an integrated bacterial system for the discovery of ...10.1038/s41551-017-0144... · supplementary...

ArticlesDOI: 10.1038/s41551-017-0144-3

© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. © 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

An integrated bacterial system for the discovery of chemical rescuers of disease-associated protein misfoldingIlias Matis1,2, Dafni Chrysanthi Delivoria1,2, Barbara Mavroidi3, Nikoletta Papaevgeniou1,4, Stefania Panoutsou1,5, Stamatia Bellou1, Konstantinos D. Papavasileiou1,10, Zacharoula I. Linardaki1,6, Alexandra V. Stavropoulou5, Kostas Vekrellis7, Nikos Boukos8, Fragiskos N. Kolisis2, Efstathios S. Gonos1,9, Marigoula Margarity6, Manthos G. Papadopoulos1, Spiros Efthimiopoulos5, Maria Pelecanou3, Niki Chondrogianni1 and Georgios Skretas1*

1Institute of Biology, Medicinal Chemistry and Biotechnology, National Hellenic Research Foundation, 11635 Athens, Greece. 2School of Chemical Engineering, National Technical University of Athens, 15780 Athens, Greece. 3Institute of Biosciences and Applications, National Center for Scientific Research “Demokritos”, 15310 Athens, Greece. 4Faculty of Biology and Pharmacy, Institute of Nutrition, Friedrich Schiller University of Jena, 07743 Jena, Germany. 5Department of Biology, National and Kapodistrian University of Athens, 15701 Athens, Greece. 6Department of Biology, University of Patras, 26504 Patras, Greece. 7Department of Neuroscience, Center for Basic Research, Biomedical Research Foundation of the Academy of Athens, 11527 Athens, Greece. 8Institute of Nanoscience and Nanotechnology, National Center for Scientific Research “Demokritos”, 15310 Athens, Greece. 9Medical School, Örebro University, 70182 Örebro, Sweden. Present address: 10Institute of Nanoscience and Nanotechnology, National Center for Scientific Research “Demokritos”, 15310 Athens, Greece. *e-mail: [email protected]

SUPPLEMENTARY INFORMATION

In the format provided by the authors and unedited.

NAturE BIoMEDICAL ENGINEErING | www.nature.com/natbiomedeng

mailto:[email protected]

http://www.nature.com/natbiomedeng

1

SUPPLEMENTARY INFORMATION

for the Article

An integrated bacterial system for the discovery of chemical rescuers

of disease-associated protein misfolding

Ilias Matis1,2, Dafni Chrysanthi Delivoria1,2, Barbara Mavroidi3, Nikoletta Papaevgeniou1,4,

Stefania Panoutsou1,5, Stamatia Bellou1, Konstantinos D. Papavasileiou1,, Zacharoula I.

Linardaki1,6, Alexandra V. Stavropoulou5, Kostas Vekrellis7, Nikos Boukos8, Fragiskos N.

Kolisis2, Efstathios S. Gonos1,9, Marigoula Margarity6, Manthos G. Papadopoulos1, Spiros

Efthimiopoulos5, Maria Pelecanou3, Niki Chondrogianni1, Georgios Skretas1*

1Institute of Biology, Medicinal Chemistry & Biotechnology, National Hellenic Research

Foundation, Athens, Greece; 2School of Chemical Engineering, National Technical University of

Athens, Athens, Greece; 3Institute of Biosciences & Applications, National Center for Scientific

Research “Demokritos”, Athens, Greece; 4Faculty of Biology and Pharmacy, Institute of Nutrition,

Friedrich Schiller University of Jena, Jena, Germany; 5Department of Biology, National and

Kapodistrian University of Athens, Athens, Greece; 6Department of Biology, University of Patras,

Patras, Greece; 7Department of Neuroscience, Center for Basic Research, Biomedical Research

Foundation of the Academy of Athens, Athens, Greece; 8Institute of Nanoscience and

Nanotechnology, National Center for Scientific Research “Demokritos”, Athens, Greece; 9Medical

School, Örebro University, Örebro, Sweden

Present address: Institute of Nanoscience and Nanotechnology, National Center for Scientific

Research “Demokritos”, Athens, Greece

*Correspondence to:

Georgios Skretas

Institute of Biology, Medicinal Chemistry & Biotechnology

National Hellenic Research Foundation

48 Vassileos Constantinou Ave

11635 Athens

Greece

Email: [email protected]

2

Table of Contents page

Supplementary Results 4-8

Supplementary Methods 9-22

Supplementary Figures

Figure 1 Characterization of the generated combinatorial oligopeptide library cyclo-

NuX1X2X3-X5. 23-25

Figure 2 Genetic screening for the identification of macrocyclic rescuers of disease-

associated protein misfolding. 26-27

Figure 3 CD and ThT spectra of cyclic peptide:Aβ42 solutions at 2:1 ratio. 28

Figure 4 The selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116 inhibit Αβ-induced

cytotoxicity in vitro. 29-30

Figure 5 The selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116 inhibit Αβ-induced

aggregation and toxicity in vivo in a dose-dependent manner. 31

Figure 6 Sequence analysis of the selected cyclic TXXXR pentapeptides targeting Αβ. 32-33

Figure 7 Computational modelling of ΑβC5-34 and ΑβC5-116 binding to Αβ. 34-35

Figure 8 Genetic screening for the identification of macrocyclic rescuers of mutant SOD1

misfolding and aggregation. 36-37

Figure 9 Sequence analysis of the selected cyclic TXSXW pentapeptides targeting

SOD1(A4V). 38-39

Figure 10 SOD1(A4V) purification and characterization. 40-41

Supplementary Tables

Table 1 PCR primers used in this study. 42-47

Table 2 Bacterial expression vectors used in this study. 48-49

Table 3 Sequencing results of the peptide-encoding regions of 23 randomly selected

clones from the constructed pSICLOPPS-NuX1X2X3, pSICLOPPS-NuX1X2X3X4,

and pSICLOPPS-NuX1X2X3X4X5 vector sub-libraries. 50

Table 4 High-throughput sequencing analysis of the peptide-encoding regions of ~260,000

randomly selected clones from the constructed pSICLOPPS-NuX1X2X3,

pSICLOPPS-NuX1X2X3X4, and pSICLOPPS-Nu X1X2X3X4X5 sub-libraries. 51

Table 5 Sequences and frequency of appearance of the selected cyclic TXXXR

pentapeptides. 51-55

Table 6 Sequences and frequency of appearance of the selected cyclic pentapeptides

resembling ΑβC5-34. 55

Table 7 Sequences and frequency of appearance of the selected cyclic TXXR

tetrapeptides. 55

Table 8 MM–PBSA binding free energy (ΔGbind) calculations of the Aβ-cyclic peptide

complexes. 56

Table 9 MM–PBSA calculated free-energy contributions and standard error of the mean

values of the Aβ-cyclic peptide complexes. 56

Table 10 Hydrogen-bonding interactions between ΑβC5-34 and the Aβ pentameric model

unit. 57

Table 11 Hydrogen bonding interactions between ΑβC5-116 and the Aβ pentameric model

unit. 57

3

Table 12 MM–PBSA binding free energy (ΔGbind) calculations of Aβ complexes with active

and inactive variants of the selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116. 58

Table 13 Sequences and frequency of appearance of the selected cyclic TXSXW

pentapeptides. 58-59

Table 14 Comparison of the molecular properties of the selected cyclic pentapeptides with

those of conventional drugs, oral macrocyclic (MC) drugs and non-oral MC drugs. 60

Supplementary References 61-64

4

Supplementary Results

Quality assessment of the constructed combinatorial library of random cyclic

oligopeptides. Colony PCR of 124 randomly selected clones of the combined pSICLOPPS-

NuX1X2X3-X5 library with intein-specific primers revealed that 88 of them (~71%) contained the

correct insert. Overexpression of the tetra-partite fusion in 150 randomly selected clones using

0.002% arabinose and monitoring of the production of this fusion protein by western blotting using

a mouse anti-CBD primary antibody (New England Biolabs, USA; 1:100,000 dilution) and a goat

anti-mouse HRP-conjugated secondary antibody (Bio-Rad; http://www.bio-rad.com/en-

jp/sku/1706516-goat-anti-mouse-igg-h-l-hrp-conjugate; 1:4,000 dilution), showed that 99 of them

(~66%) produced high yields of the tetra-partite fusion protein (Supplementary Fig. 1c). Among

these 99 clones that produced precursor fusion protein (molecular mass ~25 kDa), 82 clones

(~55% of total clones tested) also yielded a lower molecular weight band (molecular mass ~20

kDa), which corresponds to one of the splicing reaction products, the N-terminal domain of the

Ssp DnaE intein fused to CBD (IN-CBD), after intein splicing and cyclic peptide formation takes

place (Supplementary Figs. 1a, c; data not shown). Therefore, according to these results, the

generated bacterial libraries encoding for cyclic tetra-, penta- and hexapeptide contain

approximately 20,760,000 clones, which express tetra-partite peptide fusions at high levels and

which are capable of undergoing splicing and potentially yielding cyclic peptide products. This

diversity covers fully the theoretical diversity of our combined cyclo-NuX1X2X3, NuX1X2X3X4 and

NuX1X2X3X4X5 libraries (3×203 + 3×204 + 3×205 = 10,104,000) by more than two-fold

(Supplementary Fig. 1b).

DNA sequencing of the peptide-encoding regions of the pSICLOPPS plasmid from twenty

randomly selected clones revealed the presence of all three Nu amino acids Cys, Ser, and Thr at

position 1 and a good representation of the twenty natural amino acids at all other positions within

the tetra-, penta- and hexapeptide sequence (Supplementary Table 3). To evaluate the quality

of the constructed libraries further, we characterized in more detail the amino acid diversity

http://www.bio-rad.com/en-jp/sku/1706516-goat-anti-mouse-igg-h-l-hrp-conjugate

http://www.bio-rad.com/en-jp/sku/1706516-goat-anti-mouse-igg-h-l-hrp-conjugate

5

encoded within the generated oligopeptide libraries by performing deep sequencing analysis of

the peptide-encoding region of the combined pSICLOPPS-NuX1X2X3-X5 vector library. ~85% of

the analysed DNA sequences were unique and ~67% of those were found to encode unique cyclic

peptide sequences (Supplementary Table 4). Again, all amino acids were found to be encoded

at all positions of the generated library, albeit with an over-representation of residues

corresponding to Gly (Supplementary Fig. 1d). Taken together, these results suggest that we

have constructed a high-diversity library, which should be encoding the vast majority, if not all, of

the theoretically possible tetra-, penta- and hexapeptide cyclo-NuX1X2X3-X5 sequences.

The selected Aβ-targeting cyclic oligopeptides enhance bacterial Aβ-GFP fluorescence in

an Aβ-specific manner. The increases in bacterial Aβ-GFP fluorescence observed in the

presence of the selected Aβ-targeting cyclic oligopeptides (Table 1, top) were found to be Αβ-

specific, as the isolated pSICLOPPS-NuX1X2X3-X5 vectors from these selected clones did not

enhance the levels of cellular green fluorescence when the sequence of Αβ in the Αβ42-GFP

reporter was replaced with that of each one of two unrelated disease-associated MisPs, the DNA-

binding (core) domain of the human p53 containing a Tyr220Cys substitution (p53C(Y220C))1 and

an Ala4Val substitution of human Cu/Zn superoxide dismutase 1 (SOD1(A4V))2 (Supplementary

Fig. 2c). On the contrary, the selected pSICLOPPS-NuX1X2X3-X5 vectors were efficient in

enhancing the fluorescence of Αβ-GFP containing two additional Αβ variants, Aβ40 and the E22G

(arctic) variant of Aβ42, which is associated with familial forms of AD3 (Supplementary Fig. 2d).

Structure-activity analysis for ΑβC5-116. For the selected peptides corresponding to the

TXXXR motif, residues at positions 3 and 4 were highly variable and included the majority of

natural amino acids, with position 3 exhibiting the highest diversity (Fig. 5e; Supplementary Fig.

6b, c). At position 2, Thr, Ala, and Val were preferred, while aromatic residues (Phe, Trp, Tyr)

were completely excluded from the selected TXXXR peptide pool, in full agreement with our site-

6

directed mutagenesis studies (Fig. 5c). At the highly variable position 3, the complete absence of

the negatively charged amino acids Glu and Asp among the selected sequences was notable

(Fig. 5e; Supplementary Fig. 6b, c). In general, both negatively (Glu and Asp) and positively

charged residues (Lys, His, and Arg) were found to be strongly disfavored among the selected

TXXXR sequences at positions 2 and 3. At position 4, Ala, Asp, and Trp were found to be the

preferred residues. It is noteworthy, that Lys and Gln residues were practically absent from all

positions, while the β sheet-breaking amino acid Pro that is typically included in designed peptide-

based inhibitors of amyloid aggregation4 appeared with strikingly low frequencies (Fig. 5e;

Supplementary Fig. 6b, c).

Prediction of the binding mode of AβC5-34 and AβC5-116 to Αβ. In order to unbiasedly search

for possible binding sites, a single Molecular Dynamics (MD) simulation of 100 ns duration in the

isobaric-isothermal ensemble (NPT) ensued, having five cyclic peptides placed around the Aβ

protofilament centroid at a minimum distance of 5 Å. Spatial Distribution Function (SDF) analysis

of the produced MD trajectories was then performed in conjunction with molecular docking

calculations, which assisted in narrowing the possible binding sites of both AβC5-34 and AβC5-

116 to Αβ down to five possibilities (Supplementary Fig. 7c). Thus, a new series of MD

simulations commenced with durations of 20 ns for each of these possible positions, then

continued further for an additional 20×4 ns to provide a large ensemble of conformations for the

subsequent MM-PBSA calculations. These were performed in order to finally establish the binding

site for each cyclic peptide. The calculations involved 8,000 trajectory frames, with enthalpic (ΔH),

entropic (–TΔS) and total binding free energy (ΔGbind) contributions along with the energetic

analysis of the Αβ/AβC5-34 and Αβ/AβC5-116 complexes provided in Supplementary Tables 8-

9, respectively. Also, per residue contributions to the total enthalpy of each system were

calculated for both complexes. Calculations were performed on all five binding sites and finally

7

showed that only a single site for each cyclic peptide was found to exhibit favorable binding

energetics (Fig. 6a; Supplementary Table 8).

Hydrogen bonding analysis demonstrated that some of these residues participate in the

formation of persistent hydrogen bonds (Fig. 6b; Supplementary Tables 10, 11). Apart from the

hydrogen bond between the R5 –NH2 group of AβC5-116 with the B-L34 oxygen of Αβ (Fig. 6b;

Supplementary Table 11), the residue with the most favorable binding contributions

(Supplementary Fig. 7d), T1 of AβC5-116, also forms a hydrogen bond with D-I32 (Fig. 6b;

Supplementary Table 11). Important residues that favor binding are the hydrophobic residues

A2 and F3 of AβC5-116, along with the Aβ residues A-G33, D-I32, A-M35, B-M35 and D-M35 in

proximity, hence the significant van der Waals contributions to binding (Supplementary Fig. 7d).

