University of Groningen Steps towards de-novo life Monreal ...chaotropes [33] to destabilize the formed structures (Table S3.1). All these strate-gies led to either amorphous aggregates

University of Groningen

Steps towards de-novo lifeMonreal Santiago, Guillermo

DOI:10.33612/diss.121581426

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Monreal Santiago, G. (2020). Steps towards de-novo life: compartmentalization and feedback mechanismsin synthetic self-replicating systems. University of Groningen. https://doi.org/10.33612/diss.121581426

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 26-04-2021

https://doi.org/10.33612/diss.121581426

https://research.rug.nl/en/publications/steps-towards-denovo-life(ab0e0a2a-1728-480a-b0be-b4d6852aa04a).html

https://doi.org/10.33612/diss.121581426

Chapter 3

Complex coacervation as acompartmentalization strategyfor synthetic self-replicators

In this chapter, complex coacervates are used to compartmentalize the peptide-based self-replicator shown in Chapter 1, and its precursors. We discuss both the possibility ofusing self-replicating fibres as part of coacervates, and their partitioning into preformedcoacervate droplets. Next, we analyse the spatial distribution, selective diffusion, and self-replication of encapsulated peptide-based macrocycles, by using liquid chromatographyand confocal microscopy. Our observations show that the crowded, charge-rich environ-ment inside of the coacervates has a strong effect on the library composition, preventingthe growth of hexamer (the main component of the library in bulk aqueous solution) andallowing for the emergence of another replicator, based on trimers of the same buildingblock. We finish the Chapter by providing an outlook on some strategies to continue thisresearch, by coupling self-replication to coacervate formation.

59

CHAPTER 3. COMPLEX COACERVATES AS COMPARTMENTS FORSYNTHETIC REPLICATORS

3.1 Introduction

Compartmentalization is one of the key problems in developing de-novo life [1–3], and it must have been a crucial step in the development of prebiotic organ-isms [4–6]. In order to synthesize biomolecules and ensure their activity, it iscritical to mantain a local high concentration, and compartmentalization is a sim-ple way to achieve that [7–9]. In the early stages of life, before organisms hadthe necessary machinery to synthesize molecules such as phospholipids, simplercompartments must have performed that function. A common hypothesis (unap-pealingly named the garbage-bag world) in the prebiotic life community states thatcoacervates, charged droplets made from relatively simple molecules, could haveplayed that role [10–12]. Whether complex coacervation took place in the first lifeforms or not, it is an important process in the current ones, as it is involved in theconcentration, segregation, and transport of biomolecules [13, 14].

Complex coacervates are produced by the interaction between positively andnegatively charged molecules (typically polymers) in water1. In the right con-ditions and with the adequate pair of molecules [17], phase separation can takeplace, leading to a new liquid phase instead of a precipitate [15, 18, 19]. Thisphase contains both ions and water, and is referred to as the coacervate phase.2

This phase behaves like a viscous liquid that is immiscible with water (despitebeing water-based itself): it can flow, remain as a layer in contact with the di-luted aqueous phase, or be dispersed into a colloidal suspension of droplets thateventually coalesce into a single layer again. Coacervates typically have both ahigher concentration of charges and a lower electric permeability than the aque-ous solution in equilibrium with them. As a consequence, both hydrophilic andhydrophobic molecules tend to be more soluble in the coacervate phase than inthe surrounding solution, and partition into it [18, 20] with distribution constantshigher than 103 [21].

The properties described above, together with the similarity between the coa-cervate phase and the crowded, protein-rich environment of cells, have led to agrowing interest in their use as protocell models [18, 21–26]. Proteins [27, 28] andnucleic acids [20, 28, 29] have been succesfully compartmentalised in coacervateswithout loss of 3D structure, and it has been shown that it is possible to havetranscription and enzymatic catalysis in them [23, 28]. Their responsiveness tosalt concentration and other stimuli [22, 30] has been used to design complex be-haviours, and to synthesise them in out-of-equilibrium conditions. Despite theirpotential as protocells, self-replication or autocatalysis has still not been showninside of them.

1There are exceptions to this: Coacervates formed from a single type of molecule [15] (known assimple coacervates), or even from molecules with the same charge [16] have also been reported.

2The terminology in this field is not completely established and varies from publication to publi-cation. In this Thesis, we refer to the dense phase containing the charged molecules as coacervate phase,and to the droplets formed upon its dispersion as coacervate droplets, or simply coacervates.

60

3.1. INTRODUCTION

We considered coacervates to be a very attractive model for the compartmen-talization of our self-replicating fibres (formed from hexamers of the buildingblock XGLKFK, 1). The peptide sequence of this building block includes twolysines, that together with the -COOH terminus add up to one positive chargeper 1 unit. This amounts to six per hexamer, and 600 in a 50 nm fibre3, whichshould be enough charge density to allow for their partition into coacervates. Pre-vious reports of coacervates mantaining the secondary structure of encapsulatedproteins, and even enhancing their self-assembly in some cases [32], encouragedus even more to study the replication of these fibres in them.

NH

O

HN

O

NH

O

NH3

HN

O

NH

O

NH3

O

O

SH

HS

Partition andselective difussion

Self-replication

1

13

16

13/14

3Considering the height of each stacked hexamer to be 0.49 nm [31]

61


3.2 Preliminary results

As shown in the introduction, the chains of 1 in 1n macrocycles are positivelycharged at pHs close to 8. Due to this, fibres of 16 can be considered supramo-lecular polycations with six charges per macrocycle, while smaller macrocycleshave considerably less charges. In general, molecules with a higher number ofcharges lead to more stable coacervates [20, 30]. Therefore, with the right oligo- orpolyanion, this difference could lead to the formation of coacervates only in pres-ence of the replicator 16 but not in presence of its precursors. The initial goal ofthis project was to find such anion, and obtain a system where the self-replicationof 16 would be directly coupled to the formation of coacervates.

Initially, ATP was used as a candidate for this purpose. Upon mixing stocksolutions of ATP and 16 at the ratio of charge neutralization the mixture becameturbid, indicating polyelectrolyte complexation. However, it quickly became evi-dent that the formed structures were not coacervates, but fibrous aggregates thatcould be observed by optical microscopy. (Figure 3.1, left). As a control, poly-lysine was mixed with ATP in the same conditions, leading to the formation ofcoacervates (Figure 3.1, right). We attribute this difference in behaviour to therigidity of the 16 fibres, which facilitates their packing into hydrogel-like struc-tures upon shielding of their surface charges. On the other hand, poly-lysine ismuch more flexible and less prone to packing into an ordered microstructure. Forthis reason, we assume that coacervate formation becomes the most stable optionfor this polycation.

(a) Fibrous structures formed from ATP + 16 (b) Coacervates formed from ATP + pLys

Figure 3.1: Optical microscopy images of the structures formed upon mixing ATP andeither 16 fibres (a) or poly-lysine (b) [21] in borate buffer (pH = 8.2, 50 mM, See Methods).The concentration of ATP was 1.0 mM, and the concentration of either 16 or poly-lysinewas 3.8 mM in monomer units. Scale bars are 50 µm.

62

3.2. PRELIMINARY RESULTS

Following this, in order to obtain coacervates using 16 and anions, we neededto increase the disorder of their microstructures. Different approaches were testedfor this purpose: using shorter replicator fibres, screening charge interactionswith NaCl, changing ATP for larger and more disordered polyanions, and addingchaotropes [33] to destabilize the formed structures (Table S3.1). All these strate-gies led to either amorphous aggregates or the complete solubilization of all struc-tures, without us finding any set of conditions where complex coacervation wasachieved.

After some exploration of the parameter space, without obtaining encourag-ing results, we decided to change the goal of the project to the compartmental-ization of replicators into coacervate droplets composed of other polyions. Ouraim was to establish whether the 16 system was compatible with coacervates, andto study the effect that compartmentalization had on it. Our long-term goal, tocouple replication and compartment formation, could be achieved in a later stagethrough a different strategy (see Further experiments).

