Immunoglobulin Fc-Fused Peptide without C-Terminal Arg or Lys

Subscriber access provided by UNIV OF DURHAM

Article

An Immunoglobulin Fc-fused Peptide without C-terminal Arg or Lys Residue Augments Neuropilin-1–dependent Tumor Vascular Permeability Du-San Baek, Jeong-Ho Kim, Ye-Jin Kim, and Yong-Sung Kim Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.7b00761 • Publication Date (Web): 12 Dec 2017 Downloaded from http://pubs.acs.org on December 13, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

An Immunoglobulin Fc-fused Peptide without C-terminal Arg or Lys Residue Augments Neuropilin-1–dependent Tumor Vascular Permeability

Du-San Baek§, Jeong-Ho Kim§, Ye-Jin Kim, and Yong-Sung Kim*

Department of Molecular Science and Technology, Ajou University, Suwon 16499, Republic of Korea

*To whom the correspondence should be addressed: Yong-Sung Kim, Ph.D., Dept. of Molecular Science and Technology, Ajou University, 206 World Cup-ro, Yeongtong-gu, Suwon 16499, Republic of Korea. Tel: +82-31-219-2662; Fax: +82-31-219-1610. E-mail: [email protected]

§

These two authors contributed equally to this work.

1

ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For Table of Contents Use Only

Title : An Immunoglobulin Fc-fused Peptide without C-terminal Arg or Lys Residue Augments Neuropilin-1–dependent Tumor Vascular Permeability Authors : Du-San Baek, Jeong-Ho Kim, Ye-Jin Kim, and Yong-Sung Kim

2


Page 2 of 34

Page 3 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT: Neuropilin-1 (NRP1), which functions as a co-receptor for vascular endothelial growth factor (VEGF) and is implicated in vascular permeability and tumorigenesis, has been targeted by peptides that specifically bind to the VEGF-binding region on NRP1. Like natural VEGF ligands, all known peptides with NRP1-binding activity bind only through a carboxy (C)-terminal R/K-x-x-R/K sequence motif (x stands for any amino acids); this strict requirement is called the C-end rule (CendR). Here, we report immunoglobulin Fc-fused NRP1-specific peptides deviating from CendR. We screened a yeast surface-displayed Fc-fused non-CendR peptide library against NRP1 and isolated FcV12, wherein V12 peptide comprising 12 amino acids has a PPRV sequence at its C-terminal end. Although Fc-V12 lacked the CendR motif, it showed selective binding to the VEGFbinding region of NRP1 and triggered cellular internalization of NRP1, resulting in enhanced extravasation into tumor tissues and tumor tissue penetration of the Fc-fused peptide along with the co-injected chemical drug in tumor-bearing mice. Through a saturation mutagenesis study, we identified that the Val residue at the C-terminus of Fc-V12 is crucial for NRP1 binding. We further improved NRP1 affinity of Fc-V12 (KD = ~761 nM) through directed evolution of the upstream sequence of PPRV to obtain Fc-V12-33 (KD = ~17.4 nM), which exhibited enhanced NRP1-mediated vascular permeability as compared with Fc-V12. Our results provide functional Fc-fused non-CendR peptides, which bind to the VEGF-binding region of NRP1 and enhance vascular permeability, expanding the sequence space of NRP1targeting peptides.

KEYWORDS: C-end rule, neuropilin-1, vascular permeability, yeast surface display,

peptide engineering

3



1. INTRODUCTION Neuropilins (NRPs), NRP1 and NRP2 (NRP1/2), are homodimeric single-pass transmembrane receptors and play critical roles in the vascular, neuronal, and immune system. These act as co-receptors for various receptors with cognate ligands, including vascular endothelial growth factor receptors (VEGFRs) for VEGF family ligands and plexin receptors for secreted class 3 semaphorin (Sema3) family ligands.1, 2 NRP1/2 show distinct expression profiles each other; NRP1 is primarily found in arterial endothelial cells, whereas NRP2 is mainly expressed in venous and lymphatic endothelium.1 In particular, NRP1 is frequently expressed at high levels in tumor-associated blood vessels and epithelial tumor cells and plays critical roles in angiogenesis, vascular permeability, and tumorigenesis.2, 3 Therefore, NRP1 has been a target for the development of peptides and antibodies as anti-angiogenesis therapeutic agents as well as tumor-homing and tumor tissue penetration-enhancing agents.411

NRP1/2 are composed of extracellular regions of a1a2 domains (two complement binding motifs (CUB)), b1b2 domains (the coagulation factors V and VIII domain), and a c domain (MAM meprin domain) as well as a transmembrane domain and an intracellular PDZ (PSD-95/DIg/ZO-1 binding motif) binding domain.1,12 The upper four extracellular domains (a1, a2, b1, and b2) determine the binding specificity of multiple ligands to NRP1/2 and the last extracellular c domain, along with the transmembrane domain, is implicated in the dimerization or oligomerization between NRPs and their co-receptors.1, 12 Both VEGF and Sema3 family ligands specifically bind to the VEGF-binding region in the b1 domain of NRP1/2 through the C-terminal R/K-x-x-R/K sequence motif, where x stands for any amino acid.1, 12 In fact, all known proteins and peptides binding to the ligand-binding pocket in the NRP1-b1 domain share the sequence motif.5, 13 This basic sequence motif in the NRP14


Page 4 of 34

Page 5 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


binding ligands and peptides must be exposed at the C-terminus for binding to NRP1, with a stringent requirement for Arg (or rarely Lys) at the last C-terminal

residue;13,

14

this

requirement is called the “C-end rule” (CendR).15 The substitution of the C-terminal Arg with other amino acids or capping it with additional C-terminal amino acids completely abolished the binding to NRP1 and NRP1-mediated vascular permeability.13, 15, 16 Tumor-penetrating peptides targeting NRP1, such as RPARPAR15 and activated iRGD (RGDR)6 as well as A22p (–RNRR) and TPP11 (–TPRR), fused to the C-terminus of immunoglobulin Fc (Fc-A22p and Fc-TPP11)7-9 also comply with the CendR. However, plasma proteases, such as carboxypeptidases, can rapidly remove the C-terminal Arg/Lys residue of CendR peptides, thus promptly diminishing their active form in the blood circulation.17 In addition, the Cterminal Arg/Lys residues of CendR peptide-conjugated antibody can also be cleaved by basic carboxypeptidases during production in mammalian cell cultures.18,

19

Thus, for

practical applications, non-CendR peptides that specifically bind to the VEGF-binding site of NRP1, without the requirement for the CendR sequence motif, are highly demanding. Herein, we report the Fc-fused non-CendR peptides Fc-V12 and Fc-V12-33, which have Val residue at their C-termini instead of Arg/Lys. Hence, these peptides deviate from the CendR peptide motif but selectively bind to the VEGF-binding region of NRP1-b1 domain. We found that Fc-fused non-CendR peptides exerted NRP1-mediated vascular permeability and augmented NRP1-mediated tumor homing and tissue penetration in human tumor xenograft-bearing mice. Our results challenge the requirement of CendR for NRP1-binding peptides and demonstrate that non-CendR peptides may induce NRP1-dependent signaling.