Regarding AβC5-34, residues S1, A2 and T5 form hydrogen bonds with the A monomer residues

A21, E22, D23 and I31 of Αβ (Fig. 6b; Supplementary Table 10). Most of these amino acids

also appear as the most favorable contributors to binding, with the exception of D23, which

disfavors binding (Supplementary Fig. 7d). This is in agreement with the less pronounced van

der Waals contribution to the interaction of this cyclic peptide with Αβ.

The importance of residues found to be involved in the formation of the AβC5-34/ and

AβC5-116/Αβ complexes, as well as the validity of our computational approach, was further tested

by examining selected mutant cyclic peptide sequences: (i) AβC5-34(T5A) (cyclo-SASPA), an

AβC5-34 variant found to be inactive in the bacterial Αβ42-GFP assay (Fig. 5a) (ii) AβC5-116(A2T)

(cyclo-TTFDR), an AβC5-116 variant found to be active in the bacterial Αβ42-GFP assay (Fig. 5c);

(iii) AβC5-116(R5A) (cyclo-TADFA), an AβC5-116 variant found to be inactive in the bacterial

Αβ42-GFP assay (Fig. 5b); and (iv) cyclo-TRDFA, a scrambled variant of AβC5-116 anticipated to

be inactive. In accordance with the experimental observations, AβC5-116(A2T) exhibited

favorable free energy of binding at the same Aβ site (Supplementary Table 12). On the contrary,

all inactive cyclic peptide variants exhibited unfavorable free energies of binding at their

8

corresponding Aβ sites (Supplementary Table 12). Taken together, these results showcase the

validity of the utilized computational approach.

9

Supplementary Methods

Reagents and chemicals

All DNA-processing enzymes were purchased from New England Biolabs (USA) apart from

alkaline phosphatase FastAP, which was purchased from ThermoFisher Scientific (USA).

Recombinant plasmids were purified using NucleoSpin Plasmid from Macherey-Nagel (Germany)

or Plasmid Midi kits from Qiagen (Germany). PCR products and DNA extracted from agarose gels

were purified using Nucleospin Gel and PCR Clean-up kits from Macherey-Nagel (Germany),

respectively. All chemicals were purchased from Sigma-Aldrich (USA), unless otherwise stated.

Isopropyl-β-D-thiogalactoside (IPTG) was purchased from MP Biomedicals (Germany). Stock

solutions of the synthetic cyclic peptides were as follows: 32.5 mM in water for AβC5-34, 10 mM

in 40% DMSO for ΑβC5-116 and 30 mM in 40% DMSO for SOD1C5-4.

Cyclic oligopeptide library construction and initial characterization

Initially, we constructed nine distinct combinatorial cyclic peptide sub-libraries: the cyclo-

CysX1X2X3, cyclo-SerX1X2X3, and cyclo-ThrX1X2X3 tetrapeptide sub-libraries (pSICLOPPS-

CysX1X2X3, pSICLOPPS-SerX1X2X3, and pSICLOPPS-ThrX1X2X3 vector sub-libraries), the cyclo-

CysX1X2X3X4, cyclo-SerX1X2X3X4, and cyclo-ThrX1X2X3X4 cyclic pentapeptide sub-libraries

(pSICLOPPS-CysX1X2X3X4, pSICLOPPS-SerX1X2X3X4, and pSICLOPPS-ThrX1X2X3X4 vector

sub-libraries) and the cyclo-CysX1X2X3X4X5, cyclo-SerX1X2X3X4X5, and cyclo-ThrX1X2X3X4X5

cyclic hexapeptide sub-libraries (pSICLOPPS-CysX1X2X3X4X5, pSICLOPPS-SerX1X2X3X4X5, and

pSICLOPPS-ThrX1X2X3X4X5 vector sub-libraries) (Supplementary Table 2). These vectors

express libraries of fusion proteins comprising four parts: (i) the C-terminal domain of the split Ssp

DnaE intein (IC), (ii) a tetra-, penta-, or hexapeptide sequence, (iii) the N-terminal domain of the

split Ssp DnaE intein (IN), and (iv) a chitin-binding domain (CBD) under the control of the PBAD

promoter and its inducer L(+)-arabinose (Supplementary Fig. 1a). The libraries of genes

10

encoding these combinatorial libraries of random cyclic oligopeptides were constructed using the

degenerate forward primers GS032, GS033, GS034, GS072, GS073, GS074, GS075, GS076,

GS077, individually in pair with the reverse primer GS035, and pSICLOPPS as a template

(Supplementary Table 1). Cys, Ser, and Thr were encoded in these primers by the codons TGC,

AGC, and ACC, respectively, which are the most frequently utilized ones for these amino acids in

E. coli, while the randomized amino acids (X) were encoded using random NNS codons, where

N=A, T, G, or C and S=G or C, as described previously5. A second PCR reaction was conducted

in each case to eliminate mismatches using the aforementioned amplified DNA fragments as

templates, and the forward primers GS069, GS070 and GS071 for the peptide sub-libraries

starting with Cys, Ser, or Thr, respectively, together with the reverse primer GS035. The resulting

PCR products were digested with BglI and HindIII for 5 h and inserted into the similarly digested

and dephosphorylated auxiliary vector pSICLOPPSKanR (see below). The ligation reactions were

optimised at a 12:1 insert:vector molar ratio and performed for 4 h at 16 °C. Approximately 0.35,

0.7 and 3.5 μg of the pSICLOPPSKanR vector were used for each one of the tetra-, penta- and

hexapeptide libraries, respectively. The ligated DNA was then purified using spin columns

(Macherey-Nagel, Germany), transformed into electro-competent MC1061 cells prepared in-

house, plated onto LB agar plates containing 25 μg/mL chloramphenicol and incubated at 37 °C

for 14-16 h. This procedure resulted in the construction of the combined pSICLOPPS-NuX1X2X3-

X5 library with a total diversity of about 31,240,000 independent transformants, as judged by

plating experiments after serial dilutions.

Expression vector construction

For the construction of pETSOD1-GFP, the human SOD1 cDNA was generated by PCR-mediated

gene assembly using the primers GS100, GS101, GS102, GS103, GS104, GS105, GS106,

GS107, GS108, GS109, GS110, and GS111 (Supplementary Table 1). The assembled gene

was further amplified by PCR and the resulting product was digested with NdeI and BamHI, and

11

inserted into similarly digested pAβ42-GFP vector, in the place of Aβ42. For pETSOD1(A4V)-GFP,

SOD1 was amplified by PCR from the pETSOD1-GFP vector using the mutagenic forward primer

GS059 and the reverse primer GS060. The resulting PCR product was then digested with NdeI

and BamHI, and inserted into similarly digested pAβ42-GFP. For pETSOD1(G37R)-GFP,

pETSOD1(G85R)-GFP and pETSOD1(G93A)-GFP construction, SOD1 was mutated by overlap

extension PCR starting from pETSOD1-GFP as a template and using the following sets of primers:

GS058/GS059/GS112/GS113, GS058/GS059/GS114/GS115 and GS058/GS059/GS116/GS117,

respectively. All SOD1 PCR products were then digested with NdeI and BamHI, and inserted into

similarly digested pAβ42-GFP vector.

For the construction of pETSOD1, pETSOD1(G37R), pETSOD1(G85R) and

pETSOD1(G93A), the corresponding SOD1 genes were amplified by PCR from pETSOD1-GFP,

pETSOD1(G37R)-GFP, pETSOD1(G85R)-GFP and pETSOD1(G93A)-GFP, respectively, using

the primers SP006-SP004. For the construction of pΕΤSOD1(A4V), SOD1 was amplified from

pΕΤSOD1(A4V)-GFP using the primers SP007-SP004. All SOD1 PCR products were digested

with XbaI and BamHI, and cloned into similarly digested pET28a(+) (Novagen, USA).

In order to construct pASKSOD1-GFP, pASKSOD1(A4V)-GFP, pASKSOD1(G37R)-GFP,

pASKSOD1(G85R)-GFP and pASKSOD1(G93A)-GFP, the SOD1 genes were sub-cloned from

the corresponding pETSOD1-GFP vectors using XbaI and BamHI, and ligated into similarly

digested pASKp53-GFP (see below). Similarly, pASKSOD1, pASKSOD1(A4V),

pASKSOD1(G85R) and pASKSOD1(G93A) were constructed by sub-cloning the SOD1 genes

from the corresponding pETSOD1 vectors into pASK756 using XbaI and BamHI.

For the construction of pETp53-GFP, a truncated human TP53 gene encoding the DNA-

binding (core) domain of p53 (p53C, amino acids 94-312) was assembled by PCR using the

primers GS118, GS119, GS120, GS121, GS122, GS123, GS124, GS125, GS126, GS127,

GS128, GS129, GS130, GS131, GS132, and GS133. The PCR product was then digested with

NdeI and BamHI, and was inserted into similarly digested pETAβ42-GFP vector. For constructing

12

pETp53(Y220C)-GFP, the p53C-encoding gene was mutated by overlap PCR starting from

pETp53-GFP as a template and using primers GS003, GS004, GS007 and GS008, the PCR

product was digested with NdeI and BamHI, and inserted into similarly digested pAβ42-GFP vector.

pASKp53-GFP was generated by PCR amplification of the gene encoding for p53-GFP from

pETp53-GFP using primers GS002 and GS003 and ligated into pASK75 using the restriction sites

XbaI-HindIII.

For the construction of the pSICLOPPS vectors encoding for variants of the selected

AβC5-34 and AβC5-116 peptides, the auxiliary pSICLOPPSKanR vector was generated initially.

pSICLOPPSKanR was constructed by PCR amplification of the gene encoding aminoglycoside

3'-phosphotransferase (KanR - the enzyme conferring resistance to the antibiotic kanamycin)

from pET28a(+) using primers GS043-DG002, digestion with BglI and HindIII and insertion into

similarly digested pSICLOPPS. For the construction of the vectors pSICLOPPS-ΑβC5-34(S1C),

pSICLOPPS-ΑβC5-34(S1T), pSICLOPPS-ΑβC5-34(S3A), pSICLOPPS-ΑβC5-34(P4A) and

pSICLOPPS-ΑβC5-34(T5A), mutagenic PCR was carried out starting from pSICLOPPS-ΑβC5-

34 and using the forward primers IM033, IM034, IM036, IM037, IM038, respectively, along with

the reverse primer GS035, followed by digestion of the generated product with BglI and HindIII

and insertion into similarly digested pSICLOPPSKanR.

The vectors pSICLOPPS-ΑβC5-116(T1C), pSICLOPPS-ΑβC5-116(T1S), pSICLOPPS-

ΑβC5-116(F3A), pSICLOPPS-ΑβC5-116(D4A), pSICLOPPS-ΑβC5-116(R5A), pSICLOPPS-

ΑβC5-116(A2F), pSICLOPPS-ΑβC5-116(A2S), pSICLOPPS-ΑβC5-116(A2P), pSICLOPPS-

ΑβC5-116(A2T), pSICLOPPS-ΑβC5-116(A2Y), pSICLOPPS-ΑβC5-116(A2H), pSICLOPPS-

ΑβC5-116(A2K), pSICLOPPS-ΑβC5-116(A2E), pSICLOPPS-ΑβC5-116(A2W), pSICLOPPS-

ΑβC5-116(A2R), pSICLOPPS-ΑβC5-116(A2del), pSICLOPPS-ΑβC5-116(F3del) and

pSICLOPPS-ΑβC5-116(D4del) were generated in a similar fashion by starting from pSICLOPPS-

ΑβC5-116 as a template and using the mutagenic forward primers IM027, IM028, IM030, IM031,

IM032, IM043, IM044, IM045, IM046, IM047, IM048, IM049, IM050, IM051, IM052, IM039, IM040

13

and IM041, respectively, along with the reverse primer GS035. Digestion of the generated product

with BglI and HindIII, was followed by insertion into similarly digested pSICLOPPSKanR.

pSICLOPPS-ΑβC5-325, pSICLOPPS-ΑβC5-359, pSICLOPPS-ΑβC5-413, pSICLOPPS-

ΑβC5-479 were generated by PCR amplification using the template pSICLOPPS-Random1 and

the forward primers IM077, IM078, IM080 and IM081, respectively, along with the reverse primer

GS035, digestion with BglI and HindIII and ligation into similarly digested pSICLOPPSKanR.

Vectors pSICLOPPS(H24L;F26A)-ΑβC5-3, pSICLOPPS(H24L;F26A)-ΑβC5-2,

pSICLOPPS(H24L;F26A)-ΑβC5-17 pSICLOPPS(H24L;F26A)-ΑβC6-1,



pSICLOPPS(H24L;F26A)-Random1 and pSICLOPPS(H24L;F26A)-Random2 were constructed

by starting from the corresponding pSICLOPPS vectors encoding for the peptides TTVDR (AβC5-

3), TTYAR (AβC5-2), TTTAR (ΑβC5-17), TPVWFD (AβC6-1), TAWCR (ΑβC5-27), TTWCR

(ΑβC5-21), TAFDR (ΑβC5-116), random cyclic peptide 1 (undetermined sequence), random

cyclic peptide 2 (undetermined sequence), respectively, digestion with BglI and HindIII, and

ligation of the resulting inserts into the similarly digested auxiliary vector

pSICLOPPS(H24L;F26A)KanR. The auxiliary pSICLOPPS(H24L;F26A)KanR vector had been

generated previously using primers GS037 and DD015 to amplify and mutate the C-terminal

domain of the Ssp DnaE intein from pSICLOPPS and the resulting PCR product was digested

with NcoI and BglI and inserted into similarly digested pSICLOPPSKanR. All primer sequences

are described in Supplementary Table 1 and all constructed expression vectors are listed in

Supplementary Table 2.

Protein electrophoresis and western blot analysis

The utilized antibodies were a mouse monoclonal, horseradish peroxidase (HRP)-conjugated

anti-polyhistidine antibody (Sigma, USA;

14

http://www.sigmaaldrich.com/catalog/product/sigma/a7058?lang=en&region=GR) at 1:2,500

dilution, a mouse monoclonal anti-FLAG (Sigma, USA;

http://www.sigmaaldrich.com/catalog/product/sigma/a8592?lang=en&region=GR) at 1:1,000

dilution, a mouse anti-GFP at 1:20,000 dilution (Clontech, USA;

http://www.clontech.com/SI/Products/Fluorescent_Proteins_and_Reporters/Fluorescent_Protein

_Antibodies/ibcGetAttachment.jsp?cItemId=27547&fileId=5897131&sitex=10020:22372:US), a

mouse anti-Aβ (6Ε10) (Covance, USA;

https://www.antibodypedia.com/gene/668/APP/antibody/1457910/SIG-39320) at 1:2,000 dilution,

a mouse anti-CBD (New England Biolabs, USA; https://www.neb.com/products/e8034-anti-cbd-

monoclonal-antibody) at 1:25,000 or 1:100,000 dilution, and a HRP-conjugated goat anti-mouse

antibody (Bio-Rad, USA; http://www.bio-rad.com/en-jp/sku/1706516-goat-anti-mouse-igg-h-l-hrp-

conjugate) at 1:4,000.

