University of Groningen Structure-Function Relationships in Dynamic Combinatorial Libraries Altay, Meniz

(1)

University of Groningen

Structure-Function Relationships in Dynamic Combinatorial Libraries

Altay, Meniz

DOI:

10.33612/diss.90038152

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: 2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Altay, M. (2019). Structure-Function Relationships in Dynamic Combinatorial Libraries. University of Groningen. https://doi.org/10.33612/diss.90038152

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)

(3)

120 4. Parasitic Behavior of Self-Replicating Molecules

4.1 Introduction

T

he origins of life1–4_{and the de-novo synthesis of life are among the grand}

chal-lenges in contemporary science and an important focus in systems chemistry.5–7

Self-replicating systems8–12_{play a key role in scenarios of the origin of life and are a}

promising possible starting point for the de-novo synthesis of life. Self-replicating molecules have been developed based on DNA,13–15 _RNA,16–22 _peptides23–26 _or

synthetic molecules.27–32 The majority of these systems feature only a single self-replicating molecule. Yet approaches to the origin of life and its de-novo synthesis will inevitably involve systems in which multiple replicators co-exist and interact. Evolution involves the selection from among competing replicators and, most likely, also cooperation between replicators. Thus the dynamics that can occur in systems of co-existing replicators are an important new focus in the development of systems of self-replicators towards life. Early work in this field involves systems of replicators based on RNA,22,33 _{α-helical peptides}24,34–37 _{and synthetic replicators.}28,30,31,38

Using a dynamic combinatorial approach to self-replication,30,31,38_{based on}

pseu-dopeptide building blocks39_{we recently developed replicating cyclic disulfide oligomers}

made from dithiol building blocks.40–42 _{In a typical DCL made from an individual}

building block such as dithiol 1 (Figure 4.1a), as oxidation takes place, a mixture of diﬀerently sized macrocycles forms, that continuously interconvert through thiol-disulfide exchange (Figure 4.1b).43,44 _{When one of the library members is able to}

bind to copies of itself, this compound is stabilized and the equilibrium shifts to-wards more of that macrocycle, resulting in self-replication and the formation of stacks of the replicator. Mechanical energy can break the stacks, thereby increasing the number of ends from which the stacks grow and enabling exponential replica-tion.45,46

Unlike most other replicators, in these combinatorial systems, the structure of the building blocks does not predetermine the nature of the replicator that emerges. The ring size and building block composition of newly formed replicators are also influ-enced by mechanical agitation,40 _{the solvent environment}47 _{and pre-existing}

repli-cators.48,49 _{The latter studies revealed mechanisms of co-operation and co-existence}

by which replicators diversify and assist in each other’s formation. We now report an example where a set of newly formed replicators exhibits exactly the opposite: parasitic behavior. Emergence of the parasitic replicator relies on cross-catalysis by a structurally closely related pre-existing replicator, which is subsequently consumed by the very replicators that it brought into existence. While several reports describe the emergence of parasites in systems where enzymes mediate replication of nucleic acids,50–52 _{this is the first report of the emergence of a parasite in a system of}

(4)

Figure 4.1: a) Chemical structures of the building blocks utilized in cross-seeding exper-iments. b) Cartoon representation for the general replication mechanism for a particular building block (18 in this case). c) Proposed mechanism for the emergence of the parasitic

replicator (1n26-n) in a DCL made from building block 2 upon cross-seeding with 18. First,

a small dynamic combinatorial library of cyclic disulfides is made from oxidation of building block 2. While the cross-seed dissociates from one end, stacking of rings of one particular size (1n26-n) shifts the equilibrium in the direction of these library members. Agitation breaks

the stacks producing more ends from which the stacks can grow, giving rise to exponential replication.

