Molecular insights into NLRP1 Gly/N-degron recognition by CRL2^ZER1

We present the ZER1–NLRP1 Gly/N-degron structure, revealing a conserved binding cavity for N-terminal glycine recognition. ZER1 binds more weakly than ZYG11B due to distinct interactions at position 3, providing mechanistic insights into NLRP1 inflammasome regulation.

Abstract

The Gly/N-degron pathway is a branch of the proteasomal degradation pathway that specifically targets proteins initiated with an N-terminal glycine. The E3 ligase substrate adaptors ZYG11B and ZER1 have been identified as being responsible for recognizing the target proteins of the Gly/N-degron pathway. Previously, it has been shown that the Gly/N-degron pathway activates the human NLRP1 inflammasome by degrading the autoinhibitory N-terminal fragment of NLRP1 after cleavage by the enteroviral 3C protease. However, the recognition of the NLRP1 Gly/N-degron is not yet fully understood. Here, we determined the X-ray crystal structure of ZER1 bound to the NLRP1 Gly/N-degron at a resolution of 2.2 Å. The structural information revealed that ZER1 uses its ARM repeats to form a conserved cavity that engages the N-terminal glycine (G1) through hydrogen bonds to Asp556, Asn597 and Glu600. Structural comparisons show a shared recognition mode for Gly/N-degrons despite subtle differences in side-chain interactions. However, ZER1 exhibits weaker affinity for the NLRP1 Gly/N-degron than ZYG11B, likely due to distinct local environments surrounding position 3. This study elucidates the molecular basis of NLRP1 recognition by ZER1 and provides insights into targeting this pathway in inflammatory diseases.

1. Introduction

The N-degron pathway is a ubiquitin-dependent proteolytic system that selectively targets proteins for degradation based on their destabilizing N-terminal residues (N-degrons; Timms & Koren, 2020 ▸; Varshavsky, 2024 ▸). To date, at least five branches of the N-degron pathway have been described in eukaryotes, including the Arg/N-degron pathway, the Pro/N-degron pathway, the fMet/N-degron pathway, the Ac/N-degron pathway and the Gly/N-degron pathway (Varshavsky, 2019 ▸, 2024 ▸). As key degradation signals, the N-degrons of target proteins can be recognized by specific E3 ubiquitin ligases and subsequently targeted for degradation through the ubiquitin (Ub)–proteasome system (UPS) pathway (Varshavsky, 2019 ▸). The molecular-recognition mechanisms of the Arg/N-degron, Pro/N-degron and Gly/N-degron pathways have been elucidated through structural and biochemical studies, providing important insights into how N-terminal residues are decoded by dedicated E3 ligases (Pan et al., 2021 ▸; Choi et al., 2010 ▸; Dong et al., 2018 ▸, 2020 ▸; Yan et al., 2021 ▸; Li et al., 2022 ▸; Zhang et al., 2022 ▸; Liu et al., 2026 ▸; Lin et al., 2022 ▸). The N-degron pathway, by regulation of protein stability, plays a role in numerous biological processes, including stress sensing, chromosome segregation, spermatogenesis and immune responses (Dissmeyer, 2019 ▸; An et al., 2012 ▸; Yi et al., 2024 ▸; Jung et al., 2025 ▸). Disruptions of the N-degron pathway can result in various diseases, such as neurodegenerative diseases, epilepsy and cancer (Tasaki & Kwon, 2007 ▸; Eldeeb & Ragheb, 2020 ▸; Eldeeb et al., 2022 ▸).

As the most recently identified branch of the N-degron family, the Gly/N-degron pathway was first characterized in 2019 (Chui et al., 2019 ▸). It specifically targets proteins with unmodified N-terminal glycine residues for ubiquitin–proteasomal degradation (Robinson et al., 2020 ▸; Timms et al., 2019 ▸). Due to evolutionary pressures, naked N-terminal glycine motifs are naturally absent from metazoan proteosomes (Chui et al., 2019 ▸; Robinson et al., 2020 ▸; Timms et al., 2019 ▸). However, a serendipitous discovery revealed the presence of such motifs in the genome of SARS-CoV-2 (betacoronavirus; Cesar Ramos de Jesus et al., 2024 ▸; Zhu et al., 2024 ▸). In metazoans, an N-myristoylation modification orchestrates this phenomenon, which can occur either co-translationally when methionine aminopeptidase (MetAP) acts on an N-terminal ‘MG’ motif at the ribosome, or post-translationally at an internal site. N-Myristoyltransferase enzymes, such as NMT1 and NMT2 in human cells, perform this specific task of N-myristoylation modification, thereby establishing a surveillance mechanism (Thinon et al., 2014 ▸; Perinpanayagam et al., 2013 ▸).

