WO2018207115A1

WO2018207115A1 - Tumor-homing peptides, conjugation products thereof and their use in diagnostic and therapy

Info

Publication number: WO2018207115A1
Application number: PCT/IB2018/053225
Authority: WO
Inventors: Angelo Corti; Flavio Curnis; Anna Maria GASPARRI; Angelina SACCHI; Francesco SUDATI; Fabio Pastorino; Mirco Ponzoni
Original assignee: Fondazione Centro San Raffaele; Irccs Istituto Giannina Gaslini; Ospedale San Raffaele Srl
Priority date: 2017-05-10
Filing date: 2018-05-09
Publication date: 2018-11-15

Abstract

The present invention relates to novel tumor-homing peptides, namely to a modified NGR motif and to peptides comprising said motif. The invention also relates to conjugation products comprising said modified motif and said peptides and their use in diagnostic and therapy.

Description

TUMOR-HOMING PEPTIDES, CONJUGATION PRODUCTS THEREOF AND

THEIR USE IN DIAGNOSTIC AND THERAPY

***

Summary of the Invention

Technical Background

Tumor-homing peptides represent an important class of ligands that can be exploited for enhancing the accumulation of anticancer drugs or imaging compounds in tumors, thereby improving their therapeutic or diagnostic properties. Among the various peptide motifs identified for this purpose, the NGR (asparagine-glycine- arginine) motif (herein also "NGR") is one of the most widely used. Indeed, peptides containing said motif have proven to be useful in delivering a variety of agents to tumors, such as chemotherapeutic drugs, liposomes, gold nanoparticles, anti- angiogenic compounds, DNA complexes, viral particles, imaging compounds, cytokines, such as tumor necrosis factor (TNF)-a, interferon (ΓΡΝ)-γ, IFN-a-2a and so on.

A primary mechanism underlying the tumor-homing properties of NGR relies on the recognition of aminopeptidase N ("CD13", EC 3.4.11.2) expressed on the membrane of endothelial cells and other perivascular cells in tumors. CD 13 is not expressed in normal endothelial cells of vasculature but is overexpressed on tumor neovascular endothelial cells and in some tumor cells, where it plays an important role in tumor angiogenesis. Therefore, NGR-containing peptides are considered useful for ligand- targeted delivery of drugs and other active moieties to angiogenic blood vessels in tumors.

Remarkably, some NGR-drug conjugates are being tested in patients. For example, the CNGRCG-TNFa fusion product, referred to as NGR-TNF, is currently being tested in Phase II and III clinical studies in patients with solid tumors, with evidence of activity and good tolerability. The good results obtained with this product and also with other NGR-based products in preclinical and clinical studies highlight the usefulness and versatility of NGR peptides in the development of tumor-homing drugs and imaging agents.

A major problem associated with the use of these peptides is related to the fact that the asparagine residue of the NGR sequence has a strong propensity to undergo deamidation, which is a spontaneous reaction that converts asparagine (N) into aspartate (D) and isoaspartate (isoD), i.e. converts NGR into DGR and z^'soDGR. This reaction occurs very quickly in NGR containing peptides so that the half-life of CNGRCG is 3-5 hours in physiological solutions. The consequent structural changes are associated with a dramatic change of function, causing receptor switching from CD 13 to integrins. Indeed, the z^'soDGR, but not NGR or DGR, can mimic RGD- motif and bind to the RGD-binding pocket of various integrins, such as ανβ3, ανβ5, ανβό, ανβ8 and α5β1, with different affinity and selectivity depending on flanking residues and molecular scaffolds. Considering the rapid kinetics of NGR deamidation, this reaction may occur during NGR-drug preparation and storage, and, potentially, also in vivo, after drug administration to animals or to patients. It is therefore clear that the poor stability of NGR represents an important concern in the development and manufacture of NGR-drugs, and also implies important pharmacological and toxicological implications.

There is need for providing modified NGR motifs which are chemically stable and, in particular, do not undergo asparagine-deamidation while preserving CD 13- targeting ability avoiding "off-target" effects.

Objects of the invention

It is an object of the invention to provide an NGR modified motif and novel tumor- homing peptides containing said motif which are able to prevent asparagine- deamidation while preserving CD 13 -targeting ability.

It is another object of the invention to provide conjugation products comprising said motif and/or said peptides and their use in diagnostic and therapy.

Definitions

According to the present invention amino acids are generally referred to by their single letter code. Linear peptides are indicated by the sequence of their single letter code.

Cyclic peptide sequences (head-to-tail cyclized peptides) are indicated in braces and preceded by a "c" (c[ ]).

Amino acids, with the exception of "NGR" may be in D- or L-configuration.

The term "tumor(s)" herein indicates any type of cancer such as a solid tumor or any tumor expressing CD 13 in angiogenic tumor vasculature including, but not limited to, lung, liver, pancreas, breast, colon, larynx, gastro-intestinal, brain, skin renal, head and neck, bone and ovary cancer.

Even when not expressly indicated, the terms "diagnosis" or "diagnostic" include the imaging techniques.

Description of the invention

According to one of its aspects, a subject-matter of the present invention is an NGR motif wherein the hydrogen atom of glycine's nitrogen is substituted, i.e. a modified NGR motif of the following formula (I)

wherein Z is different from hydrogen, herein after also "NA¾GR" or "NA¾GR motif). According to a preferred embodiment, in the NGR motif, glycine is N-substituted with a substituent selected from an alkyl group, a hydroxy group and an alkoxy group.

According to a preferred embodiment, in the NGR motif, glycine is ^-alkylated (herein also "N AlkGR" or "N^AlkGR motif) of formula (II)

The term "Alk" or "Alkylated" herein indicates the residue of a lower alkyl, preferably a Ci-C4-alkyl, such as, methyl, ethyl, n-propyl, isopropyl or one of the butyl residues, preferably methyl and n-butyl.

The methyl residue is preferred and a preferred N^AlkGR motif is therefore the NA eGR motif of formula (III)

According to another of its aspect, a subject-matter of the invention is N/ zGR, N AikGR, preferably NA eGR, for use as a ligand of aminopeptidase N (EC 3.4.11.2), herein after "CD13".

According to another of its aspects, a subject-matter of the present invention is a peptide comprising an NA¾GR, N AlkGR motif (herein also "N AlkGR-containing peptide"), preferably the NA eGR motif.

According to a preferred embodiment, the N AlkR motif is bound to the other amino acids of the peptide by bonds with marked (*) atoms shown in the following formula

According to a preferred embodiment, the NA¾GR-, preferably the NwAikGR- containing peptide has the following formula (IV)

X- N^AikGR-X' (IV)

wherein X is selected from C, G, R, K, A, in "L" or "D" configuration and X' is selected from C, G, R K, A, in "L" or "D" configuration.

According to a preferred embodiment, the NAfAikGR-containing peptide is selected from c[CGNvMeGRG], c[CN^MeGRGG], c[CNiVMeGRG], c[CRGvMeGRGPD], CRGA eGRGPD, CN^MeGRC.

Preferred peptides in which the NGR sequence is modified according to the invention, are derived from the following proteins: fibrillin-2 precursor (FBN2 HUMAN), complement receptor type 1 (CR1 HUMAN); low-density lipoprotein receptor-related protein 2 (LRP2 HUMAN); Fibronectin precursor (FINC HUMAN), serine protease inhibitor Kazal-type 5 precursor (ISK5 HUMAN); low-density lipoprotein receptor-related protein 1 (LRP 1 HUMAN) ; Atrophin-1 -interacting protein 1 ( AIP 1 HUMAN); fibrillin-3 precursor (FBN3 HUMAN) ; WD-repeat protein 9 (WDR9 HUMAN); ancient ubiquitous protein 1 precursor (AUPI HUMAN); C4b-binding protein alpha chain precursor (C4BP HUMAN); signal transduction protein CBL-B (CBLB HUM AN) ; ceruloplasmin precursor (CERU HUMAN); Chromodomain helicase-DNA-binding protein 3 (CHD3 HUMAN); Complement receptor; type 2 precursor (CR2 HUM AN) ; Protocadherin Fat 2 precursor (FAT2 HUMAN); Galactoside 2- alpha-L-fucosyltransferase 1 (FUTI HUMAN); Integrin beta-1 precursor (ITB 1 HUMAN); Integrin beta-4 precursor (ITB4 HUMAN); Antigen KI- 67(KI67_HUMAN); Muscleblind-like X-linked protein (MBN3 HUMAN) ; Muscleblind-like protein (MBNL HUMAN); Metabotropic glutamate receptor 2 precursor(MGR2_HUMAN); Paired box protein Pax-6 (PAX6_HUMAN); Propionyl-CoA carboxylase beta chain, mitochondrial precursor (PCCB HUMAN); Tyrosine-protein phosphatase, non-receptor type 4 (PTN4 HUMAN); Tenascin precursor (TN) (TEN A HUM AN) ; Tenascin-N precursor (TN-N) (TENN_HUM AN) ; Vacuolar proton translocating ATPase (VPP2_HUMAN); NEDD4-like E3 ubiquitin-protein ligase WWP2 (WWP2_HUMAN); Cadherin EGF LAG seven-pass G-type receptor 1 precursor (CELR1 HUMAN); Cadherin EGF LAG seven-pass G-type receptor 3 precursor (CELR3 HUMAN); Low-density lipoprotein receptor-related protein IB (LRP1B HUMAN); Plexin Dl precursor (PLXD 1 HUMAN) ; DNA (Cytosine-5)-methyltransferase 1 (DNMT 1 HUMAN) ; Stabilin-2 precursor (FEEL-2 protein) (STAB2 HUMAN); Protocadherin gamma B l precursor (PCDGD HUMAN); Cadherin EGF LAG seven-pass G-type receptor 2 (CELR2 HUM AN) ; contactin 4 precursor (CNTN4_HUMAN); Fibroblast growth factor receptor 3 (FGFR3 HUMAN); very low-density lipoprotein receptor precursor (VLDLR HUMAN); laminin alpha- 1 chain precursor (LAMA 1 HUMAN) ; laminin alpha-5 chain precursor (LAMA5_HUMAN); polyadenylate-binding protein 3 (PABP3 HUMAN); latent transforming growth; factor-beta-binding protein 2 (LTBP2 HUMAN); lysosomal alpha-mannosidase precursor (M A2B 1 HUMAN) ; protocadherin 19 precursor (PCD19_HUMAN); Tyrosine-protein phosphatase, non-receptor type 21 (PTN21_HUMAN); receptor- type tyrosine-protein phosphatase beta precursor (PTPRB HUMAN); Receptor-type tyrosine-protein phosphatase F precursor(PTPRF HUMAN); SH3 -domain kinase binding protein 1 (SH3K1_HUMAN); Nesprin-1 (Nuclear envelope spectrin repeat protein 1) (SYNEI HUMAN); Zinc finger and BTB domain Containing protein 10 (ZBTIO HUMAN); Fibrinogen gamma chain (FIBG HUMAN); Fibrinogen beta (FIBB HUM AN) .

According to a preferred embodiment, the length of the peptide is 5 to 10 amino acids, for instance 5-6 amino acids.

A particularly preferred peptide, according to the invention, is the NA eGR-containing peptide c[CGNA eGRG], which is herein also referred to as "MeNl", while the corresponding non-A^-methylated peptide, i.e. c[CGNGRG], is here referred to also as "Nl".

According to a preferred embodiment, a subject-matter of the invention is a N/ zGR- containing peptide, preferably a N AikGR-containing peptide or a NA eGR- containing peptide, for its use in targeted-delivery of therapeutic and/or imaging and/or diagnostic moiety to tumors.

Indeed, the N/ zGR-containing peptide, preferably the N AikGR-containing peptide or the NA eGR-containing peptide, may be conjugated with a therapeutic and/or imaging and/or diagnostic moiety, thus forming a "conjugation product".

The conjugation product as above defined constitutes another subject-matter of the invention, as well as its use in the diagnosis and/or imaging and/or treatment of tumors.

