University of Groningen Steps towards de-novo life Monreal Santiago, Guillermo

(1)

Steps towards de-novo life

Monreal 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 from it. Please check the document version below.

Document Version

Publisher'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 mechanisms in synthetic self-replicating systems. University of Groningen. https://doi.org/10.33612/diss.121581426

Copyright

Other 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 the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

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

(2)

Chapter 2

Mechanistic studies

uncovering a new pathway for

supramolecular

polymerization and replication

In this chapter, we use evidence from high speed atomic force microscopy and molecular dynamics to gain insight into the self-replication process of peptide-based macrocycles. In doing so, we discover a novel mechanism for supramolecular polymerization, involv-ing the accumulation of precursors on the side of the supramolecular polymer, and their subsequent diffusion to the growing ends. We confirm this pathway with kinetic mea-surements, and discover the existence of another state of the precursors: an off-pathway assembly separated from the fibres that does not contribute to the polymerization. This mechanism, which bears some resemblance to previous findings in the growth of amyloid proteins, explains (at least partially) the role of the self-assembled structures of the pep-tide replicators in catalysing their own formation, and it can have implications for the autocatalytic growth of other supramolecular structures.

This chapter is part of a submitted article: Maity, S.; Ottelé, J.; Monreal Santiago, G.; Frederix, P. W.; Kroon, P. C.; Markovitch, O.; Stuart, M. C. A.; Marrink, S. J.; Otto, S.; Roos, W. H. Caught in the act: Mechanism of supramolecular polymerization driven self-replication visualized in real time. Submitted.

(3)

2.1 Introduction

Self-assembled supramolecular structures, such as the ones formed from 16, have

gained importance in fields from materials science to biomedicine [1–3]. Since they are based on weak intermolecular interactions and not on covalent bonds, they have historically been regarded as structures in thermodynamic equilibrium. However, it is becoming more and more evident than the formation of supramo-lecular structures can also be governed by their assembly pathway and not only by thermodynamics [4–6]. Furthermore, the focus of self-assembly research is ex-panding to out-of-equilibrium systems, where the outcome of the self-assembly is undoubtedly influenced by its mechanism [7, 8].

Despite this growing interest in understanding the mechanisms behind self-assembly, the tools that we have available for their study are limited. Current methods for real-time visualization of such systems use confocal laser scanning microscopy [9] or stochastic optical reconstruction microscopy [10], but those re-quire the use of fluorescent probes and only provide resolutions down to tens of nanometers. Electron microscopy could provide higher resolutions, but its appli-cations to real-time visualization are still in their early days [11], presenting lim-itations due to the low temperature, cell morphology, and high-energy electrons that are required by the technique. Recent advances in the field of atomic force microscopy (AFM) have enabled studying dynamic processes of (bio)molecular systems using High-Speed AFM (HS-AFM) [12] at even smaller length scales, including the configurational dynamics of proteinaceous structures, [13–15] the assembly of amyloid-like fibrils [7, 16], and the movement of synthetic molecular transporters [17] with unprecedented spatiotemporal resolution. HS-AFM allows for the continuous observation of samples at room temperature and in solution, with the only constraint being that one of the components needs to be attached to a surface so its position remains constant.

In this chapter, we use HS-AFM to study the mechanism of formation of the self-replicating fibres introduced in Chapter 1. These fibres emerge sponta-neously from stirred aqueous solutions of the monomer 1. First, the thiol groups in 1 are oxidized to disulfides, giving rise to an array of small macrocycles (mainly

13/14), continuously exchanging with each other. From this library of

macrocy-cles emerges 16, which initially forms in low concentrations but has the ability

to stack into supramolecular fibres. These fibres draw the equilibrium towards their own growth, consuming 13/14 and forming more 16 until it becomes the

main component of the library (Figures 2.1a and 2.1b). The growth mechanism of these fibres is particularly complicated to elucidate, since it involves both the conversion of 13/14 precursors to 16 and the self-assembly of 16 into fibres. We

have previously established that it is an autocatalytic process [18], but we only understand it kinetically, not mechanistically. In this chapter, HS-AFM gives us new insight into this process, by allowing us to directly visualize the growth of

(4)

2.1. INTRODUCTION

16fibres. This uncovers a new mechanism of autocatalysis and supramolecular

polymerization, that we confirm with experiments in solution and molecular dy-namics.

(a) Scheme showing the reactions converting 1 into 16fibres.

(b) Evolution over time of the main 1n

macrocycles in a 1 library (c) Mechanism proposed in this chapter for thegrowth of 16fibres

Figure 2.1: Growth of 16. (a) Formation of 16from building block 1. Upon oxidation ( 1 ) the

monomer forms a mixture of small macrocycles that can interchange building blocks with one another through disulfide exchange ( 2 ). After nucleation ( 3 ), 16fibres can elongate

by the stacking of additional 16 macrocycles ( 4 ). (b) Representative kinetic analysis of

the relative concentrations of 1n, measured over time by UPLC. Only the main species of

the library are shown. Inserts show the coarse-grained models of 1, 13, 14, 16, and a high

resolution AFM image of a single fibre on a lipid bilayer, where the twisted structure of the fibre can be observed. Amino acid side chains are not shown in the coarse-grained models. (c) Model representation of the self-assembly pathway described in this chapter.

(5)

2.2 Results and discussion

In order to monitor the growth of 16fibres by HS-AFM, we needed to attach them

to a surface. Preliminary results showed that 16fibres, positively charged at pHs

close to 7 (such as 8.2, typically used for fibre growth) were readily adsorbed to negative surfaces. In order to achieve a binding that was strong enough for adsorption, but not so strong that the fibres would not be reactive, we used a slightly negative lipid bilayer as the surface for our experiments (See Methods). The morphology of the fibres in these conditions (Figure 2.1b, inset) was similar to previous cryo-EM results, indicating that the structure of the fibres was not being altered by the surface.