Neuronal cell cultures

The utilized U87MG cells (human glioblastoma-astrocytoma, epithelial-like cell line) were a kind

gift from Dr. Maria Paravatou-Petsotas, Radiobiology Laboratory, Institute of Nuclear &

Radiological Sciences & Technology, Energy & Safety, NCSR “Demokritos” and have been

utilized previously to reproduce successfully published observations from the scientific literature.

This cell line was found to be free of mycoplasma contamination, as judged visually under

microscope observation and by regular DAPI staining of the cell cultures. The utilized

media/agents for U87MG cell cultures were obtained from Biochrom AG (Germany) and PAA

Laboratories (USA). U87MG cells were grown in Dulbecco's modified Eagle medium (DMEM),

supplemented with 10% fetal bovine serum (FBS), 2.5 mM L-glutamine, 1%

penicillin/streptomycin at 37 °C and 5% CO2. For MTT cytotoxicity studies, cells were plated at a

density of 2×104 cells per well in 96-well plates and incubated at 37 °C for 24 h to allow cells to

attach. The medium was subsequently removed and cells were rendered quiescent by incubation

15

in serum-free medium for 24 h. For cell viability measurements cells were subsequently treated

with the indicated concentrations of Aβ in the presence or absence of synthetic peptides, as

described in the corresponding paragraph of the Methods.

The human dopaminergic neuroblastoma cell line SH-SY5Y (kindly provided by Prof.

Leonidas Stefanis, University of Athens, Greece) was maintained in DMEM supplemented with

10% heat-inactivated fetal bovine serum, 2 mM glutamine and 1% non-essential amino acids

(complete medium). SH-SY5Y cells were treated with conditioned medium (CM) produced by (a)

the control cell line CHO and, (b) the Aβ-oligomer-producing cell line 7PA2 (kindly provided by

Prof. Dominic Walsh, Brigham & Women’s Hospital, USA and subsequently validated for their

cytotoxic Αβ-oligomer-secreting activity), derived upon stable transfection of CHO cells with

human APP bearing the Val717Phe familial AD mutation that leads to Aβ overproduction7. Both

CHO and CHO-7PA2 cell lines were maintained in complete medium with or without G418 (200

μg/mL - Invitrogen, USA). SH-SY5Y cells were exposed to solvent (no peptide), 5 μM AβC5-116

or 10 μΜ AβC5-34 for 24 h followed by the addition of the relevant CM derived from 7PA2 cells

(CMAβ) or CHO cells (CMcontrol). All cultures were maintained at 37 °C in a humidified 5% CO2

incubator. For CM preparation, CHO and 7PA2 cells were grown to approximately 90%

confluency, washed with PBS and incubated in serum-free DMEM for approximately 16 h. The

CM was collected and centrifuged to remove cell debris. 7PA2-derived Αβ oligomers and supplied

to primary cortical neurons were prepared, in the presence or absence of the selected cyclic

pentapeptides, in a similar manner.

Immunocytochemistry

Primary mouse cortical neurons were treated for 1 h at 37 °C with vehicle, 1 μΜ Aβ40, 1 μΜ Aβ40

+ 1 μΜ ΑβC5-34 or 1 μΜ Aβ40 + 1 μΜ ΑβC5-116. Aβ40 had been pre-aggregated for 3 d at 37 °C

in the presence or absence of the selected cyclic peptides before addition to neurons. Cell-free

cover slips were also used as negative controls to assess non-specific binding of Aβ to the glass.

16

After treatment, neurons were immunofluorescently labeled for Aβ binding to the cell surface using

the mouse monoclonal anti-Αβ antibody 6E10 (Covance, USA) (1:1000). To discriminate between

Aβ-specific labeling and labeling derived from staining full-length APP, neuronal

immunofluorescence analysis was also performed using the rabbit polyclonal C-terminal anti-APP

antibody R1(57)8 (1:1000), a kind gift from Dr. Pankj Mehta (Institute for Basic Research in

Developmental Disabilities, Staten Island, New York), which does not recognize Αβ. Briefly, after

removing the medium and rinsing with PBS twice, neurons were fixed in 4% paraformaldehyde

(in PBS, pH 7.4) for 15 min at room temperature (RT). Cells were then washed three times with

PBS and non-specific binding was blocked with 10% normal goat serum (NGS), 0.1% Triton X-

100 (in PBS) for 60 min at RT. Then, neurons were doubly immunolabeled by overnight incubation

at 4 oC with 6E10 and R1(57) diluted in PBS, 10% NGS, 0.1% Triton X-100. Subsequently, cells

were rinsed three times with PBS and incubated for 1 h at RT with Alexa Fluor 488-labeled anti-

mouse IgG (ThermoFisher Scientific, USA) and Cy3-labeled anti-rabbit IgG (ThermoFisher

Scientific, USA) secondary antibodies (1:2,500) diluted in PBS, 1% NGS, 1% Triton X-100.

Following washing, cell nuclei were counterstained with DAPI (1:1000) for 5 min, and coverslips

were mounted with mounting medium. Digital images of neurons were acquired using a 40×

objective lens in a Αxio observer Z1 fluorescence microscope (Zeiss, Germany) supported by the

ZEN 2012 (blue edition) software. 15-20 images were obtained for each experimental condition

and experiment. Images of all samples within an experiment were acquired using identical image

acquisition settings for each fluorophore.

In vivo assays in C. elegans

Paralysis assay. Synchronized L4 larvae CL2006 animals (90-120 animals per condition) were

transferred to NGM plates containing living bacteria biosynthetically producing ΑβC5-34, ΑβC5-

116 or a randomly selected cyclic peptide sequence at 20 °C. Synchronized CL4176 animals

(150-300 animals per condition) were transferred to NGM plates containing synthetic ΑβC5-34,

17

ΑβC5-116 or 0.26% DMSO at 16 °C for 48 h before transgene induction via temperature up-shift

to 25 °C. Synchronized offspring were randomly distributed to treatment plates to avoid systematic

differences in egg lay batches. Treatment and control plates were handled, scored and assayed

in parallel. Scoring of paralyzed animals was initiated at day 1 of adulthood for the CL2006 strain

and 24 h after temperature up-shift for the CL4176 strain. Nematodes were scored as paralyzed

upon failure to move their half end-body upon prodding. Animals that died were excluded. Plates

were indexed as 1, 2, 3 etc by an independent person and were given to the observer for scoring

in random order. The index was revealed only after scoring. The log-rank (Mantel–Cox) test was

used to evaluate differences between paralysis curves and to determine P values for all

independent data. n in paralysis figures is the number of animals that paralyzed over the total

number of animals used (the number of paralyzed animals plus the number of dead and censored

animals). Median paralysis values are expressed as mean ± s.e.m.

Dot blot Analysis. CL4176 animals were allowed to lay eggs for 3 h on NGM plates containing

either synthetic peptides or 0.26% DMSO. Paralysis was induced upon temperature up-shift and

the progeny were exposed to either pure peptides or 0.26% DMSO until 50% of the control

population was paralyzed. The animals were then collected and boiled in non-reducing Laemmli

buffer. For dot blot analysis, 1–5 µg of protein lysates were spotted onto 0.2 μm nitrocellulose

membranes (Bio-Rad, USA) after soaking into TBS pre-heated at 80 °C. Immunoblotting was

performed using the anti-Αβ antibody 6E10 (recognizes total Aβ) and the anti-amyloid protein,

oligomer-specific antibody AB9234 (Merck Millipore, Germany;

https://www.merckmillipore.com/INTL/en/product/Anti-Amyloid-Oligomer-Antibody%2C

%CE%B1%CE%B2%2C-oligomeric,MM_NF

AB9234?ReferrerURL=https%3A%2F%2Fwww.google.gr%2F&bd=1). Actin was used as a

loading control. Blots were developed with chemiluminescence by using the ClarityTM Western

ECL substrate (Bio-Rad, USA). Quantification of the ratio of each detected protein to actin using

https://www.merckmillipore.com/INTL/en/product/Anti-Amyloid-Oligomer-Antibody%2C%20%CE%B1%CE%B2%2C-oligomeric,MM_NF%20AB9234?ReferrerURL=https%3A%2F%2Fwww.google.gr%2F&bd=1



18

the anti-actin antibody sc-1615 (Santa Cruz, Germany; http://datasheets.scbt.com/sc-1615.pdf),

and normalization to control appears next to each representative blot.

Confocal microscopy analysis. For Aβ3-42 deposit measurements, synchronized (at the L4 larval

stage) CL2331 and CL2179 (control strain) animals exposed to solvent (0.26% DMSO), 10 μΜ

ΑβC5-34 or 5 μM ΑβC5-116 and grown at 20 C (to induce aggregation) until day 2 of adulthood

were collected. Animals were mounted onto 2% agarose pads on glass slides, anesthetized with

10 mM levamisole and observed at RT using a Leica TCS SPE confocal laser scanning

microscope (Leica Lasertechnik GmbH, Germany). The LAS AF software was used for image

acquisition. At least twenty animals/condition in three independent experiments were processed.

Images of whole worms and focused images in the posterior area of nematodes were acquired

with 10 x 0.45 and 20 x 0.70 numerical aperture, respectively.

Computational Methods

In order to study the interactions of the selected cyclic peptides with Aβ and to determine their

possible binding sites, we selected as a model the 3D structure of an Aβ42 protofibril previously

proposed based on data collected by NMR (PDB ID: 2BEG, model 10)9 (Supplementary Fig. 7a),

which has been previously utilized for similar purposes10. Missing residues were added with

Pymol11 and, in accordance to previous studies12, residues 1 to 8 were omitted given their

unimportance for fibril growth13, 14 and for the structural dynamics of the remainder residues13, 15.

The lack of the eight N-terminal residues was tackled by capping with an acetyl group (ACE),

while addition of residues 9 to 16 yielded the final sequence of the monomer, namely

ACE-GY10EVHHQKLVFF20AEDVGSNKGA30IIGLMVGGVV40IA12, bearing a net charge of –2.

Each monomer consists of two β-strands, β1 (residues G9−S26) and β2 (residues I31-A42),

connected by a U-bent turn spanning the four residues N27-A30 (Supplementary Fig. 7a). The

cyclic pentapeptides examined were initially sketched in ACD/ChemSketch,16 and 3D structures

http://datasheets.scbt.com/sc-1615.pdf

19

were consequently obtained utilizing the ANTECHAMBER module17 of the AMBER 12 suite of

programs16 (Supplementary Fig. 7b).

Molecular docking calculations. Both the Aβ pentameric model (receptor) and cyclic peptide

(ligand) structures were prepared for docking calculations in AutoDock Vina18, by means of

AutoDock Tools 1.5.619, 20. Our docking studies were performed intermediately to MD simulations,

following a scheme which is detailed below, and assisted in revealing regions with affinity towards

peptide binding. Specifically, MD simulations revealed possible regions of peptide affinity

(Supplementary Fig. 7c, top), and docking positions were then delimited by five grid boxes

(Supplementary Fig. 7c, bottom). Partial atomic charges were assigned to all systems according

to the Kollman United Atom scheme, by previously merging non-polar hydrogens to heavy atoms.

Three-dimensional affinity maps were generated by AutoGrid for each peptide atom type along

with electrostatic and desolvation maps for the Aβ pentamer unit. All five grid boxes had 1.000 Å

spacing and dimensions 22×22×22, setting the exhaustiveness equal to 200; the lowest energy

complex structures were then used as initial structures in MD simulations.

Molecular Dynamics simulations. All systems were modeled using the ff99SB force field21, adding

Na+ counter ions to achieve electroneutrality. Water solvent molecules were explicitly added to all

systems by means of the TIP3P model22, by using a truncated octahedron unit cell and by

imposing a distance of at least 15 Å between any atom of the complex and the box boundaries.

This was selected since the Aβ pentameric model shows great flexibility and large conformational

changes were expected to occur during the MD simulations13. Prior to production runs, close

contacts between atoms were removed by performing a total of 20,000 energy minimization

cycles, imposing positional restraints on solute atoms with a harmonic force constant. The

restraint was gradually lifted every 5,000 steps starting from 500 to 10, 2 and 0 kcal mol-1 Å-2,

respectively. The Particle Mesh Ewald (PME) method23, 24 was applied on long-range electrostatic

interactions. Initial heating to 310 K for 100 ps in the NVT ensemble, having a 10 kcal mol-1 Å-2

harmonic force constant restraint on Aβ and peptide atoms, was followed by a 100-ps density

20

equilibration run in the NPT ensemble, imposing the same restraint. Temperature was maintained

by Langevin thermostat25 with a collision frequency equal to 2 ps-1 25 and the non-bonded cutoff

was set to 10 Å. Then, 100-ps unrestrained equilibration runs in the NPT ensemble followed, at

constant pressure of 1 bar using the Langevin barostat. All bonds involving hydrogen were

constrained to their equilibrium bond lengths using SHAKE26 and the integration step was set to

2 fs. Simulations were performed by means of the PMEMD GPU accelerated AMBER module24,

27, 28 using the standard fixed point precision model (SPFP)28. The pentameric model unit was first

subjected to 100 ns MD simulations in order to extract its most preferable conformational state.

Clustering analysis was performed in order to find the most populated cluster of the pentamer unit

by imposing a 2.5 Å RMSD cutoff, using the gromos algorithm29 as implemented within the

g_cluster utility of the GROMACS 4.6.4 software30. The centroid cluster structure accounts for 75%

of the total population, in agreement with previous studies12, and served as the starting point for

subsequent simulations. Then, a single MD simulation of 100 ns duration in the NPT ensemble

ensued, placing five cyclic peptides in a radius of 5 Å around the Aβ pentameric model centroid

unit. Upon completion, spatial distribution functions (SDF) were evaluated by means of the

GROMACS g_spatial module30. SDF analysis revealed areas of preference towards the Aβ

surface (Supplementary Fig. 7c, top), which were then partitioned into five different grid boxes

for docking calculations (Supplementary Fig. 7c, bottom). Binding sites were individually tested

using the lowest energy pose of each one as the initial structure for additional 20-ns MD

simulations. Hydrogen bond (HB) analysis was performed by setting a donor-acceptor distance

cut-off at 3.5 Å and the angle of donor−hydrogen−acceptor at 150° by means of the CPPTRAJ

program. Figures were prepared by means of the VMD software31.

MM–PBSA calculations. The Molecular Mechanics Poisson–Boltzmann Surface Area (MM-PBSA)

method performs well for predicting relative binding energies32. It calculates the interaction energy

in the gas phase with molecular mechanics and estimates the solvation free energy by solving

the Poisson–Boltzmann equation33, 34; the equations can be found elsewhere35. Following the

21

work by Wright et al.36, where the use of multiple replica simulations of relatively short duration

showed well converged energy estimates and improved relative binding strengths of several

HIV‑1 protease inhibitors36, we performed 20×4 ns individual simulations to be used in MM-PBSA

calculations. In calculating the total binding free energy (ΔGbind) contributions of each complex,

8,000 trajectory frames were used for the enthalpy term (ΔH), while the conformational entropy

contribution (–TΔS) was evaluated by normal mode analysis over 1,000 trajectory frames, for

efficiency reasons37.

