Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2026 Sep 14;11(38):56849–56857. doi: 10.1021/acsomega.6c03333

3D Electron Diffraction Elucidates the Structure of a 1,3,5-Triazin-2,4-dione Cinnamamide Covalent Inhibitor of SARS-CoV‑2 3CLpro

Meng Ge †,*, Peizhi Yan †, Boxin Zhang ‡, Bingxing Zhang ‡, Yunfei Zhang ‡, Lei Tian ‡, Chengyuan Liang ‡,*
PMCID: PMC13628987  PMID: 42825046

Abstract

The coronavirus 3-chymotrypsin-like protease (3CLpro) is an essential target for anticoronavirus drug discovery. Eleven covalent 1,3,5-triazin-2,4-dione cinnamamides (1a–1k) bearing a Cys145-directed α,β-unsaturated Michael acceptor were synthesized and evaluated against SARS-CoV-2 3CLpro. 1a showed the highest potency (IC50 = 92.45 nM) and was further characterized by structural and docking analyses. To determine its crystal structure, we applied three-dimensional electron diffraction (3DED), which enabled atomic-resolution analysis of nanocrystals under ultralow dose conditions. Integration of three data sets improved completeness from 58.2% to 91.0%, yielding a refined structural model at 0.85 Å resolution. The resulting structure provides an experimentally derived model of the molecular conformation and solid-state organization of compound 1a and illustrates the geometry of the α,β-unsaturated cinnamamide moiety in the crystalline state. Covalent docking analysis further suggests that the cinnamamide Michael acceptor can adopt a plausible orientation toward Cys145 within the 3CLpro active site. This work highlights 3DED as a useful approach for structural analysis of beam-sensitive pharmaceutical nanocrystals and provides experimentally derived structural information for further optimization of cinnamamide-containing 3CLpro inhibitors.

graphic file with name ao6c03333_0008.webp

graphic file with name ao6c03333_0007.webp

Introduction

Since its emergence in early 2020, COVID-19 caused by SARS-CoV-2 has imposed a sustained global burden. , As of March 2026, cumulative confirmed cases have exceeded 803 million, with reported deaths surpassing 7.1 million worldwide. Although the World Health Organization has declared that COVID-19 no longer constitutes a Public Health Emergency of International Concern, SARS-CoV-2 continues to circulate and generate variants with altered transmissibility and immune-evasion characteristics, maintaining the need for effective antiviral therapeutics alongside vaccination. , A central antiviral target is the coronavirus 3-chymotrypsin-like protease (3CLpro, Mpro), a cysteine protease that cleaves the pp1a and pp1ab polyproteins at 11 sites to release NSP4-NSP16 required for replication and transcription. − The sequence and three-dimensional architecture of 3CLpro are highly conserved across diverse coronaviruses, supporting its relevance as a broadly applicable drug target. , Accordingly, inhibiting 3CLpro blocks polyprotein processing and suppresses viral RNA replication and transcription, thereby limiting viral proliferation. ,

Medicinal chemistry over the past several years has demonstrated the clinical tractability of SARS-CoV-2 3CLpro inhibition. The peptidomimetic covalent inhibitor nirmatrelvir (PF-07321332) and the nonpeptidic noncovalent inhibitor ensitrelvir (S-217622) exemplify the therapeutic potential of this target. − At the same time, these agents underscore persistent optimization needs that continue to drive new scaffold and warhead discovery, including improved metabolic stability and oral exposure, reduced drug–drug interaction liabilities, and maintenance of antiviral efficacy as the virus evolves. Accordingly, there is strong interest in expanding nonpeptidic chemical space while preserving potent target engagement. Because 3CLpro catalysis depends on the nucleophilic Cys145, covalent inhibition remains an attractive strategy to achieve high potency and sustained target occupancy when electrophile reactivity is appropriately tuned. , Beyond peptidomimetic covalent designs, nonpeptidic covalent chemotypes featuring electrophilic motifs are appealing starting points for optimization. , Supporting the feasibility of this strategy, several naturally occurring electrophilic constituents of medicinal herbs have recently been identified as covalent inhibitors of SARS-CoV-2 3CLpro. − In this context, α,β-unsaturated amides are tunable Michael-acceptor warheads whose electrophilicity and steric properties can be modulated by aryl substitution (Figure ). − On this basis, a series of 1,3,5-triazin-2,4-dione cinnamamide derivatives (1a–1k) bearing different aryl substituents was designed and synthesized to position the α,β-unsaturated amide for covalent engagement of Cys145. Structure–activity relationship (SAR) profiling of cinnamoyl aryl substitution identified the unsubstituted analogue 1a as the most potent inhibitor, leading to its selection for atomic-resolution structural analysis.

1.

1

Open in a new tab

Design rationale for a nonpeptidic covalent SARS-CoV-2 3CLpro inhibitor.

