Compartmentalised RNA catalysis in membrane-free coacervate protocells

Phase separation of mixtures of oppositely charged polymers provides a simple and direct route to compartmentalisation via complex coacervation, which may have been important for driving primitive reactions as part of the RNA world hypothesis. However, to date, RNA catalysis has not been reconciled with coacervation. Here we demonstrate that RNA catalysis is viable within coacervate microdroplets and further show that these membrane-free droplets can selectively retain longer length RNAs while permitting transfer of lower molecular weight oligonucleotides.

This paper describes the first report of ribozyme activity within a complex coacervate environment. This result is of particular interest in the context of the RNA world hypothesis for the origin of life. The authors show preferential partitioning and retention of longer RNA (ribozymes) into coacervate droplets as compared with shorter substrate strands. These results provide support for the idea that liquid-liquid phase separated droplets could serve as a method for compartmentalizing material and providing a competitive evolutionary advantage.

Specific Comments:
1. Figure S2A shows the kinetic analysis for the reaction performed in a bulk coacervate phase.
However, the data plateaus at a cleaved fraction value of approximately 0.4, while the data in Figure S3B continues up to a value of 0.7. However, the data for microdroplet experiments extend to a cleaved fraction of ~1.0. Why is such a difference observed?
My first thought was that the kinetic measurements might be inaccurate in this case due to diffusional limitations. However, taking the values of the diffusivity determined via FRAP (1-2 µm 2 /s), both the ribozyme and substrate should be able to diffuse on the order of 300-400 µm over the course of a 1000 min experiment. This should be sufficient for equilibration across the coacervate phase. Could the limitation instead be diffusion throughout the entire sample since the droplet case would be much more dispersed throughout the sample volume, while the bulk phase would be more spatially localized in the tube? Even if the samples were mixed during the experiment, there should be a significant difference in the diffusional profiles of these samples. If true, this diffusive limitation could account for the differences in the kinetic rate parameters obtained between the bulk and droplet phases, rather than differences in partitioning of the TAM-HH-min and FAM-substrate into a large vs. a small volume of the same material.
2. While I agree that the positive result of RNA catalysis inside of a coacervate phase is exciting, the wording in the manuscript almost makes it seem like such results were also unexpected. Is there some reason why this might be true? Previous studies on protein-based enzymes have shown the ability to retain and possibly enhance enzyme activity in coacervate phases. Is there some aspect of RNA-based materials that might cause this trend not to hold for these materials?
3. I assume that the use of a capillary diffusion experiment to test for material exchange between the droplets was performed in order to avoid coalescence events that would complicate the analysis. It might be worth mentioning this in the manuscript. However, I do wonder if it would have been beneficial to use the type of microfluidic technology that has been reported previously (van Swaay et al., Angewandte Chemie, (2015), 54, 8398-8401) to prepare stable dispersions of droplets and thus highlight the potential for observing a direct competitive advantage.
4. I suggest the following two references to further support the authors case about length-dependent competitive sequestration of polymers into a coacervate: • Zhang and Shklovskii, Physica A, (2005), 352, 216-238.

5.
A tremendous amount of information about the materials and procedures is available in the SI. It would be very useful to include a statement in the materials and methods section (and throughout the manuscript where relevant) describing the types of information available in the SI.
6. Much of the work in this manuscript was done in "single-turnover conditions." Does this simply mean that a stoichiometric amount of ribozyme and reactant were present in the mixture, or that each ribozyme was only capable of performing a single reaction and did not recover into a "ground state?" In the SI the concentrations appear to be 2:1 ribozyme:substrate. Please clarify.
Minor Comments: 1. Please define abbreviations like CM-Dex and PLys in the manuscript.
2. At the bottom of Page 3 of the SI, "than" should be "then".
3. At the top of Page 4 of the SI there appears to be a glitch in the following sentence: "TAM-HH-min or FAM-substrate were loaded into 150 µl dispersions containing into 3 µl of bulk coacervate phase achieve a final initial concentration of 0.36 µM (c ini )." 4. What is the error in the data shown in Figure S1B?
5. In the caption for Figure S2B "plot" should be "plotted".