2. EXPERIMENTAL SECTION 2.1. Generation and screening of Fc-fused non-CendR peptide library displayed on 5



the surface of yeast. The DNA fragment carrying the hinge-CH2-CH3 regions of human IgG1 (residue 225-447 in EU number), a 15-residue (G4S)3 linker, and the C-terminal sequence of Sema3A (749HLQENKKGRNRR760), was subcloned in-frame into the Cterminus of Aga2 gene in pCTCON of the yeast surface-display plasmid, as previously described.7 For the construction of the Fc-fused non-CendR peptide library on yeast surface, eight residues of the C-terminal Sema3A were diversified with degenerate codons of NHC (ACGT/ACT/C) and VVM (ACG/ACG/AC) for the last C-terminal residue and the upstream seven residues, respectively, as shown in Figure 1A. The transformants were selected directly in selective liquid media, as previously described.20 The library screening was performed by one round of magnetic-activated cell sorting (MACS), followed by three rounds of fluorescence-activated cell sorting (FACS) using FACS Aria III (BD Biosciences) against biotinylated NRP1-b1b2 in the presence of a 10-fold molar excess of non-biotinylated NRP1b1b2-AAA mutant as a competitor. Briefly, in MACS, 1 × 1010 freshly induced yeast cells in SG-CAA media (6.7 g/l yeast nitrogen base, 5 g/l casamino acids, 5.4 g/l Na2HPO4, 8.56 g/l NaH2PO4·H2O, and 20 g/l galactose in deionized water) were washed with 25 ml of PBSE buffer (PBS (10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4, plus 137 mM NaCl and 2.7 mM KCl) containing 5 mg/ml bovine serum albumin [BSA] and 2 mM EDTA) were incubated with 1 µM of biotinylated-NRP1-b1b2 and 10 µM of non-biotinylated-NRP1-b1b2-AAA in 10 ml PBSE for 1 h with frequent gentle shaking at room temperature. After washing twice with ice-cold 50 ml PBSE, the cells were resuspended with 5 ml PBSE and 200 µl streptavidin microbeads (Miltenyi Inc., #130-048-101) and then incubated with rotations at 4 °C for 10 min. Incubated yeast cells were subsequently pelleted and the supernatant was aspirated. After adding 50 ml PBSE, resuspended cells were loaded onto a Miltenyi Macs LS column held in place in the external magnetic field. The column was washed with 1 ml PBSE 6


Page 6 of 34

Page 7 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in every 7 ml of sample loading. The bound cells were eluted by detaching the column from the magnetic apparatus and adding 7 ml selective medium of SD-CAA, which contained the same composition as SG-CAA, except galactose was replaced with glucose. The eluted cells were grown in selective media. The enriched cells were subsequently subjected to FACS using more stringent conditions: 0.5 µM of biotinylated-NRP1-b1b2 and 5 µM of nonbiotinylated-NRP1-b1b2-AAA in round 1, 0.1 µM of biotinylated-NRP1-b1b2 and 1 µM of non-biotinylated-NRP1-b1b2-AAA in round 2, and 50 nM of biotinylated-NRP1-b1b2 and 0.5 µM of non-biotinylated-NRP1-b1b2-AAA in round 3. The cell-surface expression and binding level of biotinylated NRP1-b1b2 of the library were determined by indirect double immunofluorescence labeling of a C-terminal c-myc tag (anti-c-myc mouse Ab (9E10, dilution 1:200) with Alexa488-labeled anti-mouse goat Ab (Invitrogen, #A11001, dilution 1:600)) and streptavidin-conjugated R-phycoerythrin (SA-PE) (Invitrogen, #S866, dilution 1:600), respectively. Typically, the top 0.5-1% of target-binding cells was sorted. The finally sorted yeast cells were plated on the selective medium and individual clones were isolated and further analyzed.

2.2. Construction and purification of Fc-fused peptides. From the screened yeast cells, DNAs were rescued by Zymoprep kit (Zymo Research, #D2001) as previously described.20 The DNA fragment carrying CH3-(G4S)3-peptide region was subcloned into the modified pcDNA3.4 vector using BsrGI/HindIII sites to be expressed in format of Fc-fused peptides in mammalian cells.7 Fc and Fc-fused peptides were expressed by transient transfection into HEK293F cells (Thermo Fisher Scientific) and purified using a Protein-A agarose chromatography column (GE Healthcare), as previously described.21, 22

7



2.3. Expression and purification of recombinant NRP proteins. The cDNAs of human NRP1 and NRP2 were purchased from Thermo Fisher Scientific Inc. DNA fragments encoding NRP proteins (NRP1-b1b2 [residues 273-586], NRP2-b1b2 [residues 275-595], NRP1-b1b2-AAA [Tyr297Ala, Ser346Ala, Tyr353Ala], and NRP2-b1b2-AAA [Tyr299Ala, Ser349Ala, Tyr356Ala]), Avi-tag, and 6× histidine tag were subcloned into pET28a vector (Novagen) using NcoI/BamHI sites. For the in vivo biotinylation of NRP1-b1b2 and NPR2b1b2, plasmids encoding NRP1-b1b2-Avi-His or NRP2-b1b2-Avi-His were co-transformed into Escherichia coli Origami 2(DE3) strain with an expression plasmid for biotin protein ligase, pBirA (Avidity).23 The co-transformed bacteria were cultured in 1 L of Luria-Bertani broth containing 50 µg/mL kanamycin, 25 µg/mL chloramphenicol, and 10 mM magnesium chloride (MgCl2) at 37 °C up to an absorbance (OD600) of 0.8. For induction and biotinylation of proteins, isopropyl β-D-1-thiogalactopyranoside (IPTG) and d-biotin (Biobasic, #BB0078) in a buffer (10 mM bicine pH 8.3; Sigma, #163791) were added at final concentrations of 0.1 mM and 50 µM, respectively. After 16 h of induction at 30 °C, cells were harvested and the proteins were purified using Ni-NTA resin (Qiagen, #30210).7, 8 Non-biotinylated proteins were prepared by induction without d-biotin.

