RNA selectively modulates activity of virulent amyloid PSMα3 and host-defense LL-37 via phase separation and aggregation dynamics
- Department of Biology, Technion - Israel Institute of Technology, Israel
- German Center for Neurodegenerative Diseases, Germany
- CSSB Centre for Structural Systems Biology, Deutsches Elektronen-Synchrotron DESY, Germany
- The Life Sciences and Engineering Infrastructure Center, Technion-Israel Institute of Technology, Israel
- Ilse Katz Institute for Nanoscale Science and Technology, Ben Gurion University of the Negev, Israel
- Department of Biotechnology and Biomedicine, Technical University of Denmark DTU, Denmark
- Department for NMR-based Structural Biology, Max Planck Institute for Multidisciplinary Sciences, Germany
- Department of Neurology, University Medical Center Göttingen, Germany
- The Center for Experimental Medicine, Universitätsklinikum Hamburg-Eppendorf (UKE), Germany
- European Molecular Biology Laboratory (EMBL), Germany
Figures
Figure 1 with 2 supplements
The binding and effect of single- and double-stranded RNA on PSMα3.
EMSA illustrating the interaction between PSMα3 and RNA. The assay compares the effects of increasing concentrations of PSMα3 (0, 40, 80, 160, and 320 µM), shown in a gradient from left to right, on the mobility shift of single-stranded Poly(A) RNA (left panel) and double-stranded Poly(AU) RNA (right panel) at 40 nM. EMSA was performed in four replicates (n=4). Quantification is shown in Figure 1—figure supplement 1.
-
Figure 1—source data 1
PDF file containing the full, uncropped EMSA gels for Figure 1, with the Figure 1 panels boxed and the free and shifted (bound) RNA bands labeled.
- https://cdn.elifesciences.org/articles/109290/elife-109290-fig1-data1-v1.zip
-
Figure 1—source data 2
Original raw image files (unedited IRDye 800CW TIFF scans) for the EMSA gels displayed in Figure 1, including a README describing the lane order and indicating which gel in each scan appears in Figure 1.
- https://cdn.elifesciences.org/articles/109290/elife-109290-fig1-data2-v1.zip
Figure 1—figure supplement 1
Concentration-dependent binding of PSMα3 to Poly(AU) and Poly(A) RNA.
Percent RNA bound was quantified as a function of PSMα3 concentration for Poly(AU) (A) and Poly(A) (B) calculated from the EMSA gels shown in Figure 1. The data were obtained from four independent EMSA experiments (n=4). Data points represent mean ± SD. Solid lines indicate fits to a Hill binding model with fixed asymptotes (0–100%). The apparent dissociation constant (Kd) corresponds to the protein concentration yielding 50% RNA binding; fitted Kd and Hill coefficient (n) values ± SE are shown.
Figure 1—figure supplement 2
Turbidity measurements of PSMα3 with increasing Poly(AU) RNA concentrations.
PSMα3 concentration was maintained at 100 µM, and Poly(AU) RNA concentrations varied across 0, 10, 20, 50, 100, 200, and 400 ng/µL. Data points represent the average optical density at 400 nm (OD400) recorded within the first 30 minutes after sample preparation, with each condition tested in technical triplicate across three independent experiments (n=3). Error bars indicate the standard deviation.
Figure 2 with 3 supplements
Colocalization, droplet formation, and texture of PSMα3 mixed with varying Poly(AU) RNA concentrations.
Widefield fluorescence microscopy images of 100 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 incubated with Poly(AU) RNA (red) at 50 ng/µL (A) or 400 ng/µL (D), shown as individual fluorescence channels and merged images. (B–C) FRAP analysis of PSMα3-FITC in the presence of 50 ng/µL Poly(AU) RNA after 10 min (B) or 2 hr (C) of co-incubation. (E) FRAP analysis of PSMα3-FITC in the presence of 400 ng/µL Poly(AU) RNA after 10 min of co-incubation. For panels B, C, and E, images were acquired before photobleaching, immediately after photobleaching, and 40 s post-photobleaching. (F) 100 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 in the absence of RNA. All scale bars represent 20 µm.