Reviewer #2 (Remarks to the Author):
This paper addresses an interesting question, whether ribozyme activity is altered in coacervates, and whether ribozymes can maintain genetic identity in the coacervate phase. The authors cite the potential interest in coacervates as model protocells (as in the title), which motivates these experiments.
The main results are 1) the HH-min ribozyme works in the coacervate phase but is quite a bit slower than in aqueous buffer, 2) the HH-min ribozyme kinetics are biphasic in the coacervate phase, 3) these rates are modestly increased in droplets compared to bulk coacervate, 4) the HH-min ribozyme does not diffuse much among droplets but the shorter substrate RNA does.
Overall, my impression is that while the questions and results are interesting, more work and discussion is needed to understand the results, especially in the context of prior work. There are two main issues. First, the work (result 4) must be taken in context of Jia, Hentrich, and Szostak, 2014, which studied diffusion of RNA among coacervate droplets and concluded that exchange was rapid.
The studies presented here must reconcile their opposing results with the previous study. For example, if the difference is the composition of the coacervate, the authors could validate their experimental setup by repeating the composition of Jia et al. in their microscopic droplet mixing experiment. Along this vein, the first instance of "CM-Dex:PLys" should be defined. I was assuming this represents a carboxylated dextran and lysine; if so, the reason for choosing this composition should be given since it is not an obvious prebiotic choice. The composition should be justified in any case. Figure 1C: The curve fit equations should be given explicitly with coefficients and constants. Does the curve begin from 0? If not, why not? What is the dead time of this experiment? Is (i) showing a different experiment from (ii); if so, how are these different? How do the grey points in (ii) indicate standard deviation -why would the stdev increase over time? Figure 1D: Again, the curve fit equation should be given. These curves appear to begin at 0, but the relative amplitude of the two exponential fits should be given. Also, a supplementary figure should be included to show this graph as a semilog plot to demonstrate the two regimes. Same question as Fig. 1C with respect to stdev. Figure 3A: Axes labels on insets are needed. I assume these are distributions of intensity; some comment on the procedure for intensity normalization between samples is needed.
Why would FRAP recovery only reach 70% for HH-min? This needs at least a proposed explanation.
Semantic issue: I do not think 'protocell' is normally used to describe coacervates; this could potentially be confusing to readers who understand the word 'protocell' to be a membrane-bound compartment. I would recommend eliminating this word from the title.

Reviewer #3 (Remarks to the Author):
The study entitled "Compartmentalized RNA catalysis in membrane -free coacervate protocells" reported the ribozyme activity in coacervate and size-dependent diffusion of RNAs among coacervate microdroplets. The authors analyzed the kinetics of a ribozyme and diffusion rate in detail. The manuscript was clearly written and the measurement and analysis were well performed. However, I have two major concerns as described below and they are critically important for the significance of this study to readers in a broad field of science. Therefore, I cannot recommend this manuscript for publication in Nature Communications.
Major points 1. The first concern is about novelty. I agree that ribozyme activity has not been observed in coacervate. But there have been several studies about other biochemical reactions, such as actinorhodin polyketide synthesis and gene expression, as the authors cited in references 8-9.
Therefore, ribozyme reaction in coacervate is not surprising and I don't believe this is a significant advance. I admit that the authors precisely analyzed the kinetics and diffusion and this study is worth publishing, but a more specific journal would be suitable.
2. The second concern is about one of the main claim of this study, selective retention of a long RNA in coacervate. I understood that this claim based on the result of Figure 4. I think there are two problems in this experiment to withdraw the conclusion.
First, the localization of the substrate RNA was not directly measured. I agreed that he HH-RNAs were localized in region 1, whereas the localization of the substrate RNA was not directly determined. Instead, the authors observed the FRET signal produced by ribozyme activity, which depends on the localization of both ribozyme and substrate and thus cannot be regarded as a direct evidence of substrate localization. Direct measurement of the substrate localization should be measured to conclude the different retention time.
Second, the difference in the retention of the two RNAs can be attributed to other factors than RNA sizes, such as labeled fluorescent compound (TAM or FAM), RNA sequences, or RNA secondary structures. To verify the effect of RNA size, additional experiments with various size and sequence of RNAs are required.