NLRP1 (NACHT, LRR and PYD domains-containing protein 1) was one of the first inflammasome-forming PRRs (pattern-recognition receptors) to be identified in humans (Cesar Ramos de Jesus et al., 2024 ▸). It is highly expressed and functional in epithelial barrier tissues (Bauernfried et al., 2021 ▸). The Gly/N-degron pathway has been shown to play a crucial role in activation of the human NLRP1 inflammasome by degrading the autoinhibitory N-terminal fragment of NLRP1 following cleavage by enteroviral 3C protease (Robinson et al., 2020 ▸; Timms et al., 2019 ▸). ZYG11B and ZER1, as substrate receptors of the cullin 2-RING E3 Ub ligase (CRL2), are responsible for recognizing the 3C protease-cleaved autoinhibitory N-terminal fragment (Robinson et al., 2020 ▸; Timms et al., 2019 ▸). This sequence of events ultimately results in the activation of inflammasomes by the liberated NLRP1 C-terminal fragment, leading to the activation of caspase-1 and the secretion of inflammatory cytokines from infected airway epithelial cells (Robinson et al., 2020 ▸). Although the molecular details of NLRP1 Gly/N-degron recognition by CRL2^ZYG11B (PDB entry 9bj9) have recently been determined using cryo-EM (Liu et al., 2026 ▸), the molecular mechanism underlying CRL2^ZER1 recognition of the NLRP1 Gly/N-degron remains unknown. Here, we report the high-resolution crystal structure of CRL2^ZER1 in complex with the Gly/N-degron of NLRP1. Our structural analysis, together with isothermal titration calorimetry (ITC) assays, reveals the essential elements required for the specific recognition of the NLRP1-derived Gly/N-degron by CRL2^ZER1. This study provides further molecular insight into the Gly/N-degron by CRL2 E3 ligase and sheds light on the future development of small-molecule inhibitors of CRL2^ZER1.

2. Methods

2.1. Cloning, protein expression and purification

The coding sequences for human ZYG11B (amino acids 483–728) and ZER1 (amino acids 521–766) were PCR-amplified from cDNA templates, with an additional Ser-Glu-Arg (SER) sequence incorporated at their 5′ ends. These modified fragments were subsequently inserted into the pET-28-MKH8SUMO expression vector (Addgene plasmid #79526) via seamless cloning. This vector was specifically designed to carry an 8×His-SUMO tag, followed by a TEV protease recognition site (LYFQ↓G, where ↓ marks the cleavage site), facilitating proteolytic removal of the fusion tag and yielding an N-terminal glycine residue, which can function as an N-degron signal.

The recombinant plasmids were individually transformed into Escherichia coli BL21 (DE3) cells and kanamycin was used to select positive transformants. Cultures were grown in LB medium at 37°C until an OD₆₀₀ of 0.8 was reached. The incubation temperature was then reduced to 18°C, and protein expression was induced by the addition of 0.2 mM isopropyl β-d-1-thiogalactopyranoside. After 16 h of induction, the cells were collected via centrifugation (4000 rev min⁻¹, 8 min, 4°C), resuspended in lysis buffer (20 mM Tris–HCl pH 7.5, 400 mM NaCl, 2 mM β-mercaptoethanol) and disrupted by sonication on ice. The lysate was clarified by centrifugation (14 000 rev min⁻¹, 50 min, 4°C) and the supernatant was incubated with Ni–NTA affinity resin for 1 h at 4°C to bind His-tagged proteins. Nonspecifically bound proteins were removed using a wash buffer containing 25 mM imidazole, and the fusion protein was subsequently eluted with 300 mM imidazole in 20 mM Tris–HCl pH 7.5, 400 mM NaCl.

