Bacterial ribonucleoprotein bodies maintain an acidic pH environment as a mechanism of enzyme regulation

Wade E Schnorr; Moeka Sasazawa; Kathryn G Dzurik; Kaveendya S Mallikaarachchi; Hadi Yassine; Kulathungage H Dilrangi; Ryan Cho; Shelby L Millheim; Jill E Millstone; Jared M Schrader; Saumya Saurabh; W Seth Childers

doi:10.1038/s41467-026-76859-9

. 2026 Aug 27;17:10242. doi: 10.1038/s41467-026-76859-9

Bacterial ribonucleoprotein bodies maintain an acidic pH environment as a mechanism of enzyme regulation

Wade E Schnorr ¹, Moeka Sasazawa ², Kathryn G Dzurik ¹, Kaveendya S Mallikaarachchi ³, Hadi Yassine ³, Kulathungage H Dilrangi ³, Ryan Cho ², Shelby L Millheim ¹, Jill E Millstone ¹, Jared M Schrader ^3,^✉, Saumya Saurabh ^2,^✉, W Seth Childers ^1,^✉

PMCID: PMC13616956 PMID: 42802233

Abstract

Biomolecular condensates create distinct subcellular niches, but their chemical environments and impacts on clients remain poorly understood. Using ratiometric fluorescent probes in vivo, we find that bacterial ribonucleoprotein bodies (BR-bodies) exhibit an acidic dense phase (pH ~5.1) due to the scaffold’s low pI and negative charge. Single-molecule localization microscopy and fluorescence lifetime imaging reveal spatially variable, acidic nanoscale RNase E clusters in Caulobacter crescentus BR-bodies. Reducing RNase E’s negative charge yields more neutral BR-bodies, causing morphological defects and viability loss upon outgrowth from stationary phase, likely reflecting altered pH and reduced condensation. In vitro assays with C-SNARF-4F and RNase E CTD-pHluorin2 recapitulate this acidic gradient and show that the acidic microenvironment enhances PNPase activity (optimal pH 5.3–6.3). Moreover, stationary-phase BR-bodies average just below this pH at 4.8 ± 0.2, a feature consistent with arrested RNA decay observed in vitro and stationary phase RNA storage in past work.

Subject terms: Bacteriology, RNA-binding proteins, Bacterial genetics, Intrinsically disordered proteins

Ratiometric pH sensors show that phase-separated BR-bodies in Caulobacter crescentus form acidic interiors, tuned by scaffold protein charge, to regulate enzymatic activity. Disrupting this pH control impairs stationary phase recovery and causes stress-induced morphological defects.

Introduction

Biomolecular condensates are dynamic, membrane-less assemblies of proteins, nucleic acids, and other biomolecules that organize and regulate cellular biochemistry^1–3. Unlike membrane-bound organelles, biomolecular condensates rely on phase separation to form distinct sub-cellular organelle-like structures that diverge in composition and density from their cytoplasmic surroundings^4–9. Diverse biochemical pathways are regulated by biomolecular condensates in eukaryotic systems, and recent discoveries have revealed their presence and importance in bacteria, as well^1,10–12. For example, PopZ condensates in Caulobacter crescentus localize at cell poles to regulate asymmetric cell division, ensuring the differential allocation of cell fate-determining proteins^13–15. Similarly, ParB condensates are vital for chromosome segregation during cell division in several bacteria by binding and spreading along the parS sequence^16,17. In addition, McdB phase separation helps position carboxysomes in proteobacteria and cyanobacteria¹⁸. Bacteria also form ribonucleoprotein bodies (BR-bodies), analogous to eukaryotic P-bodies and stress granules, via RNase E phase separation. These condensates mediate global RNA decay by colocalizing enzymes involved in the multi-step degradation pathway (Fig. 1A), ensuring turnover and preventing the accumulation of decay intermediates. Together, these condensates illustrate select recent examples of how phase separation regulates essential physiology in bacteria^1,10,11.

Fig. 1 — A Schematic showing the two-step RNA degradation process utilizing RNase E and its bound client PNPase. B RNase E consists of an N-terminal ribonuclease domain and a C-terminal intrinsically disordered region (IDR) with alternating clusters of negatively and positively charged amino acids. An example side chain of aspartic acid is shown in its protonated and deprotonated states to highlight the buffering ability of both aspartic and glutamic acid, given the ~ 5 mM concentration present in the adjacent in vitro RNase E CTD condensates. C The cartoon representation of the BR-body structure emphasizes the high net negative charge within the condensate, which may play a role in pH regulation. The zoomed-in portion is shown with other client proteins, and RNA omitted for clarity. *Caulobacter crescentus* cell cartoon adapted from W.E. Schnorr and W.S. Childers, *ChemSystemsChem*, 2024, https://doi.org/10.1002/syst.202400011, licensed under CC BY-NC 4.0. D Negative charge is a conserved feature among RNase E homologs in Alphaproteobacteria, which have a median net charge/residue of − 0.12 and a median pI of 4.2. The net charge/residue of the *Caulobacter crescentus* proteome is 0.00, while that of known clients of BR-bodies skews negative to a median net charge/residue of − 0.02. E Structural model of the PNPase client, showing the distribution of charges on the solvent-exposed surface compared to the RNA-binding surface. Figure 1 (bioinformatic/charge analysis): W.E.S. and K.G.D. performed bioinformatic analysis of RNase E homolog pI conservation; W.E.S. performed net charge analysis of BR-body clients and structural analysis of PNPase.

A major question in biology is how cells use their 3-dimensional layout to spatially regulate biochemical pathways. For example, in eukaryotes, membrane-bound organelles create oxidizing conditions in mitochondria and acidic conditions in lysosomes, influencing enzymatic function. Microbes generally lack membrane-bound organelles, raising the question about their capacity to generate chemically distinct sub-cellular niches. Within bacteria, condensates could create such niches, regulated partly through mass action and by generating distinct chemical environments that influence protein folding and function. This potential ability of condensates to generate distinct chemical environments that diverge from the cytoplasm opens exciting ways to regulate enzyme functions. Here, we consider whether a recently discovered biomolecular condensate in bacteria^1,10,11 can locally control pH to regulate function.

Recent studies have identified several mechanisms by which condensates regulate pH without a membrane barrier. First, charged acidic residues (i.e., glutamate- and aspartate-rich stretches) can enrich protons within condensates, lowering local pH as seen in nucleolar condensates and RNA-binding proteins^19,20. Second, phase separation can alter ion distribution, generating ion fluxes and interfacial electric fields as a result of variably partitioned protons and other ions, changing buffering capacity and pH^21–23. Third, localized enzymatic activity, such as ammonia production from urease-driven urea hydrolysis, elevates condensate pH²⁴. Fourth, condensates affect water distribution and solvation zones available for buffering^25,26. Collectively, these mechanisms enable condensates to establish distinct pH-controlled microenvironments.

The RNase E C-terminal domain (RNase E CTD) features an intrinsically disordered region (IDR) with distinctive alternating clusters of negatively and positively charged residues (Fig. 1B). Within BR-bodies, sticker interactions between positively charged blocks and either RNA or negatively charged blocks stabilize the condensate, but excess negatively charged spacer regions remain unpaired (Fig. 1C). This results in a high net negative charge and a low isoelectric point (pI) that is conserved across RNase E homologs in alphaproteobacteria (Fig. 1D and Supplementary Table 1). This conservation suggests that the negative charge is crucial for BR-body function. Notably, most BR-body client proteins also exhibit a net negative charge/residue (median = -0.02) (Fig. 1D). As an example, client analysis of PNPase’s structure shows that negative charge is distributed along the solvent-exposed surface (Fig. 1E). We present evidence that ionizable side chains act as intrinsic buffers within BR-bodies, maintaining an acidic pH that drives the transition from active RNA decay to RNA storage.

Results

In vivo pH measurements of Caulobacter crescentus expressing RNase E-pHluorin2

We first measured BR-bodies pH in vivo by constructing an RNase E-pHluorin2 chimera under the native RNase E promoter at the rne locus in Caulobacter crescentus. A standard curve was generated using methylamine (pK_a= 10.7), potassium benzoate (pK_a= 4.2), and either Tris-Cl or sodium acetate/acetic acid buffer at a desired pH. Methylamine and benzoate can diffuse across cell membranes in their neutral forms, whose relative proportions are dictated by the buffer’s pH, and will either capture protons or release protons, respectively, to clamp the cytoplasmic pH near to the extracellular pH. This method yielded a quantitative pH range of ~ 4.8–7.5 (Supplementary Fig. 1A). Ratiometric imaging of C. crescentus revealed a near-neutral pH of 7.0 ± 0.4 in the dilute phase and a markedly acidic pH of 5.1 ± 0.2 in the dense BR-bodies, which comprise approximately 30% of a cell’s total RNase E (Fig. 2A, B and Supplementary Fig. 1B). Fluorescence intensity profiles confirmed that high RNase E density corresponds to lower pH (Fig. 2C and Supplementary Fig. 1C–E), with acidic zones extending beyond the brightest puncta. Small speckled pixel clusters in the dilute phase are expected artifacts of the calibration curve’s reduced sensitivity near neutral pH, and small clusters (4 pixels at pH 6 and 2 pixels at pH 5) cannot be statistically distinguished from noise (Supplementary Fig. 1F; for full discussion see Methods section and Supplementary Notes). With respect to the distinct puncta, past single-molecule localization microscopy revealed a median of four RNase E clusters per cell (ranging from 1-10)²⁷. Consistent with these findings, we observed both large and small assemblies, both of which were acidic (Fig. 2). In C. crescentus pre-divisional cells, which divide asymmetrically into distinct swarmer and stalked cells, BR-bodies were more abundant in the stalked compartment (Fig. 2D), with ~ 70% of cells containing one or two BR-bodies in the stalked region and none in the swarmer compartment (Fig. 2E). Consequently, the local environment around RNase E in the swarmer compartment remained more neutral, while the stalked compartment was enriched in dense-phase acidic conditions (Fig. 2D). When normalized to their respective compartment sizes, stalked and swarmer compartments appear to contain the same concentrations of RNase E, indicating that the greater presence of BR-bodies in stalked compartments is not simply due to increased amounts of RNase E (Supplementary Fig. 1G).

Fig. 2 — A Ratiometric and fluorescent images of several *C. crescentus* cells endogenously expressing RNase E-pHluorin2 during mid-log phase. B Dense BR-bodies (n = 530 BR-bodies) and dilute phase (n = 599 dilute regions in several hundred cells) pH values were taken from many cells over three biological replicates. Calibration curves (Supplementary Fig. 1A) indicate a dynamic range of 4.8–7.4. 3.2% of BR-bodies fell below this range and were conservatively assigned a pH value of 4.8 during analysis. Measurements show a dense phase pH of 5.1 ± 0.2 and a dilute phase pH of 7.0 ± 0.4 (unpaired two-tailed t test, p $\approx 4.39 \times 10^{- 567}$ ). The bar graph shows individual data points overlaid on bars representing the mean ± SD. C Fluorescent intensity profiles of a *C. crescentus* cell using ratiometric and fluorescent images show that high-intensity and denser BR-bodies are spatially localized to the same regions that have lower pH values. D Fluorescent and ratiometric images of a model pre-divisional cell and the associated ratiometric profile illustrate that swarmer compartments have fewer BR-bodies and exhibit a more neutral pH than their corresponding stalked compartments. E Plot of pre-divisional cells and pairings of stalked and swarmer BR-body counts (n = 29 pre-divisional cells across the same samples used in panel A). Axes are labeled according to the number of BR-bodies in each compartment. Blue overlapping dots represent cells with the overlaid numbers indicating the specific counts of cells with BR-bodies in the respective compartments. There are significantly more BR-bodies located in stalked compartments (Wilcoxon signed-rank test, ***, p = $8.5 \times 10^{- 6}$ ). All Scale bars are 2 µm. Figure 2 (in vivo pH imaging): W.E.S. designed the RNase E-pHluorin2 strain, and performed ratiometric pHluorin2 imaging of mid-log phase cells, standard curve generation, and BR-body distribution analysis in pre-divisional cells.

BR-bodies exhibit structural and biochemical heterogeneity

After confirming that BR-bodies maintain acidic microenvironments, we asked whether these interiors are uniform or structurally organized. Since biomolecular condensates often exhibit spatial heterogeneity, we hypothesized that the pH gradients detected with pHluorin2 might reflect distinct substructures that influence local enzyme activity and RNA processing. To test this notion, we used single-molecule localization microscopy and fluorescence lifetime imaging microscopy (FLIM) in live cells.

To dissect structural heterogeneities within BR-bodies, we first applied single-molecule localization microscopy²⁸ using RNase E-eYFP fusion proteins (Fig. 3A). Super-resolution imaging revealed sub-diffraction clusters of 50–100 nm (Fig. 3B), showing that BR-bodies are composed of distinct nanoscale RNase E clusters rather than uniform structures. These clusters may represent localized enrichment of client proteins, RNA, or different stages of RNA degradation within specific microenvironments²⁷.

To explore environmental heterogeneity within these structural clusters, we employed FLIM with RNase E-pHluorin2²⁹. FLIM revealed intra-cluster variability, with lifetimes ranging from ~ 1 ns-2.5 ns (Fig. 3C and Supplementary Fig. 2A). Although in vitro calibration of purified RNase E CTD-pHluorin2 indicated fluorescence lifetimes spanning from ~ 2.5 ns at low pH (~ 5) to ~ 3.3 ns at higher pH (~ 9) (Supplementary Fig. 2B), our in vivo lifetimes clustered below 2.5 ns, consistent with acidic conditions. Since absolute fluorescence lifetime values in vivo may be influenced by factors beyond pH alone, such as molecular crowding or interactions within condensates, we focused on relative trends where shorter lifetimes correspond to more acidic conditions. Variable fluorescence lifetimes among BR-bodies within the same cell indicated pronounced internal biochemical complexity. The phasor plot of an individual BR-body (Fig. 3D) revealed a broad spread of lifetime components, indicating that even within a single condensate, multiple microenvironments with distinct pH states coexist. Comparison of the average fluorescence lifetimes of a cluster, standard deviation of observed lifetimes per cluster (Supplementary Fig. 2C), and pixel entropy (Supplementary Fig. 2D) further corroborated the observations of inter- and intra-cluster heterogeneity in BR-body pH.

