Field of the Invention The invention relates to peptide beads comprising amphiphilic bead-forming peptides, a mediator peptide, and a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

Background Art

Future advanced materials will no longer rely merely on the chemical structure of a molecule per se but also on supramolecular chemistry, focusing on properties that emerge and are enabled by the assembly of a distinct number of molecular subunits. By exploiting non-covalent interactions, these systems come to possess new properties that reach beyond those of single molecules in bulk. The construction of advanced materials should exploit a bottom-up strategy that relies on self-assembly to allow large scale fabrication. Nature itself makes use of such systems, with enzymes and other proteins having well- defined secondary (a-helix, β-sheet) and quaternary structures (3D-assembly of individual subunits); each structure exhibiting distinct properties and thereby the potential for distinct functioning. Many of these possible functions may be lost without the well-controlled folding and assembly of a proteins' primary structure (sequence) into a precise secondary and quaternary structure. Small changes in the primary structure alone can perturb the assembly of the protein, along with its function. The precise folding of a protein or a small peptide from a linear sequence to a complex machine (e.g., an enzyme) may be considered self-assembly.

Based on virtually unlimited synthetic control over primary structure and ease of implementing point mutations, peptides represent fruitful molecules for the study of how self-assembly is controlled by the chemical structure of individual building blocks.

Furthermore, peptidic materials are preferred for applications in the human body, due to a high level of biocompatibility and established degradation pathways that eliminate unwanted accumulation.

Hybrid materials have been developed to combine the functionality of peptides with the versatility of hydrophobic/hydrophilic triggering of self-assembly. Based on this concept, several groups have built peptide-based supramolecular assemblies by taking a non- peptidic hydrophobic contributor such as alkyl chains, poly(butadiene) or modified amino acids (AAs) and attaching a peptidic hydrophilic part. Although such systems form supramolecular functional materials, the polymeric character rules out the possibility of point mutations and therefore the ability for fine tuning. Short peptidic amphiphiles overcome this limitation.

C. Dittrich and W. Meier, Macromolecular Bioscience 2010, 10, 1406-1415 recently reported the synthesis, purification and characterization of purely peptidic amphiphiles and provided insights into the process of self-assembly of the peptides into micelles and solid, spherical particles, termed "peptide beads" (T. B. Schuster, D. de Bruyn Ouboter, E.

Bordignon, G. Jeschke and W. Meier, Soft Matter 2M , 6, 5596-5604). The hydrophobic block is inspired by the sequence and secondary structure of gramicidin A (gA), which is consistently constructed of hydrophobic amino acids that hide its hydrophilic backbone inside a β-helical secondary structure. Moreover, due to its highly hydrophobic nature and its compatibility with the hydrophobic core of phospholipid membranes, gA is a suitable hydrophobic constituent.

Gramicidin A is a pentadecapeptide, consisting of alternating and completely hydrophobic D- and L-amino acids. The sequence of 15 amino acids is N-terminally modified with a formyl residue and C-terminally functionalized with an ethanolamine residue: formyl-L-Val-Gly-L-Ala-D-Leu-L-Ala-D-Val-L-Val-D-Val-L-Trp-D -Leu-L-Trp-D-Leu-L-Trp-D-

Leu-L-Trp-ethanolamine Gramicidin is known for its antibiotic effect on gram positive bacteria, caused by its ion channel formation for monovalent cations. Ion channel formation is enabled by a β-helical- like secondary structure, causing two molecules to dimerize (B. A. Wallace, Advances in experimental medicine and biology 1996, 398, 607-614). Novel medications will require advanced, functional delivery systems to allow efficient drug delivery, protect sensitive ingredients, reduce side-effects, and increased time in circulation. Hydrophilic drugs often possess good solubility, but are quickly degraded or eliminated from the body. Hydrophobic drugs, on the other hand, lack solubility and accumulate in hydrophobic organelles. Several diseases are known to require multi-drug treatment, leading to the necessity of delivering and releasing the hydrophobic as well as the hydrophilic drugs simultaneously to a specific target. One example of such a need is cancer treatment in a combination therapy using hydrophilic doxorubicin and hydrophobic paclitaxel.

Another important task of nano-sized delivery systems is protecting drug molecules from the environment, where they are exposed to degradation processes. Gene therapy has been promoted as having outstanding potential to cure diseases on a genetic level. The use of siRNA - small interfering RNAs composed of only 15 to 25 base pairs - can temporarily stop, i.e. silence the transcription of specific genes into their peptides/proteins (D. M. Dykxhoorn, C. D. Novina and P. A. Sharp, Nature reviews. Molecular cell biology 2003, 4, 457-467). The technology virtually enables the treatment of almost every disease that has a basis established in gene hyperactivity, whereby the full decryption of the human genome becomes of great advantage. The use of plasmids - carrying the sequences for small hairpin RNA (shRNA) that are cleaved by the cellular machinery into siRNA - can permanently silence a gene (F. Xie, M. Woodle and P. Lu, Drug Discovery Today 2006, 11, 67-73). In contrast to siRNA, a plasmid reproduces during mitosis and therefore represents a very powerful method for gene therapy. Both siRNA and the DNA plasmids are prone to degradation before reaching their destination in such an application, which seems logical, because the body naturally protects itself from foreign genetic information. This fact, however, corroborates the need for a surrounding, protective layer to enable successful delivery into cells. D. J. Gary, N. Puri and Y. Y. Won, Journal of controlled release: Official journal of the Controlled Release Society 2007, 121, 64-73 provide an overview of polymer-based siRNA delivery in a recent review.

Besides their main task of encapsulation and delivery of active components, the materials used ideally exhibit biocompatibility and biodegradability. These are very important properties and their absence has caused several potentially applicable systems to fail. Amino acid based materials are generally considered biocompatible and biodegradable, due to the fact that they are an intrinsic component of the human body and they have established degradation pathways.

IFNy is an immunomodulatory cytokine produced by T lymphocytes and natural killer cells. In the central nervous system, IFNy which is increased in chronic inflammatory disease or following injury, can activate astrocytes and microglial cells. Targeting IFNy addresses the prime question of the role of inflammation in neurological diseases and has therefore a notable therapeutic potential for diseases involving IFNy-mediated neuroinflammation, such as Amyotrophic Lateral Sclerosis (ALS), Multiple Sclerosis (MS) and Parkinson's Disease (PD).

Summary of the Invention

The invention relates to peptide beads comprising amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ²

wherein

R ¹ is cysteine, histidine, lysine, acetyl-lysine, glutamic acid, homocysteine, thiocitrulline, S- methyl-thiocitrulline, biotin, 6-hydrazino-nicotinamide, 4-formyl-benzamide, tert-butyloxy- carbonyl, 9-fluorenylmethyloxycarbonyl, formyl, acetyl, propionyl, or hydrogen;

R ² is NH ₂ , OH, OCH ₃ , OCH2CH ₃ , N HCH2CH2OH, cysteine, cysteine-amide, histidine, histidine-amide, lysine, lysine-amide, acetyl-lysine, acetyl-lysine-amide, glutamic acid, glutamic acid amide, homocysteine, homocysteine-amide, thiocitrulline, thiocitrulline- amide, S-methyl-thiocitrulline, S-methyl-thiocitrulline-amide, or biotin;

(A ¹ ) _k and (A ¹ )i are sequences composed of hydrophilic L-amino acids A ¹ , wherein

A ¹ is a hydrophilic L-amino acid selected from lysine, acetyl-lysine, cysteine, histidine, glycine, glutamic acid, aspartic acid, serine, oarmino-glycine, acetyl-oamino-glycine, α,γ- diaminobutyric acid, acetyl-a,Y-diaminobutyric acid, α,β-diaminopropionic acid, acetyl-α,β- diaminopropionic acid, ornithine, and acetyl-ornithine; and wherein A ¹ may be the same or different in a sequence (A ¹ ) _k or in (A ¹ )i or in both;

A ² is an aromatic L-amino acid selected from tryptophane, phenylalanine, tyrosine, 3- hydroxymethyl-tyrosine, 3-(3,4-dihydroxy-phenyl)-serine, 4-carboxy-phenylalanine, β-(2- thienyl)-alanine, phenylglycine, omethyl-phenylalanine, homophenylalanine, 4-azido- phenylalanine, 4-cyano-phenylalanine, 3,5-dinitro-tyrosine, 3,5-dibromo-tyrosine, 4- carboxy-phenylalanine, 1 ,2,3,4-tetrahydroisoquinoline-3-carboxylic acid, 3-(3-benzo- thienyl)-alanine, 4-nitro-phenylalanine, 4-bromo-phenylalanine, 4-tert-butyl-phenylalanine, a-methyl-tryptophan, 3-iodo-tyrosine, 3-nitro-tyrosine, 3,5-diiodo-tyrosine, 3-(1 -naphthyl)- alanine, 4-iodo-phenylalanine, 3-fluoro-phenylalanine, 4-fluoro-phenylalanine, 4-methyl- tryptophan, 5-methyl-tryptophan, 5-hydroxy-tryptophan, 3-(2-naphthyl)-alanine, 4-chloro- phenylalanine, 3,4-dichloro-phenylalanine, 2,6-difluoro-phenylalanine, N-methyl- tryptophan, 4-methyl-phenylalanine, 4-phenyl-phenylalanine, 3-(2-quinolyl)-alanine, 3- amino-tyrosine, 4-amino-phenylalanine, 3-(2-pyridyl)-alanine, and 3-(3-pyridyl)-alanine; and wherein A ² may be the same or different in (A ² - D-A ³ ) _m and (A ² ) _n ; A ³ is an aliphatic D-amino acid selected from D-alanine, D-valine, D-leucine, D-isoleucine, D-proline, dehydro-D-alanine, β-fluoro-D-alanine, β-chloro-D-alanine, β-iodo-D-alanine, D- oaminobutyric acid, β-cyano-D-alanine, β-ureido-D-alanine, D-2-allyl-glycine, tert-butyl-D- alanine, β-(1 -cyclopentenyl)-D-alanine, D-norvaline, omethyl-D-valine, 4,5-dehydro-D- leucine, allo-D-isoleucine, D-norleucine, and omethyl-D-leucine;

k and I are, independently of each other, between 0 and 12, with the proviso that at least one of k and I is different from 0;

m is between 2 and 8; and

n is 0 or 1 ; a mediator peptide of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein A ¹ , A ² , D-A ³ , R ¹ , R ² , k, I, m and n are defined as for the amphiphilic bead-forming peptide, but at least one of the substituents R ¹ , R ² and/or at least one of the amino acids A ¹ is different from the corresponding substituents R ¹ , R ² and/or amino acids A ¹ in the amphiphilic bead- forming peptide; and a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

The invention likewise relates to peptide beads comprising a mediator peptide and a hydrophilic active pharmaceutical ingredient for the delivery of this active pharmaceutical ingredient into tissue and cells, and the release of the hydrophilic active pharmaceutical ingredient in the tissue or cell causing the desired therapeutic effect.