To cleave the 8×His-SUMO fusion tag, the eluate was dialyzed overnight at 4°C against a buffer (20 mM Tris–HCl pH 7.5, 300 mM NaCl) containing home-purified TEV protease in a 1:15 molar ratio. The digestion product was then passed again through an Ni–NTA column to remove the cleaved tag and any residual protease. Further purification was achieved through a two-step chromatography process: initially using anion exchange with a HiTrap Q HP column, followed by size-exclusion chromatography with a Superdex 200 Increase 10/300 GL column (GE Healthcare). The gel-filtration column was equilibrated with a buffer consisting of 20 mM Tris–HCl pH 7.5, 150 mM NaCl, 1 mM DTT. The final protein was concentrated and stored at −80°C. Site-directed mutants were generated using the QuikChange method (An et al., 2011 ▸), with the wild-type plasmid serving as the template. All mutations were confirmed through sequencing, and the mutant proteins were purified following the same protocol as the wild type.

2.2. Protein crystallization

Protein crystals were obtained using the sitting-drop vapor-diffusion technique at 4°C. Equal volumes (1 µl each) of protein and reservoir solutions were mixed to create crystallization drops in 48-well plates. For the GSER-tagged ZER1 fragment (amino acids 521–766), optimal crystal growth was achieved in a reservoir consisting of 8%(v/v) Tacsimate pH 6.0, 20%(w/v) polyethylene glycol 3350. Before collecting X-ray data, the crystals were transferred into a cryoprotectant solution composed of the reservoir buffer supplemented with 20–25%(v/v) glycerol and subsequently flash-cooled in liquid nitrogen.

2.3. Data collection and processing

X-ray diffraction data were collected on the ID02-Microfocusing X-Ray Protein Crystallography Beamline at the High Energy Photon Source (https://cstr.cn/31138.02.HEPS.ID02). The data were processed and scaled using the XDS software suite (Kabsch, 2010 ▸). Structural elucidation of ZER1 was carried out by molecular replacement using Phaser (McCoy et al., 2007 ▸), with PDB entry 7ep3 (Yan et al., 2021 ▸) serving as the search model. The structural model was refined through multiple cycles combining manual rebuilding in Coot (Emsley et al., 2010 ▸) with automated refinement procedures in phenix.refine (Afonine et al., 2012 ▸). All three-dimensional molecular visualizations were generated using PyMOL (https://www.pymol.org). The electrostatic potentials were calculated using Adaptive Poisson–Boltzmann Solver (APBS; Baker et al., 2001 ▸). The color scale for the surface potentials has been standardized to −5 to +5 kT/e.

2.4. Isothermal titration calorimetry

Isothermal titration calorimetry was carried out at 16°C utilizing a MicroCal PEAQ-ITC instrument (Malvern Panalytical). For this assay, the ZER1 gene fragment was cloned into the pET-28-MKH8SUMO vector, while the constructs encoding wild-type or mutant ZER1 (residues 469–766) were inserted into the pNIC-CH vector (Addgene #26117). Protein purification followed the same procedure as described above, except that removal of the 6×His-tag was unnecessary in this case. Proteins (final concentration 50 µM) and peptides (final concentration 1.0 mM) were prepared in a binding buffer composed of 20 mM Tris–HCl pH 7.5, 150 mM NaCl. Each titration experiment involved 300 µl protein solution loaded into the calorimetric cell, and sequential peptide injections were carried out beginning with an initial 0.5 µl dose, followed by 15 subsequent injections of 1.5 µl each spaced at 90 s intervals for thermal equilibration. Throughout the experiment, the reference power was set at 10 µcal s⁻¹. The raw data were analyzed using the MicroCal PEAQ-ITC software (version 1.30). Binding affinities and thermodynamic parameters were extracted by fitting the data to a one-site binding mode.

3. Results

3.1. The overall structure of NLRP1 peptide bound to ZER1

Previously, we demonstrated that ZER1/ZYG11B employ their C-terminal armadillo (ARM) repeats to recognize Gly/N-degrons (Yan et al., 2021 ▸). Utilizing ITC assays on purified ARM repeats (ARM3–ARM8 for ZER1 and ARM2–ARM8 for ZYG11B) from ZER1/ZYG11B, we discovered that both ZER1 and ZYG11B could bind to the NLRP1 Gly/N-degron (peptide sequence GSERRVL). ZYG11B exhibited a strong inter­action with the NLRP1 peptide, with a K_D of 4.0 µM (Fig. 1 ▸a), whereas ZER1 showed a relatively weaker binding affinity to the NLRP1 peptide, with a K_D of 40.5 µM (Fig. 1 ▸b). It is worth noting that ZER1 exhibits an apparent substoichiometric binding (n ≃ 0.6); this aligns with the weaker binding affinity of ZER1 to the NLRP1 peptide with a deviation in the fitting. We speculate that this might be due to the presence of some transient monomeric ZER1 populations during titration.