Preferred therapeutic and/or imaging and/or diagnostic moieties are preferably selected from chemotherapeutic drugs, such as doxorubicin, melphalan, cis-platin, gemtabicine, taxol, auristatins or maytansines derivatives; antibodies; kinase inhibitors; liposomes; nanoparticles, such as nanogold; anti-angiogenic compounds; DNA complexes; viral particles; imaging compounds; and cytokines, such as interferons such as (IFN)-y, IFN-a-2a IL12, TRAIL and tumor necrosis factors such as TNF-a. Preferred moieties are T F-bearing gold nanoparticles and liposomal doxorubicin.

Preferred conjugation products comprise T F-α and doxorubicin conjugates, such as liposomal doxorubicine.

Other preferred conjugates comprise NOTA-conjugates, preferably radiolabeled, for instance with ¹⁸F, wherein NOTA represents 1,4,7- triazacyclononane-l,4,7-triacetic acid.

All the above moieties are known per se in the art and the skilled person is perfectly able to manufacture said conjugation products starting from the novel N/ zGR-containing peptide, preferably the N AlkGR-containing peptide, most preferably the NA eGR-containing peptide.

The conjugation product of the invention may be prepared by directly linking the moieties as above defined to the peptide or through a spacer. Suitable spacer are known in the art, such as for instance a single amino acid, preferably glycine, an amino acid sequence or an organic residue. When cysteine is present in the peptide, the moiety may be linked by the free thiol group of the cysteine.

According to an embodiment, the spacer is MAL-dPEGn-lipoamide.

The peptide of the invention coupled to MAL-dPEGn-lipoamide represents a further subject matter of the invention. Particularly preferred are MeNl-dPEGn-lipoamide and Isol-dPEGii-lipoamide, as shown in Figure 28.

According to another of its aspects, a subject-matter of the invention is a N/ zGR-containing peptide of formula (I), preferably the N AlkGR-containing peptide of formula (II), most preferably the NA eGR-containing peptide of formula (III), for its use in the treatment and or diagnosis of tumors.

Pharmaceutical and diagnostic compositions comprising a N/ zGR-containing peptide of formula (I), preferably the N AikGR-containing peptide of formula (II), most preferably the NA eGR-containing peptide of formula (III), or a conjugation product as above defined, constitute another aspect of the present invention. Said compositions preferably comprise conventional pharmaceutically acceptable carriers and/or excipients and are preferably formulated as injectable solutions or suspensions, especially for intravenous infusion.

The compositions of the invention may be prepared according to the methods known in the art.

According to another of its aspects, a subject-matter of the invention is the use of a N/ zGR glycine or preferably a -alkyl glycine or, most preferably, A^-methyl glycine, for preventing asparagine deamidation in the NGR motif.

According to another of its aspects, the invention relates to a method for the treatment or for the diagnosis of a tumor, which comprises administering, to a mammal in need thereof, an effective amount of a NA¾GR- or NA^AikGR-, or a NAfAikGR-containing peptide or a conjugation product as above defined. The preferred embodiments above defined, also apply to the method of the invention. As it will be shown in the Experimental Section, in a surprisingly and unexpected way, a N/ zGR-containing peptide of formula (I), preferably the N AlkGR-containing peptide of formula (II), most preferably the NA eGR-containing peptide of formula (III), are able to inhibit the enzymatic activity of CD 13 but do not bind ανβ3 integrin, thus demonstrating that glycine Af-methylation in the NGR motif prevents asparagine- deamidation and z^'soDGR formation.

Ν/ zGR-, preferably NAfAikGR-conjugates, especially NA eGR-conjugates with radioactive or fluorescence compounds, microbubbles and nanoparticles can be prepared and constitute another subject-matter of the invention as well as their use for imaging of angiogenesis in tumors and other tissues.

Diagnostic kits comprising Ν/ zGR-, preferably NAfAlkGR-conjugates, especially NA eGR-conjugates, with radioactive or fluorescence compounds also constitute another subject-matter of the invention.

Brief Description of the Figures

Figure 1 : A) Characterization of anti-linker and anti-isoDGR antisera by ELISA. Binding of anti-linker {left panel) or anti-isoDGR {right panel) antiserum to microtiterplates coated with different peptides containing isoDGR (acetyl- CNGRCGVRS S SRTPD SKY), NGR (acetyl-CNGRCGVRS S SRTPD SKY) or control peptides containing the ARA or RGD sequence (acetyl- C ARAC GVRS S SRTPD SKY and acetyl- CRGDC GVRS S SRTPD SKY) . Antibody binding was detected using a peroxidase-labeled goat-anti-rabbit antibody. Anti- linker antiserum is specific for the VRS S SRTP SDK Y sequence (corresponding to human TNFi-n sequence, underlined) and does not discriminate among NGR, isoDGR or control peptides. In contrast anti-isoDGR antiserum is specific for isoDGR peptide. The low cross-reactivity of anti-isoDGR antiserum with the CNGRC peptide is likely related to the formation of little amounts of deamidated peptide during peptide synthesis and assay execution. B) Schematic representation of 7D1-, [NGR+isoDGR]- and [isoDGRj-ELISAs for peptide-antibody conjugate detection. Peptide-antibody conjugates consist of acetyl- CNGRCGVRS S SRTPD SKY or acetyl-CisoDGRCGVRSSSRTPDSKY peptides, chemically coupled to the anti-VSl mouse mAb 7D1 (NGR-Ab and isoDGR-Ab). The ELISAs are based on VS1 and anti-T Fi-n or anti-isoDGR rabbit antiserum in the capture and detection steps, respectively, followed by anti-rabbit Igs-peroxidase (HRP) conjugate. These assays allow quantification of [NGR+isoDGR]- or [isoDGR]-antibody conjugates, respectively. C) Dose-response curves of [7D1]-, [NGR+isoDGR]- and [isoDGRj-ELISAs with mAb 7D1 (Ab), NGR-Ab and isoDGR-Ab

Figure 2: NGR-to-isoDGR transition in vivo. An NGR peptide-antibody conjugate (NGR-Ab) was injected (i.v.) in mice and plasma samples, collected after 5 min and 4 h, were tested by ELISAs specific for NGR-Ab or isoDGR-Ab. *p<0.05, **p<0.01, two-tail t-test. Table: In vivo formation of isoDGR in NGR-antibody conjugate

Figure 3 : 2D representation of cyclic peptides containing NGR, NA eGR, isoDGR, or ARA motifs, embedded in different scaffolds. Aminoacids are represented with the single letter code; SH, free thiol group; isoD, isoaspartate. Peptide codes are reported within each structure.

Figure 4: Molecular mass of cyclic (head-to-tail and disulfide-bridged) and linear peptides as determined by Matrix- Assisted Laser Desorption/Ionization (MALDI) or ElectroSpray-Ionization (ESI) mass spectrometry (MS) analysis.

Figure 5: MS analysis (deconvoluted molecular weight spectrum) of TV-methylated and non-methylated NGR peptides before and after incubation at 37 °C in 0.1 M ammonium bicarbonate buffer, pH 8.5 for 16 or 42 h. The molecular weights of monomers and disulfide-bridged dimers, formed during incubation and analysis, are shown. The +2 Da observed for Nl after 16-42 h likely accounts for complete dimer deamidation.

Figure 6: A) MS analysis (deconvoluted molecular weight spectrum) of MeN3 peptide after incubation at 37 °C in 0.1 M ammonium bicarbonate buffer, pH 8.5 for 16 or 42 h. The molecular weights of monomers and disulfide-bridged dimers, formed during incubation and analysis, are shown. No deamidation occur even after 42 h of incubation. B) Binding of peptides to ανβ3 coated-microtiter plates before (0 h) and after (16 h) of incubation, as measured by competitive ELISA. C) Binding of N3 and MeN3 to ανβ3 coated-microtiter plates before (0 h) and after (16 h) incubation, as measured by competitive ELISA.

Figure 7: Effect of different peptides on CD 13 enzymatic activity. IC50, peptide concentration that inhibit 50% of activity; n, number of independent experiments (each in duplicate). A representative experiment is shown (mean ± SE). Table: Effects of NGR peptides on CD 13 enzymatic activity.

Figure 8: Steady state kinetic analysis of CD 13 in the presence of various amounts of substrate and peptides. Representative plots of initial velocity (Vo) versus substrate concentration (upper panels) and double reciprocal plot (lower panels). The inhibitory constant (Ki) values reported in each plot are the result of two independent experiments (mean ± SE). Table: Effects of NGR peptides on CD 13 enzymatic activity

Figure 9: A) MS analysis of Nl and MeNl (30 μΜ) after incubation with or without CD13. No changes in the molecular mass were observed, suggesting that both peptides are resistant to degradation by CD13. B) Nl and MeNl peptides were coupled to fluorescent nanoparticles (Qdot605, Invitrogen) as described¹. None-Qdot corresponds to nanoparticles activated with sulfo-SMCC, but lacking the peptide. HUVEC were cultured for 48 h on chamber slides (40000 cells/well). Cell culture medium was replaced with a solution containing fluorescent nanoparticles (6 nM) in 25 mM Hepes buffer, pH 7.4, containing 130 mM sodium chloride, 3% bovine serum albumin, 1 mM magnesium chloride and 1 mM manganese chloride. After 2 h of incubation at 37 °C, cells were extensively washed with the same buffer, fixed with 4% of paraformaldehyde, counterstained with DAPI and analyzed by fluorescence microscope. Magnification, 400X; bar, 20 μπι; red, Qdots; blue, nuclear staining with DAPI.

Figure 10: A) Upper panels: Superpositions of the crystallographic structure of CNGRCG/CD13 with the binding mode of N1-CD13 (left) and MeNl/CD13 (right). N1/CD13 and MeNl /CD 13 are shown with Nl and MeNl with blue and yellow sticks, respectively, and CD 13 in purple surface and cartoon, The CNGRCG peptide (complexed with CD13, pdb code: 4ou3) is shown with cyan sticks; Zn²⁺ is represented with a dotted sphere. Lower panels, CD 13 residues involved in interactions with peptides are highlighted as sticks and labeled with the three-letter code. Arrow, Af-methylglycine. B) Representative decoy poses of Nl {left panel) and MeNl {right panel) with CD13. CD 13 is shown as purple cartoon and surface. The Y477 of CD 13 interacting with G or NMeG is highlighted in sticks and labeled with the one-letter code. The arrow indicates the Af-methylation. Zn²⁺ is represented as a sphere; peptides are represented as sticks.

Figure 11 : Intermolecular interactions are marked with lines, the ligand is displayed with a 2D representation and protein residues, which have any atom within 3 A from any ligand atom, are represented as spheres and labeled with the three-letter code. Hydrogen bonds to protein backbone or protein side chains and pi-cation interactions are shown as solid pink, dotted pink and red lines, respectively. Ligand atoms that are exposed to solvent are marked with gray spheres. The protein pocket is displayed with a line around the ligand, colored with the color of the nearest protein residue. The gap in the line shows the opening of the pocket.

Figure 12: Summary of the interactions engaged by Nl and MeNl with CD13.

Figure 13 : A) Reverse-phase-HPLC analysis of N1-, MeNl- and Isol- NOTA conjugates (50 μg) on a C18 column. B) Mass spectrometry analysis (LTQ-XL Orbitrap) of N1-, MeNl- and Isol- NOTA conjugates. C) Schematic representation of each conjugate and expected monoisotopic masses.