Brief Description of the Figures

Fig. 1 : Static light scattering (SLS) analysis of Ac-X ₃ -gT peptide beads in water

(0.25 mg/ml)

(A) Guinier Plot resulting in an R _g of 236 ± 7 nm

(B) form factor analysis (-) hard sphere model for Rs = 304.8 nm, ( ) coil model for Rg = 236.1 nm, (■) experimental data

Fig. 2: Scanning electron microscopy (SEM) and transmission electron micrographs (TEM) of Ac-X ₃ -gT peptide beads

(A) SEM of dried sample 0.5 mg/ml; (B) SEM of shock-frozen and lyophilized sample 0.5 mg/ml; (C) TEM 0.25 mg/ml; (D) TEM with uranyl acetate staining 0.25 mg/ml. Fig. 3: Surface of Ac-X ₃ -gT peptide beads in atomic force microscope (AFM),

phase contrast (A) showing structures with the same size as K ₃ -gT micelles on mica in phase contrast (B) and in topographic contrast (C). Peptide beads in AFM topographic contrast (D) shown for illustration.

Fig. 4: Scanning electron microscopy (SEM) of self-assembled peptide beads from Ac-X ₃ -gT-C (a); insights into a peptide bead revealing a structure constructed by smaller spherical subunits (b). Fig. 5: Small angle x-ray scattering (SAXS) of peptide beads showing the sizes of substructures inside the beads. Experimental data (-), Unified fit model ( ). Two-level Beaucage method (A) and PDDF with scattering intensity in inset (B).

Fig. 6: (A) Fitted small angle x-ray scattering (SAXS) curves of Ac-X ₃ -gT using a TNNLS model (grey: experimental data, black: calculated from the size distribution); (B) volume weighted size distribution (n = frequency, 0 = particle diameter); C Number weighted size distribution (n = frequency, 0 = particle diameter.

Fig. 7: Hierarchical self-assembly into peptide beads (PB): aggregated micelle model (left) and large compound micelle model (right). With micelle (M), aggregated micelles (AM), inverted micelle (IM), and large compound micelle (LCM).

Fig. 8: Multichannel confocal laser scanning microscopy (CLSM) of hepatocytes (scale bars: 10 μηι).

(A) Internalization of BODIPY633 loaded peptide beads (shown as light dots, upper right and lower right panel) into hepatocytes visualized in transmission channel (lower panels). (B) Internalization of cy3-siRNA loaded peptide beads (shown as light dots) into hepatocytes. Nucleus (grey ovals) and plasma membrane (thin, grey area) fluorescently stained for visualisation.

Fig. 9: Comparison of cell viability (by MTS absorption assay at 490 nm) between THP-1 macrophages (A) and THP-1 monocytes (B) after 3-day treatment: control experiment (Θ), peptide beads only (PB), peptide beads loaded with the anti-cancer drugs paclitaxel (tax) and doxorubicin (dox). Different dilutions shown. Plots visualize the lack of toxic effect by the pure peptide beads as seen form the "PB - Θ" samples (A+B), and the desired effect of the drugs in monocytes (B). Fig. 10: Quantitative Real-Time-PCR results of gene silencing in HuH-7 hepatocytes. The comparison between control experiments(O), peptide beads only (PB - Θ), and negative control siRNA (si-θ) / siRNA silencing GAPDH (si-GAPDH) transfection using

conventional transfection agent Lipofectamine 2000 (Lipo) and the peptide beads (PB) shown in A. Similar with negative control plasmid (sc-PP2Ac) and plasmid silencing PP2Ac (sh-PP2Ac) shown in B and C. Results show the lack of toxicity by the delivery system as well as the desired gene silencing. The effect was significantly increased by cell stability selection of the plasmid transfected cells (using peptide beads (PB) proving gene delivery effective, shown in C1 , C2).

Detailed Description of the Invention

The invention relates to the use of a mediator peptide to allow anchoring, embedding, and/or post-formation-modification in or on the peptide beads, using the hydrophobic part of the mediator peptide sequence as an anchor and an additional modified part as connector to the payload, i.e. the guest molecule. The mediator peptide may be tailored for particular interactions to the RNA, DNA, or hydrophilic active pharmaceutical ingredient.

An example of such a mediator peptide is the K ₃ -gT peptide, which may act as a mediator to complex, bind, condense, and/or embed negatively charged payloads (e.g.

nucleotides). Mixing the mediator with the payload and the bead forming peptide (e.g. peptide Ac-X ₃ -gT) then leads to effective incorporation of the payload into peptide beads.

This key aspect of the invention has not been shown in the prior art, and now for the first time allows efficient incorporation of RNA, such as siRNA, shRNA and miRNA, and of DNA, such as pieces of single stranded and double stranded DNA of variable length or of plasmid DNA and vectors, into peptide beads for efficient transfer into cells. Likewise the mediator peptide may be used to incorporate a hydrophilic drug (active pharmaceutical ingredient) other than DNA or RNA, such as doxorubicin. The beads may further contain hydrophobic drugs, such as for example paclitaxel, nutrition factors and/or nutrition additives.

The invention further relates to the directed use of modifications of the multicompartment micellar structure of the peptide beads to design and produce specific or unspecific payload embedding for therapeutic purposes. In particular the invention is directed to the embedding of active pharmaceutical ingredients, nutrition factors, nutrition additives and the like into the peptide beads by addition of a mediator supporting incorporation of this particular payload into the peptide beads.

The invention likewise relates to peptide beads comprising a mediator peptide and a hydrophilic active pharmaceutical ingredient for the delivery of this active pharmaceutical ingredient into tissue and cells, and the release of the hydrophilic active pharmaceutical ingredient in the tissue or cell causing the desired therapeutic effect. The payload (also termed cargo) is embedded within the peptide bead, and is then transported to tissues or cells where the payload is released either by diffusing out of the peptide bead, by digestion or other destabilization of the peptide bead

The invention relates to peptide beads comprising amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ²

wherein

R ¹ is cysteine, histidine, lysine, acetyl-lysine, glutamic acid, homocysteine, thiocitrulline, S- methyl-thiocitrulline, biotin, 6-hydrazino-nicotinamide, 4-formyl-benzamide, tert-butyloxy- carbonyl, 9-fluorenylmethyloxycarbonyl, formyl, acetyl, propionyl, or hydrogen;

R ² is NH ₂ , OH, OCH ₃ , OCH ₂ CH ₃ , NHCH ₂ CH ₂ OH, cysteine, cysteine-amide, histidine, histidine-amide, lysine, lysine-amide, acetyl-lysine, acetyl-lysine-amide, glutamic acid, glutamic acid amide, homocysteine, homocysteine-amide, thiocitrulline, thiocitrulline- amide, S-methyl-thiocitrulline, S-methyl-thiocitrulline-amide, or biotin;

(A ¹ ) _k and (A ¹ )i are sequences composed of hydrophilic L-amino acids A ¹ , wherein

A ¹ is a hydrophilic L-amino acid selected from lysine, acetyl-lysine, cysteine, histidine, glycine, glutamic acid, aspartic acid, serine, oarmino-glycine, acetyl-oamino-glycine, α,γ- diaminobutyric acid, acetyl-a,Y-diaminobutyric acid, α,β-diaminopropionic acid, acetyl-α,β- diaminopropionic acid, ornithine, and acetyl-ornithine; and wherein A ¹ may be the same or different in a sequence (A ¹ ) _k or in (A ¹ )i or in both;

A ² is an aromatic L-amino acid selected from tryptophane, phenylalanine, tyrosine, 3- hydroxymethyl-tyrosine, 3-(3,4-dihydroxy-phenyl)-serine, 4-carboxy-phenylalanine, β-(2- thienyl)-alanine, phenylglycine, omethyl-phenylalanine, homophenylalanine, 4-azido- phenylalanine, 4-cyano-phenylalanine, 3,5-dinitro-tyrosine, 3,5-dibromo-tyrosine, 4- carboxy-phenylalanine, 1 ,2,3,4-tetrahydroisoquinoline-3-carboxylic acid, β-(3- benzothienyl)-alanine, 4-nitro-phenylalanine, 4-bromo-phenylalanine, 4-tert-butyl- phenylalanine, a-methyl-tryptophan, 3-iodo-tyrosine, 3-nitro-tyrosine, 3,5-diiodo-tyrosine, 3-(1 -naphthyl)-alanine, 4-iodo-phenylalanine, 3-fluoro-phenylalanine, 4-fluoro- phenylalanine, 4-methyl-tryptophan, 5-methyl-tryptophan, 5-hydroxy-tryptophan, β-(2- naphthyl)-alanine, 4-chloro-phenylalanine, 3,4-dichloro-phenylalanine, 2,6-difluoro- phenylalanine, N-methyl-tryptophan, 4-methyl-phenylalanine, 4-phenyl-phenylalanine, β- (2-quinolyl)-alanine, 3-amino-tyrosine, 4-amino-phenylalanine, 3-(2-pyridyl)-alanine, and 3-(3-pyridyl)-alanine; and wherein A ² may be the same or different in (A ² - D-A ³ ) _m and (A ² ) _n ;

A ³ is an aliphatic D-amino acid selected from D-alanine, D-valine, D-leucine, D-isoleucine,

D-proline, dehydro-D-alanine, β-fluoro-D-alanine, β-chloro-D-alanine, β-iodo-D-alanine, D- oaminobutyric acid, β-cyano-D-alanine, β-ureido-D-alanine, D-2-allyl-glycine, tert-butyl-D- alanine, β-(1 -cyclopentenyl)-D-alanine, D-norvaline, a-methyl-D-valine, 4,5-dehydro-D- leucine, allo-D-isoleucine, D-norleucine, and omethyl-D-leucine;

k and I are, independently of each other, between 0 and 12, with the proviso that at least one of k and I is different from 0;

m is between 2 and 8; and

n is 0 or 1 .

