Cancer Research Annual Meeting 2010 Jordan
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 67, 9623, October 15, 2007. doi: 10.1158/0008-5472.CAN-07-2737
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jaalouk, D. E.
Right arrow Articles by Arap, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jaalouk, D. E.
Right arrow Articles by Arap, W.

Priority Reports

The Original Pathologische Anatomie Leiden-Endothelium Monoclonal Antibody Recognizes a Vascular Endothelial Growth Factor–Binding Site within Neuropilin-1

Diana E. Jaalouk1, Michael G. Ozawa1, Jessica Sun1, Johanna Lahdenranta1, Reinier O. Schlingemann2, Renata Pasqualini1 and Wadih Arap1

1 The University of Texas M.D. Anderson Cancer Center, Houston, Texas and 2 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands

Requests for reprints: Wadih Arap or Renata Pasqualini, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: 713-792-3873; Fax: 713-745-2999; E-mail: warap{at}mdanderson.org or rpasqual{at}mdanderson.org.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
For two decades, the antigen recognized by the Pathologische Anatomie Leiden-Endothelium (PAL-E) monoclonal antibody, a standard vascular endothelial cell marker, has remained elusive. Here, we used a combinatorial phage display–based approach ("epitope mapping") to select peptides binding to the original PAL-E antibody. We found that a subset of the selected panel of peptides had motifs with strong homology to an exposed site within the b1 domain of human neuropilin-1 (NRP-1). We confirmed peptide binding by ELISA and by surface plasmon resonance. We also showed that the PAL-E antigen colocalizes with NRP-1 staining in endothelial cells. Crystal structure of the b1 domain in NRP-1 suggests that the PAL-E binding site overlaps with a vascular endothelial growth factor (VEGF) binding site. Taken together, these results indicate that NRP-1 is an endothelial cell antigen recognized by the true PAL-E antibody. The consistent biochemical, morphologic, and functional features between the PAL-E antigen and NRP-1 support our interpretation. Given that NRP-1 is a VEGF receptor, these results explain the attributes of the PAL-E antibody as a marker of vascular permeability and angiogenesis. [Cancer Res 2007;67(20):9623–9]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The Pathologische Anatomie Leiden-Endothelium (PAL-E) antibody, generated over 20 years ago by injection of human melanoma lymph node metastasis into mice, has since been used as a vascular endothelial cell marker. PAL-E recognizes vascular endothelial cells lining blood capillaries, venules, and small-to-medium sized veins in human frozen tissue sections. On the other hand, it reacts only weakly (if at all) with endothelium of arterioles, arteries, and large veins and does not recognize lymphatic endothelium. Because of these attributes, the PAL-E antibody has become useful (a) for delineating phenotypic differences between distinct vascular beds, (b) for understanding the endothelial molecular diversity in diseases with an angiogenic component—such as cancer, proliferative retinopathy, or rheumatoid arthritis; and (c) for ruling out the predominance of lymphatic endothelium (1, 2).

Indeed, because the PAL-E antibody shows high reactivity to venules in lymph nodes and to tumor blood vessels, it has been extensively used for histopathologic analysis of the endothelium lining tumor vasculature in several premalignant and malignant conditions, including carcinomas and sarcomas (3–5). Moreover, the PAL-E antibody has long served to study the new blood vessel formation that occurs during tumor metastasis and wound healing (6). Notably, unlike endothelial cells in normal brain (with an intact blood-brain barrier), the endothelium in primary malignant gliomas and in secondary brain metastases stain positively with PAL-E, indicating that its reactivity may somehow relate to the altered vascular permeability state observed in angiogenesis (7).

Further insight into the functional relevance of endothelial cell staining by PAL-E antibody was gained by immunoelectron microscopy studies that showed that the PAL-E antibody stains the luminal endothelial surface in a local pattern and is mostly associated with endothelial pinocytotic plasmalemmal vesicles (6, 8). Given that such vesicles (also termed caveolae) function in trans-cellular transport, they are scarce in endothelium, separating blood from intact "sanctuary" tissue sites, such as the blood-brain or the blood-retinal barrier. Thus, studies on the subcellular localization of PAL-E staining have again suggested a role for PAL-E antigen in general trans-endothelial transport and vascular permeability in diverse pathologic conditions varying from inflammatory diseases to malignant tumors to organ transplantation rejection (9–11).

The widespread application of the PAL-E antibody as a marker of vascular endothelium has led to efforts in identifying its corresponding antigen(s); unfortunately, the identification of the elusive "PAL-E antigen" has proved challenging for the past two decades, largely because this unique monoclonal antibody can only be used under a relatively limited set of experimental conditions.

Recently, Xu et al. (12) have attempted to fulfill this objective by combining protein purification to tandem mass spectrometry analysis of tryptic peptides. Although these investigators proposed a secreted isoform of vimentin as a candidate antigen recognized (by invoking the presence of a new so-called "PAL-E–reactive vimentin"), they have actually used and reported (12) on a different isotype from the original PAL-E monoclonal antibody (1).