Next, we added a solution of fibre precursors (a preoxidized solution of 1, consisting mostly 13/14 and ~15 % of 1 to catalyse exchange)1 to the imaging

chamber of the microscope, in conditions where fibre growth would typically take place (room temperature, 2.31 mM concentration in 1 units). As soon as the solution of precursors was in contact with the surface, we observed the forma-tion of aggregates on top of the pre-attached fibres (Figure 2.2a). These aggre-gates had a diameter of ~4 nm, did not show any defined structure, and were only observed in contact with the fibres. They only appeared in presence of the precursors, clearly indicating that the small macrocycles were responsible for their formation. Interestingly, these aggregates seemed to be closely linked to the mechanism of fibre growth: We only observed elongation in fibres that had an aggregate attached. Furthermore, the growth appeared distance-dependent: it only took place in the fibres where the aggregate was close enough to the grow-ing end (Figure 2.2). Monitorgrow-ing closely the growth of sgrow-ingle fibres, we were able to observe that the aggregates were depleted as the fibres grew (Figure S2.1), and that when they were too far away from the fibre end, growth stopped until a new aggregate was attached. We also observed diffusion of the aggregates along the fibre axis, but only in their early stages of formation (Figure S2.2).

The combination of these experiments led us to start envisioning a new mech-anism for fibre elongation. This mechmech-anism starts with the accumulation of the small macrocycles, 13/14, on the surface of fibres. The interaction between small

macrocycles must be more favourable that their interaction with 16fibres, so they

tend to cluster together forming aggregates. The diffusion of those aggregates along the fibre axis is possible, but less likely as they grow in size. However, we propose that material, in the form of individual macrocycles or small clusters, can still diffuse from the immobile aggregates to the fibre ends and continue the formation of more self-assembled structure in the form of 16.2

1_{For simplicity, preoxidized libraries of 1 containing mostly 1}

3and 14will be referred to as

"pre-cursors" or "13/14" for the rest of this chapter and thesis. 2_{The step in which 1}

3/14are converted into 16is unclear in this mechanism - we have no evidence

(6)

2.2. RESULTS AND DISCUSSION

(a) 16fibres with 13/14

attached in different positions relative to the fibre ends

(b) Different fibre-precursor interactions

(c) Growth of 16with

13/14near the end

(d) Growth of 16with

13/14in the middle

(e) Growth of 16with

no 13/14attached

Figure 2.2: Precursor attachment near the fibre end is essential for growth. (a) Exam-ple of a typical AFM image of 16 fibres in presence of 13/14 precursors. Three different

fibre-precursor interactions are indicated as green circles (precursors attached near the end of the fibre, <20 nm), blue square (precursors attached in the middle of the fibre) and grey arrows (fibres without any precursor attached). (b) Schematic representation of these three fibre-precursor arrangements. Histograms over time of fibres that have precursors attached near the end (c), in the middle of the fibres (d) or without precursors (e). [13/14]

= 2.31 mM in 1 units. Imaging times = 0, 30, 60 minutes. N = 70, 139, and 108 for (c); 20, 17, and 39 for (d); and 120, 77, and 90 for (e).

To validate this mechanism, we first compared the results obtained in HS-AFM to solution-based experiments. In a library prepared in bulk containing 2.29 mM of 13/14 (88 % oxidation) and 0.1 mM of 16 (both in 1 units), we

ob-served a rate of fibre growth of 6 ± 1 nm × minute-1_{× fibre end}-1_{(see Methods}

for calculations and assumptions made). This rate fits the growth rate observed in HS-AFM for the fibres that were growing (5.2 ± 0.4 nm × min-1_{, N = 40, mean}

± S. E. M.). This indicates that, unless there is a faster reaction that we are not detecting, all the fibres in solution must be growing through this pathway and at a very similar rate. This fits previous results, where the length distribution of

16fibres was conserved after their growth [19], and gives them a mechanistic

ex-planation: since all fibre ends are growing at almost the same rate, the variation on their respective lengths introduced by growth is minimal. However, there is a discrepancy between HS-AFM and solution-based experiments: in the former a considerable number of fibres are not growing at all (Figure 2.2a), so the aver-age growth rate of all fibre ends is much lower than the one observed in solution. We attribute this difference to the attachment of fibres to the lipid bilayer. As shown by molecular dynamics [20] and microscopy (Figure 2.1b, inset), the 16

(7)

for a schematic representation). When the fibres are deposited on the lipid sur-face, part of these grooves must be directly in contact with the bilayer, "facing down". Therefore, if 13/14 bind and diffuse along the grooves, as molecular

dy-namics suggests (Figure S2.3b) they could only travel for a certain length of the fibre before being "blocked" by the surface. We believe that this is the reason why, in the HS-AFM experiments, the growth rate depends on the distance between aggregate and fibre end: if this distance is long enough for the surface to stop the movement of 13/14, the aggregates are "stuck" and cannot contribute to

elonga-tion.3 _{Both molecular dynamics and AFM imaging of samples taken from bulk}

solution confirmed this hypothesis: in both cases, without the surface present,

13/14spontaneously aggregated in the fibre surface and diffused completely

to-wards the fibre ends (Figure S2.3). The MD results indicate a tendency for the fibres to concentrate in the fibre ends and not just distribute randomly along the structure, which could explain the uneven initial distribution of aggregates in AFM experiments (Figure S2.4). This seems to indicate that the diffusion along the fibres is biased towards the fibre ends. However, at the moment we do not have enough evidence to completely support this hypothesis nor understand the mechanism through which it would take place.

We further characterized the 13/14 aggregates by studying them with other

techniques. First, Cryo-TEM confirmed the presence of aggregates with similar sizes to the ones shown in AFM (Figure 2.3a). Surprisingly, those aggregates were not only attached to the 16fibres, but also existed in solution independently

(Fig-ure 2.3b). This was further supported by DLS and Nile Red fluorescence (Fig(Fig-ures S2.5a and S2.5b), which confirmed their size and indicated the presence of hy-drophobic binding sites, respectively.

