Skip to main content
NCBI home page
ACS Central Science logoLink to ACS Central Science
. 2023 Dec 14;9(12):2358–2368. doi: 10.1021/acscentsci.3c01068

Modulating Liquid–Liquid Phase Separation of Nck Adaptor Protein against Enteropathogenic Escherichia coli Infection

Min Liu †, Chunjian Wu †,*, Rui Wang ‡, Jiaming Qiu †, Zhentao She §, Jianan Qu §, Jiang Xia †,*
PMCID: PMC10755736  PMID: 38161366

Abstract

graphic file with name oc3c01068_0008.jpg

Signaling proteins often form biomolecular condensates through liquid–liquid phase separation (LLPS) during intracellular signal transduction. Modulating the LLPS property of intracellular protein condensates will redirect intracellular signals and provide a potential way to regulate cellular physiology. Phosphorylation of multiple tyrosine residues of the transmembrane receptor nephrin is known to drive the LLPS of the adaptor protein Nck and neuronal Wiskott–Aldrich Syndrome protein (N-WASP) and form the Nck signaling complex. Phosphorylation of the translocated intimin receptor (Tir) in the host cell may recruit this enteropathogenic Escherichia coli (EPEC) virulence factor to the Nck signaling complex and lead to the entry of EPEC into the intestine cell. In this work, we first identified a phosphotyrosine (pY)-containing peptide 3pY based on the sequence similarity of nephrin and Tir; 3pY promoted the LLPS of Nck and N-WASP, mimicking the role of phosphorylated nephrin. Next, we designed a covalent blocker of Nck, peptide p1 based on the selected pY peptides, which site-selectively reacted with the SH2 domain of Nck (Nck-SH2) at Lys331 through a proximity-induced reaction. The covalent reaction of p1 with Nck blocked the protein binding site of Nck-SH2 and disintegrated the 3pY/Nck/N-WASP condensates. In the presence of membrane-translocating peptide L17E, p1 entered Caco-2 cells in the cytosol, reduced the number of Nck puncta, and rendered Caco-2 cells resistant to EPEC infection. Site-selective covalent blockage of Nck thereby disintegrates intracellular Nck condensates, inhibits actin reorganization, and shuts down the entrance pathway of EPEC. This work showcases the promotion or inhibition of protein phase separation by synthetic peptides and the use of reactive peptides as LLPS disruptors and signal modulators.

Short abstract

A designed peptide selectively reacts with a single lysine residue on Nck, disrupts phase-separated signaling coacervates, shuts down an intracellular signal, and inhibits Escherichia coli infection.

1. Introduction

Severe or persistent diarrhea is the second leading cause of death in children younger than 5 years old, accounting for 1.3 million deaths annually, especially in low- and middle-income countries.1−3 Among all the diarrheagenic Escherichia coli pathotypes that cause diarrhea, enteropathogenic Escherichia coli (EPEC) is a major pathotype highly prevalent in the community and hospitals.4−7 The emerging multidrug-resistant strains also have drastically increased the difficulty of treating EPEC-caused infections.8−10 The mechanism of EPEC infection centers around signal transduction of the adaptor protein Nck. An EPEC bacterium first latches to the surface of an intestinal cell and injects one of the virulence factors, the translocated intimin receptor (Tir), into the cell. Tir is encoded by espE located on the locus of enterocyte effacement (LEE) pathogenicity island in EPEC strains. These steps result in an attaching and effacing (A/E) lesion between the EPEC bacterium and the intestine cell.6,11 Phosphorylation of the Tir protein at Tyr 474 induces its binding with the SH2 domain of Nck (Nck-SH2), which subsequently recruits Nck-binding proteins such as WIP and N-WASP and activates the Arp2/3 complex-dependent actin polymerization pathway in the host cells.12 Actin polymerization forms a pedestal underneath the bacterium and leads to the invasion of EPEC and other microbes, such as vaccinia13,14 and other vertebrate poxviruses.15 Nck proteins (Nck1 and Nck2, also called Nckβ and Grb4) contain four domains in tandem, including one SH2 domain and three SH3 domains. Nck-SH2 is known to bind to the phosphorylated peptide sequence on the Tir protein in EPEC, the viral membrane protein A36, and nephrin.12,16,17 Because of the high degree of sequence homology of Nck1 and Nck2, Nck1 (for simplicity, Nck hereinafter) was chosen in this work. Due to the essential role of the interaction between Tir and Nck-SH2 in EPEC infection, we envision that a peptide that blocks the protein-binding site of the Nck-SH2 domain will inhibit EPEC-mediated pedestal formation, which will block the entrance of EPEC and guard the cells against EPEC infection (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Schematic illustration summarizing the main discoveries of this work. (a) Schematic illustration showing that a trimeric pY-containing peptide promotes LLPS of Nck and N-WASP, and covalent and noncovalent Nck-SH2 blockers reverse the coacervate formation. (b) Proximity-induced lysine reaction of Nck-SH2 yields covalent SH2 blockers with a selected lysine. (c) Covalent and noncovalent Nck-SH2 inhibitors disintegrate Nck coacervates in cells and prevent EPEC invasion.

Biological molecules, such as proteins or nucleic acids, form biomolecular condensates or coacervates and function as membrane-less organelles in eukaryotic cells.18,19 The condensates are highly concentrated micron-sized liquid droplets formed by liquid–liquid phase separation (LLPS) driven by physical interactions.20 The changes in physical states of biomolecules are associated with the occurrence and prevention of some diseases, such as protein aggregates involved in neurodegenerative diseases, protein accumulation21 or transcription abnormal in cancer,22 and viral replication condensates in SARS-CoV-2 infection.23 Rosen and co-workers discovered that the adaptor protein Nck serves as a platform for a multivalent interaction and forms LLPS coacervates with the actin-regulatory protein called neural Wiskott–Aldrich syndrome protein (N-WASP) and phosphorylated nephrin.24−29 Nck phase transition is associated with the activity of an actin nucleation factor, the Arp2/3 complex, and is governed by the degree of nephrin phosphorylation. However, the role of Nck-mediated LLPS in infectious diseases and whether LLPS can derive druggable targets for treating bacterial infections remain largely untapped.