Figure 14: A) Biodistribution of radiolabeled N1-, MeNl-. and Isol-NOTA¹⁸F in B16-F1 melanoma-bearing mice (3 h after administration). Cumulative results of 1 or 2 independent experiments (bars, mean ± SE; n=5-7 mice/group). B) Biodistribution (% injected dose (ID)/g of tissue) of radio-labeled peptides in B16 tumor-bearing mice (3 h after injection) (black bars). Unlabeled peptide-NOTA (at the indicated doses) was co-injected intravenously with each labeled peptide (white bars). Cumulative results of one or two independent experiments. Values are expressed as mean ± SE; n, number of mice. *, P<0.05; **, P<0.01 by two-tail t-test analysis. Figure 15: A) Reducing SDS-PAGE analysis of N1-, MeNl-. Activated-HSA (albumin activated with sulfo-SMCC but lacking the peptide) conjugates. B) MALDI-TOF mass spectra of MeNl -HSA. Activated-HSA and HSA. Dotted line corresponds to the molecular mass of HSA. The found molecular mass is indicated above the main peaks. Mass spectrometry was carried out using instrument a 4800 MALDI TOF/TOFTM instrument (AppliedBiosystems, CA, USA). The results suggest that MeNl-HSA is modified ~4 peptides/HSA molecule.

Figure 16: A) Fluorescence intensity of HSA tagged with Nl and MeNl-and labeled with IRDye680, as analyzed using the Odyssey scanner. "Activated-HSA/IRDye" corresponds to the negative control (lacking the peptide). B-C) Tumor-homing properties of Nl-HSA/, MeNl-HSA/, or activated-HSA/IRDye conjugates in the WEHI-164 fibrosarcoma model. Tumor-bearing mice were injected, i.v., with 30 μg of N1-, MeNl-, activated-HSA/IRDye. After 24 h, tumors, spleen, and kidneys were explanted and analyzed with an Odyssey CLx scanner to quantify the fluorescence uptake. Tumor fluorescence images (black-and-white) (B) and tissue fluorescence quantification (histograms) are shown (C). Each tumor was subdivided into small squared areas of equivalent size and fluorescence intensity (pixel) of each area was evaluated using the in-build software of the scanner. Bars represent the mean ± SE intensity values of the 15 areas with the highest intensity (mean maximum fluorescence); the number of mice/group is indicated on each bar. D) In vivo binding of Nl-HSA/, MeNl-HSA/. or Activated-HSA/IRDye to tumor cells, tumor endothelial cells and blood cells, as measured by FACS. The blood and the tumors of mice described in (B) were analyzed, after tissue disaggregation, by FACS using anti-CD31 and anti-CD45 antibodies. The percentage (mean ± SE) of IRDye680 positive cells gated on CD31 D45- (tumor cells), CD31⁺CD45⁺ (endothelial cells), CD45⁺ (blood cells) are shown. The number of mice/group is indicated on each bar. Figure 17: A representative spectrophotometric plot is shown. Solid line corresponds to uncoated 25-nm gold nanoparticles. Table: Characterization of TNF-bearing gold nanoparticles tagged with Nl-HSA or MeNl-HSA by cytotoxicity assay, UV- Visible spectroscopy (UV-Vis) and Dynamic Light Scattering (DLS).

Figure 18: Anti-tumor effects of TNF and TNF-coated nanoparticles (Au/TNF) tagged with Nl or MeNl, alone or in combination with MeNl-HSA, anti-CD 13 mAb (R3-63) or control mAb (IgG2a), in the WEHI-164 fibrosarcoma model. MeNl-HSA was co-administered with Nl/Au/TNF, whereas antibodies were given 2.5 h before the MeNl /Au/TNF. Effect of MeNl- and ARA-liposomal doxorubicin (-Lipo[Doxo]) on tumor growth, animal weight and survival in the B16-F1 melanoma model. Arrows, time of treatment; tumor volume (6 mice/group, mean ± SE). * (P < 0.05); ** (P < 0.01); *** (P< 0.001), by two-tail t-test (B) and log-rank (Mantel-Cox) test (G).

Figure 19: RESP charges. RESP atomic partial charges for the Af-methylated glycine are reported

Figure 20: MeNl peptide binds to CD13⁺ murine melanoma (B 16-F1) cells

A) Expression of CD13 and integrin subunits by B 16-F1 cells as evaluated by FACS analysis using anti-integrin monoclonal antibody followed by AlexaFluor 488- conjugated secondary polyclonal antibody. B) Binding of MeNl, Nl and Isol coupled to fluorescent Qdot605 nanoparticles (6 nM, 2 h at room temperature) to cultured B16-F1 (40.000 cell/well) as detected by fluorescence microscopy. None-Qdot corresponds to nanoparticles activated with sulfo-SMCC, but lacking the peptide. Magnification, 400X; bar, 20 μπι; red, Qdots; blue, nuclear staining with DAPI. C) and D) Binding of MeNl/QDot and None/Qdot to B16-F1 cells as detected by FASC analysis. Direct (C) and competitive binding (D) assays are shown. Directed binding assays were performed as follows: 5xl0⁵ cells were resuspended in a solution containing the indicated doses of MeNl/QDot or None/Qdot in Hepes-buffer containing 1% BSA, 1 mM MgCb, 1 mM MnCb (binding buffer) and incubated for 2 h at room temperature. Competitive binding assays were performed by mixing 50 nM of MeNl/QDot or None/Qdot with 200 μΜ of the indicated free peptides diluted in binding buffer. Cells were then washed, fixed, in 2% formaldehyde in PBS, and analyzed by FACS. MeNl. CGNVMeGRG; Nl, CGNGRG; N(r); CGNGrG, r; D- Arg.**, p < 0.01 t-test analysis.

Figure 21 : Binding of peptide-HSA/IRDye680 conjugates to human Kaposi's sarcoma cells (KS1767) and human breast adenocarcinoma cells (MCF-7).

A) Expression of CD 13 and ανβ3 by KS1767 and MCF-7 cells as evaluated by FACS analysis, using the indicated mouse monoclonal antibody followed by AlexaFluor 488-goat anti-mouse IgG polyclonal antibody. B) Binding of various amounts of the indicated peptide-HSA/IRDye680 conjugates to KS1767 (upper panels) and MCF-7 (lower panels) cells. Cells (40000 cells/well) were cultured for 48 h in 96-black wells plate, then peptide-HSA/IRDye680 conjugates were added to the plates and left to incubate for 1 h at 37°C, 5% C0₂. After washing, cells were fixed and the fluorescence uptake was measured using an Odyssey CLx (LI-COR) scanner. The data are showed as percentage of the binding of *HSA/IRDye680 (fluorescent-albumin lacking the peptide), where 100 % corresponds to the maximum binding measured and 0 % to the background. Cumulative results of 1 or 3 independent experiments (each in quadruplicate) are shown. Mean±SEM, n, replicates.

Figure 22. Effect of the R3-63 anti-CD 13 antibody on the tumor homing properties of MeNl -HSA/IRDye in the WEHI-164 fibrosarcoma model.

BALB/c mice bearing WEHI-164 tumors were injected with mAb R3-63 (IgG2a) or an isotype-matched control mAb (15 μg, i.p.) and 1.5 h later with MeNl -HSA/IRDye (6 μg). Twenty-four hour later, tumors, spleen, and kidneys were explanted, homogenized and analyzed with the Odyssey CLx scanner to quantify the fluorescence uptake. Bars, mean ± SE.

Figure 23 : Effect of TNF-gold nanoparticles loaded with MeNl peptide on tumor perfusion as determined by Contrast Enhanced Ultrasound (CEUS) imaging. WEHI- tumor-bearing mice (n = 10-11/group) were treated, i.v., with the indicated dose of MeNl/Au/TNF at day 5 after tumor implantation and analyzed by CEUS imaging at day 7-8. MicroMarker Contrast Agent was injected at time 0 and its uptake was recorded for 40 s. A), B) and D) Cumulative results of two independent experiments. A) Effect of MeNl/Au/TNF (5 pg) or vehicle on tumor growth before (0 day) and after (7 day) treatment as determined by caliper. B) Wash-in curves, mean ± SE; **, p<0.001, by Two-way Anova. C) Gray-scale tumor images (left) and color-coded peak enhancement (PE) maps of a representative experiment (n=5 mice/group). Red and blue areas correspond to high and low perfused tumor areas, respectively. Numbers, PE value. D) Quantitative analysis of tumor perfusion. Quantitative analysis was performed using VevoCQTM software. Perfusion parameters were calculated using a region of interest (ROI, green lines in panel C) corresponding to the entire tumor area. Perfusion index corresponds to the ratio Area Under Curve/Mean Transit Time. Box-plot with 5-95 percentile and median are shown; *, P < 0.05, by t-test. Note images displayed correspond to maximum tumor diameter (transversal or longitudinal width) depending on the animal position on the stage. Figure 24: Effect of T F-gold nanoparticles loaded with Isol peptide on tumor perfusion as determined by Contrast Enhanced Ultrasound (CEUS) imaging. WEHI- tumor-bearing mice (n = 6/group) were treated, i.v., with the indicated dose of Isol/Au/T F at day 5 after tumor implantation and analyzed by CEUS imaging at day 7-8. MicroMarker Contrast Agent was injected at time 0 and its uptake was recorded for 40 s. A. Effect of Isol/Au/TNF (5 pg) or vehicle on tumor growth before (0 day) and after (7 day) treatment as determined by caliper. B) Wash-in curves, mean±SE; ****, p<0.0001, Two-way Anova. C) Gray-scale tumor images (left) and color-coded peak enhancement (PE) maps of a representative experiment (n=5 mice/group). Red and blue areas correspond to high and low perfused tumor areas, respectively. Numbers, PE value. D) Quantitative analysis of tumor perfusion. Quantitative analysis was performed using VevoCQTM software. Perfusion parameters were calculated using a region of interest (ROI, green lines in panel C) corresponding to the entire tumor area. Perfusion index corresponds to the ratio Area Under Curve/Mean Transit Time. Box and Whisker plot (5-95 percentile and median); **, P < 0.01, *, P < 0.05, by t-test. Note images displayed correspond to maximum tumor diameter (transversal or longitudinal width) depending on the animal position on the stage.

Figure 25: FACS analysis of immune cell that infiltrate the WEHl-164 tumor after administration of TNF-gold nanoparticles loaded with Isol and MeNl peptide.

WEHI 164-tumor-bearing mice (n=6/group) were treated i.v. with 5 pg of Isol/Au/TNF or MeNl/Au/TNF at day 5 after tumor implantation. At 13-14 days tumors were recovered, cellularized and analyzed by FACS using specific antibodies that allow the cell-subtype discrimination. 7-AAD (7-amino-actinomycin D) was used to discriminate viable cells during flow cytometric analysis. Box and Whisker plot (5-95 percentile and median); **, P < 0.01 by t-test.

Figure 26: Schematic representation of Nl peptide and Nl peptide derivatives A) Nl peptide derivatives bearing different N-alkylated chain. B) Nl peptide derivatives bearing different substitutions. Aminoacids are represented with the single letter code; lowercase letters: D-aminoacids; SH, free thiol group. Peptide codes are reported within each structure.

Figure 27 - Table 2. Effects of MeNl derivatives and other control peptides on CD 13 enzymatic activity.

Figure 28 Schematic representation of head-to-tail cyclic peptides containing the NA eGR and isoDGR motifs (MeNl and Isol, A) and peptide conjugates with MAL- PEGii-lipoamide (MeN 1 -PEGi i -LP A, Isol-PEGn-LPA and Cys-PEGn-LPA, B). The monoisotopic molecular weight (expected and found) of each conjugate is shown.

Figure 29 Characterization of MeNl-LPA/Au/TNF nanoparticles prepared with different amounts of MeN 1 -PEGi i -LP A and TNF. Binding of MeNl-LPA/Au/TNF nanoparticles to microtiterplates (solid-phase) coated with (A) or without (B) anti- polyethyleneglycol antibody (anti-PEG mAb), as detected with anti-TNF antibodies (anti-mTNF pAb). MeNl-LPA/Au/TNF nanoparticles were prepared by adding the indicated amounts of TNF and MeN 1 -PEGi ι -LP A to 1 ml of gold solution (~1 U/ml at A520 nm). Bars, mean ±SE. Cumulative results of one or two independent preparation experiments, as indicated.