Due to the high internal variance of lifetimes in a cluster (0.6 to 2.0 ns across all observed assemblies, examples in Fig. 3E), standard univariate statistical comparisons of mean lifetimes are insufficient to characterize the population. Instead, we employed Principal Component Analysis (PCA on segmented pixels from 65 individual clusters; Figs. 3F, G) to multi-dimensionally resolve the parameters governing these heterogeneities, providing a statistically rigorous framework to distinguish biologically relevant microenvironment shifts from stochastic noise. Further, we applied a cosine similarity-based clustering of spatial features, including lifetime variance, spatial entropy, sphericity, centroid shift, and local high-lifetime patches. These parameters capture both biochemical heterogeneity and physical organization of lifetimes in each BR-body. This multidimensional analysis revealed five distinct groups of BR-bodies exemplified in Fig. 3E, each characterized by a combination of structural and lifetime features as elucidated by the PCA (Fig. 3F). To further interpret the defining characteristics of each group, we visualized the contribution of individual features using a spider plot (Fig. 3G), which revealed clear distinctions in lifetime heterogeneity and structural asymmetry across groups. Together, this classification captures a spectrum of condensate types, from compact, homogeneous structures with uniform lifetimes to irregular, internally heterogeneous assemblies with sharp spatial gradients. These analyses suggest that BR-bodies exist in a range of biochemical states, potentially reflecting differences in protein and RNA client occupancy, spatial positioning within the nucleoid or near the poles, or stages of RNA degradation. Indeed, past proteomic analysis suggests BR-bodies contain more than 100 clients, and some proteins like Hfq³⁰ only localize to a subset of BR-bodies.

These findings suggest that BR-bodies are not uniform liquid droplets but are spatially heterogeneous assemblies exhibiting biochemical compartmentalization. The nanoscale RNase E clusters, lifetime heterogeneity, and distinct condensate classifications is reminiscent of small-world network architectures³¹, wherein a few well-connected hubs mediate communication across a sparse, highly organized network. Resembling percolated, heterogeneous networks in vitro^32–34 of prion-like low-complexity domains, BR-bodies appear to host physically and functionally distinct subdomains that enable parallel or staged RNA processing. This internal organization may allow BR-bodies to support diverse biochemical activities within the same condensate, possibly by tuning local connectivity, client access, or catalytic microenvironments.

Increased RNase E Scaffold pI leads to a more neutral BR-Body pH in vivo

We next sought to determine whether the acidic pH of BR-bodies arises from the high net negative charge within its IDR, as has been demonstrated in recent in vitro studies^19,20,22. To test this idea, we engineered Caulobacter crescentus strains in which synthetic RNase E-pHluorin2 variants with elevated pI were expressed under the native RNase E promoter at the rne locus (Fig. 4A). As an example, this involved mutations of negatively charged residues (21 glutamates and 17 aspartates) to uncharged polar residues (21 serines and 17 threonines) to generate the RNase E–pI7 variant (Fig. 4A). Ratiometric pHluorin2 imaging of mid-log phase Caulobacter cells expressing wild-type (WT) RNase E, RNase E–pI6, or RNase E–pI7 revealed average BR-body pH values of 5.1 ± 0.2, 5.8 ± 0.3, and 6.3 ± 0.5, respectively (Fig. 4B, C). In contrast, dilute-phase cytoplasmic pH values were more similar across strains and closer to neutral (7.1 ± 0.5, 6.8 ± 0.3, and 7.4 ± 0.4, respectively) in Fig. 4B, C. In a similar manner, we observed heterologous expression in E. coli of an RNase E CTD homolog from Bartonella henselae using a corresponding pHluorin2 chimera (estimated pI = 9.5), which yielded a pH of 9.3 ± 0.2 (Supplementary Fig. 3A–C). Notably, the fraction of C. crescentus cells lacking detectable BR-bodies increased from 13.3% in wild-type to 35.0% in the pI6 strain and 53.3% in the pI7 Caulobacter strain (Fig. 4D) with BR-bodies also containing a smaller total cell fraction of RNase E in pI7 (Supplementary Fig. 1B) despite roughly equivalent total cell intensity levels (Supplementary Fig. 3E). Together, these results demonstrate that reducing the net negative charge of the RNase E IDR increases the apparent pH of BR-bodies and decreases the number of BR-bodies/cell in vivo.

Fig. 4 — A Schematic of the engineered mutations within the intrinsically disordered region (IDR) of RNase E in *Caulobacter crescentus*. Acidic residues (Asp, Glu; red) were selectively mutated to Ser or Thr to reduce net negative charge and generate variants with predicted pI values of 6 or 7. Basic residues (Arg, Lys) are shown in blue and neutral residues in gray. Colored boxes represent unperturbed binding sites of known client proteins (Green = aconitase, Magenta = RNase D, Blue = PNPase) and (B) Ratiometric and fluorescent images of mid-log phase and stationary phase cells of WT (*rne::rne-pHluorin2*), pI6 (*rne::rne-pI6-pHluorin2*), and pI7 (*rne::rne-pI7-pHluorin2*) strains. Scale bar = 2 µm. C Calibration curves (Supplementary Fig. 1A) indicate a dynamic range of 4.8–7.4. For measurements where ratios fell below the lowest point on the standard curve, mid-log phase WT cells had 3.2% that fell below a pH of 4.8. Stationary phase WT cells, pI6 cells, and pI7 cells had 48.2, 2.5, and 4.4% of BR-bodies, respectively, that fell below a pH of 4.8. In all cases, these were assigned a value of 4.8 for analysis. The resulting quantification of apparent pH within dense-phase BR-bodies (n = 2722) compared to surrounding dilute cytoplasm (n = 257 regions across several hundred cells) across three biological replicates (one-way Welch’s ANOVA, two-tailed Games-Howell’s multiple comparisons test, mid-log phase $p \approx 9.63 \times 10^{- 48}$ and stationary phase $p \approx 2.96 \times 10^{- 82}$ ), and (D) BR-bodies per cell, for *C. crescentus* WT, pI6, and pI7 strains (non-directional chi-square test of independence, χ²(30), p $= 3.1 \times 10^{- 12}$ ) across three biological replicates is shown. Bar graphs in panel C show individual data points overlaid on bars representing the mean ± SD. E Representative Differential Interference Contrast (DIC; top) and Single-Molecule Localization Microscopy (SMLM; bottom) images of *Caulobacter crescentus* cells expressing RNase E–eYFP fusions. White dashed lines indicate approximate cell boundaries derived from DIC images. Scale bar = 1 µm. F Quantification of the SMLM images reveals WT RNase E organizes into large, high-density BR-bodies, which exhibit high localization counts per cluster. In contrast, shifting the pI of the RNase E scaffold alters the condensation landscape: the pI6 mutant forms smaller, medium-density clusters, while the pI7 mutant results in small, sparse assemblies that are not present in every cell. Localization density within clusters was determined via density-based spatial clustering of applications with noise (DBSCAN) with the heatmap indicating the number of localizations per cluster. No significant differences in cluster area or localizations per cluster were detected between any pairwise strain comparisons (WT, pI6, pI7; Mann–Whitney U test and Levene’s test, all p > *0.05*; n = 110, 195, and 58 clusters, respectively; 1 biological replicate, 4–7 technical replicates per condition). Box-and-whisker plots show cluster area distributions for WT, pI6, and pI7 conditions following DBSCAN clustering (aspect ratio ≤ 1.9; cluster area 0.001–0.03 µm²). Boxes represent the interquartile range (IQR; 25th–75th percentile), with the center line indicating the median. Whiskers extend to the most extreme data point within 1.5 × IQR of the box edges; points beyond this range are classified as outliers and excluded from the plot for clarity. For WT (n = 110 clusters), the median cluster area was 0.0028 µm² (Q1 = 0.0017 µm², Q3 = 0.0059 µm², IQR = 0.0041 µm²), with whiskers spanning 0.0010–0.0120 µm² and 8 outliers (max = 0.0288 µm²). For pI6 (n = 195), the median was 0.0025 µm² (Q1 = 0.0014 µm², Q3 = 0.0072 µm², IQR = 0.0058 µm²), whiskers spanning 0.0010–0.0155 µm² and 18 outliers (max = 0.0298 µm²). For pI7 (n = 58), the median was 0.0028 µm² (Q1 = 0.0017 µm², Q3 = 0.0055 µm², IQR = 0.0038 µm²), whiskers spanning 0.0010–0.0105 µm² and 9 outliers (max = 0.0294 µm²). Individual data points are overlaid and color-coded by number of localizations per cluster (colormap range: 50–500 localizations). G Ratiometric images of free pHluorin2 lacking the *rne* fusion in *C. crescentus* indicate an acidic cytoplasm in the stationary phase relative to a neutral cytoplasm in mid-log phase growth across three biological replicates (n = 100 cells for mid-log phase and n = 217 cells for the stationary phase) (unpaired two-tailed t test, p $\approx 1.01 \times 10^{- 79}$ ). Calibration curves (Supplementary Fig. 3F) indicate a dynamic range of 5.0–6.8. Only mid-log phase had cells with ratios outside of the dynamic range with 85% of cells falling above the highest point on the calibration curve. They were assigned a pH value of 6.75 for analysis. Scale bar = 2 µm. Figure 4 (pI variant strains and in vivo imaging): W.E.S. and K.G.D. designed the RNase E-pHluorin2 chimeric constructs. K.S.M., H.Y., and K.H.D. constructed the wild-type, pI6, and pI7 *Caulobacter crescentus* strains. W.E.S. performed ratiometric imaging of these strains in mid-log and stationary phase. M.S. performed DBSCAN cluster analysis.

SMLM analysis indicates RNase E pI tunes BR-body phase properties

To investigate how the RNase E scaffold pI dictates the physical dimensions and molecular occupancy of BR-bodies, we performed super-resolution microscopy on these pI6 and pI7 variants (Fig. 4E). Super-resolution reconstructions reveal that these subtle charge modifications profoundly alter the condensation landscape: while the pI6 variant enhances assembly, the pI7 mutation severely disrupts phase separation, yielding only small, sparse, and low-density clusters that are not present in every cell. Quantitative analysis via density-based spatial clustering of applications with noise (DBSCAN) confirms these observations (Fig. 4F), showing that pI6 facilitates the formation of clusters with highly variable molecular counts that range from low-occupancy nucleation sites to high-density mature bodies. In stark contrast, the pI7 assemblies are restricted to a narrow, low-density regime. Interestingly, the increased frequency of pI6 clusters across disparate size and density scales suggests a more fluid material state that favors growth via coalescence³⁵. This likely arises from the substitution of strong hydrogen-bond acceptors (Asp, Glu) with weaker species (Ser, Thr), which weakens specific “sticker” interactions. While this weakening promotes a more pervasive and fluid assembly, there is a clear thermodynamic limit; further weakening in the pI 7 variant raises the critical concentration (C_sat) to a level that nearly abolishes stable condensate formation under physiological conditions. Collectively, these data demonstrate that scaffold pI acts as a tunable control, where slight charge shifts govern whether a protein maintains a broad distribution of functional assemblies or fails to phase separate entirely.

Stationary phase drives BR-body and cytoplasm acidification

We next asked whether BR-bodies maintain pH during the stationary phase, as BR-bodies undergo a change in function to RNA storage³⁶. Ratiometric pHluorin2 imaging of stationary phase Caulobacter cells expressing WT RNase E, RNase E–pI6, or RNase E–pI7 revealed average BR-body pH values of 4.8 ± 0.2, 5.3 ± 0.3 and 5.2 ± 0.3, respectively (Fig. 4B, C). This apparent BR-body pH was 0.3 units lower for wild-type, 0.5 units lower for pI6 and 1.1 units lower for pI7 compared to mid-log phase. Thus, during the stationary phase, BR-bodies undergo additional acidification, suggesting other factors contribute to BR-body pH regulation beyond RNase E’s net negative charge.

In comparison, dilute-phase cytoplasmic pH (Fig. 4B, C) exhibited an unexpected decrease in pH in the stationary phase for wild-type and pI6, (5.5 ± 0.2, 6.1 ± 0.5, respectively) and relative maintenance of dilute phase pH for pI7 (6.8 ± 0.6). Free pHluorin2 expression with a dynamic range of 5.0–6.8 pH (Supplementary Fig. 3F) also indicated a mean stationary-phase pH of 5.3 ± 0.1 compared to a mid-log pH of greater than 6.8 (Fig. 4G). This result indicates that the stationary phase cytoplasm becomes acidic in a manner similar to observations in yeast and Salmonella under stresses^37–41. Notably, we observed that rne-pI6-pHluorin2 and rne-pI7-pHluorin2 Caulobacter strains exhibited more cells with no BR-bodies than wild-type (29.6% and 46.8% versus 10.7% in wild-type in Fig. 4D). This loss of BR-bodies indicates that the negative charge of RNase E’s IDR also contributes to robustness in phase separation.

RNase E–pI7 is not viable when grown from stationary phase cultures

To determine whether variation in BR-body pH carries phenotypic consequences, we analyzed growth and morphology in mid-log versus stationary phase. In mid-log phase, we observed a buildup of constricted pre-divisional Caulobacter cells from 15.6% in wild-type to 25.9% in the rne-pI6-pHluorin2 strain and 56.9% in the rne-pI7-pHluorin2 strain (Fig. 5A).

To assess how variation in RNase E pI impacts growth, we compared mid-log and stationary phase cultures using an efficiency of plating assay (Fig. 5B, C). Overall, BR-body acidic pH appears dispensable for logarithmic growth in Caulobacter (Fig. 5A). However, when grown from the stationary phase, the rne-pI7-pHluorin2 strain failed to grow, while the wild-type strain grew normally (Fig. 5B, C). This result suggests that the negative charges of RNase E’s IDR, and potentially the BR-body acidic pH, are critical for cell viability when resuming growth from stationary phase. Alternatively, this phenotype may stem from the reduced BR-body phase separation observed for the rne-pI7-pHluorin2 strain (Fig. 4D–F).

One of the primary known functions of BR-bodies is to facilitate RNA decay during logarithmic phase growth, then switch to RNA storage in stationary phase³⁶. To determine if this functional switch remains intact, we used qRT-PCR to measure ctrA transcript half-life in wild-type, rne-pHluorin2, and rne-pI7-pHluorin2 Caulobacter strains during logarithmic and stationary phase growth (Fig. 5D). As expected, ctrA half-lives were consistent across all strains in mid-log phase (1.3 ± 0.1 min. for each sample). In the stationary phase, the wild-type strain exhibited a ctrA half-life of 13.6 ± 7.0 min., and fusion of pHluorin2 to the C-terminus had minimal impact (10.2 ± 2.6 min). However, the rne-pI7-pHluorin2 strain showed a 1.8-fold reduction in ctrA half-life (5.7 ± 1.4 min.) compared to the rne-pHluorin2 strain. This decrease suggests that reducing negative charges in RNase E’s IDR impairs the ability of BR-bodies to effectively store RNA transcripts during stationary phase, likely due to the combined effects of diminished BR-body phase separation and neutralization of BR-body pH (Fig. 4).