63


3.3 Results and discussion

As shown above, when 16 macrocycles were the main polycation in a mixture,their interaction with polyanions led to the precipitation of the resulting com-plex. However, we expected that if a second polycation was present in a muchhigher concentration, the phase behaviour of the resulting polyelectrolyte com-plex would be dictated by this more concentrated polymer, and the final struc-tures would not be affected by the presence of 16. We studied this by mixing 16with a pair of polyions known in literature to form coacervates, sodium poly-acrylate (pAA) and poly-diallyldimethylammoniumchloride (pDADMAC) [34](Figure 3.2a). Coacervates were succesfully prepared in these conditions, in pres-ence of 16 (Figure 3.2b) or other 1n macrocycles (Figure S3.1), and remained stablefor hours until they coalesced into a separate phase. In order to decrease the ex-tent of this coalescence, we tried to minimise the ionic strength of the prepareddispersions. To do this, we first minimized the concentration of salts and buffersused during their preparation. Second, we used centrifugation and redispersionto remove the solution in equilibrium with the coacervate phase and substituteit for ion-free ddH2O (See Methods). The redispersed coacervates had a longerlifetime as droplets (in the order of days instead of hours), although there was ahigh variability from sample to sample.

We initially studied the compartmentalization of 1n macrocycles in the formedcoacervates by fluorescence microscopy, using thioflavin T (ThT) as a fluorophore.Thioflavin T is a β-sheet - sensitive probe [35] that becomes fluorescent in presenceof amyloid-like structures. It has been shown that 16 replicators, and to a smallerextent other 1n macrocycles, are also able to trigger its fluorescence [36]. Whencoacervates were prepared in presence of ThT and either 16 or 13/14, microscopyshowed a much higher fluorescence inside of the coacervate droplets than in thebackground (Figures 3.2b and S3.1). As expected, droplets prepared in presenceof ThT but without 1n macrocycles showed almost no fluorescence (Figure S3.3a).These observations suggested that the macrocycles partitioned into the droplets,consistent with a high affinity between the charged macrocycles and the charge-rich environment of the coacervate phase.

Interestingly, while the fluorescence in the case of 13/14 was homogeneousthroughout the droplets, the 16 fibres seemed to aggregate inside of them, form-ing patterns of different fluorescence intensities (Figure 3.2b). Besides that, thedistribution of the macrocycles appeared to be similar from one droplet to an-other, as long as the macrocycles were present in solution before coacervate for-mation. However, when 16 was added to a preformed dispersion of coacervates,its distribution was irregular, suggesting slow diffusion in, out and through thecoacervate phase (Figure S3.2).

To study the partition of 1n macrocycles quantitatively, samples of coacervateswere prepared in presence of the different components of the system and cen-

64

3.3. RESULTS AND DISCUSSION

N

Cl

n

O O

Na

S

N

N

Cl

n

pDADMAC pAA

ThT

(a) Compounds used forcoacervate formation and imaging.

(b) ThT fluorescence of coacervatesprepared in presence of 16.

T h T 1 1 6 1 3 / 1 4

0 , 0

0 , 5

1 , 0

Norm

alize

d Csu

perna

tant

(c) Depletion of the supernatant inpresence of coacervates.

Component KD (ccoa/csup)

ThT (2 ± 7)*100

1 (2.2 ± 0.6)*102

16 (2.8 ± 0.8)*105

13/14 (2.5 ± 0.5)*105

(d) Partition constants intocoacervates.

Figure 3.2: (a) Structures of the compounds used in this section. (b) Fluorescence mi-croscopy image of pAA/pDADMC coacervates (50 mM each, in monomer units) preparedin presence of ThT (80 µM) and 16 (40 µM). (c) Concentration of different species in the su-pernatant in presence of coacervates, normalized against controls without them. The totalconcentrations were 50 mM for pAA and pDADMAC (in monomer units), 80 µM for ThT,200 µM for 1 and 400 µM for either 16 or 13/14. (d) Distribution constants calculated fromthe data in (c). See Methods for the calculation of errors.

trifuged to separate the coacervate phase from the bulk solution. The concentra-tion of the different species in the supernatant was then studied by UPLC, com-paring it with control samples that did not contain coacervates (Figure 3.2c). Weobserved that, for 16 and 13/14, the supernatant was completely depleted in pres-ence of coacervates. This was extremely remarkable, specially considering thesmall volume of the coacervate phase (1.5 ± 0.2 % of the total solution). A dis-tribution constant was then calculated for each of the components (Figure 3.2d),

65


showing particularly high values for 16 and 13/14.4 These results, as reportedpreviously for smaller molecules [23], seem to follow a trend where entities withmore charges are partitioned to a larger extent into the coacervate phase. As anestimate of the magnitude of this compartmentalization: in a typical sample con-taining 16 with a total concentration of 80 µM, its concentration in the coacervatephase would be higher than 5 mM.

The fact that the partition of ThT into the coacervates was minimal encour-aged us to continue using this dye as a probe. Since ThT was present in similarconcentrations inside and outside of the coacervates, the intensity of its fluores-cence should depend only on the concentration of 1 derivatives. We confirmedthis by confocal microscopy, studying images of coacervates with increasing con-centrations of either 16 or 13/14. As shown in Figure S3.3, in the range of concen-trations studied the fluorescence intensity was directly proportional to the con-centration of 1n in the droplets. It should be noted that the average fluorescenceintensity did not change upon increasing the concentration of ThT (Figure S3.3d),confirming that its concentration was high enough to bind all 1n macrocycles.

Next, we studied the exchange of 1n macrocycles between different droplets.The partition of solutes into the coacervate phase is a well-established pheno-menon, but there is still some controversy in the literature about whether thediffusion of such solutes in and out of the droplets is a slow (longer than days) orfast (shorter than minutes) process, with a large variation between very similarsystems [37, 38]. The rate at which this exchange takes place is critical for the syn-thesis of de-novo life: if it happens too fast, the molecules from different dropletscan diffuse together and spatial separation is impossible. On the other hand, ifthe process is too slow, the system cannot show behaviours that require interac-tions between different droplets (such as cooperation, parasitism or competition)and is "only" a collection of independent microreactors running in parallel. Inorder to study exchange in our system, we prepared populations of coacervatescontaining ThT and either 16, 13/14, or nothing else. We then mixed the "empty"coacervates with the droplets containing macrocycles, resulting in a dispersionwhere each coacervate contained either a high concentration of macrocycles orno macrocycles at all. In this case, instead of redispersing the droplets as in pre-vious experiments, we deposited a layer of the dispersion on a glass slide. Wemade this layer thin and diluted so, practically, no droplets would be touchingeach other. Therefore, the only possible mechanism for exchange of material be-tween them was diffusion.

This sample was then studied by confocal microscopy, measuring the averagefluorescence of a number of droplets and studying the change of the resulting dis-

4In order to confirm that these values were not artificially high due to a kinetic trap, we preparedcontrols adding 13/14 or 16 to the supernatant after the coacervates had been formed and centrifuged.These samples reached the same final composition as the ones in which the macrocycles had beenadded first. Therefore, we can conclude that the two phases were equilibrated, and the KD valuesreported correspond to equilibrium constants.