(5)

4.2 Results and Discussion

We used two closely related building blocks 1 and 2 (Figure 4.1a) featuring two thiol units to promote covalent thiol-disulfide exchange and a short peptide chain composed of alternating hydrophilic and hydrophobic amino acids to promote self-assembly through β-sheet formation. As we reported previously, building block 1 spontaneously forms a self-replicating cyclic octamer (18).41 Building block 2

con-tains an additional methylene in the amino acid that connects the peptide to the aromatic dithiol core. We reasoned that this modification would make nucleation of any replicators formed from 2 more diﬃcult by increasing the degrees of freedom in the peptide chain. Indeed, in contrast to building block 1 and most previously stud-ied peptide-based building blocks in this family,[12,14-16] the spontaneous emergence

of replicators from DCLs made from building block 2 was sluggish. When a DCL (1.0 mM in 2 in 50 mM borate buﬀer, pH=8.2) was exposed to air under constant mechanical agitation, cyclic trimers (23) and tetramers (24) emerged as the main

products (Figure 4.2a). Repeating this experiment at a constant oxidation level (65 %, ensuring suﬃcient free thiol to mediate disulfide exchange) yielded <9% cyclic hexamer replicator (26) after two months (see Figure 4.14; for evidence that 26 is a

self-replicator, vide infra).

Given that replicators derived from building block 1 assemble readily into fibers,40

we investigated whether these fibers could act as templates and cross-catalyze the formation of replicators from building block 2. Thus, we first prepared a DCL by dissolving 2 in aqueous borate buﬀer (50 mM, pH 8.2) to a concentration of 1.0 mM. After 24 h of stirring in the presence of air the library had oxidized to approximately 75%. We then added 0.2 mol eq. (with respect to building block) of replicator 18

and monitored the library composition over nine days by UPLC.53 _{A set of cyclic}

hexamer replicators 1n26-nemerged rapidly and grew to dominate the mixture after

four days (Figure 4.2b). Repeating this experiment using 0.5 mol eq. of replicator 18 led the somewhat faster emergence of 1n26-n (Figure 4.2c), suggesting a

cross-catalytic role of 18. To confirm that the emergence of the hexameric replicators was

indeed promoted by 18 we set up a negative control experiment from an equimolar

mixture of 1 and 2 ([1]=[2]=0.5 mM) to which we did not add any 18. We did not

observe any cyclic hexamers in this sample even after seven days (see Figure 4.27). Remarkably, the emergence of the set of hexameric replicators is accompanied by diminishing of the amount of 18, to the point that this replicator was no longer

detectable after 3 days in the experiments shown in Figure 4.2b and c. Repeating the experiment with 1.0 mol eq. 18 confirmed this behavior, although a small amount

of 18 was still left at the point that the disulfide exchange ceased due to complete

(6)

repli-Figure 4.2: Product distribution over time monitored by UPLC for agitated DCLs that are a) non-seeded; mixed with b) 0.2 mol eq. 18 on day 1, c) 0.5 mol eq. 18and d) 1.0 mol

eq. 18on day 0.

cators act as parasites: they grow at the expense of the original octameric replicators to which they owe their existence. This conclusion was supported by MS analysis of the UPLC peak that contains the co-eluting 1n26-n macrocycles with diﬀerent

com-position (see Figure 4.53, Figure 4.56, Figure 4.61) including up to six units of 2. In contrast, no mixed cyclic octamers (1n28-n) could be detected in the experiments

shown in Figure 4.2.

In order to prove that 1n26-n, including 16, are replicators and to compare their

replication eﬃciencies, we performed a set of serial transfer seeding experiments (Fig-ure 4.3). A second generation sample was prepared by transferring an aliquot (0.2 mol eq.) of the sample corresponding to Figure 4.2b to a DCL made from building block 2. Finally, a third generation sample was prepared by transferring 0.2 mol eq. from the second-generation sample to a fresh DCL prepared from 2.

(7)

Figure 4.3: a) Schematic representation of the serial transfer seeding experiments. Product distribution over time monitored by UPLC for DCLs that are b) second-generation and c) third-generation samples. The product distribution for the first-generation sample is shown in Figure 4.2b.

Through these serial transfer experiments, we were able to obtain almost pure 26

in the second generation as the mass spectrum shows (see Figure 4.61). Therefore, the third generation seeding mainly probes the autocatalytic behavior of 26. After

12 days, 26 amounted to 40 % of the overall library composition. Comparing these

data with that for the spontaneous emergence of 26 (Figure 4.2a) shows that 26 is

indeed a replicator. However, comparing the kinetic data for the growth of 1n26-nin

samples with decreasing content of 1 shows that 26 is a less eﬃcient replicator than

the set of mixed-building-block 1n26-n replicators.