(A ² - D-A ³ ) _m - A ² comprises the gramicidin inspired hydrophobic sequence. One characteristic attribute of the gramicidin family is the unique secondary structure. The gramicidin family is based on the primary sequence: LV-G-LA-DL-LA-DV-LV-DV-LW-DL-LX- DL-LW-DL-LW while X at position 1 1 can be tryptophan (LW), phenylalanine (LF) or tyrosine (LY). All these isomers form an ion conducting channel, i.e. the secondary structure is not significantly altered by the selection of the amino acid in position 1 1 .

Therefore the properties of the gramicidin family are rather based on their secondary structure than on their primary sequence. However the alternating D - L configuration of the amino acids is obviously crucial for the secondary structure. Consequently variation in the primary sequence as described for position 1 1 can be transferred to position 9, 13, and 15, and the variation can be extended to the mentioned non-natural occurring derivates of the corresponding amino acids. Similarly, conservation of the secondary structures allows the replacement of Dl_ at position 10, 12 and 14 by other aliphatic amino acids, such as proline (DP), alanine (DA), valine (DV), and isoleucine (Dl) and non-natural occurring derivates of the corresponding amino acids.

In particular the invention relates to peptide beads comprising

amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² as defined hereinbefore; a mediator peptide of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein at least one of the substituents R ¹ , R ² and/or at least one of the amino acids A ¹ is different from the corresponding substituents R ¹ , R ² and/or amino acids A ¹ in the amphiphilic bead- forming peptide; and

a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

The relative amount of mediator peptide to the amount of amphiphilic bead-forming peptides is preferably between 1 to 1000 and 1 to 5, preferably between 1 to 100 and 1 to 10, such as around 1 to 50 or around 1 to 30.

More particularly the invention relates to such peptide beads wherein the mediator peptide preferentially interacts with the payload. Interaction may be by chemical binding or preferably, by hydrogen binding or electrostatic interaction. A particular example is a peptide bead wherein the mediator peptide carries amino functions NH ₂ that can protonate and are then able to interact with payloads selected from RNA, DNA and a hydrophilic drug.

More specifically the invention relates to a peptide bead comprising

amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² wherein

R ¹ is cysteine, histidine, lysine, acetyl-lysine, glutamic acid, homocysteine, thiocitrulline, S- methyl-thiocitrulline, biotin, 6-hydrazino-nicotinamide, 4-formyl- benzamide, tert-butyloxy- carbonyl, 9-fluorenylmethyloxycarbonyl, formyl, acetyl, propionyl, or hydrogen;

R ² is NH ₂ , OH, OCH ₃ , OCH ₂ CH ₃ , N HCH2CH2OH, cysteine, cysteine-amide, histidine, histidine-amide, lysine, lysine-amide, acetyl-lysine, acetyl-lysine-amide, glutamic acid, glutamic acid amide, homocysteine, homocysteine-amide, thiocitrulline, thiocitrulline- amide, S-methyl-thiocitrulline, S-methyl-thiocitrulline-amide, or biotin;

(A ¹ ) _k and (A ¹ )i are sequences composed of hydrophilic L-amino acids A ¹ , wherein

A ¹ is a hydrophilic L-amino acid selected from lysine and acetyl-lysine; and wherein A ¹ may be the same or different in a sequence (A ¹ ) _k or in (A ¹ )i or in both;

A ² is an aromatic L-amino acid selected from tryptophane, phenylalanine, and tyrosine; and wherein A ² may be the same or different in (A ² - D-A ³ ) _m and (A ² ) _n ;

A ³ is an aliphatic D-amino acid selected from D-alanine, D-valine, D-leucine, D-isoleucine, D-norvaline, and D-norleucine; k and I are, independently of each other, between 0 and 12, with the proviso that at least one of k and I is different from 0;

m is between 2 and 8; and

n is 0 or 1 ;

a mediator peptide of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein at least one of the substituents R ¹ , R ² and/or at least one of the amino acids A ¹ is different from the corresponding substituents R ¹ , R ² and/or amino acids A ¹ in the amphiphilic bead- forming peptide, and which carries an amino function NH ₂ ; and

a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

Even more specifically the invention relates to a peptide bead comprising

amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² wherein

R ¹ is cysteine, lysine, acetyl-lysine, biotin, acetyl, or hydrogen;

R ² is NH ₂ , OH, cysteine, lysine, acetyl-lysine, or biotin;

(A ¹ ) _k and (A ¹ )i are sequences composed of hydrophilic L-amino acids A ¹ , wherein

A ¹ is a hydrophilic L-amino acid selected from lysine and acetyl-lysine; and wherein A ¹ may be the same or different in a sequence (A ¹ ) _k or in (A ¹ )i or in both;

A ² is tryptophane;

A ³ is an aliphatic D-amino acid selected from D-alanine, D-valine, D-leucine, D-isoleucine, k is between 1 and 5

I is 0 or 1 ;

m is between 2 and 4; and

n is 0 or 1 ;

a mediator peptide of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein at least one of the substituents R ¹ , R ² and/or at least one of the amino acids A ¹ is different from the corresponding substituents R ¹ , R ² and/or amino acids A ¹ in the amphiphilic bead- forming peptide and carries an amino function NH ₂ ; and

a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

Particularly preferred is payload selected from siRNA, shRNA, miRNA, and plasmid DNA.

Most preferred is a peptide bead comprising

amphiphilic bead-forming peptides of the formula

R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein

R ¹ is acetyl; R ² is NH ₂ ; A ¹ is acetyl-lysine; A ² is tryptophan; A ³ is D-leucine;

k is 3; I is 0; m is 3; and n is 1 ;

a mediator peptide of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein R ¹ is hydrogen; R ² is NH ₂ ;

A ¹ is lysine; A ² is tryptophan; A ³ is D-leucine;

k is 3; I is 0; m is 3; and n is 1 ; and

a payload selected from RNA, DNA and a hydrophilic active pharmaceutical ingredient.

The invention also relates to the particular peptide beads comprising amphiphilic bead- forming peptides of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ²

described in the Examples (Table 3).

Preferred are such particular peptide beads comprising a multitude of amphiphilic bead- forming peptides of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein R ¹ is acetyl; R ² is NH ₂ ;

A ¹ is acetyl-lysine; A ² is tryptophan; A ³ is D-leucine;

k is 2, 4 or 5; I is 0; m is 3; and n is 1 .

Likewise preferred are such particular peptide beads comprising amphiphilic bead-forming peptides of the formula R ¹ - (A ¹ ) _k - (A ² - D-A ³ ) _m - (A ² ) _n - (A ¹ ), - R ² , wherein

R ¹ is hydrogen; R ² is NH ₂ ;

A ¹ is acetyl-lysine; A ² is tryptophan; A ³ is D-leucine;

k is 2, 3, 4, and 5, in particular 3; I is 0; m is 3; and n is 1 . Hydrophilic active pharmaceutical ingredients considered are doxorubicin, topotecan, and irinotecan.

Particular RNA or DNA considered are siRNA, shRNA, miRNA, plasmid DNA, and vector DNA. Examples of such RNA or DNA are: GAPDH siRNA, silencing the transcription of the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH);

psiSTRIKE-Neomycin vector (containing a hairpin target sequence of PP2Aca that allows expression of siRNA against PP2Aca and a sequence coding for antibiotic resistance); and a DNA sequence encoding the extracellular portion of IFNyRI fused to a coiled-coil domain of cartilage oligomeric matrix protein (COMP); and also corresponding negative control siRNA and negative control vector DNA. Peptide beads engineered to incorporate active pharmaceutical ingredients, such as therapeutic anti-neuroinflammatory payloads, present several advantages. Their surface can be functionalized to deliver the therapeutic instruction to specific tissues, e.g. those affected in neurological disorders or those to become a therapeutic source of conversion of diffusible proteins. Furthermore, these high-load nanostructure carriers may allow more efficient gene delivery, leading to high levels of expression. In addition, the use of a non- viral vector system has the potential of being much safer than viral-based systems, with reduced immunogenicity and cytotoxicity along with much greater biodegradability. A delivery system that does not trigger any inflammatory reaction is of utmost importance in the present application, which specifically aims at neutralizing pro-inflammatory reactions.

In a particular embodiment the designed approach endeavors to block the interaction of IFNy with IFNy receptor 1 (IFNyRI) and/or IFNy receptor 2 (IFNyRII), to antagonize the biological activity of IFNy and/or of IFNy receptors and/or of IFNy variants or fragments thereof, and to reach the broadest number of neurons in the affected areas.

For this purpose, peptide beads may incorporate and deliver a payload comprising a nucleic acid sequence coding for the extracellular portion of IFNyRI (SEQ ID NO:9, nucleic acid) coding for SEQ ID NO:10 (amino acid) for mouse, SEQ ID NO:1 1 (nucleic acid) coding for SEQ ID NO:12 (amino acid) for human, or a variant thereof being at least 80% identical to SEQ ID NO:10 or SEQ ID NO:12, fused to an oligomerization domain, such as a coiled-coil domain of cartilage oligomeric matrix protein (COMP), a leucine zipper, a trimerization domain of the NCI domain of collagen XVIII or collagen XV, the Fc fragment of an immunoglobulin, or the C-terminal dimerization domain of the TNF family member osteoprotegerin, under the control of a ubiquitous, a neuronal- or a glial-specific promoter, including a phosphoglycerokinase (PGK1 ) (SEQ ID NO:13), GfaABC1 -D (SEQ ID NO:14), cytomegalovirus promoter (SEQ ID NO:15), or chimeric CMV-chicken β-actin (CBA) promoter (SEQ ID NO:16), allowing efficient and sustained expression. The peptide beads comprising DNA encoding an IFNy antagonist as a payload may be administered in any manner including, but not limited to, intravenous, intra-arterial, intraperitoneal, subcutaneous, intramuscular, intracerebroventricular, intraparenchymal, intracisternal, intrathecal, intranasal, intravitreal, transcleral, epidural, or oral

administration. They may also be administered in the form of an implant, allowing the controlled release of the compositions. The peptide beads comprising DNA encoding an IFNy antagonist as a payload can be administered alone or in combination with a co-agent useful in the treatment of a neurological disorder, e.g. riluzole, dexpramipexole or AMPA receptor antagonists for ALS; levodopa (L-DOPA), dopa decarboxylase inhibitor, COMT inhibitor, dopamine agonists, or MAO-B inhibitors for PD; or corticosteroids, interferon beta-1 a, interferon beta-1 b, glatiramer acetate, non-steroidal immunomodulators, or lymphocyte-targeting antibodies for MS.