Figure 30 Characterization of Isol-LPA/Au/TNF nanoparticles prepared with different amounts of Isol-PEGn-LPA and TNF. Binding of Isol-LPA/Au/TNF nanoparticles to microtiterplates (solid-phase) coated with (A) or without (B) an anti- polyethyleneglycol antibody (anti-PEG mAb), as detected with anti-TNF antibodies (anti-mTNF pAb). Binding of Isol-LPA/Au/TNF and Isol-HSA/Au/TNF nanoparticles to microtiterplates (solid-phase) coated with ανβ3 (C and D), as detected with anti-TNF antibodies (anti-mTNF pAb). Isol-LPA/Au/TNF nanoparticles were prepared by adding the indicated amounts of TNF and Isol- PEGn-LPA to 1 ml of gold solution (~1 U/ml at A520 nm). Isol-HSA/Au/TNF was prepared as described previously ^{5 41}. Bars, mean ±SE. Cumulative results of one to three independent preparation experiments as indicated.

Figure 31 Characterization of Cys-LPA/Au/TNF nanoparticles prepared with different amounts of Cys-PEGn-LPA and TNF. Binding of Cys-LPA/Au/TNF nanoparticles to microtiterplates (solid-phase) coated with anti-PEG mAb and detected with anti-TNF antibodies (anti-mTNF pAb). Isol-LPA/Au/TNF nanoparticles were prepared by adding the indicated amounts of TNF and Cys- PEGn-LPA to 1 ml of gold solution.

Figure 32 Characterization of Isol-LPA/Au and Isol-LPA/Au/TNF nanoparticles by UV-VIS spectrophotometry. UV-VIS absorption spectra of naked gold nanoparticles (upper and lower panels, dashed lines), Isol-LPA/Au (upper panels, solid lines) and Isol-LPA/Au/TNF (lower panels, solid lines) prepared by adding to 1 ml of gold solution the amounts of Isol-PEGn-LPA and TNF indicated in each panel.

Experimental Section

The invention is herein below disclosed by way of examples which in no way limit the scope of the protection of the present invention.

The technical methods disclosed are conventional and well known to the skilled in the art. anyway, specific literature references will be given.

Experimental Assays

NGR-to-isoDGR transition

To assess whether the NGR-to-isoDGR transition can indeed occur in vivo, the inventors injected in mice a conjugate consisting of a peptide encompassing the CNGRCG and TNFi-n sequences coupled to an anti-vasostatin-1 monoclonal antibody. The plasma levels of this peptide-antibody conjugate were then analyzed at different time points using two ELISAs based on the use of vasostatin-1 in the capture step, and antibodies against the TNFi-n sequence (thus unable to discriminate between NGR and isoDGR), or antibodies specific for isoDGR, in the detection steps (Figure 1). A marked reduction of NGR-antibody conjugate (NGR- Ab) occurred 4 h after injection (Figure 2, left panel), likely consequent to peptide proteolysis and antibody clearance. However, a marked increase of isoDGR-antibody conjugate (isoDGR- Ab) was also observed (Figure 2, right panel and Table), indicating that the NGR-to-isoDGR transition can indeed occur in vivo.

In the attempt to develop a more stable peptide, the inventors hypothesized that glycine Af-methylation in NGR (Figure 3) can prevent asparagine deamidation and, consequently, improve peptide stability. To test this hypothesis, the inventors synthetized various cyclic peptides containing NGR and their Af-methylated derivatives (NA eGR) embedded in different scaffolds (see Figure 3 and Figure 4). Mass spectrometry (MS) analysis of the head-to-tail cyclized peptide c[CGNGRG] (called Nl) before and after incubation for 16 or 42 h in 0.1 ammonium bicarbonate, pH 8.5 (a condition known to favor NGR deamidation), showed a gain of 1 Da, pointing to complete deamidation (Figure 5, upper panel, left). In contrast, and surprisingly, no change of molecular mass was observed with the methylated peptide c[CGNA eGRG] (called MeNl) (Figure 5, upper panel, right). Glycine -methylation blocked deamidation also in other NGR peptides, such as CNGRC (disulfide- bridged) (Figure 5, bottom panel) and c[CNGRGG] (head-to-tail cyclized, called N3) (Figure 6A). Integrin binding assays showed that Nl, N3 and CNGRC, but not MeNl, MeN3 and CA eGRC, could bind ανβ3 integrin after 16 h of incubation (Figure 6B and Figure 6C). Thus, glycine Af-methylation in NGR peptides completely prevents asparagine deamidation and isoDGR formation.

Glycine Af-methylation and CD 13 recognition

To assess whether glycine Af-methylation affects CD 13 recognition the inventors performed enzyme inhibition assays with methylated and non-methylated peptides. Both Nl and MeNl could inhibit CD 13 (aminopeptidase N) enzymatic activity (Figure 7). Steady-state enzyme kinetic analyses (Figure 8) showed that these peptides could inhibit the enzymatic activity of CD 13 with similar potency (Nl, Ki = 12.4 ± 5.4 μΜ; MeNl. Ki = 8.9 ± 1.8 μΜ). Interestingly, CNGRC was about 15-fold less potent than MeNl (CNGRC. Ki = 134 ± 28 μΜ). Nl, MeNl. and CNGRC affected the Michaelis constant (Km), but not the maximum velocity (Vmax) (Figure 8, bottom panels). These data are consistent with the competitive inhibition model, suggesting that Nl, MeNl and CNGRC can bind within, or close to, the active site of the enzyme. No change in the molecular mass of Nl and MeNl were observed upon 40 min incubation with CD 13 (Figure 9 A), indicating that no peptide cleavage occurred and, therefore, that these peptides were not CD13 substrates. Finally, both Nl and MeNl. coupled to Qdot fluorescent nanoparticles, could bind CD13⁺ endothelial cells (HUVEC) with similar potencies (Figure 9B). These data, overall, suggest that the methyl-group of MeNl does not impair the recognition of CD 13, both in solution and on endothelial cells.

To provide a structural rationale for this hypothesis, Nl and MeNl were docked into the CD13 crystallographic structure (PDB code: 4fyr)², as described in "Material and Methods" Section. Both peptides could adopt a binding mode reminiscent of that displayed by CNGRCG in complex with CD 13 (PDB code: 4ou3)³ (Figure 10). Analysis of the docking poses of MeNl and Nl revealed that the asparagine is involved in the coordination of the zinc ion and in stable interactions with the GXMEN catalytic motif of CD13. Moreover, the cysteine and the arginine establish polar interactions with the enzyme IV domain. Finally, the glycine and its methylated form interact with Y477 of CD 13, an important residue involved in the catalytic transition state⁴, through polar and hydrophobic interactions, respectively. No other significant contacts were detected with residues of the large internal cavity of the enzyme. These results suggest that both peptides can accommodate into the enzyme cavity and that Af-methylation is well tolerated (Figures 10, 11 and 12).

Tumor-homing properties of Nl and MeNl

To compare the tumor-homing properties of Nl and MeNl the inventors prepared ¹⁸F-radiolabeled Nl- and MeNl -NOTA conjugates (Figure 13) and analyzed their biodistribution in B16-F1 melanoma-bearing mice, 3 h after injection. In parallel, to assess the contribution of deamidated products, the inventors performed a similar study with c[CGisoDGRC] (called Isol, see Figure 3 and Figure 13). The results showed 4-fold accumulation of MeNl -NOTA¹⁸F in tumors compared to blood and other tissues (Figure 14A). Similar accumulation was observed with Nl- and Isol- NOTA¹⁸F. However, in these cases a certain degree of accumulation was observed also in other organs (e.g. stomach, spleen, liver, intestine, femur), pointing to a lower specificity of Nl and Isol . Co-administration of labeled peptides with an excess of the corresponding free peptide-NOTA showed a significant inhibition of tumor uptake in the case of MeNl and Isol, but not in the case of Nl (Figure 14B). One possible explanation for the lack of inhibition in the latter case is that a small amount of Isol was formed in Nl-NOTA¹⁸F, which could not be competed by free Nl . These results suggest that the tumor-homing properties of MeNl are similar to those of Nl in terms of tumor accumulation and even better in terms of tumor specificity. According to this view, other in vivo assays performed with fluorescence-labeled albumin (HSA) and peptide-albumin conjugates (Figure 15 and 16 A) showed that Nl and MeNl could increase the accumulation of albumin in tumors, but not in spleen and kidney, to a similar extent (Figure 16B and 16C). Interestingly, FACS analysis of cells obtained from these tumors showed that Nl and MeNl increased albumin accumulation on tumor and endothelial cells, but not on blood cells (Figure 16D). These data further support the concept that glycine methylation in Nl does not abrogate its capability to target tumor and endothelial cells.

Potential of MeNl to deliver drug-loaded nanoparticles

The inventors then investigated the potential of MeNl to deliver drug-loaded nanoparticles, such as TNF-bearing gold nanoparticles and liposomal doxorubicin, to tumors. The inventors have shown previously that the therapeutic properties of TNF- bearing gold nanoparticles (Au/TNF) can be enhanced by coupling with Nl (Nl/Au/TNF)⁵. Notably, systemic administration of Nl/Au/TNF bearing 5 pg of bioactive TNF are sufficient to induce anti-tumor effects in the WEHI-164 fibrosarcoma model, whereas 5 pg of non-targeted Au/TNF or free TNF were almost totally inactive⁵. To assess whether also MeNl can be exploited for nanodrug delivery to tumors the inventors prepared Nl/ and MeNl /Au/TNF. and analyzed their physicochemical properties, their cytotoxic activity in vitro and their anti-tumor activity in vivo. In vitro studies showed that these products were characterized by similar physicochemical and biological properties (Figure 17). In vivo studies showed that doses equivalent to 5 pg of TNF of both nanodrugs induced similar effects in the WEHI-164 model (Figure 18, upper panels). Interestingly, the antitumor activity of Nl/Au/TNF was completely inhibited by an excess of MeNl-HSA (Figure 18, middle panel, left), suggesting that Nl and MeNl can bind the same receptor in vivo, likely CD 13, and that NGR deamidation (and integrin targeting) was not necessary for activity. Furthermore, the anti-tumor activity of MeNl /Au/TNF was significantly inhibited by the anti-CD 13 mAb R3-63 (Figure 18, middle panel, right), supporting the hypothesis that CD 13 is an important receptor of MeNl in vivo.

Next, the inventors compared the anti-tumor activity MeNl -tagged liposomal doxorubicin (MeNl-Lipo[doxo], 60 μg, three times) with that of liposomal doxorubicin tagged with a negative control peptide (ARA-Lipo[doxo]) in the B 16-F 1 melanoma model. Treatment with MeNJ_.-Lipo[doxo] significantly prolonged the survival of mice compared to ARA-Lipo[doxo], with no evidence of increased toxicity, as judged from animal loss of body weight (Figure 18, lower panel). These results suggest that MeNl can be exploited as vehicle for delivering T F-bearing gold nanoparticles and liposomal doxorubicin to tumors, thereby enhancing their anti-tumor activity.

Materials and Methods

Cell lines and reagents

Murine B16-F1 melanoma and WEFQ-164 fibrosarcoma cells were cultured in DMEM supplemented with 10% fetal bovine serum, 2 mM glutamine, 50 μg/ml streptomycin, 100 U/ml penicillin and 0.25 μg/ml amphotericin-B. Sulfosuccinimidyl-4-[N-maleimidomethyl]cyclohexane-l-carboxylate (sulfo-SMCC) was from PIERCE (Rockford, IL); human serum albumin (HSA) was from Baxter (Deerfield, IL); bovine serum albumin (BSA) was from Sigma. Murine tumor necrosis factor-alpha (TNF) was prepared as described previously ⁶. Human ανβ3 integrin was from Immunological Sciences (Rome, Italy). Mab R3-63 (anti-mouse CD13, rat IgG2a) was from Acris (Herford, Germany); rat IgG2a isotype control antibody was from BioLegend (San Diego, CA). Anti-vasostatin-1 (VS1) mouse monoclonal antibody (mAb) 7D1 was described previously ⁷ . Polyclonal antiserum anti-CisoDGRC was raised in rabbit by immunization with acetylated-CisoDGRCK coupled to Keyhole Limpet Hemocyanin (PRIMM, Milan, Italy) and characterized by ELISA. The results show that anti-CisoDGRC antiserum can recognize CisoDGRCG and little or not CNGRCG (Figure 1A). Rabbit polyclonal antiserum anti-human T Fi-n (called anti-linker) was prepared as described ⁸.