The existence of aggregates without fibres raised a question about the mech-anism: were the 13/14 aggregates detected in solution the same aggregates

ob-served in HS-AFM, or were they a different reservoir competing for the concen-tration of 13/14? In order to answer this question, we decided to study the growth

rate of 16fibres in presence of 13/14 precursors under and over their critical

ag-gregation concentration (CAC). The CAC represents the minimal concentration of 13/14required for the formation of aggregates in solution, so if those were

nec-essary for the formation of the aggregates attached to the fibres, there should be no growth under it.4

We determined the critical aggregation concentration (CAC) of 13/14by

titra-tion of a Nile Red solutitra-tion (Figure S2.5c), obtaining a value of 23 ± 5 µM (in 1 units, mean ± standard deviation, N = 3). Then, we measured the initial growth

3_{The distance of half a twist of the fibre structure, 20 nm, was used as a threshold to classify}

fibres as "aggregates attached in the middle" and "aggregates attached in the end" during this text, in accordance with this hypothesis.

4_{The aggregates formed in the side of 1}

6fibres might also have a CAC. We cannot detect this

concentration directly (as Nile Red will also give a signal due to 16fibres), but we expect it to be much

(8)

2.2. RESULTS AND DISCUSSION

(a) Cryo-TEM micrograph of 16+ 13/14,

showing fibres (orange) and aggregates (red)

Diameter: 2,6±0,6 nm (100 measurements)

(b) Cryo-TEM micrograph of 13/14, showing

aggregates (red) 0 1 0 2 0 3 0 4 0 0 , 0 0 0 , 0 5 0 , 1 0 0 , 1 5 0 , 2 0 0 , 2 5 R e p e a t 1 R e p e a t 2 [1 ₃/1 ₄] (µM ) G ro w th r a te ( µ M /m in )

(c) Initial growth of 16at concentrations of

13/14over and under their CAC (blue)

0 8 0 0 1 6 0 0 2 4 0 0 0 1 0 2 0 3 0 4 0 5 0 G ro w th r a te ( µ M /m in ) [1 ₃/1 ₄] (µM )

(d) Initial growth of 16at concentrations of

13/14higher than their CAC (blue)

Figure 2.3: The fibre precursors 13/14form off-pathway aggregates in solution. Cryo-EM

images of 13/14(4 mM, in 1 units) in presence (a) and absence (b) of 16fibres (4 mM, in 1

units). To aid interpretation, arrows have been added indicating the location of aggregates (red), fibres (orange), and ice contamination (blue). Scale bars are 10 nm. Panels (c) and (d) show the initial growth rate of 16fibres at different concentrations of precursors, over

and under their critical aggregation concentration, (blue area). Panel (c) shows two series of experiments, with a concentration of 16of 5 µM in 1 units, which were monitored by

UPLC. Panel (d) shows three repeats (red circles) and their average and standard deviation (black squares and error bars). In this case the concentration of 16was 60 µM in 1 units,

(9)

rate of fibres in samples with the same initial concentration of 16 and different

concentrations of 13/14, encompassing values under and over their CAC (Figure

2.3c). In these experiments the growth rate was proportional to the concentration of precursors, but we did not observe any effect when going to concentrations un-der the value of the CAC. As explained above, this is evidence that the aggregates in solution are not necessary for fibre growth, therefore they must be a different state than the aggregates attached to the fibres.

When we repeated this experiment, expanding the range of 13/14

concentra-tions to higher values (Figure 2.3d), we observed that after a certain point the growth rate started to level off. This observation supports the theory that solution aggregates do not contribute to growth - after surpassing the CAC an increase in the concentration of 13/14increases the concentration of solution aggregates, but

not directly the concentration of free 13/14 and therefore not the growth rate.5

While it is true that the growth rate starts to plateau at concentrations higher that the CAC, this difference could be due to the amount of 13/14 that is removed

from the solution by the fibre surface.

Taking this last piece of evidence into account, we proposed the mechanism shown in Figure 2.1c. This mechanism combines the steps mentioned before (ac-cumulation of 13/14on the surface of 16fibres, diffusion along the fibre axis, and

growth) with an off-pathway state, the aggregates of 13/14 in solution. This

off-pathway state competes with the fibres for the concentration of its precursors, limiting the maximum concentration of 13/14 available and therefore the

maxi-mum growth rate.

5_{We discarded that the plateau in growth rate was due to the saturation of the fibre surface since,}

(10)

2.3. CONCLUSIONS

2.3 Conclusions

In this chapter we show, for the first time, direct observation of a self-replication process by microscopy. The results show the formation of aggregates of precur-sors of the replicator, that then diffuse along its fibres to elongate them. This has some similarities to secondary nucleation, a process described in amyloid litera-ture where fibres catalyse the formation of other fibres that grow from their side. [21] However, it is radically different, as the mechanism described here involves the diffusion of the precursors to the fibre end, instead of the formation of new structures.

This mechanism explains the autocatalysis of 16 fibres, or at least a

compo-nent of it. Here, the accumulation of precursors on the surface of the fibre has two effects that accelerate fibre growth. First, it increases the local concentra-tion of small macrocycles, compensating the entropic penalty of the formaconcentra-tion of larger disulfide rings (such as 16). Secondly, it facilitates the contact between the

fibre ends and new material from which they can grow - since the diffusion of macrocycles happens through the fibre grooves, it becomes a 1D problem instead of a 3D one. Furthermore, preliminary evidence obtained through the course of this work indicates that the structure of the fibres might actively guide those pre-cursors to the fibre ends, which would further accelerate the process.

This mechanism leaves some questions open about the replication of 16. It is

still unclear whether the conversion of 13/14 → 16happens in the aggregates or

if it is also catalysed by a templating effect of the fibre end. It is also unknown whether the formation of these aggregates is merely a physical process due to hydrophobicity, or whether there is some recognition between the fibres and the precursors. Previous results indicate that replicators of this type show specific interactions in competition experiments [22, 23], but we still have not discovered whether this specificity is related to the fibre ends or to another type of interaction between precursors and fibres such as the one shown here.