The surface of the peptide beads comprising DNA encoding an IFNy antagonist as a payload may or may not be functionalized by incorporating further compounds useful to facilitate brain and spinal cord targeting, in particular peptides, proteins, immunoglobulins, or immunoglobulin fragments, such as, but not restricted to SynB1 , Brain Homing peptides, anti-transferrin receptor antibodies, transferrin, angiopep-2, lactoferrin, thiamine, penetratin antennapedia peptide, poloxamer, or polysorbate 80. Furthermore, the surface of the peptide beads may or may not be functionalized to target specific cellular populations, such as astrocytes, to be converted into platforms to secrete, on the long term, the therapeutic IFNy antagonists and minimizing peripheral transgene expression. Such functionalization may, for example, involve antibodies or antibody fragments targeting Glutamate Aspartate Transporter (GLAST1 ), Monocarboxylate transporter 1 , or Aquaporine 4. Using a highly astrocyte-specific promoter will, in addition, specifically drive transgene expression in astrocytes.

Finally, the surface of the beads may or may not be modified with suitable compounds in order to decrease degradation and immunogenicity and improve the circulation time as well. Compounds useful for that purpose are sialic or glucuronic acid, proteins such as human serum albumin and IgA, glycolipids, such as GM1 and GM type III glycolipids, hydrophilic polymers such as polyethylene glycol, polyvinyl pyrrolidone), poly(acryl amide), poly(2-methyl-2-oxazoline), poly(2-ethyl-2-oxazoline), poly[N-(2-hydroxypropyl)- methacrylamide, and polyvinyl alcohol, or D- and L-amino acid based polymer, such as poly(hydroxyethyl asparagine) and poly(hydroxyethyl glutamine). Preferred compound for surface modification is polyethylene glycol.

It is known that the peptide K ₃ -gT formed by N-terminal addition of three lysines to a truncated amino acid sequence of gA (gramicidin A), which corresponds to the last seven amino acids of gA self-assembles into spherical micelles in aqueous solution, while the self-assembly of the fully acetylated analogue Ac-X ₃ -gT starts with the amphiphilic peptide in a dissolved monomeric state in pure ethanol. Here, increasing water concentration induces the formation process of small micelles (R _h ~ 10 nm). During further solvent exchange the threshold of solvent polarity is exceeded, leading to aggregation of the micelles into highly solvent-swollen particles (R _h ~ 160 nm). These particles then condense further and fuse with residual small micelles below ethanol concentrations of 20 wt%, ultimately leading to rigid solid peptide particles that preserve their shape and dimensions (R _h ~ 350 nm) even after isolation from the aqueous dispersion. The final size of the solid, spherical peptide particles, termed "peptide beads", can be controlled by the initial concentration of the peptide in ethanol (T. B. Schuster, D. de Bruyn Ouboter, E. Bordignon, G. Jeschke and W. Meier, Soft Matter 2010, 6, 5596-5604).

Table 1 : Code, sequence, and charge of amphiphilic peptides and the wildtype gA

Code Sequence ^[a| Charge' wt-gA formy/-LV-G-LA-DL-LA-DV-LV-DV-[LW-DL] ₃ -LW-ei/?ano/am/ ^' ne 0

K ₃ -gT H-LK ₃ -[LW-DL] ₃ -LW-/VH ₂ 4+

Ac-X ₃ -gT /\C-[LKC4C ] ₃ -[LW-DL]3-LW-/VH ₂ 0

^[a| one letter code, ^[b| at pH 7, X = LK(AC)

The following description of the hierarchical self-assembly of this amphiphilic peptide is important to understand in which respect the internal structure of the peptide beads will influence the embedding of payloads. The self-assembled peptide beads were investigated in pure water by static light scattering (SLS) and dynamic light scattering (DLS) and showed a narrow size distribution. Guinier-Plot analysis (Fig. 1A) revealed a size of R _g = 236 nm during the process of bead formation. The p-parameter (Rg/Rh) of 0.79, calculated using the measured hydrodynamic radius (Rh) of 300 nm from DLS data analysis, is close to the theoretical value of 1/V(5/3) = 0.78 for solid spheres. When the angular dependence of the particle scattering factor (Fig. 1 B) is compared to the theoretical functions for homogenous, hard spheres (black line) and monodisperse polymer coils (gray line), respectively, a close fit to the hard sphere model is obtained. The spherical shape of the formed particles was visualized by scanning electron microscopy (SEM) of a dried sample as well (Fig. 2A). The typical contrast gradient of solid particles in the transmission electron micrographs (TEM) and the SEM micrographs of a peptide bead broken-up by shock freezing proved the solid sphere character (Fig. 2B and 2C). TEM of uranyl acetate stained peptide beads (Fig. 2D) revealed smaller structures within the beads made visible by the contrast enhancement and thereby indicates a substructure, possibly composed of micelles within the final peptide beads. The measured radius of 10.7 ± 1.8 nm, gained from image analysis of the underlying smaller species within the stained peptide beads, is clearly in the range of micelle sizes.

To obtain topography-based data, atomic force microscope (AFM) was used to compare K ₃ -gT micelles with the surface of Ac-X ₃ -gT peptide beads. Since the image quality in AFM relies on the flatness of the sample, the size of the peptide beads was increased by using an initial peptide concentration in ethanol of 1 .5 mg/ml. This led to beads with diameters larger than 1000 nm and therefore lower surface curvature. The scans performed on top of the beads revealed an astonishing similarity with the scans of the micelles (Fig. 3).

Although the imaging techniques showed micelles within the peptide beads, they allowed only a limited view inside. To disturb the packing of micelles slightly, the initial peptide sequence was elongated with a cysteine, resulting in Ac-X ₃ -gT-C. This modification led to a rougher surface and a looser packing of the formed peptide beads (Fig. 4).

Lyophilisation of these loosely packed beads allowed a view into partially opened beads (Fig. 4b) and also revealed a structure constructed of smaller, spherical subunits. In terms of a functional model, the packing of the micelles into the peptide beads can be compared to a sintering process, commonly known as relating to metal particles or ceramics. The diminished sintering effect in the case of Ac-X ₃ -gT-C can be explained by more form- stable micelles, presumably due to stronger / ^' niramicellar binding of newly formed disulfide bridges.

Although these images qualitatively demonstrate an assembly process that makes use of small, spherical structures that form the resulting peptide bead structures, i.e. multicompartment micelles, they are not associated with significant statistical data for the inside of the peptide beads. Therefore, small angle x-ray scattering (SAXS) was performed on a concentrated aqueous Ac-X ₃ -gT peptide bead sample, using a synchrotron source. Even if the beads are larger than the experimentally available maximal size for SAXS - based on theory, but also indicated by a missing Guinier plateau and a missing Porod regime (that is, q does not scale with q ^'4 at higher q) - the collected data can be used to estimate the size of possible internal substructure.

The scaling of the experimental data with q ^"3 indicates a spherical structure. The absence of a Porod regime also supports the presence of a diffuse interface, since a clear electron density difference between the dispersant and the substructure's surface is missing. Moreover, this behavior is an agreement with a mass fractal behavior. A classical data analysis using a single solid sphere distribution failed. This is interpreted as the

consequence of the aggregated and polydisperse nature of the micelles as shown by microscopy. The use of a bimodal model is not possible without introducing broad errors. Therefore a two-level Beaucage fitting method was used in which the size of the internal substructure was estimated to have a radius of gyration R _g of 9.8 ± 0.2 nm that corresponds to the size of aggregated micelles with a spherical radius of 12.7 ± 0.3 nm on average (Fig. 5). The radius corresponded well with the radii found in AFM, SEM, and TEM. Moreover, the pair distance distribution function (PDDF) showed the expected shape for solid particles' which, in this case, are micelles. To prove consistency, the averaged substructure using this model also had a diameter of approx. 12 nm. Based on the PDDF shape, the presence of inverted micelles with hollow core shell particles in this size can be ruled out. Deduced from the presented microscopic and scattering data, the peptide beads are formed by aggregation of smaller, spherical subunits of -10 nm. The size is in the normal known range for micelles. Based on this fact, three-dimensional models of the peptide beads were built, where the well-known crystal structure of gramicidin-A (B. M. Burkhart, R. M. Gassman, D. A. Langs, W. A. Pangborn and W. L. Duax, Biophysical Journal 1998, 75, 2135-2146) was used for the gT part in combination with a simple, random coil for the three acetylated lysines. The model quickly revealed that a classical approach with one layer of molecules forming a micelle is not possible, since the radius of 10 nm cannot be spanned by one molecule. Ac-X ₃ -gT, with a preserved gramicidin-like secondary structure, has a length of approximately 2 nm. Even when assuming a structural change towards the most space demanding β-sheet alignment of 0.35 nm per amino acid, the peptide has a maximum length of 3.5 nm. Thus, another model than the classical one-layer-approach is necessary to understand the creation of micelles of the measured size.

Defying the lack of further methods to resolve the micellar structure within the beads, a Total Non-Negative Least Square analysis (TNNLS) of the X-ray scattering data for the peptide beads was performed. Basically, the process consists of creating a computed- generation population of nano-objects, calculating the corresponding scattering curve and matching it iteratively with an experimental curve. Results are shown in Fig. 6.

Interestingly, the size distribution curve shows a polydisperse system. There are three diameters visible: one at 6.8 nm, one at 15.6 nm, and one at 21 .2 nm. The 6.8 nm fits well with the dimension of two peptides self-assembled in a tail-to-tail or head-to-head mode, and is commensurate with the peptide size and the classical one layer approach to form a micelle or an aggregate of 7 nm in size. Both the 15.6 nm and the 21 .2 nm can be interpreted as the diameter of the small species found with the other techniques, both being multiples of the 7 nm species. They are polydisperse, as shown by the absence of a clear minimum in the scattering curve. This is not surprising, as due to the aggregation, the dispersant and the object lack a clearly defined border. This additional data supports a model, having three stages of self-assembly: first peptide with peptide in small aggregates, these then form larger micelles, which then aggregate up to the final peptide beads.