The solid-state structure of 1a was determined to define its molecular conformation, functional-group orientation, and Michael-acceptor geometry. Single-crystal X-ray diffraction (SCXRD) remains the benchmark technique for atomic-level structure determination; however, many pharmaceutical compounds crystallize only as micro- or nanocrystalline solids that are unsuitable for conventional SCXRD. Powder X-ray diffraction (PXRD), while applicable to such samples, often suffers from peak overlap and limited sensitivity to light atoms, complicating accurate structure solution and refinement. Three-dimensional electron diffraction (3DED), also referred to as microcrystal electron diffraction (MicroED), has emerged as a powerful crystallographic approach for structure determination from crystals in the nanometer to submicrometer size range. Unlike conventional SCXRD, 3DED exploits the strong interaction between electrons and matter, enabling high-quality diffraction data to be collected from crystals that are several orders of magnitude smaller than those required for SCXRD. As a result, 3DED has been successfully applied to the structural elucidation of a wide variety of crystalline materials, including small-molecule organic compounds, pharmaceuticals, metal organic frameworks, covalent organic frameworks, zeolites, and related crystalline systems. − A key advantage of 3DED lies in its ability to perform true single-crystal analysis within multiphasic or heterogeneous samples. Individual nanocrystals can be selectively targeted and analyzed in a transmission electron microscope, allowing the crystal structures of distinct phases to be determined even when present in minute amounts. This capability is particularly valuable for pharmaceutical and materials systems where minor phases, polymorphs, or poorly crystalline products may evade detection by conventional PXRD techniques.

Building upon recent advances in 3DED, we report here the first atomic-level crystal structure of compound 1a. The compound crystallizes exclusively as nanometer-sized crystals, rendering conventional SCXRD impractical. High-quality 3DED data were therefore collected using an ultrashort exposure time of 0.5 s frame–1, with a total acquisition time of less than 3 min per data set. To overcome the inherent limitations in reciprocal-space coverage associated with electron diffraction, multiple data sets obtained from individual nanocrystals were merged, yielding a data set of sufficient completeness and quality to enable reliable structure solution and refinement at subångström resolution.

The resolved crystal structure defines the solid-state molecular conformation and packing of the cinnamamide scaffold, providing a structural basis to rationalize its electrophile geometry and to provide structure-informed considerations for potential covalent engagement of the catalytic Cys145 in 3CLpro. When integrated with enzymatic inhibition and cytotoxicity data, the experimentally determined crystal structure provides a concrete structural context for interpreting the activity profile and for generating testable structure–activity hypotheses. Together, these results establish a structure-informed basis for further optimization of cinnamamide-based covalent 3CLpro inhibitors toward improved antiviral lead properties.

Results and Discussion

Synthesis and Chemical Characterization

To improve 3CLpro binding and achieve sustained target engagement, Shionogi’s nonpeptidic, noncovalent inhibitor S-217622 and our previously reported triazin-2,4-dione-based inhibitors were selected as lead chemotypes. Cinnamoyl groups with varied aryl substitution were introduced by N-acylation of the 1,2,3-triazole moiety, affording compounds 1a–1k bearing an α,β-unsaturated amide warhead targeting Cys145. The analogues retained the same triazinane–dione–triazole core and differed only in cinnamoyl phenyl substitution, enabling direct evaluation of the electronic and steric effects on covalent inhibition (Figure ). N-Alkylation of 3-tert-butyl-6-(ethylthio)-1,3,5-triazin-2,4­(1H,3H)-dione (2) with 2,4,5-trifluorobenzyl bromide (3) in the presence of K2CO3 in refluxing acetonitrile afforded intermediate 4. TFA-mediated removal of the tert-butyl group in dichloromethane gave 5, followed by N-propargylation with propargyl bromide (6) to furnish 7. The reaction of 7 with 6-chloro-2-methyl-2H-indazol-5-amine (8) in tert-butanol and acetic acid afforded alkyne 9. CuSO4/sodium ascorbate-catalyzed azide–alkyne cycloaddition with TMSN3 in DMA/H2O yielded 1,2,3-triazole 10. Cinnamoyl chlorides 13a–13k, generated in situ from cinnamic acid derivatives 12a–12k and thionyl chloride (11), were subsequently coupled with 10 in the presence of TEA in DMF/THF to afford target compounds 1a–1k. All compounds were characterized by 1H and 13C NMR spectroscopy and high-resolution mass spectrometry (HRMS; Figures S1–S33).

2.

2

Open in a new tab

Synthetic route of compounds 1a–1k.

Inhibition of SARS-CoV-2 3CLpro and Cytotoxicity

Compounds 1a–1k were evaluated against wild-type SARS-CoV-2 3CLpro using a FRET-based enzymatic assay. Covalent inhibitor PF-07321332 and noncovalent inhibitor S-217622 were included as reference inhibitors, with IC50 values of 37.25 and 31.93 nM, respectively (Figure A,B). The unsubstituted analogue 1a was the most potent compound, with an IC50 of 92.45 nM. Substitution of the cinnamoyl phenyl ring consistently reduced potency. Among the substituted analogues, meta substitution was best tolerated. 1b (3-CN, 139.06 nM), 1c (3-NO2, 161.75 nM), 1e (3-CF3, 207.58 nM), and 1f (3-F, 276.19 nM) retained submicromolar activity. The 3-Cl,4-F analogue 1j showed comparable potency (186.61 nM), indicating that limited meta or para substitution was tolerated. Comparison of the regioisomers 1e (3-CF3, 207.58 nM) and 1g (4-CF3, 371.83 nM) further demonstrated a preference for meta substitution. In contrast, ortho substitution markedly impaired activity. The 2-F analogue 1d was essentially inactive (IC50 > 1000 nM), whereas the 2-F,3-Br analogue 1k showed weak inhibition (IC50 = 893.70 nM). Multiple or sterically demanding substituents were also unfavorable, as observed for 1h (3,4-di-F, 482.69 nM) and 1i [3,5-bis­(CF3), 736.12 nM]. Overall, the SAR favored the unsubstituted cinnamamide of 1a, whereas meta substitution was moderately tolerated and ortho or extensive substitution substantially reduced potency.

3.