Minor point
The subsection of "Preparation of bulk coacervate phase…" of Materials and Method section, was difficult to understand. I believe the problems is lack of explanation of several phases: coacervate phase, polymer phase, bulk polymer phase, supernatant phase. Is the "coacervate phase" same as polymer phase? Clarification of these terms would be kind for readers.

General Comments:
This paper describes the first report of ribozyme activity within a complex coacervate environment. This result is of particular interest in the context of the RNA world hypothesis for the origin of life. The authors show preferential partitioning and retention of longer RNA (ribozymes) into coacervate droplets as compared with shorter substrate strands. These results provide support for the idea that liquid-liquid phase separated droplets could serve as a method for compartmentalizing material and providing a competitive evolutionary advantage.
Specific Comments: 1. Figure S2A shows the kinetic analysis for the reaction performed in a bulk coacervate phase. However, the data plateaus at a cleaved fraction value of approximately 0.4, while the data in Figure S3B continues up to a value of 0.7. However, the data for microdroplet experiments extend to a cleaved fraction of ~1.0. Why is such a difference observed? My first thought was that the kinetic measurements might be inaccurate in this case due to diffusional limitations. However, taking the values of the diffusivity determined via FRAP (1-2μm2/s), both the ribozyme and substrate should be able to diffuse on the order of 300-400 μm over the course of a 1000 min experiment. This should be sufficient for equilibration across the coacervate phase. Could the limitation instead be diffusion throughout the entire sample since the droplet case would be much more dispersed throughout the sample volume, while the bulk phase would be more spatially localized in the tube? Even if the samples were mixed during the experiment, there should be a significant difference in the diffusional profiles of these samples. If true, this diffusive limitation could account for the differences in the kinetic rate parameters obtained between the bulk and droplet phases, rather than differences in partitioning of the TAM HH-min and FAM-substrate into a large vs. a small volume of the same material.
We thank the reviewer for the comment regarding the system being diffusion limited within the coacervate phase and microdroplet environment. To address this, we have calculated the concentration of ribozyme and substrate within the bulk coacervate phase and in the coacervate microdroplet (Shown in the table below) based on our partitioning experiments and observe an almost 50 x increase in the concentration of the ribozyme and substrate within the coacervate microdroplets compared to the bulk coacervate phase.
[ Therefore, we attribute the difference in cleaved fraction at specific time points as observed in Figure S3A and 4B (previous Figure S2A and 3B) to a dilution effect or up-concentration effect within the bulk coacervate phase. It would be expected that with increased concentrations, as observed in other enzymatic reactions in solution, the distances of molecular diffusion will be reduced. This will lead to increased collision and binding events between the enzyme and substrate and increased rates of reaction. With the lower concentrations in the bulk coacervate phase, it would be expected that 100% cleavage will take place over a long time frame. Therefore, the differences in the kinetic rate parameters are attributed to secondary effects from the up concentration of the enzyme and substrate and this has been described in the following text starting on page 5, line 122. 2. While I agree that the positive result of RNA catalysis inside of a coacervate phase is exciting, the wording in the manuscript almost makes it seem like such results were also unexpected. Is there some reason why this might be true? Previous studies on proteinbased enzymes have shown the ability to retain and possibly enhance enzyme activity in coacervate phases. Is there some aspect of RNA-based materials that might cause this trend not to hold for these materials?
As the referee states correctly, enzyme catalysis and in vitro transcription and translation within the coacervate environment have been shown previously. However, in these examples, the enzymatic machineries are large and highly evolved macromolecules, which are adapted to fold and function in crowded environments. They have highly evolved molecular scaffolds and often auxiliary domains, which enable the stable formation of active sites and interfaces (depending on their function). In the case of primitive RNA catalysis (such as the minimal form of our hammerhead ribozyme), there are no or only very few stabilizing domains. Thus, the highly charged and crowded interior of the coacervate phase might very well disrupt or destabilize primitive ribozymes and thus inhibit their activity. Therefore, despite previous results, it would not be expected that minimal forms of ribozymes are active within the coacervate microenvironment. Therefore, additional text has been added to the manuscript to explain why the result cannot be assumed.
On page 2, line 46, we have now written: "Therefore, coacervate protocells based on Carboxymethyl Dextran sodium salt (CM-Dex) and Poly-L-Lysine (PLys) ( Figure S1) were chosen as the model system due to their proven ability to encapsulate and support complex biochemical reactions catalysed by highly evolved enzymes 9,10 . In contrast to these enzymes, structurally simple ribozymes, which are thought to have played a key role during early biology, lack structural stability and therefore may be rendered inactive by interactions within the highly charged and crowded interior. Herein, we directly probe the effect of the coacervate microenvironment on primitive RNA catalysis, and show the ability of the coacervate microenvironment to support RNA catalysis whilst"…. 3. I assume that the use of a capillary diffusion experiment to test for material exchange between the droplets was performed in order to avoid coalescence events that would complicate the analysis. It might be worth mentioning this in the manuscript. However, I do wonder if it would have been beneficial to use the type of microfluidic technology that has been reported previously (van Swaay et al., Angewandte Chemie, (2015), 54, 8398-8401) to prepare stable dispersions of droplets and thus highlight the potential for observing a direct competitive advantage.
The reviewer is correct and the capillary diffusion experiment was undertaken to test for material exchange. There are some advantages to using the microfluidic device as described in the 2015 Van Swaay paper i.e. droplets prepared by this microfluidic devices are stable to coalescence over a time frame of weeks and the droplets show increased monodispersity compared to droplets prepared by hand. However, there is no added value in using the microfluidic devices for the capillary diffusion experiments as the droplets prepared by hand are stable over the course of the experiment (600 mins) with no evidence of droplet coalescence after droplet loading As the reviewer has suggested we have mentioned the use of the capillary experiments to allow for material exchange whilst avoiding coalescence. The sentence below was in the materials and methods but has now been moved to the main text: " in such a way as to prevent droplet mixing whilst permitting passive diffusion of molecules through the length of the channel." The main text now reads (page 6, line 162): …"in such a way as to prevent droplet mixing whilst permitting passive diffusion of molecules through the length of the channel (see materials and methods)"….