On the other hand, Niemelä et al. (13) have shown that the plasmalemmal vesicle 1/fenestrated endothelial-linked structure (PV-1/FELS) protein is the antigen recognized by 174/2, a monoclonal antibody generated by immunizing BALB/c mice with vessels from human lymph nodes and fusing lymphocytes with myeloma cells. Next, based on immunostaining and biochemical experiments, the authors have surmised that 174/2 acts as a "PAL-E–like" monoclonal antibody; as such, they have plausibly proposed PV-1/FELS as another PAL-E candidate antigen (13). However, neither monoclonal antibody inhibits the other in competitive staining experiments (13), bringing such data interpretation into question.

As the identity of the PAL-E monoclonal antibody originally generated by Schlingemann et al. (1) evidently still remains elusive, we used an entirely different approach for the identification of the PAL-E antigen. By screening a phage display random peptide library on the true PAL-E monoclonal antibody, we selected and isolated PAL-E–binding peptides. We reasoned that such selected peptide motifs would mimic the antigenic binding site recognized by the antibody. Indeed, combinatorial epitope mapping of monoclonal antibodies has long served to reveal the identity of target antigens (14) and our group has even adapted this methodology to fingerprint the circulating pool of patient-derived antibodies (15, 16).

Here we show (a) that PAL-E–binding consensus peptide motifs reveal an exposed protein site in neuropilin-1 (NRP-1), (b) that the original PAL-E antibody binds to NRP-1 and colocalizes with NRP-1 staining in endothelial cells, and (c) that the PAL-E binding site within NRP-1 does overlap with a vascular endothelial growth factor (VEGF) binding site. Together, these data indicate that NRP-1 is an endothelial cell antigen recognized by the original PAL-E monoclonal antibody. Given that NRP-1 is a VEGF receptor, these results may explain the attributes of PAL-E as a marker of vascular permeability and angiogenesis.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Antibodies, recombinant proteins, and synthetic peptides. Schlingemann et al. (1) generated the original hybridoma corresponding to the monoclonal antibody PAL-E (an isotype IgG2a mouse clone) used throughout this study. A control mouse monoclonal antibody with the same isotype as PAL-E (mIgG2a) and monoclonal antibodies against NRP-1 and NRP-2 were purchased (DAKO and R&D Systems). Polyclonal antibodies against the synthetic cyclic peptide CTQYAMHLC were generated by serial immunization of rabbits with keyhole limpet hemocyanin (KLH)–conjugated peptide (17). Standard IgG purification from rabbit serum (preimmune and anti-CTQYAMHLC IgGs) was accomplished by using Protein A Sepharose Cl-4B resin (Amersham Biosciences). Recombinant NRP-1, NRP-2, semaphorin-3A (Sema3A), and VEGF-A165 were commercially obtained (R&D Systems). Unless otherwise specified, all peptides were synthesized, cyclized, and purified at AnaSpec.

Cell culture. Human umbilical vascular endothelial cells (HUVEC) were from Clonetics and cultured in complete endothelial cell basal medium from Cambrex at 37°C in 5% CO2. To minimize tissue culture drifting, only early passage cells (defined as passages 3 and 4) were used.

Selection of combinatorial peptide libraries on immobilized PAL-E. We used a phage display library with an insert displaying random peptides with the general cyclic arrangement CX7C (C, cysteine; X, any amino acid residue). Increasing concentrations of PAL-E and control mIgG2a were immobilized in Protein A–coated strip 96-well plates (Pierce) at 4°C overnight. PBS containing 3% bovine serum albumin (BSA) was used for blocking at 37°C for 1 h. After two washing steps with PBS, 2 x 109 transducing units (TU) of CX7C phage library were precleared on mIgG2a at room temperature for 1 h, then incubated with the original PAL-E monoclonal antibody (1) for 3 h after which unbound phage was discarded. After multiple washes in PBS, PAL-E–bound phage was rescued by infection with 200 µL of stationary-phase K91Kan Escherichia coli at room temperature for 1 h and then amplified by incubation in 20 mL Luria-Bertani (LB) broth supplemented with 0.2 µg/mL tetracycline and 100 µg/mL kanamycin at 37°C/shaker (225 rpm) for 1 h. Serial dilutions were then plated in triplicates on LB plates supplemented with tetracycline and kanamycin, then incubated at 37°C overnight. Colony counting was done the next day to determine the number of TU. Multiple rounds of panning were done in a similar manner so that the phage population recovered from one round was used for the subsequent panning round and more stringent washing was applied as the panning progresses.