3

Open in a new tab

In vitro 3CLpro inhibitory activity and cytotoxicity of the cinnamamide series. (A) 3CLpro inhibitory potency of 1a–1k and the reference inhibitors PF-07321332 and S-217622. (B) Representative dose–response inhibition curves for 1a, PF-07321332, and S-217622 against 3CLpro. (C-E)­Cytotoxicity of 1a, PF-07321332, and S-217622 in A549 (C), HEK293 (D), and HepG2 (E) cells.

As the most potent analogue, 1a was evaluated for cytotoxicity in A549, HEK293, and HepG2 cells. Cell viability remained above 80% at 400 nM, the highest concentration tested, in all three cell lines (Figure C–E). The estimated CC50 values were 3.24 mM, 229 μM, and 1.17 μM, respectively. No appreciable cytotoxicity was observed within the tested concentration range. The nonpeptidic scaffold, defined enzymatic SAR, and low cytotoxicity of 1a supported its selection for subsequent structural studies.

3DED Data Collection and Structure Solution

Due to the nanometer- to submicrometer-sized crystals obtained for compound 1a, conventional SCXRD could not be applied. Instead, 3DED using a continuous-rotation electron diffraction (cRED) protocol was employed for structure determination. By taking advantage of continuous rotation and low-dose data acquisition, a complete cRED data set could be collected within less than 3 min using an ultrashort exposure time of 0.5 s frame–1, yielding diffraction data with a maximum resolution of 0.85 Å (Figure ). These data-collection conditions effectively minimized electron-beam-induced damage and preserved the structural integrity of the beam-sensitive organic crystals during measurement.

4.

4

Open in a new tab

Representative 3DED pattern of compound 1a recorded with an exposure time of 0.5 s frame–1. The resolution ring at 0.85 Å and selected indexed reflections are marked.

The cRED data allowed reliable indexing of the crystals in the monoclinic space group Pc (No. 7), with unit-cell parameters a = 12.568 Å, b = 16.682 Å, c = 6.864 Å, α = 90.00°, β = 96.01°, and γ = 90.00°. To improve reciprocal-space coverage and data quality, three independent cRED data sets were collected from different nanocrystals. Each data set was processed individually using the program XDS for indexing and integration. The individual data sets showed completeness values of 68.4%, 58.2%, and 62.3%, respectively, reflecting the intrinsic limitations imposed by the goniometer tilt range in the transmission electron microscope.

To obtain a data set suitable for reliable refinement, the three data sets were subsequently merged using XSCALE, resulting in a combined completeness of 91.0% and a merged resolution of 0.87 Å. The merged data set exhibited significantly improved data statistics and yielded markedly enhanced electrostatic potential maps, with well-defined atomic peaks and reduced anisotropy compared to those obtained from single data sets.

Structure solution was carried out by direct methods implemented in the SHELX program suite, followed by refinement against the merged electron diffraction intensities using SHELXL with electron scattering factors. The refinement converged to final residuals of R 1 = 0.280 and wR 2 = 0.5075, values that are typical for organic structures refined against 3DED data where dynamical scattering effects are not explicitly modeled. Despite the relatively high residuals, the refined structural model is chemically reasonable, with sensible bond lengths and angles, and an electrostatic potential map fully consistent with the assigned atomic positions, supporting the correctness of the structure solution. Representative electrostatic potential maps around the cinnamamide Michael acceptor and aromatic moieties are shown in Figure S34.

The 3DED-determined structure reveals that compound 1a adopts an overall extended molecular conformation in the solid state (Figure ). In particular, the cinnamamide moiety displays a largely planar and extended geometry in the solid state. This experimentally resolved conformation provides a structural basis for interpreting the observed inhibitory activity of compound 1a and offers a solid foundation for subsequent structure-informed optimization.

5.

5

Open in a new tab

Final refined crystal structure of compound 1a obtained from the merged 3DED data set. (A) Molecular structure of compound 1a. (B–D) Crystal packing structures of 1a viewed along the a, b, and c axes, respectively. Carbon, oxygen, nitrogen, chlorine, and fluorine atoms are shown in gray, red, blue, green, and light green, respectively.

Data Completeness, Data Set Merging, and Refinement Strategy

To provide an overview of the experimental basis for the subsequent analysis, the data collection parameters and crystallographic characteristics of the three individual cRED data sets are summarized in Table . The data sets were collected from different nanocrystals using comparable acquisition conditions, including similar rotation ranges, exposure times, and total data collection durations. All three data sets were indexed in the same monoclinic space group, Pc (No. 7), and exhibit closely matching unit-cell parameters, indicating a high degree of isomorphism among the analyzed crystals. Despite this consistency, the individual data sets show limited and variable completeness, which precludes reliable structure refinement based on any single data set alone. These characteristics provide the basis for evaluating reciprocal-space coverage limitations and motivate the data set-merging strategy discussed in the following sections.

1. Data Collection and Crystallographic Information of the Three Data Sets of Compound 1a .

  data set
  1 2 3
rotation range (deg) 97.30 113.85 101.19
rotation step (deg) 0.38 0.39 0.39
exposure time (s frame–1) 0.5 0.5 0.5
no. of frames 229 262 232
total data collection time (min) 2.3 2.6 2.3
crystal system monoclinic
space group Pc (No. 7)
a (Å) 12.568 12.206 12.418
b (Å) 16.682 16.794 16.895
c (Å) 6.864 7.030 6.812
α (deg) 90 90 90
β (deg) 96.01 95.68 96.16
γ (deg) 90 90 90
Open in a new tab

Reciprocal-Space Completeness Limitation

A well-recognized limitation of 3DED arises from the restricted tilt range of the transmission electron microscope goniometer, which often leads to incomplete sampling of reciprocal space. While such data sets are frequently sufficient for ab initio structure solution, they may not provide the level of completeness required for a stable and precise structural refinement. Previous studies on beam-sensitive crystalline materials, including metal–organic frameworks and organic compounds, have demonstrated that limited completeness can compromise the quality of electrostatic potential maps and hinder reliable refinement. Consequently, improving data completeness is a critical step in achieving chemically meaningful structural models from 3DED data.