Because Nck-SH2 binds to phosphorylated Tir, and this interaction drives EPEC invasion, we reason that blocking Nck-SH2 will disintegrate the Nck signaling complex and shut down the Nck-mediated actin rearrangement. However, the SH2 domains in mammalian cells are highly promiscuous, so selective blocking of the Nck-SH2 domain among ∼120 SH2 domains in the proteome is difficult.30 Because the phosphorylated tyrosine (pY) residue provides approximately half of the total binding energy in the pY-peptide–SH2 complex, it is challenging to design high-affinity and high-selectivity blockers specifically for the Nck-SH2 domain while sparing other SH2 domains or pY-binding proteins in the cytoplasm. In previous work, we developed proximity-driven cysteine-selective reactions between pY-containing peptides with SH2 domains using the reaction between an α-chloroacetyl group and a cysteine. This allowed us to differentiate different SH2 domains and achieve high-affinity covalent inhibition of specific SH2 domains selectively.30 The site-selective cysteine reactions were also brought inside cells, leading to intracellular domain-specific irreversible inhibition of a PDZ domain and an SH3 domain and effective blockage of signal transduction inside cells and even in animals.31,32 These covalently reactive peptides may be precursors of covalent drugs.33−36 Here, we first validate that Nck forms phase-separated condensates with N-WASP and a pY peptide trimer (Figure 1a). Then, a structure-guided design leads to a reactive peptide that can covalently react with the Nck-SH2 domain at a selected lysine residue to interfere with the Nck/N-WASP LLPS (Figure 1b). We also report the intracellular delivery of this reactive peptide, the disintegration of Nck coacervates, and the blockage of EPEC infection in intestine cells (Figure 1c).

2. Results and Discussion

A. Peptide Promoter of Nck Phase Separation in Vitro

Nck, N-WASP, and p-nephrin are known to undergo liquid–liquid phase separation (LLPS).24−29 N-WASP consists of 5 domains: WH1 (WASP-homology 1), BR1 (a highly basic region), GBD (GTPase-binding domain), and proline-rich and VVCA (verprolin homology, central hydrophobic region, acidic region) domains. Here, the GBD-P-VCA tridomain of N-WASP was expressed and purified, and it was called N-WASP thereafter in this work.27 In the absence of p-nephrin, Nck and N-WASP formed micrometer-sized microdroplets in aqueous PBS buffer at very high concentrations, i.e., 40 μM of Nck and 60 μM of N-WASP (Figure 2a). The droplets showed liquid-like characteristics based on a fluorescence recovery assay after photobleaching (FRAP) assay (Figure 2b), proving to be coacervates driven by LLPS.

Figure 2.

Figure 2

Open in a new tab

Phosphopeptide 3pY promotes Nck/N-WASP phase separation. (a) Microscopic image showing Nck and N-WASP form droplets at high concentrations (Nck, 40 μM; N-WASP, 60 μM). (b) Recovery of fluorescence after bleaching, suggesting the Nck/N-WASP droplets have liquid-like properties. (c) Design of a 3pY peptide based on the sequence alignment of Tir and nephrin. The Nck-SH2 binding sequences are underscored, and the Tyr sites that are potentially phosphorylated are marked in red. (d) Macroscopic and microscopic images showing that 3pY promoted the microdroplet formation of Nck/N-WASP, whereas 3Y did not. Peptides, 10 μM; Nck, 20 μM; N-WASP, 10 μM. Scale bar: 20 μm. (e) Phase diagrams of protein condensate formation at a fixed 3pY concentration of 10 μM (left) and a fixed Nck concentration of 20 μM (right). (f) Time-lapse microscopy images showing the fusion of two microdroplets. Scale bar: 2 μm. (g) Recovery of fluorescence after bleaching suggesting Nck/N-WASP/3pY droplets have liquid-like properties. 3pY, 10 μM; Nck, 20 μM; N-WASP, 10 μM. Scale bar: 2 μm.

Phosphorylated Tir is known to bind to Nck-SH2. A 12-residue pY-containing Tir peptide EEHIpYDEVAADP was identified as a strong binder to Nck-SH2, and the crystal structure of the complex was reported.17 Also, the sequence of nephrin contains three Tyr-containing sequences that are similar to those of the Tir peptide: HLYDEV, PLYDEV, and GIYDQV (the potential phosphorylation sites are underscored), with a hydrophobic residue at the pY-1 site, and negatively charged residues at pY+1 and/or pY+2 (Figure 2c). Therefore, we reason that phosphorylation at these Tyr turns nephrin into p-nephrin, a multivalent Nck-SH2 binder, which drives the LLPS of Nck-SH2 and other binding proteins.27 To prove this hypothesis, we designed a trimeric phosphopeptide 3pY containing three LpYDEV linked through PEG linkers—all with L at pY-1 and D and E at pY+1 and pY+2 sites, to mimic the multivalent phosphorylated nephrin or Tir (Figure 2c). A nonphosphorylated version 3Y was also synthesized as a control. Adding 3pY to Nck and N-WASP turned the solution to opalescent, and micron-sized spherical droplets were visible under the microscope at significantly lower protein concentrations (Figure 2d). Under a time-lapse microscope, we observed that the microdroplets could fuse into larger condensates (Figure 2e). A phase diagram revealed that the minimal phase-forming concentrations of Nck, N-WASP, and 3pY were 10 μM, 5 μM, and 1 μM, respectively (Figure 2f). Next, we fluorescently labeled N-WASP with Cy3 and Nck with Cy5. The Cy3 and Cy5 signals colocalized in the Nck/N-WASP/3pY system under the fluorescent microscope (Figure S1 in the Supporting Information). The fluorescence recovery after photobleaching (FRAP) assay showed rapid fluorescence recovery after photobleaching the fluorescent signal within 30 s (Figure 2g). Taken together, these results show that 3pY promotes the condensation of Nck and N-WASP through the interaction with Nck-SH2 similar to that of p-nephrin, and the condensates formed in the Nck/N-WASP/3pY system have liquid-like properties, meeting the properties, of liquid–liquid phase separation (LLPS). We reason that the multivalent feature and two flexible PEG4 linkers are essential for 3pY to promote the LLPS of the Nck/N-WASP system.27

