LUBAC and NF‐κB trigger a nuclear response from mitochondria

Plain language summary

Mitochondria are key signaling hubs for innate immune responses. In this issue, Wu et al (2022) report that remodeling of the outer mitochondrial membrane by the linear ubiquiting chain assembly complex (LUBAC) facilitates transport of activated NF‐κB to the nucleus in response to TNF signaling.

Mitochondria and linear ubiquitination take prime position in promoting nuclear translocation of activated NF‐kB.

The NF‐κB signaling pathway is a key component of innate immune responses. NF‐κB activation regulates different innate immune signaling pathways in response to different stimuli including pathogen‐associated molecular patterns (PAMPs) or TNFα. The recognition of PAMPs by TLRs or RIG‐I‐like receptors triggers the formation of multimeric protein complexes that converges on the mitochondrial transmembrane protein MAVS to regulate downstream NF‐κB signaling that induces interferons and proteins involved in antigen presentation. Another NF‐κB pathway depends on TNF, a cytokine that upon binding to its receptor (TNFR) at the plasma membrane promotes cell survival through the transcriptional upregulation of anti‐apoptotic NF‐κB target genes. As a result, apoptotic or necroptotic cell death upon TNF stimulation can only proceed when the NF‐κB prosurvival transcriptional response is blocked or caspase‐8‐mediated inhibition of necroptosis is absent (Peltzer & Walczak, 2019). However, TNF downstream effects can have cell type‐ and context‐dependent readouts, which are particularly evident in the central nervous system (CNS).

Binding of TNF to TNFR leads to the formation of a complex composed of TRADD, RIPK, TRAF2/5, and the E3 ligase cIAP1/2 (primary complex I in Fig 1). cIAP1/2 forms K63‐linked ubiquitin chains, which prompts the linear ubiquitin chain assembly complex (LUBAC) to complex I to form M1‐linked ubiquitin chains. Linear ubiquitin chain assembly complex is a multimeric protein complex composed of hem‐oxidized IRP2 ubiquitin ligase‐1 (HOIL‐1), HOIL‐1 interacting protein (HOIP), and the Shank‐associated RH domain‐interacting protein (SHARPIN) (Emmerich et al, 2013). LUBAC‐mediated M1 ubiquitination can be reversed by the OTULIN deubiquitinase (Rivkin et al, 2013). A crucial step in this pathway is the binding of the adaptor protein NEMO to M1‐linked ubiquitin chains and its own modification by LUBAC. NEMO is the core regulatory component of the NF‐κB inhibitor (I‐κB) kinase (IKK) complex. To allow for canonical NF‐κB activation, modification of NEMO with M1‐linked ubiquitin causes a conformational change, resulting in the activation of the kinases IKKα and IKKβ. These kinases phosphorylate I‐κB leading to its degradation by the proteasome to enable the translocation of NF‐κB heterodimers to the nucleus (Fujita et al, 2014). Linear ubiquitin chain assembly complex therefore acts as a control station and enhancer of TNF signaling.

Figure 1. Parkin and LUBAC‐mediated ubiquitination determine starvation and TNFR‐mediated readouts from their mitochondrial localization.

While Parkin‐mediated K63 ubiquitination activates autophagic elimination of mitochondria, LUBAC‐mediated M1 ubiquitination of the mitochondrial environment triggers protective NF‐κB signaling and mitochondria–nucleus tethering. The role of tethering proteins such as TSPO or mitofusin‐2 in this spatial arrangement during TNFR signaling remains to be determined.