Beyond the implications for replicators of the type of 16, this pathway could

be present in other systems. Since it only depends on the ability of precursors and structure to bind to each other and it does not depend on their chemical composition, we expect it to play a role for the polymerization of other supramo-lecular fibres (or other non-discrete structures, such as sheets or ribbons!). If this mechanism is proven to be general, we envision that it could be used not only to explain but also to design new autocatalytic supramolecular structures, with no other requirements than the formation of a surface to which their precursors can be attached.

(11)

2.4 Material and methods

All reagents, solvents, and buffer salts were purchased from commercial sources and used without further purification. Building block 1 (XGLKFK) was obtained from Cambridge Peptides Ltd (Birmingham, UK).

UPLC analysis

UPLC analysis was performed on a Waters Acquity UPLC H-class, equipped with a 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 colcol-umn tem-perature was kept at 35 °C, and the sample plate was kept at 25 °C, unless oth-erwise specified. UV absorbance was monitored at 254 nm. For each injection, 10 µL of sample was injected. All peaks were assigned to their corresponding compounds by comparison with previous results.

The following solvents and gradient were used for UPLC analysis. The flow rate 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

Buffer preparation

All the samples in this chapter were prepared using borate buffer as a solvent, with a pH of 8.2 and a total concentration of 50 mM in boron atoms.This buffer was prepared from B2O3, purchased from Sigma-Aldrich (Purity ≥ 99.5%)

(12)

2.4. MATERIAL AND METHODS

Formation of 1

6

fibres

A stock solution of 16was prepared by adding building block 1 to a 1 mL HPLC

vial (12 x 32 mm) containing a Teflon coated magnetic stirring bar (5 x 2 mm, VWR). The building block was dissolved in borate buffer prepared from 25 mM B2O3 and adjusted to a pH of 8.1 to a final concentration of 1.54 mM in 1 units

and was kept at elevated temperatures while mechanical agitation was applied (1200 rpm, 45 °C). The sample was subjected to periodic UPLC analysis and kept at the conditions described above until the sample contained >90 % 16. The 16

fibres were then stored at room temperature while stirring and could be used up to 8 weeks after preparation without observing any significant changes in sample composition.

Seed preparation by mechanical shearing

Samples of 16 fibres were mechanically sheared using a modified protocol

pre-viously published by our group [19]. A 150 µL aliquot of a prepre-viously prepared solution of 16was placed in a Couette cell (Rcup= 20.25 mm, Rbob= 20.00 mm,

av-erage radius (R) = 20.125 mm). The sample was mechanically sheared by rotation of the inner cylinder, with a frequency of 4000 rpm (corresponding to a shear rate of 33702 s-1_{). The resulting fibres were used within 48 h of preparation.}

Fibre end estimation

The average fibre length of the sheared seeds was analyzed using transmission electron microscopy. Using the ImageJ software [24], the length of 994 sheared seeds was measured, resulting in an average length of 34.8 ± 15.2 nm (mean ± standard deviation). The average height of a single 16macrocycle is 0.485 nm

[20], therefore we find an average of 71.8 ±31.3 macrocycles in a sheared seed fi-bre. We emphasize that these intervals do not represent experimental errors, but the width of the fibre length distribution.

Preparation of stock solutions of 1

3

/1

4

A stock solution of a mixture of 1, 13 and 14 was prepared by adding building

block 1 to a HPLC vial (12 × 32 mm) and transferring it to a glovebox. Then, building block 1 was oxidized using sodium perborate (0.80-0.85 equiv.) in borate buffer to obtain a final concentration of 2.31 mM in 1 units. The resulting mixture was analysed by UPLC, and could be used up to three days if no mechanical agitation was supplied.

(13)

Determination of the average fibre growth rate by UPLC

A glass insert was placed in a 1 mL HPLC vial. Of an oxidized 1 solution (88 % oxidation, 2.31 mM in 1 units in 50 mM borate buffer, pH 8.12), 95 µL was added to the insert. Two minutes before the UPLC injection, 5 µL of sheared 16seeds

(2.0 mM in 1 units) were added and the sample was mixed thoroughly, resulting in a final concentration of 2.29 mM in 1 building blocks. The mixture was kept without any mechanical agitation at a constant temperature of 25 °C, and the library composition was monitored by UPLC every 18 min for 128 min. The initial growth rate was determined by linear regression, in units of %16 × min-1. This

rate was converted to 16molecules × min-1, normalized by the number of fibre

ends using the average fibre length determined above (71.8 nm), and converted to nm × minute-1 × fibre end -1 _{considering the width of a 1}

6 molecule to be

0.485 nm [20]. The data reported is the average and standard deviation of four independent measurements.

Elongation experiments monitored by fluorescence

Samples containing sheared 16seeds (60 µM in building block), thioflavin T (500

µM), and increasing concentrations of 13/14 precursors (0-2.3 mM in 1 units, 86

% oxidation) were prepared in a 96-well plate using borate buffer as a solvent. The samples were shaken (orbital shaking for 30 s) at a controlled temperature of 25 °C, and the fluorescence of thioflavin T (λexc = 440 nm, λem= 500 nm) was

measured every 5 min using a Synergy|H1 microplate reader (BioTek, USA). Si-multaneously, samples containing thioflavin T (500 µM) and increasing concen-trations of 16 seeds or 13/14 were monitored in the same way. These samples

were used as calibration to correlate the fluorescence signal with concentration of both 13/14and 16, and to monitor the photobleaching of thioflavin T (which

remained always lower than 5 % of the initial signal). The fluorescence intensity at every time point was converted to the concentration of 16after subtracting the

signal coming from 13/14, and the initial growth rate was calculated by linear

re-gression of the first 5 points of each sample. This experiment was repeated three times.