SOD1 purification and preparation of stocks and solutions

SOD1 or mutants thereof were overexpressed from the appropriate pET-SOD1 or pASK-SOD1

vectors in E. coli Origami 2(DE3) cells in LB medium containing 50 μg/mL kanamycin (for pET-

SOD1) or 100 μg/mL ampicillin (for pASK-SOD1), 200 μM CuCl2, and 200 μΜ ZnCl2 by the

addition of 0.01 mM IPTG (for pET-SOD1) or 0.2 μg/mL anhydrotetracycline (aTc) (for pASK-

SOD1), either at 37 °C for 2-3 h or at 18 °C for about 16 h. Origami 2(DE3) cells were utilized in

order to provide an oxidizing cytoplasmic environment and to promote correct formation of

disulfide bonds38, which are required for proper SOD1 folding and function39. Under these

conditions, bacterially produced SOD1 is produced in dimeric and enzymically active form

(Supplementary Fig. 10), while it simultaneously co-exists with misfolded, soluble and insoluble

SOD1 oligomeric/aggregated species (Supplementary Fig. 8c; Supplementary Fig. 10). Thus,

as described previously, the acquired protein is found in a state that resembles the conditions

encountered in human cells under stressful or pathogenic conditions40. The appearance of

misfolded SOD1 oligomers/aggregates is enhanced with increasing incubation temperatures

(Supplementary Fig. 10). Thus, for assays that are more appropriate for monitoring the early

steps of SOD1 oligomerization/aggregation, such as dynamic light scattering (DLS), we utilized

SOD1 produced at 18 °C, whereas for assays that are more appropriate for monitoring the later

22

steps of SOD1 aggregation, such as filter retardation, ThT staining and CD spectroscopy, we

utilized SOD1 produced at 37 °C.

Cell pellets from 1 L cultures were collected after SOD1 overexpression by centrifugation,

re-suspended in 20 mL lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, pH 8.0)

and lysed by brief sonication steps on ice. The soluble cell lysate was collected after centrifugation

at 13,000 × g for 15 min at 4 °C, and was mixed with 1 mL Ni-NTA agarose resin (Qiagen,

Germany) for 1 h at 4 °C on a roller mixer before loading onto a 5 mL polypropylene

chromatography column (Pierce, USA). Column-bound protein was washed twice with 10 mL

wash buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8.0) and eluted in three

fractions, each consisting of 1 mL elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM

imidazole, pH 8.0). Imidazole was removed after protein purification by dialysis against buffer

(100 mM Tris, 300 mM NaCl, 200 μΜ CuSO4 and 200 μM ZnSO4, pH 6.8) overnight at 4 °C. The

purified protein was quantified using the assay described by Bradford41. SOD1 used for DLS

analyses was further purified by size-exclusion chromatography (SEC) using a Superdex75

10/300GL column (GE Healthcare, USA), to isolate the dimeric protein fraction in TBS buffer, pH

7.4. DLS analysis was performed on the purified sample for verification of the dimeric state of

SOD1 (Supplementary Fig. 10). Purified SOD1 was incubated with or without the selected cyclic

peptides in 40 μM solutions in PBS under the indicated conditions.

23

Supplementary Figures

a

b

Peptide type General formulaa Theoretical diversity Actual library

coverage

Tetrapeptides cyclo-NuX1X2X3 3×203 = 24,000

Pentapeptides cyclo-NuX1X2X3X4 3×204 = 480,000

Hexapeptides cyclo-NuX1X2X3X4X5 3×205 = 9,600,000

Combined library cyclo-NuX1X2X3-X5 10,104,000 ×2 aNu=C, S, or T; X=anyone of the twenty natural amino acids

c

25

Supplementary Figure 1. Characterization of the generated combinatorial oligopeptide library

cyclo-NuX1X2X3-X5. (a) (Left) Schematic of the pSICLOPPS-NuX1X2X3-X5 vector library encoding

the combinatorial oligopeptide library cyclo-NuX1X2X3-X5. Nu: Cys (C), Ser (S), or Thr (T); X: any

of the 20 natural amino acids; NNS: randomized codons, where N=A, T, C or G and S=G or C; IC:

C-terminal fragment of the split Ssp DnaE intein; IN: N-terminal fragment of the split Ssp DnaE

intein; CBD: chitin-binding domain. (Right) Intein-mediated peptide cyclization using SICLOPPS.

The tetra-partite fusion undergoes intein splicing upon intein fragment re-association, leading to

peptide cyclization and the production of the cyclo-NuX1X2X3-X5 library. (b) Theoretical and actual

diversity of the constructed combinatorial cyclo-NuX1X2X3-X5 oligopeptide library, determined as

described in the Supplementary Results and Supplementary Methods sections. (c) Indicative

western blot analysis using an anti-CBD antibody of fourteen randomly selected individual clones

from the constructed cyclo-NuX1X2X3X4X5 hexapeptide sub-library, demonstrating that individual

clones can exhibit variable levels of expression. Lanes 4, 6, 11 and 12 correspond to clones that

contain stop codons or frameshifts and, thus, do not express a full-length IC-peptide-IN-CBD

tetrapartite fusion or generate cyclic peptide product. (d) Heat maps depicting the amino acid

distribution at each position of the constructed cyclo-CysX1X2X3-X5, cyclo-SerX1X2X3-X5, and

cyclo-ThrX1X2X3-X5 oligopeptide libraries, as determined by deep sequencing analysis of the

peptide-encoding region of the generated pSICLOPPS-NuX1X2X3-X5 vector library.

26

a

b

27

c d

Supplementary Figure 2. Genetic screening for the identification of macrocyclic rescuers of

disease-associated protein misfolding. (a) Schematic of the utilized MisP-GFP genetic system for

monitoring MisP folding and misfolding and for identifying macrocyclic rescuers of MisP misfolding

in E. coli cells. (b) SDS-PAGE/western blot analysis using an anti-CBD antibody of the ten

individual selected clones investigated in Table 1 (top). The upper band of ~25 kDa corresponds

to the IC-peptide sequence-IN-CBD precursor, while the lower band of ~20 kDa corresponds to

the processed IN-CBD product, whose appearance is an indication of successful cyclic peptide

formation. CBD: chitin-binding domain. (c) Fluorescence of BL21(DE3) cells co-expressing Αβ42-

GFP, SOD1(A4V)-GFP or p53(Y220C)-GFP, from the vectors pETΑβ42-GFP, pETSOD1(A4V)-

GFP or pETp53(Y220C)-GFP, respectively, together with the cyclic peptides encoded by the

selected clones 7 and 10, which were isolated after the second round of FACS sorting depicted

in Table 1 (top). For each fusion, the fluorescence of the cell population producing a random

cyclic peptide was arbitrarily set to 100. Experiments were carried out in replica triplicates (n=1

independent experiment) and the reported results correspond to the mean value ± s.e.m. (d)

Fluorescence of BL21(DE3) cells co-expressing Αβ-GFP fusions that include the indicated Αβ

variants together with the cyclic peptides encoded by the selected clones 7 and 10, which were

isolated after the second round of FACS sorting depicted in Table 1 (top). Τhe Aβ42-GFP

fluorescence of the cell population producing a random cyclic peptide was arbitrarily set to 100.

Experiments were carried out in replica triplicates (n=1 independent experiment) and the reported

results correspond to the mean value ± s.e.m.

28

a

b

c

Supplementary Figure 3. CD and ThT spectra of cyclic peptide:Aβ42 solutions at 2:1 ratio. (a)

CD spectra of 50 µM Aβ42 in phosphate buffer (10 mM, pH 7.33) in the presence of 100 µM of

ΑβC5-34 or ΑβC5-116. Spectra were collected for a period of 30 days at 33 °C. (b) ThT

fluorescence spectra after ThT addition to the aged (30 d) CD solutions shown in (a). (c) CD

spectra of Αβ-free solutions of ΑβC5-116, ΑβC5-34, and SOD1C5-4 at 50 (black) and 100 µM

(blue) solutions in phosphate buffer (10 mM, pH 7.33) after 30 d at 33 °C.

29

a

b

c

30

d

Supplementary Figure 4. The selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116 inhibit Αβ-

induced cytotoxicity in vitro. (a) Effect of pre-incubation time on the cytotoxicity of Aβ solutions. Cell

viability of primary hippocampal neurons, as determined by the MTT assay, treated for 24 h at 37 °C

with 1 μΜ solutions of Aβ40 or Aβ42, pre-aggregated for time periods ranging from 0–25 d. An MTT

stock solution in Neurobasal-A complete medium was added to each well at a final concentration of

0.5 mg/mL and incubated for 3 h at 37 °C. (b) Cell viability as determined by the MTT assay of

serum-starved U87MG cells treated for 24 h at 37 °C without Aβ (white bars) or with 1 μΜ solutions

of Aβ40 or Aβ42, previously aggregated in the presence or absence of 1 and 2 μΜ of ΑβC5-34 (left)

or ΑβC5-116 (right). A MTT stock solution in DMEM complete medium was added to each well at

a final concentration of 1 mg/mL and incubated for 4 h at 37 °C. Before addition to the cells, all Aβ40

solutions were pre-aggregated for 3 d and all Aβ42 solutions were pre-aggregated for 1 d. (c) Cell

viability as determined by crystal violet staining of SH-SY5Y human neuroblastoma cells exposed

to conditioned medium (CM) containing secreted human Aβ42 oligomers from cultured 7PA2 cells7

(CMΑβ) and treated without cyclic peptide (no peptide), with 10 μΜ AβC5-34, or with 5 μM AβC5-

116, for 2 h. The viability of SH-SY5Y cells exposed το CMΑβ in the absence of cyclic peptide was

set to 100%. The addition of the selected cyclic peptides to CM from control CHO cell cultures did

not affect SH-SY5Y viability. (d) Cell viability of primary mouse hippocampal neurons (left) and

serum-starved U87MG cells (right) as determined by the MTT assay after treatment for 24 h at

37 °C without Aβ (white bars) or with 1 μΜ solutions of Aβ40 or Aβ42, which had been pre-aggregated

in the presence or absence of 1 μΜ SOD1C5-4 for 3 and 1 d, respectively. In (a), (b), and (d), results

are expressed as the percentage of MTT reduction, assuming that the absorbance of control

(untreated) cells was 100%, and the mean values ± s.e.m. of three independent experiments (n=3)

with six replicate wells for each condition are reported. Statistical significances of the differences in

the levels of viability between cells untreated and treated with Αβ (a) or between cells treated with

Αβ in the presence and absence of the selected cyclic peptides (b-d) are presented.

31

Supplementary Figure 5. The selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116 inhibit Αβ-

induced aggregation and toxicity in vivo in a dose-dependent manner. Paralysis curves of C.

elegans CL4176 expressing human Aβ42 and treated with synthetic AβC5-34 (blue) and AβC5-

116 (red) at the indicated concentrations. The “No peptide” sample (control) is common for all

experiments and contains the same amount of solvent (DMSO) as in the samples containing

synthetic cyclic peptides (0.26% final plate concentration). No peptide: mean=29.00±0.1,

n=651/659. AβC5-34 (2 μM): mean=29.20±0.1, n=144/147, ns; AβC5-34 (5 μΜ): mean=

28.78±0.2, n=600/606, P<0.01; ΑβC5-34 (10 μΜ): mean=31.0±0.2, n=789/806, P<0.0001; AβC5-

34 (15 μΜ): mean=28.79±0.2, n=588/599, ns. AβC5-116 (2 μM): mean=29.77±0.2, n=726/736,

P<0.001; ΑβC5-116 (5 μΜ): mean=31.0±0.1, n=733/743, P<0.0001; AβC5-116 (10 μΜ):

mean=32.65±0.2, n=564/571, P<0.001; AβC5-116 (15 μΜ): mean=33.57±0.1, n=151/153,

P<0.001; AβC5-116 (30 μΜ): mean=33.66±0.2, n=80/87, P<0.001

32

a

b

33

c

Supplementary Figure 6. Sequence analysis of the selected cyclic TXXXR pentapeptides

targeting Αβ. (a) Fluorescence of E. coli BL21(DE3) cells co-expressing Aβ42-GFP and four

individual cyclic peptide sequences appearing after the second round of FACS sorting (Fig. 1b)

in the selected population only at low frequencies as shown in Supplementary Table 5. The

fluorescence of the cell population producing a random cyclic peptide (random 1) was arbitrarily

set to 100. Experiments were carried out in replica triplicates (n=1 independent experiment) and

the reported values correspond to the mean value ± s.e.m. (b) Frequency of appearance of

codons corresponding to the twenty natural amino acids at positions 2, 3, and 4 of the peptide-

encoding region of the pSICLOPPS-NuX1X2X3-X5 vectors contained in the bacterial clones

isolated after the second round of FACS sorting (Fig. 1b) that encoded for TXXXR cyclic

pentapeptides (1,901,945 reads corresponding to TXXXR cyclic pentapeptides out of 4,530,567

total reads that appeared more than 50 times in the sorted peptide pool). (c) Frequency of

appearance of the twenty natural amino acids at positions 2, 3, and 4 of the unique TXXXR cyclic

pentapeptides encoded by the pSICLOPPS-NuX1X2X3-X5 vectors contained in the bacterial

clones isolated after the second round of FACS sorting (Fig. 1b) (159 unique peptide sequences

corresponding to TXXXR cyclic pentapeptides out of 605 total unique selected peptide sequences

that appeared more than 50 times in the sorted peptide pool).

34

a

b

c

35

d

Supplementary Figure 7. Computational modelling of ΑβC5-34 and ΑβC5-116 binding to Αβ. (a)

Initial (left) and 100 ns MD simulation (right) cluster centroid structures of the Aβ protofilament

model used in this study. A-E indicate each one of the five Αβ chains comprising the utilized

pentameric Aβ protofilament model. (b) Energy-minimized structures of the selected cyclic

pentapeptides AβC5-34 and AβC5-116 according to the AMBER ff99SB force field21. Each amino

acid is designated by its single-letter abbreviation. (c) (top) Spatial distribution function (SDF)

isosurfaces of the cyclic peptides ΑβC5-34 (red) and ΑβC5-116 (blue) around the Aβ protofibril.

(bottom) SDF-guided partitioning of the Aβ protofibril into five grid boxes, common for both cyclic

peptides, in order to examine possible binding positions. (d) Energy decomposition analysis of

the Aβ/ΑβC5-34 (top) and Aβ/ΑβC5-116 (bottom) complexes. Amino acids with favorable and

unfavorable binding contributions are indicated in blue and red color, respectively.

36

a b

c

d

e f

37

g

Supplementary Figure 8. Genetic screening for the identification of macrocyclic rescuers of

mutant SOD1 misfolding and aggregation. (a) Relative fluorescence of E. coli Origami 2(DE3)

cells overexpressing chimeric SOD1-GFP fusions from the corresponding pETSOD1-GFP

vectors (left), following the addition of 0.01 mM IPTG and growth at 37 °C for 2 h, and from the

corresponding pASKSOD1-GFP vectors (right), by the addition of 0.2 μg/mL aTc and growth at

37 °C for 2 h,. Wild-type (WT) SOD1-GFP fluorescence was arbitrarily set to 100. Mean values ±

s.e.m. are reported (n=3 independent experiments, each one performed in triplicate). (b) Solubility

analysis of SOD1-GFP fusions overexpressed as in (a, left) by SDS-PAGE/western blotting using

an anti-polyHis antibody. Representative data from n=2 independent experiments are presented.

