English

• Membraneless organelles, also known as biomolecular condensates, are essential to cellular organization, concentrating or sequestering biomolecules to regulate biological reactions [1,2]. The extensive engagement of membraneless organelles in cellular functions has stimulated the creation of synthetic molecules emulating their characteristics [3–10]. These artificial structures, often leveraging self-assembling peptides, have provided new avenues for manipulating biological processes [11–20]. Nevertheless, the achievement of comparable outcomes with small non-peptidic organic molecules remains challenging [21–23]. Creating artificial membraneless organelles using small organic molecules offers simple models for recreating or controlling cellular events and unique opportunities for drug discovery.

  A promising source of such small molecules is a collection of organic compounds that self-assemble or aggregate in aqueous solutions. To construct such a focused library, we initially selected 1000 divergent small molecules with molecular weights exceeding 700 Da, enriched in aromatic, hydrophobic, or symmetric structures, from a comprehensive collection of approximately 1 million compounds in the Chinese National Compound Library (Shanghai). Self-assembling properties of these molecules were profiled by a method we previously reported using two fluorescent probes, 8-anilino-1-napthlenesufonic acid (ANS) and Nile Red [24]. These environment-sensitive probes, typically nonfluorescent in aqueous solutions, experience fluorescence augmentation when encapsulated in hydrophobic pockets. Given the propensity of self-assembling or aggregating small molecules to form higher-order structures with multiple hydrophobic pockets, the observed increase in fluorescence intensity serves as an indicator for monitoring the presence of molecular condensates in aqueous solutions. Out of the initial 1000 molecules, 44 demonstrated a significant increase in the fluorescence of both probes in phosphate-buffered saline (PBS), prompting their selection for subsequent investigations.

  Naturally occurring membraneless organelles typically concentrate or sequester a subset of proteins. We examined the ability of the 44 molecules to interact with selective cellular proteins (Fig. 1A). Each compound (250 µmol/L) was incubated with HeLa cell lysates for 0.5 h. After centrifugation and extensive wash, the proteins co-precipitated with small-molecule assemblies were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Silver staining of the gels showed that the majority of the 44 self-assembling molecules displayed similar patterns of co-precipitated proteins, which either non-specifically interacted with the assembled materials or denatured during the experiments. One compound, D008 (molecule 1), exhibited selective protein bands within the molecular weight range of 35–40 kDa (Figs. 1B and C). The structures of the compounds used in Fig. 1B are listed in Table S1 (Supporting information). For subsequent in-depth studies, we chemically synthesized D008 (1) and confirmed its ability to reproduce the screening result.

  Figure 1

  

  Figure 1.  Discovery of D008 (1). (A) Screening of self-assembling molecules. (B) A representing SDS-PAGE gel of the library screening. The arrows indicate bands specific to the D008 sample. (C) Chemical structure of D008 (1).

  DownLoad: Full-Size Img PowerPoint

  Dynamic light scattering (DLS) analysis of D008 confirmed its capability to form assemblies in both PBS and water (Fig. S1 in Supporting information). The diameters of D008 particles in PBS exhibited an increase with concentrations, reaching micrometer sizes (1000–2000 nm). In contrast, the particle diameter of D008 remained nearly constant (~200 nm) with increasing concentration in pure water. These characteristics were further corroborated by confocal microscopy, wherein Nile Red-stained D008 particles were detected in both PBS and water (Fig. S2 in Supporting information). Particle aggregation occurred in PBS with increasing concentrations, whereas dispersed nanoparticles were observed in water. This observation aligns with cryo-transmission electron microscopy (cryo-TEM) results, which depicted spherical particles of D008 in water and their aggregation in PBS (Fig. S3 in Supporting information). Collectively, these results suggest that D008 forms nanoparticles in aqueous solutions, and their higher-order assembly patterns are contingent upon solution conditions.