66


01 02 0

n = 1 9 1

Numb

er of

drople

ts

0 ha v g = 1 0 2 1

01 02 0

2 h

n = 1 7 3

n = 1 8 2

a v g = 1 2 4 0

01 02 0

3 h

n = 1 5 5

a v g = 1 0 5 8

0 1 0 0 0 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 00

1 02 0

A v e r a g e f l u o r e s c e n c e i n t e n s i t y( c o u n t s / p x )

2 4 ha v g = 1 2 1 9

(a) Diffusion of 16

01 02 0 0 h

n = 1 2 1a v g = 9 1 8

n = 1 5 6a v g = 5 2 1

n = 5 8a v g = 8 2 3

n = 7 4a v g = 3 1 6

02 55 0

2 h

07

1 4

3 h

0 1 0 0 0 2 0 0 0 3 0 0 0 4 0 0 0 5 0 0 00

1 53 0

A v e r a g e f l u o r e s c e n c e i n t e n s i t y ( c o u n t s / p x )

2 4 h

(b) Diffusion of 13/14

Figure 3.3: Exchange of 1n macrocycles between coacervates. The histograms represent thedistribution of average fluorescence intensities in coacervates containing ThT and either16 (left) or 13/14 (right) after mixing them with coacervates containing only ThT. The con-centrations used for the preparation of the coacervates were 40 µM for 1n macrocycles andThT, and 25 mM for pAA and pDADAMAC. Immediately after mixing the coacervatescontaining 1n and ThT with the ones containing only ThT, an aliquot of each sample wasdiluted 4 times, placed in a covered glass slide at room temperature and imaged over timeusing confocal microscopy. Only the coacervates with an average fluorescence intensityhigher than a threshold of 50 counts/px are shown.

tribution over time (Figure 3.3). Our aim was to conclude from this observationwhether diffusion between droplets was taking place: if the macrocycles werebeing exchanged, the distribution would shift to lower values as their concentra-tion would average out between the "full" and "empty" droplets. In the case of 16,this was not the case, as the distribution did not change significantly over time(Figure 3.3a). However, for 13/14 we observed that the distribution was slowlychanging, as both the number of droplets with high concentration and the aver-age fluorescence were decreasing (Figure 3.3b). A Spearman rank-order correla-tion was run to confirm these relationships. In this statistical analysis, rs valuesclose to 0 indicate no correlation between two variables, and positive or negative

67


values indicate a correlation of the same sign (stronger the closer that rs is to +1or -1). We found no correlation between time and fluorescence for 16 (rs = 0.060,p = 0.11), and a negative correlation between time and fluorescence for 13/14 (rs= -0.476, p < 0.0001). This is a clear indication that, while 16 fibres remain trappedin coacervate droplets, small macrocycles can be exchanged between them. Thisopens the door to the preparation of systems with different "species" of replica-tors, which could coexist in independent droplets while having an influence oneach other through the exchange of small macrocycles - in example competing forthem.

The results so far focus on the individual components of the 1n system. How-ever, this study would not be complete without addressing the dynamics of thesystem and how different macrocycles convert into one another. We used UPLCanalysis for this purpose, measuring over time the composition of libraries of 1prepared in presence of coacervates. This allowed us to study the relative concen-trations of the different components of the system, however it had a limitation:In the conditions of this experiment, the coacervates slowly sediment towardsthe bottom of the vial, which results in a constant loss of peak area in UPLC thatcan be partially recovered upon sonication. Since oxidized species partition com-pletely into the coacervate phase, the peak area of all of them decrease to thesame extent through this process, and we can expect that their relative amounts,as observed by UPLC, correspond to the overall ones. However, 1 is significantlypresent both inside and outside of coacervates (Figure 3.2d), so the observed ratiobetween it and 1n changes as sedimentation progresses (For a clear example, seeFigure S3.4a). For this reason, the percentage of 1 has been omitted for most of thecalculations in this study. The graphs where it is plotted must only be interpretedqualitatively.

Two unexpected behaviours were observed in 1n libraries inside of coacer-vates: An accelerated oxidation of thiols to disulfides, and a change in the final li-brary composition, going from almost exclusively 16 to almost exclusively 13 (Fig-ures 3.4a and 3.4b). Controls in presence of both polyions (Figures 3.4d and 3.4c)showed that the accelerated oxidation was also taking place in presence of pDAD-MAC alone. To investigate the nature of this process, we prepared a sample inpresence of coacervates and, after it had been completely oxidized, we reducedit partially using TCEP. After the reduction was completed, the rate of oxidationwas not enhanced anymore (Figure S3.4a). These observations would fit a sto-chiometric reaction between 1 and an impurity in the pDADMAC stock solution.We initially suspected of tert-butylperoxide (a common initiator for pDADMACpolymerization [39]), but peroxides were not detected in the pDADMAC stocksolution using a standard test. Further research will need to be done in order toidentify the impurity or phenomenon responsible of this behaviour, and discernif other polycations will have the same effect too.

68


0 2 0 4 0 6 00

5 0

1 0 0 1 1 4 1 6 1 3

% of

1 n mac

rocycl

es

T i m e ( h )(a) Library in presence of

coacervates.

0 2 0 4 0 6 00

5 0

1 0 0

% of

1 n mac

rocycl

es 1 1 4 1 6 1 3

T i m e ( h )(b) Control without polyions.

0 2 0 4 0 6 00

5 0

1 0 0 1 1 4 1 6 1 3

% of

1 n mac

rocycl

es

T i m e ( h )(c) Library in presence of pAA.

0 2 0 4 0 6 00

5 0

1 0 0

1 1 4 1 6 1 3

% of

1 n mac

rocycl

es

T i m e ( h )(d) Library in presence of

pDADMAC.

Figure 3.4: Effect of coacervates and their components in 1 libraries. The libraries wereset up by mixing stock solutions of 1, pAA and/or pDADMAC and diluting them withddH2O to final concentrations of 200 µM for 1 and 25 mM for pAA and pDADMAC.The libraries remained unstirred at 40 °C, and were periodically sonicated before UPLCanalysis. See above for an important note on the calculation of percentages.

The emergence of 13 as the main component of the library made us suspectthat this macrocycle was also behaving as a self-replicator. Previous studies inour group have shown that, in conditions where 16 is not stable (such as highconcentrations of denaturants), 13 can emerge as a self-assembled self-replicatingmacrocycle [40]. Interestingly, we observed that the environment of the coacer-vate phase was having a similar effect in our system: preformed seeds of 16 werenot growing inside of the coacervates (Figures 3.5a and S3.5), but 13 was, andits growth was autocatalytic (Figure 3.5b). Negative stain electron microscopy ofsamples of 13 formed in this way showed fibrous structures (Figure 3.5c), con-sistently with other self-assembled self-replicators based on 1 and similar build-

69


ing blocks. Since 13 was also growing after partial reduction of the library (Fig-ure S3.4b), where disulfide exchange was possible for longer times, we concludethat this replicator is not a kinetic trap, but the thermodynamic product of 1nlibraries in the coacervate phase.

0 1 2 3 4 50

2 0

4 0

% of

1 n mac

rocycl

es

T i m e ( h )

1 4 1 6 1 3

(a) 1 and 13/14 seeded with 16

0 2 0 0 4 0 00

5 0

1 0 0

��% of

1 3

��

��

��

(b) 1 seeded with 13 (c) Electron micrograph of 13

Figure 3.5: Effects of the coacervate phase in 1n replicators. (a) Evolution of 1 libraries,(200 µM, 50% oxidized using NaBO3), seeded with 16 (5 % of the total concentration). Thedata shown is the average of three repeats, and the error bars show the standard deviation.(b) Growth of 13 in libraries of 1 (200 µM), seeded with coacervates containing preformed13 (33% of the total volume), and without them. In both cases, the libraries were kept at40 °C without stirring and sonicated before each injection. (c) Supramolecular assembliesof 13, observed by negative stain TEM upon destruction of the coacervates by addition of0.05 % TFA. The scalebar is 50 nm

70

3.4. CONCLUSIONS

3.4 Conclusions

The goal of this project was to determine whether self-replication could be com-bined with compartmentalization in coacervate droplets. This encapsulation strat-egy appears promising: coacervates are easy to prepare, chemically compatiblewith our system, and due to the high charge density of both coacervates andmacrocycles, encapsulation is simple and almost complete.