PCR and sequencing. Colonies from third round of panning on PAL-E antibody were transferred into 30 µL of double-distilled water containing 10% glycerol and stored at –20°C for PCR. PCR mix was prepared such that a 20 µL reaction sample included 2 µL bacteria preserved in double-distilled water containing 10% glycerol, 8 pmol of each of fUSE5 forward primer (5'-TAATACGACTCACTATAGGGCAAGCTGATAAACCGATACAATT-3') and reverse primer (5'-CCCTCATAGTTAGCGTAACGATCT-3'), 1 µL of 2.5 mmol/L deoxynucleotide triphosphate mix (Promega), 2 µL of 10x Taq polymerase buffer (Promega), and 2.5 units of Taq DNA polymerase (Promega). PCR was done by using 94°C annealing for 3 min, followed by 30 cycles of 94°C for 10 s, 60°C for 30 s, and 72°C for 30 s that were followed by 72°C for 5 min. PCR products were verified by DNA sequencing.

Phage binding assays. PAL-E or mIgG2a were coated at 5 µg in 50 µL PBS per well in Protein A–precoated plates (Pierce) at 4°C overnight. Similarly, recombinant Sema3A or VEGF-A165 were coated at 0.25 µg/well. PBS containing 3% BSA served for blocking at 37°C for 1 h followed by two tandem PBS washing steps. Control Fd-tet (insertless) phage or selected PAL-E–binding phage clones were added at 109 TU/well and incubated at room temperature for 3 h. After multiple wash steps with PBS, bound phage were recovered by infection in 200 µL of stationary-phase K91Kan E. coli at room temperature for 1 h and amplified by overnight culture in 20 mL LB supplemented by tetracycline and kanamycin. Serial dilutions were then plated in triplicates on LB plates supplemented by tetracycline and kanamycin and incubated at 37°C for colony counts overnight to determine the number of TU.

ELISA. Recombinant proteins and synthetic peptides were coated at 0.5 µg/well in 96-well plates (Maxi Sorp, Nunc) at 4°C overnight. PBS containing 3% BSA was used for blocking at room temperature for 1 h. Unless otherwise specified, primary antibody was added at 25 ng/µL in PBS containing 1% BSA and incubated for 2 h at room temperature. After three washing steps with PBS containing 1% BSA plus 0.05% Tween 20 (Sigma), and 100 µL/well of either 1:2,000 dilution of horseradish peroxidase (HRP)–conjugated rabbit anti-total mouse IgG (Zymed) or 1:4,000 dilution of HRP-conjugated goat anti-rabbit IgG (Zymed) prepared in PBS containing 1% BSA was incubated at room temperature for 1 h. Secondary antibody binding was followed by three washing steps as described above, then the signal was developed by using 100 µL/well of TMB soluble substrate (Calbiochem) that was incubated at room temperature while shaking at 75 rpm for 10 min. To stop the reaction, 50 µL of 0.5 N H2SO4 was used per well. Absorbance was determined at 450 nm (Power WaveX340; BIO-TEK Instruments).

Surface plasmon resonance. PAL-E antibody was captured at 5 µg/mL to an anti-mouse-Fc–coated CM5-chip (Biacore, Inc.) by injecting it for 15 min over the chip at 10 µL/min flow rate. Interactions were allowed to stabilize for 5 min then 10 µg/mL of NRP-1 or NRP-2 were captured to PAL-E for 15 min. After injection, dissociation was followed for 5 min. Chips were regenerated by 10 mmol/L glycine (pH 1.7) at 100 µL/min.

Cell staining. HUVEC cells were plated at 104 per well in a chamber slide (Lab-Tek II; Nalge Nunc International) in culture medium and allowed to adhere at 37°C, 5% CO2 for 48 h. Cells were washed twice with PBS, then fixed and permeabilized with precooled acetone at –20°C for 10 min. After three PBS washing steps, 5% normal goat serum (NGS) was used for blocking at room temperature for 20 min. Then, 5 µg/250 µL 1% NGS of either mIgG2a or PAL-E were incubated at 4°C overnight. After washing with PBS, cells were incubated with 1:400 µL/well of Cy3-conjugated goat anti-mouse IgG (Jackson ImmunoResearch) at room temperature for 30 min. Vectashield Mounting Medium supplemented with 4',6-diamidino-2-phenylindole (DAPI; Vector Labs, Inc.) was used for nuclear staining and to mount the slides with coverslips. Slides were then evaluated by using a fluorescent microscope and images were acquired at x200 magnification with the Magna Fire IX70 software.

Statistical analysis. Results are expressed as the mean of three independent experiments ± SEM. Student's t test was used for generation of P values and determination of statistical significance.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Combinatorial peptide selection on the original PAL-E monoclonal antibody. To identify peptides that bind to the variable fragment of the PAL-E monoclonal antibody, we used a CX7C peptide library to profile the surface of PAL-E IgGs immobilized onto Protein A–coated wells via their Fc region. After the third round of serial selection (Fig. 1A ), we observed marked phage binding to PAL-E compared with negative controls (91-fold relative to BSA and 16-fold relative to mIgG2a isotype control). We next did a comprehensive protein homology analysis of peptide sequences selected and isolated (N = 85) from the third round of panning on the PAL-E antibody (Supplementary Table S1). We identified the SQYSTNW motif spanning residues 295 to 301 in the b1 domain of NRP-1 as a strong match such that several PAL-E–binding phage clones share ≥50% homology with this motif (Supplementary Table S2).