2.4. Binding analysis by enzyme-linked immunosorbent assay (ELISA). Binding specificity of Fc and Fc-fused peptides for the purified NRP proteins was analyzed by indirect ELISA, as previously described.8 Briefly, 200 ng of biotinylated proteins (NRP1b1b2, NRP2-b1b2, NRP1-b1b2-AAA, and NRP2-b1b2-AAA) were added to wells of 96-well plates (Corning, #3690) and incubated at 25 °C for 1 h. The wells were subsequently blocked with a blocking buffer (1% BSA diluted in PBS, pH 7.4). After washing thrice with PBS (pH 7.4), wells were treated with serially diluted Fc or Fc-fused peptides in blocking buffer at 8


Page 8 of 34

Page 9 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


25 °C for 1 h. After washing thrice with PBS, bound proteins were detected by adding 50 µL of horseradish peroxidase (HRP)-conjugated anti-human Fc antibody (dilution 1:8000; Sigma, #I2136). Enzymatic reactions were carried out at 25 °C by adding 3,3′,5,5′tetramethylbenzidine (TMB) substrate for 5-15 min, followed by the addition of 1 N sulfuric acid (H2SO4) to stop the reaction. The absorbance at 450 nm wavelength was read with a 96well plate reader. For competitive ELISA with Fc-fused peptides, wells of 96-well plates were coated with 200 ng of soluble NRP1-b1b2 domains and blocked with blocking buffer as described above. Mixtures of diluted Fc or Fc-fused peptides at various concentrations and 1 nM biotinylated VEGF-A165 (R&D, #BT293) were added to the wells and incubated at 25 °C for 1 h. After washing thrice, bound biotinylated VEGF-A165 was detected with HRPconjugated streptavidin (dilution 1:16000; Thermo Fisher Scientific, #N100). The binding data were fitted by a nonlinear regression model using GraphPad PRISM (GraphPad software, Inc.).

2.5. Surface plasmon resonance (SPR) analysis. Kinetic interactions of Fc-fused peptides with NRP proteins (NRP1-b1b2 and NRP2-b1b2) were monitored by Biacore 2000 SPR biosensor (GE Healthcare), as previously described.9 After immobilization of the indicated Fc-fused peptides onto the CM5 sensor chip at a level of about 1,000 response units (RUs), NRP proteins (NRP1-b1b2 and NRP2-b1b2) were injected into the flow cells at a flow rate of 30 µL/min for 3 min, followed by 3-min dissociation per cycle. Regeneration of the flow cells was allowed by flowing 20 mM sodium hydroxide (NaOH; pH 10) for 2 min. The binding kinetic data were evaluated by analyzing fitted sensorgrams for each concentration using the 1:1 binding model within BIA evaluation software provided by the manufacturer.

9



2.6. Endothelial permeability assay. The in vitro transwell assay was performed to measure permeability across endothelial cell monolayer of human umbilical vein endothelial cells (HUVECs), as previously described.7, 8 Small interfering RNAs (siRNAs) were synthesized by Bioneer Co. (Daejeon, Korea). HUVECs (5 × 105 cells/well in 6-well plate) were transfected with control siRNA (100 nM), NRP1 siRNA (100 nM), or NRP2 siRNA (100 nM) using Lipofectamine RNAiMAX (Invitrogen, USA) following the manufacture’s protocol and incubated for 24 h. The sequences of the siRNAs are as follows: control scrambled siRNA 5′AAU UCU CCG AAC GUG UCA CGU-3′, NRP1 siRNA 5′-GGA UUU UCC AUA CGU UAU-3′ and NRP2 siRNA 5′-AAA GGC UGG AAG UCA GCA CUA AUU U-3′.7, 8 For the permeability assay, HUVECs or siRNA-transfected HUVECs (2 × 105 cells/well) were seeded into the 24-well transwell chamber (0.4 µm pore size; Corning Costar, USA) and grown for 2 days. After 6 h of serum starvation, proteins were added to both upper and lower chamber as specified in the figure legends. After 30 min, 50 µg fluorescein isothiocyanate (FITC)-conjugated dextran (approximately 40 kDa, Sigma-Aldrich, #FD40) was added to the upper chamber for 30 min at 37 °C. The fluorescence of samples from the lower chamber was measured by a fluorescence plate reader (Cytation 3, Bioteck, USA). Fc-treated control samples were used for normalization of data.

2.7. Confocal fluorescence microscopy of cells. Confocal fluorescence microscopic analyses of cells were performed as previously described.7, 9 Briefly, cells (5 × 104 cells per well) grown on coverslips in 24-well culture plates were washed and treated with proteins diluted in a serum-free medium at 4 or 37 °C, as specified in the figure legends. For detection of surface-bound or internalized proteins, cells were stained with Alexa 488-conjugated goat anti-human IgG antibody (Invitrogen, #A11013) for 1 h at 25 °C. NRP1 proteins were 10


Page 10 of 34

Page 11 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


visualized with a primary rabbit anti-NRP1 antibody (Abcam, #ab81321) and then secondary tetramethylrhodamine (TRITC)-conjugated anti-rabbit antibody (Sigma-Aldrich, #T6778) for 1 h at 25 °C. Center-focused single z-section images were obtained using a Zeiss LSM710 confocal microscope with ZEN software (Carl Zeiss).

2.8. Xenograft tumor models. All mouse experimental protocols were approved by the Animal and Ethics Review Committee of Ajou University and the experiments were carried out following the guidelines established by the Institutional Animal Care and Use Committee. FaDu (5× 106 cells/mouse) cells in 150 µL of a 1:1 mixture of PBS/Matrigel (BD Biosciences, USA) were subcutaneously inoculated into the right thigh of 4-week-old female BALB/c nude mice (Nara bio, Korea) weighing 15-20 g.7, 9 When the mean tumor volume reached approximately 250-300 mm3, mice were randomly assigned to treatment agents (Fc proteins and/or doxorubicin [Sigma-Aldrich]), as specified in the figure legends.

2.9. Confocal fluorescence microscopy of tumor tissues. The cryosectioning and immunofluorescence staining of tumor tissues were performed as previously described.7, 22 Briefly, 20 µm-thick cryosectioned tumor tissues were prepared and incubated with freshly prepared blocking solution (2% (w/v) BSA in PBS, pH 7.4) for 1 h at 25 °C. The tissue sections were then stained with rat anti-mouse CD31 antibody (BD Biosciences, #553370) for 1.5 h at 4 °C, washed thrice with PBST (PBS, pH 7.4, 0.1% Tween-20) for 5 min, and incubated with goat anti-rat TRITC-conjugated antibody (Millipore, #AP136R). The injected Fc proteins were stained with Alexa 488-conjugated goat anti-human IgG antibody (Invitrogen, #A11013) for 1 h at 25 °C. The nuclei were stained with Hoechst 33342 for 5 min at 25 °C. For the analysis of tumor tissue distribution of doxorubicin (DOX) (Sigma, 11



Page 12 of 34

#44583), cryosectioned tumor tissues were stained for blood vessels with the rat anti-mouse CD31 antibody and then rabbit anti-rat FITC-conjugated antibody (Sigma, #F1763). DOX was detected using the autofluorescence (red). Tissue sections were washed thrice with PBST, mounted on slides with Perma Fluor aqueous mounting medium (Thermo Fisher Scientific, #TA-030-FM), and examined under a Zeiss LSM710 system with ZEN software (Carl Zeiss). Areas positively stained for Fc proteins in the acquired fluorescence images of each tissue were quantified using ImageJ software.