Given that the rne-pI7-pHluorin2 Caulobacter strain fails to resume growth from stationary phase (Fig. 5B, C), we next considered the strain’s growth in glucose-supplemented spent media (Fig. 5E, F), which reflects the nutrient-depleted and metabolite-accumulated media environment characteristic of late stationary phase. We observed that wild-type grew in spent media with no obvious morphological defects. In comparison, about 20% of the rne-pI7-pHluorin2 Caulobacter cells exhibited severe morphological defects with noticeable blebbing from the cell surface (Fig. 5E, F). These results suggest that the reduction of RNase E IDR’s negative charges leads to a growth defect under nutrient-challenged and metabolite-accumulated conditions.

Ethanol stress is characterized by a decrease in BR-body pH

Having established that BR-body pH decreases during stationary phase growth (Fig. 4B, C), we next asked whether a well-studied environmental stress similarly affects BR-body pH. Previous work has shown that growth of Caulobacter in 5% ethanol causes severe cell filamentation⁴² and that BR-body phase separation is critical for fitness under these conditions⁵. We therefore investigated whether BR-body pH changes during ethanol stress. Consistent with prior observations, wild-type cells grown in 5% ethanol for 24 hours exhibited cell lengths of 13.2 ± 5.3 µm, while the rne-pI7-pHluorin2 strain showed increased filamentation at 17.6 ± 6.5 µm under the same conditions (Fig. 6A, B). We then asked whether ethanol exposure also alters BR-body pH, as observed during stationary phase (Fig. 4B, C). Wild-type cells exposed to 5% ethanol for 24 hours showed a significant reduction in apparent BR-body pH, from 5.1 ± 0.3 to 4.8 ± 0.1, with 86.2% of BR-bodies displaying apparent pH values below the lower limit of our standard curve (pH 4.8), suggesting their true pH may be considerably more acidic (Fig. 6C). The rne-pI7-pHluorin2 strain BR-bodies similarly acidified under ethanol stress, decreasing from 5.9 ± 0.7 to 5.2 ± 0.4 (Fig. 6A, C, FLIM analysis in Supplementary Fig. 3G). Wild-type cells expressing free pHluorin2 showed a decrease in cytoplasmic pH under stresses from stationary phase after 24 hours (Fig. 4G) that was not altered by the presence of ethanol (Supplementary Fig. 3H). Together, these results demonstrate that ethanol stress drives BR-body acidification in both wild-type and rne-pI7-pHluorin2 cells in concurrence with a general reduction of cytoplasmic pH in the stationary phase after 24 h.

Fig. 6 — A Representative phase-contrast, fluorescence, and ratiometric pHluorin images of WT (*rne::rne-pHluorin2*), pI6 (*rne::rne-pI6-pHluorin2*), and pI7 (*rne::rne-pI7-pHluorin2*) strains grown for 24 h in PYE without stress or in PYE supplemented with 5% ethanol. Scale bar is 2 µm. B Quantification of cell length under no-stress (n = 834 cells) and ethanol stress conditions (n = 237 cells) across three biological replicates (unpaired two-tailed t-test, p $= 4.00 \times 10^{- 10}$ for No Stress and p $= 2.51 \times 10^{- 7}$ for 5% EtOH stress). Graphs show individual data points where the horizontal line represents the mean and the error bars represent the SD. C Histograms of BR-body pH in WT and pI7 strains with and without 5% ethanol stress conditions (unpaired two-tailed t-test, p $\approx 1.39 \times 10^{- 30}$ for WT vs. WT EtOH and p $\approx 6.07 \times 10^{- 34}$ for pI7 vs. pI7 EtOH). Of the 260 WT BR-bodies examined, 43.4% fell below the standard curve dynamic range of 4.8–7.4 (Supplementary Fig. 1A). For the 334 WT EtOH BR-bodies, 86.2% fell below this range. For the 658 pI7 BR-bodies, 0.1% fell below the range. Lastly, for the 1479 pI7 EtOH BR-bodies, 27.7% fell below the range. In all cases, these data points were assigned the conservative value of 4.8. Figure 6 (ethanol stress): W.E.S. performed ethanol stress assays.

C-SNARF and pHluorin2 reveal an RNase E condensate pH gradient in vitro

To determine if the RNase E scaffold of BR-bodies was minimally sufficient to generate an acidic microenvironment, we measured RNase E condensate pH in vitro. Past studies utilizing C-SNARF dyes and pHluorin2 estimated the pH of polyaspartate and polyglutamate coacervates¹⁹, FUS²², BSA condensates containing urease²⁴, the nucleolus²⁰, and synthetic resilin- and elastin-like condensates²¹. We, therefore, compared two ratiometric pH probes to understand the pH environment within RNase E CTD condensates in vitro: the free small molecule C-SNARF-4F (Fig. 7A) and the fluorescent protein pHluorin2 (Fig. 7B). C-SNARF-4F exists in equilibrium between two protonation states with a pK_a of 6.4 and a dynamic range of ~5-8 (Supplementary Fig. 4A). One benefit of the dye is that it can be used in dilute conditions, reducing quenching from crowded dye. Notably, Keating and colleagues proposed that anionic dyes entering condensates may interact with positive polyions, leading to pK_a shifts that affect their response to proton concentration¹⁹.

Fig. 7 — A Representative DIC and ratiometric images of RNase E CTD condensates incubated with pH-sensitive C-SNARF-4F dye with quantitative analysis of average condensate (n = 105 condensates) pH (yellow) relative to the dilute phase (cyan) across three biological replicates (unpaired two-tailed t test, p $\approx 4.18 \times 10^{- 87}$ ) All points fell within the calibration range of 5.9–8.1 (Supplementary Fig. 4A). Bar graphs show individual data points overlaid on bars representing the mean ± SD. B Representative DIC and ratiometric images from three biological replicates and quantitative analysis of RNase E CTD-pHluorin2 condensates (n = 645 condensates) exposed to increasing concentrations of Tris-Cl buffer. Higher concentrations of the Tris-Cl buffer result in the loss of the pH gradient between the dense and dilute phases (one-way Welch’s ANOVA, two-tailed Games-Howell’s multiple comparisons test, p $\approx 6.20 \times 10^{- 456}$ ). All points fell within the calibration range of 6.2–8.1 (Supplementary Fig. 4B). Bar graphs show individual data points overlaid on bars representing the mean ± SD. C Representative images and quantitative analysis of RNase E CTD-pHluorin2 condensates (n = 39 condensates) with 20 mM MgCl₂ (n = 157 condensates) and 500 mM NaCl (n = 207 condensates). Adding MgCl₂ leads to a modest increase in condensate pH from 6.5 ± 0.1 to 6.8 ± 0.2 (unpaired two-tailed t test, p $\approx 6.41 \times 10^{- 65}$ ) while adding NaCl results in a statistically insignificant change to 6.4 ± 0.2 (unpaired two-tailed t test, p = *0.14* for comparison to 100 mM NaCl shown in Supplementary Fig. 4F). With a calibration range of 6.2-8.1 (Supplementary Fig. 4B), only the 500 mM NaCl condition had points below the lower bound of the standard curve with 4.4% having a pH below 6.2 to which they were conservatively assigned. Bar graphs show individual data points overlaid on bars representing the mean ± SD. D Profiles of RNase E CTD-pHluorin2 with and without 20 mM MgCl₂ show the diminished pH gradient in the presence of Mg²⁺ and a loss of a distinct zone of intermediate pH between the dense and dilute phases. E Representative images and quantitative analysis from three biological replicates of RNase E CTD-pHluorin2 with PNPase, aconitase, and poly(A) (n = 224 condensates). The addition of these clients does not result in a significant change in the pH of the dense and dilute phases, with values of 6.4 ± 0.1 and 7.4 ± 0.2 (unpaired two-tailed t test, p $\approx 2.10 \times 10^{- 107}$ ), respectively. With a calibration range of 6.2–8.1 (Supplementary Fig. 4B), 5.7% of condensates recorded values below 6.2. They were conservatively assigned this value during analysis. The bar graph shows individual data points overlaid on bars representing the mean ± SD. All scale bars are 5 µm. Figure 7 (in vitro condensate pH): W.E.S. purified proteins, reconstituted in vitro condensates, and performed C-SNARF-4F and RNase E CTD-pHluorin2 pH measurements, buffer titration experiments, ion effect experiments, and client protein pH impact assays. S.L.M. performed zeta potential measurements.

We also employed an RNase E CTD-pHluorin2 chimera, which features a pH-sensitive SYG chromophore with a pK_a of 7.0. The β-barrel of pHluorin2 shields the chromophore from negatively charged polyions, reducing direct charge interactions with the chromophore and enhancing sensitivity to proton concentrations. Based on our in vitro pH titration standard curves, we found that pHluorin2 has a dynamic range of ~ 6–8 (Supplementary Fig. 4B).

C-SNARF-4F measurements indicated that RNase E condensates assembled in a 20 mM Tris-Cl buffer of pH 7.5 have an internal pH of 6.7 ± 0.2, compared to a dilute phase pH of 7.7 ± 0.1. Similarly, RNase E CTD-pHluorin2 condensates exhibited a dense phase pH of 6.5 ± 0.1 and a dilute phase pH of 7.6 ± 0.2. Both biosensors consistently detected a pH difference of 1.1 ± 0.2 between dense and dilute phases (Fig. 7A, B). These results from two biosensors with distinct pH-sensing mechanisms collectively support that RNase E CTD condensates maintain a more acidic interior in vitro.

Given the role of excess negative charge that resulted in acidic microenvironments in other condensate systems^19,20, we hypothesized that the high density of aspartates and glutamates in RNase E CTD’s dense phase plays a role in buffering condensate pH. We investigated whether increasing Tris-Cl buffer concentrations, which raise the buffer-to-scaffold ratio within RNase E CTD condensates, could counteract the scaffold’s influence to neutralize the pH gradient. At Tris-Cl concentrations between 20–50 mM, the condensate maintained a lower pH of ≥ 0.80 pH units. In comparison, we observed an elimination of the condensate pH gradient at Tris-Cl buffer concentrations of 100-500 mM (Fig. 7B). This suggests that Tris-Cl is at sufficient concentrations within the condensate to act as a buffer, or it may impact crowding or the degree of hydration within the condensate.

Next, we examined how ions regulate condensate pH because phase separation could mediate selective partitioning of ions between dense and dilute phases, leading to charge asymmetry and varied zeta potentials (Supplementary Fig. 4C, D) as a result of induced interfacial electric potentials²³. To test the impact of ions on the observed pH gradient, we examined the effect of adding MgCl₂ and NaCl on RNase E CTD condensate pH. The addition of MgCl₂ led to a statistically significant but mild increase in pH of 6.5 ± 0.1 to 6.8 ± 0.2 (Fig. 7C and Supplementary Fig. 4E), indicating more specific interactions with negatively charged residues. Moreover, the larger pH gradient without MgCl₂ displays a gradual shift from the dilute to dense phase pH (Fig. 7D), suggesting a possible effect on the interfacial region pH as MgCl₂ alters ionic strength and Debye length⁴³. In contrast, NaCl concentrations as high as 500 mM had no statistically significant impact on the condensate pH (Fig. 7C and Supplementary Fig. 4F). The insensitivity of BR-body pH to NaCl concentrations up to 500 mM suggests that Na⁺ ions may not compete for carboxylate proton binding sites the way Tris or Mg²⁺ does, indicating the intrinsic buffering capacity of the densely packed acidic IDR residues is robust to monovalent salt perturbation. This observation is consistent with DFT calculations showing that replacing divalent cations with Na⁺ in carboxylate-rich environments is energetically unfavorable⁴⁴ and experimental evidence that Mg²⁺ forms stable contact ion pairs with glutamate carboxylates while Na⁺ is much less stable⁴⁵.

RNase E’s primary clients do not alter RNase E condensate pH in vitro

Biomolecular condensates contain scaffolds that drive phase separation and client proteins that dynamically associate within the assemblies, provoking the question of how clients impact RNase E condensate pH since BR-bodies are scaffolded by RNase E and possess more than 100 clients³⁰. These clients have a range of net charges and molecular weights that could modulate condensate pH. Interestingly, over two-thirds of the clients carry a net negative charge, with a median charge/residue of -0.02 (Fig. 1D). Examples include three well-characterized clients that bind at near stoichiometric amounts: the exoribonuclease PNPase, the citric acid metabolic enzyme aconitase, and unstructured RNA such as poly(A). Here, we examine the impact of client proteins on the buffering capacity of RNase E CTD condensates.

Without client proteins, RNase E CTD-pHluorin2 displayed an average dense and dilute phase pH of 6.5 ± 0.1 and 7.6 ± 0.2, respectively (Fig. 7B). The average condensate pH was also 6.5 ± 0.1 with 5 µM PNPase, 5 µM aconitase, or 80 ng/µL poly(A) (Supplementary Fig. 4G). Titrating PNPase from 0-40 µM found statistical significance between 0–20 µM PNPase but with an effect size of 0.1–0.2 pH units, indicating large overlaps in their pH distributions (Supplementary Fig. 4H). With PNPase, aconitase, and poly(A) assembled simultaneously with RNase E CTD, the mean dense and dilute pH also did not change with values of 6.4 ± 0.1 and 7.4 ± 0.2, respectively (Fig. 7E). These results indicate no significant change of RNase E CTD condensate pH upon client addition in vitro, suggesting that the RNase E CTD condensate acidic environment is robust to alterations in its primary client composition that include changes in RNA composition during RNA degradation.

Acidic condensate environments enhance the activity of PNPase

We considered how this robust acidic environment impacts the activity of associated enzymatic clients. We focused on using Thioflavin T, which increases its fluorescence (Ex = 450 nm, Em = 482 nm) upon binding poly(A) due to restricted rotation between its benzothiazole and benzylamine groups⁴⁶. Its emission intensity increases with higher poly(A) concentrations and decreases as PNPase degrades poly(A), reducing binding sites (Fig. 8A). Notably, when PNPase is inactive due to missing Mg²⁺, the emission remains unchanged (Supplementary Fig. 5A).