Notable in these seeding experiments is the absence of any octamer replicators (1n28-n). So cross-catalysis appears to be strictly unidirectional: octamers promote

the formation of hexamers but not the other way around.54 _{This conclusion was}

confirmed in experiments in which we added 0.2 mol eq. 26 or 1n26-n as seed to

an agitated DCL made from building block 1 (1.0 mM). After 5 days the library composition was dominated by trimers and tetramers and no 18 was detected (see

(8)

nated DCLs (Figure 4.4b). The β-sheet structure is more pronounced in second generation serial transfer samples dominated by 26 than in samples of 18.

Analy-sis by TEM showed that 18 formed laterally associated short fibers (around 100 nm)

(Figure 4.4c). In the course of the serial transfer experiments the average fiber length increased to around 150 nm for the first-generation replicators and to 350 nm for the

Figure 4.4: a) CD spectra (recorded at identical concentrations), b) Thioflavin T emission spectra for DCLs made from only cross-seed 18, from peptide 2 without cross-seed and the

first and the second generation of seeding. TEM micrographs for c) cross-seed 18, d) first

(9)

second generation. We tentatively attribute the increased fiber length and enhanced β-sheet structure of fibers of 26 (as compared to those of 18) to the stronger

hy-drophobic interactions within the stacks arising from the additional methylene unit in 2 as compared to 1. Since the rate of replication depends on the number fiber ends,45_{and since longer fibers means fewer fiber ends, the increase in length of fibers}

of 26can (partially) account for the reduced rate of replication of these fibers relative

to those of 18.

Finally, we investigated the extent to which structurally related peptides are able to show similar cross-catalytic eﬀects. First, we probed whether replicators other than 18 can also induce the formation of replicators from building block 2. We

seeded DCLs from 2 with 0.2 mol eq. of replicators 46, 56and 65but failed to detect

any replicators, despite the fact that the ring size of the replicator seeds 46 and 56

now matches the ring size of the 26 replicators, while the spacer length in 6 matches

that in 2. Only mixtures of trimers and tetramers coexisting with the seeds were obtained in these seeding experiments (see Figure 4.8). Second, we investigated the eﬀect of elongating the spacer in 2 by an additional methylene unit to give building block 3. We prepared DCLs from building block 3 and seeded these with 0.2 mol eq. 18or 26. Again, we did not observe the emergence of any new replicators (see Figure

4.9). Thus, it appears that cross-catalysis of formation of 1n26-nby 18 is specific to

these particular peptide sequences.

4.3 Conclusion

In conclusion, we observed how a set of 6-ring replicators emerged, aided by a pre-existing 8-ring replicator, only to consume the 8-rings to which the new replicator owed its existence. We speculate that the 6-ring replicator fibers nucleate at some of the ends of the 8-ring replicator fibers (Figure 4.1c). The fact that the 8-ring replicators are eﬃciently broken down (a process that occurs at the fiber ends),46

suggests that the 8-ring fibers remain exposed to the solution at at least one of their fiber ends (i.e. for most 8-ring fibers not more than one fiber end is capped with 6-ring replicators). This behavior is reminiscent of parasitic behavior as it occurs in biology: the set of six-ring replicators benefit from cross-catalysis by the 8-ring replicator in a non-mutualistic way as the 6-ring replicators do not cross-catalyze the formation of 8-ring replicator. Like in biology and in previous RNA-based systems50–52 _the

parasite is smaller than its host (albeit not much), replicates faster and extracts resources (building blocks) from its host, causing it some harm. Unlike in biology, the host replicator nor its parasite has a metabolism. Notably, the 6-ring replicator even causes the (partial) demise of the 8-ring replicator and utilizes the building block that were previously contained in the 8-rings for its own growth, which starts to resemble

4.4. Acknowledgements 127

predatory behavior. These unique observations illustrate the rich dynamics that multi-replicator systems can exhibit. Appreciating and understanding such dynamics is essential for directing the evolution of multi-replicator systems towards the de-novo synthesis of life.