Single-Data Set Solution versus Refinement Instability

In the present study, the initial 3DED data sets exhibited completeness values ranging from 58.2% to 68.6% (Table ), reflecting the intrinsic limitations imposed by the restricted goniometer tilt range during TEM data collection. At this level of completeness, the crystal structure of compound 1a could be readily solved by direct methods, allowing unambiguous determination of molecular connectivity. However, refinement against any individual data set did not converge to a stable model. This instability arises from insufficient reciprocal-space sampling, which leads to pronounced correlations among refined parameters and prevents reliable optimization of atomic displacement parameters. Consequently, refinement based on single data sets was not pursued further, and subsequent structural analysis focused exclusively on merged data sets with substantially improved completeness.

Data Set Merging and Isomorphism Assessment

To overcome the limited reciprocal-space coverage inherent to individual 3DED data sets, three independent cRED data sets collected from different nanocrystals were scaled and merged using XSCALE. During the scaling procedure, XSCALE evaluates the correlation coefficients of common reflection intensities (CC I ) between each pair of data sets, which provides a quantitative measure of their internal consistency and crystallographic isomorphism (Table ). High CC I values indicate that the data sets share comparable intensity distributions and unit-cell metrics, and can therefore be meaningfully combined without introducing systematic inconsistencies.

2. Correlation Coefficients of Common Reflection Intensities (CC I ) between Pairs of cRED Data Sets for Compound 1a (Calculated by XSCALE).

data set no. data set no. no. of common reflections CC I
1 2 948 0.918
1 3 868 0.913
2 3 719 0.899
Open in a new tab

As summarized in Table , each pair of data sets for compound 1a shares a substantial number of common reflections (719–948), with corresponding CC I values ranging from 0.899 to 0.918. These consistently high correlation coefficients demonstrate that the three data sets originate from the same crystalline phase and exhibit compatible scaling behavior despite being collected from different nanocrystals. On this basis, the data sets were deemed suitable for merging. The resulting merged data set significantly improves reciprocal-space sampling, increasing the overall completeness to 91.0%, and provides a sufficiently robust experimental basis for subsequent structural refinement against a single, internally consistent data set.

Although moderate variations in unit-cell parameters were observed among the three individual data sets, all data sets were indexed in the same monoclinic space group Pc and showed highly consistent intensity distributions. The pairwise CC I values between common reflections were 0.899–0.918, supporting their crystallographic compatibility. Detailed merging statistics of the combined 3DED data set are summarized in Table S1. Therefore, the data sets were merged after correlation-based isomorphism assessment rather than simply combined to improve completeness. The merged data set contained 9529 measured reflections and 3892 unique reflections, with an overall completeness of 90.9%, R int of 0.2602, and R σ of 0.3002 at 0.87 Å resolution.

Refinement Outcome and Interpretation of R Values

Refinement against the merged 3DED data set enabled the construction of a well-defined and chemically consistent structural model. Due to the limited reciprocal-space coverage of individual data sets, reliable refinement could not be achieved when using any single data set alone, although ab initio structure solution was possible. In contrast, merging three isomorphous cRED data sets substantially increased data completeness and reduced parameter correlations, thereby enabling stable refinement of atomic displacement parameters and reliable evaluation of geometric parameters.

The crystallographic statistics of the merged data set are summarized in Table . The merged data set reaches a completeness of 91.0% at a resolution of 0.87 Å and can be indexed in the monoclinic space group Pc (No. 7), with unit-cell parameters a = 12.56(3) Å, b = 16.67(3) Å, c = 6.88(1) Å, and β = 95.97(3)°, corresponding to a unit-cell volume of 1432.7(5) Å3. In total, 9529 reflections were measured, yielding 3892 unique reflections, of which 2129 satisfy the criterion I > 2σ­(I).

3. Crystallographic Data and Refinement Details for the Merged Data Set.

  merged 3DED data
no. of data sets merged 3
wavelength (Å) 0.02508
resolution (Å) 0.87
crystal system monoclinic
space group Pc (No. 7)
unit cell a (Å), b (Å), c (Å), β (deg) 12.56(3), 16.67(3), 6.88(1), 95.97(3)
volume (Å3) 1432.7(5)
completeness (%) 91.0
no. of measured reflections (cRED) 9529
no. of unique reflections 3892
no. of observed reflections [I > 2σ(I)] 2129
R 1 [I > 2σ(I)] 0.192
R 1 (all reflections) 0.280
Goof 1.382
Open in a new tab

Refinement against the merged data set converged to R 1 values of 0.192 for observed reflections and 0.280 for all reflections, with a goodness-of-fit of 1.382 (Table ). Although these residual indices are higher than those typically obtained from SCXRD, they are consistent with values commonly reported for 3DED refinements, where dynamical scattering effects are not explicitly accounted for within kinematical refinement models. Importantly, the refined bond lengths, bond angles, and overall molecular geometry are chemically reasonable, supporting the reliability of the final structural model. Selected geometric parameters around the cinnamamide moiety are provided in Tables S2 and S3.

