Impact of wet-dry cycling on the phase behavior and compartmentalization properties of complex coacervates

Wet-dry cycling on the early Earth is thought to have facilitated production of molecular building blocks of life, but its impact on self-assembly and compartmentalization remains largely unexplored. Here, we investigate dehydration/rehydration of complex coacervates, which are membraneless compartments formed by phase separation of polyelectrolyte solutions. Solution compositions are identified for which tenfold water loss results in maintenance, disappearance, or appearance of coacervate droplets. Systems maintaining coacervates throughout the dehydration process are further evaluated to understand how their compartmentalization properties change with drying. Although added total RNA concentrations increase tenfold, RNA concentration within coacervates remains steady. Exterior RNA concentrations rise, and exchange rates for encapsulated versus free RNAs increase with dehydration. We explain these results in light of the phase diagram, with dehydration-driven ionic strength increase being particularly important in determining coacervate properties. This work shows that wet-dry cycling can alter the phase behavior and protocell-relevant functions of complex coacervates.


Supplementary Figures and Tables
The presence of salts was also required to fit the suggested ionic content of geothermal fields or ponds. 16 Sodium chloride was chosen as the main salt and magnesium chloride was added in anticipation of introducing ribonucleic acids, as it has been shown to enhance their function by possibly organizing water molecules around the RNA and interacting with the phosphate oxygens. [17][18][19] The starting concentrations of the main ions were around 0.4 -0.9 mM for Mg 2+ and between 1 -50 mM for NaCl. The choice of these values is explained further in Figure 1 and Note 3 below. These concentrations are equivalent to around 9.6 mg/L -21.6 mg/L (9,600 -21,600 ppb) of Mg 2+ and 23 mg/L -1,150 mg/L (23,000 -1,150,000 ppb) of Na +ranges that fall mostly within, or on a similar order of magnitude, of those present in the waters of various inland ponds as reported by Mulkidjanian et al. 16 While pH could be low in these ponds, we added a buffer, HEPES, in our experiments to ensure that the pH would not vary, placing the emphasis on the change in hydration level and the accompanying concentration effect. The pH was tested in the experiments below.
The water of primordial oceans is thought to have been more concentrated in sodium and magnesium (~0.4 M and ~0.01 M, respectively) 16 . Such locations are not, thus, relevant to the wetdry cycles studied here. Nonetheless, the literature demonstrates that coacervates can have stronger interaction strengths, depending on the chemistry of their constituents, which would translate into improved salt resistance. 20 Such stronger-interacting coacervates could be, in the future, studied in the context of wet-dry cycles occurring in tide pools or ocean water.
Supplementary Note 2: Troubleshooting the drying step and quantification of the dried mixture ( Supplementary Figures 1, 2, 3, 4) For drying, the goal was to find a manually controllable procedure that would allow the simultaneous drying of multiple samples. While initial attempts included using an oil bath with temperature control, drying directly on a hot plate, and immersing tubes in a heated water bath, the most fitting process was to use a heat block. Empty tubes and samples were weighed before and after each step. A thermometer equipped with a thermocouple was used to measure the temperature of the samples and the heater. A separate meter recorded the ambient humidity and temperature which were not found to impact the drying rate and final volumes. Supplementary   Figures 1 -3 show a series of troubleshooting experiments that were used to validate the quantification of the tubes content and the drying rates (see Methods section for more details).
A series of drying trials was performed to choose the best method to calculate the coacervate content (Supplementary Figure 1). The drying rate of a coacervate suspension prepared from the same charge concentration of polymers in different solvents was found to be identical to the evaporation rate of water (Supplementary Figure 1a). Two different quantification techniques were used to obtain the volume of the content (Supplementary Figure 1b). The first relied on visually comparing the content of the tubes with a calibration curve made with different volumes of water, where the content was quantified using the ImageJ software (Supplementary Figure 2). The second was based on converting the net weight of the tube content into volume using the density of water at room temperature. Both gave identical volumes, and the weight technique was adopted since it was more straightforward. Finally, the volumes of different coacervate compositions (with different salt and polymer concentrations) were compared to that of water at the same time points and were found to be identical which showed, experimentally, that the concentrations are low enough for a density of 1 to still be valid for weight-to-volume conversions (Supplementary Figure   1c).
The evaporation rate was shown to decrease from around 4 to 1.8 µL min −1 over the course of a range of time points where the volume decreased by a factor of ~ 10 (from ~ 500 to ~ 50 µL). This could be attributed to the conical shape of the tube which caused a reduction in the surface area of evaporation with time ( Supplementary Figure 3).
The effect of temperature on the PDADMA/PAA coacervate was evaluated in the presence of water only as well as in the presence of buffer (50 mM NaCl, 25 mM HEPES, and 4.3 mM MgCl2) in the range of 0 -95 °C. This was done by measuring the turbidity after holding the coacervate at the chosen temperature for 10 min. As it can be seen from Supplementary Figure 4, the turbidity did not change across this range of temperatures. In addition to turbidity, further measurements performed in this study support our assertion that temperature does not seem to affect the coacervate. These are: 1) volume measurements (Figures 5,Supplementary Figures 22,23), which show that the volume ratio is consistent between dried samples (at high temperature) and nondried samples; 2) weight measurements (Figure 4), which show that the same weight of the dry material is regained after multiple wet-dry cycles at high temperature.