Fig. 8 — A Representation of enhanced enzyme activity due to condensate acidic microenvironment using Thioflavin T to track poly(A) degradation by PNPase. B The percentage of poly(A) degradation over three biological replicates by PNPase was measured at several pH values with a maximal activity of 68.0% ± 6.9% degraded at pH 5.3 (one-way ANOVA, Tukey’s HSD, p $= 1.12 \times 10^{- 9}$ ). Bar graphs show individual data points overlaid on bars representing the mean ± SD. C Partitioning of polyA-Cy5 (n = 494 condensates) and PNPase-mCherry (n = 611 condensates) into RNase E condensates with low Tris-Cl (acidic interior) versus high Tris-Cl (neutral interior) (unpaired two-tailed t test, p = *0.06* for PNPase, p = *0.20* for poly(A)). Bar graphs show individual data points overlaid on bars representing the mean ± SD. D Fluorescence imaging and relative degradation rates reveal that poly(A) degradation is strongly localized within phase-separated condensates that maintain a pH gradient (low Tris-Cl), while minimal activity is observed in the phase-separated condensates that lack a pH gradient (high Tris-Cl). E Quantitative analysis of condensates (n = 47 condensates) individually reveals that the relative rate of Thioflavin T intensity change (dense/dilute ratio) shows an increased relative degradation of 13.5 ± 2.8 within phase-separated condensates that maintain a pH gradient compared to the phase-separated condensates lacking a pH gradient that show a relative degradation of 1.4 ± 0.4 (unpaired two-tailed t test, p $\approx 7.14 \times 10^{- 27}$ ). Bar graphs show individual data points overlaid on bars representing the mean ± SD. F Fluorescence imaging and relative polymerization rates reveal that poly(A) polymerization is strongly localized within phase-separated condensates that maintain a pH gradient (low Tris-Cl), while minimal activity is observed in the phase-separated condensates that lack a pH gradient (high Tris-Cl). G Analyzing the condensates (n = 22 condensates) individually reveals that the relative rate of Thioflavin T intensity change (dense/dilute ratio) shows an increased relative poly(A) polymerization of 1.8 ± 0.2 within phase-separated condensates that maintain a pH gradient (low Tris-Cl) compared to the phase-separated condensates lacking a pH gradient (high Tris-Cl) that show a relative degradation of − 0.2 ± 0.6 (unpaired two-tailed t test, p = $7.3 \times 10^{- 8}$ ). However, the phase-separated condensates that lack a pH gradient (high Tris-Cl) have a relative polymerization rate that is not significantly different than 0 (one-sample t test, p = *0.24*). Scale bars are 5 µm. The bar graph shows individual data points overlaid on bars representing the mean ± SD. Figure 8 (enzyme activity assays): W.E.S. performed Thioflavin T binding assays, plate reader activity assays, and epifluorescence imaging of PNPase degradation and polymerization activity.Writing and supervision: W.E.S. wrote the original manuscript draft and prepared figures. W.E.S., W.S.C., J.M.S., and S.S. reviewed and edited the manuscript. W.S.C. supervised W.E.S. and K.G.D.; J.M.S. supervised K.S.M., H.Y., and K.H.D.; S.S. supervised M.S. and R.C.; and J.E.M. supervised S.L.M. W.S.C., J.M.S., and S.S. secured funding as described in the Funding Statement. All authors contributed to the final manuscript.

We measured fluorescence from bound Thioflavin T:poly(A) before and after 30-second incubations with PNPase using a bulk fluorescence plate reader assay. Applying pH-dependent standard curves (Supplementary Fig. 5B), we observed PNPase activity between pH 3.5-7.5, with a peak at acidic levels (pH 5.3–6.2) like those found in BR-bodies at mid-log phase (Fig. 2B). Specifically, poly(A) degradation was highest at pH 5.3 with 68.0% ± 6.9% degraded but sharply decreased to 11.0% ± 3.2% at pH 4.9. At pH values above 5.3, the decrease was gradual, with 51.8% ± 5.5% degradation at pH 5.5 and 47.1% ± 4.9% at pH 6.2 (Fig. 8B). Notably, the average stationary-phase BR-body pH of 4.8 ± 0.2 falls at or just below the observed PNPase activity range of pH 4.9 and 5.3 (Fig. 8B), placing stationary-phase BR-bodies in a regime where small pH fluctuations could toggle between residual and negligible RNA decay activity.

Bulk assays reveal overall reaction dynamics but cannot distinguish activity in the alkaline/neutral dilute phase and the acidic dense phase. To address this, we used epifluorescence to monitor Thioflavin T fluorescence, which reflects poly(A) concentration changes due to PNPase activity. First, to assess for partitioning-based differences in subsequent poly(A) degradation measurements, we found that poly(A) and PNPase partition equally in low Tris-Cl (acidic condensate interior) versus high Tris-Cl (neutral condensate interior), which indicates that Tris-Cl and RNase E condensate pH does not regulate partitioning of PNPase or its substrate (Fig. 8C). Under these conditions where PNPase and poly(A) phase separated with RNase E CTD, we calculated the average poly(A) degradation rate at each pixel. By normalizing each pixel’s rate to the dilute background, we created a relative rate image that spatially maps PNPase activity (Fig. 8D). Under low Tris-Cl conditions, which maintain a lower pH inside the condensates, the dense phase sometimes degraded poly(A) over 26-fold faster than the dilute phase, with an average 13.5 ± 2.8-fold increase. In contrast, under high Tris-Cl conditions, where the pH is equal in both phases, the increase was only 1.4 ± 0.4-fold (Fig. 8E). These results directly demonstrate that the acidic environment within RNase E condensates enhances PNPase degradation activity.

We also visualized the rate of PNPase polymerase activity, where ADP is polymerized into poly(A), using the same imaging and relative rate analysis (Fig. 8F). Starting with only ADP monomers (lacking poly(A) and inorganic phosphate), our relative rates revealed that under low Tris-Cl conditions, which maintain a lower dense phase pH, polymerization is enhanced by a factor of 1.8 ± 0.2 compared to the dilute background. Under high Tris-Cl conditions, with a uniform pH, the rate was not significantly different from zero (Fig. 8G). These results indicate that a low pH environment, facilitated by RNase E phase separation, enhances PNPase polymerase activity. However, RNA breakdown in the condensates remains more pronounced, supporting the role of RNase E condensates as RNA decay sites.

Discussion

This work illuminates that the chemical compositions of biomolecular condensates can play critical roles in regulating biochemical pathways and reinforces emerging concepts of condensate pH regulation. Choi et al. found that condensates formed from model polycation-polyanion mixtures exhibited apparent pH differences between dense and dilute phases that increased with polyion length mismatch¹⁹. In a complementary finding, King et al. showed that negatively charged IDR tracts establish nucleolar pH gradients through proton accumulation²⁰. Further studies developed a thermodynamic model demonstrating that dense-phase pH shifts toward the scaffold pI to minimize electrostatic repulsion²². Collectively, these studies directly predict our observation that the highly negative RNase E IDR (net charge – 38, pI ~ 4.7) maintains the pH of BR-bodies to ~ 5.1, nearly 2 units below the cytoplasm. Thus, scaffold net charge and pI appear to be predictors of condensate pH.

A major open question in the field is whether condensate pH impacts cell physiology. Here, we demonstrated that genetic perturbation of RNase E IDR charge in living Caulobacter progressively raises BR-body pH and impairs phase separation robustness. Our work showed that while acidic BR-body pH is dispensable for logarithmic growth, the rne-pI7-pHluorin2 strain fails to resume growth from stationary phase (Fig. 5B, C), which is consistent with observed impaired RNA storage (Fig. 5D), and displays morphological defects in spent media (Fig. 5E, F) and ethanol stress (Fig. 6). Together, these findings reveal that the RNase E scaffold’s net negative charge influences BR-body pH and BR-body assembly, both of which may be important features that enable Caulobacter to survive nutrient limitation and environmental stress. We propose a model in which stress-driven BR-body acidification lower than a threshold pH, combined with material property changes, may inhibit RNA decay activity within BR-bodies, potentially converting BR-bodies from active sites of RNA turnover during logarithmic growth into protective RNA storage compartments during stress as described in past work³⁶. In addition, since RNase E catalysis requires two conserved aspartate residues to coordinate Mg²⁺ for phosphodiester cleavage⁴⁷, stress-induced BR-body acidification to below pH 5 may directly silence the rate-limiting endonucleolytic step that licenses downstream mRNA decay, though this remains to be tested. In total, this control of condensate pH is associated with cell viability upon return to favorable growth conditions, though the relative contributions of pH and phase separation capacity remain to be disentangled (Fig. 9).

Fig. 9 — During rapid mid-log growth, the BR-body dense phase maintains a pH that supports active RNA decay. Upon transition to stationary phase or exposure to stress (e.g., ethanol), BR-bodies undergo localized acidification and increased material solidification. These changes shift RNA decay client enzymes within the dense phase into a pH regime that suppresses their catalytic activity. Importantly, the cytoplasmic dilute phase remains within a pH range that permits RNase E activity. As a result, transcripts partitioned into BR-bodies are protected and stored, whereas RNAs remaining in the dilute phase continue to be degraded. This spatial separation of pH-dependent enzymatic activity enables BR-bodies to function as stress-responsive regulators that toggle between RNA decay and RNA storage states. *Caulobacter crescentus* cell cartoon adapted from W.E. Schnorr and W.S. Childers, *ChemSystemsChem*, 2024, https://doi.org/10.1002/syst.202400011, licensed under CC BY-NC 4.0.

Unexpectedly, we observed cytoplasmic acidification occurs in the stationary phase that may be linked to the observation of the C. crescentus cytoplasm having glassy material properties in stationary phase⁴⁸. Moreover, cytoplasmic acidification may share similarities to the cytoplasmic acidification that occurs during stress in yeast and Salmonella^{37,39–41,49}. Establishing the generality of stress-induced cytoplasmic pH changes remains an important avenue for future investigation.

Condensate pH has become an emergent chemical property of biomolecular condensates, and work here and by others suggests it may play a critical role in enzymatic regulation. To understand the significance of condensate pH, new tools are needed to change condensate pH and measure its functional significance. Here, we manipulated BR-body pH by altering the net negative charge of the RNase E IDR scaffold, providing a tractable approach to study the functional consequences of condensate pH. However, a limitation of our study is that reducing the IDR’s negative charge also decreased BR-bodies per cell. As a result, the phenotypes observed in the rne-pI7-pHluorin2 strain likely reflect a combination of altered BR-body pH and reduced phase separation capacity. In addition, measurements near the boundaries of our empirical ratiometric dynamic range (i.e., pH 4.8–7.5 for rne-pHluorin2 and pH 5.0–6.8 for free pHluorin2) should be interpreted with caution, as sensor sensitivity diminishes outside of this regime. Understanding the consequences of condensate pH in diverse systems fully will require the development of strategies to measure and manipulate condensate pH independently of scaffold sequence, such as recruitment of enzymes whose products could alkalinize²⁴ or acidify condensate pH.

Super-resolution imaging and FLIM reveals that BR-bodies exhibit structural complexity, with distinct nanoscale clusters of RNase E that maintain heterogeneous acidic microenvironments rather than uniform internal conditions (Fig. 3A–D). The distribution of fluorescence lifetimes observed within single BR-bodies demonstrates that multiple pH microenvironments are present simultaneously on a sub-condensate scale (Fig. 3E), suggesting the presence of functionally specialized subdomains. Principal component analysis further classified BR-bodies into five groups with diverse structural and lifetime features. These groups ranged from compact and homogeneous to irregularly shaped heterogeneous condensates (Fig. 3F, G). These variations indicate that biomolecular condensates may adopt a variety of organizational and compositional states that may be critical for their multifunctional roles in RNA metabolism. Indeed, past proteomic studies of BR-bodies suggest that BR-bodies may exhibit significant compositional heterogeneity³⁰.

While condensates are well known to accelerate chemical reactions through mass action, we also found their internal pH can regulate enzyme activity. In vitro, within acidic RNase E CTD condensates, PNPase activity increased 13.5-fold compared to the dilute phase, whereas neutral RNase E condensates only led to a 1.4-fold enhancement (Fig. 8D, E) despite similar levels of PNPase and poly-A partitioning (Fig. 8C). This pH sensitivity may arise from changes in PNPase’s conformational or oligomeric state, RNA binding, or protonation of active site residues. Notably, its active site contains two aspartates, a lysine, and a histidine, critical for RNA interaction, Mg²⁺ coordination, and phosphate binding, which may influence the enzyme’s pH- dependent activity^50,51. In a similar manner, Stoffel et al. demonstrated in vitro that condensate pH shifts broaden enzyme functional range and enable otherwise-incompatible cascade reactions⁵². Our work, combined with discoveries of Stoffel et al., establishes condensate pH buffering as a physiologically important regulatory strategy that may be common within biomolecular condensates throughout biology.

Methods

This study involved only bacterial cultures and in vitro/vivo assays and did not include human participants or vertebrate animals. All experiments were conducted in accordance with institutional biosafety regulations at the University of Pittsburgh, Indiana University Bloomington, and New York University.

Imaging Caulobacter crescentus expressing RNase E-pHluorin2

Caulobacter crescentus strains expressing rne::rne-pHluorin2 or carrying dual broad-host-range plasmids for free pHluorin2 expression⁵³ were cultured in either M2G or PYE media. For mid-log phase studies and stationary phase measurements shown in Fig. 4B–D, cells were grown in M2G to OD₆₀₀ = 0.2. For all other stationary phase or ethanol stress experiments, cells were grown in PYE to OD₆₀₀ = 1.2 over 24 h.

Spent media was prepared by culturing Caulobacter in PYE to OD₆₀₀ = 1.2, leaving the culture undisturbed for 3 days, then pelleting the cells by centrifugation and collecting the supernatant. The supernatant was sterile filtered through a 0.22 µm membrane and supplemented with 0.2% glucose prior to use.

For experiments in spent media or 5% ethanol, starter cultures were first grown in PYE to OD₆₀₀ = 0.2–0.4, then back-diluted into the spent media or ethanol-supplemented media to OD₆₀₀ = 0.05 and grown for 24 h. For strains expressing free pHluorin2, ratiometric pH measurements were collected at OD₆₀₀ = 0.2 (mid-log) or OD₆₀₀ = 1.2 (stationary phase) in PYE.

A 1% agarose M2G imaging pad was prepared on a glass slide (VWR), and 2 µL of the cell suspension was placed on the pad, dried for 2 min., and covered with a glass coverslip (VWR). Cells were immediately imaged using a Nikon Eclipse Ti-E inverted microscope with a Plan-Apo-λ 100x/1.45 oil objective and 518 F immersion oil (Zeiss). Samples were excited at 379 nm (protonated) and 470 nm (deprotonated), with emission collected between 505–545 nm using an Andor Ixon Ultra 897 EMCCD camera. Ratiometric images were generated in ImageJ (National Institutes of Health, Bethesda, MD, USA) by dividing the 379 nm fluorescence by the 470 nm image. Phase contrast images were used to apply masks, and a yellow LUT was used for visualization. Dense BR-bodies and dilute regions were selected and measured using the Microbej⁵⁴ plugin.