(c) Solubility analysis of SOD1 variants overexpressed as in as in (a, left) by SDS-PAGE/western

blotting using an anti-polyHis antibody. Representative data from n=2 independent experiments

are presented. The assay was performed using E. coli Origami 2(DE3) cells overexpressing GFP-

free SOD1 from the corresponding pETSOD1 vectors by the addition of 0.01 mM IPTG at 37 °C

for 2 h. (d) SDS-PAGE/western blot analysis using an anti-polyHis antibody of E. coli Origami

2(DE3) cells overexpressing GFP-free SOD1 from the corresponding pASKSOD1 vectors by the

addition of 0.2 μg/mL aTc at 37 °C for 2 h. (e) Western blot analysis using an anti-CBD antibody

of the four individual selected clones investigated in Table 1 (bottom). The upper band of ~25

kDa corresponds to the IC-peptide sequence-IN-CBD precursor, while the lower band of ~20 kDa

corresponds to the processed IN-CBD product, whose appearance is an indication of successful

cyclic peptide formation. CBD: chitin-binding domain. (f) Fluorescence of E. coli Origami 2(DE3)

cells co-expressing SOD1(A4V) or Αβ42-GFP from the vectors pETSOD1(A4V)-GFP or pETΑβ42-

GFP, respectively, together with the cyclic peptides encoded by the selected clones 1-4

investigated in Table 1 (bottom). SOD1(A4V)-GFP fluorescence of the cell population producing

a random cyclic peptide was arbitrarily set to 100. Experiments were carried out in replica

triplicates (n=1 independent experiment) and the reported data correspond to the mean value ±

s.e.m. (g) Solubility analysis of SOD1(A4V)-GFP overexpressed with/without the four selected

cyclic peptide sequences shown in Fig. 7c by SDS-PAGE/western blotting using an anti-polyHis

antibody.

39

b

Supplementary Figure 9. Sequence analysis of the selected cyclic TXSXW pentapeptides

targeting SOD1(A4V). (a) Frequency of appearance of codons corresponding to the twenty

natural amino acids at positions 2 and 4 of the peptide-encoding region of the pSICLOPPS-

NuX1X2X3-X5 vectors contained in the bacterial clones after the fourth round of FACS sorting (Fig.

7b) that encoded for TXSXW cyclic pentapeptides (3,939,406 reads corresponding to TXSXW

cyclic pentapeptides out of 4,243,704 total reads that appeared more than 50 times in the sorted

peptide pool). (b) Frequency of appearance of the twenty natural amino acids at positions 2 and

4 of the unique TXSXW cyclic pentapeptides encoded by the pSICLOPPS-NuX1X2X3-X5 vectors

contained in the bacterial clones isolated after the fourth round of FACS sorting (Fig. 7b) (46

unique peptide sequences corresponding to TXSXW cyclic pentapeptides out of 367 total unique

selected peptide sequences that appeared more than 50 times in the sorted peptide pool).

40

a

b

41

c

d

Supplementary Figure 10. SOD1(A4V) purification and characterization. (a) DLS analysis of

SOD1(A4V) immediately after IMAC purification (as purified) following overexpression in E. coli

Origami 2(DE3) cells at 18 °C (blue) or 37 °C (orange). (b) SEC analysis of SOD1(A4V) as in (a).

(c) SDS-PAGE, native PAGE, and in-gel activity42 analysis of SOD1(A4V) fractions collected after

SEC as described in (b). The esterolytic enzyme EstDZ243 (MW=28.4, pI=5.96) was utilized as a

MW marker for native PAGE. (d) DLS analysis of SOD1(A4V) immediately after SEC following

overexpression in E. coli Origami2(DE3) cells at 18 °C and IMAC purification.

42

Supplementary Tables

Supplementary Table 1. PCR primers used in this study.

Name Primer sequence (5'à3') Use

GS032 GGAATTCGCCAATGGGGCGATCGCCCACAATTGC(NNS)3TGCTTAAGTTTTGGC

Degenerate forward primer for the construction of the CysX1X2X3 sub-library. BglI site is underlined.

GS033 GGAATTCGCCAATGGGGCGATCGCCCACAATAGC(NNS)3TGCTTAAGTTTTGGC

Degenerate forward primer for the construction of the SerX1X2X3 sub-library. BglI site is underlined.

GS034 GGAATTCGCCAATGGGGCGATCGCCCACAATACC(NNS)3TGCTTAAGTTTTGGC

Degenerate forward primer for the construction of the ThrX1X2X3 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the CysX1X2X3X4 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the SerX1X2X3X4 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the ThrX1X2X3X4 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the CysX1X2X3X4X5 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the SerX1X2X3X4X5 sub-library. BglI site is underlined.


Degenerate forward primer for the construction of the ThrX1X2X3X4X5 sub-library. BglI site is underlined.

GS035 AAAAAAAAGCTTTCATTGAAGCTGCCACAAGG

Reverse primer annealing to CBD. HindIII site is underlined.

GS069 AAAAAAGCCAATGGGGCGATCGCCCACAATTGC

Forward zipper primer for the construction of the Cys sub-libraries. BglI site is underlined.

GS070 AAAAAAGCCAATGGGGCGATCGCCCACAATAGC

Forward zipper primer for the construction of the Ser sub-libraries. BglI site is underlined.

GS071 AAAAAAGCCAATGGGGCGATCGCCCACAATACC

Forward zipper primer for the construction of the Thr sub-libraries. BglI site is underlined.

GS100 ATGGCGACGAAGGCCGTGTGCGTGCTGAAGGGCGACGGCCCAGTGCAGGGCATCATC

Forward primer for the SOD1 gene assembly (segment 1).

43

GS101 CACACCTTCACTGGTCCATTACTTTCCTTCTGCTCGAAATTGATGATGCCCTGCACTGGG

Reverse primer for the SOD1 gene assembly (segment 1).

GS102 GGACCAGTGAAGGTGTGGGGAAGCATTAAAGGACTGACTGAAGGCCTGCATGGATTCC


GS103 CTGGTACAGCCTGCTGTATTATCTCCAAACTCATGAACATGGAATCCATGCAGGCC


GS104 CAGCAGGCTGTACCAGTGCAGGTCCTCACTTTAATCCTCTATCCAGAAAACACGG


GS105 GTCTCCAACATGCCTCTCTTCATCCTTTGGCCCACCGTGTTTTCTGGATAGAGG


GS106 GAGAGGCATGTTGGAGACTTGGGCAATGTGACTGCTGACAAAGATGGTGTGGCCG


GS107 CCTGAGAGTGAGATCACAGAATCTTCAATAGACACATCGGCCACACCATCTTTGTC


GS108 CTGTGATCTCACTCTCAGGAGACCATTGCATCATTGGCCGCAC


GS109 GCCCAAGTCATCTGCTTTTTCATGGACCACCAGTGTGCGGCCAATGATGC


GS110 GCAGATGACTTGGGCAAAGGTGGAAATGAAGAAAGTACAAAGACAGGAAACGC


GS111 TTGGGCGATCCCAATTACACCACAAGCCAAACGACTTCCAGCGTTTCCTGTCTTTGTAC


GS059 AAAAAAGGATCCACTAGTTTGGGCGATCCCAATTACACC

Reverse primer for the construction of pETSOD1-GFP, pETSOD1(A4V)-GFP, pETSOD1(G37R)-GFP, pETSOD1(G85R)-GFP and pETSOD1(G93A)-GFP. BamHI site is underlined.

GS060 AAAAAACATATGGCGACGAAGGTGGTGTGCGTGCTG

Forward primer for the construction of pETSOD1(A4V)-GFP. NdeI site is underlined.

GS058 AAAAAACATATGGCGACGAAGGCCGTGTGCGTG

Forward primer for the construction of pETSOD1-GFP, pETSOD1(G37R)-GFP, pETSOD1(G85R)-GFP and pETSOD1(G93A)-GFP. NdeI site is underlined.

GS112 GTGGGGAAGCATTAAAcGACTGACTGAAGGCC

Forward overlap primer for SOD1(G37R) mutagenesis.

GS113 GGCCTTCAGTCAGTCgTTTAATGCTTCCCCAC

Reverse overlap primer for SOD1(G37R) mutagenesis.

GS114 CATGTTGGAGACTTGcGCAATGTGACTGCTG

Forward overlap primer for SOD1(G85R) mutagenesis.

44

GS115 CAGCAGTCACATTGCgCAAGTCTCCAACATG

Reverse overlap primer for SOD1(G85R) mutagenesis.

GS116 CTGCTGACAAAGATGcTGTGGCCGATGTGTC

Forward overlap primer for SOD1(G93A) mutagenesis.

GS117 GACACATCGGCCACAgCATCTTTGTCAGCAG

Reverse overlap primer for SOD1(G93A) mutagenesis.

SP006 AAAAAATCTAGAAGGAGGAAACGATGGACTACAAGGACGACGATGACAAGGCGACGAAGGCCGTGTGCGTG

Forward primer for the construction of pETSOD1(WT), pETSOD1(G37R), pETSOD1(G85R) and pETSOD1(G93A). XbaI site and FLAG-tag are underlined.

SP004 AAAAAAAGCTTGGATCCTTAGTGGTGGTGGTGGTGGTGTTGGGCGATCCCAATTACACC

Reverse primer for the construction of pETSOD1, pETSOD1(A4V)-GFP, pETSOD1(G37R), pETSOD1(G85R) and pETSOD1(G93A). HindIII, a BamHI site and 6×His-tag are underlined.

SP007 AAAAAATCTAGAAGGAGGAAACGATGGACTACAAGGACGACGATGACAAGGCGACGAAGGTGGTGTGCGTG

Forward primer for the construction of pETSOD1(A4V). XbaI site and FLAG-tag are underlined.

GS002 AAAAAAAAGCTTCTCGAGttaGTGGTGGTGGTGGTGGTGTTTGTAGAGTTCATCCATGCC

Reverse primer for the construction of GFP fused protein constructs. HindIII site and 6×His-tag are underlined.

GS118 ATGTCATCTTCTGTCCCTTCCCAGAAAACCTACCAGGGCAGCTACGGTTTCCGTCTGGGC

Forward primer for TP53 gene assembly (segment 1).

GS119 GGAGTACGTGCAAGTCACAGACTTGGCTGTCCCAGAATGCAAGAAGCCCAGACGGAAACC

Reverse primer for TP53 gene assembly (segment 1).

GS120 GACTTGCACGTACTCCCCTGCCCTCAACAAGATGTTTTGCCAACTGGCCAAGACC


GS121 CCGGGCGGGGGTGTGGAATCAACCCACAGCTGCACAGGGCAGGTCTTGGCCAGTTGGC


GS122 GATTCCACACCCCCGCCCGGCACCCGCGTCCGCGCCATGGCCATCTACAAGCAGTCACAG


GS123 CAGCGCTCATGGTGGGGGCAGCGCCTCACAACCTCCGTCATGTGCTGTGACTGCTTGTAG


GS124 CCCACCATGAGCGCTGCTCAGATAGCGATGGTCTGGCCCCTCCTCAGCATCTTATC


GS125 CCAAATACTCCACACGCAAATTTCCTTCCACTCGGATAAGATGCTGAGGAGGG


GS126 GCGTGTGGAGTATTTGGATGACAGAAACACTTTTCGACATAGTGTGGTGGTGCCC


45

GS127 GTGGTACAGTCAGAGCCAACCTCAGGCGGCTCACAGGGCACCACCACACTATG


GS128 GGCTCTGACTGTACCACCATCCACTACAACTACATGTGTAACAGTTCCTGCATG


GS129 GTGTGATGATGGTGAGGATGGGCCTCCGGTTCATGCCGCCCATGCAGGAACTGTTAC


GS130 CCTCACCATCATCACACTGGAAGACTCCAGTGGTAATCTACTGGGACGGAACAGCTTTG


GS131 GTGCGCCGGTCTCTCCCAGGACAGGCACAAACACGCACCTCAAAGCTGTTCCGTCCCAG


GS132 GAGAGACCGGCGCACAGAGGAAGAGAATCTCCGCAAGAAAGGGGAGCCTCACCACG


GS133 GGTGTTGTTGGACAGTGCTCGCTTAGTGCTCCCTGGGGGCAGCTCGTGGTGAGGCTCCCC


GS003 AAAAAATCTAGAAGGAGGAAACGCATATGTCATCTTCTGTCCCTTCCCAG

Forward primer for the construction of p53 protein constructs. XbaI site is underlined.

GS004 AAAAAAGGATCCCTGCAGGGTGTTGTTGGACAGTGCTCG

Reverse primer for the construction of the p53 protein fusion with GFP. BamHI site is underlined.

GS007 AGTGTGGTGGTGCCCTgTGAGCCGCCTGAGGTTG

Forward point mutagenesis primer for the construction of the p53(Y220C)-GFP fusion. Lower case indicates point mutation.

GS008 CAACCTCAGGCGGCTCAcAGGGCACCACCACACT

Forward point mutagenesis primer for the construction of the p53(Y220C)-GFP fusion. Lower case indicates point mutation.

GS043 AAAAAAGCCAATGGGGCATGAGCCATATTCAACGGGAAAC

Forward KanR primer. BglI site is underlined.

DG002 TTTTTTAAGCTTTTAGAAAAACTCATCGAGC

Reverse KanR primer. HindIII site is underlined.

IM033 CTAGCCAATGGGGCGATCGCCCACAATtgcGCCTCGCCGACGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-34(S1C). Lower case indicates modification. BglI site is underlined.

IM034 CTAGCCAATGGGGCGATCGCCCACAATaccGCCTCGCCGACGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-34(S1T). Lower case indicates modification. BglI site is underlined.

IM036 CTAGCCAATGGGGCGATCGCCCACAATAGCGCCgcgCCGACGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-34(S3A). Lower case indicates modification. BglI site is underlined.

46

IM037 CTAGCCAATGGGGCGATCGCCCACAATAGCGCCTCGgcgACGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-34(P4A). Lower case indicates modification. BglI site is underlined.

IM038 CTAGCCAATGGGGCGATCGCCCACAATAGCGCCTCGCCGgcgTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-34(T5A). Lower case indicates modification. BglI site is underlined.

IM027 CTGCTAGCCAATGGGGCGATCGCCCACAATtgcGCGTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(T1C). Lower case indicates modification. BglI site is underlined.

IM028 CTGCTAGCCAATGGGGCGATCGCCCACAATagcGCGTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(T1S). Lower case indicates modification. BglI site is underlined.

IM030 CTGCTAGCCAATGGGGCGATCGCCCACAATACCGCGgcgGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(F3A). Lower case indicates modification. BglI site is underlined.

IM031 CTGCTAGCCAATGGGGCGATCGCCCACAATACCGCGTTCgcgCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(D4A). Lower case indicates modification. BglI site is underlined.