A tremendous amount of information about the materials and procedures is available in the SI. It would be very useful to include a statement in the materials and methods section (and throughout the manuscript where relevant) describing the types of information available in the SI.
As recommended by the editor, all the materials and methods have been moved to the main manuscript. Additionally, on the reviewer's advice, we have added in, where appropriate, additional references to the materials and method in the main text.

On page 3, lines 57-66: "We developed a real-time fluorescence resonance energy transfer (FRET) assay (see materials and methods) to investigate the effect of the coacervate microenvironment on catalysis of a minimal version of the hammerhead ribozyme derived from satellite RNA of tobacco ringspot virus (HH-min) 29 . HH-min and its FRET-substrate (Figure 1A, materials and methods) were incubated within a bulk polysaccharide / polypeptide coacervate phase or within coacervate microdroplets under single turnover conditions (see materials and methods). Cleavage of the FRET-substrate strand by HH-min increases the distance between 6-carboxyfluorescein (FAM) and Black Hole quencher 1 (BHQ1), resulting in increased fluorescence intensity. We further developed an inactive control ribozyme (HHmut) by introducing two point mutations at the catalytic site (see materials and methods)."
On page 3, line 74: 6. Much of the work in this manuscript was done in "single-turnover conditions." Does this simply mean that a stoichiometric amount of ribozyme and reactant were present in the mixture, or that each ribozyme was only capable of performing a single reaction and did not recover into a "ground state?" In the SI the concentrations appear to be 2:1 ribozyme:substrate. Please clarify.
Single turnover conditions means that there is stoichiometric amounts or an excess or enzyme. Under these conditions, the ribozyme would perform a single reaction and, in an ideal situation, would return to its ground state with full dissociation of the substrate/product.  Figure S2B with the single exponential fit. In addition, the text has been modified in the main text to state the error and the materials and methods modified to describe the fitting procedure:

Changes to materials and methods,
The kinetics of HH-min were determined by integrating the band intensities of the cleaved and uncleaved bands and performing a global fit to a single exponential (Equation1), yielding a first order rate constant of 0.38 ± 0.05 min -1 (see materials and methods). Data points are an average of six independent measurements. Note that the in-gel fluorescent intensities of cleaved and uncleaved FAM-labelled RNA are different and are attributed to the FRET effect ( Figure S13B). Figure S2B "plot" should be "plotted". "The increase in fluorescence intensity from cleaved product (grey data) was plotted as a function of time and the thickness of the grey data points are from the standard deviation from five repeats."