Figure 1.

ITC analysis of ZYG11B or ZER1 binding to the NLRP1 Gly/N-degron. (a) ITC titration and fitting curve of ZYG11B (residues 483–728) binding to the NLRP1 Gly/N-peptide (sequence GSERRVL; K_D = 4.0 µM). (b) ITC titration and fitting curve of ZER1 (residues 521–766) binding to the NLRP1 Gly/N-peptide (sequence GSERRVL; K_D = 40.5 µM).

To investigate the binding details of ZER1/ZYG11B with the NLRP1 Gly/N-degron, we generated His-SUMO fusion proteins by tandemly fusing a known Gly/N-degron tetrapeptide (sequence GSER), derived from the N-terminal fragment of NLRP1 cleaved by Tobacco etch virus protease (TEV protease), to the C-terminus of the SUMO tag (hereafter referred to as the Gly/N-degron; Yan et al., 2021 ▸). Meanwhile, the truncated ZER1/ZYG11B (residues 483–728 and residues 521–766 for ZYG11B and ZER1, respectively) was fused to the C-terminus of the NLRP1 Gly/N-degron tetrapeptide (Fig. 2 ▸a). Through cleavage by TEV protease, the N-terminal Gly/N-degron of NLRP1 can be exposed by the action of TEV protease on an engineered cleavage site (Fig. 2 ▸a). Subsequently, we conducted protein purification and crystal-growth experiments to attempt to obtain complex crystals and determine their structures.

Figure 2.

Crystal structure of ZER1 bound to the NLRP1 Gly/N-degron. (a) Schematic representation of the fused ZER1-Gly/N-degron tetrapeptide design. A tetrtapeptide Gly/N-degron (sequence GSER), derived from the 3C protease-cleaved autoinhibitory N-terminal fragment of NLRP1, was fused to the N-terminus of ZER1 and exposed by TEV protease cleavage. (b) Structural conformation of the NLRP1 Gly/N-degron within the ZER1-peptide complex structure. The peptide is depicted as cyan sticks. The 2F_o − F_c electron-density map (blue mesh) of the peptide is contoured at the 1.0σ level. (c) The composite omit map of the ZER1-peptide in the complex. The peptide is depicted as cyan sticks. The composite omit map (green mesh) of the peptide is contoured at the 1.0σ level. (d) Dimeric structure of the ZER1–NLRP1 Gly/N-degron complex. The N-terminus of ZER1 (molecule A; orange-pink) inserts into the substrate-recognition pocket of the opposing molecule (molecule B; cyan), and vice versa. (e) Superposition of two ZER1 molecules bound to NLRP1 peptide molecules. The two ZER1 molecules are depicted as orange-pink and cyan cylinders, respectively. Each NLRP1 Gly/N-degron peptide is shown as sticks, with the peptide coloration matching the corresponding ARM repeats (orange-pink or cyan).

Although the crystallization of ZYG11B bound to the NLRP1 Gly/N-degron failed, we successfully obtained the crystal structure of the ZER1–NLRP1 degron complex. The structure was determined at a resolution of 2.2 Å (Figs. 2 ▸b and 2 ▸c and Table 1 ▸). During the purification process of the fusion protein, we found that ZER1-NLRP1 degron complex eluted earlier than ZER1 in size-exclusion chromatography (Supplementary Fig. S1), indicating that the fusion protein exists as a dimer in solution. The crystal structure confirms our result as the ZER1-fused NLRP1 Gly/N-degron (sequence GSER) forms a dimer, with the N-terminus inserting into the substrate-recognition pocket of the symmetrical molecule (Fig. 2 ▸d). These two complex structures are almost identical, with a root-mean-square deviation (r.m.s.d.) of 0.156 Å over 216 C^α atoms (Fig. 2 ▸e). The overall structure of the NLRP1 peptide–ZER1 complex exhibits an arcuate conformation, with the NLRP1 peptide situated within the groove of the ARM repeats (ARM3–ARM8; Fig. 2 ▸e). This indicates that ZER1/ZYG11B adopts a similar conformation whether it binds to a GSER Gly/N-degron peptide or to other Gly/N-degron peptides such as GGFN or GFLH (Yan et al., 2021 ▸; Li et al., 2022 ▸).