Surface preparation for AFM studies

To immobilize the fibres on the surface, we have used a lipid bilayer deposited on top of freshly cleaved mica. The lipid bilayer was formed by absorption of large unilamellar vesicles (LUVs) onto a freshly cleaved mica surface. LUVs were prepared using a lipid mixture composed of 60 % dioleoyl-phosphatidylcholine (DOPC) and 40 % dioleoyl- phosphatidyl-serine (DOPS) (mol:mol) from Avanti polar. The lipid mixture containing 1 mg/mL of total lipids was mixed in 200 µl chloroform in a small glass vial. Next chloroform was evaporated using Argon

(14)

gas while slowly rotating the vial to produce a lipid film on the glass wall. The film was kept in a vacuum desiccator for 30-45 min. After drying the lipid film, 200 µL of a buffer composed of 10 mM HEPES, pH 7.4, 100 mM NaCl and 50 mM sucrose was added and vortexed for 30 seconds. The mixture was freeze-thawed three times using liquid nitrogen. The LUVs were stored at -20 °C for further use within one month. For deposition on a mica surface, we have used 0.2 mg/mL concentration of the stock preparation (diluted in the same buffer) and incubated on top of freshly cleaved mica (HS-AFM sample holder) for 15-30 min. The surface was then cleaned 3-5 times with 50 mM borate buffer, pH 8.1.

HS-AFM experiments

All the AFM studies were done using HS-AFM (RIBM, Japan) in amplitude mod-ulation tapping mode in liquid [15, 25, 26]. Short cantilevers (USC-F1.2-k0.15, NanoWorld, Switzerland) with spring constant of 0.15 N/m, resonance frequency around 0.6 MHz and a quality factor of ~2 in buffer, were used. The cantilever free amplitude was set to 1 nm, and the set-point amplitude for the cantilever os-cillation was set around 0.9 nm. Images were taken at 0.2-0.5 frame/s depending on the size of the image. A mica surface of diameter 1.5 mm glued on top of a 5 mm high glass rod was used as the AFM sample stage. The glass rod was then attached to the scanner Z-piezo using a small amount of wax. After formation of the lipid bilayer (as mentioned above), the short preassembled fibres were incu-bated for 30 seconds and then cleaned with borate buffer. The scanner head was then put upside down into a small liquid chamber containing the cantilever and filled with 120 µL of the recording solution. The HS-AFM works as sample scan-ning system and a minimum imaging force (<100 pN) was applied throughout all experiments. We have performed three different sets of AFM experiments as mentioned in the main text. For classic AFM experiments, the sheared 16 seeds

were immobilized on a lipid surface and the chamber was filled with a 2.3 mM

13/14solution (in 1 units). AFM images were taken at different time points while

keeping the seeds incubated in the 13/14solution. Due to mechanical drift,

imag-ing the exact place after several minutes was not possible, therefore, we have estimated the elongation looking to a bulk batch of fibres in different positions (at least 10) on the surface. For dynamic studies, we used a similar setup, but af-ter localizing a seed with precursor aggregates attached near its end, we zoomed in and imaged it continuously at 0.5 frames. Due to low growth rate, mechanical drift of imaging and small piezo limit (900 nm × 900 nm), we were only able to follow a fibre for a maximum of 10 min. For in-solution studies, we pre-incubated the seed and the 13/14 precursors at a 5 % to 95 % molar ratio in a sealed glass

bottle. For every pre-decided time, we then took a small amount of the mixture and incubated it on a freshly cleaved mica surface for 30 s, which was then rinsed with borate buffer. The AFM imaging was done immediately afterwards in borate buffer.

(15)

AFM data analysis

For AFM data analysis we have used Igor-pro software with built in script from RIBM (Japan) and ImageJ software with additional home written plugins. The HS-AFM images/movies were only processed minimally, through tilt correction, drift correction and brightness correction. The kymographs were obtained from the cross-section at a fixed scale (marked for each image) over the entire movie. It represents the height distribution (in terms of intensity) along the line cross sec-tion as a funcsec-tion of time. For all different experimental condisec-tions, we obtained and reported the results from several days of experiments.

Cryo-transmission electron microscopy

An aliquot (3 µL) of solutions containing 16, 13/14, or both, (4 mM in 1 units,

prepared in borate buffer) was deposited on holey carbon-coated grids (3.5/1 Quantifoil Micro Tools, Jena, Germany) that were previously glow-discharged for 15 seconds. After blotting the excess liquid for 4 seconds, the grids were vitrified in liquid ethane using a Vitrobot (FEI, Eindhoven, the Netherlands) and transferred to a FEI Tecnai T20 electron microscope equipped with a Gatan model 626 cryo-stage operating at 200 keV. Micrographs were recorded under low-dose conditions with a slow-scan CCD camera.

Dynamic light scattering

Dynamic light scattering measurements were performed on a NanoBrook 90Plus PALS Particle Size Analyzer (Brookhaven, NY), using a 659 nm laser at a 90° de-tection angle. Samples were prepared in borate buffer and filtered through a 0.2 µm pore size filter. The refractive index used for the particles was 1.5, but no sig-nificant differences were observed when changing it from 1.4 to 1.6. A set of 10 repeats were recorded for each sample.

Fluorescent probe measurements

A borate buffer solution containing Nile Red (15 µM) was titrated with a con-centrated solution of 13/14 (4 mM in building block, 85 % oxidized). After each

addition of precursors, the sample was homogenised by immersing it in an ultra-sound bath for 1 minute, and its fluorescence spectra was recorded using a JASCO FP6200 fluorimeter (λexc= 553 nm). The titration was repeated three times,

mea-suring in each of them the point when the fluorescence band started blue-shifting and increasing in intensity [27].