In the caption for
This paper addresses an interesting question, whether ribozyme activity is altered in coacervates, and whether ribozymes can maintain genetic identity in the coacervate phase. The authors cite the potential interest in coacervates as model protocells (as in the title), which motivates these experiments.
The main results are 1) the HH-min ribozyme works in the coacervate phase but is quite a bit slower than in aqueous buffer, 2) the HH-min ribozyme kinetics are biphasic in the coacervate phase, 3) these rates are modestly increased in droplets compared to bulk coacervate, 4) the HH-min ribozyme does not diffuse much among droplets but the shorter substrate RNA does.
Overall "As other studies have shown that RNA rapidly exchanges from PLys and Adenosine Triphosphate (ATP, Figure S1) coacervate microdroplets into the surrounding environment 39 , we also tested this coacervate system for selective localization of RNA. To this end, localization experiments were undertaken as previously described (see materials and methods) with PLys : ATP coacervate microdroplets (4:1 molar ratio) at pH 8: Droplets containing either TAM-HH min or FAM-substrate were loaded into one end of a capillary channel, and coacervate droplets containing HH-mut were loaded into the other end of the capillary channel in such a way as to prevent droplet mixing ( Figure S11). Fluorescence optical microscopy images obtained in the middle of the channel (region 2, Figure S11B) showed no change in the fluorescence intensity of TAM-HH-min ( Figure S11C, i) over the course of the experiment (500 min). In contrast, a small increase in the fluorescence intensity from FAM-substrate ( Figure  S11C, ii) was observed after 300 min, suggesting a higher exchange rate of the 12-mer with the environment. Whole droplet FRAP experiments of PLys / ATP coacervate microdroplets containing either TAM-HH-min (39-mer), FAM-substrate (12-mer) or cleaved FAM-substrate (6-mer) ( Figure S12) confirmed a consistent trend in RNA retention based on RNA length with an order of magnitude difference in t between the different oligonucleotides ( Figure S12).
Whilst the general trends are consistent with those observed with CM-Dex : PLys coacervate microdroplets, a direct comparison of whole droplet FRAP recovery times (Table S4)

Along this vein, the first instance of "CM-DexLys" should be defined. I was assuming this represents a carboxylated dextran and lysine; if so, the reason for choosing this composition should be given since it is not an obvious prebiotic choice. The composition should be justified in any case.
We thank the reviewer for this comment and have added an additional few sentences to define CM-Dex and PLys (page 2, line 147). For clarity we have included the molecular structures in the SI (new Figure S1). A justification for the composition has also been included in the introduction on page 2, line 46: "Therefore, coacervate protocells based on Carboxymethyl Dextran sodium salt (CM-Dex) and Poly-L-Lysine (PLys) ( Figure S1) were chosen as the model system due to their proven ability to encapsulate and support complex biochemical reactions catalysed by highly evolved enzymes 9,10 . In contrast to these enzymes, structurally simple ribozymes, which are thought to have played a key role during early biology, lack structural stability and therefore may be rendered inactive by interactions within the highly charged and crowded interior….." The additional figure S1, in supplementary information, page 3, line 36  (results 1-3). I believe more attention, at least discussion, and experiments if possible, regarding possible mechanisms for each of these observations is needed to understand the potential importance of these results to the field.

For example, a statistically significant acceleration of ~5-fold in rate constants was observed in microdroplets vs. bulk phase. The text states a 'slight alternation of the material properties' but the mechanism of such an effect, even if only proposed but not verified, is not obvious to me. Do the authors mean the viscosity? If so, this could probably be measured.
A similar concern relates to the biphasic curve (some mention is given of alternate conformers -perhaps this could be tested by chemical probing or even spectroscopically, e.g., CD). An explanation for the quite significant slowing of the ribozyme in the coacervate is also needed given the observation in cited literature of a hammerhead ribozyme accelerated in an ATPS.
We thank the reviewer for this comment and have now included additional discussion regarding the observations of different rate constants. Taking into account the comments from reviewer 1, comment 1, we have also considered that the difference in the rate constants may be attributed  With regard to the change in the transition from the mono-exponential to biphasic behavior within the coacervate phase compared to the buffer solution. We have undertaken CD spectra, which show a decrease in the secondary structure and a small shift in the peak maxima. This supports the hypothesis that there are heterogeneous ribozyme populations with different secondary structure, conformational and equilibrium states. Therefore, we have changed the text accordingly (page 4, lines 90-97).