IM032 CTGCTAGCCAATGGGGCGATCGCCCACAATACCGCGTTCGACgcgTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(R5A). Lower case indicates modification. BglI site is underlined.

IM043 CTAGCCAATGGGGCGATCGCCCACAATACCtttTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2F). Lower case indicates modification. BglI site is underlined.

IM044 CTAGCCAATGGGGCGATCGCCCACAATACCagcTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2S). Lower case indicates modification. BglI site is underlined.

IM045 CTAGCCAATGGGGCGATCGCCCACAATACCccgTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2P). Lower case indicates modification. BglI site is underlined.

IM046 CTAGCCAATGGGGCGATCGCCCACAATACCaccTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2T). Lower case indicates modification. BglI site is underlined.

IM047 CTAGCCAATGGGGCGATCGCCCACAATACCtatTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2Y). Lower case indicates modification. BglI site is underlined.

IM048 CTAGCCAATGGGGCGATCGCCCACAATACCcatTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2H). Lower case indicates modification. BglI site is underlined.

IM049 CTAGCCAATGGGGCGATCGCCCACAATACCaaaTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2K). Lower case indicates modification. BglI site is underlined.

IM050 CTAGCCAATGGGGCGATCGCCCACAATACCgaaTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2E). Lower case indicates modification. BglI site is underlined.

47

IM051 CTAGCCAATGGGGCGATCGCCCACAATACCtggTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2W). Lower case indicates modification. BglI site is underlined.

IM052 CTAGCCAATGGGGCGATCGCCCACAATACCcgtTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2R). Lower case indicates modification. BglI site is underlined.

IM039 CTAGCCAATGGGGCGATCGCCCACAATACCTTCGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(A2del). BglI site is underlined.

IM040 CTAGCCAATGGGGCGATCGCCCACAATACCGCGGACCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116 (F3del). BglI site is underlined.

IM041 CTAGCCAATGGGGCGATCGCCCACAATACCGCGTTCCGGTGCTTAAGTTTTGGC

Forward primer for the construction of pSICLOPPS-AβC5-116(D4del). BglI site is underlined.

IM077 CTAGCCAATGGGGCGATCGCCCACAATaccaccaccgtgcgtTGCTTAAGTTTTGGCACCGAAATTTTAACCG

Forward primer for the construction of pSICLOPPS-AβC5-479. Lower case indicates peptide DNA sequence. BglI site is underlined.

IM078 CTAGCCAATGGGGCGATCGCCCACAATaccgcgatgtggcgtTGCTTAAGTTTTGGCACCGAAATTTTAACCG


IM080 CTAGCCAATGGGGCGATCGCCCACAATaccgtgtggattcgtTGCTTAAGTTTTGGCACCGAAATTTTAACCG


IM081 CTAGCCAATGGGGCGATCGCCCACAATaccagccatgcgcgtTGCTTAAGTTTTGGCACCGAAATTTTAACCG


GS037 CTATAACTATGGCTGGAATG Forward primer annealing to the pSICLOPPS backbone, before the 5’-end of the C-terminal domain of the Ssp DnaE intein.

DD015 TTTTTTGCCCCATTGGCTAGCAGAgcATTAaGGTCTTGGGGAAGACCAATATC

Reverse primer for the H24L/F26A mutagenesis of the C-terminal domain of the Ssp DnaE intein (IC). Lower case indicates modification. BglI site is underlined.

48

Supplementary Table 2. Bacterial expression vectors used in this study.

Plasmid Encoded Protein Marker Origin of

replication Source

pΕΤΑβ42-GFP Aβ42-GFP KanR ColE1 Prof. M. H.

Hecht

pΕΤΑβ42(F19S;L34P)-GFP Αβ42(F19S;L34P)-GFP KanR ColE1 Prof. M. H.

Hecht

pETSOD1-GFP SOD1-GFP KanR ColE1 This work

pETSOD1(A4V)-GFP SOD1(A4V)-GFP KanR ColE1 This work

pETSOD1(G37R)-GFP SOD1(G37R)-GFP KanR ColE1 This work

pETSOD1(G85R)-GFP SOD1(G85R)-GFP KanR ColE1 This work

pETSOD1(G93A)-GFP SOD1(G93A)-GFP KanR ColE1 This work

pETSOD1 FLAG-SOD1-6×His KanR ColE1 This work

pETSOD1(A4V) FLAG-SOD1(A4V)-6×His KanR ColE1 This work

pETSOD1(G37R) FLAG-SOD1(G37R)-6×His KanR ColE1 This work

pETSOD1(G85R) FLAG-SOD1(G85R)-6×His KanR ColE1 This work

pETSOD1(G93A) FLAG-SOD1(G93A)-6×His KanR ColE1 This work

pETp53-GFP p53C-GFP KanR ColE1 This work

pETp53(Y220C)-GFP p53C(Y220C)-GFP KanR ColE1 This work

pASKp53-GFP p53C-GFP-6×His AmpR ColE1 This work

pASKSOD1-GFP SOD1-GFP-6×His AmpR ColE1 This work

pASKSOD1(A4V)-GFP SOD1(A4V)-GFP-6×His AmpR ColE1 This work

pASKSOD1(G37R)-GFP SOD1(G37R)-GFP-6×His AmpR ColE1 This work

pASKSOD1(G85R)-GFP SOD1(G85R)-GFP-6×His AmpR ColE1 This work

pASKSOD1(G93A)-GFP SOD1(G93A)-GFP-6×His AmpR ColE1 This work

pASKSOD1 FLAG-SOD1-6×His AmpR ColE1 This work

pASKSOD1(A4V) FLAG-SOD1(A4V)-6×His AmpR ColE1 This work

pASKSOD1(G85R) FLAG-SOD1(G85R)-6×His AmpR ColE1 This work

pASKSOD1(G93R) FLAG-SOD1(G93A)-6×His AmpR ColE1 This work

pSICLOPPS IC-SGGYLPPL-IN-CBD CmR ACYC Prof. S.

Benkovic

pSICLOPPS-CysX1X2X3 sub-library IC-CysX1X2X3-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-SerX1X2X3 sub-library IC-SerX1X2X3-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-ThrX1X2X3 sub-library IC-ThrX1X2X3-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-CysX1X2X3X4 sub-library

IC-CysX1X2X3X4-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-SerX1X2X3X4 sub-library

IC-SerX1X2X3X4-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-ThrX1X2X3X4 sub-library

IC-ThrX1X2X3X4-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-CysX1X2X3X4X5 sub-library

IC-CysX1X2X3X4X5-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-SerX1X2X3X4X5 sub-library

IC-SerX1X2X3X4X5-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-ThrX1X2X3X4X5 sub-library

IC-ThrX1X2X3X4X5-IN-CBD sub-library

CmR ACYC This work

pSICLOPPS-ΑβC5-34 IC-SASPT-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-34(S1C) IC-CASPT-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-34(S1T) IC-TASPT-IN-CBD CmR ACYC This work

49

pSICLOPPS-ΑβC5-34(S3A) IC-SAAPT-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-34(P4A) IC-SASAT-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-34(T5A) IC-SASPA-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116 IC-TAFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(T1C) IC-CAFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(T1S) IC-SAFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(F3A) IC-TAADR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(D4A) IC-TAFAR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(R5A) IC-TAFDA-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2F) IC-TFFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2W) IC-TWFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2Y) IC-TYFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2S) IC-TSFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2T) IC-TTFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2E) IC-TEFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2R) IC-TRFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2H) IC-THFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2K) IC-TKFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(A2P) IC-TPFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(delA2) IC-TFDR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(delF3) IC-TADR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-116(delD4) IC-TAFR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-325 IC-TVWIR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-359 IC-TAMWR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-413 IC-TSHAR-IN-CBD CmR ACYC This work

pSICLOPPS-ΑβC5-479 IC-TTTVR-IN-CBD CmR ACYC This work

pSICLOPPS-Random1 IC-unknown peptide sequence1-IN-CBD

CmR ACYC This work

pSICLOPPS-Random2 IC-unknown peptide sequence2-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-2 IC(H24L;F26A)-TTYAR-IN-CBD CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-3 IC(H24L;F26A)-TTVDR-IN-CBD CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-17 IC(H24L;F26A)-TTTAR-IN-CBD CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-21 IC(H24L;F26A)-TTWCR-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-26 IC(H24L;F26A)-TAWCR-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-34 IC(H24L;F26A)-SASPT-IN-CBD CmR ACYC This work

pSICLOPPS(H24L;F26A)-ΑβC5-116

IC(H24L;F26A)-TAFDR-IN-CBD CmR ACYC This work

pSICLOPPS(H24L;F26A)-AβC6-1 IC(H24L;F26A-TPVWFD-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-SOD1C5-2

IC(H24L;F26A)-TASFW-IN-CBD

CmR ACYC This work


IC(H24L;F26A)-TWSVW-IN-CBD

CmR ACYC This work


IC(H24L;F26A)-TFSMW-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-Random1

IC(H24L;F26A)-unknown peptide sequence1-IN-CBD

CmR ACYC This work

pSICLOPPS(H24L;F26A)-Random2

IC(H24L; F26A)- unknown peptide sequence2-IN-CBD

CmR ACYC This work

50

Supplementary Table 3. Sequencing results of the peptide-encoding regions of 23 randomly

selected clones from the constructed pSICLOPPS-NuX1X2X3, pSICLOPPS-NuX1X2X3X4, and

pSICLOPPS-NuX1X2X3X4X5 vector sub-libraries.

Clone number DNA sequence of peptide-encoding

gene Encoded peptide

sequence

C4-1 TGC GGC AAG GTG CGKV

C4-2 TGC CGC CAC CGG CRHR

C4-3 AGC GCG TCC GGG SASG

C4-4 AGC ACG CGC CGG STRR

C4-5 ACC AAC TGG GTC TNWV

C4-6 ACC AGG GCC TCC TRAS

C4-7 AGC CGG GTG CTC SRVL

C4-8 ACC AAC TGG CCG TNWP

C5-1 TGC AAC TTG GTC TGG CNLVW

C5-2 TGC TGC GCG GCG GGG CCAAG

C5-3 TGC GCG TCG CGG GGG CASRG

C5-4 AGC TTC GTG GAG GGG SFVEG

C5-5 ACC TGC CCC GTG TAG TCPV*

C5-6 ACC CCG GCG CGG TGC TPARC

C5-7 ACC TCG GGC GCG TAG TSGA*

C6-1 TGC GGG CGG GGG TGG ACG CGRGWT

C6-2 TGC TGC AGC GGC TGC CGG CCSGCR

C6-3 TGC AAG TCG GGG CAC GGC CKSGHG

C6-4 AGC TTG GTG CCG TAC CTG SLVPYL

C6-5 AGC GCC TAG GGC GGG CCC SA*GGP

C6-6 AGC GAG GGG GGG GGG G Frame shift

C6-7 ACC TCG CTC TAG TCC CAC TSL*SH

C6-8 ACC AGG GGG GGC AGG GGG TRGGRG

*: stop codon

51

Supplementary Table 4. High-throughput sequencing analysis of the peptide-encoding regions

of ~260,000 randomly selected clones from the constructed pSICLOPPS-NuX1X2X3,

pSICLOPPS-NuX1X2X3X4, and pSICLOPPS-Nu X1X2X3X4X5 sub-libraries.

Number of reads

Unique DNA sequences

Unique peptide sequences

cyclo-NuX1X2X3 tetrapeptides

3,448 2,648 (77%) 2,034 (59%)

cyclo-NuX1X2X3X4

pentapeptides 6,348 5,519 (87%) 4,537 (71%)

cyclo-NuX1X2X3X4X5

hexapeptides 250,756 213,763 (85%) 168,003 (67%)

Combined cyclo-NuX1X2X3-X5 library

260,552 221,930 (85%) 174,574 (67%)

Supplementary Table 5. Sequences and frequency of appearance of the selected cyclic TXXXR

pentapeptides as determined by high-throughput sequencing of the isolated pSICLOPPS-

NuX1X2X3-X5 vectors after the second round of bacterial sorting for enhanced Αβ42-GFP

fluorescence.

Nu

mb

er

Peptide name

Aminoacid sequence Number of reads

Reads/Total TXXXR

reads (%)

Reads/Total pentapeptide

reads (%)

Reads/Total peptide

reads (%)