  To identify D008-binding proteins at a proteome level, we performed a comprehensive liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis as outlined in Fig. S4 (Supporting information). HeLa cell lysates were treated with 50 µmol/L D008 in PBS, followed by centrifugation and washing. The proteins coprecipitated with D008 particles were analyzed by LC-MS/MS for label-free quantification. The criteria for protein enrichment were set as log₂[D008/DMSO] > 1 and −log₁₀P-value > −log₁₀0.05, resulting in the identification of 174 enriched proteins (Fig. 2A, refer also to Supporting information Excel file 1). To gain insights into the functional relevance of the identified D008-binding proteins, a gene ontology (GO) enrichment analysis was performed. The analysis of the 174 D008-binding proteins revealed an enrichment of RNA-binding proteins (Fig. 2B).

  Figure 2

  

  Figure 2.  LC-MS/MS analysis of D008-precipitated proteins. (A) Volcano plots (log₂(D008/DMSO) vs. −log₁₀(P-value)). Dashed line represents threshold ((log₂D008/DMSO) > 1, P-value < 0.05). Enriched proteins over the threshold are indicated as red dots. (B) GO enrichment analysis (molecular function) was conducted for the 174 proteins.

  DownLoad: Full-Size Img PowerPoint

  Among the highly significant D008-binding proteins, we noticed hnRNPA2B1, an RNA-binding protein known to exhibit a doublet band at 36 and 38 kDa on an SDS-PAGE gel, consistent with the bands observed during the library screening. hnRNPA2B1 plays essential roles in mRNA processing, export, translation, and the innate immune response [25–29]. To corroborate the binding of D008 to hnRNPA2B1, the protein samples coprecipitated with D008 were subjected to Western blot analysis using an antibody specific to hnRNPA2B1 (Fig. 3A). The results demonstrated the binding of D008 to hnRNPA2B1. Intriguingly, treatment with RNase abolished this interaction, suggesting the dependence on RNA for the observed binding. These results collectively suggest that D008 interacts selectively with a subset of mRNA-binding proteins, including hnRNPA2B1, and this interaction is mediated through RNA.

  Figure 3

  

  Figure 3.  The interaction of D008 and molecule 2 with hnRNPA2B1. The HeLa cell protein samples co-precipitated with D008 (A) or molecule 2 (B) were subjected to Western blot analysis using an antibody specific to hnRNPA2B1. The impact of RNA removal by RNase was also assessed.

  DownLoad: Full-Size Img PowerPoint

  The interaction of D008 with RNA-binding proteins prompted us to examine its ability to modulate the translation of mRNA in vitro. The impact of D008 on translation efficacy was evaluated through a luciferase translation assay using the Human Cell-Free Protein Expression System (Fig. 4A) [30]. Human cell lysates were subjected to incubation with varying concentrations of D008, spanning from 50 µmol/L to 500 µmol/L, over a duration of 3 h. Subsequently, firefly luciferase mRNA was introduced into the lysate solution. Following an additional 3-h incubation to allow for luciferase translation, the level of luciferase expression within each cell lysate was estimated by measuring the intensity of luciferase luminescence. The results demonstrated a concentration-dependent reduction in luciferase translation by D008, indicating that D008 particles impede mRNA translation, at least within an in vitro cell-lysate environment, through the sequestration of RNA-binding proteins.

  Figure 4

  

  Figure 4.  Effects of D008 and its analogs on in vitro translation of mRNA. (A) Human cell lysates were subjected to incubation with varying concentrations of D008 (50 µmol/L to 500 µmol/L) for 3 h. Subsequently, firefly luciferase mRNA was introduced into the lysate solution. Following an additional 3-h incubation to allow for luciferase translation, the level of luciferase expression within each cell lysate was estimated by measuring the intensity of luciferase luminescence. (B) Chemical structures of D008 and 2. (C) Effects of the hydroxyl group. Human cell lysate was treated with 200 µmol/L of each compound, and the luminescence intensity was subsequently monitored. (D) Effects of molecule 2 on the in vitro translation of luciferase mRNA at varying concentrations.