For pH calibration, cells were treated for 25 min. at 30 °C with 10 mM methylamine, 10 mM potassium benzoate, and 50 mM buffer. For RNase E-pHluorin2, Tris-Cl (pH 6.2, 7.5) or sodium acetate/acetic acid (pH 4.8, 5.6) were used. For cells expressing free pHluorin2, Tris-Cl (pH 7.4, 8.1), MES (6.4) and sodium acetate/acetic acid (pH 5.6, 5.0) were used. Treated cells were placed on agarose pads buffered with 50 mM of the corresponding buffer. Imaging and processing were performed as described above.

To analyze pH distribution, a 2-pixel-wide line was drawn across representative cells in ImageJ, measuring fluorescence intensity and ratio values along the line. Data were plotted using the ImageJ measurement function.

Efficiency of plating assay

NA1000, JS819 (wild-type rne-pHluorin2), and JS825 (pI 7.03 rne-pHluorin2) were grown overnight at 28 °C in PYE liquid media. PYE cultures were then serially diluted into M2G and incubated at 28 °C overnight. For comparing log phase cultures, overnight M2G cultures that reached an OD ~ 0.3–0.6 were used for the efficiency of plating assay. For comparing stationary phase cultures, overnight M2G cultures incubated with shaking for three days after reaching an OD of 1.2 were used for the efficiency of plating assay. Optical density at 600 nm (OD₆₀₀) was measured in a cuvette using a UV-1600PC spectrophotometer. For efficiency of plating, cultures at each growth phase were diluted to an OD of 0.05, followed by five 1:10 serial dilutions. 5 μL from each dilution were then spotted on an M2G plate. Following cell spotting, plates were allowed to dry agar-side down for 30 min. before being incubated inverted at 28 °C for 1–3 days. Three biological replicates were tested for each strain in each condition. Plates were imaged using an iBright imaging system. Colonies were counted from the serially diluted plates, and the CFU/mL for each replicate was calculated using the formula: ((number of colonies counted/volume spotted)/total dilution factor). Finally, the averages of the CFU/mL were then plotted.

RT-qPCR mRNA half-live measurements

NA1000, JS819 (wild type rne-pHluorin2), and JS825 (pI 7 rne-pHluorin2) were grown overnight at 28 °C in PYE liquid media supplemented with spectinomycin. C. crescentus cultures grown in PYE were then serially diluted into liquid minimal M2G media supplemented with spectinomycin and incubated at 28 °C overnight until log phase (OD 0.3-0.6) was reached. Optical density at 600 nm (OD₆₀₀) was measured in a cuvette using a UV-1600PC spectrophotometer. M2G log phase cultures were then diluted to an OD of 0.05 into fresh 25 mL liquid M2G. For early stationary phase, cultures were grown to an OD ~ 1.2 and then incubated for an additional 7 hours prior to RNA sample collection. The following is the rifampicin transcription shut-off time-series RNA sample collection procedure described for both log and stationary phase cells: 25 mL cultures in the desired condition, 1 mL of cells was collected and added to 2 mL of RNAprotect Bacteria Reagent (Qiagent) and vortexed for 5 s prior to rifampicin addition. This is referred to as time point 0. Rifampicin was then immediately added to the log phase 25 mL cultures at a concentration of 200 μg/mL, and a timer was started. At each of the following time points (1, 2, 4, and 8 min.), 1 mL of culture was collected and added to 2 mL of RNAprotect Bacteria Reagent and vortexed for 5 s. Cells were incubated in the RNAprotect Bacteria reagent at room temperature for at least 5 minutes. Cells were then pelleted and resuspended in 1 mL of 65 °C pre-heated TRizol (Ambion) and incubated for 10 min. at 65 °C. 200 μL of chloroform was added to the TRIzol samples, and the tubes were inverted 6-7 times. The samples were incubated for 5 minutes at room temperature. Samples were then centrifuged at 14,000 rpm for 10 min at 4 °C. The top aqueous layer for each sample was pipetted into a new 1.5 mL Eppendorf tube. 2 μL of 1 mg/mL glycogen and 700 μL isopropanol were added to each tube. The samples were then vigorously vortexed and incubated at − 80 °C overnight. The tubes were then centrifuged for 60 minutes at 14,000 rpm at 4 °C. The RNA pellets were then washed with 80% ethanol and centrifuged for 10 minutes at 14,000 rpm at 4 °C. Following the washing step, the RNA pellets were allowed to air-dry and were resuspended in 50 μL elution buffer (10 mM Tris-HCl, pH = 7.0, and 0.1 mM EDTA in autoclaved water). Using the Luna Universal One-Step RT-qPCR, a RT-qPCR master mix was freshly prepared containing 1X Luna Universal One-Step Reaction Mix, 0.4 μM of forward and reverse primers, water, and 1X Luna WarmStart RT Enzyme Mix. Each RT-qPCR tube contained 1 μL of 100 ng/μL of RNA and 19 μL of the RT-qPCR master mix. The tubes were mixed well and quickly spun down. To conduct the RT-qPCR experiments, a QuantStudio Real-Time PCR machine was utilized. RT-qPCRs were done to study the half-lives of both 5S rRNAs and CtrA mRNAs. The forward and reverse primers used for the experiments are as follows: ctrA mRNA (CtrA F actgatgctgaagtctgaagg and CtrA R gattgaggtcgagcaggataag), and 5S rRNA (5S F ccattccgaactcggtcgttaag and 5S R tggcggcgacctactct). Following a standard curve, the quantity of RNA in each sample was determined from its Ct value. RNA abundance at each time point represents the average of three biological replicates, each measured with three technical replicates. To calculate the RNA half-life, at each time point, a linear curve was fitted to the ln (fraction of RNA remaining). The natural logarithm of the percentage of RNA remaining at each time point was normalized to the natural logarithm of RNA levels at time point 0. Linear regression was performed on the resulting values, and the slopes of the fitted lines were used to calculate mRNA half-lives according to the equation: mRNA half-life = −ln(2)/slope.

Super-resolution microscopy and FLIM of RNaseE-pHluorin2

Caulobacter crescentus cells expressing rne::rne-eYFP (wild type, pI6 and pI7 mutants) or rne::rne-pHluorin2 were grown overnight in M2G medium, diluted to OD₆₀₀ = 0.05, and incubated at 30 °C with shaking at 200 rpm. Once cultures reached OD₆₀₀ = 0.2–0.25, cells were imaged on a 1% agarose M2G pad.

Single-molecule localization microscopy was performed on an inverted microscope (Nikon Ti2) equipped with DIC optics, an oil immersion objective (Nikon PlanApo, 100x, 1.45 NA), and an sCMOS camera (Photometrics Prime 95B) with a system magnification of 0.11 µm/ pixel. A 514 nm continuous wave laser was used for exciting RNase E-eYFP in live Caulobacter cells. This method relies on spontaneous blinking of eYFP after initial photobleaching, which allows for active control of emissive states to obtain optimal single molecule densities in the cell⁵⁵. Accordingly, the sample was pre-bleached at high laser power (800 W/cm²) to get to a single molecule density, followed by imaging 1000 frames at 20 ms exposure time. Data was processed using the ThunderSTORM⁵⁶ plugin in Fiji⁵⁷. Resolution enhancement was quantified by comparing the areas of super-resolved clusters (0.001-0.029 µm²) against the theoretical diffraction-limited area ( ~ 0.040 µm²) calculated for our detection window (525–545 nm), confirming that all localized assemblies reside significantly below the diffraction limit.

To quantify the spatial organization of RNase E variants, Density-Based Spatial Clustering of Applications with Noise (DBSCAN)⁵⁸ was performed on the localization coordinates exported from ThunderSTORM. To ensure robust identification of condensates while minimizing noise, the clustering parameters were set to an ε of 20 nm that represented the maximum nearest-neighbor distance, with a minimum threshold of 20 localizations per cluster. To validate the fidelity of the clustering algorithm, the resulting DBSCAN scatter plots were cross-referenced against the original SMLM localization maps for every image. This dual-verification process ensured that identified clusters accurately corresponded to high-density localization events and that no significant molecular assemblies were omitted or artificially merged during the sampling process.

Fluorescence lifetime imaging was performed on a Nikon Eclipse Ti2 microscope equipped with a confocal scanning system (Abberior). A 100x, NA 1.45 oil-immersion objective and a 1 Airy-unit pinhole was used. The confocal detectors were connected to a Time-Correlated Single-Photon Counting (TCSPC) module (Becker & Hickl). For the measurement of pHluorin2 lifetime, 488 nm excitation pulsed diode laser (40 MHz) at 5.15% power; emission at 505–550 nm was employed. All FLIM data were acquired in line-scanning mode (pixel dwell time 10 µs) using SPCM Data Acquisition software (v9.89; Becker & Hickl), with 90 s acquisition per frame. Images were segmented by pixel intensity and analyzed for decay curves and phasor plots in SPCImage NG Data Analysis software (v8.87.0; Becker & Hickl). For FLIM of Caulobacter crescentus cells expressing RNaseE–pHluorin2 (wild type, pI6 and pI7 mutants), cells were grown in PYE to stationary phase (OD₆₀₀ = 1.35), spotted on 1% M2G agar pads, and washed twice in M2G before imaging. Wild-type and pI6 cells yielded > 150 photons per image scan, providing > 100 photons for the brightest pixel in the exponential fit, whereas pI7 cells produced ~ 40 photons initially and < 30 photons at the corresponding pixel over 6 min. measurement, which was insufficient for reliable lifetime estimation. The observed lifetime variations within individual BR-bodies (1–2 ns) significantly exceed the intrinsic variance of the pHluorin2 sensor measured in calibrated buffers (0.3–0.4 ns), and the stress-dependent shifts in these distributions further indicate that the intra-cluster heterogeneity is biologically regulated rather than an instrumental artifact.

FLIM Cluster analysis

Image data was first processed through a binary mask, which utilized lifetime statistics to identify spatial heterogeneity and shape. Quantitative feature extraction was applied utilizing ‘RegionProps’ module from ‘skimage.measure’ in Python. Over 15 features were extracted, but they were selectively eliminated after data visualization via pair plots and a heatmap revealed collinearity. The goal of this pre-processing was to turn image data into a tabular CSV file with distinct features for further statistical analysis. The final five features to determine heterogeneity were lifetime standard deviation, sphericity, centroid shift, number of strong patches, and entropy per pixel.

After pre-processing, the CSV data was then processed through a cosine similarity program. The cosine similarity results were then utilized as the basis of a hierarchical clustering model. This model grouped all data points on feature similarity. Principle Component Analysis (PCA) was then used to reduce the dimensionality of our data to 2D for cleaner visualization. Hierarchical clustering data was then overlaid on the PCA graph utilizing a hue. Once groups were established with hierarchical clustering and PCA, F values and P values were extracted via Python and used for further analysis regarding which traits contributed most to heterogeneity and cluster groupings.

RNase E CTD, RNase E CTD-pHluorin2 Expression and Purification

RNase E(451-898) (RNase E CTD) and RNase E CTD-pHluorin2, both with N-terminal His₆-MBP tags, were expressed in BL21(DE3) cells transformed with pTEV5-RNase E CTD or pTEV6-RNase E CTD-pHluorin2. Cultures were grown to OD₆₀₀ = ~ 0.5, induced with 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG), and incubated for 3 h at 37 °C. Cells were harvested (4 °C, 4000 × g, 10 min), resuspended in HEPES pH 8.0, 500 mM NaCl, pelleted again, and stored at − 80 °C.

Thawed cells were resuspended in 100 mL lysis buffer (500 mM NaCl, 20 mM Tris-Cl, pH 7.4, 20 mM imidazole, pH 7.0, 1 mM DTT, 0.1% Triton X-100) supplemented with 20 U Pierce™ Universal Nuclease (Thermofisher™), SIGMAFAST protease inhibitor (MilliporeSigma), and 0.1 mg/mL lysozyme. Lysis was performed via an Avestin Emulsiflex-C3 at 15,000 psi for 30 min. Debris was removed by centrifugation (4 °C, 20,000 × g, 45 min).

The supernatant was purified using a 5 mL HisTrap™ FF column (Cytiva) on a Cytiva Äkta™ pure 25. After equilibration and 10 column volume washes, proteins were eluted with 20 mM Tris-Cl, pH 7.4, 200 mM NaCl, 200 mM imidazole, 1 mM DTT. The eluted protein was concentrated to ~ 20 mg/mL using a 50 kDa MWCO Sartorius Vivaspin® concentrator, then incubated with TEV protease (1 mg per 20 mg protein, 2 h at RT). Cleaved protein was separated via a HiPrep Sephacryl S-200 HR size exclusion column (Cytiva) equilibrated in 20 mM Tris-Cl pH 7.5, 200 mM NaCl, 1 mM DTT. Final protein fractions ( ~ 8 mg/mL) were concentrated (30 kDa MWCO Sartorius Vivaspin®), aliquoted, and stored at − 80 °C.

Imaging RNase E with C-SNARF-4F

RNase E CTD was diluted to 20 µM in a buffer containing 20 mM Tris-Cl (pH 7.5), 100 mM NaCl, 10% PEG-8000, and 10 µM C-SNARF-4F (ThermoFisher) at room temperature. Components were added sequentially: water, Tris-Cl, NaCl, RNase E, C-SNARF-4F, and PEG-8000. The solution was pipetted into a 1 mm adhesive spacer well (Electron Microscopy Sciences) on a microscope slide (VWR) and sealed with a glass coverslip (VWR). Samples were inverted and incubated for 20 min. at room temperature, then imaged using the same microscope setup as live-cell fluorescence measurements. Excitation was at 470 nm, with emission collected at 579–631 nm (protonated state) and 669–741 nm (deprotonated state) to generate ratiometric images. Image analysis in ImageJ involved background subtraction (rolling ball radius = 50) followed by image division.

For standard curve generation, PEG-8000 was omitted, and buffers at defined pH values were used: Tris-Cl (pH 6.6, 7.1, 7.7, 8.1) or sodium acetate/acetic acid (pH 3.5, 4.0, 5.0, 5.5, 5.9). Images were acquired under identical conditions and fit using nonlinear regression of the form:

R a t i o = (R a t i o_{o} - P l a t e a u) e^{- k (p H)} + P l a t e a u

Imaging RNase E-pHluorin2 with Tris-Cl, MgCl₂, and NaCl titrations

For the Tris-Cl titration, RNase E CTD-pHluorin2 (20 µM) was diluted into buffers containing 20, 50, 100, or 500 mM Tris-Cl (pH 7.5), 100 mM NaCl, and 10% PEG-8000. For the MgCl₂ titration, RNase E CTD-pHluorin2 (20 µM) was diluted into a buffer containing 20 mM Tris-Cl (pH 7.5), 100 mM NaCl, and 0, 20, 50, 75, 100, 200, or 500 mM MgCl₂, with 10% PEG-8000. For the NaCl titration, RNase E-pHluorin2 (20 µM) was diluted into a buffer containing 20 mM Tris-Cl (pH 7.5), 100, 250, or 500 mM NaCl, and 10% PEG-8000.