1 ΑβC5-2 T T Y A R 304,753 16.023 7.506 6.727

2 ΑβC5-3 T T V D R 214,461 11.276 5.282 4.734

3 ΑβC5-5 T T T W R 175,510 9.228 4.323 3.874

4 ΑβC5-7 T T L H R 134,018 7.046 3.301 2.958

5 ΑβC5-8 T T F A R 96,700 5.084 2.382 2.134

6 ΑβC5-9 T V L D R 89,669 4.715 2.209 1.979

7 ΑβC5-12 T T W A R 65,929 3.466 1.624 1.455

8 ΑβC5-13 T A L D R 62,792 3.301 1.547 1.386

9 ΑβC5-15 T A N V R 47,855 2.516 1.179 1.056

10 ΑβC5-17 T T T A R 40,135 2.110 0.989 0.886

11 ΑβC5-18 T T I A R 37,150 1.953 0.915 0.820

12 ΑβC5-19 T V W D R 37,091 1.950 0.914 0.819

13 ΑβC5-20 T T I S R 37,044 1.948 0.912 0.818

14 ΑβC5-21 T T W C R 36,295 1.908 0.894 0.801

15 ΑβC5-22 T V L W R 35,820 1.883 0.882 0.791

16 ΑβC5-25 T T L A R 28,989 1.524 0.714 0.640

17 ΑβC5-26 T A W C R 28,391 1.493 0.699 0.627

18 ΑβC5-27 T T S A R 28,188 1.482 0.694 0.622

19 ΑβC5-29 T T L E R 27,514 1.447 0.678 0.607

20 ΑβC5-30 T S T A R 27,456 1.444 0.676 0.606

21 ΑβC5-35 T V R D R 25,428 1.337 0.626 0.561

22 ΑβC5-41 T G W A R 21,784 1.145 0.537 0.481

23 ΑβC5-44 T A W A R 20,807 1.094 0.512 0.459

52

24 ΑβC5-45 T T W V R 20,798 1.094 0.512 0.459

25 ΑβC5-46 T L L W R 19,957 1.049 0.492 0.440

26 ΑβC5-47 T T I D R 19,735 1.038 0.486 0.436

27 ΑβC5-50 T A L A R 19,433 1.022 0.479 0.429

28 ΑβC5-51 T S V D R 19,249 1.012 0.474 0.425

29 ΑβC5-53 T T V W R 18,669 0.982 0.460 0.412

30 ΑβC5-66 T T H W R 14,304 0.752 0.352 0.316

31 ΑβC5-67 T A R D R 14,213 0.747 0.350 0.314

32 ΑβC5-73 T T R D R 12,894 0.678 0.318 0.285

33 ΑβC5-80 T S V H R 10,181 0.535 0.251 0.225

34 ΑβC5-82 T A V W R 9,781 0.514 0.241 0.216

35 ΑβC5-83 T T G C R 9,362 0.492 0.231 0.207

36 ΑβC5-89 T A T D R 7,984 0.420 0.197 0.176

37 ΑβC5-94 T V L F R 7,442 0.391 0.183 0.164

38 ΑβC5-102 T T Y N R 6,067 0.319 0.149 0.134

39 ΑβC5-105 T V R W R 5,450 0.287 0.134 0.120

40 ΑβC5-116 T A F D R 4,243 0.223 0.105 0.094

41 ΑβC5-117 T T R C R 4,237 0.223 0.104 0.094

42 ΑβC5-118 T T F W R 4,216 0.222 0.104 0.093

43 ΑβC5-121 T I K D R 3,970 0.209 0.098 0.088

44 ΑβC5-123 T T V H R 3,371 0.177 0.083 0.074

45 ΑβC5-126 T T L L R 3,016 0.159 0.074 0.067

46 ΑβC5-129 T T L F R 2,630 0.138 0.065 0.058

47 ΑβC5-130 T A Y H R 2,594 0.136 0.064 0.057

48 ΑβC5-136 T A L H R 2,026 0.107 0.050 0.045

49 ΑβC5-139 T T S P R 1,904 0.100 0.047 0.042

50 ΑβC5-146 T T W S R 1,612 0.085 0.040 0.036

51 ΑβC5-147 T A M H R 1,611 0.085 0.040 0.036

52 ΑβC5-155 T S L D R 1,251 0.066 0.031 0.028

53 ΑβC5-158 T T G A R 1,172 0.062 0.029 0.026

54 ΑβC5-162 T S V W R 1,094 0.058 0.027 0.024

55 ΑβC5-173 T T H A R 953 0.050 0.023 0.021

56 ΑβC5-176 T A G W R 945 0.050 0.023 0.021

57 ΑβC5-177 T A T A R 925 0.049 0.023 0.020

58 ΑβC5-184 T V L A R 818 0.043 0.020 0.018

59 ΑβC5-185 T T F N R 800 0.042 0.020 0.018

60 ΑβC5-188 T G M R R 768 0.040 0.019 0.017

61 ΑβC5-189 T T V A R 757 0.040 0.019 0.017

62 AβC5-190 T L C L R 739 0.039 0.018 0.016

63 ΑβC5-192 T G L A R 720 0.038 0.018 0.016

64 ΑβC5-198 T S W C R 679 0.036 0.017 0.015

65 ΑβC5-209 T T R A R 580 0.030 0.014 0.013

66 ΑβC5-215 T T P W R 524 0.028 0.013 0.012

53

67 ΑβC5-218 T V L H R 497 0.026 0.012 0.011

68 ΑβC5-223 T G L D R 464 0.024 0.011 0.010

69 ΑβC5-230 T T S D R 442 0.023 0.011 0.010

70 ΑβC5-239 T T M H R 384 0.020 0.009 0.008

71 ΑβC5-242 T T S T R 376 0.020 0.009 0.008

72 ΑβC5-244 T T R V R 366 0.019 0.009 0.008

73 ΑβC5-245 T T R F R 364 0.019 0.009 0.008

74 ΑβC5-248 T T T H R 339 0.018 0.008 0.007

75 ΑβC5-250 T H A W R 334 0.018 0.008 0.007

76 ΑβC5-252 T V I W R 331 0.017 0.008 0.007

77 ΑβC5-253 T T W F R 327 0.017 0.008 0.007

78 ΑβC5-255 T T S R R 325 0.017 0.008 0.007

79 ΑβC5-258 T T S C R 301 0.016 0.007 0.007

80 ΑβC5-260 T T W T R 295 0.016 0.007 0.007

81 ΑβC5-262 T T S S R 286 0.015 0.007 0.006

82 ΑβC5-263 T H L A R 284 0.015 0.007 0.006

83 ΑβC5-264 T S G A R 282 0.015 0.007 0.006

84 ΑβC5-266 T T L R R 274 0.014 0.007 0.006

85 ΑβC5-270 T A T W R 266 0.014 0.007 0.006

86 ΑβC5-272 T C M W R 254 0.013 0.006 0.006

87 ΑβC5-275 T A H V R 249 0.013 0.006 0.005

88 ΑβC5-276 T S W A R 249 0.013 0.006 0.005

89 ΑβC5-278 T T W L R 241 0.013 0.006 0.005

90 ΑβC5-291 T T L D R 213 0.011 0.005 0.005

91 ΑβC5-294 T T P H R 207 0.011 0.005 0.005

92 ΑβC5-298 T T R G R 201 0.011 0.005 0.004

93 ΑβC5-299 T T V G R 200 0.011 0.005 0.004

94 ΑβC5-301 T T T R R 191 0.010 0.005 0.004

95 ΑβC5-304 T S I N R 182 0.010 0.004 0.004

96 ΑβC5-305 T T A D R 181 0.010 0.004 0.004

97 ΑβC5-315 T T S E R 158 0.008 0.004 0.003

98 ΑβC5-316 T T C A R 157 0.008 0.004 0.003

99 ΑβC5-317 T T A W R 156 0.008 0.004 0.003

100 ΑβC5-320 T T V E R 150 0.008 0.004 0.003

101 ΑβC5-321 T T T F R 148 0.008 0.004 0.003

102 ΑβC5-323 T A V D R 147 0.008 0.004 0.003

103 ΑβC5-325 T V W I R 144 0.008 0.004 0.003

104 ΑβC5-329 T T V R R 141 0.007 0.003 0.003

105 ΑβC5-333 T H V R R 137 0.007 0.003 0.003

106 ΑβC5-343 T N L D R 125 0.007 0.003 0.003

107 ΑβC5-344 T T P G R 125 0.007 0.003 0.003

108 ΑβC5-348 T T L T R 119 0.006 0.003 0.003

109 ΑβC5-355 T A T V R 115 0.006 0.003 0.003

54

110 ΑβC5-359 T A M W R 110 0.006 0.003 0.002

111 ΑβC5-361 T T K W R 108 0.006 0.003 0.002

112 ΑβC5-362 T T W D R 107 0.006 0.003 0.002

113 ΑβC5-364 T T M A R 106 0.006 0.003 0.002

114 ΑβC5-365 T T G G R 106 0.006 0.003 0.002

115 ΑβC5-366 T T M V R 105 0.006 0.003 0.002

116 ΑβC5-375 T N L A R 97 0.005 0.002 0.002

117 ΑβC5-376 T I R D R 96 0.005 0.002 0.002

118 ΑβC5-378 T T T G R 96 0.005 0.002 0.002

119 ΑβC5-379 T R L G R 95 0.005 0.002 0.002

120 ΑβC5-381 T T H T R 93 0.005 0.002 0.002

121 ΑβC5-382 T T I T R 92 0.005 0.002 0.002

122 ΑβC5-384 T T Y T R 90 0.005 0.002 0.002

123 ΑβC5-385 T T L Y R 90 0.005 0.002 0.002

124 ΑβC5-389 T H L D R 89 0.005 0.002 0.002

125 ΑβC5-391 T L L I R 88 0.005 0.002 0.002

126 ΑβC5-392 T T C D R 87 0.005 0.002 0.002

127 ΑβC5-393 T T G R R 87 0.005 0.002 0.002

128 ΑβC5-394 T T V S R 86 0.005 0.002 0.002

129 ΑβC5-395 T T Q H R 85 0.004 0.002 0.002

130 ΑβC5-396 T T T P R 84 0.004 0.002 0.002

131 ΑβC5-399 T A F A R 82 0.004 0.002 0.002

132 ΑβC5-405 T T S H R 78 0.004 0.002 0.002

133 ΑβC5-410 T V L G R 76 0.004 0.002 0.002

134 ΑβC5-411 T T Q R R 75 0.004 0.002 0.002

135 ΑβC5-413 T S H A R 74 0.004 0.002 0.002

136 ΑβC5-415 T T T C R 74 0.004 0.002 0.002

137 ΑβC5-422 T A W R R 72 0.004 0.002 0.002

138 ΑβC5-428 T T C G R 69 0.004 0.002 0.002

139 ΑβC5-434 T T S G R 65 0.003 0.002 0.001

140 ΑβC5-438 T T T S R 62 0.003 0.002 0.001

141 ΑβC5-440 T A T G R 61 0.003 0.002 0.001

142 ΑβC5-441 T A W D R 61 0.003 0.002 0.001

143 ΑβC5-443 T T H H R 60 0.003 0.001 0.001

144 ΑβC5-448 T A Y A R 58 0.003 0.001 0.001

145 ΑβC5-449 T A N A R 58 0.003 0.001 0.001

146 ΑβC5-450 T R D V R 58 0.003 0.001 0.001

147 ΑβC5-452 T H V D R 58 0.003 0.001 0.001

148 ΑβC5-453 T L F W R 57 0.003 0.001 0.001

149 ΑβC5-459 T T A A R 55 0.003 0.001 0.001

150 ΑβC5-463 T V V D R 54 0.003 0.001 0.001

151 ΑβC5-464 T T P A R 54 0.003 0.001 0.001

152 ΑβC5-469 T T I G R 53 0.003 0.001 0.001

55

153 ΑβC5-472 T M Y A R 51 0.003 0.001 0.001

154 ΑβC5-473 T H V A R 51 0.003 0.001 0.001

155 ΑβC5-474 T T W P R 51 0.003 0.001 0.001

156 ΑβC5-475 T T G D R 51 0.003 0.001 0.001

157 ΑβC5-479 T T T V R 50 0.003 0.001 0.001

158 ΑβC5-481 T V F G R 50 0.003 0.001 0.001

159 ΑβC5-483 T R V G R 50 0.003 0.001 0.001

Sum 1,901,945 100 46.847 41.980

Supplementary Table 6. Sequences and frequency of appearance of the selected cyclic

pentapeptides resembling ΑβC5-34 as determined by high-throughput sequencing of the isolated

pSICLOPPS-NuX1X2X3-X5 vectors after the second round of bacterial sorting for enhanced Αβ42-

GFP fluorescence.

Supplementary Table 7. Sequences and frequency of appearance of the selected cyclic TXXR

tetrapeptides as determined by high-throughput sequencing of the isolated pSICLOPPS-

NuX1X2X3-X5 vectors after the second round of bacterial sorting for enhanced Αβ42-GFP

fluorescence.

Nu

mb

er

Peptide name

Amino acid sequence

Number of reads

Reads/ Total TXXR

reads (%)

Reads/ Total tetra-

peptide reads (%)

Reads/ Total

peptide reads (%)

1 ΑβC4-9 T T C R 258 34.492 1.428 0.006

2 ΑβC4-11 T T R R 248 33.155 1.372 0.005

3 ΑβC4-31 T T S R 67 8.957 0.371 0.001

4 ΑβC4-34 T R G R 63 8.422 0.349 0.001

5 ΑβC4-35 T T G R 61 8.155 0.338 0.001

6 ΑβC4-41 T R R R 51 6.818 0.282 0.001

Sum 748 100 4.139 0.017

Nu

mb

er

Peptide name

Amino acid sequence

Number of reads

Reads/ Total SASPT-like reads (%)

Reads/ Total

pentapeptide reads (%)

Reads/ Total peptide

reads (%)

1 ΑβC5-34 S A S P T 25673 97.349 0.632 0.567

2 ΑβC5-216 S I C P T 516 1.957 0.013 0.011

3 ΑβC5-380 S I T P T 94 0.356 0.002 0.002

4 ΑβC5-387 S H S P T 89 0.337 0.002 0.002 Sum 26,372 100 0.645 0.578

56

Supplementary Table 8. MM–PBSA binding free energy (ΔGbind) calculations of the Aβ-cyclic

peptide complexes (units are in kcal mol-1). Enthalpy (ΔH) values correspond to calculations

performed on a total of 8,000 frames of 20×4 ns trajectories, while entropy (–TΔS) was calculated

for 1,000 frames corrected from the same trajectories. Total binding energy (ΔGbind) and standard

error of the mean values are also provided.

Cyclic peptide ΔH –TΔS ΔGbind

ΑβC5-34 –23.68 ± 0.06 18.79 ± 0.15 –4.88 ± 0.08

ΑβC5-116 –29.68 ± 0.04a 23.50 ± 0.14a –6.18 ± 0.06b aStandard error of the mean, 𝑠 = (𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛)/√𝑁, where N is the number of trajectory

frames used in calculations (8,000 for enthalpy (𝑛1) and 1,000 for entropy (𝑛2), respectively).

bPooled standard error of the mean, 𝑆 = √[(𝑛1 − 1)(𝑠1)2 + (𝑛2 − 1)(𝑠2)2]/(𝑛1 + 𝑛2 − 2)

Supplementary Table 9. MM–PBSA calculated free-energy contributions and standard error of

the mean values of the Aβ-cyclic peptide complexes.

Cyclic peptide Energy (kcal mol–1)

ΑβC5-34 ΔEvdW –31.60 ± 0.03

ΔEelec –41.73 ± 0.11

ΔEMM, gas –73.33 ± 0.07

ΔGPB 52.35 ± 0.09

ΔGelec(tot)a 10.62 ± 0.10

ΔGNP –2.70 ± 0.00

ΔGsolv 49.65 ± 0.05

ΑβC5-116 ΔEvdW –42.04 ± 0.04

ΔEelec –29.79 ± 0.06

ΔEMM, gas –68.83 ± 0.05

ΔGPB 43.05 ± 0.05

ΔGelec(tot)a 16.26 ± 0.06

ΔGNP –3.90 ± 0.00

ΔGsolv 39.16 ± 0.03 aΔGelec(tot)= ΔEelec + ΔGPB

57

Supplementary Table 10. Hydrogen-bonding interactions between ΑβC5-34 and the Aβ

pentameric model unit. Aβ participating residues are designated by their monomer unit letter.

Interaction Acceptor Acceptor Atom

Donor Donor atom

Occurrence (%)

1 ΑβC5-34-T5 OG1 A-A21 N 67.4

2 A-A21 O ΑβC5-34-T5 N 40.3

3 ΑβC5-34-S1 O A-D23 N 38.7

4 ΑβC5-34-A2 O A-I31 N 29.6

5 A-A21 O ΑβC5-34-S1 N 29.5

6 A-E22 OE1 ΑβC5-34-S1 OG 18.4

7 A-E22 OE1 ΑβC5-34-S1 OG 17.6

Supplementary Table 11. Hydrogen bonding interactions between ΑβC5-116 and the Aβ

pentameric model unit. Aβ participating residues are designated by their monomer unit letter.

Interaction Acceptor Acceptor Atom

Donor Donor atom Occurrence (%)

1 D-I32 O ΑβC5-116-T1 OG1 80.2

2 B-L34 O ΑβC5-116-R5 NH2 67.7

3 A-G26 O ΑβC5-116-R5 NH1 59.6

58

Supplementary Table 12. MM–PBSA binding free energy (ΔGbind) calculations of Aβ complexes

with active and inactive variants of the selected cyclic pentapeptides ΑβC5-34 and ΑβC5-116

(units are in kcal mol-1). Enthalpy (ΔH) values correspond to calculations performed on a total of

8,000 frames of 20×4 ns trajectories, while entropy (–TΔS) was calculated for 1,000 frames

corrected from the same trajectories. Total binding energy (ΔGbind) and standard error of the mean

values are also provided.