Taken together, these results demonstrate that while a single 3DED data set with moderate completeness can be sufficient for ab initio structure solution, merging multiple data sets to achieve completeness above 90% is essential for obtaining a robust and chemically meaningful refined structure. Moreover, correlation-based assessment of data set isomorphism provides a rational criterion for data set selection, ensuring that improvements in completeness are not offset by inconsistencies in data quality.

Covalent Docking Analysis of 1a with SARS-CoV-2 3CLpro

To investigate the covalent binding mode of 1a, covalent docking was performed in the active site of SARS-CoV-2 3CLpro, with Cys145 specified as the attachment residue. The resulting model placed the cinnamamide Michael acceptor adjacent to Cys145, allowing 1,4-addition of the cysteine thiolate to the electron-deficient double bond. A thioether linkage was formed between the Cys145 Sγ atom and the β-carbon of the Michael acceptor, with an Sγ–Cβ distance of 1.81 Å and a Cα–Cβ–Sγ angle of 113°. The tetrahedral geometry at Cβ and the modeled bond length were compatible with formation of a covalent Michael adduct (Figure ). 1a extended across the substrate-binding cleft and occupied the S1, S1′, S2, and S4 subsites (Figure A). The binding pose was further stabilized by multiple noncovalent interactions (Figure B). Hydrogen-bonding contacts were observed with Glu166, Asn142, the backbone of Cys145, Gln189, and Ser144, thereby engaging the oxyanion-loop and S1-pocket regions. Additional contacts were formed with Met165 in the S2 subsite, Gln189 in the S4 subsite, and Leu27 near the S1′ subsite. The aromatic moiety of 1a also formed a T-shaped π-stacking interaction with His41. Thus, both residues of the catalytic dyad were engaged in the modeled complex, with His41 contributing aromatic recognition and Cys145 forming the covalent linkage.

6.

6

Open in a new tab

Predicted covalent binding mode of 1a in the SARS-CoV-2 3CLpro active site, PDB ID: 7VU6. (A) Molecular surface representation of 3CLpro (gray) with 1a (salmon sticks) bound in the substrate-binding cleft, occupying the S1, S1′, S2, and S4 subsites. (B) Close-up of the interaction network. 1a forms a covalent thioether bond between the β-carbon of its cinnamamide Michael acceptor and the Sγ atom of Cys-145.

Overall, the docking model indicated that the ensitrelvir-derived scaffold retained noncovalent recognition of the 3CLpro substrate-binding cleft, whereas the appended cinnamamide Michael acceptor was appropriately oriented toward Cys145. This combination of noncovalent anchoring and modeled covalent engagement of the catalytic cysteine provides a plausible binding rationale for the potent inhibition of 3CLpro by 1a.

Conclusions

In this study, 3DED enabled atomic-level structure determination of a beam-sensitive, nanocrystalline pharmaceutical compound that is not amenable to conventional SCXRD. By combining low-dose cRED with a data set-merging strategy, a high-completeness data set was obtained, enabling reliable structure solution and refinement at subångström resolution despite intrinsic limitations in reciprocal-space coverage. The resolved structure defines the solid-state conformation of a cinnamamide covalent inhibitor of SARS-CoV-2 3CLpro and establishes the geometry of the Michael acceptor, demonstrating that 3DED can deliver chemically informative structural detail that is not readily accessible from molecular design or biochemical assays alone. More broadly, this work highlights the growing potential of 3DED as a versatile crystallographic tool for structurally challenging organic and pharmaceutical systems and underscores its value in structure-informed molecular design and discovery.

Experimental Section (Methods)

Transmission Electron Microscopy (TEM)

Transmission electron microscopy (TEM) samples were prepared using a solvent-free dry-deposition method. A small amount of the crystalline powder was placed between two clean glass slides and gently ground to obtain a fine, well-dispersed nanocrystalline powder. A copper TEM grid coated with a carbon support film was then lightly placed onto the ground powder, allowing individual crystallites to adhere to the carbon film. Excess material was removed by gentle tapping of the grid. This dry-deposition approach avoids the use of dispersing solvents and thereby minimizes the risk of partial dissolution and solution- or solvent-mediated phase transformations, which could otherwise alter the original solid-state form prior to electron-diffraction measurements.

Continuous-rotation electron diffraction (cRED) data were collected on a JEOL F200 transmission electron microscope equipped with a Schottky field-emission gun and a CEOS CEFID energy filter, operated at an accelerating voltage of 200 keV. Zero-loss diffraction data were recorded using an ASI Cheetah M3 retractable hybrid pixel detector, generating 16-bit image stacks with a frame size of 512 × 512 pixels. The detector was operated in continuous low-gain mode to minimize detector saturation and beam-induced damage.

All cRED data sets were acquired under cryogenic conditions using a Gatan Elsa cryo-transfer holder, which provided a goniometer tilt range from −80° to +80°. Automated cRED data acquisition was controlled using the Instamatic , software package, enabling continuous-rotation data collection with precise synchronization of goniometer rotation and detector readout.

Inhibition of SARS-CoV-2 3CLpro Enzymatic Activity

SARS-CoV-2 3CLpro activity was measured with a commercial kit (Beyotime, P0312S, Shanghai, China) following the manufacturer’s instructions. Briefly, recombinant 3CLpro (1 μL) in assay buffer (90 μL) was preincubated with 5 μL of each test compound prepared in DMSO (Sigma-Aldrich, Beijing, China) at 37 °C with shaking for 15 min. The reaction was initiated by adding 4 μL of the FRET substrate Dabcyl-KTSAVLQSGFRKME-Edans (20 μM) and allowed to proceed for 10 min at 37 °C in the dark. Fluorescence was recorded at excitation/emission 340/490 nm on a multimode microplate reader (Varioskan Flash, Thermo Fisher Scientific, Vantaa, Finland). PF-07321332 (nirmatrelvir, covalent) and S-217622 (Ensitrelvir, noncovalent) were run in parallel as reference inhibitors. Wells containing enzymes with vehicle (DMSO) served as positive controls, and wells lacking enzyme served as blanks. IC50 values were obtained by nonlinear regression in GraphPad Prism 10.