Supplementary Note 3: Troubleshooting the buffer and MgCl2 concentrations (Supplementary Figures 5 and 6)
The compositions and drying procedure were chosen as to not lead to a HEPES and MgCl2 concentrations that surpassed 25 mM and 5 mM, respectivelyboth within the high turbidity range The heterogeneity of these observations with coacervates dried on a flat surface is fascinating; we imagine the results would be even more complex on a rocky surface. Further analyses of dry and near-dry coacervates are needed to elucidate the mechanisms behind these observations. In (a), the drying rates of 1:1 PDADMA:PAA coacervates prepared with a fixed charge concentration of each polymer of 15 mM in water, in 50 mM NaCl, and in buffer (50 mM NaCl, 25 mM HEPES, 4.3 mM MgCl2) were compared with the drying rate of water.
In (b), two methods to estimate the volume of a coacervate solution were compared: estimation of the volume by visual analysis using a calibration curve of different water volumes (see below) and estimation of the volume through dividing the weight by the density of water at the weighing temperature (room temperature ~ 23 °C). The content is 1:1 PDADMA:PAA prepared at 5 mM charge concentration, with 5 mM NaCl, 2.5 mM HEPES, and 0.43 mM MgCl2. Error bars represent standard deviations obtained from 3 trials.
In (c), the volumes of different coacervate compositions obtained from the weights are compared at the same time points. The charge concentration is shown in the legend (as "Pol"), as well as the salt concentration. All compositions contained 2.5 mM HEPES and 0.43 mM MgCl2. Supplementary Figure 9. 1:1 PDADMA/PAA complex coacervate as prepared (t = 0, left) and after allowing it to rest, without drying, for 24 hours. After 24 hours, it was resuspended, and imaged on an oligoethylene glycol-functionalized slide similar to the one at t = 0. Figure 10. Fully dried 1:1 PDADMA:PAA complex coacervate samples on microscope slides. Drying was performed directly on slides at 95 °C and the deposited mixture was imaged before and after drying, when no more liquid could be discerned with the naked eye. Coacervates were initially prepared at 15 mM charge concentration each (with respect to the monomer units) in a) 50 mM NaCl, 25 mM HEPES, and 4 mM MgCl2, b) water, and c) 50 mM NaCl. Panels show the coacervate before drying (left image) and 3 different locations on the same slide after drying (on the right, except c). In c) the first two images after drying are from different locations on the same slide while the third one (on the right, marked with an asterisk *) is from a different sample prepared on an oligoethylene glycol-functionalized slide. All scale bars are 10 µm. Figure 11. Experimental data (left) showing the turbidity phase diagram and schematic representation (right) showing a simplified representation of an approximate binodal curve and tie-lines that could describe the drying-induced movement on the phase diagram from the separate perspectives of the dilute phase and the polymer-rich phase which constitute the overall sample, located experimentally using turbidity measurements. The red line on the turbidity diagram (left) is used to define the approximate binodal or phase coexistence curve between the two-phase and one-phase regions. molecules. 21 This measurement discounted the argument that a more negative coacervate could be repelling the nucleic acid and causing K to decrease. Zeta potential measurements on a series of different charge ratios PDADMA/PAA coacervates showed that the system became neutral around 1:2 PDADMA:PAA (Supplementary Figure 16b). Such a change in behavior has been observed before with coacervate systems that contain various charged components. 22 Zeta potential at various ratios, such as 1:1 and 1:4 PDADMA:PAA, agree with previously reported measurements with the same system. 1, 2

Supplementary Note 6: Troubleshooting the whole-droplet exchange experiment with equal [U15]
We wanted to examine whether the increase in diffusion stemmed from adding U15 in increasing    Bdil). In our system, these changes do not appear to be on an equal magnitude as the overall sample concentration (Atotal → Btotal change seems to be bigger than Apol → Bpol change).
Simultaneously, revisions to the Voorn-Overbeek theory on phase separation 23 predict a tilt in the tie-lines (downward slope) which means slightly more salt could be present in the dilute phase than in the polymer-rich phase. As we demonstrate in Figure 5, the volume that the coacervate phase occupies in the total sample volume increases. These parallel changes in volume and composition of the polymer-rich phase lead to an increase in its hydration, which explains the observed partitioning and diffusion outcomes. As we mention in the main text, this is usually accompanied by self-suppression, which is shown by the decrease of the polymer concentration in the coacervate phase. We note here that these changes in the polymer-  Single-factor ANOVA test: • Among all samples: P = 0.14 (P > 0.05, difference of partitioning coefficients among cycles is not significant). F = 2.59; df1 = 5; df2 = 6.