However, the compartmentalization of our system comes with a significantnumber of challenges too. As we have seen in this chapter, the behaviour oflibraries in presence of coacervates is completely different than without them,leading to the formation of another self-replicating species. This is perhaps un-surprising, due to the charge-rich and crowded nature of the coacervate phase.Since the behaviour of our system relies so much on mechanical processes andinteractions between library members (See Chapter 2), it is understandable that achange in environment of this magnitude would greatly affect it. We have shownthat it is possible to achieve self-replication inside of coacervates, but the self-replicating species is not the one that we had predicted. This is an interestingprocess by itself (self-replicating species that change in different environmentsare not common [41]), but an obstacle for the design of systems that use coacer-vates as a compartment. As the results of this Chapter show, the properties ofreplicators in bulk solution cannot be directly translated to the coacervate phase,and therefore any studies performed in bulk would need to be repeated in coac-ervates before designing any further experiments.

The analysis of coacervates is also far from trivial. Confocal microscopy hasbeen a useful analytical tool for these systems: it allows for the analysis of eachdroplet separately, and it can be performed in a thin film of solution, which pre-vents coalescence. It is also a fast technique, and the analysis of its output is easyto automatize, allowing for a fast analysis of large populations of coacervateswith minimal input. However, it is limited by our current fluorescent probe, as itonly provides one single channel of information: its fluorescence intensity. There-fore, it does not allow us to distinguish between the different species in a library,or between building blocks with a different chemical composition.

UPLC has the opposite strengths and weaknesses. It does give informationabout the exact composition of the library at each moment, and it can distinguishbetween replicators with different chemical structures, but it does not distinguishbetween droplets. Furthermore, it requires preparing samples in a larger volume,and as the droplets sediment and coalesce the analysis loses reliability. A com-bination of both techniques was enough to study the single-component systemdescribed in this Chapter, but for more advanced studies we will need to devisenew analytical solutions.

A promising analytical technique to study this system could be confocal mi-croscopy combined with more elaborate fluorescent probes, that can discriminate

71


between different macrocycles and building blocks through the combination ofseveral fluorophores. These probes have been described in the literature for amy-loid proteins [42], and our group is working towards their implementation in oursystems.

Despite the challenges, significant progress has been made towards our finalgoal. We have confirmed the high efficiency of the encapsulation of replicatorsinside of coacervate droplets, and the fact that they are still able to bind and ac-tivate ThT is an indication that their structure is, to some extent, conserved. Wehave also described how these coacervates can discriminate between large self-assembled macrocycles, and smaller ones, retaining the first but allowing for thediffusion of the others. We foresee that this will be critical in future experiments,where replicators in different droplets would be able to interact with each otherthrough the exchange of material, but not directly mixing and forming interme-diate species.

Lastly, we have described for the first time self-replication inside of coacer-vates. Besides the milestone that this represents in the de-novo synthesis of life,the self-replicator discovered in this work is also worthy of further investigationby itself: Can it also be obtained at room temperature? If its formation is triggeredby the coacervate environment, can it be formed in other coacervate systems orsimply in crowded solutions with high ionic strength? In the seeding experiment,coacervates with seed were mixed with coacervates without it and an effect wasobserved: is this due to the diffusion of the replicator, the diffusion of the pre-cursors, the diffusion of both, or simply to coalescence? The work in this chapterlays the groundwork for the combination of self-replication and compartmen-talization, but more characterization and mechanistic studies will be required tounderstand this phenomenon completely and bring it to the next level.

72

3.5. FURTHER EXPERIMENTS

3.5 Further experiments

As we said in the introduction, the ultimate goal of this project is to couple self-replication and coacervate formation. Since this could not be done directly (SeePreliminary results), another connection between these two properties needed tobe developed. With this goal in mind, we protected the tetraamine sperminewith 2-sulfo-Fmoc groups ("Smoc", see Figure 3.6a). Past research in our grouphas shown that self-replicators of the type of 16 are able to deprotect amines thathave been protected with this motif [43, 44]. Coacervate formation with sperminehas also been described [29], and we were able to prepare coacervates in our labby mixing spermine and pAA (not shown).

Based on these observations, we envisioned a system where replicator - medi-ated deprotection of Smoc-spermine, followed by polyelectrolyte complexationwith a suitable polyanion, would lead to the formation of coacervates as self-replication progressed.

Our results showed that 16 was indeed able to deprotect spermine. However,coacervate formation could not be achieved in these conditions, as the deprotec-tion was inhibited by the presence of pAA (Figure 3.6b). Nonetheless, this doesnot necessarily mean that the reaction is incompatible with polyanions (Smoc-spermine is an oligoanion itself), so a different anion could show the desired be-haviour.

H3NH2N

NH2

H3N

SmocNSmocN

NSmoc

SmocN Smoc =O

O

O3S16

O3S

+ 4 x

(a) Proposed reaction for the16-mediated synthesis of spermine

0 4 5 0 9 0 00 , 0

0 , 5

1 , 0

Abs 3

08 nm

(a.u.

)

T i m e ( m i n )

b l a n k 1 6 + p A A 1 6

(b) Deprotection of spermine inpresence and absence of pAA

Figure 3.6: Strategy for a replicator-mediated synthesis of coacervates. (a) Reaction catal-ysed by the replicator and compounds involved. (b) Spermine deprotection in presenceand absence of pAA, monitored by UV-Vis. The absorbance at 308 nm corresponds to 2-sulfo-dibenzofulvene, a side-product of the deprotection. The concentrations used were100 µM for Smoc-spermine and 40 µM for both 16 and pAA.

73


3.6 Materials and methods

All reagents, solvents, and buffer salts were purchased from commercial sourcesand used without further purification. Building block 1 (XGLKFK) was obtainedfrom Cambridge Peptides Ltd (Birmingham, UK). pDADMAC and pAA wereobtained from Sigma-Aldrich, pDADMAC as a 20% (w/w) solution in water(molecular weight and purity not analysed), and pAA as a sodium salt (8-14 %of water as an impurity, average Mw ∼ 5100). For both polymers, the concen-trations used in the text refer to their monomer residues. ThT was also obtainedfrom Sigma Aldrich (dye content ≥ 65%). Image analysis was performed withImageJ (Using the Fiji distribution [45]). Statistical analysis was performed withOriginPro 8 (OriginLabs, USA). 1H-NMR was performed using a 400 MHz spec-trometer.

UPLC analysis

UPLC analysis was performed on a Waters Acquity UPLC H-class, equipped witha PDA detector. All analyses were performed using a reversed-phase UPLC col-umn (Aeris Peptide 1.7 µm XB-C18 x 2.10 mm, Phenomenex). The column tem-perature was kept at 35 °C, and the sample plate was kept at 25 °C, unless other-wise specified. UV absorbance was monitored at 254 nm. For each injection, 10µL of sample was injected.

The following solvents and gradient were used for UPLC analysis. The flowrate was kept at 0.3 ml/h.

Time(min) % MeCN /0.1 % TFA

%H2O/0.1 % TFA

0 10 90

1 10 90

1.3 25 75

3 28 72

11 40 60

11.5 95 5

12 95 5

12.5 10 90

17 10 90

74

3.6. MATERIALS AND METHODS

Buffer preparation

Two different buffers were used in this project, both at pH 8.2 and based on borateions, but each of them with different ionic strengths. For the Preliminary results,we used borate buffer with a total concentration of 50 mM in boron atoms. For therest of experiments, borate buffer with a total concentration of 200 mM in boronatoms was used. The low concentration borate buffer (50 mM) was preparedfrom B2O3, and the high concentration borate buffer (200 mM) was prepared fromNa2B4O7 · 10 H2O, both purchased from Sigma-Aldrich (Purity ≥ 99.5%)

Preparation of 16

Stock solutions of 16 were typically prepared by dissolving building block 1 inborate buffer to a final concentration of 4.0 mM, and stirring the library at 40 °C.The library was monitored by UPLC analysis until completion.