Peptide synthesis

CNGRC and H₂-terminal acetylated peptides CNGRCG, CisoDGRCG, CRGDCG, CARACG fused to the VRSSSRTPDSKY sequence (human TNFi-n sequence, underlined) were prepared by chemical synthesis as previously described ^{9 10}. Head- to-tail cyclized peptides (Figure 3) and NGRAHA were prepared by the solid-phase Fmoc method ¹¹ All peptides were dissolved in water and stored in aliquots at -20 °C. The molecular mass of each peptide was checked by MALDI-TOF or ESI-MS mass spectrometry (Table 4) using a LTQ-XL Orbitrap (Thermo Scientific, Bremen, Germany) or a Voyager-DE STR MALDI-TOF MS (Sciex) instruments. Thiol- containing peptides were quantified using the Ellman's assay.

Peptide-antibody conjugate preparation and detection

NGR- and isoDGR-antibody conjugates (called NGR-Ab and isoDGR-Ab, respectively) were prepared by coupling acetyl-CNGRCGVRSSSRTPDSKY and acetyl-CisoDGRCGVRSSSRTPDSKY to mAb 7D1 using the bifunctional cross- linker bis[sulfosuccinimidyl] suberate (BS3, Pierce). These conjugates were characterized using 3 different ELISAs, called: [7D1]-, [NGR+isoDGR]- and [isoDGR]-ELISA, which are based on the use of VSl in the capture step and 3 different antisera in the detection steps (anti-mouse-Igs-HRP, or anti-T Fi-n linker, or anti-CisoDGRC antisera, see Figure IB for a schematic representation of the assays). 7D1-[ELISA] (Figure IB and 1C, left panel) was used for total antibody quantification in plasma samples, whereas the [NGR+isoDGR]- and [isoDGRj-ELISAs (Figure IB and 1C, middle and right panel) were used for the quantification of specific peptide-antibody conjugates.

To assess the NGR deamidation reaction in vivo NGR-Ab was intravenously injected in BALB/c mice (150 μg/mouse, 3 mice) and left to circulate for 5 min or 4 h. Plasma samples were then collected and the levels of peptide-antibody conjugates were detected by using the [NGR+isoDGR]- and [isoDGR]-ELISAs, as follows: microtiterplates were coated with vasostatin-1 (VSl) (10 μg/ml in 50 mM sodium phosphate buffer, pH 7.4, containing 150 mM sodium chloride (PBS, 50 μΐ/well) and incubated overnight. After washing with PBS, the plates were blocked with 3% BSA in PBS (200 μΐ/well, 2 h at r.t.) and washed again. Each well was filled with plasma samples (1 : 10 in PBS containing 0.5% BSA, 0.05% Tween-20, 2 mg/ml EDTA, 1 :500 protease inhibitor cocktail (Set-Ill, Calbiochem) (50 μΐ/well, 1 h at r.t.). After washing with PBS containing 0.05% Tween 20 (PBS-T), the wells were incubated with anti-TNFi-n linker ([NGR+isoDGR]-ELISA) or anti-CisoDGRC ([isoDGR]- ELISA) antisera (1 : 1000 in PBS containing 0.5% BSA (PBS-BT), 50 μΐ/well, 1 h at r.t.). The plates were washed again with PBS-T and incubated with peroxidase- labeled secondary antibody (1 : 1000 in PBS-BT, 50 μΐ/well, 1 h at r.t.). After washing with PBS-T, bound peroxidase was detected by adding o-phenylendiamine chromogenic substrate and measuring the absorbance at 490 nm using an ELISA plate reader.

CD 13 enzymatic activity assay

The effect of peptides on CD 13 enzymatic activity was carried out as described previously ⁵. The IC50 was calculated by non-linear regression analysis of competitive inhibition data using the GraphPad Prism Software (GraphPad Software, Version 6.00 San Diego, California, USA). Steady-state enzyme kinetic analyses were carried out at room temperature (23 °C) in 96-well, using increasing concentration of L-alanine-p-nitroanilide (0.1-1 mM), and fixed concentrations of MeNl. Nl and CNGRGC ranging from 0 to 1000 μΜ. The initial velocities (Vo) were calculated from the slopes of the first 5 min of the reaction. The enzyme inhibitory constants (Ki) were calculated using the competitive enzyme-inhibition model of the GraphPad Prism Software.

Binding of peptides to ανβ3 integrin by competitive ELISA.

Binding of peptides, before and after forced degradation at 37 °C in ammonium bicarbonate buffer, was measured by using a competitive binding assay based on ανβ3 integrin coated plates and a isoDGR-HRP conjugate peptide, as described ¹. Molecular docking

Af-methylglycine parameterization

The atomic partial charges of Af-methylated glycine atoms have been derived using the R.E.D. Ill (RESP ESP charge Derive) package ¹¹. Specifically, two different conformations of Af-methylglycine dipeptide (i.e. the Af-methylated glycine capped with the acetyl and Af-methyl groups, respectively) have been generated with Maestro Software package ¹³. Geometry optimization in gas phase and the computation of the molecular electrostatic potentials (MEP) have been performed with GAMES S package at HF/6-31G* level of theory with the Connolly surface algorithm ¹⁴. Additionally, four different molecular orientations for each optimized geometry have been considered using the rigid-body reorientation algorithm implemented in the R.E.D. tool. Finally, the two-stages RESP method was used for the fitting of the atomic charges following the procedure originally published by Kollman et al. ¹⁵, where intra-molecular constraints were imposed to set the charges of the acetyl and Af-methyl capping groups to zero. Charges equivalences were imposed to hydrogens of methyl and methylene groups. The derived RESP partial charges for N- methylglycine atoms are reported in Figure 19. The topology files to perform MD simulations with the AMBERff99sb-ildn force field ¹⁶ were generated using the Antechamber tool ¹¹.

Simulation set-up

Since conformations of cyclic peptides are often poorly and/or unrealistically sampled by algorithms commonly implemented in docking software ^{18 19}, we decided to exploit an enhanced sampling technique, named Bias-Exchange Metadynamics (BE-META) ²⁰'²¹, to better account for their intrinsic conformational heterogeneity. In particular, initial conformations for the cyclic head-to-tail hexapeptides c[CGNGRG] (Nl) and c[CGNA eGRG] (MeNl) have been generated and energy minimized exploiting the Maestro software package. A second minimization run was performed using Gromacs. Each system was then solvated in a cubic box with explicit TIP3P water molecules extending it at least 12 A from the peptide to the edges ²². The net charge of the peptides was neutralized with chloride ions. The initial energy minimization was followed by a three-stages equilibration procedure consisting in: i) 2 ns equilibration in the NVT ensemble, where the v-rescale thermostat ² was used to maintain temperature at 300 K; ii) 2 ns in the NPT ensemble, where Berendsen barostat ²⁴, and v-rescale thermostat were employed to control pressure and temperature (1 bar and 300 K); iii) 2 ns in the NPT ensemble, where the Berendsen barostat was replaced by the Parrinello Rahman one ²⁵. Positional restraints on the peptides' heavy atoms were employed in the first two steps of equilibration and were then released in the last stage. During the equilibration the LINCS algorithm ²⁶ was applied to constraint all the bond lengths, the md integrator was used with a time step of 2 fs. A cut-off of 1.0 nm was used to truncate both the van der Waals and the electrostatic interactions; long range electrostatic interactions beyond the cut-off were treated with Particle-Mesh Ewald method (Fourier grid spacing of 0.12 nm and interpolation order of 4) ^27,28. BE-META simulations were performed in the NVT ensemble, setting the temperature at 300 K by means of the v-rescale thermostat (coupling time constant of 100 fs) and using the same settings described for equilibration. During this production run all the bond lengths involving hydrogen atoms have been constrained. Gaussian hills were added every 1 ps (height: 0.4 kJ mol-1, width: 0.2 rad) and exchanges between pairs of replica were attempted every 15 ps. Herein an acceptance rate between 17% and 26% was observed. Molecular Dynamics simulations have been performed using the Gromacs-5.0.4 software package (www.gromacs.org) ²⁹ and the Plumed 2.1.3 plugin ⁰ using Amber ff99SB- ILDN force field ¹⁶. For each peptide a twelve-replicas bias-exchange simulation was performed (for a total of 1440 ns), imposing all the φ and ψ dihedral angles describing the cycle as collective variables, resulting in the bias of 12 CVs. Convergence was checked monitoring the mono-dimensional bias potential of each replica. Specifically, bias potential profiles obtained by averaging on the two halves of the simulations after an equilibration time of 1 ns were compared. In all the cases the energy-difference between the two halves was lower than 2kBT, demonstrating that the potential was stable and that the simulation reached convergence. The analysis and the reconstruction of the free energy landscape were performed using the Metagui tool ³¹, in a six-dimensional sub space defined by the six ψ dihedral angles, where it was possible to identify the main minima and to evaluate their population. The criteria used to select these CVs were: i) the microstates had to be characterized by low internal RMSD between conformations; ii) all the low-energy microstates (i.e. the ones with energy lower than 2kBT) had to be connected. Structures belonging to minima populated more than 10% (for a total of 2289 structures) were extracted and used as input for docking calculations.

Docking Calculations

The human receptor structure of CD 13 in complex with bestatin (pdb code 4fyr, res 1.91 A) ² was prepared exploiting Protein Preparation Wizard, available in Maestro Software Package: hydrogens were added, the orientation of the hydroxyl groups of Ser, Thr and Tyr, the side chains of Asn and Gin residues were optimized, the protonation states were chosen according to neutral pH and a minimization employing an opls 2005 force field ^[2f>1 with a Root Mean Square Deviation (RMSD) tolerance on heavy atom of 0.3 A was performed. Bestatin was used to determine the grid center and subsequently removed. The default cubic grid dimensions as defined in Glide was used and no constrains were applied. As input peptides structures, conform ers extracted from BE-META minima populated more than 10% were used. Since an exhaustive conformational sampling was performed during the BE-META simulations, neither minimization nor conformational search of the single peptides were performed. All the ligand-receptor complexes were generated by using the extra-precision mode of Glide (XP) considering a flexible ligand (i.e., flexible side chains), and saving at most five possible poses for each ligand input structure. The obtained decoy poses were then clustered based on geometric criteria, exploiting the clustering algorithm described by Daura et al ². The RMSD was computed on all the heavy atoms of the peptides. For the clustering an RMSD cutoff of 3 A was employed and clusters containing less than 10 poses were excluded. The final docking poses were selected based on their geometric similarity with CNGRCG in complex with CD 13 (pdb code 4ou3) (Figure 10). 2D-ligand interaction diagrams, generated with Maestro using a cutoff of 3 A, are shown in Figure 11. Interactions details are reported in Figure 19, Table. Molecular graphics were produced with Pymol (PyMOL Molecular Graphics System, Version 1.7 Schrodinger, LLC).

Preparation of MeNl-. Nl- and isol-NOTA conjugates.