Figure 1
View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Phage panning on PAL-E antibody and matched sequence alignment. A, multiple rounds of panning on PAL-E antibody with CX7C random phage library resulted in significant enrichment with 91-fold stronger phage binding to PAL-E compared with BSA control after round III. B, validation of PAL-E phage clones with NRP-1 sequence homology by phage binding assay (n = 3). Compared with Fd-tet insertless phage, all PAL-E phage clones showed strong and specific binding to PAL-E antibody compared with mouse IgG2a isotype or to BSA control. Phage clone with the cyclic insert sequence CSQWNMLLC exhibited the strongest binding to PAL-E antibody with (93.2 ± 15.7) x 104 TU, which is significantly higher compared with (0.5 ± 0.2) x 104 TU with Fd-tet control phage. Columns, average; bars, SEM (P = 0.01).

 
Next, to validate the selected peptide motifs, we evaluated the binding of three individual phage clones (displaying the inserts CSQYSFNWC, CTQYAMHLC, and CSQWNMLLC) to the immobilized PAL-E monoclonal antibody. In comparison with control insertless phage, all three selected phage clones showed marked binding to PAL-E antibody relative to BSA and mIgG2a (Fig. 1B). By arbitrarily setting the binding of the negative control (insertless phage) to PAL-E to one relative TU, CSQYSFNWC-displaying phage yielded 68 ± 18 relative TU, CTQYAMHLC-displaying phage yielded 108 ± 44 relative TU, and CSQWNMLLC-displaying phage yielded 186 ± 32 relative TU (all representing mean ± SEM). These results establish that peptide motifs with NRP-1 homology mediate binding to the PAL-E monoclonal antibody and suggest that PAL-E may recognize NRP-1.

PAL-E binds to NRP-1. We used ELISA to determine whether the PAL-E monoclonal antibody binds to recombinant NRP-1 in vitro (Fig. 2A ). We also evaluated whether PAL-E binds to recombinant NRP-2, a related receptor that shares 44% identity with NRP-1 at the protein sequence level (Fig. 2B). Relative to negative control IgG2a, PAL-E reacted strongly against NRP-1 (A450 = 2.77 ± 0.09; t test, P = 0.002). Moreover, PAL-E showed a much weaker binding to NRP-2 (A450 = 0.69 ± 0.05; t test, P = 0.005). We next used surface plasmon resonance–based technology (Biacore) as a second independent biochemical method to confirm PAL-E binding to NRP-1. We again observed strong binding of the PAL-E antibody to NRP-1 (Fig. 2C). Consistently with our ELISA data showing much lower affinity of PAL-E for NRP-2, no significant binding to NRP-2 was detected by surface plasmon resonance methodology.


Figure 2
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. PAL-E antibody recognizes recombinant NRP-1. A and B, assessing the binding of PAL-E to recombinant NRP-1 to recombinant NRP-2 by ELISA (n = 3). Either recombinant protein was coated at 0.5 µg/well. PAL-E antibody, the IgG2a isotype control, anti–NRP-1, or anti–NRP-2 was assayed at 25 ng/µL. Compared with IgG2a, PAL-E reacted positively with both NRP-1 (P = 0.002) and NRP-2 (P = 0.005) with much higher absorbance measurements with NRP-1. C, assessing the binding of PAL-E to recombinant NRP-1 and NRP-2 by Biacore. PAL-E (5 µg/mL) was captured to an anti–mouse-Fc coated CM5-chip. NRP-1 and NRP-2 (10 µg/mL each) were captured to PAL-E. The results show strong binding of PAL-E antibody to NRP-1, whereas no significant binding to NRP-2 was detected.

 
PAL-E staining colocalizes with NRP-1. Furthermore, acetone-fixed HUVEC cells stained positively with PAL-E (Fig. 3A ). Comparison of double staining in HUVEC cells shows that the fluorescence pattern resulting from PAL-E immunostaining colocalizes to a great extent with that resulting from immunostaining with a commercial anti–NRP-1 antibody, suggesting that PAL-E can also bind to cellular NRP-1 (Fig. 3B). Taken together, these data show that NRP-1 is an antigenic target for PAL-E antibody likely through direct recognition of the primary amino acid sequence comprising the SQYSTNW motif. Alternatively, PAL-E may recognize secondary or higher-order protein structure around this motif.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. PAL-E stains NRP-1 in HUVEC cells. A, acetone-fixed HUVEC cells stain positively with PAL-E. B, immunofluorescence patterns indicate that PAL-E staining colocalizes with NRP-1 staining in HUVEC cells adherent in chamber slides. Goat anti–mouse-FITC or Cy3 was used as secondary antibody. Vectashield with DAPI was used for mounting and nuclear staining.