With the presented data the persistence of spherical micelles of approximately 20 nm in diameter within the peptide beads is proven. The evaluated data allows two possible models for the structure of the micelles (Fig. 7): According to the aggregated micelle model, the peptides first form micelles of -7 nm diameters, these then aggregate into 21 nm objects due to reduced solubility of the micelles with decreasing ethanol content, which also reduces the flexibility of the objects. Further decrease in ethanol content leads to the aggregation of the objects into the peptide beads until the aggregation process ends due to maturing of the particles into solid peptide beads, which occurs below 20 wt% ethanol. According to the large compound micelle model, the peptides first form inverted micelles (-7 nm) at very high ethanol concentrations, based on the good solubility of the gT sequence in ethanol. With decreasing ethanol content these inverted micelles, presumably stabilized due to H-bonding of the lysine part, start aggregating (-14 nm) and are then surrounded by a layer of remaining peptide to reveal the hydrophilic part to the now aqueous environment (-21 nm). Further decreasing ethanol content leads to aggregation of the 21 nm large compound micelles into peptide beads.

A terminology "multicompartment micelles" is considered best to describe the present system. Multicompartment micellar structures are promoted to entrap and release payloads with different hydrophilic and hydrophobic properties. The herein presented hierarchical self-assembled peptide beads potentially find application in drug delivery, due to the combination of the multicompartmentized structure, the high performance regarding encapsulation efficiency, and the biodegradable peptide-based origin.

The peptide Ac-X ₃ -gT is soluble in ethanol and forms peptide beads with controllable radii in the range of 100 to 800 nm when transferred to water. It was showed with both, microscopic and scattering techniques, that the peptide beads are solid spheres, self- assembled in a hierarchical manner by aggregation of smaller spherical subunits of -10 nm in radius. The size of these is comparable with micelles formed by the charged peptide analogue, K ₃ -gT. The peptide beads hold segregated hydrophobic compartments in form of the micellar subunits and a continuous, rather hydrophilic micellar coronae compartment. Hence, the peptide beads offer compartments for payloads with different affinities, a property generally advertised for multicompartment micelles. In contrast to vesicular structures, which offer a very large volume/entity for hydrophilic payloads but relatively small volume/entity for hydrophobic payloads, multicompartment micelles are advantageous since they offer almost equal space for both species. Due to the small size, and the short distances between the hydrophobic regimes, the presently claimed multicompartment micelles also possess optimal properties to embed macromolecules that have a hydrophilic backbone, sputtered with hydrophobic sections/residues, like for example DNA and RNA. Since the peptide beads are solid in their matured form, payloads are protected from the environment overall until their release by, e.g. degradation of the peptide or diffusion release mechanisms. Nevertheless, a representation of a small fraction of the payload that is embedded close to the outside is not excluded and is useful for functionalization and recognition applications (e.g., targeting, vaccine). Furthermore the solid character allows lyophilization and storage of the peptide beads, which is a strong advantage in contrast to vesicle systems. For all these reasons, as well as for biodegradability, the described peptide beads have useful drug delivery applications.

Peptide bead formation and payload embedding

The charged peptide K ₃ -gT forms micelles with a radius of -10 nm in aqueous solutions. The acetylated and uncharged analogue Ac-X ₃ -gT is soluble in ethanol and forms spherical peptide particles, termed "peptide beads," having adjustable diameters between 200 and 1500 nm when solvent is displaced by water. The beads exhibit solid character at ethanol concentrations below -20 % wt. Above this concentration, the peptide is a loose aggregate of micelles or individual micelles.

With the embedding method as used - a straight forward co-encapsulation - the payload is dissolved in ethanol or water, followed by mixing with the ethanol solution of the peptide to obtain solutions of at least 50 % ethanol. Subsequent dialysis against water yields peptide beads with embedded payloads. As reported earlier for the hydrophobic and hydrophilic dyes, rose bengal and carboxy fluorescein (C. Dittrich and W. Meier, Macromolecular Bioscience 2010, 10, 1406-1415), respectively, the payloads are not only embedded but considerably enriched inside the peptide beads.

It has now be found that other payloads may be used, namely BODIPY630 (hydrophobic), or Alexa488 (hydrophilic). Confocal laser scanning micrographs of peptide beads with embedded dyestuffs can be used to demonstrate proper incorporation.

It has further been found that doxorubicin, paclitaxel, siRNA, and plasmids (DNA), can be loaded. No measurable influence on the self-assembled structure of the peptide was observed within the range of the concentrations used for hydrophobic or for RNA DNA payloads. For hydrophilic doxorubicin, however, an impairment was found in self- assembly when more than 50 % of payload mass was used relative to peptide. Above such concentrations, the size of the resulting peptide beads is drastically reduced, presumably due to shielding of the stabilizing peptide-peptide ττ-ττ-interactions by the payload. However, such high payload-to-carrier ratios are usually not necessary.

The enrichment of payloads in the peptide beads reduces the need for a method to separate non-embedded payloads from the aqueous dispersion. However, the fact that the peptide beads have a higher density than the surrounding water (1 .4 g/cm ³ ) allows easy separation of non-embedded molecules by multiple centrifugation and decantation.

Embedding efficiency (%ee) is defined as %ee = [c] _emb e _dd e _d /[c]totai ^* 100. In case of

RNA DNA payloads embedded in Ac-X ₃ -gT peptide beads, the measured embedding efficiency was only 2 - 4 %ee, which speaks for a passive embedding effect (embedding the available payload with the same concentration as on the surrounding solution in the volume of the carrier).

Premixing RNA with the positively charged peptide analog K ₃ -gT followed by peptide bead formation with Ac-X ₃ -gT enabled active embedding and raised the embedding efficiency to as high as 99.8 %ee (enrichment of the available payload within the carrier). Within the precision of the experiment, this represents a complete embedding of all available RNA, making any further step for separation of payload-filled beads and free payload dispensable. The positively charged peptide, comprising a hydrophobic anchor (gT) and a part that interacts with the payload (K ₃ ), was therefore used as a mediator molecule to enhance the embedding efficiency. The previously low embedding efficiency when using Ac-X ₃ -gT only was obviously caused by the repulsive negative charges of the base pairs. However, the use of both the charged and the uncharged peptide (K ₃ -gT and Ac-X ₃ -gT) was found to influence peptide bead self-assembly: An excess of positive charges (from the K ₃ -gT) also led to the formation of peptide beads of the usual dimensions, but within 24 hours the beads aggregated and developed a strong affinity for glass. A balanced, overall charge, and therefore an equilibrated, mutual compensation of the positive charges from the peptide and the negative charges of the RNA, was found to be necessary to yield finely dispersed but heavily loaded peptide beads.

Payloads such as the siRNA have a length of approximately 6 nm, whereas plasmids are larger and have a spherical diameter of -500 nm. However, neither was seen in TEM of the payload-embedded peptide beads. In contrast, the loaded peptide beads looked similar to the unloaded beads. Thus, the large plasmids obviously curled-up during embedding, fitting into one bead, a phenomenon that was induced by the charge compensation and the peptide amphiphilicity.

Cell internalization

Three different cell lines were used in order to test cellular uptake of the K ₃ -gT/Ac-X ₃ -gT peptide beads: THP-1 (human acute monocytic leukemia cell line) monocytes, THP-1 macrophages, and HuH-7 hepatocytes (hepato cellular carcinoma cells). To visualize the internalization of the potential drug delivery vehicles, cy3-labeled siRNA (250 nM, 99%ee) as well as BODIPY633 (75 nM) were used as payloads for 0.25 mg/ml and 0.5 mg/ml peptide beads, respectively. The cells were incubated with the peptide beads in a 1 :5 dilution in cell growth media for 48 h. As revealed by confocal laser scanning microscopy (CLSM), shown by way of example in Fig. 8, both kinds of beads were internalized by the cells. Internalization of the carrier is a necessary step in the development of a drug delivery system. However, after internalization, the functional payload preserved in this way has to be released at its site of action. In the case of siRNA, this is the cytosol.

Diffuse fluorescence in cells indicated a release of the payload from the peptide beads to the cytosol. However, in the case of the siRNA, whether this diffuse fluorescence in the cytosol is from intact siRNA or from the remaining cy3 molecules after the decomposition of the siRNA cannot be determined. Therefore, it could not be determined whether functional siRNA was released. In order to clarify this, specific gene silencing using siRNA or a plasmid, as well as the toxic effect of doxorubicin and paclitaxel were tested. Toxicity and therapeutic effect

Pure (unloaded) peptide beads are compared with control samples (no beads added) during all experiments. Similar results were always obtained within the limits of experimental error. The resulting absence of any toxic effect by the peptide beads (see Fig. 9 and 10) is an indication of the biocompatibility of the purely peptidic amphiphiles used.

The therapeutic effect of potentially delivered drugs was tested in order to determine not only the internalization of the peptide beads but also a functional drug delivery system at a cellular level. Therefore, cell viability tests were performed with doxorubicin, paclitaxel, and doxorubicin/paclitaxel-loaded peptide beads using THP-1 macrophages and THP-1 monocytes. To minimize non-specific delivery of the drugs, the prepared peptide beads were separated from non-embedded drugs by triple centrifugation/decantation and offered to the cells for 24 h and 72 h in two different concentrations. As Fig. 9 reveals, all drug- loaded samples significantly reduced cell viability, as expected from doxorubicin and paclitaxel residing in monocytes. Since the mechanisms of action of both drugs take place during cell multiplication, no effect was monitored in the non-multiplying macrophages, also indicating no toxic effect of the material used.

In principle, paclitaxel and doxorubicin can enter a cell without a drug vehicle. Although such behavior was minimized by separating free drug from the beads, it cannot be ruled out completely. Therefore the specific silencing of a target gene with siRNA as well as the delivery of a vector in the form of a plasmid was tested in hepatocytes. Neither of these payloads can enter a cell without the help of a vehicle, because the material would be degraded. Gene silencing by the siRNA as well as effects of the delivered plasmid sequence (siRNA production and antibiotic resistance) are capable of demonstrating successful delivery and therefore show the usefulness of the peptide beads as a drug delivery system.

The following payloads were used:

GAPDH siRNA, silencing the transcription of the housekeeping gene glyceraldehyde 3- phosphate dehydrogenase (GAPDH).

Negative control siRNA: scrambled sequence, causing no silencing.

psiSTRIKE-Neomycin vector (containing a hairpin target sequence of PP2Aca that allows expression of siRNA against PP2Aca, and a sequence coding for antibiotic resistance). Negative control: the same vector with a scrambled sequence for the siRNA (causing no silencing).

Overexpression of PP2Aca in hepatocytes plays an important role in liver cancer development. Although the silencing of this gene was used as proof of principle, it can be directly applied in hepatitis C-induced liver cancer treatment.