  DownLoad: Full-Size Img PowerPoint

  D008 has two hydroxyl groups attached to the stereocenters. To investigate their effects, we synthesized an achiral analog lacking hydroxyl groups (molecule 2) (Fig. 4B). Its ability to modulate the translation of mRNA in vitro was similarly assessed by the luciferase translation assay (Fig. 4C). Human cell lysate was treated with 200 µmol/L of each compound, and the luminescence intensity was subsequently monitored. The results showed that molecule 2, lacking hydroxyl groups, exhibited less activity than D008 (Figs. 4C and D), underscoring the importance of hydroxyl groups. DLS analysis of 2 in PBS revealed that it maintains the ability to form particles akin to D008 (Fig. S5 in Supporting information). To evaluate the binding affinity of 2 to hnRNPA2B1, protein samples coprecipitated with 2 were subjected to Western blot analysis. The results revealed no detectable interaction between molecule 2 and hnRNPA2B1 (Fig. 3B). These results suggest that the hydroxyl groups of D008 are dispensable for its self-assembly but play a crucial role in facilitating binding to RNA-binding proteins.

  Encouraged by the in vitro results, we extended our study to evaluate the impact of D008 in cultured human cells. To monitor the protein biosynthesis, we employed azidohomoalaine (AHA), an azide-containing methionine analog that is incorporated into newly synthesized proteins. Through a click reaction between the azido group in the incorporated AHA and an alkyne-tagged fluorescent molecule, the newly synthesized proteins can be visualized by fluorescent scanning of an SDS-PAGE gel [31]. While cycloheximide (CHX), a gold-standard translation inhibitor, diminished the fluorescent intensity of protein bands, D008 had no obvious effects (Fig. 5). However, a closer examination of the band intensities suggested that D008 had a marginal effect, if any, on protein translation. The limited cellular activity observed for D008 may be due to the challenges associated with the cell permeability of rapidly formed particle aggregates of D008. Precise control over D008 self-assembly, including the development of a prodrug, would enhance the cellular activity of D008 in future investigations.

  Figure 5

  

  Figure 5.  SDS-PAGE analysis of newly synthesized proteins in the presence of 1% (v/v) DMSO, D008 (50 µmol/L), or cycloheximide (CHX, 50 µmol/L). The figure displays fluorescence scanning images (left) and Coomassie brilliant blue (CBB)-stained images (right). Mean intensities of bands are shown below.

  DownLoad: Full-Size Img PowerPoint

  The present study discovered D008 as a self-assembling small molecule that sequesters a subset of RNA-binding proteins to impede mRNA translation. This discovery echoes McKnight and co-workers’ serendipitous discovery of b-isox, a self-assembling molecule that forms hydrogel in vitro and co-precipitates with RNA-binding proteins containing RNA granule constituents [21]. Although the cellular activity of b-isox remains unexplored, the b-isox-induced RNA granule-like assemblies facilitated in vitro analysis of mRNA components within RNA granules [32]. A comparison of b-isox-binding proteins with D008-binding proteins reveals a certain level of overlap, but key RNA granule proteins, including G3GPs, ATXN2, and PABPs, were not found in D008-binding proteins (Fig. S6 in Supporting information). This observation suggests D008’s potential utility in isolating a distinct set of RNA-binding proteins. The biological significance of this unique set of RNA-binding proteins remains undetermined. Nevertheless, the present identification of another self-assembling organic molecule that co-precipitates RNA-binding proteins may offer a unique opportunity for analyzing the condensations of RNA-binding proteins and for developing an unprecedented class of small molecules that control gene expression.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Hyoseok Kim: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Changyi Cui: Methodology, Investigation, Formal analysis, Data curation. Kohei Toh: Methodology, Investigation, Formal analysis, Data curation. Genyir Ado: Methodology, Investigation, Formal analysis, Data curation. Tetsuya Ogawa: Methodology, Investigation, Formal analysis. Yixin Zhang: Investigation, Formal analysis. Shin-ichi Sato: Supervision. Yong-Beom Lim: Supervision, Funding acquisition. Hiroki Kurata: Supervision. Lu Zhou: Supervision, Resources, Project administration, Funding acquisition, Formal analysis. Motonari Uesugi: Writing – review & editing, Writing – original draft, Supervision, Resources, Funding acquisition, Formal analysis, Conceptualization.