B. Peptide Inhibitor of Nck/N-WASP/3pY Coacervation

Next, we explored the intervention of Nck/N-WASP/3pY LLPS using pY peptides (Figure 3a). Three pY-containing peptides extracted from different Tir sequences were synthesized, namely, peptides Y474, Y751, and Y147 (Figure 3b). When peptide Y474 was added to the Nck/N-WASP/3pY system at a high concentration of 50 μM, protein microdroplets disappeared; such an effect was not observed when peptides Y751 and Y147 of the same concentration were added to the coacervates (Figure 3c). Using solution turbidity as a measure of coacervate formation, peptide Y474 caused a rapid decrease in the turbidity of the Nck/N-WASP/3pY solution. In contrast, the effect of the other two peptides was significantly less pronounced (Figure 3d). The inhibitory effect of peptide Y474 was also found to be concentration dependent: 50 μM Y474 caused more turbidity decrease than 5 μM y474 (Figure 3e). Also, peptide Y474 was found to bind with Nck-SH2 with a KD of 70 nM, based on the measurement by microscale thermophoresis (MST) (Figure S2 in the Supporting Information). These data show that monovalent Nck-SH2 binding peptides disrupt Nck LLPS at high concentrations by competing with multivalent binders for Nck-SH2 binding sites.

Figure 3.

Figure 3

Open in a new tab

pY peptides disrupt Nck/N-WASP/3pY LLPS. (a) Sequences of synthetic pY peptides. (b) Microscopic images showing the droplet disappearance after adding peptide Y474 (50 μM) to the Nck/N-WASP/3pY system. Nck: 20 μM; N-WASP: 10 μM; 3pY: 10 μM. Scale bars, 20 μm. (c) Peptide Y474 rapidly decreases the turbidity of the Nck/N-WASP/3pY solution in around 20 min. Nck: 20 μM, N-WASP: 10 μM, 3pY: 10 μM. (d) 50 μM Y474 more effectively decreased the turbidity of the Nck/N-WASP/3pY solution than 5 μM Y474. Nck: 20 μM, N-WASP: 10 μM, 3pY: 10 μM.

C. Reactive Peptides Based on Proximity-Induced Lysine Reaction

Although peptide Y474 can block Nck-mediated LLPS, the requirement of a high concentration of the blocker peptide is hard to achieve inside the cell. We then sought to design covalent, irreversible blockers, anticipating that, at a significantly lower concentration, the covalent blocker can effectively inhibit Nck coacervation in the cell. The cocrystal structure of the Tir peptide and Nck-SH2 (PDB ID: 2CI9) shows that several lysine residues of Nck (including Lys328, Lys331, and Lys369) form a positively charged surface to host negatively charged Glu and Asp at pY+1 and pY+2 sites of the ligand EEHIpYDEVAAD (shown in red) (Figure 4a). The side chains of the Nck lysine residues are pointing toward the residues at pY+1 and pY+2 positions with distances of only 4 to 6 Å. This unique feature gives us an opportunity to design covalent blockers for Nck-SH2 by installing electrophiles at the pY+1 or pY+2 positions to enable a proximal conjugation reaction with one of the lysine residues. Covalent blockage through the characteristic lysine residues will maximize the specificity of the inhibitor toward Nck-SH2. As sulfonyl fluoride has been reported to react with proteinaceous nucleophiles such as lysine, histidine, or cysteine,34 we installed a sulfonyl fluoride group on the side chain of diaminopropionic acid, giving an unnatural amino acid X1 at different positions such as pY+1 (Asp6) and pY+3 (Val8) of peptide EEHIpYDEVAAD. Besides the amino acid X1, we also included α-chloroacetyl-carrying amino acids X2 and X3, which may also react with proteineous nucleophiles. Electrophile-containing reactive peptides p1 to p11 were synthesized with a biotin tag at the N-termini (Figure 4b). After the reactive peptides were incubated with recombinant Nck protein at a 5:1 ratio at 37 °C in PBS for 1 h, the formation of covalently linked Nck–peptide conjugates could be observed, shown as new protein bands with molecular weights higher than that of Nck based on denatured SDS-PAGE and Western blotting analysis against biotin (Figure 4c). The SDS-PAGE gel showed that among all the peptides, p1 gave the highest cross-linking efficiency with Nck, with >90% of Nck protein converted to conjugates. Even though the ratio of p1 to Nck is 5 fold in excess, only one Nck–peptide conjugate band was observed in the SDS-PAGE; the 1:1 reaction ratio suggests that the reaction most likely happens at the peptide-binding groove of Nck-SH2, i.e., with the projected lysine residues. All the peptides carrying sulfonyl fluoride (p1, p5, p7, p10, and p11) showed noticeable reactivity with Nck, whereas among the chloroacetyl-containing peptides, only p2 and p9 showed reactivity. Microscale thermophoresis (MST) analysis also showed that p1 outperformed p7 (as one example), giving a significantly lower EC50 value, 0.75 vs 6.2 μM respectively (Figure S3 in the Supporting Information). At 37 °C in PBS buffer, the 1:1 reaction efficiency reached 50% in 2 min (Figure S4 in the Supporting Information). These data show that peptides with side-chain-modified sulfonyl fluoride are viable reactive peptides for Nck, and the Nck-peptide binding induces a covalent cross-linking reaction between Nck and the irreversible blocker.

Figure 4.

Figure 4

Open in a new tab

Covalently reactive peptides designed based on a Tir peptide. (a) Crystal structure of Nck-SH2 in complex with a peptide EEHIpYDEVAAD from EPEC protein Tir (PDB: 2CI9). (b) List of synthetic reactive peptides. (c) Unnatural amino acids in the peptides. (d) SDS-PAGE (top) and Western blotting (bottom) analyses showing covalent conjugation of the peptides with Nck. Briefly, biotin-labeled peptides were mixed with recombinant Nck protein at a peptide-to-protein ratio of 5:1 at 37 °C for 1 h before denaturation and resolution by SDS-PAGE.