2.10. Statistical analysis. Data represent the mean ± standard deviation for representative data from two to three independent experiments, unless otherwise specified. Comparisons of data from tests and controls were analyzed for statistical significance by a two-tailed, unpaired Student's t-test using GraphPad Prism 5 software (GraphPad). A value of P < 0.05 was considered significant.

3. RESULTS AND DISCUSSION 3.1. Design and construction of Fc-fused non-CendR peptide library. The crystal structures of NRP1-b1 domain in complex with either VEGF14 or CendR peptide Tuftsin (TKPR)24 showed that the last Arg residue of the CendR motif binds to a pocket composed of Asp320, Tyr297, Ser346, Thr349, and Tyr353 residues on the NRP1-b1 domain through electrostatic interactions and multiple hydrogen bonds. This region on NRP1-b1 domain for binding of VEGF and CendR peptide is referred as an arginine-binding pocket.14,

24

In

addition, approximately eight residues before the last C-terminal Arg in VEGF specifically interact with residues within and around the arginine-binding pocket, contributing to the selective binding of VEGF to NRP1 instead of NRP2.25 We reasoned that the substitution of 12


Page 13 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the last C-terminal Arg residue of CendR motif with non-Arg/Lys residues and additional optimization of the upstream sequences may generate non-CendR peptides that specifically bind to the arginine-binding pocket onNRP1-b1 domain. We sought to engineer the Cterminal sequence of furin-processed Sema3A into non-CendR peptides in the fusion at the C-terminus of Fc to generate homodimeric Fc-mediated bivalent NRP1-binding form, and thus, exert NRP1-mediated biological activity7-9 similar to that observed for the homodimeric VEGF and Sema3A ligands.12 The chosen template peptide carries 12 residues derived from the C-terminal basic tail of Sema3A (Figure 1A), which is thought to bind to the argininebinding pocket of both NRP1/2.26 Based on the template sequence, we generated a Fc-fused non-CendR peptide library using yeast surface display technique, wherein the last C-terminal Arg residue was randomized with NHC degenerate codon (encoding Ala, Asp, Phe, His, Ile, Leu, Asn, Pro, Ser, Thr, Val, Tyr) and the seven upstream residues were randomized with a VVM degenerate codon (encoding Ala, Asp, Glu, Gly, His, Lys, Asn, Pro, Gln, Arg, Ser, Thr) (Figure 1A). The library diversity, determined by number of colonies grown on the selective plates20, 21, was approximately 4.3 × 108.

3.2. Isolation and characterization of NRP1-specific Fc-V12. To rule out the isolation of false-positive peptides binding to other surfaces of NRP1 located away from the argininebinding pocket during screening, we designed a screening competitor, NRP1-b1b2-AAA mutant, which displayed three mutations of Tyr297Ala, Ser346Ala, and Tyr353Ala to abolish the arginine-binding pocket (Figure 1A). The yeast library was screened by one round of MACS and three rounds of FACS using the biotinylated NRP1-b1b2 protein in the presence of a 10-fold molar excess of non-biotinylated NRP1-b1b2-AAA mutant as a competitor. As a consequence, we isolated one unique NRP1 binder designated as Fc-V12, in which the V12 13



peptide had a sequence deviating from CendR (Figure 1B). The isolated Fc-V12 was expressed in HEK293F mammalian cells. The purified Fc-V12 was the correctly assembled monomeric form without non-native oligomers (Figure 1B and Supplementary Figure S1). Indirect ELISA revealed that Fc-V12 selectively bound to NRP1b1b2 but exhibited negligible binding to NRP2-b1b2 and NRP1-b1b2-AAA mutants (Figure 1C), indicative of its selective binding to the arginine-binding pocket on the NRP1-b1 domain. SPR assay revealed the affinity of Fc-V12 for NRP1 with a dissociation constant (KD) of approximately 761 nM (Table 1 and Supplementary Figure S2). The previously reported two Fc-fused CendR peptides showed the expected binding characteristics; Fc-A22p bound to both NRP1 (KD = ~58 nM) and NRP2 (KD = ~96 nM)8 and Fc-TPP11 retained approximately 50-fold higher selective affinity to NRP1 (KD = ~28 nM) than NRP2 (KD = ~1.4 µM)7 (Figure 1C, Table 1 and Supplementary Figure S2). We evaluated the binding of Fc-V12 to the surface of HUVECs expressing NRP1 and found that Fc-V12 bound to the cells; however, the binding activity was abolished by the knockdown of NRP1, but not NRP2, demonstrating the specific binding of Fc-V12 to the endogenously expressed NRP1 on the cell surface (Figure 1D). To evaluate whether Fc-V12 shares the binding region on NRP1-b1 domain with VEGF, we conducted competitive ELISA with the 165-amino acid isoform of VEGF A (VEGF-A165). Indeed, Fc-V12 competed with VEGF-A165 for NRP1 binding, with the inhibitory concentration for 50% reduction in binding (IC50) of ~9.7 µM (Figure 1E). Though Fc-V12 competed less efficiently than Fc-TPP11 (IC50 = ~85 nM) (Figure 1E), the result indicates that Fc-V12 specifically binds to VEGF-binding region in NRP1-b1 domain, similar to other CendR peptides.

14


Page 14 of 34

Page 15 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.3. Fc-V12 homes to tumor-associated blood vessels and penetrates into tumor tissues. As NRP1-mediated vascular permeability is accompanied with cellular internalization of NRP1 after forming a complex with multivalent CendR peptide,7, 9, 27 we monitored NRP1 internalization by Fc-V12 with confocal fluorescence microscopy. Both Fc-V12 and FcTPP11 co-localized with NRP1 on the surface of HUVECs without triggering NRP1 internalization at 4 °C; however, both were internalized into cells together with NRP1 at 37 °C (Figure 2A). This result suggests that Fc-V12 triggers the cellular internalization of NRP1 in an energy-dependent manner, as shown with NRP1-binding CendR peptides in multivalent format.7-9, 27 The ability of Fc-V12 to induce cellular internalization of NRP1 led us to investigate whether Fc-V12 enhances endothelial permeability across HUVEC monolayer in an in vitro transwell assay. Fc-V12 exhibited dose-dependent enhancement in the passage of fluorescein isothiocyanate (FITC)-dextran through HUVEC monolayer as compared with Fc, though the extent was modest as compared with Fc-TPP11 (Figure 2B). This result demonstrates that FcV12 induced NRP1-mediated vascular permeability in vitro, similar to that detected with Fcfused CendR peptide Fc-TPP11.7 The tumor-homing and tumor tissue-penetration abilities of Fc-V12 were compared with those of Fc and Fc-TPP11 by a single intravenous (i.v.) injection into mice with preestablished human head and neck FaDu tumors.8 After 24 h of circulation, tumors were taken from mice and analyzed to detect Fc-V12 by immunofluorescence staining. In comparison with Fc, Fc-V12 showed approximately three-fold higher detection in and around tumor vessels as well as inside the tumor tissues away from the vessels (Figure 2C). Furthermore, i.v. co-administration of Fc-V12 with the anti-cancer drug doxorubicin (DOX) into mice carrying FaDu tumors exhibited a deeper and broader penetration of DOX into the 15



extravascular tumor tissue as compared with DOX penetration observed with Fc co-injection (Figure 2D). However, these activities of Fc-V12 were much lower than those of Fc-TPP11, which showed 26-fold higher NRP1-binding affinity (Figure 2C and D). Nonetheless, the above results indicate that Fc-V12 possesses the ability to home to tumor vessels and trigger NRP1-mediated vascular permeability of Fc-fused CendR peptide Fc-TPP11,7 resulting in the enhanced tumor tissue penetration.