Covalent Docking Analysis

Covalent docking analysis was performed using AutoDockFR to evaluate the potential binding mode of compound 1a within the active site of SARS-CoV-2 3CLpro. The crystal structure of SARS-CoV-2 3CLpro in complex with a reported inhibitor (PDB ID: 7VU6) was used as the receptor model. Prior to docking, the bound ligand and crystallographic water molecules were removed, hydrogen atoms were added, and partial charges were assigned. The catalytic residue Cys145 was designated as the covalent reaction site. The structure of compound 1a was constructed in Chem3D 21.0 and geometrically optimized using the MM2 force field. The docking grid was centered on the inhibitor-binding pocket of 3CLpro. Covalent docking calculations were carried out using the link-atom protocol implemented in AutoDockFR, and ligand poses were sampled using the Lamarckian genetic algorithm. The final binding pose was selected based on docking score and covalent-bond geometry and subsequently analyzed using PyMOL 3.1 and Discovery Studio Visualizer 4.0.

Cytotoxicity

The cell viability was measured using the CCK-8 assay. HepG2, HEK293, and A549 cells (ATCC) were maintained in DMEM supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. Cells were seeded in 96-well plates at 1 × 104 cells well–1 in 100 μL of medium and allowed to adhere overnight. Serial dilutions of test compounds in complete medium were added, and cells were incubated for 48 h at 37 °C. CCK-8 reagent (10 μL) was then added to each well and incubation continued for 4 h. Absorbance at 450 nm was recorded on a multimode reader (Varioskan Flash, Thermo Fisher Scientific, Vantaa, Finland). PF-07321332 and S-217622 were included as reference compounds. The cell viability was calculated as

viability(%)=A(candidate)−A(blank)A(control)−A(blank)×100

where A(candidate) is the absorbance of drug-treated wells, A(blank) is the absorbance of wells without cells, and A(control) is the absorbance of untreated cell wells. CC50 values were obtained by nonlinear curve fitting of the concentration–response data in GraphPad Prism 10.

Statistical Analysis

Data are expressed as mean ± SD. Group differences were evaluated by one-way ANOVA followed by Tukey’s multiple-comparison test. Analyses were performed in GraphPad Prism 10 (GraphPad Software, USA). Statistical significance was defined as P < 0.05.

Deposition Number 2534628 contains the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service (www.ccdc.cam.ac.uk/structures).

Supplementary Material

ao6c03333_si_001.cif (337.4KB, cif)
ao6c03333_si_002.pdf (2.5MB, pdf)

Acknowledgments

The study was supported by Schweizerischer Nationalfonds zur Foerderung der Wissenschaftlichen Forschung (Grant TMPFP3_210032 to M.G.), National Natural Science Foundation of China (Grant 82373843 to C.L.), Science and Technology Department of Shaanxi Province (Grant S2024-YF-ZDCXL-ZDLSF-0008 to C.L.), Xi’an Key Laboratory for Antiviral and Antimicrobial-Resistant Bacteria Therapeutics Research (Grant 2024JH-ZDSYJX-0018 to C.L.), Research Center for the Development of Therapeutic Drugs for Serious Infectious Diseases in Shaanxi Province’s Universities and Colleges (grant to C.L.), and Xi’an International Science and Technology Cooperation Base for Anti-Infective Drug Research and Development (grant to C.L.).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c03333.

  • Additional crystallographic information (CIF)

  • 1H and 13C NMR and HRMS spectra of compounds 1a–1k; electrostatic potential maps of compound 1a; merging statistics of the combined 3DED data set; selected bond lengths and bond angles around the cinnamamide moiety (PDF)