Figure 2—figure supplement 1
Encapsulation of Poly(AU) RNA within PSMα3-FITC droplets.
Figure 2—figure supplement 2
FRAP analysis of PSMα3–Poly(AU) RNA condensate dynamics.
Normalized fluorescence recovery after photobleaching (FRAP) measured for condensates formed by 20% PSMα3-FITC (green) and 80% unlabeled 100 µM PSMα3 in the presence of 50 ng/μL Poly(AU) RNA, corresponding to the experiment shown in Figure 2. Blue circles represent measured fluorescence intensity normalized to pre-bleach levels, and the orange curve indicates a single-exponential fit to the recovery phase. The bleach event is marked by the orange arrow. Recovery reaches a plateau corresponding to a mobile fraction of 68% and an immobile fraction of 32%. The green marker and dashed line denote the half-time of recovery (t½=3.1 s). Horizontal dashed lines indicate pre-bleach intensity (top), plateau recovery level (middle), and immediate post-bleach intensity (bottom).
Figure 2—figure supplement 3
Colocalization analysis of PSMα3 and Poly(AU) RNA at different RNA concentrations.
Quantitative comparison of colocalization metrics for 100 µM PSMα3 (20% -PSMα3-FITC and 80% unlabeled) and Poly(AU) RNA at 50 ng/μL (blue) and 400 ng/μL (orange), corresponding to the experiment shown in Figure 2. Pearson’s correlation coefficient was calculated using above-threshold pixels only, with thresholds determined automatically by the Costes method. Additional metrics include Spearman’s rank correlation coefficient (ρ), Li’s intensity correlation quotient (ICQ), and Manders’ colocalization coefficients (tM1 and tM2) calculated using Costes automatic thresholding. Numerical values for each metric are indicated on the plot. The dashed vertical line at 0.5 marks a commonly used reference value for strong colocalization.
Figure 3 with 2 supplements
TEM and TIRF visualization of PSMα3 aggregation and morphology with different Poly(AU) RNA concentrations and incubation times.
(A) TEM micrographs of 100 µM PSMα3 incubated with or without Poly(AU) RNA at varying concentrations of 10 ng/µL, 50 ng/µL, and 400 ng/µL for 2 hr (top row) and 24 hr (bottom row). Scale bars represent 500 nm. (B) TIRF microscopy images showing 100 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 co-incubated with 50 ng/µL Poly(AU) RNA and the amyloid indicator AT630 (magenta) for 30 min and 2 hr. Scale bars represent 20 μm. (C) TIRF microscopy images of 100 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 co-incubated with 400 ng/µL Poly(AU) RNA and AT630 (magenta) for 30 min. Scale bars represent 20 µm.
Figure 3—figure supplement 1
PSMα3 amyloid formation is enhanced by Poly(AU) RNA in a time- and concentration-dependent manner.
The graphs show the calculated AmyTracker630 (AT630) fluorescence intensity of 100 µM PSMα3 incubated with Poly(AU) RNA under the indicated conditions, corresponding to the experiment shown in Figure 3. Boxplots show the interquartile range with the center line indicating the median; black diamonds denote the mean. Individual points represent independent measurements. Statistical significance was assessed using a Kruskal–Wallis test followed by Dunn’s post hoc test with Bonferroni correction. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Figure 3—figure supplement 2
RNA enhances α-helical structural features of PSMα3 over time.
Solid-state circular dichroism (ssCD) spectra of 100 µM PSMα3 incubated alone or with Poly(AU) RNA at concentrations of 50 ng/µL and 400 ng/µL. Spectra were collected immediately after preparation (A), and after 2 hr of incubation (B). Measurements were recorded in the far-UV range (180–250 nm) to assess changes in secondary structure under the indicated conditions.
Figure 4
Impact of Poly(AU) RNA on PSMα3 cytotoxicity and antibacterial activity.