Preparation of 40 nm 16 fibres

16 fibres were mechanically sheared using a modified protocol previously pub-lished by our group [46]. A 150 µL aliquot of a previously prepared solution of 16(4.0 mM) was placed in a Couette cell (Rcup = 20.25 mm, Rbob = 20.00 mm, aver-age radius (R) = 20.125 mm). The sample was mechanically sheared by rotationof the inner cylinder, with a frequency of 4000 rpm (corresponding to a shear rateof 33702 s-1). The resulting fibres were used within 48 h of preparation.

Preparation of 13/14

Stock solutions of oxidized 1 libraries, named "13/14" in the text by their two maincomponents, were typically prepared by dissolving building block 1 in boratebuffer, adding 1 equivalent of NaBO4, and diluting the library to a final concen-tration of 4.0 mM in residues of 1. The libraries were monitored by UPLC untilcompletion, which was typically done in less than 1 hour.

Polyelectrolyte complexes using 16 and polyanions

Stock solutions of 16 or 13/14 (3.8 mM) were directly mixed with stock solutionsof the different polyanions and chaotropes in the concentrations indicated in thetext. Stock solutions were prepared in borate buffer (pH = 8.2, 50 mM in boron),and doubly distilled water was used for further dilutions. To ensure that theobserved structures were the thermodynamic products in the samples containingadditives, either the additive was added first or the sample was submerged inan ultrasound bath for at least 1 minute before observation. All images wererecorded using a Nikon Coolpix 5000 camera attached to a Nikon eclipse TS100

75


microscope, with a 40x magnification. The scalebar was set by comparison witha reference slide.

General procedure for the preparation of pAA/pDADMACcoacervates

Stock solutions of pDADMAC, pAA, 1n and ThT were mixed in the desired ratiosfor coacervate formation. Except for the 1n stock (which was prepared in boratebuffer, pH = 8.2, 200 mM in boron), ddH2O was used as a solvent. The volumesadded were calculated for concentrations of 50 mM for pAA and pDADMAC, 80µM for ThT and 0-80 µM for 1n macrocycles. Unless otherwise specified in thetext, the addition order was ddH2O, pDADMAC, ThT, 1n, and pAA, homogeniz-ing the solution before adding pAA.

Typically, samples were centrifugued (RCF = 3969×g, 5 minutes) after coa-cervate formation, the supernatant was carefully decanted, ddH2O was added tothe desired concentrations, and the samples were redispersed using a standardvortex mixer for 1 minute.

Fluorescence and confocal microscopy

Fluorescence and confocal microscopy experiments were performed using a Mi-croTime 200 setup (PicoQuant, Germany), using an Olympus IX73 microscopewith a 100x oil immersion objective. For the fluorescence images, a standard 120W mercury lamp (X-Cite®series 120 Q) was used as an irradiation source, to-gether with a fluorescence cube with exciter and emitter wavelengths of 466 ± 20and 525 ± 25 nm, respectively (Semrock GFP-4050B). For confocal images, a 440nm laser was used together with a 10%R/90%T Beamsplitter, an interference fil-ter with wavelengths of 482 ± 35 nm, and a standard Single-Photon CountingModule (Excelitas Technologies, USA). The laser power remained under 10 µW,and was kept constant for each series of experiments.

Glass slides were coated with PEG to prevent interactions with the coacer-vate phase, using a protocol modified from [47]: 3-[Methoxy(polyethyleneoxy)-propyl]trimethoxysilane (6-9 PE-units, ACBR GmbH) (0.5% v/v) and acetic acid(1% v/v) were dissolved in ethanol to make the coating reagent. 50 µL of thissolution was placed on top of each microscope slide, placing another slide on topof it to spread the solution through both surfaces. More slides and droplets ofcoating reagent were put on top of each other in this way, and then heated in anoven to 100 °C for at least 30 minutes. The slides were then submerged in ddH2O,sonicated, rinsed with more ddH2O, dried, and stored for use.

76


Partition constants of 1n macrocycles in pAA/pDADMACcoacervates

To calculate the volume of the coacervate phase, solutions of pDADMAC andpAA were mixed to a final concentration of 50 mM and volume of 15 mL. Thesesamples were then centrifuged, and the volume of the supernatant was then mea-sured using a standard micropipette (error = 1 µL). The volume of the coacervatephase was then calculated by difference with the original volume, and the datareported is the average and standard deviation of four independent repeats.

pAA/pDADMAC coacervates were then prepared following the general pro-cedure, but after centrifugation the supernatant was collected and analysed byUPLC. The total concentrations used were 80 µM for ThT, 200 µM for 1, and 400µM for 1n oxidized macrocycles. Three repeats and three controls were preparedfor each of the components studied (in the controls, ddH2O was added instead ofpolyions). Aliquots of these controls were taken close to the surface immediatelyafter centrifugation, and analysed by UPLC.

The concentration in the supernatant of each component was calculated bycomparison of its average peak area with the average peak area of its controlsamples, and the average concentration in the coacervates was calculated fromthe difference between this value and the total concentration. Finally, the distri-bution constant was calculated for each component as the ratio between its con-centration in the coacervate phase and its concentration in the supernatant. Anerror value was calculated for each of the measured magnitudes (peak area in thesupernatant, peak area in the controls, coacervate volume) as the standard devia-tion of the three repeats, and the error reported for KD was calculated combiningthose values.

In order to determine whether the 13/14 and 16 samples had reached equi-librium, two controls were prepared following the same procedure, but addingeach macrocycle after the coacervate phase had been formed and centrifuged.These samples were left to equilibrate at room temperature for 20 days, followedby UPLC analysis of their supernatant. For both controls, the peak area valuescorresponding to 1n were in the same range as for the the samples shown in thetext.

Validation of ThT as a fluorescence probe for 1n

Coacervates were prepared following the general procedure, containing increas-ing amounts of either 16 or 13/14. After removal of the supernatant, ddH2O wasadded so the final concentrations of pAA and pDADMAC were 25 mM and the fi-nal concentration of 1n was 0-40 µM. Aliquotes of each sample were deposited onpretreated glass slides, and measured by confocal microscopy. The droplets wereselected manually and their average fluorescence (counts/px) was measured us-ing ImageJ.

77


Diffusion of 1n between coacervate droplets.

Coacervates containing ThT (40 µM) and either 16, 13/14, or nothing else, wereprepared following the general procedure, excluding centrifugation and redisper-sion. The coacervate dispersion containing only ThT was mixed with each of theother two separately, and an aliquote of the mixture was diluted 4 times, placedin a pretreated glass slide and covered using a SecureSealTM imaging spacer (9 x0.12 mm, Grace Bio-Labs, USA) and a cover slip. Each sample was kept at roomtemperature and studied periodically using confocal microscopy, recording threedifferent pictures in different spatial regions for each time point in each sample.Three micrographs in total (one corresponding to 2h of diffusion of 16, the othertwo corresponding to 3h of diffusion of 13/14) showed unusually high noise, andwere discarded before analysis.

The remaining micrographs were then analysed automatically using an Im-ageJ script (Appendix A) to detect each droplet and measure its average fluores-cence. The process was initially optimized, changing the circularity and thresh-old parameters, to ensure that no artifacts were measured. Then, once that aworking set of parameters was found, all images were analysed using the samescript. The coacervates not containing macrocycles (average fluorescence < 50counts/px) were not reliably detected neither by the program nor manually, sothey were removed before analysis.

Dynamic combinatorial libraries of 1n in coacervates analysed byUPLC

Libraries of 1 and its derivatives were prepared in presence of coacervates fol-lowing the general procedure, excluding centrifugation and redispersion. Thelibraries were kept at 40 °C inside of the UPLC sample holder, and they weresonicated for 30 seconds before injection. 10 µL of sample were injected directlyfrom the dispersion for analysis. The controls with individual polymers and with-out polymers were kept in exactly the same conditions and sonicated before eachanalysis as well.