MeNl. Nl and Isol were coupled to maleimide-NOTA (CheMatech, Dijon, France) as follows: maleimide-NOTA (33.3 mg in 3.33 ml of 50 mM sodium phosphate, pH 8) was mixed with peptide solution (10 mg in 9.67 ml of 10 mM phosphate buffer, pH 7.4, containing 138 mM sodium chloride, 2.7 mM potassium chloride, 5 mM EDTA) and left to react for 3 h at r. . The mixtures were acidified with 50% (vol/vol) orthophosphoric acid (50 μΐ/conjugate) and purified by semi-preparative reverse-phase UPLC using a C18 column (LUNA, 250 xlO mm, 10 μπι, Phenomenex), as follows: mobile phase A, 0.1% trifluoroacetic acid (TFA) in water; mobile phase B, 0.1% TFA in 95% acetonitrile; 0% B (7 min), linear gradient 5-10% B (40 min), 100% B (7 min), 0% B (7 min); flow rate, 5 ml/min. The products were lyophilized and analyzed by reverse-phase UPLC using a CI 8 column (LUNA, 250 x 4.6 mm, 5 μπι; Phenomenex) as follows: 0% B (7 min), linear gradient 5-15% B (40 min), 100% B (7 min), 0% B (7 min), flow rate, 0.5 ml/min. Product identity was assessed by mass spectrometry analysis (LTQ-XL Orbitrap). ¹⁸F radiolabeling of peptide-NOTA conjugates and biodistribution studies

MeNl-, Nl- and Isol-NOTA conjugates were radiolabeled with ¹⁸F as previously described ³. Briefly, 2-4 GBq ¹⁸F-sodium fluoride, prepared using a CTI RDS-111 Eclipse Cyclotron (Siemens, Knoxville, TN), was loaded onto an anion-exchanger cartridge (Sep-Pak Accell Plus QMA Plus Light, Waters, Italy). The column was washed with 3 ml of water and eluted with 500 μΐ of 0.4 M potassium carbonate. The products, collected into a siliconized tube, were adjusted to pH 4 with metal-free acetic acid, mixed with 2 mM aluminum acetate, pH 4 (25 μΐ), peptide-NOTA conjugate (2 mg/ml in 0.1% TFA in water, 45 μΐ), pure ethanol (200 μΐ), and incubated at 100°C for 15 min. After cooling, the products were brought to 10 ml with deionized water, loaded onto a CI 8 cartridge (Sep-Pak Plus Waters), washed with 15 ml of water and eluted with 1 ml of acetonitrile. The products were dried at 40°C, under helium-flow, and finally rehydrated with PBS containing 100 μg/ml HSA. The chemical and radiochemical purity of each product was checked by analytical RP-HPLC using a CI 8 column (Chromolith Performance, 100 x 4.6, μπι, Merck, Italy) connected to an HPLC system equipped with a radiochemical counter using the following method: buffer A, 0.1% TFA in water; buffer B, 0.1% TFA in 90% acetonitrile; flow, 1 ml/min; linear gradient 0-18 min 15% B, 18-25 min 0% B. Radiochemical purity, >97%; specific activity, 2000 Ci/mmole.

Biodistribution studies

All procedures involving mice were conducted in conformity with institutional guidelines and Ethics Committee regulations of San Raffaele Hospital (Milan, Italy). B16-F1 tumor-bearing animals were injected with 3.1±0.7 MBq (i.v.) of peptide- NOTA¹⁸F conjugates before tumors reached a diameter of 1.0-1.5 cm (12-14 days after tumor implantation). After 3 h, mice were sacrificed and organs (blood, tumor, muscle, stomach, spleen, liver, lungs, kidneys, femur, skin, heart and intestine) were collected, washed with saline solution and weighted. The radioactivity in the tissues was measured using a gamma-counter (LKB Compugamma CS 1282).

Preparation and characterization of Nl- and MeNl-HSA conjugates

Peptide Nl and MeNl were coupled to human serum albumin (HSA). HSA was activated with the hetero-bifunctional cross-linker sulfosuccinimidyl-4-[N- maleimidomethyl]cyclohexane-l-carboxylate (sulfo-SMCC) (1 :9, HSA/sulfo-SMCC molar ratio), as described previously ⁵. Aliquots of activated-HSA were mixed with the peptides (1 :9, HSA/peptide molar ratio) and, after reaction, purified by gel- filtration chromatography. An aliquot of activated-HSA, blocked with cysteine, was also prepared and used as negative control. Protein concentration in each conjugate was determined spectrophotometrically by measuring the absorbance at 280 nm (extinction coefficient: 1 mg/ml = 0.54 units).

Reducing SDS-PAGE analysis of Nl- and MeNl -conjugates (called Nl- and MeNl- HSA) and of HSA activated with sulfo-SMCC, but lacking the peptide (called "Activated-HSA") showed similar electrophoretic mobility (Figure 15 A); mass spectrometry analysis suggested that 4 peptides/HSA molecule were coupled (Figure 15B).

Aliquots of peptide-HSA conjugates were dialyzed against 25 mM Hepes, pH 7.4, containing 150 mM sodium chloride and labeled with IRDye680 (LI-COR Biosciences, Lincoln, E), an infrared fluorescent dye, as described previously ⁵.

Fluorimetric analysis of the conjugates showed that they were equally modified with the dye (Figure 16A).

Quantification of peptide-HSA/IRDye680 conjugate up-take in tumor and tissue extracts

WEHI-164 tumor-bearing mice were injected with Nl-HSA/, MeNl -HSA/ or activated-HSA/ IRDye680 (i.v., 30 μg) or vehicle. After 24 h, tumor, spleen and kidney were collected and analyzed with an Odyssey CLx scanner to assess the uptake of fluorescence (settings: focus offset, 0 mm; laser intensity, 1.5; filter channel, 700 nm; resolution, 169 μπι).

Preparation of gold nanoparticles bearing peptide-HSA and TNF

Bifunctional gold nanoparticles bearing peptide-HSA, recombinant murine TNF and metoxi-PEG (20 KDa) were prepared essentially as described previously ⁵, except that the pH of colloidal gold (25 nm, Aurion, The Nederland) during the conjugation was 5.5 and the TNF/peptide-HSA ratio was 160 μg and 8 μg, respectively.

The physicochemical properties of each nanodrug were characterized by UV-visible spectrophotometry using an UltroSpec 2100 spectrophotometer (Amersham Biosciences) and 1 cm path-length quartz cuvette.

The number of bioactive TNF molecules loaded onto nanoparticle was determined by murine fibroblast LM cytotoxicity assay ⁶ (Figure 17, Table).

Preparation of liposomal-doxorubicin nanoparticles decorated with MeNl and ARA Liposomal-doxorubicin nanoparticles loaded with MeNl and ARA were prepared and characterized essentially as described ⁴. Liposomes had a diameter of 120 ± 10 nm, a polydispersity of 0.07 ± 0.01, and a Z-potential value of -30 ± 1 mV in water and - 2.6 ± 0.5 mV in PBS. The doxorubicin entrapment efficiency was 95 %.

Characterization of the pharmacological and toxicological properties of nanodrugs C57-BL/6J and BALB/c female mice (6-7 weeks old; Charles River Laboratories, Calco, Italy) were challenged with subcutaneous injections in the left flank of 2x105 B16-F1 cells or 1.5 x 106 WEHI-164 cells. Mice were injected i.v. with liposomal- doxorubicin diluted in 100 mM Hepes buffer pH 7.4 or gold nanoparticles in 0.9% sodium chloride containing 100 μg/mL HSA. Tumor growth was monitored by measuring tumor size with calipers. The tumor volume was estimated by calculating rl x r2 x r3 x 4/3π, where rl and r2 are the longitudinal and lateral radii, and r3 is the thickness of the tumor protruding from the surface of normal skin. Animals were sacrificed before tumors reached a diameter of 1.0-1.5 cm. Statistical analysis was performed by calculating the area under the curve for each tumor-bearing mouse using GraphPad Prism. Differences between calculated areas were evaluated by two- tail t-tests. * (P < 0.05); ** (P < 0.01); *** (P< 0.05).

MeNl/Qdot binds to murine melanoma B16-F1 cells

The binding properties MeNl, Nl and Isol peptide coupled to quantum dots (Qdot) were investigated using the B16-F1, a murine melanoma cell line expressing CD 13 and RGD-integrins, including ανβ3 (Figure 20A). Qdot lacking the peptide were also prepared in parallel and used as negative control. Fluorescence microscopy experiments showed binding of all peptide-conjugates to adherent B16-F1 cells, although with different extent, being the Isol/Qdot the most potent (Figure 20B). In contrast, little or no binding was observed with None/Qdot (Figure 20B). FACS analysis of non-adherent B16-F1 cells after incubation of MeNl/Qdot and None/Qdot confirmed that MeNl/Qdot could bind to these cells more efficiently than None/Qdot (Figure 20C). The binding of MeNl/Qdot, but not that of None/Qdot, was efficiently inhibited by an excess of free MeNl and Nl peptide (Figure 20D). Little or no inhibition was observed with N(r)l peptide, a poor ligand of CD13, used as negative control peptide. This data suggests that MeNl and Nl share the same receptor (i.e. CD 13) and glycine N-methylation does not prevent binding of MeNl .

Fluorescent MeNl-HSA. Nl-HSA. Isol-HSA but *HSA. bind to human Kaposi's sarcoma cells.

The capability of NGR peptide-albumin conjugates to recognize CD 13 was also investigated using the KS1767 cells (human Kaposi's sarcoma), previously shown to be capable of binding NGR, like endothelial cells ^{5 6}. The binding properties of Nl- HSA, MeNl-HSA and Isol-HSA conjugates (labeled with the fluorescent compound IRDye680) were investigated using these cells and MCF-7 cells (human breast adenocarcinoma cells, CD13 negative) (Figure 21A). Cell binding assays showed that Nl-HSA and MeNl-HSA could bind to KS1767 (CD13⁺) more efficiently than *HSA (activated albumin lacking the peptide) (Figure 2 IB, upper panels). In contrast, Nl-HSA, MeNl-HSA and *HSA could bind MCF-7 cells (CD 13^") similarly (Figure 2 IB, lower panels). Taken together these data confirm that MeNl-HSA and Nl-HSA bind to Kaposi's sarcoma cells via Nl or MeNl peptide. As excepted, Isol/HSA could bind both cell lines, although with different potency (Figure 21B, right panels), the binding for KS1767 cells being greater than that of MCF-7. The binding to MCF-7 by Isol-HSA is likely related to ανβ5 and α5β1 expressed by these cells.

The homing of MeNl-HSA to tumors is partially inhibited by an anti-CD 13 antibody We have demonstrated that fluorescent (IRDye680) MeNl-HSA and Nl-HSA home to the tumor more efficiently than fluorescent *HSA. To demonstrate that the accumulation in tumor is mediated by CD13, we performed in vivo CD 13 neutralization experiments using the antibody R3-63, an anti-murine CD 13 antibody previously used to neutralize the anti-tumor activity of NGR-TNF and NGR- IFNgamma ⁷'³⁸. To this aim, WEHI-164 tumor-bearing mice were pretreated with mAb R3-63 or with isotype matched control antibody (15 μg or 100 pmol/mouse). Antibodies were left to circulate for 1.5 h, and then the animals were administered with fluorescent MeNl-HSA (6 μg or 85 pmol/mouse). After 24 h, tumor, spleen and kidney were explanted, homogenized and organ associated fluorescence was quantified using a scanner. A reduced accumulation of MeNl-HSA in tumors (P=0.16), but not in control organs (spleen and kidney) was observed in mice pre- treated with the anti-CD 13 (Figure 22). Although, the inhibition is not statistically significant, this data suggests a role of CD 13 in the accumulation of MeNl-HSA. Another important factor that need to be taken into consideration is that these experiments were performed by administering equimolar amounts of reagents, a condition that may not lead to complete competition. Given the complexity of in vivo neutralization experiments that require high amount of competitor (i.e. anti-CD13), no further experiments were performed.

MeNl Au/TNF and Isol/Au/TNF reduce WEHI-164 tumor perfusion

To provide additional information on the antitumor mechanism of MeNl Au/TNF and

Isol/Au/TNF, we have studied the effect of these nanodrugs on tumor vasculature of tumor-bearing mice by contrast-enhanced ultrasound (CEUS) imaging technique. To this aim WEHI-164 tumor bearing animals were treated with 5 pg of each nondrugs and 48-72 h later were analyzed by CEUS. The results showed that both nanodrugs significantly reduced tumor perfusion (Figure 23A and 24B). In particular, quantitative analyses of CEUS data showed that both nanodrugs could significantly reduce the peak enhancement, wash-in-rate and perfusion index, i.e. parameters that quantify the blood volume. No effects on parameters that quantify the blood speed were observed (Figure 23D and 24D). These results suggest that the antitumor activity of both nanodrugs is presumably related to the vasculature damaging activity of TNF.