D. Mass Spectrometric Evidence for a Single-Site Lysine Reaction in Nck-SH2

We next identified the residues in Nck-SH2 that are involved in the reaction with p1 by mass spectrometry. LC-MS analysis first confirmed that only one p1 was conjugated with Nck (Figure S5 in the Supporting Information). The Nck–p1 conjugate was digested with trypsin, and peptide fragments were analyzed with an in-line EASY-spray source and nano-LC UltiMate 3000 HPLC system interfaced with an Elite mass spectrometer, operated in the data-dependent acquisition (DDA) mode with one full MS scan at R = 60,000 (m/z = 200) mass followed by HCD MS/MS scans. A new peak corresponding to the precursor mass (Mr + 3H)3+ of biotinylated p1 peptide (α-chain) conjugated with Nck-SH2 (β-chain) was observed at 2631.1402 Da after searching by pLink v2.3.11.37 The fragmentation spectral analysis of [EEHIpYXEVAAD(α)] + [HFKVQLK(β)] extracted by pLabel v2.4.3.0 (Figure 5a and Figure S6 in the Supporting Information)38,39 confirmed the ligand-directed conjugation of Nck-SH2 protein with p1 peptide. The underlined lysine residue in the β-chain sequence indicated the cross-linked site on the predicted Nck Lys331 residue. Next, we mutated Nck Lys331 to Ala, expressed and purified the Nck-SH2K331A mutant. In contrast to wild-type Nck-SH2 and Nck, Nck-SH2K331A did not show noticeable reactivity with peptide p1 (Figure 5b). These data identify Lys331 as a reaction site for p1 on Nck-SH2, suggesting that the proximity-driven lysine reaction occurred in a residue-selective manner.

Figure 5.

Figure 5

Open in a new tab

Peptide p1 reacts with Lys 331. (a) MS/MS analysis of the conjugated site between p1 peptide and Nck-SH2 at Lys331. The corresponding mass data are shown in Figure S6 in the Supporting Information. (b) Mutation of Lys331 to Ala abolished the covalent reaction of Nck with p1. Purified recombinant proteins were incubated with peptides in PBS buffer at 37 °C for 1 min (left) or 10 min (right). Molecular weight markers are shown in the middle. Red arrows indicate the covalent peptide–protein complex.

In addition, we performed the reaction between p1 and Nck in a cell lysate, representing a complex solution. We incubated peptide p1 with the lysate of Caco-2 cells overexpressing Nck, and the reaction mixture was resolved on SDS-PAGE and analyzed with Western blotting against an anti-Nck antibody. A complete band shift was observed in the reaction system with p1, whereas for other peptides only a partial band shift was observed (Figure S7 in the Supporting Information). Next, we incubated biotinylated peptide Y474 and Nck solutions with or without p1 peptide, and pulled down proteins that bound with biotinylated Y474. In the absence of p1, we clearly observed that Nck bound with Y474. However, in the presence of p1, the Nck protein did not bind with Y474, suggesting that p1 competitively blocked the peptide-binding site of Nck-SH2 (Figure S8 in the Supporting Information). These data show that p1 competitively blocks the protein binding site of Nck-SH2.

E. Nck-p1 Reaction Disrupts LLPS

To explore whether intracellularly delivered peptide p1 could disrupt the Nck-mediated phase separation in mammalian cells (Figure 6a), we first examined whether Nck can form coacervates with Nck in mammalian cells. Confocal microscopic images of HeLa cells overexpressing fluorescently tagged Dsred-Nck and GFP-N-WASP revealed the formation of micrometer-sized puncta with colocalized Dsred and GFP signals (Figure 6b). An intracellular FRAP analysis using a home-built two-photon fluorescence microscope at room temperature revealed the fluidlike property of the puncta, showing they are not precipitates (Figure 6c). After photobleaching the GFP fluorescence in puncta to 40% of its original fluorescence intensity, gradual fluorescence recovery to 70% in 50 s was observed. This shows that the Nck/N-WASP condensates are dynamic phase-separated droplets.

Figure 6.

Figure 6

Open in a new tab

Peptide p1 inhibits the formation of the Nck/N-WASP condensate in mammalian cells. (a) Schematic illustration showing inhibition of intracellular Nck coacervation by p1. (b) Confocal microscopic images showing the formation of Nck/N-WASP puncta in HeLa cells that overexpress GFP-N-WASP and dsRed-Nck. Scale bar: 10 μm. (c) FRAP analysis showing the Nck/N-WASP condensates in HeLa cells have fluidity properties based on the recovery of the GFP fluorescent signal after photobleaching in a two-photon microscope with excitation at 920 nm. Scale bar: 5 μm. (d) Confocal microscopic images showing that L17E/p1-treatment reduced the numbers of Nck punta in Caco-2 cells transfected to overexpress dsRed–Nck. L17E, 40 μM; p1, 1 μM. Statistical analyses of puncta number per cell in the control and experiment (L17E/p1). N = 20 cells.

Delivering p1 into mammalian cells across the plasma membrane is a challenge, as the negatively charged p1 can not spontaneously enter mammalian cells. After exploring several delivery systems, we finally selected the L17E peptide to deliver p1 into mammalian cells. The lipid-sensitive endosomolytic peptide L17E (IWLTALKFLGKHAAKHEAKQQLSKL-amide) was discovered based on peptide screening and was found to bring cargo into the cytosol through endosomal escape by sampling mixing with the cargo.40 Fluorescently labeled p1 was efficiently delivered into Caco-2 cells by mixing with L17E, and the fluorescence signal was distributed in the cytosol (Figure S9 in the Supporting Information). We also confirmed that L17E-delivered p1 did not cause noticeable cytotoxicity based on the CCK8 assay (Figure S10 in the Supporting Information). Western blotting using an Nck antibody identified an Nck complex with higher molecular weight in the lysate of L17E/p1-treated cells, showing L17E-delivered p1 reacted with the endogenous Nck in the cytosol of Caco-2 cells (Figure S11 in the Supporting Information). The extent of the molecular weight increase in this case was lower than that in vitro, likely due to the intracellular degradation of the p1 peptide. Despite the fact that p1 may react with a range of nucleophiles in the cells, here we provide evidence of the reaction between endogenous Nck and p1, which paves the way for the following function-based experiments. When Caco-2 cells overexpressing dsRed-Nck were treated with the p1 peptide (1 μM) and L17E (40 μM) for 2 h, we found a 50% decline in puncta numbers in the cells, showing a significant decrease of phase-separated Nck condensates (Figure 6d). This data show that Nck and N-WASP form phase-separated condensates in cotransfected HeLa cells, and L17E-delivered p1 peptide may disrupt the Nck/N-WASP coacervates.