All samples were prepared and imaged using the same slides, coverslips, spacers, incubation times, and imaging system as with C-SNARF-4F. Excitation was performed at 379 nm and 470 nm, with emission collected between 505–545 nm for ratiometric image generation.

Imaging RNase E-pHluorin2 with clients

RNase E CTD-pHluorin2 (20 µM) was diluted in a buffer containing 20 mM Tris-Cl (pH 7.5), 100 mM NaCl, either 5 µM PNPase, 5 µM aconitase, or 80 ng/µL poly(A) (or combined), and 10% PEG-8000. Components were added in the following order: water, Tris-Cl, NaCl, RNase E-pHluorin2, client(s), and PEG-8000. All solutions were imaged using the same slides, coverslips, spacers, incubation times, and imaging system as with C-SNARF-4F and the RNase E CTD-pHluorin2 titration experiments.

Samples were excited at 379 nm, corresponding to the protonated state, and 470 nm, corresponding to the deprotonated state. Emission was collected between 505–545 nm for ratiometric image generation. The final ratiometric images were created by dividing the fluorescence image captured at 379 nm by the image obtained at 470 nm (after background subtraction with a rolling ball radius of 50) using ImageJ software. A custom red-to-blue LUT was applied in ImageJ to visualize the resulting ratios.

Zeta potential measurements

Zeta potential measurements were collected on an Anton Paar Litesizer 500, at 25 °C, with solvent refractive index settings set to 1.3318 for concentrated salt solution. The protein samples were loaded into a clean, dust-free, Omega Cuvette Mat. No. 225288, and 600 μL of the sample was added. After a 2 min. equilibration time, a total of 60 scans were run, and the Smoluchowski approximation was used to analyze the data.

PNPase activity assay

PNPase activity measurements were conducted using a plate reader (Tecan Infinite M1000 Pro). Samples were prepared by diluting PNPase (2.5 µM) in a buffer containing 20 mM Tris-Cl (for pH 6.2 and 7.5) or 20 mM sodium acetate/acetic acid (for pH 3.5, 4.6, 4.9, 5.3, and 5.5), 100 mM NaCl, 20 ng/µL poly(A), 10 µM Thioflavin T, 500 µM MgCl₂, and 4 mM phosphate. Components were added in the following order: water, buffer, NaCl, PNPase, poly(A), MgCl₂, Thioflavin T, and phosphate.

For standard curves, each pH condition included poly(A) concentrations of 20, 15, 10, 5, and 0 ng/µL, with all other components unchanged except for the exclusion of MgCl₂. Samples (70 µL) were loaded into a 384-well plate, and fluorescence intensity was measured with an excitation wavelength of 450 nm (5 nm bandwidth) and emission at 485 nm (5 nm bandwidth). Each sample received eight readings per timepoint, measured sequentially. Standard curve fits were either exponential or linear, depending upon comparisons between the models.

The plate reader temperature was maintained at 26–27.5 °C. For activity measurements including MgCl₂, all components except phosphate were added to the wells, and phosphate was mixed for ~ 8 s before initiating sequential measurements.

Calibration curve for in vivo Caulobacter crescentus pH measurements

The C. crescentus strain containing the rne::rne-pHluorin2 construct was grown in M2G media to OD₆₀₀ = 0.3. Cells were pelleted by centrifugation at 5000 × g for 10 min. at room temperature, and the clarified supernatant was transferred to a fresh tube. This supernatant was used to prepare calibration media by supplementing it with 10 mM monomethylamine and 10 mM potassium benzoate. To establish specific pH conditions, 50 mM Tris-Cl was added for pH 6.2 and 7.5, while 50 mM sodium acetate/acetic acid was used for pH 4.8 and 5.6.

Cells were resuspended in an equal volume of calibration media to maintain the original OD₆₀₀ and incubated at 30 °C for 15 min. in the corresponding pH buffer. Each condition was prepared separately to ensure consistent timing.

Simultaneously, 1% agarose imaging pads were prepared, supplemented with the same 50 mM buffer used for each pH condition. Following incubation, 2 µL of the cell suspension was applied to the imaging pad, dried, and imaged.

Samples were excited at 379 nm and 470 nm, with emission collected between 505–545 nm to generate ratiometric images. The fluorescence image obtained at 379 nm was divided by that at 470 nm. Regions of cells without apparent BR-bodies were analyzed for fluorescence ratios, and these data were used to construct the calibration curve, which was fitted to the following equation:

R a t i o = (R a t i o_{o} - P l a t e a u) e^{- k (p H)} + P l a t e a u

Construction of a false-positive probability color map

A two-dimensional false-positive probability map was constructed to evaluate the statistical likelihood that a fluorescence ratio-derived pH measurement of a given pixel cluster arises from background noise rather than a true acidic compartment. The map encodes the joint probability $p^{n}$ as a function of two variables: the inferred pH threshold (x-axis, 4.8–7.9 in 0.1-unit steps) and the cluster size in pixels (y-axis, 1–20 pixels).

Intracellular pixel pH was calculated from the ratios measured in Caulobacter crescentus cells and using the empirically produced calibration curve as described above. To establish the null distribution, fluorescence ratios were measured from pixels classified as diffuse (non-punctate) signal within bacterial cells imaged under experimental conditions as described in the imaging methods of Caulobacter crescentus in previous sections. This diffuse population represents the expected background ratio distribution in the absence of a true acidic compartment. The distribution was well described by a Gaussian with mean $μ = 1.3944$ ratio units and standard deviation $σ = 0.11$ ratio units. These parameters define the null model against which all pixel-level measurements are compared.

For any given pH threshold value, the corresponding fluorescence ratio was first computed using the calibration curve for in vivo scenarios using RNase E-pHluorin2 chimeras expressed endogenously. The single-pixel false-positive probability $p$ was then defined as the probability that a pixel drawn from the background (diffuse) distribution would produce a ratio at or below the ratio corresponding to that pH:

p (pH) = Φ (\frac{R (pH) - μ}{σ})

where $Φ$ is the standard normal cumulative distribution function. This quantity represents the probability that a single background pixel would be misclassified as belonging to a compartment with pH at or below the specified threshold.

Under the assumption that individual pixels within a candidate cluster are independent draws from the background distribution, the joint false-positive probability for a cluster of $n$ pixels all falling at or below a given pH threshold is:

P_{FP} (n, pH) = p {(pH)}^{n}

This expression gives a conservative upper bound: it represents the probability of observing a cluster of $n$ pixels whose ratios are all consistent with or more extreme than the stated pH threshold, given only background signal. If individual pixels within the cluster have ratios that are substantially lower than the threshold (i.e., more acidic than the cap value), the true joint probability is often much smaller.

The joint probability $p^{n}$ was computed across a grid of pH values from 4.8 to 7.9 (step 0.1) and cluster sizes from 1 to 20 pixels, yielding a 20 × 32 matrix. Values were log-transformed ( $\log_{10}$ ) to span the full dynamic range from $10^{- 20}$ to $1.0$ .

The resulting matrix was visualized as a heatmap using Plotly (Python). A diverging perceptual colorscale was applied across the log-transformed values, ranging from deep purple (near-zero probability, highly significant) through green and yellow to red (probabilities near 1.0, indistinguishable from background). Color stops were placed at $\log_{10} (p^{n})$ values of 0, − 1, − 2, − 3, − 5, − 10, − 15, and − 20 to highlight the region of practical significance.

For each pH column, a gray triangle annotation was placed at the minimum cluster size for which $p^{n} < 0.001$ , marking the threshold below which a cluster is considered statistically unlikely to arise from background signal noise alone.

All calculations were performed in Python 3 using NumPy and SciPy (stats.norm.cdf). Visualization was produced with Plotly.

Electrostatic visualization of PNPase structure

The three-dimensional structure of PNPase was obtained from the Protein Data Bank (PDB). Solvent molecules, including crystallographic water, were removed using PyMOL (Schrödinger, LLC). To determine the protonation states of titratable residues (histidine, aspartic acid, glutamic acid, lysine, and arginine) at the experimental pH, the protein structure was processed using pdb2pqr⁵⁹, which assigned appropriate protonation states based on the PROPKA algorithm. The processed structure was saved in PQR format for further electrostatic calculations.

The electrostatic potential of the protein surface was computed using the Adaptive Poisson-Boltzmann Solver (APBS)⁶⁰. The input PQR file was used to generate the electrostatic potential map, with the following parameters: a dielectric constant of 2 for the protein interior and 78.5 for the solvent, a solvent radius of 1.4 Å, and a grid spacing optimized to balance resolution and computational efficiency. The Poisson-Boltzmann equation was solved under physiological ionic conditions, and the resulting electrostatic potential was saved for visualization.

PyMOL was used to visualize the electrostatic potential on the protein surface. The PDB structure was loaded, followed by the DX file generated by APBS. The “APBS Electrostatics” plugin within PyMOL was utilized to map the electrostatic potential onto the molecular surface, applying a color gradient from red (negative charge) to blue (positive charge). Visualization settings, including surface transparency, color ramp adjustments, and contouring, were optimized to enhance the clarity of electrostatic distributions.

RNase E homolog and client charge

Homologs were identified by BLAST-searching the C. crescentus RNase E (RNE) protein sequence against the National Center for Biotechnology Information (NCBI) database^61,62. Intrinsically disordered regions within the protein sequences were predicted using Metapredict, while amino acid composition was analyzed using ProtParam^63,64. BR-body client proteins were identified by Nandana et al.³¹. Client protein sequences and amino acid composition were processed identically to the homologs.

Designing of synthetic RNase E pI6, 7, and 10

The synthetic RNase E constructs with isoelectric points of ~ 6, ~ 7, ~ 10 were generated by modifying the Caulobacter crescentus RNase E C-terminal domain (CTD). Aspartic acid (D) and glutamic acid (E) residues within three key negatively charged regions were substituted with neutral residues.

pI6

D570T, D571S, E572T, E573S, E574T, E575S, D576T, D577S, D581T, D582S, E583T,D584T,D585S, E589T, D590S, E634T, E636T, E638S, D646T, D648S, D654S, D829S, E833S, E837T, E842S, E845S, E847T, E850S, E859S, E863T, E869S, E872T, E873S, E878T, D880S

pI7

D570T, D571S, E572T, E573S, E574T, E575S, D576T, D577S, D581T, D582S, E583T,D584T,D585S, E589T, D590S, D592S, D593T, D594S, E634T, E636T, E638S, D646T, D648S, D654S, D829S, E833S, E837T, E842S, E845S, E847T, E850S, E859S, E863T, E869S, E872T, E873S, E878T, D880S

pI10

E559S, E560T, D561S, D562T, D563S, D564T, D570S, D571T, E572S, E573T, E574S, E575T, D576S, D577T, D581S, D582T, E583S, D584T, D585S, E589T, D590S, D592S, D593T, D594S, D595T, E632S, E634T, E636T, E638S, D646T, D648S, D654S, D655T, E826T, D829S, E833S, E837T, E842S, E845S, E847T, E850S, E859S, E863T, E869S, E872T, E873S, E878S, D880T, E883S. These substitutions were designed to reduce the net negative charge while preserving known client-binding motifs: Aconitase at 675-725, PNPase at 889-898 and RNase D at 760-803.

Microscopy analysis of PNPase activity

For poly(A) degradation, RNase E CTD (20 µM) was diluted in a buffer containing 20 mM or 100 mM Tris-Cl (pH 7.5), 100 mM NaCl, 500 µM MgCl₂, and 10% PEG-8000. Additional components included PNPase (2.5 µM), poly(A) (20 ng/µL), Thioflavin T (10 µM), and phosphate (4 mM). Components were added sequentially in the following order: water, Tris-Cl, NaCl, RNase E, PNPase, poly(A), MgCl₂, Thioflavin T, PEG-8000, and phosphate.

Since PNPase requires phosphate for activity, phosphate was omitted until just before imaging. Samples without phosphate were deposited on a 22 × 60 mm glass coverslip (VWR), incubated for 10 min in a humidity chamber containing a moist paper towel to minimize evaporation, and then placed on the microscope stage. Phosphate was added in situ, and images were acquired every 10 s with 440 nm excitation, 480 nm emission, and a 50 ms exposure time.

For poly(A) polymerization, RNase E CTD (20 µM) was diluted in a buffer containing 20 mM or 100 mM Tris-Cl (pH 7.5), 100 mM NaCl, 500 µM MgCl₂, and 10% PEG-8000. Additional components included PNPase (2.5 µM), ADP (50 µM), and Thioflavin T (10 µM). Components were added sequentially in the following order: water, Tris-Cl, NaCl, RNase E, PNPase, ADP, Thioflavin T, PEG-8000, and MgCl₂.

Since PNPase also requires MgCl₂ for activity, MgCl₂ was omitted until just before imaging. Samples without MgCl₂ were deposited on a 22 × 60 mm glass coverslip (VWR), incubated for 10 minutes in a humidity chamber containing a moist paper towel to minimize evaporation, and then placed on the microscope stage. MgCl₂ was added in situ, and images were acquired every 10 s with 440 nm excitation, 480 nm emission, and a 50 ms exposure time.

Relative rate images were generated by selecting frames 30 s apart for degradation and 10 seconds apart for polymerization. Pixel intensity differences between frames were divided by the respective time intervals to compute pixel-specific mean intensity rates of change. These rates were then normalized by dividing the mean background intensity rate of change for each pixel. Finally, the relative rates of multiple condensates were averaged to enable quantitative comparisons between low and high Tris-Cl conditions.

PNPase-mCherry and poly(A)-Cy5 partitioning into RNase E CTD condensates

RNase E CTD (20 µM) was diluted in a buffer containing 20 mM Tris-Cl (pH 7.5), 100 mM NaCl, 5 µM PNPase-mCherry or 20 ng/µL poly(A)-Cy5, and 10% PEG-8000. Components were added in the following order: water, Tris-Cl, NaCl, RNase E CTD, PNPase-mCherry or poly(A)-Cy5, and PEG-8000. The solution was pipetted into a 1 mm adhesive spacer well (Electron Microscopy Sciences) on a microscope slide (VWR) and sealed with a glass coverslip (VWR). Samples were inverted and incubated for 20 min at room temperature, then imaged using the same microscope setup as live-cell fluorescence measurements. PNPase-mCherry and poly(A)-Cy5 were imaged using excitation/emission wavelengths of 549/650 nm and 640/705 nm, respectively. Condensates were segmented using MicrobeJ. Partitioning coefficients were calculated as the mean fluorescence intensity within condensates divided by the mean fluorescence intensity of the surrounding dilute phase. Three independent biological replicates were performed for each condition, and statistical significance was assessed using an unpaired two-sample t test to compare the high and low Tris-Cl conditions for PNPase-mCherry and poly(A)-Cy5.