3.4. Essential elements of Fc-V12 for binding to NRP1. To determine the contribution of the last C-terminal Val residue of V12 non-CendR peptide for binding to NRP1, we conducted saturation mutagenesis by substituting Val of Fc-V12 with the other 19 amino acids. All 19 Fc-V12 mutants were purified in the native form (~56 kDa without oligomers; Figure 3A). When evaluated for their binding to HUVECs expressing NRP1 on the cell surface, only Fc-V12 bound to the cells, while the other 19 mutants failed to show significant binding activity (Figure 3B). This result indicates that Val residue at the C-terminal end of V12 plays critical role in binding to NRP1. Next, we hypothesized that the last five residues at the C-terminus (8KPPRV12) in V12 may be important for binding to NRP1 due to the sequence similarity with CendR motif. To test this hypothesis, we generated a mutant, Fc-GS-V12, which retained the C-terminal five residues (8KPPRV12) of Fc-V12, but the upstream sequence (1HLQESPG7) was replaced with 1

GSGSGSG7. In comparison with Fc-V12, Fc-GS-V12 mutant maintained the selective

binding activity for NRP1, but not for NRP2, as observed in indirect ELISA (Figure 3C). This result also demonstrates that the C-terminal five residues are highly significant in term of its selectivity for NRP1 over NRP2. Taken together, these results indicate that the interactions between Fc-V12 and NRP1 mainly occur from five residues at C-terminus of V12 and the 16


Page 16 of 34

Page 17 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


last Val residue plays a essential role in the binding of Fc-V12 to NRP1.

3.5. Affinity-matured Fc-V12 derivative shows more potent vascular permeability. Although the Fc-fused non-CendR V12, Fc-V12, binds to the VEGF-binding region on the NRP1-b1 domain, its ability to trigger NRP1-mediated vascular permeability was much weaker as compared with that of Fc-TPP11 (Figure 2). Therefore, we conducted affinity maturation of V12 to investigate whether the non-CendR peptide shows improved binding affinity to NRP1 and enhanced biological activity, comparable to those of CendR peptides. As the C-terminal five residues (8KPPRV12) of Fc-V12 were critical for NRP1 binding, we constructed a yeast surface-displayed Fc-fused V12 peptide-mutated library, wherein the upstream seven residues (1HLQESPG7) of V12 were diversified by VVM degenerate codon while conserving 8KPPRV12 region. Screening of the library against biotinylated NRP1-b1b2 in the competition mode using non-biotinylated NRP1-b1b2-AAA mutant with one round of MACS and three rounds of FACS yielded four unique high-affinity Fc-V12 variants (Fc-V1214, 15, 33, and 44) (Figure 4A and Table 1). However, only Fc-V12-33 showed selective binding to NRP1 expressed on HUVECs, whereas the other three clones showed crossreactivity with NRP2 (Figure 4B). These results indicate that, in addition to the C-terminal five residues (8KPPRV12), the upstream seven residues of the non-CendR peptides significantly contribute to the selective binding activity for NRP1 and NRP2, as shown with VEGF-A145 and VEGF-A165 with the CendR motif.14, 28 Since we intended to isolate NRP1specific non-CendR binder, we chose Fc-V12-33 for the further study. The purified Fc-V12-33 exhibited approximately 44-fold higher affinity (KD = ~17 nM) for NRP1 as compared with the parental Fc-V12 (KD = ~761 nM) (Table 1 and Supplementary Figure S2). NRP1 affinity of Fc-V12-33 was comparable with that of Fc-TPP11 (KD = ~28 17



nM). In competitive ELISA, Fc-V12-33 competed more efficiently with VEGF-A165 (IC50 = ~68 nM) than Fc-V12 (IC50 = ~9.7 µM) for NRP1 binding, showing competition efficiency similar to that of Fc-TPP11 (IC50 = ~85 nM) (Figure 4C). This result suggests that Fc-V12-33 was affinity matured while conserving the binding epitope for the VEGF-binding region of NRP1. When evaluated in the streptavidin-mediated tetravalent format, the synthesized V12 and V12-33 peptides with the N-terminal biotinylation also exhibited the specific binding activity for NRP1-b1b2 domain in the concentration-dependent manner, but not for NRP2b1b2 and NRP1-b1b2-AAA mutant protein. Further, the selective binding of the non-CendR peptides was abolished in the presence of VEGF-A165 (Supplementary Figure S3). These results further confirmed that the non-CendR V12 and V12-33 peptides specifically bind to NRP1, particularly to the VEGF-binding region. To assess whether the affinity improvement of Fc-V12-33 augments NRP1-mediated vascular permeability, we performed transwell assay using HUVECs. In comparison with FcV12, Fc-V12-33 induced approximately two-fold increase in the passage of FITC-dextran through the HUVEC cell monolayer, showing a magnitude similar to that of VEGF-A165 and Fc-TPP11 (Figure 4D). Knockdown of NRP1 by siRNA treatment completely abolished the induced endothelial permeability (Figure 4D). These results indicate that the high-affinity non-CendR Fc-V12-33 triggers greater NRP1-mediated vascular permeability than Fc-V12, which is in good agreement with the observations for Fc-fused CendR peptides (Fc-A22p versus Fc-TPP11) and ligands (VEGF-A121 versus VEGF-A165).7, 29

4. CONCLUSION Studies have reported that CendR peptides bind to the VEGF-binding region on the b1 domain of NRP1/2,8,

15, 30

and lose their binding activity and biological function upon 18


Page 18 of 34

Page 19 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


substitution of their last C-terminal Arg or Lys residue with another amino acid.13, 15, 16 In this study, we successfully isolated Fc-fused non-CendR peptides, Fc-V12 and Fc-V12-33, which selectively bind to the VEGF-binding region of NRP1 and display minimal binding activity for NRP2, despite the absence of a CendR sequence motif. Similar to CendR peptides,6-8, 29 Fc-fused non-CendR Fc-V12 triggered cellular internalization of NRP1, resulting in increased extravasation into tumor tissues and tumor tissue penetration of the peptide as well as the co-administered drug DOX. Thus, the biological activities of Fc-V12 were comparable with those of CendR peptides. In comparison with the parent Fc-V12 (KD = ~761 nM), the affinity-matured NRP1-selective Fc-V12-33 (KD = ~17 nM) showed enhanced NRP1dependent vascular permeability in HUVECs. We found that the C-terminal five residues of Fc-V12 are sufficient to confer selectivity toward NRP1 over NRP2 and the last Val residue is essential for the binding to the VEGF-binding region of NRP1. For complete understanding of the structural basis, we need to further determine the complex structure of NRP1 with V12 or V12-33 peptides. The development of Fc-fused non-CendR peptides is essential for the future practical applications because the C-terminal Arg and Lys residues of Fc-fused CendR peptides may be removed by carboxypeptidases in cell culture or blood circulation after systemic administration, as shown with antibody.18, 19 To the best of our knowledge, this is the first report on a non-CendR peptide that selectively binds to the VEGF-binding region of NRP1. Our results provide a new avenue for the development of optimized NRP1-targeting peptides without the restriction of CendR peptides.