F. L17E/p1 Protects Cells against EPEC Infection

Lastly, to examine whether pretreatment of Caco-2 cells with L17E/p1 could prevent EPEC infection, we first established a bacterial infection assay (Figure 7a). Caco-2 cells were pretreated with L17E/p1 for 2 h. After excess peptides were removed, Caco-2 cells were then infected with EPEC at different multiplicity of infection (MOI) values (MOI is commonly defined as the ratio of infectious bacteria to cells in a culture) for 3 h, and extra bacterial cells were removed from the culture. Next, gentamicin was added to the cell culture to remove EPEC that are merely attached to the surface of Caco-2 cells, leaving only engulfed EPEC inside Caco2 cells to be counted (gentamicin treatment will not kill intracellular bacteria due to its cellular impermeability). After gentamicin was washed away, cells were lysed to release EPEC cells inside, and the cell lysates were diluted to count the number of EPEC cells as colony-forming units (cfu) on agar plates. L17E/p1 treatment reduced the CFU numbers in a dose-dependent manner, with 5 μM of p1 reducing about 50% of the EPEC infection (Figure 7b). Comparing the covalent blocker p1 and the noncovalent blocker Y474 peptide, both blockers can disintegrate phase-separated Nck droplets and reduce the turbidity of the solution (Figure S12 and S13 in the Supporting Information). Although both peptides, in conjunction with L17E-mediated delivery, can block EPEC invasion to Caco-2 cells, covalent blocker p1 showed significantly higher efficacy than Y474 at lower concentrations (Figure 7c).

Figure 7.

Figure 7

Open in a new tab

L17E/p1 treatment safeguards Caco-2 cells against EPEC infection. (a) Schematic illustration showing the procedure of the infection assay. (b) Anti-infection efficacy of L17E/p1 treatment at different concentrations of p1 using CFU as an indicator for infection. (c) Comparison of p1 and Y474 peptides in the anti-infection assay. (d) Representative microscopic images of L17E/p1-treated Caco-2 cells showing the absence of actin pedestals, whereas in the absence of p1 peptide, EPEC induces actin pedestal formation to attach to Caco-2 cells. The concentration of L17E in all of the experiments was 40 μM.

Lastly, to confirm that the inhibitors conferred resistance through actin rearrangement and pedestal formation, Tir and actin were visualized in inhibitor-treated cells. Briefly, infected Caco-2 cells were fixed with paraformaldehyde, stained with immunofluorescent anti-Tir antibody, and treated with phalloidin labels actin. EPEC cells were found to be located on the surface of Caco-2 cells and supported by the actin network, showing that EPEC infection induced actin pedestal formation (Figure 7d).36 Treatment with L17E/p1 significantly reduced the number of EPEC cells that attach to the surface of Caco-2: nearly no noticeable EPEC cells can be found under the microscope. Also, actin is evenly distributed on the surface of Caco-2 cells without signs of forming pedestals. These data show that L17E-delivered p1 prevented the invasion of EPEC into Caco-2 cells by inhibiting Nck-mediated pedestal formation.

3. Discussion

Nck bridges the externally introduced EPEC virulence factor Tir with a downstream actin rearrangement signal in an intestine cell and makes the cell a recipient of EPEC. To do so, Nck utilizes its four protein-binding domains, three SH3 and one SH2, to bind with designated proteins upstream and downstream and form a signaling complex. Our work shows that the signaling complex formed between Nck and N-WASP may exist as phase-separated condensates, where phosphorylated nephrin (mimicked by a synthetic peptide 3pY here) promotes the Nck/N-WASP phase separation and reduces the minimal protein concentrations required for phase separation. Based on this observation, we designed noncovalent and covalently reactive peptides to block the pY-binding site of Nck-SH2. A sulfonyl fluoride-containing pY peptide p1 covalently reacted with Nck-SH2 at Lys331 and effectively blocked the Nck-mediated phase separation. Through an L17E delivery system, the covalent blocker p1 reversed the phase separation of Nck and N-WASP and disallowed EPEC from entering Caco-2 cells. Unlike 3pY, which promoted the LLPS of Nck and N-WASP by providing multivalent interaction sites and flexible internal PEG4 linkers, Y474-derived monovalent binders at high concentrations competed with the multivalent binders and disrupted the LLPS.

Blockage of a specific SH2 domain is challenging because the binding of pY to the pY-binding pockets on the SH2 domains contributes to about 50% of the binding energy, which makes SH2–peptide binding to be promiscuous. In a previous report, we identified SH2 domains that have special cysteine residues at the peptide-binding site and developed reactive peptides containing an α-chloroacetyl group at selected sites that are adjacent to the cysteine residue. Proximity-driven reactivity enables selective reaction at the cysteine.30 Here, we show a proximity-driven lysine reaction at Lys331 of the Nck-SH2 domain. Covalently reactive peptide p1 showed high reactivity with Nck-SH2 both in the cell lysate in vitro and in the cytosol. Admittedly, the sulfonyl fluoride on p1 also reacts with other nucleophiles inside cells, but these side reactions did not cause noticeable changes in the cell physiology or toxicity. On another note, delivery of the reactive peptide p1 to the cytosol while retaining its reactivity is not trivial. Covalent linkage of p1 with positively charged cell-penetrating peptides invalidated the activity of p1 possibly because of shielding of the pY residue (data not shown). Importantly, the L17E delivery system does not need a covalent linkage with p1, and the endosomolytic activity of L17E allows p1 to escape endosome entrapment and be distributed in the cytosol. However, the drawback of this delivery method is that it is not economical due to the high concentration of L17E peptide required in the delivery: For example, for the delivery of the 1 μM p1 peptide, 40 μM L17E is required.