Isol/Au/TNF treatments promote infiltration of CD1 lb⁺ cells in WEHI-164 tumors To provide additional information of the mechanism of actions of nanodrugs, we have characterized the immune cell population infiltrating WEHI-164 tumor after 8-9 day from administration of Isol/Au/TNF or MeNl /Au/TNF (5 pg dose). The following cell sub-types were characterized by FACS: myeloid cells (CDl lb⁺); monocytes, neutrophils and eosinophils (CDl lb⁺, Gr-1⁺); macrophages (CDl lb⁺, F4/80⁺), T-lymphocytes (CD3⁺), CD4 and CD8 lymphocytes (CD3⁺, CD4⁺; CD3⁺, CD8⁺). The results showed that Isol/Au/TNF could significantly increase CDl lb⁺ cell population, (Figure 25). A trend in the increase of CD4+ cell population was also observed, although it was not significant (+ 30 %, P=0.13). Similarly, MeNl/Au/TNF showed a trend in the increase of CDl lb⁺ cells (+ 20 % respect to the untreated control animals). No difference in the change of infiltrate was observed for the other cell sub-types. These data suggest that the nanodrugs can stimulate the recruitment of CD1 lb⁺, which could play a role in the control of tumor growth. This observation is in line with previous published results showing that NGR- TNF (a recombinant protein with NGR peptide fused to TNF) could significantly increase the recruitment of immune system cells, such as CD8⁺ T lymphocytes after 2-72 h from NGR- TNF administration ⁹'⁴⁰.

The observations that nanodrugs increased CDl lb⁺ cells in tumors, but not CD8⁺ T lymphocytes, may be related to the fact that these experiments were conducted using different experimental settings (i.e. tumor model and time of analysis of infiltrates). Effects of MeNl derivatives on CD13 enzymatic activity

Molecular docking results obtained with MeNl and CD 13 suggest that the binding pocket of the CD 13 could in principle accommodate a MeNl derivative with a longer chain than the simple TV-methyl group. To verify this hypothesis, we synthetized 3 MeNl derivatives in which the 7V-methylglycine was replaced by an V-ethylglycine (called ethylNl), V-(«orwa/)propylglycine (called propylNl) or with an N- («orwa/)butylglycine (called butylNl) and tested their capability to bind CD 13 in a competitive CD 13 enzymatic assay. In parallel, also a new batch of MeNl was prepared to be used as positive control. Figure 26A shows the schematic representation of these peptides. The results of competitive binding to CD 13 showed that all the peptides could inhibit the enzymatic activity with similar potency, the IC₅o being comprised between 10 and 20 μΜ (2017 batch's), Figure 27.

These data are consistent with the original hypothesis that the binding pocked of CD 13 can accommodate MeNl derivatives with a longer chain (up to a butyl) without impairing CD 13 recognition.

In addition, the overall IC50 of MeNl (prepared by 2 different suppliers, 4 independent synthesis) was 42.5 ± 10 μΜ (mean±SE), suggesting that the protocol that we have developed for the preparation is robust and reproducible.

Non-deamidating NGR- and isoDGR-linker conjugates for direct coupling to TNF gold nanoparticles

Reagents

Human serum albumin (HSA) was from Baxter (Deerfield, IL). Murine tumor necrosis factor-alpha (TNF) was prepared as describe previously ⁶. Alphavbeta3 (ανβ3) integrin was from Immunological Science (Societa Italiana Chimici, Italy). Anti-polyethyleneglycol (PEG) antibody (anti-PEG mAb, clone 26A04) was from Abeam (Cambridge, UK).

Peptide synthesis and characterization

Cyclic head-to-tail c(C GNA eGRG) (called MeNl) and c(CGisoDGRG) (called Isol) were prepared and characterized as described ^{1 41} (see Figure 28). All peptides were dissolved in water and stored in aliquots at -80°C.

Assays

The alphavbeta3/anti-TNF antibody sandwich assays of nanoparticles coated with or without TNF and Isol were carried out using microtiterplates coated with 1 μg/ml ανβ3 in the capture step and a rabbit anti-TNF polyclonal antiserum (1 : 1000) in the detection step (followed by a polyclonal goat anti-rabbit-HRP conjugate), essentially as described ⁵.

The anti-PEG mAb/anti-TNF antibody sandwich assays of nanoparticles coated with or without TNF and Isol or MeNl was carried out using microtiterplates coated with 5 μg/ml anti-PEG mAb in the capture step and a rabbit anti-TNF polyclonal antiserum (1 : 1000) in the detection step (followed by a polyclonal goat anti-rabbit- HRP conjugate), essentially as described ⁵.

Coupling of MeNl to MAL-dPEGu-lipoamide.

MeNl was coupled to MAL-dPEGu-lipoamide (CAS #: 1334172-73-4, TM, Quanta Biodesign, Plain City, Ohio) using a ratio of 1 mol peptide / 1.2 mol MAL-dPEGn- lipoamide, as follows: 8 mg of MeNl in 50 μΐ of phosphate buffer, pH 7.8, was mixed with 50 μΐ of acetonitrile (final pH: ~7.0). The resulting solution was chilled on ice and mixed with 16 mg of MAL-dPEGu-lipoamide in 100 μΐ of 50% acetonitrile and left to react for 1 h under gentle shaking at 15°C. Control conjugates were also prepared in parallel using either Isol or a cysteine residue.

The products were loaded onto a Shimadzu Shimpack GWS CI 8 column (5 μπι, 4.6 mm x 150 mm) connected to Shimadzu Prominence HPLC. The column was eluted with mobile phase A (0.1% trifluoroacetic acid in water) and mobile phase B (70% acetonitrile, 0.1% trifluoroacetic) using the following chromatographic methods: 0% B (2 min), linear gradient (0-50% B, in 15 min), 100% B (4 min), 0% B (2 min); flow rate, 1 ml/min. The final products were lyophilized, resuspended in water and stored at -20°C. The identity of each conjugate (called MeNl-PEGn-LPA, Isol- PEGn-LPA and Cys-PEGn-LPA) was checked by mass spectrometry analysis using an LTQ-XL Orbitrap mass spectrometer.

Preparation of gold nanopar tides functionalized with MeNl-PEGn-LPA and TNF To prepare nanodrugs loaded with MeNl-PEGn-LPA and recombinant murine TNF we exploited a two-step procedure. First, 1 ml aliquots of colloidal gold (25 nm-gold nanoparticles, Aurion, The Netherlands, pH adjusted to 7.5 with sodium hydroxide) were mixed with different amounts of TNF (0, 8 or 16 μg) in 100 μΐ of 5 mM sodium phosphate buffer, pH 7.33. The mixtures (final pH: ~ 7.5) were incubated at room temperature for 30 min under shaking. Then, various amounts of MeNl-PEGn-LPA (ranging from 0.125 to 64 μg) in 100 μΐ of water were added to the mixtures to obtain various TNF and peptide concentrations (see Figure 29) and left to incubate at room temperature for 30 min. Finally, 100 μΐ aliquots of 0.5% HSA were added to the mixtures and left to incubate for 10 min at room temperature to saturate gold nanoparticles. The products were then centrifuged at 13000xg for 15 min. The pellets were resuspended in 5 mM sodium phosphate buffer, pH 7.33 containing 0.05% HSA {storage buffer). The centrifugation/washing steps were repeated twice. The final products (named MeNl-LPA/Au/TNF) were resuspended with 1 ml of storage buffer and stored at -80°C.

Preparation of gold nanoparticles functionalized with Isol-PEGn-LPA, or with Cys- PEGn-LPA and TNF

To prepare nanodrugs loaded with Isol-PEGn-LPA and TNF, we exploited a one- step procedure. In this case, aliquots of colloidal gold (prepared as describe above) were added to pre-mixed solutions containing different amounts of TNF (0, 8, 12 or 16 μg) in 50 μΐ of 5 mM sodium phosphate buffer, pH 7.33, and various amounts of Isol-PEGn-LPA (0.5 to 32 μg) in 50 μΐ of 5 mM sodium phosphate buffer, pH 7.33, to obtain different TNF and peptide concentrations (see Figure 30). The mixtures were then incubated for 30 min at room temperature, blocked with HSA, washed and centrifuged as described above for the two-step procedure. The final products were named Isol-LPA/Au/TNF. Control gold nanoparticles functionalized with Cys- PEGn-LPA and TNF (Cys-LPA/Au/TNF) were prepared accordingly to the one-step procedure using 1 ml gold aliquots, 16 μg of TNF in 50 μΐ of 5 mM sodium phosphate buffer, pH 7.33, and various amounts of Cys-PEGn-LPA (0.5 to 16 μg) in 50 μΐ of 5 mM sodium phosphate buffer, pH 7.33 (see Figure 31). The mixtures were then incubated for 30 min at room temperature, blocked with HSA, washed and centrifuged as described above for the one-step procedure. The final products were named Cys-LPA/Au/TNF.

RESULTS

Preparation of peptide -MAL-PEGi i-lipoamide conjugate

To have at hand peptides that could be directly attached to gold nanoparticles we have coupled MeNl or Isol to the MAL-PEGn-lipoamide cross-linking reagent. MAL-PEGii-lipoamide contains a maleimide group (MAL), which forms a thioether bond with the sulfhydryl group of the peptide, and a disulphide-containing molecule (lipoamide), which enables the conjugation to gold through the formation of dative bonds. In parallel, we also prepared a control ligand consisting of cysteine coupled to MAL-PEGii-lipoamide. The molecular weights of the purified peptide-linker conjugates (called MeNl-PEGn-LPA, Isol-PEGn-LPA and Cys-PEGn-LPA), measured by mass spectrometry analysis, were consistent with the expected values (Figure 28B).

Characterization of gold nanoparticles bearing different amounts of TNF and peptide -PEGa -LP A

To assess the presence of peptide and TNF on nanoparticles loaded with different amounts of MeNl-PEGn-LPA, or Isol-PEGn-LPA, or Cys-PEGn-LPA and TNF, we measured their capability to bind an anti-PEG mAb (as a probe for the targeting ligand) and an anti-TNF pAb (as a probe for TNF) using the anti-PEG mAb/anti- TNF pAb sandwich assay. MeNl -LPA/Au/TNF, Isol-LPA/Au/TNF and Cys- LPA/Au/TNF showed bell-shaped dose-response curves (Figure 29A, 30A and 31) suggesting that the preparation of bifunctional nanoparticles is feasible when optimal doses of peptide and TNF are used. In particular, maximal binding of MeNl- LPA/Au/TNF was obtained with nanoparticles prepared with 16 μg of TNF and 1-2 μg of MeNl-PEGn-LPA per ml of gold. Similarly, the maximal binding of Isol- LPA/Au/TNF and Cys-LPA/Au/TNF was obtained with nanoparticles prepared with 16 μg of TNF and 2-4 μg of peptide. No binding was observed when the assays were carried out using microtiterplates lacking the anti-PEG mAb (Figure 29B and 3 OB), suggesting that the binding of these nanoparticle was specific.

To verify the presence of functional Isol on nanoparticles we analyzed the binding of Isol-LPA/Au/TNF to ανβ3 using the avP3/anti-TNF pAb sandwich assay. Isol- LPA/Au/TNF could bind to ανβ3 (Figure 30C), pointing to the presence of functional Isol . The maximum binding was achieved when the nanoparticles were prepared using 16 μg of TNF and 4 μg of Isol-PEGn-LPA (Figure 30C). Of note, the binding properties of Isol-LPA/Au/TNF were similar to those of Isol- HSA/Au/TNF. i.e. of gold nanoparticles loaded Isol-HSA and TNF described previously ¹ (Figure 30D).

Characterization of Isol-LPA/Au/TNF by UV-vis spectrophotometry.