Antimicrobial activity of PSMα3 against E. coli using the PrestoBlue cell viability assay (A) and its cytotoxicity against HeLa cells using the LDH colorimetric assay (B) were assessed with and without Poly(AU) RNA at varying concentrations. The experiments were performed in technical triplicate across three independent biological experiments (n=3) to ensure result reliability. Cytotoxicity and bacterial cell viability percentages were calculated as the mean of all replicates, with error bars representing the standard deviation. Statistical significance was determined using one-way ANOVA for normally distributed data in GraphPad Prism (version 11). Significance levels are indicated as follows: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Figure 5 with 1 supplement
Colocalization of PSMα3 with nucleic acids in HeLa cells.
(A) Confocal microscopy images showing the localization and colocalization of 20 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 and nucleic acids stained with PI (red) within the nucleoli of HeLa cells (indicated by arrows). The left panel illustrates the distribution of PSMα3 within the cell. The middle panel shows the nucleic acids stained with PI. The right panel is a composite image that demonstrates the colocalization of PSMα3 with nucleic acids. Scale bars represent 15 µm. (B) FRAP analysis of 20 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 inside the nucleolus (indicated by the arrow) of HeLa cells, showing fluorescence recovery after 60 s. Scale bars represent 10 µm.
Figure 5—figure supplement 1
FRAP analysis of PSMα3 dynamics within the nucleolus of HeLa cells. Normalized FRAP of 20% PSMα3-FITC and 80% unlabeled 20 µM PSMα3 measured within the nucleolus of HeLa cells, corresponding to the experiment shown in Figure 5.
Blue circles represent fluorescence intensity values normalized to the pre-bleach signal, and the orange line shows a single-exponential fit to the recovery kinetics. The photobleaching event is indicated by the orange arrow. Recovery approached ~64% within the 60 s acquisition window, suggesting the presence of both mobile and slowly exchanging populations. The recovery did not reach a clear plateau within the measured time frame, indicating the presence of a slowly exchanging or partially immobilized PSMα3 population within the nucleolus.
Figure 6 with 4 supplements
EGCG modulates PSMα3 aggregation and reduces toxicity against HeLa cells.
(A) Cytotoxicity of PSMα3 against HeLa cells in the presence and absence of EGCG and Poly(AU) RNA at two different concentrations, assessed via LDH assay. The experiment was performed in triplicate and repeated on three separate days for consistency. Cytotoxicity percentages were averaged across all replicates, with error bars representing the SD. Statistical significance: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (one-way ANOVA, GraphPad Prism v11). (B) TEM micrographs of 100 μM PSMα3 incubated for 24 hr, without (left) and with (right) a fivefold molar excess of EGCG. Scale bars: 500 nm. (C) Live-cell confocal microscopy of HeLa cells treated with 20 μM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 without (top) and with EGCG (bottom), imaged immediately after preparation (t=0) and after 10 min. Hoechst 33342 (blue) marks the nuclei, while PI staining (red) indicates membrane disruption. Scale bars: 10 μm.
Figure 6—figure supplement 1
EGCG inhibits PSMα3 fibrillation. Fibrillation kinetics of 100 µM freshly dissolved PSMα3, monitored by Thioflavin-T (ThT) fluorescence.
The data compares PSMα3 fibrillation in the absence (blue curve) and presence (red curve) of EGCG. The graph displays the mean fluorescence intensities from technical triplicate measurements, across three independent experiments (n=3), with error bars representing the SD.
Figure 6—figure supplement 2
EGCG reduces the antimicrobial activity of PSMα3 against E. coli.
(A) Super-resolution light microscopy images of E. coli treated with 20 μM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 in the absence (left) and presence (right) of EGCG at a 1:5 molar ratio. Scale bars: 5 μm. (B) Antimicrobial activity of PSMα3 against E. coli, evaluated using the PrestoBlue cell viability assay, with and without EGCG (1:5 molar ratio). The experiments were performed in technical replicates across three independent biological experiments (n=3) to ensure result reliability. Bacterial viability percentages were calculated as the mean of all replicates, with error bars representing the SD. Statistical significance was determined using one-way ANOVA for normally distributed data in GraphPad Prism (version 11). Significance levels are indicated as follows: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Figure 6—video 1
Live-cell imaging of PSMα3-induced membrane permeabilization in HeLa cells.
HeLa cells were incubated with 20 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 and monitored by time-lapse fluorescence microscopy. Nuclei were stained with Hoechst (blue). PI (red) was present in the imaging medium to detect loss of plasma membrane integrity. Time is shown in minutes: seconds (mm:ss) from peptide addition. Scale bar, 10 µm.
Figure 6—video 2
Live-cell imaging of EGCG-modulated PSMα3 activity in HeLa cells.
HeLa cells were incubated with 20 µM of 20% PSMα3-FITC (green) and 80% unlabeled PSMα3 in the presence of EGCG at a fivefold molar excess relative to PSMα3 (PSMα3: EGCG = 1:5) and monitored by time-lapse fluorescence microscopy. Nuclei were stained with Hoechst (blue). PI (red) was present in the imaging medium to detect loss of plasma membrane integrity. Time is shown in mm:ss from peptide addition. Scale bar, 10 µm.
Figure 7
Residue-specific interactions between PSMα3 and EGCG.
(A) One-dimensional (1D) 1H NMR spectra of 1.0 mM PSMα3 alone (red) and in complex with 0.5 mM EGCG (blue), recorded at 35 °C. Specific residues, including Met1, Glu2, and V4/N21, show chemical shift changes suggestive of direct interaction with EGCG (highlighted in the upper left). Dashed boxes mark proton signals corresponding to EGCG. Peak broadening of H1/H2 protons, compared to the EGCG-only reference sample (light grey), indicates interaction from the EGCG side. Slight opalescence observed in the sample suggests potential aggregate formation. (B) Two-dimensional (2D) 1H–1H TOCSY and NOESY spectra of the PSMα3:EGCG complex at a 2:1 ratio, recorded at 35 °C. The cross-peaks in the TOCSY spectrum allow the identification of the spin system and direct connections through the scalar coupling between the proton amide (HN) and the alpha protons (Hα) of the same residue. The cross-peaks in the NOESY spectrum establish a sequential connection between neighboring residues (i+1), allowing spectral assignment. The remaining cross-peaks (i+3 or i+4) in the spectrum support PSMα3 secondary structure in the experimental conditions. (C) Temporal stability of the PSMα3:EGCG sample over 3 days.
Figure 8 with 5 supplements
Effect of RNA concentration on LL-37 phase separation, aggregation, and activity at pH 7.4 after heat shock.
Fluorescence microscopy images showing 20% FITC-labeled (green) and 80% unlabeled 100 μM LL-37 in the presence of increasing Poly(AU) RNA concentrations of 100 ng/μL (A), 200 ng/μL (B), and 400 ng/μL (C) of Poly(AU) RNA (PI-stained, red) after heat shock at 65 °C for 15 min at pH 7.4. Scale bars represent 20 µm. (D) LL-37 cytotoxicity, with and without Poly(AU) RNA and EGCG at varying concentrations, was assessed in HeLa cells using the LDH colorimetric assay. The experiment was performed in triplicate and repeated on 3 separate days for reproducibility. Cytotoxicity percentages represent the average of all replicates, with error bars indicating the SD. (E) The antimicrobial activity of LL-37, with and without Poly(AU) RNA and EGCG at varying concentrations, was evaluated against E. coli using the PrestoBlue Cell Viability assay. The experiment was performed in triplicate and repeated on 3 separate days for reproducibility. Bacterial viability percentages represent the average of all replicates, with error bars indicating the SD. Statistical significance (D–E): *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (one-way ANOVA, GraphPad Prism v11).
Figure 8—figure supplement 1
LL-37 undergoes aggregation with Poly(AU) RNA.
Light microscopy images showing 20% FITC-labeled (green) and 80% unlabeled 100 μM LL-37 (with no thermal stress) forming aggregates in the presence of 100 ng/μL Poly(AU) RNA, stained with PI (red). The composite image (right) highlights the colocalization of LL-37 and RNA (yellow), indicating RNA-induced aggregation. Scale bars: 20 μm.
Figure 8—figure supplement 2
FRAP analysis of LL-37 – Poly(AU) RNA assemblies following heat shock.