(16)

Coarse-grain molecular dynamics simulations

CG molecular dynamics simulations were performed using the Martini force field v. 2.2. [28, 29] Parameters for the dithiobenzene group were derived from the atomistic simulations by matching bond, angle, dihedral and non-bonded distri-butions. Previous work has demonstrated random coil secondary peptide struc-ture for macrocycles in solution, while fibres exhibit high Β-sheet content [20]. As such, separate parameters were used for the peptide parts of the macrocycles in the fibre and in solution. In the fibre, the parameters were taken as Β-sheet parameters with extended dihedrals from the standard Martini protein parame-ters [30], while for the single macrocycle the coil parameparame-ters were used. A fibre of 16stacked hexamer macrocycles was constructed. The structure was solvated

in a box of 10.8 x 11.7 x 14.5 nm and counter ions (96 Na+ _{, 192 Cl}- _{; 261 mM)}

were added. 10 % of the water beads were replaced with Martini “anti-freeze” particles to avoid possible freezing of the water in the confined geometry of the simulation box. The system was equilibrated for 85 ns with 0.52 nm distance restraints with a force constant of 100 kJ/nm2_{between backbone beads of}

neigh-bouring peptides. Afterwards the fibre was simulated for 1 µs without distance restraints. Separately, a single trimeric macrocycle was solvated in a box of 6.4 × 5.1 × 5.8 nm, together with counter ions (3 Na+_{, 6 Cl}-_{); 79 mM) and 10 %}

“anti-freeze” particles. The system was equilibrated for 75 ns, before a 1 µs production simulation. 400 binding simulations were performed. They were set up by taking a random frame from the 1 µs fibre simulation and from the 1 µs macrocycle sim-ulation. The macrocycle was inserted in the box of the fibre at a random place in the XY plane at the middle for the fibre at 2 to 2.5 nm from the fibre surface. The original solvent was removed, the resulting structure was resolvated and counter ions were added (99 Na+, 198 Cl-; 269 mM) together with 10 % “anti-freeze” par-ticles. The system was equilibrated for 5 ns with 100 kJ/nm2 _{position restraints}

on the backbone beads of both the fibre and the macrocycle. The system was then simulated for 500 ns without restraints. To generate time-dependent density plots the density of all macrocycle atoms was averaged in blocks of 2.5 ns / 25 frames. Coarse-grained simulations were performed using GROMACS versions 5.1 and 2018 [29]. In all cases the fibre was maintained at its original orientation and position along the z-axis by means of roto-translational center of mass mo-tion removal as for the atomistic simulamo-tions. A timestep of 10 fs was used. The production runs were performed in the NPT ensemble with the velocity-rescaling thermostat [31] (τT = 1.0 ps) and the Parrinello-Rahman barostat [32] (τp= 36 ps)

keeping the temperature at 298 K and the pressure at 1.0 bar, respectively. Other simulation parameters used are described by De Jong et al. [33]

(17)

2.5 Supplementary material

(a) AFM micrographs over time of a growing fibre

(b) Area of 13/14

aggregates

(c) Height of the area defined in (b) Fibre lengt h (nm) R elati ve v olume of aggrega te

(d) Fibre length and volume of 13/14aggregates over time

Figure S2.1: The material consumed in fibre growth comes from the 13/14aggregates. (a)

Snapshots of a video, recorded with HS-AFM, of a single fibre growing over time. The kimograph of panel (d) was recording following the green dotted line. (b) Definition of the area of 13/14 for image processing. From left to right, the background was masked

and the area highlighted in black was selected for the measurements in the other panels. (c) Height distribution histogram (bin size = 0.08 nm) of the aggregate shown in (b). (d) Kymograph (top) and volume of precursors (bottom) of a fibre growing over time. The kymograph shows the height of the line drawn in panel (a) over time, showing (from dark to bright) background, fibres, and aggregates.

(18)

2.5. SUPPLEMENTARY MATERIAL

(a) AFM images showing growth of a fibre, stagnation, accumulation of more precursors and new growth

(b) Line studied in (c)

Fibre lengt

h (nm)

(c) Kymograph showing the different growth stages of the 16fibre shown in

panels (a) and (c)

Figure S2.2: Attachment of a new aggregate causes 16fibre to grow again. (a) Snapshots

of an HS-AFM video showing different stages of fibre growth. The cyan arrows indicate the position of the first active aggregate, the purple arrows indicate the growing end of the fibre, and the green arrows indicate the position of the second aggregate that is formed. (c) Representative image of the growing fibre from panel A showing the line (in cyan) selected to construct the time resolved intensity kymograph in panel (c). Scale bars in (a) and (b) are 20 nm. (c) Kymograph along the line shown in (b). Dashed lines mark the different stages of the fibre: growth (1), inactivity as the precursor becomes too distant from the growing end (2), formation of a new aggregate and movement along the fibre axis (3), slow growth and accumulation of more precursors (4) and fast growth again (5). The immobile fibre that is encountered by the growing one in phase 2 was removed from the kymograph for clarity.

(19)

(a) Position of 13/14aggregates in the fibre surface over time in a bulk experiment (b) Diffusion of a 13 molecule along a 16 fibre (c) Average position of 13molecules in a 16fibre at t = 0 (d) Average position of 13molecules in a 16fibre at t = 60 ns

(e) Average position of 13

molecules in a 16fibre at

t = 500 ns

Figure S2.3: Diffusion of 13/14towards the fibre ends in absence of surface. (a)

Represen-tative AFM images of fibres taken at different times during an experiment mixing 16and

13/14(~85 % oxidation). Scale bar 50 nm. In green insets show the cross-section side view

of the fibre. The arrows indicate the position of the 13/14aggregates. (b) Simulated coarse

grain (CG) fibre structure (16 16s, 8 nm) with one molecule of 13showing the diffusion at

different times over a 500 ns simulation. Different colours (from red to blue) indicate the position of the 13molecule at different times (from 0 to 500 ns). The fibre is represented as

a grey surface. (c-e) CG MD simulation of a 16fibre (outlined in black) and single 13

mole-cules over time. The relative density of 13was averaged over 400 simulations at the start of

the simulation (0 ns), after 60 ns and after 500 ns, respectively. The color bar representing the density plots in c-e was normalized from 0 to 1.