  Acknowledgments

  This work was supported by JSPS (No. 22H00350 to M.U.), Ministry of Health & Welfare, Republic of Korea (Korea Health Technology R&D Project through the Korea Health Industry Development Institute; No. HI19C1234 to H.K.), and JST (the Establishment of University Fellowships towards the Creation of Science Technology Innovation; No. JPMJFS2123 to K.T.). This work was partially supported and inspired by the International Collaborative Research Program of Institute for Chemical Research, Kyoto University (No. 2024–84), Kyoto University On-Site Lab (Fudan-Kyoto Shanghai Lab), and the international and interdisciplinary environments of JSPS CORE-to-CORE Program “Asian Chemical Biology Initiative”.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110135.

  
  
• 1. [1]

     D. Hnisz, K. Shrinivas, R.A. Young, et al., Cell 169 (2017) 13–23. doi: 10.1016/j.cell.2017.02.007

  2. [2]

     J.A. Ditlev, L.B. Case, M.K. Rosen, J. Mol. Biol. 430 (2018) 4666–4684. doi: 10.1016/j.jmb.2018.08.003

  3. [3]

     L.B. Case, X. Zhang, J.A. Ditlev, et al., Science 363 (2019) 1093–1097. doi: 10.1126/science.aau6313

  4. [4]

     C.D. Reinkemeier, G.E. Girona, E.A. Lemke, Science 363 (2019) eaaw2644. doi: 10.1126/science.aaw2644

  5. [5]

     M. Heidenreich, J.M. Georgeson, E. Locatelli, et al., Nat. Chem. Biol. 16 (2020) 939–945. doi: 10.1038/s41589-020-0576-z

  6. [6]

     L. Yang, R. Peltier, M. Zhang, et al., J. Am. Chem. Soc. 142 (2020) 18150–18159. doi: 10.1021/jacs.0c08463

  7. [7]

     M. Yoshikawa, T. Yoshii, M. Ikuta, et al., J. Am. Chem. Soc. 143 (2021) 6434–6446. doi: 10.1021/jacs.0c12375

  8. [8]

     Y. Dai, M. Farag, D. Lee, et al., Nat. Chem. Biol. 19 (2023) 518–528. doi: 10.1038/s41589-022-01252-8

  9. [9]

     Z.G. Qian, S.C. Huang, X.X. Xia, Nat. Chem. Biol. 18 (2022) 1330–1340. doi: 10.1038/s41589-022-01203-3

  10. [10]

      M.V. Garabedian, W. Wang, J.B. Dabdoub, et al., Nat. Chem. Biol. 17 (2021) 998–1007. doi: 10.1038/s41589-021-00840-4

  11. [11]

      H. Wang, Y. Song, W. Wang, et al., J. Am. Chem. Soc. 146 (2024) 330–341. doi: 10.1021/jacs.3c09202

  12. [12]

      S. Kim, J.-B. Chae, D. Kim, et al., J. Am. Chem. Soc. 145 (2023) 21991–22008. doi: 10.1021/jacs.3c06898

  13. [13]

      B. Jana, S. Jin, E.M. Go, et al., J. Am. Chem. Soc. 145 (2023) 18414–18431. doi: 10.1021/jacs.3c04467

  14. [14]

      S. Liu, Q. Zhang, A.N. Shy, et al., J. Am. Chem. Soc. 143 (2021) 15852–15862. doi: 10.1021/jacs.1c07923

  15. [15]

      L. Hu, Y. Li, X. Lin, et al., Angew. Chem. Int. Ed. 60 (2021) 21807–21816. doi: 10.1002/anie.202103507

  16. [16]

      W. Tan, Q. Zhang, J. Wang, et al., Angew. Chem. Int. Ed. 60 (2021) 12796–12801. doi: 10.1002/anie.202102601

  17. [17]

      H. He, S. Liu, D. Wu, et al., Angew. Chem. Int. Ed. 59 (2020) 16445–16450. doi: 10.1002/anie.202006290

  18. [18]

      Z. Feng, H. Wang, X. Chen, et al., J. Am. Chem. Soc. 139 (2017) 15377–15384. doi: 10.1021/jacs.7b07147

  19. [19]

      Z. Feng, H. Wang, B. Xu, J. Am. Chem. Soc. 140 (2018) 16433–16437. doi: 10.1021/jacs.8b10542

  20. [20]

      H. Wang, Z. Feng, B. Xu, J. Am. Chem. Soc. 141 (2019) 7271–7274. doi: 10.1021/jacs.9b03346

  21. [21]