As Fig. 10A and 10B show, the siRNA as well as the plasmid caused gene silencing compared to the negative control siRNA, negative control plasmid, and the control, measured by quantitative real time polymerase chain reaction (qRT-PCR).

To prove the statistical significance of the silencing - which, at 15%, was highest for the plasmid - the cells were cultivated for three weeks after being kept in the presence of plasmid-loaded peptide beads for 96 h (Fig. 10C). Plasmid transfected cells also replicated the plasmid itself during multiplication. To prove the successful delivery, cells transfected with plasmid were selected by addition of antibiotic in culture medium, which caused the non-transfected cells to die. The survival of the selected cells, resulting from the delivered antibiotic resistance, alone indicates the capacity of the peptide bead system to deliver a payload into the cells. The performed qRT-PCR of the stability-selected cells showed a significant silencing of 33%, caused by the siRNA produced by the delivered plasmids. The effective gene delivery capacity of the peptide beads was therefore demonstrated.

Examples

Abbreviations

AA amino acid

ACN acetonitrile

AFM atom force microscope

Boc tert-butoxycarbonyl

CCA a-cyano-4-hydroxycinnamic acid

CD circular dichroism

CLSM confocal laser scanning microscope

cmc critical micelle concentration

cw continuous wave

Da Dalton, molecular weight DA degree of acetylation

DCC N,N'-dicyclohexylcarbodiimide

DCM dichloro methane

dd H ₂ 0 double distilled water

DIPEA Λ/,/V-diisopropylethylamine

DLS dynamic light scattering

DMF Λ/,/V-dimethylformamide

DMPC dimyristoylphosphatidylcholine

DNA deoxyribonucleic acid

DQC double-quantum coherence

EDT ethanedithiol

EPR electron paramagnetic resonance

Eq equivalent

EtOH ethanol

Fmoc 9-fluorenylmethyloxycarbonyl

gA gramicidin A

GAPDH glyceraldehyde-3-phosphate dehydrogenase

GPC gel permeation chromatography

HCTU 2-(6-chloro-1 H-benzotriazole-1 -yl)-1 ,1 ,3,3-tetramethylam hexafluorophosphate

HOBt N-hydroxybenzotriazole

HPLC high pressure (performance) liquid chromatography IPE diisopropylether

IR infrared spectroscopy

K _sv Stern-Volmer constant

LCM large compound micelle

LCST low critical solution temperature

MALDI-TOF-MS matrix assisted laser desorption/ionization - time of flight spectroscopy

MCM multicompartment micelle

Mw molecular weight

NCA N-carboxy anhydride

NIR near infrared

NMP N-methyl-2-pyrrolidone

PAA poly(acrylic acid)

PDDF pair distance distribution function PDMS poly(dimethylsiloxane)

PEG poly ethyleneglycol

PEO poly(ethylene oxide)

PGA poly glutamic acid

PLGA poly(L glutamic acid)

PLL poly(L lysine)

PMOXA poly(2-methyloxazoline)

PNIPAM poly (N-isopropylacrylamide)

PS polystyrene

PyBOP benzotriazol-1 -yloxy) tripyrrolidinophosphonium hexafluorophosphate qRT-PCR quantitative real time polymerase chain reaction

R _g radius of gyration

R _h hydrodynamic radius

RNA ribonucleic acid

RP reversed phase

R _s spherical radius

RT room temperature

SAXS small angle X-ray scattering

SEM scanning electron microscope

shRNA small hairpin RNA / short hairpin RNA

siRNA short interfering RNA / small interfering RNA / silencing RNA

SLS static light scattering

SPPS solid-phase peptide synthesis

TBTU 0-(benzotriazol-1 -yl)-N,N,N',N'-tetramethyluronium tetrafluoroborate TEAP triethyl ammonium phosphate

TEM transmission electron microscope

TES triethylsilane

TFA trifluoroacetic acid

TFE trifluoro ethanol

THF tetrahydrofuran

TIS triisopropylsilane

TNBS 2,4,6-trinitrobenzenesulphonic acid

TNNLS Total Non-Negative Least Square analysis

Tris tris-(hydroxymethyl)aminomethan Amino acids (one letter code) and sequences

Ac Acetylated

DL D-leucine

LC L-cysteine

LK L-lysine

LW L-tryptophan

LX acetylated L-lysine

-gA gramicidin A inspired hydrophobic part of the amphiphile

-gT truncated version of gA

Materials

Materials and reagents were of the highest commercially available grade and were used without further purification, unless indicated. HCTU, Rink Amide AM resin (0.61 mmol/g) and Fmoc-Trp(Boc)-OH was purchased from IRIS Biotech GmbH. All other amino acids were obtained from Novabiochem. Dichloromethane and ethanol (96%) F15 were provided by Brenntag Schweizerhall AG, DMF by J. T. Baker and acetonitrile (ACN) by Fischer Scientific. Fluorescent dyes were purchased from Invitrogen Inc. DMF was treated with aluminum oxide to reduce free amines prior to application in peptide synthesis.

Solvent exchange was carried out in 24-well sitting-drop crystallization plates (HR3-158, Hampton Research) or by dialysis (Spectrum, cellulose ester (CE), MWCO=500-1000 Da) against double-distilled (dd) water. Fluorescently labeled and unlabeled GAPDH-siRNAs (AM4649, AM4624) were purchased from Ambion (Applied Biosystems). The psiSTRIKE- Neomycin vector and MTS-assay were purchased from Promega AG (Switzerland). The used vector was equipped with a hairpin target sequence for PP2Aca to express siRNA against PP2Aca. The sequences of the hairpin oligonucleotide for PP2aca messenger RNA were 5 '-AC C G G G ATAC C GTTTAATTTAATTC AAG AG ATTAAATTAAAC G GTAT CCCTTTTTC-3' (SEQ ID NO:1 ) and 5'-TGCAGAAAAAGGGATACCGTTTAATTTAA

TCTCTTGAATTAAATTAAACGGTATCC-3' (SEQ ID NO:2, loop is underlined). For control, scrambled sequences for the siRNA sequence designed to human PP2Aca were used. The oligonucleotide sequences were 5'-ACCGTTAATGGCTACGAATTAT

TTCAAGAGAATAATTCGTAGCCATTAACTTTTTC-3' (SEQ ID NO:3) and 5'-TGCAG AAAAAGTTAATG G CTACG AATTATTCTCJ GAAATAATTCGTAG CCATTAA-3 ' (SEQ ID NO:4). Remaining chemicals and buffers were purchased from Sigma-Aldrich. Peptide Synthesis

All peptides were synthesized on a Syro I Peptide Synthesizer (MulitSynTech GmbH, Witten, Germany) on solid phase using Fmoc-strategy. 2-(6-chloro-1 H-benzotriazole-1 -yl)- 1 ,1 ,3,3-tetramethylaminium hexafluorophosphate (HCTU) as the coupling reagent and N- ethyldiisopropylamine (DIPEA) dissolved in N-methyl-2-pyrrolidone (NMP) as the base were used to couple a-N-Fmoc-protected amino acids to the resins (220 mg, Rink Amide AM, 0.61 mmol/g, equals 1 equiv.). For elongation, Fmoc-Xxx-OH (0.5 mol/L, 4 equiv.), HCTU (0.5 mol/L, 4 equiv.) dissolved in dimethylformamide (DMF), and DIPEA (12 equiv.) were added to the resin. The mixture was agitated for 1 h and washed with DMF (3x3 ml_). Fmoc deprotection was performed with 20% piperidine in DMF followed by 3 min agitation, draining, and repetition of deprotection for 10 min. Resin was subsequently washed with DMF (5x3 ml_). Acetylation of unreacted amine groups was performed following each coupling using acetic anhydride/DIPEA (3 mol/L, 5 equiv.) in DMF. After synthesis, the peptide resin was washed alternating with DMF (3x6 ml_), isopropanol (3x6 ml_), and dichloromethane (3x6 ml_), and dried under vacuum. For Ac-E ₆ -gA and Ac-K ₆ -gA, N-terminal acetylation was performed after the completed reaction with 30 equivalents of Ac ₂ 0/DIPEA for 1 hour. Partially acetylated peptides were produced by using Fmoc-Lys(Ac) and Fmoc-Lys(Boc) during synthesis. Ac-X ₈ -gA, the precursor for Ac-X ₈ -gA-OEt, was synthesized by post purification acetylation from K ₈ -gA.

In a scaled up synthesis using 5 g of Rink Amide AM resin, the same protocol was followed in manual steps. For washing isopropanol was used alternating with DMF. Due to the size, reactions were carried out in a 500 ml glass solid phase reactor using only 3 equiv. of amino acids and coupling reagents for the reaction. The pH was controlled and kept above 9 during reaction, after each amino acid coupling and cleavage of the Fmoc- protection group, the resin was probed for free amine groups using a Kaiser- and TNBS (trinitrobenzene sulfonate )-test. There was no need for NMP as a co-solvent.

Cleavage from the resin and removal of protective groups was performed with a mixture of 95% trifluoroacetic acid (TFA), 2.5% triisopropylsilane (TIS), and 2.5% H ₂ 0. In case of cysteine containing peptides a mixture of 94% TFA, 2.5% TIS, 2.5% ethanedithiol and 1 % H ₂ 0 was used. The cleavage cocktail was filtered and resin was washed additional two times with cold cleavage cocktail (2 mL) and precipitated in 40 mL cold diisopropylether (IPE). The precipitated crude peptide was washed with cold IPE three times by centrifugation and decantation, before being dried under vacuum.

Peptide purification, post modification and characterization

Purification and analysis of the peptides were carried out by HPLC (Shimadzu

Prominence 20A, Japan) on reverse phase (RP) columns (Merck Chromolith, RP-18e, 100 mm x 10 mm and 100 mm x 4.6 mm; Merck LiChrospher 100, RP-18e, 5 μηη, 250 mm x 10 mm and 250 mm x 4.6 mm). The ground crude peptides were dissolved in 2 ml_ DMF, diluted with H ₂ 0 (0.1 % TFA, 8 M urea) to a final volume of 20 ml_. If necessary, acetonitrile (ACN) was added to dissolve peptide assemblies. The resulting solution was filtered through a 0.45 μηη PTFE syringe filter and pumped to the HPLC. Two runs were necessary to yield sample purities higher than 95%. Linear gradients of solvent A (ACN) and solvent B (first run: 0.1 % TFA, second run: 2 % AcOH or 0.23% formic acid in dd H ₂ 0). The eluted sample was collected in fixed volume fractions which were subsequently analyzed for purity and molecular weight. Fractions containing more than 80% product (HPLC with 280 nm absorption) were applied to the second purification run. AcC-X ₃ -gT and C-K ₃ -gT disulfides were reduced with 1 ,3-propanedithiol (5 equiv.) in methanol or tris(2-carboxyethyl)phospine (5 equiv.) in water/ACN solution before the second run. Final fractions with purities higher than 95% were combined, neutralised with ammonia and lyophilized until leaving a fluffy, white final product.