4.4 Acknowledgements

(10)

seeded DCLs from 2 with 0.2 mol eq. of replicators 46, 56and 65but failed to detect

any replicators, despite the fact that the ring size of the replicator seeds 46 and 56

now matches the ring size of the 26 replicators, while the spacer length in 6 matches

that in 2. Only mixtures of trimers and tetramers coexisting with the seeds were obtained in these seeding experiments (see Figure 4.8). Second, we investigated the eﬀect of elongating the spacer in 2 by an additional methylene unit to give building block 3. We prepared DCLs from building block 3 and seeded these with 0.2 mol eq. 18or 26. Again, we did not observe the emergence of any new replicators (see Figure

4.9). Thus, it appears that cross-catalysis of formation of 1n26-n by 18 is specific to

these particular peptide sequences.

4.3 Conclusion

In conclusion, we observed how a set of 6-ring replicators emerged, aided by a pre-existing 8-ring replicator, only to consume the 8-rings to which the new replicator owed its existence. We speculate that the 6-ring replicator fibers nucleate at some of the ends of the 8-ring replicator fibers (Figure 4.1c). The fact that the 8-ring replicators are eﬃciently broken down (a process that occurs at the fiber ends),46

suggests that the 8-ring fibers remain exposed to the solution at at least one of their fiber ends (i.e. for most 8-ring fibers not more than one fiber end is capped with 6-ring replicators). This behavior is reminiscent of parasitic behavior as it occurs in biology: the set of six-ring replicators benefit from cross-catalysis by the 8-ring replicator in a non-mutualistic way as the 6-ring replicators do not cross-catalyze the formation of 8-ring replicator. Like in biology and in previous RNA-based systems50–52 _the

parasite is smaller than its host (albeit not much), replicates faster and extracts resources (building blocks) from its host, causing it some harm. Unlike in biology, the host replicator nor its parasite has a metabolism. Notably, the 6-ring replicator even causes the (partial) demise of the 8-ring replicator and utilizes the building block that were previously contained in the 8-rings for its own growth, which starts to resemble

(11)

4.5 Materials and Methods

In all experiments, doubly distilled water was used. For preparing buﬀer solutions, boric anhydride was purchased from Sigma Aldrich and sodium hydroxide (Merck Chemicals) was utilized for pH adjustment when necessary. Peptides were purchased from Cambridge Peptides Ltd. (Birmingham, UK) with_{ě95 % purity. During UPLC} measurements, UPLC grade water, acetonitrile and trifluoroacetic acid (Biosolve BV) were used. For sample dilutions peptide synthesis grade dimethyl formamide (Bio-solve BV) was used.

4.5.1 Library and Sample Preparation

Building blocks were dissolved to a concentration of 1.0 mM in borate buﬀer (50 mM, pH 8.2). Libraries were equilibrated in HPLC vials (12_{ˆ32 mm) with Teflon caps} and were left stirring on an IKA RCT hot plate stirrer at 1200 rpm.

Prior to UPLC and LC-MS analyses, 10µL from each (1.0 mM) sample was diluted 5 times with 5µL dimethyl formamide and 35 µL doubly distilled water. The same sampling protocol was followed for all UPLC and LC-MS analyses.

4.5.2 Seeding Experiments

For the cross-seeding experiments, DCLs were prepared by dissolving peptides 2 and 3 separately in borate buﬀer (50 mM, pH 8.2) to a concentration of 1.0 mM. DCLs were stirred overnight in the presence of oxygen from the air, ensuring that oxidation levels with respect to building blocks wereě75 %. Then varying (0.2 to 1.0 mol eq.) amounts of seed from libraries dominated by 18, 46, 56, or 65 was added.

For the reverse cross-seeding experiments, DCLs were prepared by dissolving pep-tide 1 in borate buﬀer (50 mM, pH 8.2) to a concentration of 1.0 mM and oxidizing at least 80% of the thiols by adding sodium perborate (40 mM in water) to give a DCL that contained mostly monomer (11), trimer (13), and tetramer (14). Then 0.2

mol eq. 26or 1n26-nwas added as cross-seed.