Cyclic peptide ΔH –TΔS ΔGbind

ΑβC5-34(T5A) –10.72 ± 0.03a 16.03 ± 0.15a 5.31 ± 0.06b

ΑβC5-116(A2T) –28.03 ± 0.04 24.90 ± 0.17 –3.13 ± 0.07

ΑβC5-116(R5A) –21.14 ± 0.04 23.87 ± 0.17 2.73 ± 0.07

Scrambled ΑβC5-116 (cyclo-TRDFA)

–16.40 ± 0.04 22.48 ± 0.16 6.08 ± 0.07

aStandard error of the mean, 𝑠 = (𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛)/√𝑁, where N is the number of trajectory

frames used in calculations (8,000 for enthalpy (𝑛1) and 1,000 for entropy (𝑛2), respectively).

bPooled standard error of the mean, 𝑆 = √[(𝑛1 − 1)(𝑠1)2 + (𝑛2 − 1)(𝑠2)2]/(𝑛1 + 𝑛2 − 2).

Supplementary Table 13. Sequences and frequency of appearance of the selected cyclic

TXSXW pentapeptides as determined by high-throughput sequencing of the isolated

pSICLOPPS-NuX1X2X3-X5 vectors after the fourth round of bacterial sorting for enhanced

SOD1(A4V)-GFP fluorescence.

Nu

mb

er

Peptide name

Aminoacid sequence Number of reads

Reads/ Total

TXSXW reads (%)

Reads/Total pentapeptide

reads (%)

Reads/ Total

peptide reads (%)

1 SOD1C5-1 T A S W W 1255761 31.877 30.963 29.591

2 SOD1C5-2 T A S F W 744622 18.902 18.36 17.547

3 SOD1C5-3 T S S F W 700047 17.77 17.261 16.496

4 SOD1C5-4 T W S V W 543999 13.809 13.413 12.819

5 SOD1C5-5 T A S H W 330358 8.386 8.146 7.785

6 SOD1C5-6 T F S M W 208879 5.302 5.15 4.922

7 SOD1C5-7 T A S M W 108582 2.756 2.677 2.559

8 SOD1C5-9 T V S F W 31319 0.795 0.772 0.738

9 SOD1C5-11 T L S F W 3069 0.078 0.076 0.072

10 SOD1C5-13 T A S R W 1485 0.038 0.037 0.035

11 SOD1C5-14 T A S S W 1459 0.037 0.036 0.034

12 SOD1C5-18 T A S L W 1054 0.027 0.026 0.025

13 SOD1C5-20 T S S S W 966 0.025 0.024 0.023

14 SOD1C5-23 T G S V W 751 0.019 0.019 0.018

15 SOD1C5-25 T W S L W 683 0.017 0.017 0.016

16 SOD1C5-27 T L S M W 619 0.016 0.015 0.015

17 SOD1C5-31 T W S A W 576 0.015 0.014 0.014

18 SOD1C5-32 T G S W W 563 0.014 0.014 0.013

19 SOD1C5-33 T R S V W 554 0.014 0.014 0.013

59

20 SOD1C5-39 T S S L W 432 0.011 0.011 0.01

21 SOD1C5-44 T A S T W 361 0.009 0.009 0.009

22 SOD1C5-46 T S S V W 356 0.009 0.009 0.008

23 SOD1C5-53 T T S W W 295 0.007 0.007 0.007

24 SOD1C5-65 T A S V W 245 0.006 0.006 0.006

25 SOD1C5-74 T C S W W 208 0.005 0.005 0.005

26 SOD1C5-75 T P S F W 208 0.005 0.005 0.005

27 SOD1C5-76 T T S F W 203 0.005 0.005 0.005

28 SOD1C5-80 T F S T W 171 0.004 0.004 0.004

29 SOD1C5-82 T S S M W 164 0.004 0.004 0.004

30 SOD1C5-89 T V S W W 144 0.004 0.004 0.003

31 SOD1C5-93 T D S W W 136 0.003 0.003 0.003

32 SOD1C5-105 T S S W W 112 0.003 0.003 0.003

33 SOD1C5-106 T R S W W 106 0.003 0.003 0.002

34 SOD1C5-109 T W S M W 99 0.003 0.002 0.002

35 SOD1C5-116 T A S G W 96 0.002 0.002 0.002

36 SOD1C5-130 T P S W W 82 0.002 0.002 0.002

37 SOD1C5-135 T R S F W 79 0.002 0.002 0.002

38 SOD1C5-140 T S S Y W 76 0.002 0.002 0.002

39 SOD1C5-143 T L S V W 72 0.002 0.002 0.002

40 SOD1C5-148 T Y S W W 71 0.002 0.002 0.002

41 SOD1C5-160 T F S V W 62 0.002 0.002 0.001

42 SOD1C5-164 T C S V W 60 0.002 0.001 0.001

43 SOD1C5-167 T V S S W 59 0.001 0.001 0.001

44 SOD1C5-168 T R S H W 58 0.001 0.001 0.001

45 SOD1C5-182 T G S A W 53 0.001 0.001 0.001

46 SOD1C5-188 T A S Y W 52 0.001 0.001 0.001

Sum 3,939,406 100 97.134 92.829

60

Supplementary Table 14. Comparison of the molecular properties of the selected cyclic

pentapeptides with those of conventional drugs, oral macrocyclic (MC) drugs and non-oral MC

drugs.

Propertya Conventional

drugs Oral MC drugsb

Non-oral MC drugsb

AβC5-34c AβC5-116c SOD1C5-4c

MW ≤500 600 to1200 600 to1300 443 590 659

cLogP ≤5 -2 to 6 -7 to 2 -1.9 -3.5 2.3

PSA (Å2) ≤140 180 to 320 150 to 500 197 265 210

HBDs ≤5 ≤12 ≤17 7 10 9

HBAs ≤10 12 to 16 9 to 20 8 10 9

NRB ≤10 ≤15 ≤30 3 10 7 aAbbreviations - MW: molecular weight; cLogP: calculated octanol/water partition coefficient; PSA: polar surface area; HBD: hydrogen bond donor; HBA: hydrogen bond acceptor; NRB: number of rotatable bonds. bAccording to Villar et al.44 cAs determined using PerkinElmer ChemBio3D

61

Supplementary References

1. Wang, G. & Fersht, A.R. First-order rate-determining aggregation mechanism of p53 and its

implications. Proc. Natl. Acad. Sci. U.S.A. 109, 13590-13595 (2012).

2. Stathopulos, P. et al. Cu/Zn superoxide dismutase mutants associated with amyotrophic

lateral sclerosis show enhanced formation of aggregates in vitro. Proc. Natl. Acad. Sci. U.S.A.

100, 7021-7026 (2003).

3. Nilsberth, C. et al. The 'Arctic' APP mutation (E693G) causes Alzheimer's disease by

enhanced Aβ protofibril formation. Nat. Neurosci. 4, 887-893 (2001).

4. Soto, C. et al. β-sheet breaker peptides inhibit fibrillogenesis in a rat brain model of

amyloidosis: implications for Alzheimer's therapy. Nat. Med. 4, 822-826 (1998).

5. Tavassoli, A. & Benkovic, S.J. Split-intein mediated circular ligation used in the synthesis of

cyclic peptide libraries in E. coli. Nat. Protoc. 2, 1126-1133 (2007).

6. Skerra, A. Use of the tetracycline promoter for the tightly regulated production of a murine

antibody fragment in Escherichia coli. Gene 151, 131-135 (1994).

7. Walsh, D.M. et al. Naturally secreted oligomers of amyloid β protein potently inhibit

hippocampal long-term potentiation in vivo. Nature 416, 535-539 (2002).

8. Fotinopoulou, A. et al. BRI2 interacts with amyloid precursor protein (APP) and regulates

amyloid β (Aβ) production. J. Biol. Chem. 280, 30768-30772 (2005).

9. Lührs, T. et al. 3D structure of Alzheimer's amyloid-β(1-42) fibrils. Proc. Natl. Acad. Sci. U.S.A.

102, 17342-17347 (2005).

10. Bieschke, J. et al. Small-molecule conversion of toxic oligomers to nontoxic β-sheet–rich

amyloid fibrils. Nat. Chem. Biol. 8, 93-101 (2012).

11. Schrödinger, L.L.C. The PyMOL Molecular Graphics System, Version 1.8 (2015).

12. Kahler, A., Sticht, H. & Horn, A.H. Conformational Stability of Fibrillar Amyloid-Beta

Oligomers via Protofilament Pair Formation - A Systematic Computational Study. PloS One

8, e70521 (2013).

13. Horn, A.H. & Sticht, H. Amyloid-β42 oligomer structures from fibrils: a systematic molecular

dynamics study. J. Phys. Chem. B 114, 2219-2226 (2010).

14. Paravastu, A.K., Petkova, A.T. & Tycko, R. Polymorphic fibril formation by residues 10-40 of

the Alzheimer's β-amyloid peptide. Biophys. J. 90, 4618-4629 (2006).

15. Takeda, T. & Klimov, D.K. Probing the effect of amino-terminal truncation for Aβ1-40 peptides.

J. Phys. Chem. B 113, 6692-6702 (2009).

62

16. Case, D. A. et al. AMBER 12. (http://www.ambermd.org) (University of California, San

Francisco, 2012)

17. Wang, J.M., Wang, W., Kollman, P.A. & Case, D.A. Automatic atom type and bond type

perception in molecular mechanical calculations. J. Mol. Graph. Model. 25, 247-260 (2006).

18. Trott, O. & Olson, A.J. AutoDock Vina: Improving the speed and accuracy of docking with a

new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31, 455-

461 (2010).

19. Morris, G.M. et al. AutoDock4 and AutoDockTools4: Automated Docking with selective

receptor flexibility. J. Comput. Chem. 30, 2785-2791 (2009).

20. Sanner, M.F. Python: A programming language for software integration and development. J.

Mol. Graph. Model. 17, 57-61 (1999).

21. Hornak, V. et al. Comparison of multiple Amber force fields and development of improved

protein backbone parameters. Proteins 65, 712-725 (2006).

22. Jorgensen, W.L., Chandrasekhar, J., Madura, J.D., Impey, R.W. & Klein, M.L. Comparison

of simple potential functions for simulating liquid water. J. Chem. Phys. 79, 926-935 (1983).

23. Darden, T., York, D. & Pedersen, L. Particle mesh Ewald: an N∙Log(N) method for Ewald

sums in large systems. J. Chem. Phys. 98, 10089-10092 (1993).

24. Salomon-Ferrer, R., Götz, A.W., Poole, D., Le Grand, S. & Walker, R.C. Routine Microsecond

Molecular Dynamics Simulations with AMBER on GPUs. 2. Explicit Solvent Particle Mesh

Ewald. J. Chem. Theory Comput. 9, 3878-3888 (2013).

25. Izaguirre, J.A., Catarello, D.P., Wozniak, J.M. & Skeel, R.D. Langevin stabilization of

molecular dynamics. J. Chem. Phys. 114, 2090-2098 (2001).

26. Ryckaert, J.P., Ciccotti, G. & Berendsen, H.J.C. Numerical integration of the cartesian

equations of motion of a system with constraints: molecular dynamics of n-alkanes. J.

Comp. Phys. 23, 327-341 (1977).

27. Götz, A.W. et al. Routine Microsecond Molecular Dynamics Simulations with AMBER on

GPUs. 1. Generalized Born. J. Chem. Theory Comput. 8, 1542-1555 (2012).

28. Le Grand, S., Götz, A.W. & Walker, R.C. SPFP: Speed without compromise-A mixed

precision model for GPU accelerated molecular dynamics simulations. Comput. Phys.

Commun. 184, 374-380 (2013).

29. Daura, X. et al. Peptide Folding: When Simulation Meets Experiment. Angew. Chem. Int. Ed.

Engl. 38, 236-240 (1999).

http://www.ambermd.org/

63

30. Hess, B., Kutzner, C., van der Spoel, D. & Lindahl, E. GROMACS 4: Algorithms for highly

efficient, load-balanced, and scalable molecular simulation. J. Chem. Theory Comput. 4, 435-

447 (2008).

31. Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 14,

33-38 (1996).

32. Hou, T.J., Wang, J.M., Li, Y.Y. & Wang, W. Assessing the Performance of the MM/PBSA and

MM/GBSA Methods. 1. The Accuracy of Binding Free Energy Calculations Based on

Molecular Dynamics Simulations. J. Chem. Inf. Model. 51, 69-82 (2011).

33. Kollman, P.A. et al. Calculating Structures and Free Energies of Complex Molecules:

Combining Molecular Mechanics and Continuum Models. Acc. Chem. Res. 33, 889-897

(2000).

34. Gohlke, H., Kiel, C. & Case, D.A. Insights into Protein–Protein Binding by Binding Free

Energy Calculation and Free Energy Decomposition for the Ras–Raf and Ras–RalGDS

Complexes. J. Mol. Biol. 330, 891-913 (2003).

35. Leonis, G., Steinbrecher, T. & Papadopoulos, M.G. A Contribution to the drug resistance

mechanism of Darunavir, Amprenavir, Indinavir, and Saquinavir complexes with HIV-1

protease due to flap mutation I50V: A systematic MM–PBSA and thermodynamic integration

study. J. Chem. Inf. Model 53, 2141-2153 (2013).

36. Wright, D.W., Hall, B.A., Kenway, O.A., Jha, S. & Coveney, P.V. computing clinically relevant

binding free energies of HIV-1 protease inhibitors. J. Chem. Theory Comput. 10, 1228-1241

(2014).

37. Miller, B.R. et al. MMPBSA.py: An efficient program for end-state free energy calculations. J.

Chem. Theory Comput. 8, 3314-3321 (2012).

38. Bessette, P.H., Åslund, F., Beckwith, J. & Georgiou, G. Efficient folding of proteins with

multiple disulfide bonds in the Escherichia coli cytoplasm. Proc. Natl. Acad. Sci. U.S.A. 96,

13703-13708 (1999).

39. Valentine, J.S., Doucette, P.A. & Zittin Potter, S. Copper-zinc superoxide dismutase and

amyotrophic lateral sclerosis. Annu. Rev. Biochem. 74, 563-593 (2005).

40. Banerjee, V. et al. Superoxide Dismutase 1 (SOD1)-derived peptide inhibits amyloid

aggregation of familial amyotrophic lateral sclerosis SOD1 mutants. ACS Chem. Neurosci. 7,

1595-1606 (2016).

41. Bradford, M.M. A rapid and sensitive method for the quantitation of microgram quantities of

protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254 (1976).

64

42. Kuo, W., Huang, C., Shih, C. & Jinn, T. Cellular extract preparation for superoxide dismutase

(SOD) activity assay. Bio Protoc. 3, e811 (2013).

43. Zarafeta, D. et al. Metagenomic mining for thermostable esterolytic enzymes uncovers a new

family of bacterial esterases. Sci. Rep. 6 (2016).

44. Villar, E.A. et al. How proteins bind macrocycles. Nat. Chem. Biol. 10, 723-731 (2014).

an integrated bacterial system for the discovery of ...10.1038/s41551-017-0144... · supplementary...

Documents