Purification of Ac-E ₆ -gA was carried out by HPLC on a PRP-3 column (10 μηη 300 A, 4.1 x 250 mm, Hamilton) for peptide/protein/DNA purification. Also here, two runs were necessary and following conditions were used: in run one, ACN/0.5% aqueous TRIS-HCI solution (pH 7), in run two, ACN/50 mM NH ₄ OAc/HOAc (pH 7).

For verification, final products were again analyzed by HPLC for purity and MALDI-TOF- MS for mass confirmation. Examples are illustrated in Table 2. Products were stored under argon at -18°C. Table 2: Exemplary mass confirmation by MALDI-TOF-MS and HPLC purities of selected peptides.

Code ^[al Mass ^[bl Mass Purity ^[cl

(g mol- ¹ ) (MALDI-TOF-MS)

Ac-Xs-gA 3328.1 3329.9 >99.5% ^*

Ac-Xs-gA-OEt 3358.1 3357.1 >99.5% ^* Ac-E ₆ -gA 2684.0 2682.5 95%

Ac-K ₆ -gA 2678.3 2677.4 96%

[a] one letter code with abbreviations for gramicidin (gA), acetylated lysine (X) and acetylated N-terminus (Ac) [b] calculated with Pinsoft32 [c] Area% on an analytical RP- HPLC [ ^* ] HPLC purity from precursor K ₈ -gA.

Overall yields of the synthesis and purification were generally between 10-15 %, overall yields of the synthesis, purification, post-purification acetylation and end-purification were generally between 6-10 %, and therefore in the normal range for solid phase peptide synthesis and HPLC purification.

Post-purification acetylation of the free N-terminal and lysine amines was performed on the purified peptide dissolved in DMF using a 40-fold excess of acetic anhydride and DIPEA. Completeness of the reaction was controlled by MALDI-TOF-MS. The reaction mixture was repurified according to the procedure described above.

Transesterification of the Ac-X ₈ -gA amide to the Ac-X ₈ -gA-OEt was performed in ethanol/0.1 M HCI at 70°C for 3 x 10 min and intermediate ultrasonic treatment for 3 x 30 s at 70°C.

MALDI-TOF-MS was performed on Voyager-DE™ System (Applied Biosystems, USA) with a-cyano-4-hydroxycinnamic acid as the matrix in positive ion reflector mode for lysine-rich peptides and negative ion reflector mode for glutamate-rich peptides, respectively, an accelerating voltage of 25 kV, grid voltage of 75% and 300 ns extraction delay time. These standard values were optimized according to the sample. A 100-well stainless steel or gold plate was used. Hydrophobicity was calculated using Pinsoft32 (Chiron Technologies Pty Ltd., Australia) which uses the hydrophobicity algorithm from J. L. Fauchere and V. Pliska, Eur. J. Med. C em. 1983, 18, 369-375. (for X = Lys(Ac) a value of - 0.55 was used). Nanostructure formation

Either the peptide or the corresponding mixtures (e.g. AcC-X ₃ -gT / AcC(sl)-X ₃ -gT, Ac-X ₃ - gT / K ₃ -gT) were dissolved in ethanol at the appropriate concentration and filtered through a 0.2 μηη hydrophilic PTFE filter. The self-assembly process was initiated by solvent exchange from the organic solvent to dd H ₂ 0, the corresponding dd H ₂ 0/ethanol mixture, or buffer [NaCI (140 mM), EDTA (0.5 mM), TRIS (10 mM), NaN ₃ (0.003 mM) adjusted to pH 7.4 with HCI and filtered through a 0.2 μηη hydrophilic PTFE filter)]. A change in polarity was usually accompanied by opalescence. Solvent exchange (in dialysis tubes or on crystallization plates) was performed within 24 h with dd H ₂ 0 exchanged three times. Since K ₃ X ₅ -gA to K ₈ -gA are water soluble, they were dissolved directly in water or buffer for surface tension measurements. Composite gold nanoparticles were produced by dissolving the cysteinated peptide in ethanol at 1 mg/ml and filtered through a 0.2 μηη hydrophilic PTFE filter. Aqueous dispersions of gold nanoparticles (5 nm or 25 nm, Sigma- Aldrich) were mixed added to the peptide solution 1 :1 and shaken during reaction for one week. The remaining non-reacted peptide was removed by centrifugation and

decantation, and refilled with fresh ethanol. Self-assembly was induced by solvent exchange as described above.

Nanostructure characterization

Microscopic methods.

Transmission electron microscopy (TEM) images were taken on an FEI Morgani 268D operated at 80 keV. 5 μΙ_ of the sample was deposited on carbon-coated, parlodion (2% in n-butyl acetate )-covered copper grids, blotted after 2 min with a filter paper and dried in air. If necessary, staining was performed with uranylacetat (2%) for 10 s before drying.

Scanning Electron Microscopy (SEM) was performed on a Philips XL 30 ESEM operating between 2 and 5 kV. The aqueous peptide samples were frozen with liquid nitrogen and lyophilized directly on the SEM grid, or lyophilized powder was put on a double-sided carbon sticker before being sputtered with gold or platinum (2-5 nm). Atomic force microscopy (AFM) measurements were carried out using a 5100 Agilent system (formerly PicoLE Molecular Imaging) equipped with a multi-purpose scanner. Images were acquired using a silicon cantilever (type-NCHR PointProbe® Plus, force constant 42 N/m) as indicated by the manufacturer for acoustic mode images. Samples were prepared by exposing aqueous peptide sample to a freshly cleaved mica surface for 5 min, removing the supernatant solution, and drying.

Confocal laser scanning microscopy measurements were performed on a Zeiss Confocor 2 LSM equipment. For the peptide vesicles, BODIPY 650/665 and Alexa488 solutions were added to an ethanol solution with Ac-E ₆ -gA (0.5 mg mL ^"1 ) to reach a concentration of 0.2 μηηοΙ L ^"1 and 0.5 μηηοΙ L ^"1 respectively before vesicle formation and dialysis

(MWCO=100 kDa). For peptide beads and cell imaging, the appropriate filter/channel settings were applied and combined to fit the demands of the fluorophores: tryptophan (ex. 405 nm, em. 420 nm), Cy3-labeled siRNA (Ex. 543 nm, Em. 573 nm), Hoechst 33342 (Ex. 350 nm, Em. 461 nm), CellMask DeepRed (Ex. 649 nm, Em. 666 nm), Lystotracker DND-99 (Ex. 577 nm, Em. 590 nm), and transmission channel (DIC). Cell staining was performed directly before the microscopy measurements according to the manufacturer's protocol. Shortly, the tempered dye solutions were added simultaneously to the cells in 2 ml PBS buffer and incubated for the needed time at 37°C (Hoechst 33342: 6 μΙ 16.2 mM 30 min, Lysotracker: 2 μΙ 1 mM 30 min, DeepRed: 2 μΙ 5 mg/ml 5 min). The cells were rinsed with fresh PBS buffer and instantly used for microscopy at room temperature.

Scattering methods

Dynamic and static light scattering (DLS and SLS) measurements were performed with an ALV/CGS-8F platform-based goniometer system equipped with an ALV/-5000/E correlator and a HeNe-laser with a wavelength of 633 nm (35 mW). Measurements were made at 20°C and at a scattering angle 0 from 30° - 150° in steps of 10°.

DLS correlation functions were fitted with the 2 ^nd cumulant function or the CONTIN algorithm. The resulting hydrodynamic radii (mass weighted) were extrapolated to zero wave vector (q = (4πη ₀ /λ ₀ ) sin (Θ/2)) and to zero concentration. The Stokes-Einstein equation was used to calculate the hydrodynamic radius. To determine the size dependence of the peptide beads, 2 ^nd cumulant fit analysis at 90° with measurement times of 300 s was used. For the pH dependent size evolution of the peptide vesicles, the initial peptide concentration was 0.34 mg mL ^"1 . Here, also the 2 ^nd cumulant function (90°, 300 s) was used. Aliquots of 2-10 μΙ_ of corresponding molarities (10 ^"3 to 1 mol L ^"1 ) NaOH or HCI were used to adjust the pH of the 800 μΙ_ peptide solution. SLS data analysis was done with ALVStat 4.31 software (ALV, Langen, Germany) by constructing a Guinier plot. A general description of the methods used can be found in J. Pencer and F. R. Hallett, Langmuir 2003, 19, 7488-7497; O. Stauch, R. Schubert, G. Savin and W. Burchard, Biomacromolecules 2002, 3, 565-578; and W. Burchard, /Advances in Polymer Science 1983, 48, 1 -124.

Small-Angle X-ray scattering (SAXS) was performed using the Spot Beamline at BESSY (O. Paris, C. Li, S. Siegel, G. Weseloh, F. Emmerling, H. Riesemeier, A. Erko and P. Fratzl, J. AppI. Crystallogr. 2007, 40, s466-s470), using a wavelength of λ = 1.0000 A. Samples were measured using an ultrasonic levitator. The scattering pattern of silver behenate served as an external calibration standard, and 2D data were converted using the program FIT2D (A. P. Hammersley, S. O. Svensson, M. Hanfland, A. N. Fitch and D. Hausermann, High Pressure Research: An International Journal 1996, 14, 235 - 248). Bidistilled water was used as blank, and data were analyzed using a two level Beaucage method as implemented in the Irena SANS package (J. Ilavsky and P. R. Jemian, J. AppI. Crystallogr. 2009, 42, 347-353) and Igor Pro 6.1 1 (Wavemetrics). IPG-TNNLS (Internal Point Gradient - Total Non-Negative Least Square) analysis was performed as

implemented in the Irena Package version 2.38 using Igor Pro 6.1 1 (Wavemetrics). Data were fitted for a nanoparticle population with sizes starting from 3 nm to 50 nm and a logarithmic binning. The NNLS approach parameter was set to 0.5 and the maximum number of iterations was set to 400, which proved to be satisfactory to obtain a convergence.