Table 1. Data-collection and refinement statistics for ZER1–GSER (PDB entry 9uyb).

Values in parentheses are for the highest resolution shell.

Data collection
Space group	P212121
a, b, c (Å)	67.64, 68.20, 249.29
α, β, γ (°)	90.00, 90.00, 90.00
Resolution (Å)	48.03–2.20 (2.28–2.20)
Rsym or Rmerge	0.07254 (0.1883)
〈I/σ(I)〉	6.52 (3.42)
Completeness (%)	99.91 (99.74)
Multiplicity	2.0 (2.0)
CC1/2	0.982 (0.914)
Refinement
Resolution (Å)	48.03–2.20 (2.28–2.20)
No. of reflections	59562 (5850)
Rwork/Rfree	0.2026/0.2411
No. of atoms
Protein	7784
Water	463
Ligands	6
B factors (Å2)
Protein	37.96
Water	39.70
Ligands	36.72
R.m.s. deviations
Bond lengths (Å)	0.008
Bond angles (°)	0.97
Ramachandran statistics
Favored (%)	97.21
Allowed (%)	2.79
Outliers (%)	0.0

3.2. The recognition mechanism of the NLRP1 Gly/N-degron by ZER1

Structural comparison reveals that the complex structure of ZER1 bound to the NLRP1 peptide is analogous to that of other Gly/N-degrons (Yan et al., 2021 ▸). Overall, the N-terminal glycine (G1) of the NLRP1 peptide is deeply buried within the narrow cavity, fitting perfectly into the negatively charged pocket (Fig. 3 ▸a). The serine at position 2 (S2) aligns at the waist of the cavity, whereas the glutamate at position 3 (E3) and the arginine at position 4 (R4) are situated at the entrance to the cavity (Figs. 3 ▸a and 3 ▸b).

Figure 3.

Characterization of NLRP1 Gly/N-degron recognition by ZER1. (a) The NLRP1 degron-binding pocket, formed by ZER1, is depicted as an electrostatic potential surface. Electrostatic potentials were calculated by solving the Poisson–Boltzmann equation using the Adaptive Poisson–Boltzmann Solver (APBS). The electrostatic potential was mapped onto the molecular surface (probe radius = 1.4 Å) using PyMOL. The surface was colored according to the electrostatic potential in the range −5 to +5 kT/e, with red representing negative potential and blue representing positive potential. The degron is represented by cyan sticks. (b) Cross-sectional view of the GSER-binding pocket of ZER1. The degron is depicted as cyan sticks, while water molecules are depicted as green spheres. (c, d, e) Close-up views of the interactions between G1, S2 and E3 (the glutamine residue at position 3) of the NLRP1 Gly/N-degron and ZER1. The peptide is depicted as cyan sticks, the interacting residues in ZER1 are depicted as orange-pink sticks, water molecules are depicted as green spheres and hydrogen bonds are depicted as black dashed lines. (f) ITC fitting curves for the wild-type (WT) and mutant ZER1 (residues 521–766) titrated with the NLRP1 peptide (sequence GSERRVL). The corresponding mutant proteins and their binding affinities are indicated.