All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

References

  1. Zhu N., Zhang D., Wang W., Li X., Yang B., Song J., Zhao X., Huang B., Shi W., Lu R.. et al. A Novel Coronavirus from Patients with Pneumonia in China, 2019. N Engl J. Med. 2020;382(8):727–733. doi: 10.1056/NEJMoa2001017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Nicola M., Alsafi Z., Sohrabi C., Kerwan A., Al-Jabir A., Iosifidis C., Agha M., Agha R.. The socio-economic implications of the coronavirus pandemic (COVID-19): A review. Int. J. Surg. 2020;78:185–193. doi: 10.1016/j.ijsu.2020.04.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. World Health Organization . WHO COVID-19 Dashboard. https://data.who.int/dashboards/covid19/cases (accessed 2026-07-18).
  4. Harvey W. T., Carabelli A. M., Jackson B., Gupta R. K., Thomson E. C., Harrison E. M., Ludden C., Reeve R., Rambaut A., Peacock S. J., Robertson D. L.. SARS-CoV-2 variants, spike mutations and immune escape. Nat. Rev. Microbiol. 2021;19(7):409–424. doi: 10.1038/s41579-021-00573-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Tao K., Tzou P. L., Nouhin J., Gupta R. K., de Oliveira T., Kosakovsky Pond S. L., Fera D., Shafer R. W.. The biological and clinical significance of emerging SARS-CoV-2 variants. Nat. Rev. Genet. 2021;22(12):757–773. doi: 10.1038/s41576-021-00408-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Anand K., Ziebuhr J., Wadhwani P., Mesters J. R., Hilgenfeld R.. Coronavirus main proteinase (3CLpro) structure: basis for design of anti-SARS drugs. Science. 2003;300(5626):1763–1767. doi: 10.1126/science.1085658. [DOI] [PubMed] [Google Scholar]
  7. Jin Z., Du X., Xu Y., Deng Y., Liu M., Zhao Y., Zhang B., Li X., Zhang L., Peng C.. et al. Structure of M­(pro) from SARS-CoV-2 and discovery of its inhibitors. Nature. 2020;582(7811):289–293. doi: 10.1038/s41586-020-2223-y. [DOI] [PubMed] [Google Scholar]
  8. Ullrich S., Nitsche C.. The SARS-CoV-2 main protease as drug target. Bioorg. Med. Chem. Lett. 2020;30(17):127377. doi: 10.1016/j.bmcl.2020.127377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Pillaiyar T., Manickam M., Namasivayam V., Hayashi Y., Jung S. H.. An Overview of Severe Acute Respiratory Syndrome-Coronavirus (SARS-CoV) 3CL Protease Inhibitors: Peptidomimetics and Small Molecule Chemotherapy. J. Med. Chem. 2016;59(14):6595–6628. doi: 10.1021/acs.jmedchem.5b01461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Owen D. R., Allerton C. M. N., Anderson A. S., Aschenbrenner L., Avery M., Berritt S., Boras B., Cardin R. D., Carlo A., Coffman K. J.. et al. An oral SARS-CoV-2 Mpro inhibitor clinical candidate for the treatment of COVID-19. Science. 2021;374(6575):1586–1593. doi: 10.1126/science.abl4784. [DOI] [PubMed] [Google Scholar]
  11. Shang W., Dai W., Yao C., Xu L., Tao X., Su H., Li J., Xie X., Xu Y., Hu M.. et al. In vitro and in vivo evaluation of the main protease inhibitor FB2001 against SARS-CoV-2. Antiviral Res. 2022;208:105450. doi: 10.1016/j.antiviral.2022.105450. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Tian L., Qiang T., Yang X., Gao Y., Zhai X., Kang K., Du C., Lu Q., Gao H., Zhang D.. et al. Development of de-novo coronavirus 3-chymotrypsin-like protease (3CLpro) inhibitors since COVID-19 outbreak: A strategy to tackle challenges of persistent virus infection. Eur. J. Med. Chem. 2024;264:115979. doi: 10.1016/j.ejmech.2023.115979. [DOI] [PubMed] [Google Scholar]
  13. Marzolini C., Kuritzkes D. R., Marra F., Boyle A., Gibbons S., Flexner C., Pozniak A., Boffito M., Waters L., Burger D.. et al. Recommendations for the Management of Drug-Drug Interactions Between the COVID-19 Antiviral Nirmatrelvir/Ritonavir (Paxlovid) and Comedications. Clin Pharmacol Ther. 2022;112(6):1191–1200. doi: 10.1002/cpt.2646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Singh J., Petter R. C., Baillie T. A., Whitty A.. The resurgence of covalent drugs. Nat. Rev. Drug Discov. 2011;10(4):307–317. doi: 10.1038/nrd3410. [DOI] [PubMed] [Google Scholar]
  15. Bauer R. A.. Covalent inhibitors in drug discovery: from accidental discoveries to avoided liabilities and designed therapies. Drug Discov Today. 2015;20(9):1061–1073. doi: 10.1016/j.drudis.2015.05.005. [DOI] [PubMed] [Google Scholar]
  16. Hoffman R. L., Kania R. S., Brothers M. A., Davies J. F., Ferre R. A., Gajiwala K. S., He M., Hogan R. J., Kozminski K., Li L. Y.. et al. Discovery of Ketone-Based Covalent Inhibitors of Coronavirus 3CL Proteases for the Potential Therapeutic Treatment of COVID-19. J. Med. Chem. 2020;63(21):12725–12747. doi: 10.1021/acs.jmedchem.0c01063. [DOI] [PubMed] [Google Scholar]
  17. Zhu G. H., Zhang Y. N., Xiong Y., Hou X. D., Zhang Q. G., Zhang Z. Q., Zhuang X. Y., Zhang W. D., Ge G. B.. Uncovering the covalent inhibitors of SARS-CoV-2 Mpro in Tibetan edible herb Rhodiola crenulata and their synergistic anti-Mpro mechanism. J. Pharm. Anal. 2026;16(2):101224. doi: 10.1016/j.jpha.2025.101224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hu Q., Zhang Y. W., Zhang Y. N., Zhu G. H., Chen P. C., Liu W., Hu X. P., Song F. F., Pan Z. F., Zheng S. L.. et al. Uncovering the naturally occurring covalent inhibitors of SARS-CoV-2 Mpro from the Chinese medicine sappanwood and deciphering their synergistic anti-Mpro effects. J. Ethnopharmacol. 2025;342:119397. doi: 10.1016/j.jep.2025.119397. [DOI] [PubMed] [Google Scholar]