(A) Representative fluorescence images of 20% FITC-labeled (green) and 80% unlabeled 100 μM LL-37 assemblies formed in the presence of 100 ng/µL Poly(AU) RNA, shown before photobleaching (Pre-Bleach), immediately after photobleaching (Bleach), and 60 s after bleaching (+60 seconds). Scale bars in all images: 5 μm. (B) Normalized FRAP corresponding to the bleached region shown in (A). Blue circles represent fluorescence intensity normalized to pre-bleach levels, and the orange curve shows a single-exponential fit to the recovery data. The bleach event is indicated by the orange arrow. Recovery reaches a limited plateau corresponding to a mobile fraction of 10% and an immobile fraction of 90%. Horizontal dashed lines denote pre-bleach intensity (top), plateau recovery level (middle), and immediate post-bleach intensity (bottom).
Figure 8—figure supplement 3
Effect of RNA concentration on PSMα3 phase separation and aggregation at pH 4 after heat shock.
Fluorescence microscopy images showing the effects on 20% PSMα3-FITC (green) and 80% unlabeled 100 μM PSMα3 in the presence of increasing concentrations of Poly(AU) RNA (labeled with PI, red) following heat shock at 65 °C for 15 min at pH 4. At 50 ng/μL (A), 100 ng/μL (B), 200 ng/μL (C), and 400 ng/μL (D). At a low RNA concentration of 50 ng/μL (A), clear phase separation of PSMα3 was observed, as evidenced by the presence of well-defined, spherical droplets. As the RNA concentration increased to 100 ng/μL, the droplets began to lose their spherical structure and showed more morphological irregularities (B). At RNA concentrations of 200 ng/μL and 400 ng/μL, a distinct transition to aggregation with irregular, amorphous clusters was observed (C and D, respectively). The composite images revealed an overlap between the PSMα3 and RNA, indicating colocalization during both phase separation and amorphous cluster formation. Scale bars in all images represent 20 µm.
Figure 8—figure supplement 4
PSMα3 aggregates with Poly(AU) RNA after heat shock at physiological pH.
Light microscopy images showing 20% PSMα3-FITC (green) and 80% unlabeled 100 μM PSMα3 forming aggregates with 50 ng/μL Poly(AU) RNA, stained with PI (red), after heat shock at 65 °C for 15 min in 50 mM HEPES, 150 mM NaCl, at physiological pH (7.4). The composite image (right) highlights the colocalization of PSMα3 and RNA (yellow), indicating RNA-induced aggregation under heat stress. Scale bars: 20 μm.
Figure 8—figure supplement 5
Poly(AU) RNA modulates LL-37 aggregation dynamics before and after heat shock.
TEM micrographs of 100 µM LL-37 incubated alone or with 100 ng/µL or 400 ng/µL Poly(AU) RNA for 2 hr or 24 hr. (A) Samples imaged before heat shock at the indicated time points. (B) Samples subjected to a 65 °C heat shock for 15 min, followed by further incubation for 2 or 24 hr before imaging. Scale bars in all images represent 500 nm.