      M. Kato, T.W. Han, S. Xie, et al., Cell 149 (2012) 753–767. doi: 10.1016/j.cell.2012.04.017

  22. [22]

      G. Ado, N. Noda, H.T. Vu, et al., Chem. Sci. 13 (2022) 5760–5766. doi: 10.1039/D1SC07151C

  23. [23]

      D. Maity, M. Howarth, M.C. Vogel, et al., J. Am. Chem. Soc. 143 (2021) 3086–3093. doi: 10.1021/jacs.0c09967

  24. [24]

      S. Jin, H.T. Vu, K. Hioki, et al., Angew. Chem. Int. Ed. 60 (2021) 961–969. doi: 10.1002/anie.202011604

  25. [25]

      C.R. Alarcón, H. Goodarzi, H. Lee, et al., Cell 162 (2015) 1299–1308. doi: 10.1016/j.cell.2015.08.011

  26. [26]

      H. Goodarzi, H.S. Najafabadi, P. Oikonomou, et al., Nature 485 (2012) 264–268. doi: 10.1038/nature11013

  27. [27]

      C. Villarroya-Beltri, C. Gutiérrez-Vázquez, F. Sánchez-Cabo, et al., Nat. Commun. 4 (2013) 2980. doi: 10.1038/ncomms3980

  28. [28]

      X. Wang, X. Fan, J. Zhang, et al., Cell Rep. 43 (2024) 113769. doi: 10.1016/j.celrep.2024.113769

  29. [29]

      L. Wang, M. Wen, X. Cao, Science 365 (2019) eaav0758. doi: 10.1126/science.aaw7565

  30. [30]

      S. Mikami, T. Kobayashi, M. Masutani, et al., Protein Expr. Purif. 62 (2008) 190–198. doi: 10.1016/j.pep.2008.09.002

  31. [31]

      D.C. Dieterich, A.J. Link, J. Graumann, et al., Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 9482–9487. doi: 10.1073/pnas.0601637103

  32. [32]

      T.W. Han, M. Kato, S. Xie, et al., Cell 149 (2012) 768–779. doi: 10.1016/j.cell.2012.04.016

• 

  Figure 1  Discovery of D008 (1). (A) Screening of self-assembling molecules. (B) A representing SDS-PAGE gel of the library screening. The arrows indicate bands specific to the D008 sample. (C) Chemical structure of D008 (1).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  LC-MS/MS analysis of D008-precipitated proteins. (A) Volcano plots (log₂(D008/DMSO) vs. −log₁₀(P-value)). Dashed line represents threshold ((log₂D008/DMSO) > 1, P-value < 0.05). Enriched proteins over the threshold are indicated as red dots. (B) GO enrichment analysis (molecular function) was conducted for the 174 proteins.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The interaction of D008 and molecule 2 with hnRNPA2B1. The HeLa cell protein samples co-precipitated with D008 (A) or molecule 2 (B) were subjected to Western blot analysis using an antibody specific to hnRNPA2B1. The impact of RNA removal by RNase was also assessed.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Effects of D008 and its analogs on in vitro translation of mRNA. (A) Human cell lysates were subjected to incubation with varying concentrations of D008 (50 µmol/L to 500 µmol/L) for 3 h. Subsequently, firefly luciferase mRNA was introduced into the lysate solution. Following an additional 3-h incubation to allow for luciferase translation, the level of luciferase expression within each cell lysate was estimated by measuring the intensity of luciferase luminescence. (B) Chemical structures of D008 and 2. (C) Effects of the hydroxyl group. Human cell lysate was treated with 200 µmol/L of each compound, and the luminescence intensity was subsequently monitored. (D) Effects of molecule 2 on the in vitro translation of luciferase mRNA at varying concentrations.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  SDS-PAGE analysis of newly synthesized proteins in the presence of 1% (v/v) DMSO, D008 (50 µmol/L), or cycloheximide (CHX, 50 µmol/L). The figure displays fluorescence scanning images (left) and Coomassie brilliant blue (CBB)-stained images (right). Mean intensities of bands are shown below.

  下载: 全尺寸图片 幻灯片
•