Interestingly, like MAVS‐dependent antiviral signaling, LUBAC‐mediated control of NF‐κB signaling also converges on mitochondria. Here, LUBAC stabilizes the PTEN‐induced kinase 1 (PINK1). PTEN‐induced kinase 1 phosphorylates M1‐linked ubiquitin chains, which increases their stability as OTULIN cannot efficiently hydrolyze phosphorylated linear ubiquitin chains. Moreover, PINK1 phosphorylates the ubiquitin‐like domain of Parkin to activate this mitochondrial K63 ubiquitin ligase (Iguchi et al, 2013). If associated with mitochondrial depolarization, Parkin results in K63‐, K48‐, K6‐, and K11‐linked mitochondrial ubiquitination and mitophagy (Ordureau et al, 2018), while in the presence of TNF, Parkin ubiquitinates NEMO, which triggers M1‐ubiquitination of NEMO by LUBAC (Muller‐Rischart et al, 2013). Linear ubiquitin chain assembly complex therefore closely integrates the consequences of M1 and K63 ubiquitination on mitochondrial substrates.

Wu et al (2022) have now significantly expanded the spatial relationships and consequences of LUBAC and OTULIN‐controlled M1 ubiquitination. They show that from their location to the outer mitochondrial membrane, both regulators of NF‐κB signaling components protect mitochondria from apoptosis, but also control NF‐κB signal progression in the early phases of TNF signaling, before the induction of the NF‐κB transcriptional response. An important question is why cells employ mitochondria as a platform for NF‐κB signaling. Wu et al (2022) proposed a hypothesis that mitochondria might equip NF‐κB signaling with a mobile platform that interacts with other organelles, including the nucleus. For these interactions, mitochondria typically use interorganellar tethers, as well as mitochondrial motor proteins to form temporary or stable membrane contact sites (MCS) with the endoplasmic reticulum (ER), lysosomes or the nucleus. The changing composition and extent of these MCS allows mitochondria to achieve distinct goals. For instance, the plastic nature of mitochondria‐ER contacts (MERCs) responds to Ca²⁺ flux that arrests mitochondrial motor proteins and to ROS that oxidizes tethered proteins with the goal to optimize bioenergetics or trigger apoptosis. However, less information is available about the plasticity of mitochondria–nucleus contacts, although these contacts are expected to be important for the mitochondrial retrograde response (MRR). Previously, the translocator protein TSPO had been identified as a requirement for the MRR, following the activation of ROS and Ca²⁺‐mediated signaling (Desai et al, 2020). Another regulatory protein of mitochondria–nucleus communication is mitofusin‐2. During cell proliferation, the mitochondrial tethering regulator mitofusin‐2 promotes nuclear–mitochondria contacts to shuttle the pyruvate dehydrogenase complex into the nucleus without the help of the nuclear pore (Zervopoulos et al, 2022). This function may be connected to NF‐κB, given Mfn2 promotes cytokine production and the activation of NF‐κB in macrophages (Tur et al, 2020). Consistent with a role of mitochondria–nucleus contacts for their findings, Wu et al (2022) indeed observe increased HOIP‐dependent tethering signals as measured with a fluorescent mitochondria–nucleus SPLICS probe upon TNFR signaling. This suggests that LUBAC‐controlled TNFR signaling activates mitochondria–nucleus tethers, perhaps TSPO, or mitofusin‐2. These functional links remain to be established.

Taken together, the findings by Wu et al (2022) establish LUBAC as a key control step for the switch from mitochondrial apoptotic to mitochondrial prosurvival signaling. This is associated with mitochondria–nucleus tethering and facilitated import of NF‐κB into the nucleus. Moreover, LUBAC‐mediated M1‐ubiquitination emerges as a signal for increased proximity between mitochondria and the nucleus and the activation of NF‐κB at mitochondria. Future research will have to address how LUBAC is recruited to mitochondria upon TNF stimulation and whether mitochondrial tether proteins are targets of LUBAC or OTULIN. These TNFR downstream signaling components may also lead to differential activation of PINK1 or NEMO to regulate mitochondria–nucleus tethering. However, the relative, time‐dependent role of other signaling pathways, including ROS and Ca²⁺, in LUBAC‐OTULIN‐mediated control of NF‐κB signaling remains to be investigated.