(20)

2.5. SUPPLEMENTARY MATERIAL

(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

Figure S2.4: Representative AFM images of 16fibres immobilized on the surface, in

pres-ence of 2.3 mM precursors (in 1 units) after 0 (a), 30 (b), and 60 (c) minutes of incubation at room temperature. Histogram showing the length of fibres with precursors attached in different positions at 0 (d), 30 (e) minutes and 60 (f) minutes. >200 fibres were measured in different pictures for each time point. Relative frequency of fibres having precursors attached to their fibre ends, middle, and not at all at 0 (g), 30 (h) and 60 (i) minutes. The error bar in panels C, F and I represent the standard deviation of the distribution. Statisti-cal analysis using a two-sample Kolmogorov-Smirnov test confirmed that at t=60 minutes the lengths of fibres with precursors attached to their ends are significantly different from those with precursors attached to their middle (p-values < 5 × 10-7_{). Evaluating the}

elon-gation of fibres between t = 0, 30 and 60 minutes confirmed that only the fibres with pre-cursors attached at their ends changed significantly (p-values < 3 × 10-7), while the group of fibres with precursors attached at the middle and that without precursors did not sig-nificantly grew in length (p-values > 5 × 10-2). The maximum distance between aggregate and fibre end used to classify a fibre as "precursors near the end" was 20 nm.

(21)

1 10 100 0 50 100 Counts Diameter (nm) 2.8 3.3 4.0 3 ± 2 nm (10 measurements)

(a) DLS of aggregates of 13/14in solution

and their average diameter

N O O N 600 620 640 660 680 700 720 740 0 5 10 15 20 25 Fluoresc ence intensi ty Wavelength (nm) Buffer Nile Red+ 1 ₃/1 ₄ Nile Red 1 ₃/1 ₄ λmax = 640 nm λmax = 657 nm

(b) Fluorescence of Nile Red in presence and absence of 13/14 0 10 20 30 40 625 630 635 640 645 650 655 W av elen gth (nm ) 20 40 60 80 100 120 Fluo rescenc e inte nsity (a.u .) CAC = 23 ± 5 µM (3 repeats) [13/14] (µM)

(c) Intensity and peak emission wavelength (λmax) of Nile

Red in presence of increasing concentrations of 13/14

Figure S2.5: (a) Size distribution of a precursor sample, as measured by DLS. The accu-racy of DLS is relatively low for the size range measured, so 10 individual measurements were recorded (from 10 repeats each) and averaged, obtaining a diameter of 3 ± 2 nm. (b) Fluorescence spectra of Nile Red in presence and absence of 13/14(15 µM, in 1 units).

(c) Titration of Nile Red with increasing concentrations of 1314, showing both the

inten-sity and peak wavelength of its fluorescence band. The titration was repeated three times, measuring the CAC as the point when the band started to increase in intensity and blue-shifted directly. The concentration of Nile Red in (b) and (c) was 15 µM, in 1 units.

(22)

2.6. FURTHER READING

2.6

2.7 Acknowledgements

Sourav Maity performed AFM experiments and data analysis. Jim Ottelé and Guillermo Monreal Santiago performed kinetic analyses by UPLC and ThT fluo-rescence. Guillermo Monreal Santiago and Marc C. A. Stuart performed EM anal-ysis. Guillermo Monreal Santiago performed DLS and fluorescence. Peter Kroon performed coarse-grain MD simulations. The original publication was written by Sourav Maity and Jim Ottelé, with all authors discussing the results and com-menting on the manuscript. The version of the manuscript shown in this thesis was edited by Guillermo Monreal Santiago. Pim Frederix and Omer Markovitch are gratefully acknowledged for performing atomistic MD simulations and ki-netic modelling, which can be found in the original publication. Sourav Maity, Jim Ottelé, Sijbren Otto, and Walter H. Roos designed and implemented the study, which was supervised by Wouter H. Roos, Sijbren Otto, Siewert J. Marrink and Sourav J.Maity.

2.8 References

[1] Aida, T.; Meijer, E. W.; Stupp, S. I. Functional supramolecular polymers. Sci-ence 2012, 335, 813–817.

[2] Zhang, S. Emerging biological materials through molecular self-assembly. Biotechnol. Adv. 2002, 20, 321–339.

[3] Grzelczak, M.; Vermant, J.; Furst, E. M.; Liz-Marzán, L. M. Directed self-assembly of nanoparticles. ACS Nano 2010, 4, 3591–3605.

(23)

[4] Kumar, M.; Ing, N. L.; Narang, V.; Wijerathne, N. K.; Hochbaum, A. I.; Ulijn, R. V. Amino-acid-encoded biocatalytic self-assembly enables the for-mation of transient conducting nanostructures. Nat. Chem. 2018, 10, 696–703. [5] Ter Huurne, G. M.; De Windt, L. N.; Liu, Y.; Meijer, E. W.; Voets, I. K.; Pal-mans, A. R. Improving the folding of supramolecular copolymers by control-ling the assembly pathway complexity. Macromolecules 2017, 50, 8562–8569. [6] Fukui, T.; Kawai, S.; Fujinuma, S.; Matsushita, Y.; Yasuda, T.;

Saku-rai, T.; Seki, S.; Takeuchi, M.; Sugiyasu, K. Control over differentiation of a metastable supramolecular assembly in one and two dimensions. Nat. Chem.

2017, 9, 493–499.

[7] Milhiet, P. E.; Yamamoto, D.; Berthoumieu, O.; Dosset, P.; le Grimellec, C.; Verdier, J. M.; Marchal, S.; Ando, T. Deciphering the structure, growth and assembly of amyloid-like fibrils using high-speed atomic force microscopy. PLoS One 2010, 5.

[8] Mattia, E.; Otto, S. Supramolecular systems chemistry. 2015.

[9] Onogi, S.; Shigemitsu, H.; Yoshii, T.; Tanida, T.; Ikeda, M.; Kubota, R.; Hamachi, I. In situ real-time imaging of self-sorted supramolecular nanofi-bres. Nat. Chem. 2016, 8, 743–752.

[10] Da Silva, R. M.; Van Der Zwaag, D.; Albertazzi, L.; Lee, S. S.; Meijer, E. W.; Stupp, S. I. Super-resolution microscopy reveals structural diversity in molecular exchange among peptide amphiphile nanofibres. Nat. Commun.

2016, 7.

