Structural and functional characterization of NEMO cleavage by SARS-CoV-2 3CLpro

Abstract

In addition to its essential role in viral polyprotein processing, the SARS-CoV-2 3C-like protease (3CLpro) can cleave human immune signaling proteins, like NF-κB Essential Modulator (NEMO) and deregulate the host immune response. Here, in vitro assays show that SARS-CoV-2 3CLpro cleaves NEMO with fine-tuned efficiency. Analysis of the 2.50 Å resolution crystal structure of 3CLpro C145S bound to NEMO_226-234 reveals subsites that tolerate a range of viral and host substrates through main chain hydrogen bonds while also enforcing specificity using side chain hydrogen bonds and hydrophobic contacts. Machine learning- and physics-based computational methods predict that variation in key binding residues of 3CLpro-NEMO helps explain the high fitness of SARS-CoV-2 in humans. We posit that cleavage of NEMO is an important piece of information to be accounted for, in the pathology of COVID-19.

Fig. 1. Truncated NF-κB essential modulator (NEMO) is cleaved by 3CLpro in enzymatic assays.

a NEMO includes the α-helical domain 1 (Hlx1), the coiled-coil domain 1 (CC1), the α-helical domain 2 (Hlx2), the coiled-coil domain (CC2), a leucine zipper (LZ) domain, and the C-terminal zinc-finger (ZF). Human NEMO (hNEMO) truncated at site 215 and 247 was used in enzymatic assays with 3CLpro. A recognition site of cleavage is found at Gln231. b Cleavage of hNEMO_215–247 at 0.053 μg/μL (~13 μM) by 3CLpro at two concentrations. Reactions were incubated at 25 °C and aliquots were quenched at different times for analysis. The extent of proteolysis was quantified by LC-MS/MS. Apparent % cleavage was calculated by dividing the product peak area by the sum of the substrate and product peak areas. Error bars represent the range of duplicate enzymatic reactions. Statistics have been derived for n = 2 biologically independent experiments. c Multiple sequence alignment of peptide sequences of SARS-CoV-2 polyprotein and hNEMO. P1 site glutamine residues are shown in red. The peptide (P6 to P4’) used in the crystal structure is indicated beneath the sequences. d Coiled-coil pitch per residue computed with TWISTER for NEMO in the PDB structures 6MI3 (region I), 3CL3 (region II), and 6YEK (region III) is shown. Dashed lines indicate the regions I-III corresponding to these structures. The cleavage sites Gln83, Gln205, Gln 231, Gln304, and Gln313 are indicated with red arrows. The region corresponding to hNEMO_226–235, used in our X-ray structure determination is pointed out (violet arrow). e PAIRCOIL prediction of coiled-coil propensity per residue for human and mouse NEMO (mNEMO). Lower P-scores implies greater likelihood of coiled-coil. Potential disordered binding regions predicted by ANCHOR are shown in brown. The cleavage sites are depicted as in d. The region corresponding to hNEMO_226–235 used in our X-ray structure is depicted as in d.

Fig. 2. Structure of the hNEMO-bound 3CLpro C145S homodimer.

a The hNEMO_226–234-bound 3CLpro dimer. N- (NH3 + ) and C-termini (COO-) are labeled. hNEMO_226–234 (NEMO B - orange) binds into chain B (purple) and is surrounded by an omit map of electron density (1.0 σ contour level and 1.9 Å carving radius). Acetylated Lys226 and Tyr234 at the N- and C-termini of the hNEMO_226–234 peptide, respectively, are labeled. The C-terminal tail of chain D (teal) from a crystallographic symmetry equivalent binds into the substrate-binding site of chain A (light pink). Domains I, II and III are labeled. b Structure of the hNEMO-bound C145S variant asymmetric unit. The C-terminal tail of chain A (pink) binds into the substrate-binding site of chain D (teal). hNEMO peptides bound into chain B (purple) (NEMO B - orange) and chain C (light blue) (NEMO C - wheat) are superimposed with an omit map (1.0 σ contour level and 1.9 Å carving radius) of their electron density. c Interactions of hNEMO_226–234 with the substrate-binding groove of 3CLpro C145S. hNEMO_226–234 is colored orange. Residues Lys226 to Tyr234 are fully labeled. The surface of chain B (purple) is shown. Substrate-binding residues in chain B are portrayed as sticks and labeled. Catalytically relevant residues are labeled in bold. Hydrogen bonds are depicted as dashed yellow lines. The hydrogen bond predicted to form between Cys145 (in WT) or Ser145 (in C145S) and His41 is depicted as a dashed green line. d Interactions of the C-terminal tail of chain A with the substrate-binding groove of 3CLpro C145S. The C-terminal tail of chain A is colored light pink. Residues Ser301 to Gln306 are fully labeled. The surface of chain D (teal) is shown. Interacting residues and their interactions are portrayed as in c. The oxygen atom of water 109 is depicted (red sphere). e The C-terminal tail at the 3CLpro dimer interface. The C-terminal tail of chain B (purple) is depicted as sticks, juxtaposed over its cartoon representation. Density from the omit map (1.0 σ contour level and 1.9 Å carving radius) is shown around the C-terminal tail. Arg298 to Phe305 and residues that interact with the C-terminal tail of chain B are labeled. Inset, schematic of chains A and B, showing the position of the C-terminal tail at the 3CLpro dimer interface.