In addition to the well fitted binding cavity, the NLRP1 peptide was bound by ZER1 through several intermolecular interactions (Figs. 3 ▸c–3 ▸e). Firstly, the α-amino group of G1 is directly hydrogen-bonded to Asn597, Asp556 and indirectly to Glu600 and Asp556 via three water molecules, while the α-carbonyl group of G1 forms a hydrogen bond to Asn597 of ZER1 (Fig. 3 ▸c). These interactions are essentially consistent across all of the Gly/N-degrons we have previously reported. G1 is stabilized by five conserved hydrogen bonds in the narrow and deep cavity. The only difference is that there are more water interactions with G1 in the ZER1–NLRP1 degron complex. Secondly, serine at position 2 (S2) forms two main-chain hydrogen bonds with Asn679 of ZER1, while the hydroxyl side chain of S2 is hydrogen-bonded to a water molecule which also forms hydrogen bonds to Asn679 and Asp556 simultaneously (Fig. 3 ▸d). Additionally, the glutamate residue at position 3 (E3) utilizes its backbone amide to form a direct hydrogen bond to Asn553 (Fig. 3 ▸e), while the side-chain carboxyl group of E3 (the glutamate residue at position 3) is stabilized by two hydrogen bonds from Ser549 and Arg589. Finally, the entirety of R4 extends outwards and does not engage in any apparent interactions with the binding pocket (Fig. 3 ▸b). Together, the initial three residues of the NLRP1 peptide primarily utilize their main-chain amino and carboxyl groups to establish hydrogen bonds to ZER1, thereby stabilizing the binding. This stabilization is further supported by multiple water molecules within the binding pocket.

To determine the binding role of key residues of ZER1 in mediating Gly/N-degron binding, we generated a series of single-point mutants of ZER1 and examined their binding affinities for the NLRP1 peptide using ITC (Fig. 3 ▸f). In line with our structural analysis, the substitution of Asp556, Asn597 or Glu600 with alanine, which would disrupt hydrogen bonds to the α-amino group of G1, resulted in a complete loss of Gly/N-degron binding (Fig. 3 ▸f). These results confirm the critical role of hydrogen bonding in G1 recognition. Likewise, the replacement of Asn553 with alanine, which disrupts the hydrogen bond to the backbone amide of E3 (the glutamate at position 3), exhibited negligible interaction with the Gly/N-degron (Fig. 3 ▸f). These results confirmed our findings from structural studies and demonstrated the critical role of the key residues in ZER1-mediated Gly/N-degron recognition.

3.3. Residue preference of peptide position 3 in ZER1 binding

Since the affinity of the NLRP1 peptide for ZER1 is relatively weaker (sequence GSER; K_D of 40.5 µM) compared with that of the ZNF701 peptide (sequence GFLH; K_D of 3.8 µM) and the GIMAP5 peptide (sequence GGFQ; K_D of 4.8 µM) (Yan et al., 2021 ▸), we compared the binding modes of ZER1 in complex with the NLRP1 Gly/N-degron and the ZNF701 Gly/N-degron (Fig. 4 ▸a). Superposition of these two complexes shows that the conformation and mode of interaction are highly similar, with an r.m.s.d. of 0.323 Å over 214 C^α atoms, although there is a slight wobble in the main chain of peptide position 3 and position 4 (Fig. 4 ▸a). Meanwhile, ZER1 uses a largely identical set of key residues for Gly/N-degron recognition and stabilization. We focused our attention on the significant angle deviation of the side chains of E3 (the glutamate residue at position 3) and R4 in the NLRP1 Gly/N-degron compared with L3 (the leucine residue at position 3) and H4 (the histidine residue at position 4) in the ZNF701 Gly/N-degron. We first excluded an influence of the fourth residue in the substrate peptide on ZER1 affinity, since R4 is completely exposed to the solvent at the exterior of the cavity and thus contributes little to NLRP1 Gly/N-degron recognition (Figs. 3 ▸b and 4 ▸a) and our previous results showed that substitution of Q4 (the glutamine residue at position 4) by arginine in the GIMAP5 peptide showed almost no attenuation of ZER1 binding (Yan et al., 2021 ▸). Previously, our ITC assays have shown that position 3 of ZER1 shows an apparent preference for hydrophobic residues (such as phenylalanine, tyrosine, tryptophan and proline) and a particular disfavor for charged residues such as glutamate, aspartate and arginine. However, ZYG11B can tolerate many types of residues at position 3 of the substrate (Yan et al., 2021 ▸; Li et al., 2022 ▸). Comparing the electrostatic potentials of ZER1 and ZYG11B reveals a distinct difference in the local electrostatic environment around position 3 of the bound peptide, which likely accounts for their divergent residue preferences. Marked differences in the electrostatic properties of the residues flanking the binding pocket are clearly observed (Figs. 4 ▸b and 4 ▸c). Specifically, ZER1 harbors two negatively charged residues near position 3, Glu545 and Glu586, which generate a localized negative electrostatic environment that is unfavorable for accommodating negatively charged side chains such as glutamate (Fig. 4 ▸b). In contrast, ZYG11B features Lys515 and Ser556 at the corresponding positions, creating a more positive and complementary environment that is suitable for accommodating a negatively charged residue at position 3 (Fig. 4 ▸c). This distinct divergence in the local electrostatic environment explains why ZER1 exhibits low tolerance for an acidic residue at position 3, whereas ZYG11B possesses broader substrate tolerance. To validate this explanation, we introduced the E586A mutation in ZER1 and evaluated its effect on binding affinity using ITC (Fig. 4 ▸d). While wild-type ZER1 exhibited a K_D of 40.5 µM, the E586A mutant showed an approximately twofold increase in binding affinity, with a K_D of 20.2 µM. This suggests that Glu586 exerts electrostatic repulsion against the position 3 glutamate side chain, and that relieving this repulsion enhances substrate binding.