ACKNOWLEDGEMENTS 19



Page 20 of 34

This work was supported by grants from the Mid-career Researcher Program (2016R1A2A2A05005108) and the Pioneer Research Center Program (2014M3C1A3051470) from the National Research Foundation funded by the Korean government.

Conflict of Interest Y.S.K., D.S.B., and J.H.K. are listed as inventors on pending patent applications related to technology described in this work. Y.J.K. declare no competing financial interests.

REFERENCES 1.

Prud'homme, G. J.; Glinka, Y.

Neuropilins are multifunctional coreceptors involved

in tumor initiation, growth, metastasis and immunity. Oncotarget 2012, 3, (9), 921-39. 2.

Chaudhary, B.; Khaled, Y. S.; Ammori, B. J.; Elkord, E.

Neuropilin 1: function and

therapeutic potential in cancer. Cancer Immunol Immunother 2014, 63, (2), 81-99. 3.

Goel, H. L.; Mercurio, A. M.

VEGF targets the tumour cell. Nat Rev Cancer 2013,

13, (12), 871-82. 4.

Pan, Q.; Chanthery, Y.; Liang, W. C.; Stawicki, S.; Mak, J.; Rathore, N.; Tong, R. K.;

Kowalski, J.; Yee, S. F.; Pacheco, G.; Ross, S.; Cheng, Z.; Le Couter, J.; Plowman, G.; Peale, F.; Koch, A. W.; Wu, Y.; Bagri, A.; Tessier-Lavigne, M.; Watts, R. J.

Blocking neuropilin-1

function has an additive effect with anti-VEGF to inhibit tumor growth. Cancer Cell 2007, 11, (1), 53-67. 5.

von Wronski, M. A.; Raju, N.; Pillai, R.; Bogdan, N. J.; Marinelli, E. R.; Nanjappan,

P.; Ramalingam, K.; Arunachalam, T.; Eaton, S.; Linder, K. E.; Yan, F.; Pochon, S.; Tweedle, M. F.; Nunn, A. D.

Tuftsin binds neuropilin-1 through a sequence similar to that encoded by

exon 8 of vascular endothelial growth factor. The Journal of biological chemistry 2006, 281, (9), 5702-10. 6.

Sugahara, K. N.; Teesalu, T.; Karmali, P. P.; Kotamraju, V. R.; Agemy, L.; Greenwald,

D. R.; Ruoslahti, E.

Coadministration of a tumor-penetrating peptide enhances the efficacy 20


Page 21 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of cancer drugs. Science 2010, 328, (5981), 1031-5. 7.

Kim, Y. J.; Bae, J.; Shin, T. H.; Kang, S. H.; Jeong, M.; Han, Y.; Park, J. H.; Kim, S.

K.; Kim, Y. S.

Immunoglobulin Fc-fused, neuropilin-1-specific peptide shows efficient

tumor tissue penetration and inhibits tumor growth via anti-angiogenesis. Journal of

controlled release : official journal of the Controlled Release Society 2015, 216, 56-68. 8.

Shin, T. H.; Sung, E. S.; Kim, Y. J.; Kim, K. S.; Kim, S. H.; Kim, S. K.; Lee, Y. D.;

Kim, Y. S.

Enhancement of the tumor penetration of monoclonal antibody by fusion of a

neuropilin-targeting peptide improves the antitumor efficacy. Molecular cancer therapeutics 2014, 13, (3), 651-61. 9.

Kim, Y. J.; Jung, K.; Baek, D. S.; Hong, S. S.; Kim, Y. S.

receptor

and

neuropilin-1

overcomes

cetuximab

resistance

Co-targeting of EGF in

pancreatic

ductal

adenocarcinoma with integrin beta1-driven Src-Akt bypass signaling. Oncogene 2017, 36, (18), 2543-2552. 10.

Kadonosono, T.; Yamano, A.; Goto, T.; Tsubaki, T.; Niibori, M.; Kuchimaru, T.;

Kizaka-Kondoh, S.

Cell penetrating peptides improve tumor delivery of cargos through

neuropilin-1-dependent extravasation. Journal of controlled release : official journal of the

Controlled Release Society 2015, 201, 14-21. 11.

Yan, Z.; Yang, Y.; Wei, X.; Zhong, J.; Wei, D.; Liu, L.; Xie, C.; Wang, F.; Zhang, L.;

Lu, W.; He, D.

Tumor-penetrating peptide mediation: an effective strategy for improving

the transport of liposomes in tumor tissue. Molecular pharmaceutics 2014, 11, (1), 218-25. 12.

Pellet-Many, C.; Frankel, P.; Jia, H.; Zachary, I.

Neuropilins: structure, function

and role in disease. The Biochemical journal 2008, 411, (2), 211-26. 13.

Parker, M. W.; Linkugel, A. D.; Vander Kooi, C. W.

Effect of C-terminal sequence

on competitive semaphorin binding to neuropilin-1. J Mol Biol 2013, 425, (22), 4405-14. 14.

Parker, M. W.; Xu, P.; Li, X.; Vander Kooi, C. W.

Structural basis for selective

vascular endothelial growth factor-A (VEGF-A) binding to neuropilin-1. The Journal of

biological chemistry 2012, 287, (14), 11082-9. 15.

Teesalu, T.; Sugahara, K. N.; Kotamraju, V. R.; Ruoslahti, E.

C-end rule peptides

mediate neuropilin-1-dependent cell, vascular, and tissue penetration. Proceedings of the

National Academy of Sciences of the United States of America 2009, 106, (38), 16157-62. 16.

Guo, H. F.; Li, X.; Parker, M. W.; Waltenberger, J.; Becker, P. M.; Vander Kooi, C. W.

Mechanistic basis for the potent anti-angiogenic activity of semaphorin 3F. Biochemistry 21



Page 22 of 34

2013, 52, (43), 7551-8. 17.

Matthews, K. W.; Mueller-Ortiz, S. L.; Wetsel, R. A.

Carboxypeptidase N: a

pleiotropic regulator of inflammation. Mol Immunol 2004, 40, (11), 785-93. 18.