 
Identification of the antigenic epitope in NRP-1 that is recognized by PAL-E. We used the synthetic peptides CSQYSFNWC, CTQYAMHLC, and CSQWNMLLC corresponding to the cyclic insert sequences in PAL-E–binding phage clones to assess their respective binding to PAL-E antibody by ELISA (data not shown). Relative to IgG2a isotype control, PAL-E reacted positively with CTQYAMHLC peptide, resulting in an average absorbance of 0.64 ± 0.02 (P = 0.002). PAL-E showed no significant binding to either CSQYSFNWC or CSQWNMLLC peptide. To evaluate whether the CTQYAMHLC motif represents the antigenic epitope in NRP-1 that is targeted by PAL-E, we generated polyclonal anti-CTQYAMHLC antibodies by immunization of rabbits with the KLH counterpart of the peptide. Relative to preimmune sera, the resultant sera were first analyzed for their specific binding to their respective peptide antigen (Fig. 4A ). The unrelated CGSPGWVRC peptide was used as control. Then, we assessed whether purified polyclonal anti-CTQYAMHLC IgGs can bind to recombinant NRP-1 (Fig. 4B). Results show that relative to preimmune sera, anti-CTQYAMHLC IgGs bind positively to NRP-1 with a mean absorbance of 0.90 ± 0.05 (P = 0.004). Interestingly, much stronger binding of anti-CTQYAMHLC IgGs to recombinant NRP-2 was observed with a mean absorbance of 2.11 ± 0.12 (P = 0.003), indicating a stronger affinity of the polyclonal antibody to NRP-2 (Fig. 4C). Together, these findings suggest that the CTQYAMHLC motif or its homologous domain in NRP-1 represents, at the least in its secondary peptide structure, the antigenic epitope that is targeted by PAL-E.


Figure 4
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Identification of the PAL-E–binding epitope in NRP-1. A, rabbit anti-CTQYAMHLC sera strongly and specifically react with the corresponding cyclic peptide (n = 3). The unrelated CGSPGWVRC peptide was used as control. B and C, purified polyclonal anti-CTQYAMHLC IgGs bind to both recombinant NRP-1 (P = 0.004) and NRP-2 (P = 0.003) with stronger affinity to NRP-2. Either recombinant protein was coated at 0.1 µg/well. Preimmune and anti-CTQYAMHLC IgGs purified on Protein A Sepharose Cl-4B resin were assayed at 3 ng/µL.

 
PAL-E binding epitope in NRP-1 overlaps with the VEGF-A165 binding site. The crystal structure of the b1 domain in NRP-1 (Fig. 5A ) shows that the SQYSTNW motif (in blue; side chain amino acid residues in yellow) corresponding to the PAL-E binding site in NRP-1 is well exposed on the outer surface of the receptor (18). This suggests that it is accessible not only to circulating antibodies but also to receptor-binding ligands. Because it is known that the NRP-1 ligand VEGF-A165 binds to the b1 domain (19), we reasoned that the PAL-E binding site may actually overlap with the VEGF-A165 binding region. Thus, we evaluated whether PAL-E–binding phage clones displaying peptide motifs that share homology with NRP-1 also bind to VEGF-A165. Relative to Fd-tet insertless phage control, the clones with the peptide insert sequences CSQYSFNWC and CTQYAMHLC showed 73 ± 10-fold and 24 ± 3-fold more binding to human VEGF-A165, respectively. This interaction seems to be specific because binding of these phage clones to another NRP-1 ligand (human Sema3A) was negligible (Fig. 5B). These data strongly suggest that the PAL-E binding site in NRP-1 overlaps with the binding site for VEGF-A165.


Figure 5
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. PAL-E–binding epitope in NRP-1 maps to the VEGF165 binding site. A, PAL-E binding site in NRP-1. A ribbon diagram resembling the crystal structure of the b1 domain in NRP-1 (red; http://www.rcsb.org/pdb/cgi/explore.cgi, pdbId=1KEX). The SQYSTNW motif corresponding to the PAL-E binding site in NRP-1 (blue). The side-chain amino acid residues (yellow). B, select phage clones that bind to PAL-E also bind to the NRP-1 ligand VEGF165. Compared with Fd-tet insertless phage, clones with the insert sequences CSQYSFNWC and CTQYAMHLC showed strong binding to recombinant human VEGF-A165. Recombinant human Sema3A, another NRP-1 ligand, was used as control.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
To identify the PAL-E antigen, we used a phage display–based combinatorial approach to select peptide motifs that preferentially bind to the original PAL-E antibody. We found that these mimics of the antigenic epitope share strong homology with NRP-1. Results from both ELISA-based and surface plasmon resonance–based assays show that PAL-E indeed does bind to NRP-1. Consistently, we show the colocalization of PAL-E and anti–NRP-1 staining in endothelial cells. Notably, the functional and expression profiles of NRP-1 are entirely harmonious with the fact that the endothelial cell entity targeted by PAL-E must play an important role in vascular leakage and angiogenesis. NRP-1 (also known as Npn-1, NP-1, A5-antigen) is a transmembrane glycoprotein that serves versatile roles in the nervous and in the cardiovascular systems mediated by an ability to bind distinct ligands via partnering with other receptors (20). In the developing nervous system, NRP-1 forms functional multireceptor complexes along with NRP-2 and plexins and binds with high affinity to the three classes of structurally related semaphorins (namely semaphorin-3, semaphorin-IV, and semaphorin-E), which function in neuronal guidance and axonal outgrowth (21, 22). In the cardiovascular system, NRP-1 partners with VEGFR-1 and VEGFR-2 and binds to several VEGF family members, including VEGF-A165, VEGF-B, VEGF-E, and placental growth factor-2. Consequently, NRP-1 plays a central role in cardiovascular development (20, 23–25). Finally, in endothelial cells, NRP-1 amplifies the mitogenic and chemotactic effects of VEGF-A165 by allowing it to bind more efficiently to VEGF receptor-2, resulting in endothelial cell survival, activation, and angiogenesis (26).

