Research | Hur lab

Our Research

Our lab investigates two fundamental aspects of the vertebrate immune system.

First, we study how the innate immune system discriminates between self and non-self based on RNA sequence, structure, modifications, and other physicochemical properties. We focus on receptors that recognize viral double-stranded RNAs—such as RIG-I-like receptors, PKR, OASes, and ZBP-1—as well as other host proteins, including TRIM family members, that directly or indirectly detect and restrict viral RNAs.

Second, we explore gene regulatory mechanisms in both innate and adaptive immunity. We are particularly interested in transcriptional regulators that recognize and silence foreign genetic elements—including viral genes, endogenous retroviruses, and other repetitive or parasitic DNA elements—and those that promote self-tolerance during T cell development. These mechanisms form a fundamental basis for immune homeostasis.

Together, our work has broad implications for advancing the understanding and treatment of viral infections and autoimmune diseases.

Innate Immunity

Discrimination between self and non-self nucleic acids is a central feature of pathogen detection by the innate immune system. Unlike bacterial systems such as CRISPR, which directly sense foreign nucleic acid sequences, vertebrate innate immune receptors rely on alternative features to distinguish self from non-self. Our structural and biochemical studies of RIG-I-like receptors (RLRs)—a key family of RNA sensors—revealed that receptor polymerization and clustering underlie the detection of viral RNA and subsequent immune activation (Figure 1).

We discovered that RLRs form filaments on RNA, enabling the receptors to recognize a range of viral and host RNA features—such as secondary structure and chemical modifications—and to integrate this information for effective nucleic acid discrimination. By reconstituting the signaling complex with purified components, we determined structural snapshots of RLRs, RNAs and cofactors throughout their activation processes (Figure 1). We also demonstrated that specific mutations in RLRs and their regulators can lower the immunological threshold for self-tolerance, resulting in aberrant activation by self-RNAs and contributing to autoimmune conditions such as lupus.

Building on this framework, our current research expands to additional RNA sensors—including PKR, OASes/RNase L, and TRIM proteins—to define shared and distinct principles of viral RNA recognition and immune activation. We are particularly interested in how these sensors maintain tolerance to self RNA, and how this tolerance can be disrupted under pathological or stress conditions. For example, we aim to identify cellular states that lead to the accumulation of self-derived “foreign” RNAs capable of aberrantly triggering the innate immune response.

Fig 1. Our findings on RLRs (A) Schematic of the RLR signalosome assembly. RLRs form filaments upon cognate viral dsRNA binding. RLR filaments then cluster upon interaction with the E3 ligases RIPLET and TRIM65. In this clustered environment, the signaling domains (CARDs) of RLR form a homo-tetramer, which then nucleates filament formation of MAVS. The MAVS filament functions as a signaling platform to recruit further downstream signaling molecules (e.g. TRAFs) and to activate antiviral immune responses. (B) Each step of receptor multimerization functions as a checkpoint to filter out self RNAs and to ensure that only non-self RNAs activate MAVS. Once the MAVS filament is nucleated, antiviral signal amplifies through MAVS filament propagation.

Adaptive Immunity

Our laboratory investigates transcriptional regulatory mechanisms that ensure maintenance of T cell self-tolerance and immune homeostasis.

FoxP3 is a well-conserved forkhead transcription factor (TF) that plays a critical role in development and function of regulatory T cells (Tregs), thereby maintaining immune homeostasis. While FoxP3 has long been assumed to function by recognizing the canonical forkhead consensus motif (TGTTTAC), this sequence is minimally enriched at FoxP3-bound genomic loci. In a striking departure from the classical model of transcription factor–DNA interaction, we discovered that FoxP3 instead recognizes microsatellites containing tandem TnG repeats (n = 2–5) by assembling into higher-order multimers. These multimers bridge 2–4 helices of double-stranded DNA, thereby stabilizing local chromatin loops and reinforcing higher-order chromatin architecture (Figure 2). This mechanism not only challenges conventional paradigms of transcription factor binding and activity, but also reveals an unexpected regulatory function for microsatellites—previously considered non-functional genomic elements. Our ongoing work focuses on further elucidating FoxP3 function by identifying co-factors recruited to these multimeric assemblies and visualizing the complexes using super-resolution imaging and genomic approaches. We are also investigating whether other transcription factors exploit microsatellites as scaffolds for gene regulation.

Fig 2. Our findings on FoxP3 (A) Structure of the FoxP3 multimer bound to microsatellite DNA containing T3G repeats. (B) Proposed model: FoxP3 shapes global chromatin architecture by assembling into higher-order multimers on microsatellites, bridging distal DNA elements and stabilizing chromatin loops.

Fig 3. Our findings on Aire (A) Aire forms transcriptional hubs as pre-existing enhancers (e.g. RIC8A and SETD1B loci) to activate their transcription. These functional hubs are distinct from the chromatin-detached condensates formed by the loss-of-function mutant ∆PHD1, which are transcriptionally inactive. (B) Proposed model: Aire’s PHD1 domain transiently interacts with the ubiquitous histone mark H3K4me0, keeping Aire in a diluted, monomeric state and suppressing premature polymerization. This regulation ensures that Aire polymerizes only upon proper recruitment to specific enhancer regions, enabling productive hub formation

Aire is a key transcription regulator for T cell central tolerance by inducing ectopic expression of thousands of tissue-specific antigens (TSAs) in medullary thymic epithelial cells. This process exposes developing thymocytes to a wide array of “self” antigens, enabling negative selection of autoreactive T cells or their diversion into Tregs. Although Aire has long been known to form large nuclear condensates, the function of these structures and the mechanisms driving their assembly have remained unclear. We discovered that Aire assembles into filamentous polymers that nucleate transcriptional hubs by consolidating multiple pre-existing enhancers—for example, at the RIC8A or SETD1B loci (Figure 3A). These hubs amplify transcriptional activity at bound enhancers and simultaneously activate distal TSA genes that are otherwise silent. Importantly, this process requires tight regulation: Aire polymerization must be suppressed until it encounters the appropriate chromatin environment. When this regulatory mechanism is disrupted—such as by deleting the PHD1 domain (Figure 3A)—Aire forms spontaneous, chromatin-unlinked polymers that generate transcriptionally inactive nuclear foci. Currently, we are investigating how a limited number of Aire hubs per cell coordinate transcriptional activation across thousands of TSA loci, and the biochemical principles that govern these long-range interactions. Moving forward, we aim to define the ultrastructural organization and dynamic behavior of Aire transcriptional hubs using cryo-ET and advanced imaging approaches.

Gene Regulation in Immunity

In addition to the projects described above, we are broadly interested in gene regulatory mechanisms involving both RNA and DNA, with the focus on understanding how the host identifies foreign genetic elements, such as endogenous retroviruses and other repetitive and parasitic DNA, both at the transcriptional and post-transcriptional levels.

Last update, July 2025