Targeting the phase separation property of the signaling protein complex is emerging as a possible way to interfere with cellular signal transduction. Condensate-modifying therapeutics (c-mods), including peptides and small molecules, have been designed to alter condensate behaviors with functional consequences in cell-based studies. For example, Zhou and co-workers identified a peptide targeting the dimerization domain of the SARS-CoV-2 nucleocapsid protein (SARS2-NP) and found it disrupted the SARS2-NP LLPS and inhibited virus replication.41 Instead of directly acting on bacterial proteins, the inhibitory blocker in our approach targets an adaptor protein inside the cytosol of the host cells. Blockage of the Nck-SH2 binding site inhibits the Nck LLPS, shuts down Nck-mediated signal transduction, and inhibits bacterial infection. The inhibition of the Nck-SH2 binding site directly results in two events: the disintegration of Nck coacervates and the blockage of EPEC infection. According to Banjade and Rosen, p-nephrin/Nck/N-WASP condensates promote Arp2/3 complex-dependent actin assembly.28 Arp2/3-dependent actin polymerization is also known to play a critical role in EPEC infection.16 Therefore, disruption of the Nck condensates may affect Arp2/3-dependent actin polymerization and subsequently inhibit EPEC infection. Although direct evidence for the linkage between these events is yet to be acquired, modulating the LLPS of adaptor proteins by blocking essential protein–protein interactions in the host cells may provide a new antibacterial mechanism against the infection of EPEC.

On another note, unlike p-nephrin, the role of phosphorylated Tir (p-Tir) in the Nck LLPS is yet unclear, as neither 3pY nor peptide Y474 represents the membrane-bound protein p-Tir. Because peptide Y474 at low concentrations did not disintegrate the Nck/N-WASP/3pY coacervates, it is possible that the membrane-bound p-Tir (with the C-terminal Y474 phosphorylated) at low concentrations may join the Nck/N-WASP/p-nephrin coacervates. Banjade and Rosen’s work suggests that intimin on the surface of EPEC may cluster p-Tir and promote the formation of Nck coacervates,28 which consequently leads to pedestal formation during the EPEC infection. The role of p-Tir clearly needs more investigation.

Acknowledgments

We acknowledge the assistance from Prof. Michael K. Rosen (University of Texas, Southwestern Medical Center at Dallas, UTSW) with the expression and purification of N-WASP protein. This work was partially funded by the Research Grants Council of Hong Kong (GRF Grants N_CUHK422/18, 14304320 and 14306222), and internal grants from CUHK (Collaborative Research Impact Matching Scheme Grant and Impact Case Seed Grant).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscentsci.3c01068.

  • Methods and experimental procedures; colocalization of Cy3-N-WASP and Cy5-Nck in Nck/N-WASP/3pY microdroplets; microscale thermophoresis (MST) experiment to measure the binding of Y474 to Nck; MST experiment measured the EC50 of peptide 1 and 7; reaction kinetics; LC-MS analysis of Nck and Nck-p1 conjugate; mass data of the selected peaks in Figure 5a; reaction of biotin-p1 to cell lysate; a competitive pull-down experiment to show Nck-p1 complex failed to bind to peptide Y474; L17E delivers p1 peptide into the cytosol of Caco-2 cells; cell viability analysis by CCK8 assay; in cell reaction; peptide p1 reverses the phase separation of Nck/N-WASP/3pY; peptide Y474 reverses the phase separation of Nck/N-WASP/3pY; chemical structure of all the peptides and the mass spectrum data (PDF)

Author Contributions

Min Liu: Methodology, Investigation, Data curation, Formal analysis, Writing - original draft. Chunjian Wu: Methodology, Investigation, Data curation. Rui Wang: Methodology, Investigation, Data curation. Jiaming Qiu: Investigation. Zhentao She: Investigation. Jianan Qu: Supervision. Jiang Xia: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing - review and editing.

Author Contributions

⊥ M.L. and C.W. contributed equally to this work.

The authors declare no competing financial interest.

Supplementary Material

oc3c01068_si_001.pdf (2.9MB, pdf)