Statistics and reproducibility

No statistical method was used to predetermine sample size, and sample sizes were chosen based on prior experience with similar bacterial experiments and practical constraints on culture growth and imaging. All available data that met predefined quality control criteria (e.g., successful growth, correct genotype, and usable images without acquisition artifacts) were included in the analyses. No data were excluded from the analyses.

The number of replicates (n), the type of replicates (biological vs. technical), the statistical tests used, and whether tests were one- or two-sided are reported in the figure legends.

Quantitative analyses (e.g., intensity measurements, condensate properties, growth or stress-response assays) were performed in GraphPad Prism or Python 3. Parametric tests (unpaired and Welch-corrected t-tests, one-way ANOVA, Welch’s ANOVA, and Brown–Forsythe tests) were used when data met the relevant assumptions. When homogeneity of variance was violated, variance-robust procedures (e.g., Welch’s ANOVA or Brown–Forsythe) and appropriate post hoc tests (e.g., Games–Howell, Tukey’s test) were applied. For extremely small p-values that Prism reported only as thresholds (e.g., “P < 0.001”), exact p-values were obtained by evaluating the corresponding test distributions (t, F, or chi-square) numerically. Reproducibility was assessed by repeating key experiments on independent days and/or with independent bacterial cultures.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information^{(1.1MB, pdf)}

Transparent Peer Review file^{(762.7KB, pdf)}

Reporting Summary^{(87.7KB, pdf)}

Source data

Source Data^{(380.9KB, zip)}

Author contributions

W.E.S. and W.S.C. conceived the original hypothesis and designed the overall study.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Funding

This work was supported by the National Institute of Health grants (R01GM136863 to W.S.C., R35GM124733 to J.M.S. and 1R35GM157103 to S.S). Wade Schnorr and Kathryn Dzurik were supported by the US Department of Education GAANN grant, award number: P200A240158.

Data availability

The source data generated in this study are provided in the Source Data file. The imaging data used in this study are available in the Figshare database at [10.6084/m9.figshare.32095177]. Source data are provided in this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Jared M. Schrader, Email: jaschrad@iu.edu

Saumya Saurabh, Email: saumya@nyu.edu.

W. Seth Childers, Email: wschild@pitt.edu.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-76859-9.

References

1.Azaldegui, C. A., Vecchiarelli, A. G. & Biteen, J. S. The emergence of phase separation as an organizing principle in bacteria. Biophys. J.120, 1123–1138 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Banani, S. F., Lee, H. O., Hyman, A. A. & Rosen, M. K. Biomolecular condensates: organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol.18, 285–298 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Fare, C. M., Villani, A., Drake, L. E. & Shorter, J. Higher-order organization of biomolecular condensates. Open Biol.11, 210137 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Al-Husini, N. et al. BR-Bodies provide selectively permeable condensates that stimulate mRNA decay and prevent release of decay intermediates. Mol. Cell78, 670–682.e8 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Nadra, A.-H., Tomares, D. T., Bitar, O., Childers, W. S. & Schrader, J. M. -proteobacterial RNA degradosomes assemble liquid-liquid phase separated RNP bodies. Mol. Cell71, 1027–1039.e14 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Ladouceur, A.-M. et al. Clusters of bacterial RNA polymerase are biomolecular condensates that assemble through liquid–liquid phase separation. Proc. Natl. Acad. Sci. USA. 117, 18540–18549 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Heinkel, F. et al. Phase separation and clustering of an ABC transporter in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA116, 16326–16331 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Harami, G. M. et al. Phase separation by ssDNA binding protein controlled via protein−protein and protein−DNA interactions. Proc. Natl. Acad. Sci. USA117, 26206–26217 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Robles-Ramos, M. Á et al. Assembly of bacterial cell division protein FtsZ into dynamic biomolecular condensates. Biochim. Biophys. Acta Mol. Cell Res.1868, 118986 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Schnorr, W. E. & Childers, W. S. Biomolecular condensates: from bacterial compartments to incubator spaces of emergent chemical systems in matter-to-life transitions. ChemSystemsChem6, e202400011 (2024). [Google Scholar]
11.Sasazawa, M., Tomares, D. T., Childers, W. S. & Saurabh, S. Biomolecular condensates as stress sensors and modulators of bacterial signaling. PLoS Pathog.20, e1012413 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Passos, C. et al. BR-Bodies facilitate adaptive responses and survival during copper stress in Caulobacter crescentus. J. Biol. Chem. 301, 110648 (2025). [DOI] [PMC free article] [PubMed]
13.Lu, N. et al. Scaffold-scaffold interaction facilitates cell polarity development in Caulobacter crescentus. mBio14, e0321822 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Ahmed, Y. M., Brown, L. M., Varga, K. & Bowman, G. R. Phospho-signaling couples polar asymmetry and proteolysis within a membraneless microdomain in Caulobacter crescentus. Nat. Commun.15, 9282 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Lasker, K. et al. Selective sequestration of signalling proteins in a membraneless organelle reinforces the spatial regulation of asymmetry in Caulobacter crescentus. Nat. Microbiol.5, 418–429 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Connolley, L., Schnabel, L., Thanbichler, M. & Murray, S. M. Partition complex structure can arise from sliding and bridging of ParB dimers. Nat. Commun.14, 4567 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Zhao, Y. et al. Phase-separated ParB enforces diverse DNA compaction modes and stabilizes the parS-centered partition complex. Nucleic Acids Res.52, 8385–8398 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.MacCready, J. S., Tran, L., Basalla, J. L., Hakim, P. & Vecchiarelli, A. G. The McdAB system positions α-carboxysomes in proteobacteria. Mol. Microbiol.116, 277–297 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Choi, S., Knoerdel, A. R., Sing, C. E. & Keating, C. D. Effect of polypeptide complex coacervate microenvironment on protonation of a guest molecule. J. Phys. Chem. B. 127, 5978–5991 (2023). [DOI] [PubMed] [Google Scholar]
20.King, M. R. et al. Macromolecular condensation organizes nucleolar sub-phases to set up a pH gradient. Cell 187, 1889–1906 (2024). [DOI] [PMC free article] [PubMed]
21.Dai, Y. et al. Biomolecular condensates regulate cellular electrochemical equilibria. Cell187, 5951–5966 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Ausserwöger, H. et al. Biomolecular condensates sustain pH gradients at equilibrium through charge neutralization. Nat. Chem.18, 246–257 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Dai, Y. et al. Interface of biomolecular condensates modulates redox reactions. Chem9, 1594–1609 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Testa, A. et al. Sustained enzymatic activity and flow in crowded protein droplets. Nat. Commun.12, 6293 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Joshi, A. et al. Hydrogen-bonded network of water in phase-separated biomolecular condensates. J. Phys. Chem. Lett.15, 7724–7734 (2024). [DOI] [PubMed] [Google Scholar]
26.Mukherjee, S. & Schäfer, L. V. Thermodynamic forces from protein and water govern condensate formation of an intrinsically disordered protein domain. Nat. Commun.14, 5892 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Bayas, C. A. et al. Spatial organization and dynamics of RNase E and ribosomes in Caulobacter crescentus. Proc. Natl. Acad. Sci. USA115, E3712–E3721 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Gahlmann, A. & Moerner, W. E. Exploring bacterial cell biology with single-molecule tracking and super-resolution imaging. Nat. Rev. Microbiol.12, 9–22 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Joron, K. et al. Fluorescent protein lifetimes report densities and phases of nuclear condensates during embryonic stem-cell differentiation. Nat. Commun.14, 4885 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Nandana, V. et al. The BR-body proteome contains a complex network of protein-protein and protein-RNA interactions. Cell Rep. 42, 10.1016/j.celrep.2023.113229 (2023). [DOI] [PMC free article] [PubMed]
31.Farag, M. et al. Condensates formed by prion-like low-complexity domains have small-world network structures and interfaces defined by expanded conformations. Nat. Commun.13, 7722 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Wu, T. et al. Single-fluorogen imaging reveals distinct environmental and structural features of biomolecular condensates. Nat. Phys.21, 778–786 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.von Hofe, J. et al. Multivalency controls the growth and dynamics of a biomolecular condensate. J. Am. Chem. Soc.147, 25242–25253 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Sumrall, E. R., Husna, S. M. A., Halder, B., Ahnoud, A. & Walter, N. G. Single-molecule microscopy of biomolecular condensates: From liquid droplets to heterogeneous networks. Curr. Opin. Struct. Biol.98, 103271 (2026). [DOI] [PubMed] [Google Scholar]
35.Chattaraj, A. & Shakhnovich, E. I. Multi-condensate state as a functional strategy to optimize the cell signaling output. Nat. Commun.15, 6268 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Ortiz-Rodríguez, L. A. et al. Stress changes the material state of a bacterial biomolecular condensate and shifts its function from mRNA decay to storage. Nat. Commun.16, 10019 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Munder, M. C. et al. A pH-driven transition of the cytoplasm from a fluid- to a solid-like state promotes entry into dormancy. ELife5, e09347 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Joyner, R. P. et al. A glucose-starvation response regulates the diffusion of macromolecules. ELife.https://elifesciences.org/articles/9376 (2016). [DOI] [PMC free article] [PubMed]
39.Currie, S. L. et al. Quantitative reconstitution of yeast RNA processing bodies. Proc. Natl. Acad. Sci. USA120, e2214064120 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Gonçalves, D., Santiago, A. & Morano, K. A. When pH comes to the rescue. ELife9, e62022 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Goetz, S. K. et al. Structural reorganization underlying stress-induced cytoplasmic solidification in yeast. Preprint at 10.64898/2025.12.22.696049 (2025). [DOI]
42.Heinrich, K., Sobetzko, P. & Jonas, K. A kinase-phosphatase switch transduces environmental information into a bacterial cell cycle circuit. PLOS Genet.12, e1006522 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Kesler, V., Murmann, B. & Soh, H. T. Going beyond the debye length: overcoming charge screening limitations in next-generation bioelectronic sensors. ACS Nano14, 16194–16201 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Dudev, T. & Lim, C. Competition among Ca2+, Mg2+, and Na+ for model ion channel selectivity filters: determinants of ion selectivity. J. Phys. Chem. B116, 10703–10714 (2012). [DOI] [PubMed] [Google Scholar]
45.Friesen, S., Kruchinin, S. E., Fedotova, M. V. & Buchner, R. Cation-binding of glutamate in aqueous solution. J. Phys. Chem. B. 128, 5746–5755 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Verma, S., Ravichandiran, V. & Ranjan, N. Beyond amyloid proteins: Thioflavin T in nucleic acid recognition. Biochimie190, 111–123 (2021). [DOI] [PubMed] [Google Scholar]
47.Islam, M. D. et al. Impact of pseudouridylation, substrate fold, and degradosome organization on the endonuclease activity of RNase E. RNA27, 1339–1352 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Parry, B. R. et al. The bacterial cytoplasm has glass-like properties and is fluidized by metabolic activity. Cell156, 183–194 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Joyner, R. P. et al. A glucose-starvation response regulates the diffusion of macromolecules. ELife5, e09376 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Nurmohamed, S. et al. Polynucleotide phosphorylase activity may be modulated by metabolites in Escherichia coli. J. Biol. Chem.286, 14315–14323 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Symmons, M. F., Jones, G. H. & Luisi, B. F. A duplicated fold is the structural basis for polynucleotide phosphorylase catalytic activity, processivity, and regulation. Structure8, 1215–1226 (2000). [DOI] [PubMed] [Google Scholar]
52.Stoffel, F. et al. Enhancement of enzymatic activity by biomolecular condensates through pH buffering. Nat. Commun.16, 6368 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Greenwich, J. L., Alakavuklar, M. A. & Fuqua, C. An Inducible T7 polymerase system for high-level protein expression in diverse gram-negative bacteria. Microbiol. Resour. Announc.12, e01119–e01122 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Ducret, A., Quardokus, E. M. & Brun, Y. V. MicrobeJ, a tool for high throughput bacterial cell detection and quantitative analysis. Nat. Microbiol.1, 1–7 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Saurabh, S., Perez, A. M., Comerci, C. J., Shapiro, L. & Moerner, W. E. Super-resolution imaging of live bacteria cells using a genetically directed, highly photostable fluoromodule. J. Am. Chem. Soc.138, 10398–10401 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Ovesný, M., Křížek, P., Borkovec, J., Švindrych & Hagen, Z. G. M. ThunderSTORM: a comprehensive ImageJ plug-in for PALM and STORM data analysis and super-resolution imaging. Bioinformatics30, 2389–2390 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Hahsler, M., Piekenbrock, M. & Doran, D. dbscan: Fast density-based clustering with R. J. Stat. Softw.91, 1–30 (2019). [Google Scholar]
59.Dolinsky, T. J., Nielsen, J. E., McCammon, J. A. & Baker, N. A. PDB2PQR: an automated pipeline for the setup of Poisson–Boltzmann electrostatics calculations. Nucleic Acids Res.32, W665–W667 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Baker, N. A., Sept, D., Joseph, S., Holst, M. J. & McCammon, J. A. Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA98, 10037–10041 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol.215, 403–410 (1990). [DOI] [PubMed] [Google Scholar]
62.Sayers, E. W. et al. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res.50, D20–D26 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Emenecker, R. J., Griffith, D. & Holehouse, A. S. Metapredict: a fast, accurate, and easy-to-use predictor of consensus disorder and structure. Biophys. J.120, 4312–4319 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Wilkins, M. R. et al. Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol.112, 531–552 (1999). [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Information^{(1.1MB, pdf)}