In agreement with the PAL-E staining profile, VEGF-A and its cognate receptors are only weakly expressed in quiescent endothelium but are up-regulated in activated endothelium under inflammatory, ischemic, or malignant disease (27). Moreover, increases in PAL-E reactivity in leaky retinal endothelium strongly correlate with increased local expression of VEGF-A and its receptor VEGFR-2 (28). It is also worth noting that VEGF regulation of vascular permeability is mediated in part by the formation of cell membrane fenestrations associated with caveolae (29); such membrane vesicle structures stain strongly positive with the PAL-E antibody in the vascular endothelium (1, 6, 8). Interestingly, both the PAL-E antigen and NRP-1 are differentially expressed between blood vessel and lymphatic endothelium in that NRP-1 does not partner with VEGFR-3, which is mainly expressed on lymphatic endothelial cells (30, 31). With regard to PAL-E as a tumor cell marker, several studies have shown increased expression of NRP-1 in human prostate, breast, and colorectal cancer cells (32, 33). In these reports, NRP-1 expression correlates with tumor metastasis and aggressiveness by promoting tumor cell growth, survival, invasiveness, and tumor-related angiogenesis.

At a protein-protein biochemical interaction level, we were also able to map the residues likely to comprise a PAL-E binding site. Because the polyclonal antibody raised against the identified motif could bind to both NRP-1 and to the related receptor NRP-2, one might suggest that PAL-E recognizes a secondary structure at the binding site rather than the primary protein sequence. Moreover, our binding results indicate that the PAL-E binding site in NRP-1 overlaps with a VEGF ligand-receptor site. It is known that the binding specificities of the various ligands to NRP-1 are determined by its 860-residue extracellular fragment composed of two tandem repeats of each of complement-like CUB motifs (a1/a2 domains) and coagulation factor V/VIII–like motifs (b1/b2 domains) followed by a short COOH-terminal domain that contains the meprin-like and µphosphatase-like MAM motif. Studies with NRP-1 deletions have shown that Sema3A binding to NRP-1 is mediated through the a1/a2 and b1/b2 domains, whereas only the b1/b2 domains are required for binding of VEGF-A165 (19). It is also known that the b1/b2 domains contain an 18-residue stretch that mediates the heterophilic cell-cell adhesion activity of NRP-1; however, this site is evidently distinct from binding sites for Sema3A and VEGF-A165 (34). Our own results show that the SQYSTNW motif (corresponding to the PAL-E binding site within the b1 domain of NRP-1) binds to recombinant VEGF-A165 but not to Sema3A, suggesting that the PAL-E antigenic epitope on NRP-1 may encompass, overlap, or contain the VEGF-A165 binding site. Consistently, the SQYSTNW motif actually maps to the ligand binding site predicted from the crystal structure analysis of the b1 domain within human NRP-1 (18). From detailed analysis of this structure, the ligand-binding site consists of a polar cleft in the b1 domain that is formed by six juxtaposed loops (L1–L6) flanked by electronegative surfaces. In this context, we found that the SQYSTNW motif (spanning residues 295–301) comprises the first loop in the structure in which the Trp301 residue falls deep within the formed cleft. Additionally, Tyr297, which lies in perpendicular orientation to Trp301, decorates the side of the cleft along with Ser298, Thr299, and Asn300 residues in L1. However, the full understanding of the interaction attributes among the receptor, corresponding known native ligands, synthetic peptides, and PAL-E will have to await elucidation of X-ray crystal structures of the protein complexes. Of course, one cannot entirely rule out the possibility that the true PAL-E antibody may recognize more than one antigen (12, 13), perhaps in different settings. Indeed, cross-reaction of monoclonal antibodies with multiple antigens has long been shown (35, 36). Be that as it may, the common biochemical, morphologic, and functional features between PAL-E and NRP-1 support the interpretation of our experimental findings.