4.5.3 Circular Dichroism

Samples were diluted to a concentration of 0.2 mM with respect to building block concentration. Spectra were recorded at room temperature by using a JASCO J715 spectrophotometer and HELMA quartz cuvettes with a path length of 1 mm. All spectra were recorded at room temperature from 190 nm to 300 nm, with 2 nm step interval and 3 scans with a speed of 200 nm/min. Solvent spectra were subtracted from all spectra.

4.5. Materials and Methods 129

4.5.4 Thioflavin T (ThT) Fluorescence Assay

A ThT stock solution (2.2 mM) was prepared in phosphate buffer (50 mM with 150 mM NaCl, pH 8.2). A working solution was prepared by diluting the stock solution to 22µm. The fluorescence intensity of a 450 µL ThT solution (blank) was measured by excitation at 440 nm (5 nm slit width) and emission between 480 -700 nm (5 nm slit width). An aliquot of 80µL of the 100 µm peptide solution was added to HELMA 10ˆ2 mm quartz cuvette and incubated for 2 minutes before the measurement. All fluorescence measurements were performed by averaging 3 repeats on a JASCO FP6200 spectrophotometer with 480 nm cut-off filter on the emission channel to avoid high order diffractions resulting from the excitation.

4.5.5 Transmission Electron Microscopy (TEM)

All images were obtained after negatively staining the sample on the grid. Samples were diluted 20-fold prior to deposition. A small aliquot (5µL) of diluted samples was deposited on a 400-mesh copper grid covered with a thin carbon film (Agar Scientific). After 30 s of deposition, the droplet was gently blotted on filter paper. The sample was stained twice (4µL each time) with a 2% uranyl acetate solution and blotted again on filter paper after deposition (30 s each time). The grids were observed on a Philips CM12 transmission electron microscope operating at 120 kV. All images were recorded on a slow scan CCD camera.

4.5.4 Thioflavin T (ThT) Fluorescence Assay

A ThT stock solution (2.2 mM) was prepared in phosphate buffer (50 mM with 150 mM NaCl, pH 8.2). A working solution was prepared by diluting the stock solution to 22µm. The fluorescence intensity of a 450 µL ThT solution (blank) was measured by excitation at 440 nm (5 nm slit width) and emission between 480 -700 nm (5 nm slit width). An aliquot of 80µL of the 100 µm peptide solution was added to HELMA 10_{ˆ2 mm quartz cuvette and incubated for 2 minutes before the} measurement. All fluorescence measurements were performed by averaging 3 repeats on a JASCO FP6200 spectrophotometer with 480 nm cut-off filter on the emission channel to avoid high order diffractions resulting from the excitation.

4.5.5 Transmission Electron Microscopy (TEM)

4.5.4 Thioflavin T (ThT) Fluorescence Assay

4.5.5 Transmission Electron Microscopy (TEM)

4.5.4 Thioflavin T (ThT) Fluorescence Assay

4.5.5 Transmission Electron Microscopy (TEM)

(12)

4.5 Materials and Methods

In all experiments, doubly distilled water was used. For preparing buﬀer solutions, boric anhydride was purchased from Sigma Aldrich and sodium hydroxide (Merck Chemicals) was utilized for pH adjustment when necessary. Peptides were purchased from Cambridge Peptides Ltd. (Birmingham, UK) with_{ě95 % purity. During UPLC} measurements, UPLC grade water, acetonitrile and trifluoroacetic acid (Biosolve BV) were used. For sample dilutions peptide synthesis grade dimethyl formamide (Bio-solve BV) was used.

4.5.1 Library and Sample Preparation

Building blocks were dissolved to a concentration of 1.0 mM in borate buﬀer (50 mM, pH 8.2). Libraries were equilibrated in HPLC vials (12_{ˆ32 mm) with Teflon caps} and were left stirring on an IKA RCT hot plate stirrer at 1200 rpm.

Prior to UPLC and LC-MS analyses, 10µL from each (1.0 mM) sample was diluted 5 times with 5µL dimethyl formamide and 35 µL doubly distilled water. The same sampling protocol was followed for all UPLC and LC-MS analyses.