Transparent Peer Review file^{(762.7KB, pdf)}

Reporting Summary^{(87.7KB, pdf)}

Source Data^{(380.9KB, zip)}

Data Availability Statement

[CR1] 1.Azaldegui, C. A., Vecchiarelli, A. G. & Biteen, J. S. The emergence of phase separation as an organizing principle in bacteria. Biophys. J.120, 1123–1138 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Banani, S. F., Lee, H. O., Hyman, A. A. & Rosen, M. K. Biomolecular condensates: organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol.18, 285–298 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Fare, C. M., Villani, A., Drake, L. E. & Shorter, J. Higher-order organization of biomolecular condensates. Open Biol.11, 210137 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Al-Husini, N. et al. BR-Bodies provide selectively permeable condensates that stimulate mRNA decay and prevent release of decay intermediates. Mol. Cell78, 670–682.e8 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Nadra, A.-H., Tomares, D. T., Bitar, O., Childers, W. S. & Schrader, J. M. -proteobacterial RNA degradosomes assemble liquid-liquid phase separated RNP bodies. Mol. Cell71, 1027–1039.e14 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Ladouceur, A.-M. et al. Clusters of bacterial RNA polymerase are biomolecular condensates that assemble through liquid–liquid phase separation. Proc. Natl. Acad. Sci. USA. 117, 18540–18549 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Heinkel, F. et al. Phase separation and clustering of an ABC transporter in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA116, 16326–16331 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Harami, G. M. et al. Phase separation by ssDNA binding protein controlled via protein−protein and protein−DNA interactions. Proc. Natl. Acad. Sci. USA117, 26206–26217 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Robles-Ramos, M. Á et al. Assembly of bacterial cell division protein FtsZ into dynamic biomolecular condensates. Biochim. Biophys. Acta Mol. Cell Res.1868, 118986 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Schnorr, W. E. & Childers, W. S. Biomolecular condensates: from bacterial compartments to incubator spaces of emergent chemical systems in matter-to-life transitions. ChemSystemsChem6, e202400011 (2024). [Google Scholar]

[CR11] 11.Sasazawa, M., Tomares, D. T., Childers, W. S. & Saurabh, S. Biomolecular condensates as stress sensors and modulators of bacterial signaling. PLoS Pathog.20, e1012413 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Passos, C. et al. BR-Bodies facilitate adaptive responses and survival during copper stress in Caulobacter crescentus. J. Biol. Chem. 301, 110648 (2025). [DOI] [PMC free article] [PubMed]

[CR13] 13.Lu, N. et al. Scaffold-scaffold interaction facilitates cell polarity development in Caulobacter crescentus. mBio14, e0321822 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Ahmed, Y. M., Brown, L. M., Varga, K. & Bowman, G. R. Phospho-signaling couples polar asymmetry and proteolysis within a membraneless microdomain in Caulobacter crescentus. Nat. Commun.15, 9282 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Lasker, K. et al. Selective sequestration of signalling proteins in a membraneless organelle reinforces the spatial regulation of asymmetry in Caulobacter crescentus. Nat. Microbiol.5, 418–429 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Connolley, L., Schnabel, L., Thanbichler, M. & Murray, S. M. Partition complex structure can arise from sliding and bridging of ParB dimers. Nat. Commun.14, 4567 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Zhao, Y. et al. Phase-separated ParB enforces diverse DNA compaction modes and stabilizes the parS-centered partition complex. Nucleic Acids Res.52, 8385–8398 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.MacCready, J. S., Tran, L., Basalla, J. L., Hakim, P. & Vecchiarelli, A. G. The McdAB system positions α-carboxysomes in proteobacteria. Mol. Microbiol.116, 277–297 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Choi, S., Knoerdel, A. R., Sing, C. E. & Keating, C. D. Effect of polypeptide complex coacervate microenvironment on protonation of a guest molecule. J. Phys. Chem. B. 127, 5978–5991 (2023). [DOI] [PubMed] [Google Scholar]

[CR20] 20.King, M. R. et al. Macromolecular condensation organizes nucleolar sub-phases to set up a pH gradient. Cell 187, 1889–1906 (2024). [DOI] [PMC free article] [PubMed]

[CR21] 21.Dai, Y. et al. Biomolecular condensates regulate cellular electrochemical equilibria. Cell187, 5951–5966 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Ausserwöger, H. et al. Biomolecular condensates sustain pH gradients at equilibrium through charge neutralization. Nat. Chem.18, 246–257 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Dai, Y. et al. Interface of biomolecular condensates modulates redox reactions. Chem9, 1594–1609 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Testa, A. et al. Sustained enzymatic activity and flow in crowded protein droplets. Nat. Commun.12, 6293 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Joshi, A. et al. Hydrogen-bonded network of water in phase-separated biomolecular condensates. J. Phys. Chem. Lett.15, 7724–7734 (2024). [DOI] [PubMed] [Google Scholar]

[CR26] 26.Mukherjee, S. & Schäfer, L. V. Thermodynamic forces from protein and water govern condensate formation of an intrinsically disordered protein domain. Nat. Commun.14, 5892 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Bayas, C. A. et al. Spatial organization and dynamics of RNase E and ribosomes in Caulobacter crescentus. Proc. Natl. Acad. Sci. USA115, E3712–E3721 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Gahlmann, A. & Moerner, W. E. Exploring bacterial cell biology with single-molecule tracking and super-resolution imaging. Nat. Rev. Microbiol.12, 9–22 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Joron, K. et al. Fluorescent protein lifetimes report densities and phases of nuclear condensates during embryonic stem-cell differentiation. Nat. Commun.14, 4885 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Nandana, V. et al. The BR-body proteome contains a complex network of protein-protein and protein-RNA interactions. Cell Rep. 42, 10.1016/j.celrep.2023.113229 (2023). [DOI] [PMC free article] [PubMed]

[CR31] 31.Farag, M. et al. Condensates formed by prion-like low-complexity domains have small-world network structures and interfaces defined by expanded conformations. Nat. Commun.13, 7722 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Wu, T. et al. Single-fluorogen imaging reveals distinct environmental and structural features of biomolecular condensates. Nat. Phys.21, 778–786 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.von Hofe, J. et al. Multivalency controls the growth and dynamics of a biomolecular condensate. J. Am. Chem. Soc.147, 25242–25253 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Sumrall, E. R., Husna, S. M. A., Halder, B., Ahnoud, A. & Walter, N. G. Single-molecule microscopy of biomolecular condensates: From liquid droplets to heterogeneous networks. Curr. Opin. Struct. Biol.98, 103271 (2026). [DOI] [PubMed] [Google Scholar]

[CR35] 35.Chattaraj, A. & Shakhnovich, E. I. Multi-condensate state as a functional strategy to optimize the cell signaling output. Nat. Commun.15, 6268 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Ortiz-Rodríguez, L. A. et al. Stress changes the material state of a bacterial biomolecular condensate and shifts its function from mRNA decay to storage. Nat. Commun.16, 10019 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Munder, M. C. et al. A pH-driven transition of the cytoplasm from a fluid- to a solid-like state promotes entry into dormancy. ELife5, e09347 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Joyner, R. P. et al. A glucose-starvation response regulates the diffusion of macromolecules. ELife.https://elifesciences.org/articles/9376 (2016). [DOI] [PMC free article] [PubMed]

[CR39] 39.Currie, S. L. et al. Quantitative reconstitution of yeast RNA processing bodies. Proc. Natl. Acad. Sci. USA120, e2214064120 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Gonçalves, D., Santiago, A. & Morano, K. A. When pH comes to the rescue. ELife9, e62022 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Goetz, S. K. et al. Structural reorganization underlying stress-induced cytoplasmic solidification in yeast. Preprint at 10.64898/2025.12.22.696049 (2025). [DOI]

[CR42] 42.Heinrich, K., Sobetzko, P. & Jonas, K. A kinase-phosphatase switch transduces environmental information into a bacterial cell cycle circuit. PLOS Genet.12, e1006522 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Kesler, V., Murmann, B. & Soh, H. T. Going beyond the debye length: overcoming charge screening limitations in next-generation bioelectronic sensors. ACS Nano14, 16194–16201 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Dudev, T. & Lim, C. Competition among Ca2+, Mg2+, and Na+ for model ion channel selectivity filters: determinants of ion selectivity. J. Phys. Chem. B116, 10703–10714 (2012). [DOI] [PubMed] [Google Scholar]

[CR45] 45.Friesen, S., Kruchinin, S. E., Fedotova, M. V. & Buchner, R. Cation-binding of glutamate in aqueous solution. J. Phys. Chem. B. 128, 5746–5755 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Verma, S., Ravichandiran, V. & Ranjan, N. Beyond amyloid proteins: Thioflavin T in nucleic acid recognition. Biochimie190, 111–123 (2021). [DOI] [PubMed] [Google Scholar]

[CR47] 47.Islam, M. D. et al. Impact of pseudouridylation, substrate fold, and degradosome organization on the endonuclease activity of RNase E. RNA27, 1339–1352 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Parry, B. R. et al. The bacterial cytoplasm has glass-like properties and is fluidized by metabolic activity. Cell156, 183–194 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Joyner, R. P. et al. A glucose-starvation response regulates the diffusion of macromolecules. ELife5, e09376 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Nurmohamed, S. et al. Polynucleotide phosphorylase activity may be modulated by metabolites in Escherichia coli. J. Biol. Chem.286, 14315–14323 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Symmons, M. F., Jones, G. H. & Luisi, B. F. A duplicated fold is the structural basis for polynucleotide phosphorylase catalytic activity, processivity, and regulation. Structure8, 1215–1226 (2000). [DOI] [PubMed] [Google Scholar]

[CR52] 52.Stoffel, F. et al. Enhancement of enzymatic activity by biomolecular condensates through pH buffering. Nat. Commun.16, 6368 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Greenwich, J. L., Alakavuklar, M. A. & Fuqua, C. An Inducible T7 polymerase system for high-level protein expression in diverse gram-negative bacteria. Microbiol. Resour. Announc.12, e01119–e01122 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Ducret, A., Quardokus, E. M. & Brun, Y. V. MicrobeJ, a tool for high throughput bacterial cell detection and quantitative analysis. Nat. Microbiol.1, 1–7 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Saurabh, S., Perez, A. M., Comerci, C. J., Shapiro, L. & Moerner, W. E. Super-resolution imaging of live bacteria cells using a genetically directed, highly photostable fluoromodule. J. Am. Chem. Soc.138, 10398–10401 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Ovesný, M., Křížek, P., Borkovec, J., Švindrych & Hagen, Z. G. M. ThunderSTORM: a comprehensive ImageJ plug-in for PALM and STORM data analysis and super-resolution imaging. Bioinformatics30, 2389–2390 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Hahsler, M., Piekenbrock, M. & Doran, D. dbscan: Fast density-based clustering with R. J. Stat. Softw.91, 1–30 (2019). [Google Scholar]

[CR59] 59.Dolinsky, T. J., Nielsen, J. E., McCammon, J. A. & Baker, N. A. PDB2PQR: an automated pipeline for the setup of Poisson–Boltzmann electrostatics calculations. Nucleic Acids Res.32, W665–W667 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Baker, N. A., Sept, D., Joseph, S., Holst, M. J. & McCammon, J. A. Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA98, 10037–10041 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. J. Mol. Biol.215, 403–410 (1990). [DOI] [PubMed] [Google Scholar]

[CR62] 62.Sayers, E. W. et al. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res.50, D20–D26 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Emenecker, R. J., Griffith, D. & Holehouse, A. S. Metapredict: a fast, accurate, and easy-to-use predictor of consensus disorder and structure. Biophys. J.120, 4312–4319 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Wilkins, M. R. et al. Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol.112, 531–552 (1999). [DOI] [PubMed] [Google Scholar]

PERMALINK

Bacterial ribonucleoprotein bodies maintain an acidic pH environment as a mechanism of enzyme regulation

Wade E Schnorr

Moeka Sasazawa

Kathryn G Dzurik

Kaveendya S Mallikaarachchi

Hadi Yassine

Kulathungage H Dilrangi

Ryan Cho

Shelby L Millheim

Jill E Millstone

Jared M Schrader

Saumya Saurabh

W Seth Childers

Abstract

Introduction

Fig. 1. The BR-body chemical environment is defined by a high density of scaffolding and client proteins with a high density of negative charge.

Results

In vivo pH measurements of Caulobacter crescentus expressing RNase E-pHluorin2

Fig. 2. BR-bodies form distinct acidic subcellular environments in Caulobacter crescentus cells.

BR-bodies exhibit structural and biochemical heterogeneity

Fig. 3. Super-resolution and fluorescence lifetime imaging (FLIM) reveal structural and biochemical heterogeneity in BR-bodies.

Increased RNase E Scaffold pI leads to a more neutral BR-Body pH in vivo

Fig. 4. Neutral-pI RNase E scaffolds in C. crescentus exhibit more neutral BR-body pH and reduced phase separation properties.

SMLM analysis indicates RNase E pI tunes BR-body phase properties

Stationary phase drives BR-body and cytoplasm acidification

RNase E–pI7 is not viable when grown from stationary phase cultures

Fig. 5. Neutral-pI RNase E scaffolds in C. crescentus exhibit severely reduced viability when grown from stationary phase.

Ethanol stress is characterized by a decrease in BR-body pH

Fig. 6. C. crescentus strains expressing the RNase E-pI7 variant exhibited altered morphological response when grown in 5% ethanol stress.

C-SNARF and pHluorin2 reveal an RNase E condensate pH gradient in vitro

Fig. 7. RNase E condensates act as a spatially regulated acidic pH buffer in vitro.

RNase E’s primary clients do not alter RNase E condensate pH in vitro

Acidic condensate environments enhance the activity of PNPase

Fig. 8. Phase separation with a pH gradient enhances poly(A) degradation and polymerization by PNPase.

Discussion

Fig. 9. Model for Stress-driven acidification and solidification of BR-bodies that inhibits RNA decay and enables RNA storage.

Methods

Imaging Caulobacter crescentus expressing RNase E-pHluorin2

Efficiency of plating assay

RT-qPCR mRNA half-live measurements

Super-resolution microscopy and FLIM of RNaseE-pHluorin2

FLIM Cluster analysis

RNase E CTD, RNase E CTD-pHluorin2 Expression and Purification

Imaging RNase E with C-SNARF-4F

Imaging RNase E-pHluorin2 with Tris-Cl, MgCl2, and NaCl titrations

Imaging RNase E-pHluorin2 with clients

Zeta potential measurements

PNPase activity assay

Calibration curve for in vivo Caulobacter crescentus pH measurements

Construction of a false-positive probability color map

Electrostatic visualization of PNPase structure

RNase E homolog and client charge

Designing of synthetic RNase E pI6, 7, and 10

pI6

pI7

pI10

Microscopy analysis of PNPase activity

PNPase-mCherry and poly(A)-Cy5 partitioning into RNase E CTD condensates

Statistics and reproducibility

Reporting summary

Supplementary information

Source data

Author contributions

Peer review

Peer review information

Funding

Data availability

Competing interests

Footnotes

Contributor Information

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Imaging RNase E-pHluorin2 with Tris-Cl, MgCl₂, and NaCl titrations