In summary, we show that a motif within a VEGF binding site in NRP-1 is an antigenic epitope recognized by the original PAL-E antibody. Our findings and reagents enable a suitable interpretation of PAL-E staining of the vascular endothelium and may provide mechanistic insights into the design of new small molecule peptidomimetic leads and other monoclonal antibodies targeting VEGF-mediated pathways.


   Acknowledgments
 
Grant support: Department of Defense and the NIH, awards from the Gillson-Longenbaugh Foundation (W. Arap and R. Pasqualini), and the Kimberly Patterson Fellowship in Leukemia Research (D.E. Jaalouk).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Drs. Ricardo Giordano and Paul Mintz for technical advice and helpful insights.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Received 7/18/07. Accepted 8/ 3/07.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Schlingemann RO, Dingjan GM, Emeis JJ, Blok J, Warnaar SO, Ruiter DJ. Monoclonal antibody PAL-E specific for endothelium. Lab Invest 1985;52:71–6.[Medline]
  2. Paavonen K, Mandelin J, Partanen T, et al. Vascular endothelial growth factors C and D and their VEGFR-2 and 3 receptors in blood and lymphatic vessels in healthy and arthritic synovium. J Rheumatol 2002;29:39–45.[Abstract/Free Full Text]
  3. Nadimi H, Saatee S, Armin A, Toto PD. Expression of endothelial cell markers PAL-E and EN-4 and Ia-antigens in Kaposi's sarcoma. J Oral Pathol 1988;17:416–20.[CrossRef][Medline]
  4. Nakamura S, Muro H, Suzuki S, et al. Immunohistochemical studies on endothelial cell phenotype in hepatocellular carcinoma. Hepatology 1997;26:407–15.[CrossRef][Medline]
  5. Schlingemann RO, Bots GThAM, van Duinen SG, Ruiter DJ. Differential expression of endothelium-specific antigen PAL-E in vasculature of brain tumors and preexistent brain capillaries. Ann N Y Acad Sci 1988;529:111–4.[CrossRef]
  6. Schlingemann RO, Rietveld FJ, Kwaspen F, van de Kerkhof PC, de Waal RM, Ruiter DJ. Differential expression of markers for endothelial cells, pericytes, and basal lamina in the microvasculature of tumors and granulation tissue. Am J Pathol 1991;138:1335–47.[Abstract]
  7. Leenstra S, Das PK, Troost D, Bosch DA, Claessen N, Becker AE. PAL-E, monoclonal antibody with immunoreactivity for endothelium specific to brain tumours. Lancet 1990;335:671.[CrossRef][Medline]
  8. Sauter B, Foedinger D, Sterniczky B, Wolff K, Rappersberger K. Immunoelectron microscopic characterization of human dermal lymphatic microvascular endothelial cells. Differential expression of CD31, CD34, and type IV collagen with lymphatic endothelial cells vs blood capillary endothelial cells in normal human skin, lymphangioma, and hemangioma in situ. J Histochem Cytochem 1998;46:165–76.[Abstract/Free Full Text]
  9. Hengstenberg C, Rose ML, Olsen EG, Maisch B. Immune response to the endothelium in myocarditis, dilated cardiomyopathy and rejection after heart transplantation. Eur Heart J 1991;12:144–6.[Abstract/Free Full Text]
  10. Rose M, Page C, Hengstenberg C, Yacoub M. Immunocytochemical markers of activation in cardiac transplant rejection. Eur Heart J 1991;12:147–50.[Abstract/Free Full Text]
  11. Hofman P, Blaauwgeers HG, Vrensen GF, Schlingemann RO. Role of VEGF-A in endothelial phenotypic shift in human diabetic retinopathy and VEGF-A-induced retinopathy in monkeys. Ophthalmic Res 2001;33:156–62.[CrossRef][Medline]
  12. Xu B, deWaal RM, Mor-Vaknin N, Hibbard C, Markovitz DM, Kahn ML. The endothelial cell-specific antibody PAL-E identifies a secreted form of vimentin in the blood vasculature. Mol Cell Biol 2004;24:9198–206.[Abstract/Free Full Text]
  13. Niemelä H, Elima K, Henttinen T, Irjala H, Salmi M, Jalkanen S. Molecular identification of PAL-E, a widely used endothelial-cell marker. Blood 2005;106:3405–9.[Abstract/Free Full Text]
  14. Scott JK, Smith GP. Searching for peptide ligands with an epitope library. Science 1990;249:386–90.[Abstract/Free Full Text]
  15. Mintz PJ, Kim J, Do KA, et al. Fingerprinting the circulating repertoire of antibodies from cancer patients. Nat Biotechnol 2003;21:57–63.[CrossRef][Medline]