4.5.2 Seeding Experiments

For the cross-seeding experiments, DCLs were prepared by dissolving peptides 2 and 3 separately in borate buﬀer (50 mM, pH 8.2) to a concentration of 1.0 mM. DCLs were stirred overnight in the presence of oxygen from the air, ensuring that oxidation levels with respect to building blocks wereě75 %. Then varying (0.2 to 1.0 mol eq.) amounts of seed from libraries dominated by 18, 46, 56, or 65 was added.

For the reverse cross-seeding experiments, DCLs were prepared by dissolving pep-tide 1 in borate buﬀer (50 mM, pH 8.2) to a concentration of 1.0 mM and oxidizing at least 80% of the thiols by adding sodium perborate (40 mM in water) to give a DCL that contained mostly monomer (11), trimer (13), and tetramer (14). Then 0.2

mol eq. 26or 1n26-n was added as cross-seed.

4.5.3 Circular Dichroism

Samples were diluted to a concentration of 0.2 mM with respect to building block concentration. Spectra were recorded at room temperature by using a JASCO J715 spectrophotometer and HELMA quartz cuvettes with a path length of 1 mm. All spectra were recorded at room temperature from 190 nm to 300 nm, with 2 nm step interval and 3 scans with a speed of 200 nm/min. Solvent spectra were subtracted from all spectra.

4.5.4 Thioflavin T (ThT) Fluorescence Assay

4.5.5 Transmission Electron Microscopy (TEM)

added to HELMA 10_{ˆ2 mm quartz cuvette and incubated for 2 minutes before the} measurement. All fluorescence measurements were performed by averaging 3 repeats on a JASCO FP6200 spectrophotometer with 480 nm cut-oﬀ filter on the emission channel to avoid high order diﬀractions resulting from the excitation.

4.5.5 Transmission Electron Microscopy (TEM)

solution to 22µm. The fluorescence intensity of a 450 µL ThT solution (blank) was measured by excitation at 440 nm (5 nm slit width) and emission between 480 -700 nm (5 nm slit width). An aliquot of 80µL of the 100 µm peptide solution was added to HELMA 10ˆ2 mm quartz cuvette and incubated for 2 minutes before the measurement. All fluorescence measurements were performed by averaging 3 repeats on a JASCO FP6200 spectrophotometer with 480 nm cut-oﬀ filter on the emission channel to avoid high order diﬀractions resulting from the excitation.

4.5.5 Transmission Electron Microscopy (TEM)

4.5.4 Thioflavin T (ThT) Fluorescence Assay

4.5.5 Transmission Electron Microscopy (TEM)

(13)

4.5.6 Kinetic profiles

Figure 4.5: a) Relative UPLC peak area (%) for species in an agitated DCL prepared from equimolar ([1]=[2]=0.5 mM) mixture of peptides 1 and 2. Initially the DCL contained mostly mixtures of cyclic trimers (1n23-n) and tetramers (1n24-n). To that mixture 0.2 mol

eq. 1n26-n (from the first generation seeding) was added at day 0 and overall distribution

was monitored for 7 days. Unlike the non-seeded DCL, 1n26-n can replicate in the presence

(14)

Figure 4.6: Total peak area of the species over days for DCLs made from peptide 2 and seeded with a) 0.2 mol eq. 18 (first generation cross-seeding), b) 0.5 mol eq. 18, c) 1.0 mol

eq. 18. Total peak area of the species over days for d) second generation cross-seeding and

E) third generation seeding (probes the auto-catalytic behaviour of 26).

Figure 4.7: a) Relative UPLC peak area (%) for species in an agitated DCL made from peptide 2 and kept under an inert atmosphere. b) Total peak area of the species over days.

(15)

Figure 4.8: Relative UPLC peak area (%) for species in DCLs prepared from peptide 2 and cross-seeded with a) 0.2 mol eq. 46, b) 0.2 mol eq. 56 and c) 0.2 mol eq. 65.

