with MF-15, the highest loading capacity was achieved with Ac-Dex (4.2 wt%). None of the MF-15-loaded particles elicited a biologic response (, 15-lipoxygenase (LOX)-1 activation) in human M2 monocyte-derived macrophages (MDMs). The consistent failure across all MF-15 formulations, despite differences in polymer composition, drug loading, and enzymatic degradation profiles, suggests that encapsulation inherently interferes with MF-15's ability to activate 15-LOX-1, irrespective of the carrier system. In contrast, all BRP-201-loaded formulations demonstrated potent anti-inflammatory effects in human neutrophils. Overall, our findings demonstrate that polymer-drug miscibility and favorable physicochemical properties alone are insufficient to predict  efficacy, highlighting the importance of kinetic and formulation-dependent factors in the successful delivery of anti-inflammatory agents.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

Hornung, Franziska; SureshKumar, Harini K.; Klement, Laura; Reisser, Yasmina; Wernike, Christoph; Nischang, Vivien; Jordan, Paul M.; Werz, Oliver; Hoffmann, Carsten; Löffler, Bettina; Deinhardt-Emmer, Stefanie

High-fat diet impairs microbial metabolite production and aggravates influenza A infection Journal Article

In: Cell Communication and Signaling, vol. 23, no. 1, pp. 359, 2025, ISSN: 1478-811X.

Links | BibTeX

2024

Löhneysen, Sarah; Spicher, Thomas; Varenyk, Yuliia; Yao, Hua-Ting; Lorenz, Ronny; Hofacker, Ivo; Stadler, Peter F

Phylogenetic and chemical probing information as soft constraints in RNA secondary structure prediction. Journal Article

In: Journal of computational biology: a journal of computational molecular cell biology, vol. 31, iss. 6, no. 6, pp. 549-563, 2024.

Links | BibTeX

Günther, Kerstin; Ehrhardt, Christina; Werz, Oliver; Jordan, Paul M

Protocol for lipid mediator profiling and phenotyping of human M1-and M2-monocyte-derived macrophages during host-pathogen interactions Journal Article

In: STAR protocols, vol. 5, no. 3, pp. 103142, 2024.

Links | BibTeX

Jordan, Paul M; Günther, Kerstin; Nischang, Vivien; Ning, Yuping; Deinhardt-Emmer, Stefanie; Ehrhardt, Christina; Werz, Oliver

Influenza A virus selectively elevates prostaglandin E2 formation in pro-resolving macrophages Journal Article

In: iScience, vol. 27, iss. 1, no. 1, pp. 108775, 2024.

Links | BibTeX

2023

Klapproth, Christopher; Zötzsche, Siegfried; Kühnl, Felix; Fallmann, Jörg; Stadler, Peter F; Findeiß, Sven

Tailored machine learning models for functional RNA detection in genome-wide screens. Journal Article

In: NAR Genom Bioinform, vol. 5, iss. 3, no. 3, pp. lqad072, 2023.

Links | BibTeX

2022

Jordan, Paul M; Werz, Oliver

Specialized pro-resolving mediators: Biosynthesis and biological role in bacterial infections Journal Article

In: The FEBS journal, vol. 289, no. 14, pp. 4212–4227, 2022.

Links | BibTeX

2020

Jordan, Paul M; Gerstmeier, Jana; Pace, Simona; Bilancia, Rossella; Rao, Zhigang; Börner, Friedemann; Miek, Laura; Gutiérrez-Gutiérrez, Óscar; Arakandy, Vandana; Rossi, Antonietta; Ialenti, Armando; González-Estévez, Cristina; Löffler, Bettina; Tuchscherr, Lorena; Serhan, Charles N.; Werz, Oliver

Staphylococcus aureus-derived α-hemolysin evokes generation of specialized pro-resolving mediators promoting inflammation resolution Journal Article

In: Cell Rep, vol. 33, no. 2, 2020.

Links | BibTeX

2011

Lorenz, Ronny; Bernhart, Stephan H; Siederdissen, Christian Höner Zu; Tafer, Hakim; Flamm, Christoph; Stadler, Peter F; Hofacker, Ivo L

ViennaRNA Package 2.0 Journal Article

In: Alg Mol Biol, vol. 6, pp. 26, 2011.

Links | BibTeX

2004

Thurner, Caroline; Witwer, Christine; Hofacker, Ivo; Stadler, Peter F.

Conserved RNA secondary structures in Flaviviridae genomes Journal Article

In: J Gen Virol, vol. 85, pp. 1113-1124, 2004.

Links | BibTeX

1998

Hofacker, I L; Fekete, M; Flamm, C; Huynen, M A; Rauscher, S; Stolorz, P E; Stadler, Peter F.

Automatic detection of conserved RNA structure elements in complete RNA virus genomes Journal Article

In: Nucleic Acids Res, vol. 26, iss. 16, pp. 3825–3836, 1998, (Santa Fe Institute Preprint 98-03-020).

Links | BibTeX