From:
"This may be attributable to heterogeneous ribozyme populations with alternative conformational and equilibrium states, as observed for some HH systems in aqueous buffer conditions. 31,32 It is possible that the charged and crowded coacervate microenvironment affects the structure of HH-min, restricting substrate binding, sterically hindering substrateenzyme complex formation and/or spatially restricting diffusion of the cleavage assay components. " ([q]) for HH-mut within bulk coacervate phase compared to aqueous buffer with a small commensurate shift in the peak maxima from 265 nm to 268 nm respectively ( Figure S4). These results show that the fold of HH-mut is altered in the polyelectrolyte-rich bulk coacervate phase, with an overall loss of secondary structure that could affect catalytic activity."

To: " This may be attributable to heterogeneous ribozyme populations with variations to secondary structure and / or alternative conformational and equilibrium states, as observed for some HH systems in aqueous buffer conditions. 31,32 Circular dichroism (CD) spectra show a reduction in Molar Ellipticity
A new figure showing the CD-data has been placed in the SI as Figure S4. . The shaded region shows the standard deviation from 3 spectra, each an average of 5 repeats (orange and yellow) and 13 spectra, each an average of 10 repeats (blue). All spectra are shown with the background (either cleavage buffer or bulk polymer phase) removed.

Additional protocals for the materials and methods have been described on page 19, line 486-495:
"Circular Dichroism. To investigate the secondary structure of the hammerhead ribozyme, HH-mut was mixed into either buffer (cHH = 1 or 2 µM) or bulk coacervate phase (cHH = 2 µM) (CM-Dex : PLys, 4:1 molar ratio) and loaded into a 1 mm Special Quartz Cuvette (200 µL). CD spectra were measured using a Chirascan TM -Plus CD Spectrometer (Applied Photophysics), with data collected every 1 s at 25 o C from 320 to 200 nm with a resolution of 1 nm. Either 5 or 10 repeat spectra were measured for buffer and bulk coacervate samples respectively. Background spectra of buffer alone and bulk coacervate phase were taken under the same conditions and were subtracted from the spectra of HH-mut in buffer and bulk coacervate phase respectively. All spectra were offset at 320 nm, normalised for the cuvette pathlength and converted from DA to Molar Ellipticity ([q]) (deg.cm 2 .dmol -1 ). " Some additional comments are given below.

"(C) Real-time cleavage kinetics in 10 mM Tris•HCl pH 8.3 and 4 mM MgCl2. (i) A monoexponential fit (grey line) to kinetic data (grey dots) and residuals of the fit (inset) (ii) Blue line is the mean of the individual fits (blue line). Grey data points represents the standard deviation (N ≥ 5) from experimental data. "
To:

"(C) (i) A monoexponential fit (materials and methods, Equation 3) (grey line) to kinetic data (grey dots) and residuals of the fit (inset) (ii) mean of the individual fits to each experiment (Blue line) with the standard deviation of the mean of the fits (grey data points) (N = 5). "
In the materials and methods, all the terms have additionally described: Where I (t) is the normalized intensity, A1 is the amplitude of the first 1-A1 the amplitude of the second exponential, t is the time in min, t0 is the dead time between sample preparation and the first measurement (separately measured), t1 or t2 are the fitted time constants. The corresponding rate constants k1 or k2 are obtained from t1 or t2 where k=1/t." Figure 1D: Again, the curve fit equation should be given. These curves appear to begin at 0, but the relative amplitude of the two exponential fits should be given. Also, a supplementary figure should be included to show this graph as a semilog plot to demonstrate the two regimes. Same question as Fig. 1C with respect to stdev.
As discussed previously, it is not usual practice to place the fit equations within the figure legends, therefore, the curve fit equation has been referred to within the figure legend (see below). In addition, to clarify the meaning of the standard deviations, the figure legend has been rewritten: Page 26, line 632.