Figure 4.9: Relative UPLC peak area (%) for species in DCLs made from peptide 3 and cross-seeded with a) 0.2 mol eq. 18, b) 0.2 mol eq. mixture of hexamers (1n26-n) and c) 0.2

(16)

Figure 4.10: Relative UPLC peak area (%) for species in DCLs made from peptide 1 and cross-seeded with a) 0.2 mol eq. 18 (control sample), b) 0.2 mol eq. 1n26-nand c) 0.2 mol

eq. 26.

4.5.7 UPLC and UPLC-MS Analyses

UPLC analyses were performed on a Waters Acquity H-class instrument equipped with a diode array UV/Vis detector. UPLC-MS analyses were performed on a Xevo G2 UPLC/TOF with ESI ionization, manufactured by Waters. All analyses were performed at 35˝C using a reversed-phase UPLC column (Phenomenex Aeris Pep-tide, 2.1ˆ150 mm; 1.7 µm). UV absorbance was monitored at 254 nm. Positive-ion mass spectra were acquired using electro-spray ionization. Injection volumes were 10µL (UPLC) and 5 µL (UPLC-MS) of freshly aliquoted sample with 0.3 mL/min flow rate.

All of the DCLs were analyzed using the following method with a linear gradient: Solvent A: ULC/MS grade water (with 0.1 V/V % TFA as modifier)

Solvent B: ULC/MS grade acetonitrile (with 0.1 V/V % TFA as modifier)

Time, min. A% B% 0.00 90 10 1.00 90 10 1.30 78 22 11.00 68 32 16.00 59 41 16.50 5 95 17.00 5 95 17.50 90 10 20.00 90 10

(17)

Figure 4.11: UPLC trace (monitored at 254 nm) of the cross-seed (cyclic octamer 18)

made from peptide 1 (3.8 mM in phosphate buﬀer, pH 8.1) under constant mechanical agitation.

Figure 4.12: Mass spectrum of the cyclic octamer 18from the UPLC-MS analysis of an

agitated library made from peptide 1 (corresponding to Figure 4.11). Calculated isotopic profile for [M+4H]4+ (species, abundance): 1396.64 (M, 100%), 1396.89 (M+1, 95.54%), 1397.14 (M+2, 78.37%), 1397.39 (M+3, 57.46%), 1397.64 (M+4, 38.53%); m/z calcu-lated: 1862.14 [M+3H]3+, 1396.60 [M+4H]4+, 1117.28 [M+5H]5+; m/z observed: 1861.81 [M+3H]3+_{, 1396.64 [M+4H]}4+_{, 1117.52 [M+5H]}5+_.

(18)

Figure 4.13: UPLC traces (monitored at 254 nm) of the DCL made from peptide 2 (1.0 mM in 50 mM borate buﬀer, pH 8.2) a) at day 0 and b) at day 6 under constant mechanical agitation.

Figure 4.14: UPLC trace (monitored at 254 nm) of a DCL made from peptide 2 (1.0 mM in 50 mM borate buﬀer, pH 8.2) under constant mechanical agitation and under an inert atmosphere. The amount of 26 after 2 months was <9% of the overall library composition.

(19)

Figure 4.15: Mass spectrum of monomer 21from the UPLC-MS analysis of peptide 2

(cor-responding to Figure 4.13a). Calculated isotopic profile for [M+H]1+_{(species, abundance):}

714.39 (M, 100%), 715.40 (M+1, 61.67%); m/z calculated: 714.33 [M+H]1+_{; m/z observed:}

714.39 [M+H]1+_.

Figure 4.16: Mass spectrum of dimer 22from the UPLC-MS analysis of peptide 2

(corre-sponding to Figure 4.13a). Calculated isotopic profile for [M+2H]2+_{(species, abundance):}

713.39 (M, 100%), 713.89 (M+1, 48.01 %), 714.39 (M+2, 66.16%); m/z calculated: 1424.66

[M+H]1+_{, 713.33 [M+2H]}2+_{, 475.88 [M+3H]}3+_{; m/z observed: 1425.29 [M+H]}1+_{, 713.39}

[M+2H]2+_{, 475.06 [M+3H]}3+_.