References

  1. Black R. E.; Cousens S.; Johnson H. L.; Lawn J. E.; Rudan I.; Bassani D. G.; Jha P.; Campbell H.; Walker C. F.; Cibulskis R.; Eisele T.; Liu L.; Mathers C. Global, Regional, and National Causes of Child Mortality in 2008: A Systematic Analysis. Lancet. 2010, 375 (9730), 1969–1987. 10.1016/S0140-6736(10)60549-1. [DOI] [PubMed] [Google Scholar]
  2. Abba K.; Sinfield R.; Hart C. A.; Garner P. Pathogens Associated with Persistent Diarrhoea in Children in Low and Middle Income Countries: Systematic Review. BMC Infect Dis. 2009, 9 (1), 88. 10.1186/1471-2334-9-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ochoa T. J.; Contreras C. A. Enteropathogenic Escherichia Coli Infection in Children. Curr. Opin Infect Dis. 2011, 24 (5), 478. 10.1097/QCO.0b013e32834a8b8b. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Biswas R.; Nelson E. A.; Lewindon P. J.; Lyon D. J.; Sullivan P. B.; Echeverria P. Molecular Epidemiology of Escherichia Coli Diarrhea in Children in Hong Kong. J. Clin Microbiol. 1996, 34 (12), 3233–3234. 10.1128/jcm.34.12.3233-3234.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Levine M.; Nalin D.; Hornick R.; Bergquist E.; Waterman D.; Young C.; Sotman S.; Rowe B. Escherichia coli strains that cause diarrhoea but do not produce heat-labile or heat-stable enterotoxins and are non-invasive. Lancet 1978, 311 (8074), 1119–1122. 10.1016/S0140-6736(78)90299-4. [DOI] [PubMed] [Google Scholar]
  6. Kaper J. B.; Nataro J. P.; Mobley H. L. T. Pathogenic Escherichia Coli. Nat. Rev. Microbiol. 2004, 2 (2), 123–140. 10.1038/nrmicro818. [DOI] [PubMed] [Google Scholar]
  7. Mare A. D.; Ciurea C. N.; Man A.; Tudor B.; Moldovan V.; Decean L.; Toma F. Enteropathogenic Escherichia Coli—A Summary of the Literature. Gastroenterology. 2021, 12 (1), 28–40. 10.3390/gastroent12010004. [DOI] [Google Scholar]
  8. Hoban D. J.; Bouchillon S. K.; Hawser S. P.; Badal R. E. Trends in the Frequency of Multiple Drug-Resistant Enterobacteriaceae and Their Susceptibility to Ertapenem, Imipenem, and Other Antimicrobial Agents: Data from the Study for Monitoring Antimicrobial Resistance Trends 2002 to 2007. Diagn Microbiol Infect Dis. 2010, 66 (1), 78–86. 10.1016/j.diagmicrobio.2009.06.009. [DOI] [PubMed] [Google Scholar]
  9. Oteo J.; Pérez-Vázquez M.; Campos J. Extended-Spectrum β-Lactamase Producing Escherichia Coli: Changing Epidemiology and Clinical Impact. Curr. Opin Infect Dis. 2010, 23 (4), 320. 10.1097/QCO.0b013e3283398dc1. [DOI] [PubMed] [Google Scholar]
  10. Spellberg B.; Bartlett J. G.; Gilbert D. N. The Future of Antibiotics and Resistance. N Engl J. Med. 2013, 368 (4), 299–302. 10.1056/NEJMp1215093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Croxen M. A.; Law R. J.; Scholz R.; Keeney K. M.; Wlodarska M.; Finlay B. B. Recent Advances in Understanding Enteric Pathogenic Escherichia Coli. Clin Microbiol Rev. 2013, 26 (4), 822–880. 10.1128/CMR.00022-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gruenheid S.; DeVinney R.; Bladt F.; Goosney D.; Gelkop S.; Gish G. D.; Pawson T.; Finlay B. B. Enteropathogenic E. Coli Tir Binds Nck to Initiate Actin Pedestal Formation in Host Cells. Nat. Cell Biol. 2001, 3 (9), 856–859. 10.1038/ncb0901-856. [DOI] [PubMed] [Google Scholar]
  13. Snapper S. B.; Takeshima F.; Antón I.; Liu C.-H.; Thomas S. M.; Nguyen D.; Dudley D.; Fraser H.; Purich D.; Lopez-Ilasaca M.; Klein C.; Davidson L.; Bronson R.; Mulligan R. C.; Southwick F.; Geha R.; Goldberg M. B.; Rosen F. S.; Hartwig J. H.; Alt F. W. N-WASP Deficiency Reveals Distinct Pathways for Cell Surface Projections and Microbial Actin-Based Motility. Nat. Cell Biol. 2001, 3 (10), 897–904. 10.1038/ncb1001-897. [DOI] [PubMed] [Google Scholar]
  14. Weisswange I.; Newsome T. P.; Schleich S.; Way M. The Rate of N-WASP Exchange Limits the Extent of ARP2/3-Complex-Dependent Actin-Based Motility. Nature. 2009, 458 (7234), 87–91. 10.1038/nature07773. [DOI] [PubMed] [Google Scholar]
  15. Dodding M. P.; Way M. Nck- and N-WASP-Dependent Actin-Based Motility Is Conserved in Divergent Vertebrate Poxviruses. Cell Host Microbe. 2009, 6 (6), 536–550. 10.1016/j.chom.2009.10.011. [DOI] [PubMed] [Google Scholar]
  16. Donnelly S. K.; Weisswange I.; Zettl M.; Way M. WIP Provides an Essential Link between Nck and N-WASP during Arp2/3-Dependent Actin Polymerization. Curr. Biol. 2013, 23 (11), 999–1006. 10.1016/j.cub.2013.04.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Frese S.; Schubert W.-D.; Findeis A. C.; Marquardt T.; Roske Y. S.; Stradal T. E. B.; Heinz D. W. The Phosphotyrosine Peptide Binding Specificity of Nck1 and Nck2 Src Homology 2 Domains. J. Bio Chem. 2006, 281 (26), 18236–18245. 10.1074/jbc.M512917200. [DOI] [PubMed] [Google Scholar]
  18. Wang B.; Zhang L.; Dai T.; Qin Z.; Lu H.; Zhang L.; Zhou F. Liquid-Liquid Phase Separation in Human Health and Diseases. Signal Transduct Target Ther. 2021, 6 (1), 1–16. 10.1038/s41392-021-00678-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lyon A. S.; Peeples W. B.; Rosen M. K. A Framework for Understanding the Functions of Biomolecular Condensates across Scales. Nat. Rev. Mol. Cell Biol. 2021, 22 (3), 215–235. 10.1038/s41580-020-00303-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Banani S. F.; Lee H. O.; Hyman A. A.; Rosen M. K. Biomolecular Condensates: Organizers of Cellular Biochemistry. Nat. Rev. Mol. Cell Biol. 2017, 18 (5), 285–298. 10.1038/nrm.2017.7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ambadipudi S.; Biernat J.; Riedel D.; Mandelkow E.; Zweckstetter M. Liquid-Liquid Phase Separation of the Microtubule-Binding Repeats of the Alzheimer-Related Protein Tau. Nat. Commun. 2017, 8 (1), 275. 10.1038/s41467-017-00480-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Xie J.; He H.; Kong W.; Li Z.; Gao Z.; Xie D.; Sun L.; Fan X.; Jiang X.; Zheng Q.; Li G.; Zhu J.; Zhu G. Targeting Androgen Receptor Phase Separation to Overcome Antiandrogen Resistance. Nat. Chem. Biol. 2022, 18 (12), 1341–1350. 10.1038/s41589-022-01151-y. [DOI] [PubMed] [Google Scholar]