Dick, L. W., Jr.; Qiu, D.; Mahon, D.; Adamo, M.; Cheng, K. C.

C-terminal lysine

variants in fully human monoclonal antibodies: investigation of test methods and possible causes. Biotechnology and bioengineering 2008, 100, (6), 1132-43. 19.

Hu, Z.; Zhang, H.; Haley, B.; Macchi, F.; Yang, F.; Misaghi, S.; Elich, J.; Yang, R.;

Tang, Y.; Joly, J. C.; Snedecor, B. R.; Shen, A.

Carboxypeptidase D is the only enzyme

responsible for antibody C-terminal lysine cleavage in Chinese hamster ovary (CHO) cells.

Biotechnology and bioengineering 2016, 113, (10), 2100-6. 20.

Baek, D. S.; Kim, Y. S.

Construction of a large synthetic human Fab antibody

library on yeast cell surface by optimized yeast mating. Journal of microbiology and

biotechnology 2014, 24, (3), 408-20. 21.

Baek, D. S.; Kim, Y. S.

Humanization of a phosphothreonine peptide-specific

chicken antibody by combinatorial library optimization of the phosphoepitope-binding motif.

Biochemical and biophysical research communications 2015, 463, (3), 414-20. 22.

Shin, S. M.; Choi, D. K.; Jung, K.; Bae, J.; Kim, J. S.; Park, S. W.; Song, K. H.; Kim,

Y. S.

Antibody targeting intracellular oncogenic Ras mutants exerts anti-tumour effects

after systemic administration. Nature communications 2017, 8, 15090. 23.

Li, Y.; Sousa, R.

Expression and purification of E. coli BirA biotin ligase for in

vitro biotinylation. Protein expression and purification 2012, 82, (1), 162-7. 24. D. J.

Vander Kooi, C. W.; Jusino, M. A.; Perman, B.; Neau, D. B.; Bellamy, H. D.; Leahy, Structural basis for ligand and heparin binding to neuropilin B domains. Proceedings

of the National Academy of Sciences of the United States of America 2007, 104, (15), 6152-7. 25.

Parker, M. W.; Xu, P.; Li, X.; Vander Kooi, C. W.

Structural basis for the selective

vascular endothelial growth factor-A (VEGF-A) binding to neuropilin-1. The Journal of

biological chemistry 2012, 287, (14), 11082-9. 26.

Appleton, B. A.; Wu, P.; Maloney, J.; Yin, J.; Liang, W. C.; Stawicki, S.; Mortara, K.;

Bowman, K. K.; Elliott, J. M.; Desmarais, W.; Bazan, J. F.; Bagri, A.; Tessier-Lavigne, M.; Koch, A. W.; Wu, Y.; Watts, R. J.; Wiesmann, C.

Structural studies of neuropilin/antibody

complexes provide insights into semaphorin and VEGF binding. EMBO J 2007, 26, (23), 4902-12. 22


Page 23 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


27.

Pang, H. B.; Braun, G. B.; Friman, T.; Aza-Blanc, P.; Ruidiaz, M. E.; Sugahara, K. N.;

Teesalu, T.; Ruoslahti, E.

An endocytosis pathway initiated through neuropilin-1 and

regulated by nutrient availability. Nature communications 2014, 5, 4904. 28.

Gluzman-Poltorak, Z.; Cohen, T.; Herzog, Y.; Neufeld, G.

Neuropilin-2 is a

receptor for the vascular endothelial growth factor (VEGF) forms VEGF-145 and VEGF-165 [corrected]. The Journal of biological chemistry 2000, 275, (24), 18040-5. 29.

Roth, L.; Prahst, C.; Ruckdeschel, T.; Savant, S.; Westrom, S.; Fantin, A.; Riedel, M.;

Heroult, M.; Ruhrberg, C.; Augustin, H. G.

Neuropilin-1 mediates vascular permeability

independently of vascular endothelial growth factor receptor-2 activation. Science signaling 2016, 9, (425), ra42. 30.

Zanuy, D.; Kotla, R.; Nussinov, R.; Teesalu, T.; Sugahara, K. N.; Aleman, C.; Haspel,

N.

Sequence dependence of C-end rule peptides in binding and activation of neuropilin-1

receptor. Journal of structural biology 2013, 182, (2), 78-86.

23



Page 24 of 34

Tables Table 1. Kinetic binding parameters for the interaction between Fc-fused peptides and recombinant NRP1-b1b2 and NRP2-b1b2 domain, as monitored by SPR.a Proteins (analyte)

NRP1-b1b2

NRP2-b1b2

Fc-fused peptides

−1 −1

−1

kon (M s )

KD (M)

koff (s )

(immobilized) V12

(4.04±0.75)×10

V12-14

(8.10±3.75)×10

V12-15

(7.16±1.15)×10

V12-33

(1.56±0.30)×10

V12-44

(4.43±2.16)×10

A22p

(6.30±2.21)×10

TPP11

(2.62±1.01)×10

V12

NBb

V12-33

(6.01±3.15)×10

A22p

(2.70±1.64)×10

TPP11

(2.37±0.89)×10

3 3 3 5 3 4 5

(3.02±0.05)×10 (2.00±0.46)×10 (1.64±0.56)×10 (2.71±1.23)×10 (2.96±0.66)×10 (3.62±2.19)×10 (7.31±3.44)×10

−3 −3 −3 −3 −3 −3 −3

NB 3 4 3

(1.90±1.05)×10 (2.58±1.53)×10 (2.78±2.16)×10

(7.61±1.53)×10 (2.74±0.84)×10 (2.26±0.56)×10 (1.74±0.58)×10 (7.42±2.20)×10 (5.77±2.59)×10 (2.79±0.96)×10

−3 −3 −3

(3.22±1.25)×10 (9.62±0.21)×10 (1.35±1.38)×10

Each value represents the mean ± s.d. of three independent experiments.

b

NB, no significant binding.


−7 −7 −8 −7 −8 −8

NB

a

24

−7

−7 −8 −6

Page 25 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure legends Figure 1.