  16. Vidal CI, Mintz PJ, Lu K, et al. An HSP90-mimic peptide revealed by fingerprinting the pool of antibodies from ovarian cancer patients. Oncogene 2004;23:8859–67.[CrossRef][Medline]
  17. Cardó-Vila M, Arap W, Pasqualini R. {alpha}vß5 integrin-dependent programmed cell death triggered by a peptide mimic of Annexin V. Mol Cell 2003;11:1151–62.[CrossRef][Medline]
  18. Lee CC, Kreusch A, McMullan D, Ng K, Spraggon G. Crystal structure of the human neuropilin-1 b1 domain. Structure (Camb) 2003;11:99–108.[Medline]
  19. Mamluk R, Gechtman Z, Kutcher ME, Gasiunas N, Gallagher J, Klagsbrun M. Neuropilin-1 binds vascular endothelial growth factor 165, placenta growth factor-2, and heparin via its b1b2 domain. J Biol Chem 2002;277:24818–25.[Abstract/Free Full Text]
  20. Soker S, Takashima S, Miao HQ, Neufeld G, Klagsbrun M. Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endothelial growth factor. Cell 1998;92:735–45.[CrossRef][Medline]
  21. Kolodkin AL, Levengood DV, Rowe EG, Tai YT, Giger RJ, Ginty DD. Neuropilin is a semaphorin III receptor. Cell 1997;90:753–62.[CrossRef][Medline]
  22. Takahashi T, Fournier A, Nakamura F, et al. Plexin-neuropilin-1 complexes form functional semaphorin-3A receptors. Cell 1999;99:59–69.[Medline]
  23. Migdal M, Huppertz B, Tessler S, et al. Neuropilin-1 is a placenta growth factor-2 receptor. J Biol Chem 1998;273:22272–8.[Abstract/Free Full Text]
  24. Makinen T, Olofsson B, Karpanen T, et al. Differential binding of vascular endothelial growth factor B splice and proteolytic isoforms to neuropilin-1. J Biol Chem 1999;274:21217–22.[Abstract/Free Full Text]
  25. Takashima S, Kitakaze M, Asakura M, et al. Targeting of both mouse neuropilin-1 and neuropilin-2 genes severely impairs developmental yolk sac and embryonic angiogenesis. Proc Natl Acad Sci U S A 2002;99:3657–62.[Abstract/Free Full Text]
  26. Soker S, Miao HQ, Nomi M, Takashima S, Klagsbrun M. VEGF165 mediates formation of complexes containing VEGFR-2 and neuropilin-1 that enhance VEGF165-receptor binding. J Cell Biochem 2002;85:357–68.[CrossRef][Medline]
  27. Zachary I, Gliki G. Signaling transduction mechanisms mediating biological actions of the vascular endothelial growth factor family. Cardiovasc Res 2001;49:568–81.[Abstract/Free Full Text]
  28. Ruiter DJ, Schlingemann RO, Rietveld FJ, de Waal RM. Monoclonal antibody-defined human endothelial antigens as vascular markers. J Invest Dermatol 1989;93:1335–47.
  29. Roberts WG, Palade GE. Neovasculature induced by vascular endothelial growth factor is fenestrated. Cancer Res 1997;57:765–72.[Abstract/Free Full Text]
  30. Lymboussaki A, Partanen TA, Olofsson B, et al. Expression of the vascular endothelial growth factor C receptor VEGFR-3 in lymphatic endothelium of the skin and in vascular tumors. Am J Pathol 1998;153:395–403.[Abstract/Free Full Text]
  31. Paavonen K, Puolakkainen P, Jussila L, Jahkola T, Alitalo K. Vascular endothelial growth factor receptor-3 in lymphangiogenesis in wound healing. Am J Pathol 2000;156:1499–504.[Abstract/Free Full Text]
  32. Soker S, Kaefer M, Johnson M, Klagsbrun M, Atala A, Freeman MR. Vascular endothelial growth factor-mediated autocrine stimulation of prostate tumor cells coincides with progression to a malignant phenotype. Am J Pathol 2001;159:651–9.[Abstract/Free Full Text]
  33. Parikh AA, Fan F, Liu WB, et al. Neuropilin-1 in human colon cancer: expression, regulation, and role in induction of angiogenesis. Am J Pathol 2004;164:2139–51.[Abstract/Free Full Text]
  34. Shimizu M, Murakami Y, Suto F, Fujisawa H. Determination of cell adhesion sites of neuropilin-1. J Cell Biol 2000;148:1283–93.[Abstract/Free Full Text]
  35. James LC, Roversi P, Tawfik DS. Antibody multispecificity mediated by conformational diversity. Science 2003;299:1362–7.[Abstract/Free Full Text]
  36. Varga JM, Kalchschmid G, Klein GF, Fritsch P. Mechanism of allergic cross-reactions-I. Multispecific binding of ligands to a mouse monoclonal anti-DNP IgE antibody. Mol Immunol 1991;28:641–54.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jaalouk, D. E.
Right arrow Articles by Arap, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jaalouk, D. E.
Right arrow Articles by Arap, W.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online