The percentages shown in the figures represent the fraction of each of the mainspecies of the library (1, 13, 14 and 16) over the total peak area (integrated between6 and 9 minutes). For the figures where 1 is not plotted, the percentages representthe fraction of each component over the total peak area, minus the peak of 1 (7 to9 minutes). All the peaks shown were assigned by comparison of their retentiontimes with previous data.

The pH of all samples was measured upon preparation and after equilibriumhad been reached, by adding a drop of sample to pH-paper. Despite the lowconcentration of buffer, the pH did not change significantly in any experiment,remaining close to 8.

78


For the seeding experiments with 16, the seed and a solution of 1 and 13/14(that had been previously oxidized using NaBO3 to the desired ratio) were mixedimmediately before coacervate formation. For the seeding experiments with 13,a coacervate dispersion containing only 1 was mixed with one containing only13 in a proportion 2:1 (v/v). The control samples consisted only of coacervatescontaining 1.

Quantofix® peroxides test strips (Sigma-Aldrich) were used to determine thepresence of peroxides in the stock solution of pDADMAC. The concentration ofperoxides was determined to be lower than the sensitivity of the scale (1 mg/mL)in the 20% w/w stock solution. Lower concentrations than that are not enoughto oxidize 1 in the concentrations used in the experiments of this chapter.

Negative staining transmission electron microscopy of 13

An aliquot of a coacervate dispersion containing 13 was mixed in a 1:1 (v/v) ra-tio with a 0.1% v/v aqueous solution of TFA. Immediately after mixing, a smalldrop (5 µL) of the resulting solution was deposited on a 400 mesh copper gridcovered with a thin carbon film (Ted Pella Inc., USA). After 30 s, the droplet wasblotted on filter paper. The sample was then stained twice (5 µL each time) with asolution of 2 % uranyl acetate, by depositing a drop of it on the grid and blottingit on the filter paper after 30 s. The grids were observed in a Philips CM12 cryoelectron microscope operating at 120 kV. Images were recorded on a slow scanCCD camera.

Synthesis of Smoc-Spermine

9-(2-Sulfo)fluorenylmethyloxycarbonylchloride was synthesized following a pub-lished procedure [48] (Figure S3.6).

In a 25 mL round bottom flask equipped with a magnetic stirring bar, 9-(2-sulfo)fluorenylmethyloxycarbonylchloride (303 mg, 0.9 mmol) was dissolved in1.6 mL of 1,4-dioxane. Then, spermine (37 mg, 0.18 mmol) was dissolved in 1.3mL of a 10% aqueous Na2CO3 solution and transferred dropwise to the reactionflask, under cooling with a water bath. The reaction was stirred at room temper-ature and followed by TLC, using a 1:4 mixture of MeOH:DCM. After 4 hours ofstirring the mixture was then acidified to a pH of 2-3 with HCl (1 M), resultingin a yellow solution which was purified by reverse-phase flash chromatographyusing a H2O:CH3CN gradient. 1H-NMR in D2O showed only the pure product(Figure S3.7). 1H NMR (400 MHz, D2O) δ 8.09 - 7.51 (m, 8H), 7.46 – 6.50 (m, 20H),4.52 - 3.82 (m, 8H), 3.82 - 3.42 (m, 4H), 3.14 - 1.60 (m, 12H), 1.60 – 0.00 (m, 8H).

79


Deprotection of Smoc-spermineStock solutions of Smoc-spermine and pAA were mixed in a quartz cuvette (HellmaGmbH & Co. KG, germany), diluted using ddH2O, and equilibrated for 20-30minutes. Right before the measurement started, 16 was added. The final con-centrations were 100 µM of Smoc-spermine and 40 µM of 16 and pAA. The ab-sorbance at 308 nm was monitored over time by using a UV-spectrophotometerV-650 (Jasco Benelux B.V., The Netherlands).

80

3.7. SUPPLEMENTARY MATERIAL

3.7 Supplementary material

Macrocycle Anion Additive Result

16 (3.8 mM) ATP (1.0 mM) Fibrous structure

13/14 (3.8 mM) ATP (1.0 mM) Amorphous aggregate

16 (40 nm, 3.8 mM) ATP (1.0 mM) Amorphous aggregate

16 (40 nm, 0.38 mM) ATP (0.10 mM) GdmCl (1.0 M) No structures

16 (40 nm, 0.38 mM) ATP (0.10 mM) Ca(OAc)2 (1.0 M) No structure

16 (40 nm, 0.76 mM) ATP (0.20 mM) KBr (1.0 M) Amorphous aggregate

16 (40 nm, 0.76 mM) ATP (0.20 mM) Urea (1.0 M) Amorphous aggregate

13/14 (3.4 mM) ATP (0.97 mM) GdmCl (0.90 M) No structures

16 (2.2 mM) ATP (0.62 mM) GdmCl (3.6 M) Amorphous aggregate

16 (1.9 mM) ATP (0.55 mM) GdmCl (4.2 M) No structures

16 (1.9 mM) ATP (0.5 mM) NaCl (0.50 M) Amorphous aggregate

16 (1.9 mM) ATP (0.5 mM) NaCl (1.0 M) No structures

16 (1.9 mM) BSA (2.0 g/L) Fibrous structure

16 (3.8 mM) DNA 1 * Gel formation

16 (3.8 mM) DNA 2 * Fibrous structure

16 (3.8 mM) Κ-casein * Fibrous structure

16 (3.8 mM) α-casein * Fibrous structure

16 (3.8 mM) Hyaluronic acid * Gel formation

Table S3.1: Screening of conditions for the formation of coacervates in presence of 16 andits precursors. *Instead of calculating the point of charge neutralization for these anions,aliquotes of stock solutions containing them were mixed with 16 until the solution becameturbid. DNA 1 = DNA from salmon testes. DNA 2 = DNA from calf thymus.

81


Figure S3.1: Optical (left) and fluorescence (right) microscopy images of pAA/pDADMAC(25 mM) coacervates prepared in presence of ThT (40 µM) and 13/14 (40 µM). The contrastwas increased to facilitate observation.

Figure S3.2: Optical (left) and fluorescence (right) microscopy images of pAA/pDADMAC(50 mM) coacervates prepared in presence of ThT (40 µM), to which 16 (80 µM) was added.The fluorescence signal is still localized inside of the droplets, but it is much less homoge-neous that in the droplets of Figure 3.2b

82


(a) Confocal images of coacervates in presence of increasing concentrations of 1n

0 2 0 4 00

7 0 0

1 4 0 0

Avera

ge flu

oresce

nce i

n dr

oplet

s (co

unts/

px)

[ 1 6 ] ( µM )(b) Linear relationship

between 16 concentration andThT fluorescence.

0 2 0 4 00

5 0

1 0 0

1 5 0

[ 1 3 / 1 4 ] ( µM )(c) Linear relationship

between 13/14 concentrationand ThT fluorescence.

4 0 8 00 , 0

0 , 5

1 , 0

[ T h T ] ( µM )

Norm

alized

flu

oresce

nce i

n drop

lets

(d) An excess of ThT does notlead to higher fluorescence.

Figure S3.3: Validation of ThT fluorescence as a method for measuring the concentrationof 1n in coacervates. (a) Confocal micrographs of pAA/pDADMAC coacervates (25 mMin monomer units for each polymer) in presence of ThT (40 µM) and growing amountsof 16 (top images) or 13/14 (bottom images). The black-white scale in the bottom pictureswas increased by a factor of 10 to aid visualization. The average fluorescence intensityinside of the droplets is plotted against the initial concentration of 16 or 13/14 in (b) and(c), respectively. Panel (d) Compares the average fluorescence of two coacervate samplescontaining 16 (40 µM) and different initial concentrations of ThT. The data shown in panels(b) to (d) is the average and standard error of the mean of 15-20 droplets for each concen-tration. We calculate the concentrations of 1n assuming that they partition completely tothe coacervate phase (Figure 3.2d) and therefore no material is lost upon redispersion.