  23. Kliche J.; Kuss H.; Ali M.; Ivarsson Y. Cytoplasmic Short Linear Motifs in ACE2 and Integrin Β3 Link SARS-CoV-2 Host Cell Receptors to Mediators of Endocytosis and Autophagy. Sci. Signal. 2021, 14 (665), eabf1117 10.1126/scisignal.abf1117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Banjade S.; Wu Q.; Mittal A.; Peeples W. B.; Pappu R. V.; Rosen M. K. Conserved Interdomain Linker Promotes Phase Separation of the Multivalent Adaptor Protein Nck. Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (47), E6426-E6435 10.1073/pnas.1508778112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Case L. B.; De Pasquale M.; Henry L.; Rosen M. K. Synergistic Phase Separation of Two Pathways Promotes Integrin Clustering and Nascent Adhesion Formation. eLife 2022, 11, e72588 10.7554/eLife.72588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kim S.; Kalappurakkal J. M.; Mayor S.; Rosen M. K. Phosphorylation of Nephrin Induces Phase Separated Domains That Move through Actomyosin Contraction. Mol. Biol. Cell 2019, 30 (24), 2996–3012. 10.1091/mbc.E18-12-0823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li P.; Banjade S.; Cheng H.-C.; Kim S.; Chen B.; Guo L.; Llaguno M.; Hollingsworth J. V.; King D. S.; Banani S. F.; Russo P. S.; Jiang Q.-X.; Nixon B. T.; Rosen M. K. Phase Transitions in the Assembly of Multivalent Signalling Proteins. Nature. 2012, 483 (7389), 336–340. 10.1038/nature10879. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Banjade S.; Rosen M. K. Phase Transitions of Multivalent Proteins Can Promote Clustering of Membrane Receptors. Elife 2014, 3, e04123 10.7554/eLife.04123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Case L. B.; Zhang X.; Ditlev J. A.; Rosen M. K. Stoichiometry Controls Activity of Phase-Separated Clusters of Actin Signaling Proteins. Science. 2019, 363 (6431), 1093–1097. 10.1126/science.aau6313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Wang R.; Leung P. Y. M.; Huang F.; Tang Q.; Kaneko T.; Huang M.; Li Z.; Li S. S. C.; Wang Y.; Xia J. Reverse Binding Mode of Phosphotyrosine Peptides with SH2 Protein. Biochemistry. 2018, 57 (35), 5257–5269. 10.1021/acs.biochem.8b00677. [DOI] [PubMed] [Google Scholar]
  31. Yu Y.; Liu M.; Ng T. T.; Huang F.; Nie Y.; Wang R.; Yao Z.-P.; Li Z.; Xia J. PDZ-Reactive Peptide Activates Ephrin-B Reverse Signaling and Inhibits Neuronal Chemotaxis. ACS Chem. Biol. 2016, 11 (1), 149–158. 10.1021/acschembio.5b00889. [DOI] [PubMed] [Google Scholar]
  32. Yu Y.; Nie Y.; Feng Q.; Qu J.; Wang R.; Bian L.; Xia J. Targeted Covalent Inhibition of Grb2-Sos1 Interaction through Proximity-Induced Conjugation in Breast Cancer Cells. Mol. Pharmaceutics 2017, 14 (5), 1548–1557. 10.1021/acs.molpharmaceut.6b00952. [DOI] [PubMed] [Google Scholar]
  33. Singh J.; Petter R. C.; Baillie T. A.; Whitty A. The Resurgence of Covalent Drugs. Nat. Rev. Drug Discovery 2011, 10 (4), 307–317. 10.1038/nrd3410. [DOI] [PubMed] [Google Scholar]
  34. Li Q.; Chen Q.; Klauser P. C.; Li M.; Zheng F.; Wang N.; Li X.; Zhang Q.; Fu X.; Wang Q.; Xu Y.; Wang L. Developing Covalent Protein Drugs via Proximity-Enabled Reactive Therapeutics. Cell 2020, 182 (1), 85–97. 10.1016/j.cell.2020.05.028. [DOI] [PubMed] [Google Scholar]
  35. Paulussen F. M.; Grossmann T. N. Peptide-Based Covalent Inhibitors of Protein-Protein Interactions. J. Pept Sci. 2023, 29 (1), e3457 10.1002/psc.3457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Phillips N.; Hayward R. D.; Koronakis V. Phosphorylation of the Enteropathogenic E. Coli Receptor by the Src-Family Kinase c-Fyn Triggers Actin Pedestal Formation. Nat. Cell Biol. 2004, 6 (7), 618–625. 10.1038/ncb1148. [DOI] [PubMed] [Google Scholar]
  37. Chen Z.; Meng J.; Cao Y.; Yin J.; Fang R.; Fan S.; Liu C.; Zeng W.; Ding Y.; Tan D.; Wu L.; Zhou W.; Chi H.; Sun R.; Dong M.; He S. A High-Speed Search Engine PLink 2 with Systematic Evaluation for Proteome-Scale Identification of Cross-Linked Peptides. Nat. Commun. 2019, 10 (1), 3404. 10.1038/s41467-019-11337-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wang L.; Li D.; Fu Y.; Wang H.; Zhang J.; Yuan Z.; Sun R.; Zeng R.; He S.; Gao W. pFind 2.0: a software package for peptide and protein identification via tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2007, 21 (18), 2985–91. 10.1002/rcm.3173. [DOI] [PubMed] [Google Scholar]
  39. Li D.; Fu Y.; Sun R.; Ling C.; Wei Y.; Zhou H.; Zeng R.; Yang Q.; He S.; Gao W. pFind: a novel database-searching software system for automated peptide and protein identification via tandem mass spectrometry. Bioinformatics. 2005, 21 (13), 3049–50. 10.1093/bioinformatics/bti439. [DOI] [PubMed] [Google Scholar]
  40. Akishiba M.; Takeuchi T.; Kawaguchi Y.; Sakamoto K.; Yu H.-H.; Nakase I.; Takatani-Nakase T.; Madani F.; Gräslund A.; Futaki S. Cytosolic Antibody Delivery by Lipid-Sensitive Endosomolytic Peptide. Nat. Chem. 2017, 9 (8), 751–761. 10.1038/nchem.2779. [DOI] [PubMed] [Google Scholar]
  41. Wang S.; Dai T.; Qin Z.; Pan T.; Chu F.; Lou L.; Zhang L.; Yang B.; Huang H.; Lu H.; Zhou F. Targeting Liquid-Liquid Phase Separation of SARS-CoV-2 Nucleocapsid Protein Promotes Innate Antiviral Immunity by Elevating MAVS Activity. Nat. Cell Biol. 2021, 23 (7), 718–732. 10.1038/s41556-021-00710-0. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

oc3c01068_si_001.pdf (2.9MB, pdf)

Articles from ACS Central Science are provided here courtesy of American Chemical Society

RESOURCES