" (i) Biexponential fit (materials and methods Equation 2) (dark grey line) to experimental data (grey dots) with the residuals (inset) (ii) mean biexponential fit (orange) of individual fits (N ≥ 5). Grey data points represent the standard deviation (N ≥ 5) from the experimental data."
To:

"(i) Cleavage in bulk coacervate phase (normalized to the amount of cleaved product at t = 530 min from gel electrophoresis). (i) Biexponential fit (materials and methods, Equation 4) (dark grey line) to experimental data (grey dots) with the residuals (inset) (ii) mean biexponential fit (orange) of individual fits (N ≥ 5). Grey data points represent the standard deviation (N = 5) from the experimental data."
The amplitudes of the two exponential fits have been added to SI table 3: For repeat experiments, the relative amplitudes have been described in the relevant figure legend of figure S5 (supplementary information, page 7 line 100). Single exponential fits to the data show non-random residuals (inset) indicating that fits to single-exponential decays are not reliable.   For this experiment, the intensities were normalized to the amount of cleaved product, which was obtained from gel electrophoresis. This is explained in the materials and methods via the following text in the materials and methods on page 13, line 333: "Band intensities were measured using ImageQuant at a specific time point and uncleaved substrate was corrected by the FRET effect factor (1.69) ( Figure S12). The fraction of cleaved substrate of the total sum of cleaved and uncleaved substrate was determined and used to normalize kinetic data obtained from spectroscopy or microscopy." The study entitled "Compartmentalized RNA catalysis in membrane -free coacervate protocells" reported the ribozyme activity in coacervate and size-dependent diffusion of RNAs among coacervate microdroplets. The authors analyzed the kinetics of a ribozyme and diffusion rate in detail. The manuscript was clearly written and the measurement and analysis were well performed. However, I have two major concerns as described below and they are critically important for the significance of this study to readers in a broad field of science. Therefore, I cannot recommend this manuscript for publication in Nature Communications.
Major points 1. The first concern is about novelty. I agree that ribozyme activity has not been observed in coacervate. But there have been several studies about other biochemical reactions, such as actinorhodin polyketide synthesis and gene expression, as the authors cited in references 8-9. Therefore, ribozyme reaction in coacervate is not surprising and I don't believe this is a significant advance. I admit that the authors precisely analyzed the kinetics and diffusion and this study is worth publishing, but a more specific journal would be suitable.
The reviewer is correct and there have been previous studies, which have demonstrated enzymatic activities within the coacervate. However, the enzymes involved in these studies are highly evolved, large and structurally complex macromolecules and not primitive structures like the minimal form of our 39 nt hammerhead ribozyme. The ribozyme lacks the structural integrity of a protein enzyme and is therefore much more susceptible to environmental conditions. In particular, the crowded polyelectrolyte environment of the coacervate bulk phase could affect the secondary structure and therefore, it is far from obvious that the ribozyme would retain its activity under these conditions. Despite this, our studies show for the first time that primitive RNA catalysis within coacervate microdroplets. We believe that our results are novel and interesting for a broad audience including the Origins of life, synthetic biology and the biology communities.
An additional sentence has been added to the main text to explain why the results that we show cannot be assumed and therefore demonstrates novelty (page 2, from line 46).
"Therefore, coacervate protocells based on Carboxymethyl Dextran sodium salt (CM-Dex) and Poly-L-Lysine (PLys) ( Figure S1) were chosen as the model system due to their proven ability to encapsulate and support complex biochemical reactions catalysed by highly evolved enzymes 9,10 . In contrast to these enzymes, structurally simple ribozymes, which are thought to have played a key role during early biology, lack structural stability and therefore may be rendered inactive by interactions within the highly charged and crowded interior. Herein, we directly probe the effect of the coacervate microenvironment on primitive RNA catalysis,and show the ability of the coacervate microenvironment to support RNA catalysis whilst selectively sequestering ribozymes and permitting transfer of lower molecular weight oligonucleotides.".