83


0 2 4 60

5 0

1 0 0

T i m e ( h )

1 1 4 1 6 1 3

% of

1 n mac

rocycl

es

(a) Oxidation of 1 in coacervatesafter reduction with TCEP

0 2 4 60

2 5

5 0

1 4 1 6 1 3

T i m e ( h )

% of

1 n mac

rocycl

es

(b) Evolution of 1n macrocyclesin coacervates after reduction

Figure S3.4: Behaviour of a library in presence of coacervates after reduction. The libraryshown in both (a) and (b) was prepared by mixing 16 (50 µM), a 50% oxidized mixture of1 and 13/14 (200 µM), pAA and pDADMAC (25 mM each). Right after the first measure-ment, TCEP (40 µM) was added to the library to reduce it partially. The library is the samein both panels, but (a) includes 1 in the graph and takes it into account for calculations,and (b) only includes oxidized macrocycles.

0 1 2 30

2 5

5 0

% of

1 n mac

rocycl

es 1 1 4 1 6 1 3

T i m e ( h )

(a) 40 nm 16 fibres as a seed

0 1 2 30

2 5

5 0

% of

1 n mac

rocycl

es 1 1 4 1 6 1 3

T i m e ( h )

(b) Coacervates preparedwith an excess of

pDADMAC

0 1 2 30

2 5

5 0 1 1 4 1 6 1 3

% of

1 n mac

rocycl

es

T i m e ( h )

(c) Coacervates preparedwith an excess of pAA

Figure S3.5: Screening of different conditions for the growth of 16 seeds in presence ofcoacervates. Libraries were prepared by mixing 16 (30 µM), a 50% oxidized mixture of 1and 13/14 (200 µM in total), and pAA and pDADMAC, with a total concentration of 50mM and a ratio between them of 1:1 (a), 1:1.5 (b) and 1.5:1 (c).

84


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

(ppm)

1.06

1.020.99

1.890.931.911.000.99

4.444.464.474.794.804.814.834.874.894.904.927.457.457.477.477.487.497.497.507.517.517.537.697.717.947.967.968.008.028.13

OCl

O

- O3S

Figure S3.6: Structure and 1H-NMR (400 MHz, CD3CN) of 9-(2-sulfo)fluorenylmethyloxy-carbonylchloride (Smoc-Cl).

85


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

(ppm)

7.56

12.00

4.28

8.27

20.09

8.09

0.500.710.83

1.25

2.002.182.43

2.74

3.64

4.27

6.87

7.167.177.23

7.687.84

O

O

SO3-

N HN

NH N

OO

SO3-O

O

- O3S

O

O

- O3S

Figure S3.7: Structure and 1H-NMR (400 MHz, D2O) of Smoc-Spermine.

86

3.8. ACKNOWLEDGEMENTS

3.7.1 Appendix: ImageJ script used to analyse coacervatepictures

run("Duplicate...", "title=bw");run("Subtract Background...", "rolling=35");run("Duplicate...", "title=gray");selectWindow("bw");setAutoThreshold("Default dark no-reset");//run("Threshold...");setThreshold(3.500, 1000000000000000000000000000000.0000);setOption("BlackBackground", false);run("Convert to Mask");run("Watershed");run("Set Measurements...", "area mean standard median limitredirect=gray decimal=0");run("Analyze Particles...", "size=10-Infinitycircularity=0.50-1.00 show=Outlines display clear");String.copyResults();selectWindow("bw");run("Close");selectWindow("gray");run("Close");

3.8 Acknowledgements

Guillermo Monreal Santiago and Sijbren Otto conceived the project. Patricia Wolfperformed the experiments related to spermine deprotection. Guillermo MonrealSantiago performed the rest of experiments and wrote the manuscript. Viktor V.Krasnikov performed fluorescence and confocal microscopy, and participated indiscussions. Armin Kiani performed electron microscopy. Alba Algara Lópezand Omer Markovitch are gratefully acknowledged for their help with the sta-tistical analysis of droplet populations. Xinkai Qiu is gratefully acknowledgedfor the preparation of a microfluidic device that could increase the stability ofcoacervates towards coalescence [37], even if the operation of that device was toocomplicated for its practical use.

87


3.9 References

[1] Shirt-Ediss, B.; Murillo-Sànchez, S.; Ruiz-Mirazo, K. Framing major prebiotictransitions as stages of protocell development: three challenges for origins-of-life research. Beilstein J. Org. Chem. 2017, 13, 1388–1395.

[2] Ruiz-Mirazo, K.; Briones, C.; De La Escosura, A. Prebiotic systems chem-istry: New perspectives for the origins of life. Chem. Rev. 2014, 114, 285–366.

[3] Mann, S. Systems of creation: the emergence of life from nonliving matter.Acc. Chem. Res. 2012, 45, 2131–2141.

[4] Koonin, E. V. The origins of cellular life. Antonie van Leeuwenhoek, Int. J. Gen.Mol. Microbiol. 2014, 106, 27–41.

[5] Chen, I. A.; Walde, P. From self-assembled vesicles to protocells. Cold SpringHarb. Perspect. Biol. 2010, 2.

[6] Schrum, J. P.; Zhu, T. F.; Szostak, J. W. The origins of cellular life. Cold SpringHarb. Perspect. Biol. 2010, 2, a002212.

[7] Paleos, C. M. Organization and compartmentalization by lipid membranespromote reactions related to the origin of cellular life. Astrobiology 2018, 19,547–552.

[8] Kamimura, A.; Kaneko, K. Compartmentalization and cell division throughmolecular discreteness and crowding in a catalytic reaction network. Life2014, 4, 586–597.

[9] Attwater, J.; Wochner, A.; Holliger, P. In-ice evolution of RNA polymeraseribozyme activity. Nat. Chem. 2013, 5, 1011–1018.

[10] Dyson, F. Origins of life, 2nd ed.; Cambridge University Press, 1999.

[11] Baeza, I.; Ibáñez, M.; Wong, C.; Chávez, P.; Gariglio, P.; Oró, J. Possible pre-biotic significance of polyamines in the condensation, protection, encapsu-lation, and biological properties of DNA. Orig. Life Evol. Biosph. 1992, 21,225–242.

[12] Oparin, A. I. A. I. The origin of life, 2nd ed.; Dover Publications, 2003.

[13] Banani, S. F.; Lee, H. O.; Hyman, A. A.; Rosen, M. K. Biomolecular conden-sates: Organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol. 2017, 18,285–298.

88

3.9. REFERENCES

[14] Brangwynne, C. P.; Eckmann, C. R.; Courson, D. S.; Rybarska, A.; Hoege, C.;Gharakhani, J.; Jülicher, F.; Hyman, A. A. Germline P granules are liquiddroplets that localize by controlled dissolution/condensation. Science 2009,324, 1729–1732.

[15] Bungenberg de Jong, H. G.; Kruyt, H. R. Coacervation (partial miscibility incolloid systems). Proc K. Ned. Akad. Wet. 1929, 32, 849–856.

[16] Kim, S.; Huang, J.; Lee, Y.; Dutta, S.; Yoo, H. Y.; Jung, Y. M.; Jho, Y.;Zeng, H.; Hwang, D. S. Complexation and coacervation of like-chargedpolyelectrolytes inspired by mussels. Proc. Natl. Acad. Sci. 2016, 113, E847–E853.

[17] Kayitmazer, A. B. Thermodynamics of complex coacervation. Adv. ColloidInterface Sci. 2017, 239, 169–177.

[18] Nakashima, K. K.; Vibhute, M. A.; Spruijt, E. Biomolecular chemistry in liq-uid phase separated compartments. Front. Mol. Biosci. 2019, 6, 21.