Figure 4.

Substrate-binding pockets of ZER1 and ZYG11B. (a) Left: structural superposition of ZER1–NLRP1 and ZER1–ZNF701 complexes. ZER1 in ZER1–NLRP1 and ZER1–ZNF701 is shown as orange-pink and gray ribbons, respectively. Right: close-up view of the degron-binding cavity. The key residues of ZER1 are shown as orange-pink and gray sticks. The NLRP1 peptide is shown as cyan sticks, the ZNF701 peptide is shown as yellow sticks and hydrogen bonds are shown as black dashed lines. The water molecule and critical residues in the ZER1–NLRP1 and ZER1–ZNF701 complexes are shown in orange-pink and gray, respectively. (b) Close-up view of the ZER1 binding pocket accommodating the NLRP1 peptide, highlighting the local electrostatic environment surrounding position 3 of the bound peptide. The two negatively charged key residues of ZER1 near position 3, Glu545 and Glu586, are labeled. (c) Close-up view of the modeled ZYG11B binding pocket accommodating the NLRP1 peptide, highlighting the local electrostatic environment surrounding position 3 of the bound peptide. Lys515 and Ser556, the two key residues of ZYG11B near position 3 corresponding to those in ZER1, are labeled. (d) ITC titration and fitting curve of ZER1 (E586A) binding to the NLRP1 Gly/N-peptide (sequence GSERRVL; K_D = 20.2 µM).

4. Discussion

This study clarifies the molecular basis of NLRP1 Gly/N-degron recognition by CRL2^ZER1, thereby enhancing our comprehension of the Gly/N-degron pathway. Consistent with previous research, ZER1 employs its ARM repeats to create a conserved cavity that binds to the N-terminal glycine (G1) of NLRP1 via crucial hydrogen bonds to residues Asp556, Asn597 and Glu600. Compared with ZYG11B, ZER1 exhibits a tenfold lower affinity for NLRP1, which we speculate may be attributed to potential electrostatic repulsion between the glutamate side chain at peptide position 3 (E3) and the negatively charged local environment surrounding this position in ZER1, whereas this unfavorable repulsion is likely to be alleviated in the more complementary environment of ZYG11B. This underscores the nuanced adaptability of CRL2 ligases in recognizing a variety of Gly/N-degrons, which is governed by conserved structural frameworks but influenced by sequence variation.

The results underline the evolutionary specialization of Gly/N-degron recognition in regulation of inflammasome activation (Yi et al., 2024 ▸; Jung et al., 2025 ▸). NLRP1 degradation via CRL2^ZER1 is likely to ensure precise control of inflammatory responses, as dysregulation could lead to excessive cytokine release or autoimmune pathology. However, questions persist regarding how ZER1 maintains a balance between substrate promiscuity and specificity, especially considering its role in degrading various proteins that bear Gly/N-degrons. Whether auxiliary factors or post-translational modifications fine-tune the substrate selectivity of ZER1 remains to be clarified. The structural insights provided here lay a foundation for the development of small-molecule inhibitors targeting ZER1, which may offer therapeutic potential for NLRP1-driven inflammatory diseases.

Supplementary Material

PDB reference: ZER1 bound to GSER degron, 9uyb

Funding Statement

This research was supported by Natural Science Research Project of Tianjin Municipal Education Commission (Grant No. 2022KJ191).

Conflict of interest

The authors declare that they have no conflicts of interest.