[11] Parent, L. R.; Bakalis, E.; Ramírez-Hernández, A.; Kammeyer, J. K.; Park, C.; De Pablo, J.; Zerbetto, F.; Patterson, J. P.; Gianneschi, N. C. Directly observing micelle fusion and growth in solution by liquid-cell transmission electron microscopy. J. Am. Chem. Soc. 2017, 139, 17140–17151.

[12] Ando, T.; Uchihashi, T.; Scheuring, S. Filming biomolecular processes by high-speed atomic force microscopy. 2014.

[13] Kodera, N.; Yamamoto, D.; Ishikawa, R.; Ando, T. Video imaging of walking myosin V by high-speed atomic force microscopy. 2010, 2–6.

[14] Colom, A.; Redondo-Morata, L.; Chiaruttini, N.; Roux, A.; Scheuring, S. Dy-namic remodeling of the dynamin helix during membrane constriction. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 5449–5454.

[15] Maity, S.; Caillat, C.; Miguet, N.; Sulbaran, G.; Effantin, G.; Schoehn, G.; Roos, W. H.; Weissenhorn, W. VPS4 triggers constriction and cleavage of ESCRT-III helical filaments. Sci. Adv. 2019, 5.

(24)

2.8. REFERENCES

[16] Fukui, T.; Uchihashi, T.; Sasaki, N.; Watanabe, H.; Takeuchi, M.; Sugiyasu, K. Direct observation and manipulation of supramolecular polymerization by high-speed atomic force microscopy. Angew. Chemie - Int. Ed. 2018, 57, 15465– 15470.

[17] Wickham, S. F.; Endo, M.; Katsuda, Y.; Hidaka, K.; Bath, J.; Sugiyama, H.; Turberfield, A. J. Direct observation of stepwise movement of a synthetic molecular transporter. Nat. Nanotechnol. 2011, 6, 166–169.

[18] Carnall, J. M.; Waudby, C. A.; Belenguer, A. M.; Stuart, M. C.; Peyralans, J. J.; Otto, S. Mechanosensitive self-replication driven by self-organization. Sci-ence 2010, 327, 1502–1506.

[19] Pal, A.; Malakoutikhah, M.; Leonetti, G.; Tezcan, M.; Colomb-Delsuc, M.; Nguyen, V. D.; Van Der Gucht, J.; Otto, S. Controlling the structure and length of self-synthesizing supramolecular polymers through nucleated growth and disassembly. Angew. Chemie - Int. Ed. 2015, 54, 7852–7856. [20] 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. ACS Nano 2017, 11, 7858–7868.

[21] Törnquist, M.; Michaels, T. C.; Sanagavarapu, K.; Yang, X.; Meisl, G.; Co-hen, S. I.; Knowles, T. P.; Linse, S. Secondary nucleation in amyloid forma-tion. Chem. Commun. 2018, 54, 8667–8684.

[22] Altay, Y.; Altay, M.; Otto, S. Existing self-replicators can direct the emergence of new ones. Chem. - A Eur. J. 2018, 24, 11911–11915.

[23] Altay, M.; Altay, Y.; Otto, S. Parasitic behavior of self-replicating molecules. Angew. Chemie - Int. Ed. 2018, 57, 10564–10568.

[24] Rueden, C. T.; Schindelin, J.; Hiner, M. C.; DeZonia, B. E.; Walter, A. E.; Arena, E. T.; Eliceiri, K. W. ImageJ2: ImageJ for the next generation of sci-entific image data. BMC Bioinformatics 2017, 18.

[25] Uchihashi, T.; Kodera, N.; Ando, T. Guide to video recording of structure dynamics and dynamic processes of proteins by high-speed atomic force mi-croscopy. Nat. Protoc. 2012, 7, 1193–1206.

[26] Ando, T.; Kodera, N.; Naito, Y.; Kinoshita, T.; Furuta, K.; Toyoshima, Y. Y. A high-speed atomic force microscope for studying biological macromolecules in action. ChemPhysChem 2003, 4, 1196–1202.

(25)

[27] Stuart, M. C. A.; Van De Pas, J. C.; Engberts, J. B. F. N. The use of Nile Red to monitor the aggregation behavior in ternary surfactant-water-organic sol-vent systems. J. Phys. Org. Chem. 2005, 18, 929–934.

[28] Marrink, S. J.; Risselada, H. J.; Yefimov, S.; Tieleman, D. P.; De Vries, A. H. The MARTINI force field: Coarse grained model for biomolecular simula-tions. J. Phys. Chem. B 2007, 111, 7812–7824.

[29] Abraham, M. J.; Murtola, T.; Schulz, R.; Páll, S.; Smith, J. C.; Hess, B.; Lin-dah, E. Gromacs: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 2015, 1-2, 19–25. [30] De Jong, D. H.; Singh, G.; Bennett, W. F.; Arnarez, C.; Wassenaar, T. A.; Schäfer, L. V.; Periole, X.; Tieleman, D. P.; Marrink, S. J. Improved parame-ters for the martini coarse-grained protein force field. J. Chem. Theory Comput.

2013, 9, 687–697.

[31] Bussi, G.; Donadio, D.; Parrinello, M. Canonical sampling through velocity rescaling. J. Chem. Phys. 2007, 126.

[32] Parrinello, M.; Rahman, A. Polymorphic transitions in single crystals: A new molecular dynamics method. J. Appl. Phys. 1981, 52, 7182–7190.

[33] De Jong, D. H.; Baoukina, S.; Ingólfsson, H. I.; Marrink, S. J. Martini straight: Boosting performance using a shorter cutoff and GPUs. Comput. Phys. Com-mun. 2016, 199, 1–7.

[34] Maity, S.; Ottelé, J.; Monreal Santiago, G.; Frederix, P. W.; Kroon, P. C.; Markovitch, O.; Stuart, M. C. A.; Marrink, S. J.; Otto, S.; Roos, W. H. Caught in the act: Mechanism of supramolecular polymerization driven self-replication visualized in real time. Submitted.