Figure 4.17: Mass spectrum of cyclic tetramer 24 from the UPLC-MS analysis of an

ag-itated library made from peptide 2 (corresponding to Figure 4.13a). Calculated isotopic

profile (species, abundance) for [M+3H]3+_{: 950.09 (M, 100%), 950.43 (M+1, 75.77 %),}

950.76 (M+2, 50.08 %), 951.09 (M+3, 28.67 %); m/z calculated: 1424.62 [M+2H]2+_{, 949.75}

[M+3H]3+_{, 712.31 [M+4H]}4+_{; m/z observed: 1424.64 [M+2H]}2+_{, 950.09 [M+3H]}3+_{, 712.57}

[M+4H]4+_.

Figure 4.18: Mass spectrum of cyclic trimer 23from the UPLC-MS analysis of an agitated

library made from peptide 2 (corresponding to Figure 4.13a). Calculated isotopic profile

(species, abundance) for [M+3H]3+_{: 713.06 (M, 100%), 713.39 (M+1, 77.84%), 713.73}

(M+2, 47.43%), 714.06 (M+3, 23.07); m/z calculated: 1068.42 [M+2H]2+_{, 712.31 [M+3]3+,}

(20)

Figure 4.15: Mass spectrum of monomer 21from the UPLC-MS analysis of peptide 2

(cor-responding to Figure 4.13a). Calculated isotopic profile for [M+H]1+_{(species, abundance):}

714.39 (M, 100%), 715.40 (M+1, 61.67%); m/z calculated: 714.33 [M+H]1+_{; m/z observed:}

714.39 [M+H]1+_.

[M+2H]2+_{, 475.06 [M+3H]}3+_.

profile (species, abundance) for [M+3H]3+_{: 950.09 (M, 100%), 950.43 (M+1, 75.77 %),}

950.76 (M+2, 50.08 %), 951.09 (M+3, 28.67 %); m/z calculated: 1424.62 [M+2H]2+_{, 949.75}

[M+4H]4+_.

library made from peptide 2 (corresponding to Figure 4.13a). Calculated isotopic profile

(species, abundance) for [M+3H]3+_{: 713.06 (M, 100%), 713.39 (M+1, 77.84%), 713.73}

(M+2, 47.43%), 714.06 (M+3, 23.07); m/z calculated: 1068.42 [M+2H]2+_{, 712.31 [M+3]3+,}

(21)

Figure 4.19: Mass spectrum of cyclic pentamer 25 from the UPLC-MS analysis of an

ag-itated library made from peptide 2 (corresponding to Figure 4.13b). Calculated isotopic

profile (species, abundance) for [M+3H]3+_{: 11187.76 (M, 100%), 1188.09 (M+1, 71.23%),}

1188.43 (M+2, 56.98%), 1188.76 (M+3, 44.88%); m/z calculated: 1187.19 [M+3H]3+_,

890.39 [M+4H]4+_{; m/z observed: 1187.76 [M+3H]}3+_{, 891.06 [M+4H]}4+_.

Figure 4.20: Mass spectrum of cyclic hexamer 26from the UPLC-MS analysis of an

ag-itated library made from peptide 2 under an inert atmosphere (corresponding to Figure 4.13b). Calculated isotopic profile (species abundance) for [M+4H]4+_{: 1068.73 (M, 100%),}

1068.98 (M+1, 91.83%), 1069.23 (M+2, 56.59%), 1069.48 (M+3, 46.30%), 1069.73 (M+4,

30.23%); m/z calculated: 1424.62 [M+3H]3+_{, 1068.47 [M+4H]}4+_{, 854.77 [M+5H]}5+_{; m/z}

observed: 1424.63 [M+3H]3+_{, 1068.73 [M+4H]}4+_{, 855.40 [M+5H]}5+_.

Figure 4.22: Mass spectrum of monomer 31 from the UPLC-MS analysis of peptide 3

(corresponding to Figure 4.21a). Calculated isotopic profile for [M+H]1+_(species,

abun-dance): 728.40 (M, 100%), 729.40 (M+1, 729.40%); m/z calculated: 728.35 [M+H]1+_{; m/z}

(22)