Gel permeation chromatography (GPC) Measurements of gramicidin A (gA) and Ac-X ₃ -gT (c = 1 mg/mL) were performed on a Shimadzu Prominence HPLC, using an Agilent PLgel column (5 μηη, 10 ³ A, 2 x 250 mm) and isocratic elution with THF/10% H ₂ 0. Stabilizer-free solvent improved detection at the tryptophan adsorption maximum of 280 nm. For the comparison of gA, Ac-X ₈ -gA, and Ac- X ₈ -gA-OEt two GPC-columns were mounted in series (1 . CATSEC-100, 5 μιτι, 250 x 4.6 mm, 2. GPCPEP 5 μηη, 250 x 4.6 mm, Eprogen Inc.). Isocratic elution (1 mL min ^"1 ) was performed with 95% THF (without stabilizer for UV detection at 220 nm, T = 295K) and 5% dd H ₂ 0.

Biological experiments

Cell viability tests on THP-1 cells (MTS assay)

Peptide beads were produced by using filtered (0.2 μηη hydrophilic PTFE) ethanol solutions of the peptide and the payloads to produce the mixtures in ethanol with Ac-X ₃ -gT (0.25 mg/ml), Ac-X ₃ -gT (0.25 mg/ml) / paclitaxel (150 μΜ), Ac-X ₃ -gT (0.25 mg/ml) / doxorubicin (275 μΜ), and Ac-X ₃ -gT (0.25 mg/ml) / doxorubicin (138 μΜ) / paclitaxel (75 μΜ). Self-assembly was induced according to the paragraph nanostructure formation above. The payload filled peptide beads were separated from potentially free payload by triple centrifugation in Amicon Ultra 15 Centrifugal Filter Units (15 ml, 100Ό00 Da MWCO, Millipore), and refilling with dd H ₂ 0.

Cell viability tests for payload filled and unfilled peptide beads were performed using ΤΉΡ- 1 monocytes and macrophages. Monocytes were split and prepared in complete RPMI medium (2 ml well, 20Ό00 cells/well). Peptide bead samples were added to the medium according to the experimental setup in 1 :10 and 1 :100 dilutions and incubated at 37°C for 3 days. Experiment was performed in triplicate. For THP-1 macrophages, the same procedure was performed with 50Ό00 cells/well after differentiation of the THP-1 monocytes to macrophages in the presence of 100 nM PMA (phorbol 12-myristate 13- acetate, Sigma-Aldrich 79346). After incubation, MTS was added (20 μ l/well, MTS kit CellTiter 96 Aqueous One Solution Cell Proliferation Assay, PROMEGA G3580). After incubation for 1 h at 37°C, the absorbance at 490 nm was measured according to the manufacturer and compared corrected using a blank sample (medium without cells).

Gene silencing

Peptide beads were produced by using filtered (0.2 μηη hydrophilic PTFE) ethanol solutions of the peptides and aqueous solutions of the payloads to produce mixtures in ethanol with Ac-X ₃ -gT (0.25mg/ml), siRNA (250 nM) or same mass of plasmid, K ₃ -gT to neutralize the basepairs (3.57 μg/ml for 1 :1 charge compensation according to the amount of siRNA/plasmid). Self-assembly was induced according to the paragraph nanostructure formation above. Encapsulation efficiency was measured according to the fluorescence method described above.

The translucent peptide beads were added to the growth media of the HuH-7 cells in the appropriate concentrations and incubated for the indicated times at 37°C and 5% C0 ₂ .

RNA isolation, reverse transcription, and SYBR-PCR: Cells were washed twice with PBS. The total RNA was isolated from the cells using Trizol according to the manufacturer instructions (Invitrogen, USA). RNA was reverse transcribed by M-MLV reverse transcriptase (Promega, Switzerland) in presence of random hexamers (Promega) and dNTPs. After incubation for 5 min at 70°C and 1 h at 37°C, the reaction was stopped by heating at 95°C for 5 min. SYBR-PCR was performed based on SYBR-Green

Fluorescence (Applied Biosystems). To prevent influence from genomic DNA

amplification, the primers were designed across exon-exon junctions. The forward and reverse primers for GAPDH were 5'-CTCCTCCTGTTCGACAGTCA-3' (SEQ ID NO:5) and 5'-ACCTTCCCCATGGTGTCTGA-3' (SEQ ID NO:6), respectively. The forward and reverse primers for PP2Aca were 5'-CCACAGCAAGTCACACATTGG-3' (SEQ ID NO:7) and 5'-CAGAGCACTTGATCGCCTACAA-3' (SEQ ID NO:8), respectively. The ACT value was derived by subtracting the threshold cycle (CT) value for GAPDH, which served as an internal control, from the CT values for PP2Aca. For GAPDH, only arbitrary values were used. All reactions were run in duplicate using the ABI 7000 Sequence Detection System (Applied Biosystems). The mRNA expression level of PP2Aca was expressed as a fold increase or fold decrease according to the formula 2 ^{ACT(PBS)" ACT} < ^Treatment >.

Table 3: List of synthesized peptides and their bead-forming properties B = bead forming or partially bead forming

Payload

The efficacy of peptide beads to combat IFNv-involving disorders is tested through the impact of the mediated expression of an IFNy antagonist on the development of neuronal degeneration, neuroinflammation, and/or demyelination in animal models of Amyotrophic Lateral Sclerosis (ALS), Parkinson's disease (PD), and Multiple Sclerosis (MS).

For ALS, mice are used bearing the SOD1 G93A mutant variant, which is responsible for autosomal-dominant forms of ALS and triggers toxic processes within motoneurons as well as in astrocytes and microglia. Mice overexpressing these variants develop a motor syndrome with features closely replicating the human disease, including motoneuron degeneratation and hind limb tremor and weakness, which rapidly progress to paralysis and death. Astrocytes and microglial cells participate in disease progression from early motor symptoms to death. For PD, genetic rat models of the disease are used, based on the adeno-associated virus (AAV)-mediated expression of human osynuclein. Animals overexpressing human o synuclein progressively develop a number of consistent motor defects reminiscent of PD symptoms due in part to the degeneration of the nigrostriatal neuronal connections. These models also undergo early and sustained neuroinflammatory reactions similar to those reported in PD patients, notably high pro-inflammatory cytokines including IFNy, early activation of microglia, and infiltration of lymphocytes, even before overt neuronal cell death. AAV vectors directly infused in rat brain have been shown to induce minimal activation of astrocytes and microglia, mainly localized in the infusion tract.

For MS, models include toxin-induced demyelination models, viral and transgenic rodent models, as well as EAE marmosets. It is worth noting that, although EAE mice are the most widely used mouse models of MS and share similarities with the human disease, including increased IFNy secretion by autoreactive T cells, they drastically differ from human MS in the pathology of the lesions and inflammation processes. Observations also markedly vary according to genetic backgrounds or time and area of intervention. While IFNy mediates neuroinflammation and administration of IFNy leads to an exacerbation of the disease in patients with MS, several studies in EAE rodents propose a disease-limiting role for IFNy. EAE induced in primates appears to be a more reliable model to evaluate safety and efficacy for the treatment of MS. EAE marmosets closely resemble the human disease and display increased IFNy, produced by T cells infiltrating the CNS. Most importantly, some therapeutic approaches, such as NGF administration or treatment with anti-CD20 antibodies that successfully reduce MS lesions and relapses, were shown to exert their protective effect, inter alia, by downregulating the production of IFNy in the CNS.

The DNA cassettes, which are incorporated into the peptide beads, consist of the mouse IFNyRI -COMP expression module (nucleic acid SEQ ID NO:17 coding for amino acid SEQ ID NO: 18) or the human IFNyRI -COMP expression module (nucleic acid SEQ ID NO:19 coding for amino acid SEQ ID NO:20), comprising the extracellular part of mouse IFNyRI (SEQ ID NO:9 (nucleic acid) coding for SEQ ID NO:10 (amino acid)) or of human IFNyRI (SEQ ID NO:1 1 (nucleic acid) coding for SEQ ID NO:12 (amino acid)), respectively, fused to an oligomerisation domain, i.e. the COMP element (SEQ ID NO:21 (nucleic acid) coding for SEQ ID NO:22 (amino acid)) and tagged with Human influenza haemagglutin (HA) tag, under the control of either the ubiquitously expressed chimeric CMV-chicken β-actin promoter (SEQ ID NO:16) or the glial-specific GfaABC1 -D promoter (SEQ ID NO:14). A COMP element-expressing or a reporter EGFP-expressing cassette are used as control expression modules. The COMP element includes a linker peptide (PQPQPKPQPKPEPE of SEQ ID NO:23), the pentamerisation domain of the cartilage oligomeric matrix protein (COMP) (Rattus norvegicus, Genbank accession number NP_036966) and Hemagglutinin sequence (HA tag sequence) at the C-terminal part (YPYDVPDYA of SEQ ID NO:24). It promotes the pentamerisation of the IFNyRI fragment, which dramatically increases its affinity to IFNy and makes it thus efficiently compete with IFNy receptors.

The efficacy of payload incorporation is measured after dissolving the loaded peptide beads in appropriate solvent by UV spectroscopy or fluorescence, in case of fluorescently labeled payload. The experimental design comprises intra-cerebroventricular and intra-cisternal (for ALS models), intra-nigral and intra-striatal (for PD models), intra-thecal (for MS models), and intra-venous (for all models) delivery of low, medium, and high doses of peptide beads, injected before, at, and after the time of disease onset. The route of delivery and the efficacy of IFNyRI -COMP and of the control expression modules to be efficiently delivered and secreted are evaluated by immunohistochemical and biochemical analyses in brain and spinal cord using appropriate markers. Distribution in the central nervous system and peripheral expression in blood, liver, heart, spleen, kidneys, lungs, and muscles is examined by anti-HA or anti-EGFP immunostaining. The neuroprotective effects are assessed by studying the course of the disease, using the following parameters: survival, behavioural tests (e.g. swimming tank test, rotarod test, electromyography, tests for spontaneous and drug-induced motor asymmetries, animal weight measurement), and histopathological tests (e.g. impact on neuronal survival using specific neuronal and innervation markers, neuronal abnormality studies, impact on inflammatory reactions using astroglial/microglial activation markers, demyelination quantification, markers for infiltrating lymphocytes).

These experiments provide a demonstration of the therapeutic impact of the long-term IFNy blocking in the central nervous system on animal survival, neuronal degeneration and function.