2.
The second concern is about one of the main claim of this study, selective retention of a long RNA in coacervate. I understood that this claim based on the result of Figure 4. I think there are two problems in this experiment to withdraw the conclusion.
First, the localization of the substrate RNA was not directly measured. I agreed that he HH-RNAs were localized in region 1, whereas the localization of the substrate RNA was not directly determined. Instead, the authors observed the FRET signal produced by ribozyme activity, which depends on the localization of both ribozyme and substrate and thus cannot be regarded as a direct evidence of substrate localization. Direct measurement of the substrate localization should be measured to conclude the different retention time.
We thank the reviewer for this comment and agree that figure 4 shows localisation of the ribozyme and substrate. However, we also base our conclusions regarding length dependent sequestration in the coacervate phase by direct measurement of the substrate as discussed by the reviewer. This is shown in the supplementary information (previous Figure S7,    "Additionally, we find that the RNA sequence can have an impact on partitioning where an increase in purine content decreases partitioning of 12mer RNAs approximately 10-fold compared to the highly pyrimidine-rich HH-substrate ( Figure S9). For the RNAs specific to our HH ribozyme assay, the general selective retention of longer length polynucleotides with transfer of shorter length RNA can have interesting implications for ribozyme catalysis within coacervate droplets."   Minor point: The subsection of "Preparation of bulk coacervate phase…" of Materials and Method section, was difficult to understand. I believe the problems is lack of explanation of several phases: coacervate phase, polymer phase, bulk polymer phase, supernatant phase. Is the "coacervate phase" same as polymer phase? Clarification of these terms would be kind for readers. Add in explanation into the materials and methods of what these phases are.
We thank the reviewer for the comment and have included an additional explanation of the coacervate phase and the microdroplet in the materials and methods (page 11, line 279). I appreciate the obvious effort and attention to detail that went into the revised manuscript. My only remaining question relates to the observed differences between ribozyme concentration, and thus activity in bulk coacervates vs. a microdroplet dispersion. My sense is that the authors have tried to lay out a possible explanation as to why these differences have been observed. However, a lack of clarity in methodology and the way in which these explanations are presented seem to suggest an interesting observation, rather than a potential flaw in the experiment.
My confusion begins on page 4, where the authors state, "The final concentration of enzyme and substrate in the microdroplet dispersion was equivalent to the final concentration of the bulk coacervate phase under single turnover conditions (1 µM of HH-min and 0.5 µM FRET-substrate)." However, on the following page, they claim that the concentrations in the two phases are in fact different: "From the partition coefficient we calculated concentrations of 49.6 μM HH-min and 24.3 μM substrate in a single microdroplet compared to 1 μM HH-min and 0.5 μM substrate within the bulk coacervate phase." How is this possible? Coacervates are known to be equilibrium phases. Therefore, whether the material is present as a droplet or as a larger bulk phase, it should not be expected to have a different concentration.
Related to this, I was unclear as to how the samples were prepared. In the manuscript, the text suggests that the droplet dispersions were created by resuspending a bulk coacervate. However, there is no discussion of how this resuspension was achieved in the methods section for the activity experiments. Instead, there is a description of how an RNA-containing solution was mixed with either bulk coacervate or a dispersion of droplets. While care was taken to assure that the total volume of coacervate was the same between these two samples, there is no discussion about potential differences in equilibration time that would be critical based on the different geometries (and thus diffusion distances) between these two samples. I am very concerned that the observation of higher concentration and thus activity in the microdroplets is due to differences in the ability of the two samples to equilibrate because of differences in geometry, rather than an intrinsic difference between the two states of the material.
In contrast, the partitioning experiments allow equilibration of RNA with a suspension of microdroplets for 10 minutes, followed by centrifugation to form a bulk coacervate phase that is assayed. Based on this procedure, I do not see how there could be a difference in the partitioning between droplets and bulk phase as both materials are used in the assay.
The only possible way in which I could envision a difference in RNA concentration between bulk and droplet materials would be if the RNA selectively binds to the droplet surface. I was unsure whether the fluorescence images shown in Figures 3 and 4 were confocal or wide-field. However, confocal imaging would clearly highlight if there were such a concentration of RNA at the surface. Additionally, I would be surprised if the concentration of the droplet could be increased by 50x in such a manner.