UV-visible absorption spectrum of Isol-LPA/Au/TNF nanoparticles prepared with a fixed concentration of 16 μg TNF and various amount of Isol-PEGn-LPA (ranging from 4 to 32 μg) are shown in Figure 32. The results show all nanoparticles, including those prepared with optimal amounts of peptide and TNF (4/16), are homogeneous and not aggregated.

CONCLUSION

The above findings, overall, indicate that the novel NGR- and IsoDGR-linker peptides, prepared with MAL-PEGn-lipoamide as a cross-linking reagent, can be successfully exploited for the preparation of bifunctional nanoparticles bearing tumor targeting moieties and TNF effector molecules. These novel peptide-linker conjugates can overcome the need of using albumin for coupling the targeting peptide to nanogold, as previously done. This approach may therefore reduce drug complexity, heterogeneity and immunogenicity, improve drug stability, reduce production costs, and facilitate nanodrug characterization and development.

References

1. Curnis, F., et al. IsoDGR-tagged albumin: a new alphavbeta3 selective carrier for nanodrug delivery to tumors. Small 9, 673-678 (2013).

2. Wong, A.H., Zhou, D. & Rini, J.M. The X-ray crystal structure of human aminopeptidase N reveals a novel dimer and the basis for peptide processing. J. Biol. Chem. 287, 36804-36813 (2012).

3. Liu, C, Yang, Y., Chen, L., Lin, Y.L. & Li, F. A unified mechanism for aminopeptidase N-based tumor cell motility and tumor-homing therapy. J. Biol.

Chem. 289, 34520-34529 (2014).

4. Luciani, N., et al. Characterization of Glu350 as a critical residue involved in the N-terminal amine binding site of aminopeptidase N (EC 3.4.11.2): insights into its mechanism of action. Biochemistry 37, 686-692 (1998).

5. Curnis, F., et al. NGR-tagged nano-gold: A new CD 13 -selective carrier for cytokine delivery to tumors. Nano Res 9, 1393-1408 (2016).

6. Curnis, F. & Corti, A. Production and characterization of recombinant human and murine TNF. Methods Mol. Med. 98, 9-22 (2004).

7. Ratti, S., et al. Structure-activity relationships of chromogranin A in cell adhesion. Identification and characterization of an adhesion site for fibroblasts and smooth muscle cells. J. Biol. Chem. 275, 29257-29263 (2000).

8. Di Matteo, P., et al. Immunogenic and structural properties of the Asn-Gly- Arg (NGR) tumor neovasculature-homing motif. Mol. Immunol. 43, 1509-1518 (2006).

9. Curnis, F., et al. Isoaspartate-glycine-arginine: a new tumor vasculature- targeting motif. Cancer Res. 68, 7073-7082 (2008).

10. Curnis, F., et al. Critical role of flanking residues in NGR-to-isoDGR transition and CD13/Integrin receptor switching. J. Biol. Chem. 285, 9114-9123 (2010).

11. Fields, G.B. & Noble, R.L. Solid phase peptide synthesis utilizing 9- fluorenylmethoxycarbonyl amino acids. Int. J. Pept. Protein Res. 35, 161-214. (1990).

12. Dupradeau, F.Y., et al. The R.E.D. tools: advances in RESP and ESP charge derivation and force field library building. Phys. Chem. Chem. Phys. 12, 7821-7839 (2010).

13. Madhavi Sastry, G., Adzhigirey, M., Day, T., Annabhimoju, R. & Sherman, W. Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments. J. Comput-Aided. Mol. Des. 27, 221-234 (2013).

14. Schmidt, M.W., et al. General atomic and molecular electronic structure system. J. Comput. Chem. 14, 1347-1363 (1993).

15. Cornell, W.D., et al. A Second Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules J. Am. Chem. Soc. 1995, 117, 5179-5197. J. Am. Chem. Soc. 118, 2309-2309 (1996).

16. Lindorff-Larsen, K., et al. Improved side-chain torsion potentials for the Amber ff99SB protein force field. Proteins 78, 1950-1958 (2010).

17. Wang, J., Wang, W., Kollmann, P. & Case, D. Antechamber, An Accessory Software PackageFor Molecular Mechanical Calculation. J. Comput. Chem. 25, 1157-1174 (2005).

18. Rubinstein, M. & Niv, M.Y. Peptidic modulators of protein-protein interactions: progress and challenges in computational design. Biopolymers 91, 505- 513 (2009).

19. Tubert-Brohman, I, Sherman, W., Repasky, M. & Beuming, T. Improved docking of polypeptides with Glide. J. Chem. Inf. Model. 53, 1689-1699 (2013).

20. Laio, A. & Parrinello, M. Escaping free-energy minima. Proc. Natl. Acad. Sci. U. S. A. 99, 12562-12566 (2002).

21. Piana, S. & Laio, A. A bias-exchange approach to protein folding. J. Phys. Chem. B 111, 4553-4559 (2007).

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. Bussi, G., Donadio, D. & Parrinello, M. Canonical sampling through velocity rescaling. J. Chem. Phys. 126, 014101 (2007). 24. Berendsen, H.J.C., Postma, J.P.M., Gunsteren, W.F.v., DiNola, A. & Haak, J.R. Molecular dynamics with coupling to an external bath. J. Chem. Phys. 81, 3684- 3690 (1984).

25. Parrinello, M. & Rahman, A. Polymorphic transitions in single crystals: A new molecular dynamics method. J. Appl. Phys. 52, 7182-7190 (1981).

26. Hess, B., Bekker, H., Berendsen, H.J.C. & Fraaije, J.G.E.M. LINCS: A linear constraint solver for molecular simulations. J. Comput. Chem. 18, 1463-1472 (1997).

27. 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). 28. Essmann, U., et al. A smooth particle mesh Ewald method. J. Chem. Phys. 103, 8577-8593 (1995).

29. Abraham, M.J., van der Spoel, D., Lindahl, E., Hess, B. & and the GROMACS development team. GROMACS User Manual version 5.1. (2015 ).

30. Tribello, G.A., Bonomi, M., Branduardi, D., Camilloni, C. & Bussi, G. PLUMED 2: New feathers for an old bird. Comput. Phys. Commun. 185, 604-613

(2014).

31. Biarnes, X., Pietrucci, F., Marinelli, F. & Laio, A. METAGUI. A VMD interface for analyzing metadynamics and molecular dynamics simulations. Comput. Phys. Commun. 183, 203-211 (2012).

32. Daura, X., et al. Peptide folding: When simulation meets experiment. Angew. Chem. Int. Ed. 38, 236-240 (1999).

33. McBride, W.J., et al. Improved 18F labeling of peptides with a fluoride- aluminum-chelate complex. Bioconjugate Chem. 21, 1331-1340 (2010).

34. Cossu, I, et al. Neuroblastoma-targeted nanocarriers improve drug delivery and penetration, delay tumor growth and abrogate metastatic diffusion. Biomaterials

68, 89-99 (2015).

35. Arap, W., Pasqualini, R. & Ruoslahti, E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 279, 377-380 (1998).

36. Ellerby, H.M., et al. Anti-cancer activity of targeted pro-apoptotic peptides. Nat. Med. 5, 1032-1038 (1999).

37. Curnis, F., et al. Enhancement of tumor necrosis factor alpha antitumor immunotherapeutic properties by targeted delivery to aminopeptidase N (CD 13). Nat. Biotechnol. 18, 1185-1190 (2000).

38. Curnis, F., et al. Targeted delivery of IFN-gamma to tumor vessels uncouples anti-tumor from counter-regulatory mechanisms. Cancer Res. 65, 2906-2913 (2005). 39. Calcinotto, A., et al. Targeting T F-alpha to neoangiogenic vessels enhances lymphocyte infiltration in tumors and increases the therapeutic potential of immunotherapy. J. Immunol. 188, 2687-2694 (2012).

40. Porcellini, S., et al. The tumor vessel targeting agent NGR-T F controls the different stages of the tumorigenic process in transgenic mice by distinct mechanisms. Oncoimmunology 4, el 041700 (2015).

41. Corti, A., Gasparri, A.M., Ghitti, M., Sacchi, A., Sudati, F., Fiocchi, M., Buttiglione, V., Perani, L., Gori, A., Valtorta, S., Moresco, R.M., Pastorino, F., Ponzoni, M., Musco, G. & Curnis, F. Glycine N-methylation in NGR-Tagged Nanocarriers Prevents Isoaspartate formation and Integrin Binding without Impairing CD 13 Recognition and Tumor Homing. (2017) Adv Funct Mater 27.

Claims

1. A modified NGR motif of formula (I)

wherein Z is different from hydrogen.

2. A modified NGR motif according to claim 1, characterized in that Z is an alkyl group, and has the following formula (II)

3. A N^AlkGR motif according to claim 2, characterized in that glycine is alkylated by a lower alkyl group.

4. A N^AlkGR motif according to claim 3, characterized in that said lower alkyl group is selected from methyl, ethyl, n-propyl, isopropyl n-butyl, iso-butyl, sec- butyl and tert-butyl

5. A N^AlkGR motif according to claim 4, characterized in that said lower alkyl group is a methyl group and has the following formula (III)

6. The N AlkGR motif according to anyone of claims 1 to 5, for its use as a ligand of aminopeptidase N (CD 13, EC 3.4.11.2).

7. A peptide comprising a N^AlkGR motif according to anyone of claims 1 to 5.

8. The peptide according to claim 7, having the following formula (III)

X- NiVAikGR-X' (III)

wherein X is selected from C, G, R, K, A and X' is selected from C, G, R K, A.

9. The peptide according to claim 8, which is selected from c[CGNA eGRG], c[CNA eGRGG], c[CNAMeGRG], c [CRGA eGRGPD] , CRG^MeGRGPD, CN^MeGRC.

10. The peptide according to any one of claims 7 to 9, for its use in the targeted- delivery of therapeutic and/or imaging and/or diagnostic moieties to tumors.

11. A conjugation product of the peptide according to any one of claims 7 to 9, with one or more therapeutic and/or imaging and/or diagnostic moieties.

12. The conjugation product according to claim 11, characterized in that said therapeutic and/or imaging and/or diagnostic moieties are selected from chemotherapeutic drugs; liposomes; gold nanoparticles; anti-angiogenic compounds; DNA complexes; viral particles; imaging compounds; cytokines; and tumor necrosis factors.

13. The conjugation product according to claim 12, characterized in that said therapeutic and/or imaging and/or diagnostic moieties are is selected from doxorubicin, melphalan, cis-platin, gemtabicine, taxol, auristatins or maytansines derivatives, a therapeutic or diagnostic antibody, a kinase inhibitor, IFN-γ, IFN-a-2a, IL12 TRAIL and TNF-a.

14. The conjugation product according to any one of claims 11 to 13, for its use in the diagnosis and/or imaging and/or treatment of tumors.

15. The conjugation product according to any one of claims 11 to 14, characterized in said therapeutic and/or imaging and/or diagnostic moieties are directly linked to the peptide.

16. The conjugation product according to any one of claims 11 to 14, characterized in said therapeutic and/or imaging and/or diagnostic moieties are linked to the peptide linked through a spacer.

17. A pharmaceutical and/or diagnostic composition comprising a NAfAlkGR-containing peptide according to any one of claims 7 to 10 with conventional pharmaceutically acceptable carriers and/or excipients.

18. The pharmaceutical and/or diagnostic composition according to claim 17, characterized that it is in the form of an injectable solution or suspension.

19. The pharmaceutical and/or diagnostic composition according to claim 17 or 18, for its use in the diagnosis and/or imaging and/or treatment of tumors.

20. Use of a N- substituted glycine for preventing asparagine deamidation in the NGR motif.

21. Use according to claim 20, wherein the Af-substituted glycine is -alkyl glycine

22. The use according to claim 21, characterized in that said N alkyl glycine is N- methyl glycine.

23. A conjugation product according to claim 11, which comprises TNF-a.

24. A conjugation product according to claim 11, which comprises liposomes.

25. A conjugation product according to claim 11, which comprises liposomal doxorubicine.

26. A conjugation product according to claim 1 1, which comprises NOTA, optionally radiolabeled.