Figure 1. Generation and characterization of Fc-fused non-CendR peptide, Fc-V12. (A) Schematic drawing of the Fc-fused non-CendR peptide library displayed on yeast surface and library screening strategy to isolate Fc-fused non-CendR peptides binding to the arginine-binding pocket on the b1 domain of NRP1 using biotinylated NRP1-b1b2 domain in the presence of non-biotinylated NRP1-b1b2-AAA mutant as a competitor. Single-letter codes for amino acids and nucleotides are used according to the IUPAC-IUB. (B) Schematic drawing of the finally isolated Fc-fused non-CendR peptide, Fc-V12. The sequence of the non-CendR peptide V12 is shown for comparison with the sequences of CendR peptides, A22p and TPP11. (C) Indirect ELISA to determine the binding selectivity of Fc-V12 (100 nM) to a plate coated with NRP proteins, as indicated. Fc (100 nM), Fc-A22p (10 nM), and FcTPP11 (10 nM) were included as controls. ***P < 0.001 versus Fc. (D) Flow cytometric analysis to determine the binding specificity of Fc-V12 to the surface-expressed NRP1 on HUVECs as compared with that of Fc-A22p and Fc-TPP11. After transfection with NRP1, NRP2, or unrelated (control) siRNA for 24 h, the cells were stained at 4 °C with indicated proteins (1 µM) prior to FACS analysis. The Y-axis indicates relative mean fluorescence intensity (MFI) of each sample after normalization of the MFI to that of Fc. ***P < 0.001 (E) Competitive ELISA of VEGF-A165 (1 nM) to plate-immobilized NRP1-b1b2 protein in the presence of increasing concentrations of Fc, Fc-TPP11, and Fc-V12. After competition, bound biotinylated VEGF-A165 was detected by SA-HRP. In (C-E), data represent the mean ± s.d. of triplicate samples. Results are representative of three independent experiments. 25



Figure 2.

Figure 2. Fc-V12 triggers the cellular internalization of NRP1, resulting in the increased extravasation into tumor tissues and tumor tissue penetration of the peptide along with the coadministered drug. (A) Co-localization with cell surface-expressed NRP1 (red) at 4 °C and NRP1-mediated internalization at 37 °C of Fc, Fc-TPP11, or Fc-V12 (green) in HUVECs, as assessed by confocal fluorescence microscopy. The cells were treated with Fc or Fc-fused peptides (1 µM) for 30 min at 4 or 37 °C. Nuclei were counterstained with Hoechst 33342 (blue). Image magnification, ×400; scale bars, 50 µm. (B) Transwell assay to assess permeability across HUVEC monolayer by FITC-dextran passage after stimulation of cells for 30 min with 1 µM or 100 nM of Fc and Fc-fused peptides and 50 ng/mL (approximately 1.3 nM) or 5 ng/mL (approximately 0.13 nM) of VEGF-A165. Data depict the mean ± s.d. of fold-increase over Fc-treated cells, performed in triplicates. Results are representative of three independent experiments. **P < 0.01. (C) Representative immunohistochemical images showing the distribution of i.v. injected Fc, Fc-TPP11, or Fc-V12 (10 mg/kg) within the tumor tissue in FaDu xenograft mice. After 24 h of circulation, tumor tissues were taken from mice and immunostained for human Fc (green) and CD31 (red). The white boxed-areas were enlarged for better visualization. Image magnification, 200×; scale bar, 100 µm. The bar graph presents the quantified human Fc-positive area (green) using ImageJ software. Error bars, ± s.d. of five random fields for each tumor (n = 3 per group). **P < 0.01, ***P < 0.001 versus Fc. (D, E) Representative immunohistochemical images (E) showing the distribution of DOX co-injected i.v. with Fc, Fc-TPP11, or Fc-V12 in FaDu tumor tissues excised from mice (n = 3 per group) following the treatment regimen described in (D). After the indicated treatments, tumor tissues were taken from mice and immunostained for blood vessels with an anti-CD31 (green) and the native fluorescence (red) was used to detect DOX. Image 26


Page 26 of 34

Page 27 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


magnification, 200×; scale bar, 100 µm.

27



Figure 3.

Figure 3. Mutagenesis study to determine essential elements of Fc-V12 for NRP1 binding. (A) Non-reducing SDS-PAGE analysis of purified Fc-V12 and its mutants with the indicated substituted amino acid residue at the last C-terminal Val of Fc-V12. About 2 µg of each protein was analyzed on 16% SDS Bis-Tris gels stained with Coomassie brilliant Blue. (B) Flow cytometric analysis to determine the binding activity of the indicated Fc-V12 mutants to the NRP1-expressing HUVECs in comparison with Fc-V12, Fc-A22p, and Fc-TPP11. HUVECs were stained at 4 °C with indicated proteins (1 µM) prior to FACS analysis. The Yaxis indicates relative MFI of each sample after normalization of the MFI to that of Fc. Data represent the mean ± s.d. of triplicate samples. Results are representative of three independent experiments. ***P < 0.001 versus Fc. (C) Indirect ELISA to determine whether the last five C-terminal residues (8KPPRV12) of Fc-V12 may confer the binding specificity to NRP1, as performed in Figure 1C. The upper panel shows GS-V12 sequence, which has 1GSGSGSG7 sequence while retaining 8KPPRV12 sequence of V12. Data represent the mean ± s.d. of duplicate samples. Results are representative of two independent experiments. ***P < 0.001 versus Fc-V12; n.s, not significant.

28


Page 28 of 34

Page 29 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4.

Figure 4. Isolation of affinity-matured Fc-V12 derivatives and their characterization. (A) Sequence alignment of isolated non-CendR V12 variants from screening of Fc-V12 mutant library with randomization of the upstream seven residues (1HLQESPG7) of V12 for affinity maturation against NRP1. Lined region of V12 were diversified. (B) Flow cytometric analysis to compare the binding specificity of Fc-V12 variants to the surface-expressed NRP1 on HUVECs with that of Fc-V12 and Fc-TPP11. The experiments and data analysis were performed as described in Figure 1D, with an exception of 100 nM proteins used instead of 1 µM. *P < 0.05, **P < 0.01, ***P < 0.001. (C) Competitive ELISA of VEGF-A165 (1 nM) to plate-coated NRP1-b1b2 protein in the presence of increasing concentrations of indicated proteins. After competition, bound biotinylated VEGF-A165 was detected by SA-HRP. (D) Transwell assay to assess permeability across HUVEC monolayer by FITC-dextran passage after stimulation of the cells for 30 min with 100 nM of Fc and Fc-fused peptides and 50 ng/mL (approximately 1.3 nM) of VEGF-A165. Fc-V12-33 was further characterized on HUVECs, transfected with a scrambled siRNA (control) or NRP1 siRNA for 24 h before treatment. **P < 0.01, ***P < 0.001 versus cells stimulated with Fc-V12-33; n.s, not significant. In (B-D), data represent the mean ± s.d. of triplicate wells. Results are representative of three independent experiments.

29



For Table of Contents/Abstract graphic 65x34mm (300 x 300 DPI)


Page 30 of 34

Page 31 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Generation and characterization of Fc-fused non-CendR peptide, Fc-V12. 175x99mm (300 x 300 DPI)



Figure 2. Fc-V12 triggers the cellular internalization of NRP1, resulting in the increased extravasation into tumor tissues and tumor tissue penetration of the peptide along with the co-administered drug. 175x94mm (300 x 300 DPI)


Page 32 of 34

Page 33 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Mutagenesis study to determine essential elements of Fc-V12 for NRP1 binding. 140x94mm (300 x 300 DPI)



Figure 4. Isolation of affinity-matured Fc-V12 derivatives and their characterization. 140x121mm (300 x 300 DPI)


Page 34 of 34

Immunoglobulin Fc-Fused Peptide without C-Terminal Arg or Lys

Recommend Documents