Fig. 3. Molecular dynamics simulations of SARS-CoV-2 3CLpro bound to human or mouse NEMO.

a Main contacts from MD simulations and predicted hot spots in WT 3CLpro bound to human NEMO_227–234. Contacts that persist for more than 70% of the simulation time are depicted. Hot spots predicted with either KFCa or KFCb are labeled in bold and those predicted with both methods are underlined. Persistent H-bonds are shown as dashed lines (Supplementary Table 2). A representative conformation of the short (aa. 227-234) and long constructs of NEMO in the binding site is shown as solid and transparent surfaces, respectively. b Sequence Logo generated with WebLogo of NEMO (aa. 216-253) across 535 animal sequences, including those from placentals, bats, marsupials, birds, rodents, and primates. Clustal Omega was used for the multiple sequence alignment and as an input for the WebLogo analysis. c Time evolution of the distance between the catalytic S⁻ in 3CLpro Cys145 and the carbonyl C of Gln231 in hNEMO_227–234 and mNEMO_227–234 computed from MD simulations.

Fig. 4. Comparative analysis of 3CLpro from human-infecting betacoronaviruses bound to NEMO.

a–d Substrate-binding site of human-infecting betacoronaviruses with hNEMO_227–234. Contacts that persist for more than 70% of the simulation time are labeled in bold. e Ranking of predicted hNEMO_227–234-binding affinities to 3CLpro from betacoronaviruses computed with quantum mechanics (QM) and molecular dynamics/machine learning (MD/ML) approaches. SARS-CoV-2, SARS-CoV, HCoV-HKU1, and MERS-CoV are abbreviated as SARS2, SARS1, HKU1, and MERS, respectively. In the MD/ML approach, five machine learning methods were used to train the model, namely, support-vector machine (SVM), gradient-boosted trees (BT; scaled and unscaled*), and random forest (RF; scaled and unscaled*). The ranking displayed was consistent for nine of the five cases using MD conformers (five ML models applied to either all MD conformers or the three lowest-energy MD conformers). The exception was a boosted tree model trained on unnormalized features that yielded a ranking of SARS-CoV < MERS-CoV < HKU1-CoV < SARS-CoV-2 when considering just the conformers with the lowest energy of interaction with 3CLpro computed from MD simulations. *Unscaled refers to the fact that unscaled, or unnormalized, features were used in the training.

Fig. 5. Comparison of 3CLpro interactions with its N-terminal sequence and the hNEMO_226–234 peptide.

a The surface of chain B from our hNEMO_226–234-bound structure is depicted as in Fig. 2c. This is juxtaposed with residues from our structure that form interactions with the hNEMO_226–234 peptide (purple sticks) as well as N-terminal interacting residues (green sticks) from a structure of 3CLpro bound to an N-terminal peptide (PDB_ID:7N89). For 7N89, oxygens are colored light red, nitrogens are colored aquamarine and sulfurs are colored light yellow, for ease of comparison. Residues forming conserved interactions are labeled in black. Those found only in hNEMO_226–234-bound 3CLpro are labeled in purple. Those found only in 7N89 are labeled in green. Water molecules are portrayed as red spheres and are exclusive to 7N89. Leu27 forms a hydrophobic contact with P1’ of hNEMO_226–234 only. Met49 forms a hydrophobic contact with the P2 residues of both hNEMO_226–234 and the N-terminal peptide, but also forms a hydrophobic contact with P3’ of the N-terminal sequence. b Comparison of bound hNEMO_226–234 from our structure with bound N-terminal peptide from 7N89 shows a conserved substrate pose. hNEMO_226–234 is colored as in Fig. 2c. The N-terminal peptide carbons in 7N89 are colored dark brown and the oxygens and nitrogens are colored as in panel a. Substrate residues are labeled P6-P3’.