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Peptide, recombinant protein | PSMα3 (Unlabeled or with a C-terminal FITC label) | Custom synthesis; GL Biochem (Shanghai) Ltd. | P0C805 · PSMA3_STAA8 | |
| Peptide, recombinant protein | LL-37 (Unlabeled or with a C-terminal FITC label) | Custom synthesis; GL Biochem (Shanghai) Ltd. | P49913 · CAMP_HUMAN | LL-37 is generated by proteolytic cleavage of the cathelicidin antimicrobial peptide precursor (CAMP). |
| Sequence-based reagent | 30-nt oligo PolyA-IRDye 800CW and 30-nt oligo Poly (U) RNA | IDT | ||
| Sequence-based reagent | Poly(AU) RNA | Sigma-Aldrich | ||
| Chemical compound, drug | Thioflavin T (ThT) | Sigma-Aldrich | Compound CID: 16953 | |
| Chemical compound, drug | epigallocatechin gallate (EGCG) | Thermo Scientific Chemicals | Cat no. 449010100 | |
| Cell line (Homo sapiens) | HeLa | ATCC CCL-2 | RRID:CVCL_0030 | The human cervical carcinoma cell line was authenticated, and was not tested for mycoplasma contamination during the course of the study |
| Strain, strain background (Escherichia coli) | RFM795 | CGSC, Yale University (obtained as a kind gift from Prof. Sima Yaron, Technion) | CGSC# 14179 | Contains lptD4213 deletion; increased outer membrane permeability |
| Commercial assay or kit | Lactate dehydrogenase (LDH) release assay | Roche Applied Science (Sigma-Aldrich) | Cat. no. 04744934001 | Cytotoxicity assays |
| Commercial assay or kit | PrestoBlue Cell Viability Reagent | Invitrogen (Thermo Scientific Chemicals) | Cat no. A13261 | Bacterial viability assay |
| Software, algorithm | GraphPad Prism (v11) | GraphPad Software | RRID:SCR_002798 | statistical analysis |
| Software, algorithm | Fiji/ImageJ | https://imagej.net/software/fiji/ | RRID:SCR_002285 | Image processing |
| Software, algorithm | ZEISS ZEN | Carl Zeiss | RRID:SCR_013672 | TIRF and FRAP image acquisition and processing |
| Software, algorithm | Imaris (Oxford Instruments) | Oxford Instruments | RRID:SCR_007370 | Live-cell confocal movie analysis |
| Software, algorithm | TopSpin 4.1.4 | Bruker BioSpin | RRID:SCR_014227 | NMR data acquisition and processing |
| Software, algorithm | POKY | Lee et al., 2021 (ref. 121) | https://poky.clas.ucdenver.edu | NMR spectral assignment and analysis |
| Other | Scilab | Scilab Enterprises | https://www.scilab.org/ RRID:SCR_014258 | NMR numerical computation, data analysis, and custom scripting |
| Other | Transmission electron microscope | Talos F200C ThermoFisher Scientific (FEI) | Ilse Katz Institute for Nanoscale Science & Technology, Ben Gurion University of the Negev | TEM imaging of fibrils and aggregates |
| Other | Spinning-disk confocal | Ti2-E+CSU-W1 Nikon / Yokogawa | LS&E Infrastructure Center, Technion–Israel Institute of Technology | Live-cell imaging of PSMα3 in HeLa cells |
| Other | Laser-scanning confocal | LSM 710 Carl Zeiss | LS&E Infrastructure Center, Technion–Israel Institute of Technology | FRAP (in vitro condensates & nucleolar) |
| Other | Widefield fluorescence microscope | DMI8 Leica | LS&E Infrastructure Center, Technion–Israel Institute of Technology | Fluorescence imaging of peptide–RNA mixtures in vitro |
| Other | Super-resolution / TIRF microscope | Elyra 7 Carl Zeiss | LS&E Infrastructure Center, Technion–Israel Institute of Technology | TIRF (AmyTracker630) |
| Other | NMR spectrometer | 700 MHz AVANCE NEO +Prodigy cryoprobe Bruker BioSpin | Max Planck Institute for Multidisciplinary Sciences, Göttingen | 1D ¹H, 2D TOCSY, and NOESY of PSMα3–EGCG |
| Other | Multimode plate reader | FLUOstar Omega BMG Labtech | Technion–Israel Institute of Technology | ThT fibrillation kinetics, turbidity (OD₄₀₀), LDH absorbance |
| Other | Multimode plate reader | CLARIOstar BMG Labtech | Technion–Israel Institute of Technology | PrestoBlue bacterial viability assay |
| Other | CD spectropolarimeter | Chirascan (with solid sample holder CS/SSH) Applied Photophysics | SPC facility, EMBL Hamburg | Solid-state CD of PSMα3 secondary structure |
| Other | Gel imaging system | Odyssey FC LI-COR | Technion–Israel Institute of Technology | EMSA visualization (IRDye 800CW channel) |
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
RNA selectively modulates activity of virulent amyloid PSMα3 and host-defense LL-37 via phase separation and aggregation dynamics
eLife 15:RP109290.
https://doi.org/10.7554/eLife.109290.4