[19] van der Gucht, J.; Spruijt, E.; Lemmers, M.; Cohen Stuart, M. A. Polyelec-trolyte complexes: Bulk phases and colloidal systems. J. Colloid Interface Sci.2011, 361, 407–422.

[20] Frankel, E. A.; Bevilacqua, P. C.; Keating, C. D. Polyamine/nucleotide coac-ervates provide strong compartmentalization of Mg2+, nucleotides, andRNA. Langmuir 2016, 32, 2041–2049.

[21] Koga, S.; Williams, D. S.; Perriman, A. W.; Mann, S. Peptide-nucleotide mi-crodroplets as a step towards a membrane-free protocell model. Nat. Chem.2011, 3, 720–4.

[22] Aumiller, W. M.; Keating, C. D. Phosphorylation-mediated RNA/peptidecomplex coacervation as a model for intracellular liquid organelles. Nat.Chem. 2016, 8, 129–137.

[23] Yewdall, N. A.; Buddingh’, B. C.; Altenburg, W. J.; Timmermans, S. B. P. E.;Vervoort, D. F. M.; Abdelmohsen, L. K.; Mason, A. F.; van Hest, J. Physico-chemical characterization of polymer-stabilized coacervate protocells. Chem-BioChem 2019, cbic.201900195.

[24] Mason, A. F.; Buddingh, B. C.; Williams, D. S.; Van Hest, J. C. Hierarchicalself-assembly of a copolymer-stabilized coacervate protocell. J. Am. Chem.Soc. 2017, 139, 17309–17312.

[25] Drobot, B.; Iglesias-Artola, J. M.; Le Vay, K.; Mayr, V.; Kar, M.; Kreysing, M.;Mutschler, H.; Tang, T. Y. Compartmentalised RNA catalysis in membrane-free coacervate protocells. Nat. Commun. 2018, 9.

89


[26] Poudyal, R. R.; Guth-Metzler, R. M.; Veenis, A. J.; Frankel, E. A.; Keat-ing, C. D.; Bevilacqua, P. C. Template-directed RNA polymerization and en-hanced ribozyme catalysis inside membraneless compartments formed bycoacervates. Nat. Commun. 2019, 10, 490.

[27] Zhao, M.; Zacharia, N. S. Protein encapsulation via polyelectrolyte complexcoacervation: protection against protein denaturation. J. Chem. Phys. 2018,149, 163326.

[28] Sokolova, E.; Spruijt, E.; Hansen, M. M. K.; Dubuc, E.; Groen, J.;Chokkalingam, V.; Piruska, A.; Heus, H. A.; Huck, W. T. S. Enhanced tran-scription rates in membrane-free protocells formed by coacervation of celllysate. Proc. Natl. Acad. Sci. 2013, 110, 11692–11697.

[29] Aumiller, W. M.; Pir Cakmak, F.; Davis, B. W.; Keating, C. D. RNA-basedcoacervates as a model for membraneless organelles: formation, properties,and interfacial liposome assembly. Langmuir 2016, 32, 10042–10053.

[30] Nakashima, K. K.; Baaij, J. F.; Spruijt, E. Reversible generation of coacervatedroplets in an enzymatic network. Soft Matter 2018, 14, 361–367.

[31] Frederix, P. W.; Idé, J.; Altay, Y.; Schaeffer, G.; Surin, M.; Beljonne, D.; Bon-darenko, A. S.; Jansen, T. L.; Otto, S.; Marrink, S. J. Structural and spectro-scopic properties of assemblies of self-replicating peptide macrocycles. ACSNano 2017, 11, 7858–7868.

[32] McCall, P. M.; Srivastava, S.; Perry, S. L.; Kovar, D. R.; Gardel, M. L.; Tir-rell, M. V. Partitioning and enhanced self-assembly of actin in polypeptidecoacervates. Biophys. J. 2018, 114, 1636–1645.

[33] Kunz, W.; Henle, J.; Ninham, B. W. ’Zur lehre von der wirkung der salze’(about the science of the effect of salts): Franz Hofmeister’s historical papers.Curr. Opin. Colloid Interface Sci. 2004, 9, 19–37.

[34] Jha, P. K.; Desai, P. S.; Li, J.; Larson, R. G. pH and salt effects on the as-sociative phase separation of oppositely charged polyelectrolytes. Polymers(Basel). 2014, 6, 1414–1436.

[35] Levine, H. Thioflavine T interaction with synthetic Alzheimer’s disease β-amyloid peptides: detection of amyloid aggregation in solution. Protein Sci.1993, 2, 404–410.

[36] Malakoutikhah, M.; Peyralans, J. J.; Colomb-Delsuc, M.; Fanlo-Virgós, H.;Stuart, M. C.; Otto, S. Uncovering the selection criteria for the emergenceof multi-building-block replicators from dynamic combinatorial libraries. J.Am. Chem. Soc. 2013, 135, 18406–18417.

90

3.9. REFERENCES

[37] Vanswaay, D.; Tang, T. Y. D.; Mann, S.; DeMello, A. Microfluidic formationof membrane-free aqueous coacervate droplets in water. Angew. Chemie - Int.Ed. 2015, 54, 8398–8401.

[38] Jia, T. Z.; Hentrich, C.; Szostak, J. W. Rapid RNA exchange in aqueous two-phase system and coacervate droplets. Orig. Life Evol. Biosph. 2014, 44, 1–12.

[39] Butler, G. B.; Angelo, R. J. Preparation and polymerization of unsatu-rated quaternary ammonium compounds. VIII. A proposed alternatingintramolecular-intermolecular chain propagation. J. Am. Chem. Soc. 1957, 79,3128–3131.

[40] Yang, S.; Schaeffer, G.; Mattia, E.; Hussain, A. S.; Ottelé, J.; Mayer, C.;Otto, S. Chemical fueling enables complexification of assembly-driven self-replicators. Submitted

[41] Leonetti, G.; Otto, S. Solvent composition dictates emergence in dynamicmolecular networks containing competing replicators. J. Am. Chem. Soc.2015, 137, 2067–2072.

[42] Hatai, J.; Motiei, L.; Margulies, D. Analyzing amyloid beta aggregates witha combinatorial fluorescent molecular sensor. J. Am. Chem. Soc. 2017, 139,2136–2139.

[43] Ottelé, J. O.; Hussain, A. S.; Otto, S. Chance emergence of catalytic activityand promiscuity in a self-replicator. Nat. Chem. 2020, Accepted. Preprint –doi: 10.26434/chemrxiv.10002032.v1.

[44] van Esterik, K. (Sul)Fmoc deprotection as a central strategy to incorporateproto-metabolism and compartmentalisation in a system of synthetic repli-cators. Master thesis - Rijksuniversiteit Groningen 2019,

[45] Schindelin, J. et al. Fiji: an open-source platform for biological-image analy-sis. Nat. Methods 2012, 9, 676–82.

[46] Pal, A.; Malakoutikhah, M.; Leonetti, G.; Tezcan, M.; Colomb-Delsuc, M.;Nguyen, V. D.; Van Der Gucht, J.; Otto, S. Controlling the structure andlength of self-synthesizing supramolecular polymers through nucleatedgrowth and disassembly. Angew. Chemie - Int. Ed. 2015, 54, 7852–7856.

[47] Lau, A. W.; Prasad, A.; Dogic, Z. Condensation of isolated Semi-flexible fila-ments driven by depletion interactions. EPL 2009, 87, 48006.

[48] Merrifield, R. B.; Bach, A. E. 9-(2-Sulfo)fluorenylmethyloxycarbonyl chlo-ride, a new reagent for the purification of synthetic peptides. J. Org. Chem.1978, 43, 4808–4816.

91


92

Documents

University of Groningen Steps towards de-novo life Monreal ...chaotropes [33] to destabilize the formed structures (Table S3.1). All these strate-gies led to either amorphous aggregates