  19. Hu Q., Zhang Y., Chen P., Zhang Y., Zhu G., Liu W., Wang C., Zheng S., Shen N., Wang H.. et al. Discovery and characterization of naturally occurring covalent inhibitors of SARS-CoV-2 Mpro from the antiviral herb Ephedra. Chin J. Nat. Med. 2024;22(9):797–807. doi: 10.1016/S1875-5364(24)60577-7. [DOI] [PubMed] [Google Scholar]
  20. Su H., Yao S., Zhao W., Zhang Y., Liu J., Shao Q., Wang Q., Li M., Xie H., Shang W.. et al. Identification of pyrogallol as a warhead in design of covalent inhibitors for the SARS-CoV-2 3CL protease. Nat. Commun. 2021;12(1):3623. doi: 10.1038/s41467-021-23751-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Zhu G., Zhang Y., Qi S., Xu J., Xiong Y., Zhang Z., Chen X., Xie Y., Guan X., Zhang W.. et al. Highly efficient discovery of the covalent Mpro inhibitors from crude Pu-erh tea by integrating biochemical and chemoproteomic approaches. Food Sci. Hum Wellness. 2025;14(5):9250120. doi: 10.26599/FSHW.2024.9250120. [DOI] [Google Scholar]
  22. Citarella A., Moi D., Pedrini M., Perez-Pena H., Pieraccini S., Dimasi A., Stagno C., Micale N., Schirmeister T., Sibille G.. et al. Synthesis of SARS-CoV-2 Mpro inhibitors bearing a cinnamic ester warhead with in vitro activity against human coronaviruses. Org. Biomol Chem. 2023;21(18):3811–3824. doi: 10.1039/D3OB00381G. [DOI] [PubMed] [Google Scholar]
  23. Citarella A., Sibille G., Moi D., Dimasi A., Braga T., Col L. D., Ruberto L., Pieraccini S., Sironi M., Micale N.. et al. Synthesis and multitarget inhibitory effect of indole-based ethyl cinnamate derivatives against SARS-CoV-2 Mpro and cathepsins for broad-spectrum anti-coronavirus activity. Bioorg. Med. Chem. 2025;128:118258. doi: 10.1016/j.bmc.2025.118258. [DOI] [PubMed] [Google Scholar]
  24. Stille J. K., Tjutrins J., Wang G., Venegas F. A., Hennecker C., Rueda A. M., Sharon I., Blaine N., Miron C. E., Pinus S.. et al. Design, synthesis and in vitro evaluation of novel SARS-CoV-2 3CLpro covalent inhibitors. Eur. J. Med. Chem. 2022;229:114046. doi: 10.1016/j.ejmech.2021.114046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gemmi M., Mugnaioli E., Gorelik T. E., Kolb U., Palatinus L., Boullay P., Hovmoller S., Abrahams J. P.. 3D Electron Diffraction: The Nanocrystallography Revolution. ACS Cent Sci. 2019;5(8):1315–1329. doi: 10.1021/acscentsci.9b00394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Yang T., Willhammar T., Xu H., Zou X., Huang Z.. Single-crystal structure determination of nanosized metal-organic frameworks by three-dimensional electron diffraction. Nat. Protoc. 2022;17(10):2389–2413. doi: 10.1038/s41596-022-00720-8. [DOI] [PubMed] [Google Scholar]
  27. Huang Z., Willhammar T., Zou X.. Three-dimensional electron diffraction for porous crystalline materials: structural determination and beyond. Chem. Sci. 2021;12(4):1206–1219. doi: 10.1039/D0SC05731B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Huang Z., Ge M., Carraro F., Doonan C., Falcaro P., Zou X.. Can 3D electron diffraction provide accurate atomic structures of metal-organic frameworks? Faraday Discuss. 2021;225:118–132. doi: 10.1039/D0FD00015A. [DOI] [PubMed] [Google Scholar]
  29. Lightowler M., Li S., Ou X., Cho J., Liu B., Li A., Hofer G., Xu J., Yang T., Zou X., Lu M., Xu H.. Phase Identification and Discovery of an Elusive Polymorph of Drug-Polymer Inclusion Complex Using Automated 3D Electron Diffraction. Angew. Chem., Int. Ed. Engl. 2024;63(16):e202317695. doi: 10.1002/anie.202317695. [DOI] [PubMed] [Google Scholar]
  30. Okabe A., Carney D. W., Tawada M., Akther T., Aida J., Takagi T., Dougan D. R., Leffler A. E., Bell J. A., Frye L.. et al. Discovery of Highly Potent Noncovalent Inhibitors of SARS-CoV-2 Main Protease through Computer-Aided Drug Design. J. Med. Chem. 2025;68(20):21330–21345. doi: 10.1021/acs.jmedchem.5c01199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Liang, C. ; Xin, L. ; Tian, L. ; Li, J. ; Xie, X. ; Zhang, D. ; Wu, K. ; Zhang, S. ; Yang, X. ; Liu, S. ; et al. 3-Triazolyl Methyl-1,3,5-triazine-2,4-dione Compounds and Preparation Method and Application Thereof. U.S. Patent US 11,612,603 B1, 2023.
  32. Ge M., Zou X., Huang Z.. Three-Dimensional Electron Diffraction for Structural Analysis of Beam-Sensitive Metal-Organic Frameworks. Crystals. 2021;11(3):263. doi: 10.3390/cryst11030263. [DOI] [Google Scholar]
  33. Cichocka M. O., Angstrom J., Wang B., Zou X., Smeets S.. High-throughput continuous rotation electron diffraction data acquisition via software automation. J. Appl. Crystallogr. 2018;51(6):1652–1661. doi: 10.1107/S1600576718015145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Smeets S., Wan W.. Serial electron crystallography: merging diffraction data through rank aggregation. J. Appl. Crystallogr. 2017;50(3):885–892. doi: 10.1107/S1600576717005854. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao6c03333_si_001.cif (337.4KB, cif)
ao6c03333_si_002.pdf (2.5MB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES