Binding region and interaction properties of sulfoquinovosylacylglycerol (SQAG) with human vascular endothelial growth factor 165 revealed by biosensor-based assays

Abstract

Sulfoquinovosylacylglycerol (SQAG) is a sulfoglycolipid showing anti-angiogenic and radiosensitizing effects for treatment of solid tumors both in vitro and in vivo. Here we elucidated the interaction of SQAG with various growth factors and their cognate receptors for vascular formation using biosensor-based assays. The structure–binding relationship was also determined. Our results show that βSQDG selectively recognizes heparin binding domain (HBD) in human vascular endothelial growth factor 165 (hVEGF₁₆₅) with an affinity in the order of 10⁻¹¹ M. The presence of both a sulfate moiety and at least one C₁₈ length fatty acid chain is essential for binding. Conversion of anomeric configurations in SQAG did not alter the affinity with hVEGF₁₆₅. This SQAG association inhibited T7 phage-displayed HBD binding to neuropilin-1 (NRP1), a VEGF receptor on the endothelial cell surface of blood vessels that specifically recognizes HBD in hVEGF₁₆₅.

---

In a broad sense, neovascularization is primarily classified into two groups, vasculogenesis and angiogenesis. Vasculogenesis is the formation of primitive vascular networks through differentiation of endothelial progenitor cells (EPC) into endothelial cells in the developmental process,¹ while angiogenesis is the growth and sprouting of additional blood vessels from pre-existing blood vessels in postnatal vascular formation.² As well as ontogenesis or homeostasis, angiogenesis is also associated with various diseases,³ such as rheumatic inflammation,⁴ retinopathy in diabetes mellitus⁵ and tumor progression.⁶ At each stage in these vascular formation processes, growth factors and cognate receptors regulate blood endothelial and smooth muscle cells differentiation and/or growth to determine subsequent blood vessel formation.^2,7,8 These include vascular endothelial growth factor (VEGF),^2,7–9 transforming growth factor (TGF),¹⁰insulin-like growth factor (IGF),¹¹fibroblast growth factor (FGF),¹² angiopoietin (Ang),^8,13ephrin (Eph),⁸epidermal growth factor (EGF),¹⁴platelet-derived growth factor (PDGF),² bone morphogenetic protein (BMP)¹⁰ and others, as well as their cognate receptors on cell surfaces.

Sulfoquinovosylacylglycerol (SQAG) is a sulfoglycolipid that was isolated from natural resources, such as sea algae, and shown to exhibit various biological effects including inhibition of HIV reverse transcriptase or DNA polymerase (Fig. 1).^15–17 The SQAG molecule is made up of a core glycerol unit bonded with one or two long acyl chains at position 1 and/or 2. At the 3-hydroxyl group, a sulfoquinovose unit that has a sulfonate group at position 6′ is bonded via an anomeric position (Fig. 1). Several SQAG analogues were synthesized, each with a different stereochemistry at the anomeric carbon and acylation (with varying lengths of fatty acid) at the 1 and/or 2 COMPOUND LINKS

Read more about this on ChemSpider

Download mol file of compoundhydroxyl group in COMPOUND LINKS

Read more about this on ChemSpider

Download mol file of compoundglycerol.¹⁸ Some recent studies have demonstrated that SQAG shows growth suppression of certain solid tumors,^19,20 and displays anti-angiogenic^21,22 and strong radiosensitizing effects both in vitro and in vivo.^22,23 CG-0321, a synthetic analogue of SQAG, is a lead compound and is currently being scheduled to begin phase I clinical tests as an anti-angiogenic type of radiosensitizer for the treatment of solid tumors. In a pre-clinical test, this compound displayed very few side effects and gave a one-year extension to a prognosis of dogs with primary sarcoma by combination therapy with X-irradiation, abolishing the tumor angiogenesis (unpublished data). However, apart from the downregulation of Tie-2 upon treatment with SQAG,²¹ the molecular mechanisms for the anti-angiogenic and radio-sensitizing effects of this compound have not been elucidated. Indeed, it is not even known whether there is direct interaction of SQAG with growth factors or their receptors for angiogenesis.

Fig. 1 Structure of SQAGs used in this study. Sulfoquinovosyldiacylglycerol (SQDG): C₄₅H₈₅NaO₁₂S, M_w 873.2. Sulfoquinovosylmonoacylglycerol (SQMG) C₁₈: C₂₇H₅₁NaO₁₁S, M_w 606.74, C₁₀: C₁₉H₃₅NaO₁₁S, M_w 494.53. Sulfoquinovosylglycerol (SQG): C₉H₁₇NaO₁₀S, M_w 340.04. Glucosyldiacylglycerol (GDG): C₄₅H₈₆O₁₀, M_w 787.16.

The aim of this study was to elucidate the interaction between SQAG and growth factors or their receptors on vascular formation by making use of biosensor-based assays. Here, we employed two distinctive biosensors by taking advantage of each characteristic. One is a flow injection-type surface plasmon resonance (SPR)²⁴ apparatus (Biacore® 3000, GE Healthcare, Piscataway, NJ) and the other is a cuvette-type 27-MHz quartz-crystal microbalance (QCM)²⁵ device (AffinixQ, Initium, Tokyo). This SPR biosensor enables elucidation of the interactions with reversibility and is dependent on the refractive index of an analyte, which occurs by surface plasmon phenomena at a gold–liquid interface and increases proportionally to the mass on the gold. Non-linear fitting of the resulting sensorgram obtained by initial sub-minutes of association and subsequent dissociation analysis using various concentrations of analyte gives the kinetic parameters and the potential binding models.²⁶ In addition, successive injections of label-free analyte are feasible in this flow-injection system fitted with an autosampler. This biosensor is, therefore, adaptive for studies investigating the interaction with dose-dependency or structure–binding relationships for a target of interest. QCM is a biosensor whose shear modulus of quartz (AT-cut) decreases in proportion to the increase of mass on the gold electrode attached on the quartz by vapor deposition.²⁵ The interaction on the gold electrode (mass increase) is, therefore, detected as a decrease of oscillatory resonance frequency in real time. This QCM device comprises a buffer-filled cuvette, which is constantly stirred, where the sensor chip is immersed. The guest molecule is then injected into the cuvette to analyze the interaction on the gold electrode.²⁷ In addition, the sensor chip is readily detached from the device to facilitate recovery of the bound guest molecule. The multiple binding molecules can be simultaneously identified by a single affinity selection.^28,29 Thus, this biosensor is suitable for use in combination with other molecular biological techniques, such as T7 phage display.^28–31

Using the SPR biosensor, we screened the binding of 1 to growth factors or their associated receptors that are involved in vascular formation. Each protein was immobilized on a sensor chip CM5 (GE Healthcare) by an amine coupling reaction and interacted with various concentrations of 1. Of the 25 of molecules tested (Table 1), only hVEGF₁₆₅, which plays the central role in angiogenic activity, showed a response after injection of 1 ranging in concentration from 0.125 to 1 μM (Fig. 2A). The kinetic parameters were obtained from global fitting of the resulting sensorgram using BIAevaluation 3.2 software (Table 1). The SPR profile by injecting 1 showed slow association (k_a: 3.0 × 10³ M⁻¹s⁻¹) with irreversibility (k_d: 1.4 × 10⁻⁷s⁻¹) in this concentration range. The dissociation constant (K_D) between 1 and hVEGF₁₆₅, which was calculated assuming a stoichiometry of interaction of a 1 : 1, was estimated to be 4.6 × 10⁻¹¹ M, suggesting a very high level of affinity. By contrast, no response was observed with regard to other growth factors or their receptors even at 2 μM of 1 (Table 1 and Fig. S1†). In particular, hVEGF₁₂₁, a splicing variant that lacks N115-C160 (exon7) in hVEGF₁₆₅ (Fig. 3), did not show any response against 1 at the concentration range tested (Fig. 2B).

Table 1 The K_D values between compound 1 and growth factors or receptors obtained from SPR analysis in this study

Protein (human)	M w/kDa	Amino acid	K D/M^a
a The KD value was obtained from the resulting sensorgram using BIAevaluation 3.2 software (GE Healthcare). b No response by injecting compound 1 (0.125–2 μM). c The mouse PDGFRB was used. d This Mw includes the fusion GST tag.
VEGF165	19.1	A1-R165	4.6 × 10^−11
VEGF121	14	A1-R121	NR ^b
VEGFR1	60.7	S27-I328	NR
VEGFR2	116	A20-E764	NR
NRP1	90	M1-K644	NR
TGFβ2	25	A1-S112	NR
TGFβ RII	18	M1-D159	NR
IGF-I	7.6	G49-A118	NR
IGF-I R	102.9	E31-N932	NR
FGF1	15.8	M1-D141	NR
FGF2	17.2	A1-S154	NR
FGF R2α	66	R22-E378	NR
Ang-1	66	N21-F496	NR
Tie-2	100	A23-K745	NR
EphB2^c	85.3	V27-K458	NR
EphB4	57.9	L16-A539	NR
EGF	6.2	N1-R53	NR
EGFR	68.6	L25-S645	NR
PDGF-BB	12	S1-T110	NR
PDGFRB	36.7^d	L33-E133	NR
BMP-2	26	M1-R115	NR
BMP RII	41.5	A26-I151	NR
sDLL-4	54	S1-P498	NR
Jagged-1	180	S32-S1046	NR
Notch-1	80.1	A19-Q526	NR
DNA pol λ (cont.)	14.3	M1-R95	9.9 × 10^−9

---

Fig. 2 SPR sensorgram between 1 (0.125–1 μM) and hVEGF₁₆₅ (A) or 1 (0.125–2 μM) and hVEGF₁₂₁ (B). Each protein was immobilized on a sensor chip CM5 by an amine coupling reaction and then various concentrations of 1 were injected. Association: 120 s, dissociation: 120 s. Response curves were generated by subtraction of the background signals generated simultaneously on the control flow cell (protein-non-immobilized cell), the injection of vehicle, and bulk response by COMPOUND LINKS

Read more about this on ChemSpider

Download mol file of compoundDMSO. RU: resonance unit. 1 RU = 1 pg mm⁻².

Fig. 3 Schematic representations of hVEGF₁₆₅ monomer and its truncated version designed in this study.

Recent studies toward the development of anti-angiogenic agents or radiosensitizers have demonstrated hVEGF₁₆₅ to be a promising therapeutic target of angiogenesis-related diseases or tumors.^2,32,33 In particular, Bevacizumab (Avastin™), a VEGF-specific antibody,^34–36 and Pegaptanib (Macugen®), a 28-nucleotide RNA aptamer,^37–41 have been approved by the Food and Drug Administration (FDA) in the United States for treatment of colorectal cancer⁴² and age-related macular degeneration (AMD),^37–39 respectively. Intriguingly, despite the strong affinity of bevacizumab and pegaptanib for hVEGF₁₆₅ and their efficient anti-angiogenic activity both in vitro and in vivo, the site of interaction of each molecule within hVEGF₁₆₅ is discrete. Specifically, bevacizumab recognizes the VEGFR-binding region (A1-R110),⁴³ while pegaptanib binds to HBD (A111-R165).⁴⁴ In order to refine the recognition site in the case of βSQDG, we used the QCM-based affinity selection of a library of T7 phage-displayed polypeptides designated as del 1–4 (Fig. 3). Compound 1 was immobilized on the gold electrode of a sensor chip and then inserted into the buffer-filled cuvette. After stabilizing the QCM sensor, the T7 phage library was injected into the cuvette. Fig. 4A shows the representative sensorgram that was obtained by injection of the T7 phage library. The resulting frequency decrease indicated binding of the T7 phage to compound 1 on the gold electrode. After monitoring the frequency decrease for 10 min, the sensor chip was dislodged from the device. The hostEscherichia coli (BLT5615) culture was then dropped onto the electrode and further incubated for 30 min at 37 °C to recover the bound T7 phage DNA directly.^28–31 Twenty-four phage particles were randomly extracted from the resulting solution and part of the phage DNA encoding capsid protein was amplified by PCR. The products were then subjected to DNA sequencing to define the population of recovered T7 phages. Fig. 4B shows the emergence ratio obtained by a comparison between the content of the respective T7 phage particles in the resulting solution to that of the unscreened parent phage library, which was detected by an agarose gel electrophoresis. The emergence ratio of T7 phage particles displaying HBD (A111-R165; del 2 phage) and C-terminal fragment of HBD (C137-R165; del 4 phage) in hVEGF₁₆₅ increased following selection (Table S5†). This result suggests that 1 recognizes the HBD polypeptide, especially the C137-R165 region (Fig. 3).

Fig. 4 The scanning of βSQDG (1)-binding region by T7 phage display selection using a quartz-crystal microbalance (QCM). (A) A representative QCM sensorgram obtained by the injection of a T7 phage library that displays defined segments (del 1–4) of hVEGF₁₆₅. (B) Emergence of T7 phage particles after the selection by 1. The data are displayed in terms of the emergence ratio of each T7 phage of 24 arbitrarily selected T7 phage particles from the recovered solution with 1 on the gold electrode of the sensor chip and that from the unscreened parent library. The data show means ± SE of three individual experiments. 1 Hz = 0.62 pg mm⁻².

Further evidence for the specific interaction of 1 was obtained by SPR analysis using a truncated version of hVEGF₁₆₅ engineered into an E. coli expression system. The respective fragment was expressed as a His-tagged recombinant polypeptide and purified by affinity chromatography. The sensorgrams obtained from each analysis are shown in Fig. 5. Compound 1 (0.125–1 μM) showed a dose-dependent response against HBD (A111-R165) in hVEGF₁₆₅ (Fig. 5B), although the response against immobilzed del 2 was slightly decreased due to the decreased mass by deletion of A1-R110 region (i.e., partial loss of piggyback or mass enhancer effect).^45,46 Nonetheless, the experiment gave a similar profile of sensorgram and kinetic parameters to that against full length hVEGF₁₆₅ (Fig. 2A and Table 2). A response was also detected using the C-terminal fragment of HBD (C137-R165), though the strength of affinity was reduced by approximately 4 × 10⁴-fold due to the faster dissociation profile (k_d: 5.9 × 10⁻³s⁻¹) compared with full-length hVEGF₁₆₅ (k_d: 1.4 × 10⁻⁷s⁻¹) (Fig. 2A and 5D, Table 2). By contrast, the A1-R110 protein that corresponds to the VEGFR-binding region, and the A111-T142 fragment of HBD did not show any binding with 1 at the concentration range tested (Fig. 5A and C). These data are well correlated with the result of T7 phage display analysis and strongly suggest that 1 recognizes HBD in hVEGF₁₆₅ as a ‘hot spot’ of binding.

Fig. 5 SPR sensorgram of analysis between 1 (0.125–2 μM) and truncated hVEGF₁₆₅protein immobilized on a sensor chip CM5. (A) A1-R110 (del 1) in hVEGF₁₆₅. (B) 1 (0.125–1 μM) and A111-R165 (del 2). (C) A111-T142 (del 3). (D) C137-R165 (del 4).

Table 2 Summary of kinetic parameters of interaction between SQAGs and full length or truncated version of hVEGF₁₆₅ obtained from global fitting using BIAevaluation 3.2 software (GE Healthcare)^a

	k a/M^−1s^−1	k d/s^−1	K D (kd/ka)/M	R max/RU	χ ^2
a Analytical conditions: HBS-EP (10 mM HEPES pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.005% surfactant P20) −8% COMPOUND LINKS Read more about this on ChemSpider Download mol file of compoundDMSO, 25 °C. ka: association rate constant, kd: dissociation rate constant, KD: dissociation constant, Rmax: maximum binding amount, χ^2: fitting to a 1 : 1 drifted baseline model using BIAevaluation 3.2 software (complete fitting; χ^2 = 0). b No response (NR) at each concentration range tested.
1–hVEGF165	3.0 × 10^3	1.4 × 10^−7	4.6 × 10^−11	248	0.345
1–hVEGF121	NR ^b	NR	NR	—	—
1–del 1	NR	NR	NR	—	—
1–del 2	8.1 × 10^3	2.1 × 10^−7	2.5 × 10^−11	99	0.392
1–del 3	NR	NR	NR	—	—
1–del 4	4.3 × 10^3	5.9 × 10^−3	1.4 × 10^−6	7.6	0.082
2–hVEGF165	4.4 × 10^3	3.6 × 10^−7	2.5 × 10^−11	284	0.266
3–hVEGF165	1.7 × 10^3	2.6 × 10^−2	1.6 × 10^−5	37	0.23
4–hVEGF165	1.8 × 10^3	3.2 × 10^−2	1.8 × 10^−5	35.7	0.288
5–hVEGF165	NR	NR	NR	—	—
6–hVEGF165	NR	NR	NR	—	—
7–hVEGF165	NR	NR	NR	—	—

---

To define the structural features of SQAG that is responsible for binding to hVEGF₁₆₅, βSQDG (1) analogues 2–7 (Fig. 1) were prepared and subjected to SPR analysis. Of the SQAG derivatives tested in this study, both 1 and 2 displayed the strongest affinity for hVEGF₁₆₅ in the order of 10⁻¹¹ M (Table 2). Furthermore, configuration of the anomeric center did not appear to affect this interaction (Fig. 2A and 6A). SQMGs (3, 4), which lack acyl group at position 2 in compound 1 (Fig. 1), showed decreased affinity against hVEGF₁₆₅ (∼2 × 10⁵-fold increase in k_d) compared with 1 (Table 2). In addition, the interaction of 3 and 4 with hVEGF₁₆₅ was found to be reversible (Fig. 6B and C). The reduced length of the fatty acid chain from C₁₈ to C₁₀ (5), and the removal of both fatty acid chains in 1 (βSQG, 6) showed no response against hVEGF₁₆₅ at the concentration range tested (Fig. 6D and E). Intriguingly, βGDG (7), which lacks only the sulfate moiety in 1 (Fig. 1), did not respond at all (Fig. 6F). Overall, these data suggest that the presence of both a sulfate moiety and C₁₈ length fatty acid, at least in position 1 of SQAG, are essential for interaction with hVEGF₁₆₅.

Fig. 6 Structure–binding relationship studies using a SPR biosensor. hVEGF₁₆₅ was immobilized on sensor chip CM5 and various concentrations of SQAG analogues (2; 0.125–1 μM, 3–5; 1.25–20 μM, 6, 7; 0.125–2 μM) were respectively injected. (A) 2 and hVEGF₁₆₅. (B) 3 and hVEGF₁₆₅. (C) 4 and hVEGF₁₆₅. (D) 5 and hVEGF₁₆₅. (E) 6 and hVEGF₁₆₅. (F) 7 and hVEGF₁₆₅.

The hVEGF₁₆₅ is known to have a cognate protein–COMPOUND LINKS

Read more about this on ChemSpider

Download mol file of compoundtyrosine kinase receptor, VEGFR1 (Flt1)^47,48 and VEGFR2 (KDR, flk-1)⁴⁹ on the endothelial cells of blood vessels. VEGFR2 is the main signaling pathway for angiogenesis, while VEGFR1 plays a regulatory role. Both receptors bind to hVEGF₁₆₅via the receptor binding domain in A1-R110 region (Fig. 3), which lead to cellular events for new blood vessel formation. By contrast, neuropilin-1 (NRP1), a glycoprotein present on endothelial cells of blood vessels or the surface of hemocytes,⁵⁰ is an alternative receptor of hVEGF₁₆₅.^51–53 The hVEGF₁₆₅ binds to NRP1 through HBD^54–56 and mediates formation of complexes containing VEGFR2 and NRP1, resulting in an enhancement of hVEGF₁₆₅-receptor binding.⁵⁷ Our experiments suggest that SQAG shows a pegaptanib-type of binding to hVEGF₁₆₅i.e., specific recognition of HBD but no interaction with A1-R110 (del 1) or the cognate VEGFR1/2 and NRP1 (Fig. 3, 4B and 5A and Tables 1 and 2). To ascertain binding of SQAG to HBD and its subsequent influence on NRP1 binding, we carried out a binding experiment using QCM apparatus. T7 phage particles displaying HBD (A111-R165, del 2 phage) were pre-incubated with compound 1 or 4 and injected into buffer-filled cuvette where NRP1 immobilized QCM sensor chip was immersed. Fig. 7 shows the resulting data with or without the pretreatment of each compound. Compared to control (vehicle only), binding of del 2 phage to NRP1 was significantly inhibited by the pretreatment of compound 1 and 4. Thus, SQAG shows the inhibitory effect of HBD–NRP1 interaction by specifically binding to HBD.

Fig. 7 The effect of 1 μM βSQDG (1) or 10 μM αSQMG (4) for NRP1 binding of T7 phage-displayed HBD on a QCM. The data are displayed in terms of the rescued phage titer ratio (vehicle vs. compound) of T7 phage particles displaying HBD (del 2 phage) from the recovered solution on the gold electrode of the sensor chip. The data show means ± SD of three individual experiments. *P < 0.05, **P < 0.01. 1 Hz = 0.62 pg mm⁻².

Recent studies have generated inhibitors of hVEGF₁₆₅–NRP1 interaction. EG3287 is a bicyclic peptideNRP1 antagonist mimicking the C-terminal region of hVEGF₁₆₅,⁵⁸ whereas EG00229 is a small-molecule ligand that targets the hVEGF₁₆₅ binding pocket in NRP1.⁵⁹ Together with SQAGs reported in this study, these molecules are expected to be valuable therapeutic agents for treatment of solid tumors targeting hVEGF₁₆₅–NRP1 interaction. Further experiments of SQAG are currently underway.

Conclusions

In this study we have shown that SQAGs selectively recognize hVEGF₁₆₅ through interaction with HBD. Our results also show that the affinity was strongest for SQDG and the presence of both a sulfate moiety and at least one C₁₈ fatty acid in βSQDG are essential for HBD association. Furthermore, this association specifically inhibits cognate HBD–NRP1 interaction. These data provide a sound platform for further studies concerning anti-angiogenic and radiosensitizing activities as well as the development of SQAG for clinical applications.

Acknowledgements

This work was supported by the Program for Promotion of Fundamental Studies in Health Science of the National Institute of Biomedical Innovation (NIBIO), and a research grant from Cango, Co. Ltd.

Notes and references

1. T. Asahara and A. Kawamoto, Am. J. Physiol.: Cell Physiol., 2004, 287, C572–C579.
2. N. Ferrara and R. S. Kerbel, Nature, 2005, 438, 967–974.
3. R. F. Gariano and T. W. Gardner, Nature, 2005, 438, 960–966.
4. A. E. Koch, Arthritis Rheum., 1998, 41, 951–962.
5. B. P. Nicholson and A. P. Schachat, Graefe's Arch. Clin. Exp. Ophthalmol., 2010, 248, 915–930.
6. P. Carmeliet and R. K. Jain, Nature, 2000, 407, 249–257.
7. N. Ferrara, Nat. Rev. Cancer, 2002, 2, 795–803.
8. G. D. Yancopoulos, S. Davis, N. W. Gale, J. S. Rudge, S. J. Wiegand and J. Holash, Nature, 2000, 407, 242–248.
9. K. J. Kim, B. Li, J. Winer, M. Armanini, N. Gillett, H. S. Phillips and N. Ferrara, Nature, 1993, 362, 841–844.
10. N. M. Shah, A. K. Groves and D. J. Anderson, Cell, 1996, 85, 331–343.
11. R. S. Bar, M. Boes, B. L. Dake, B. A. Booth, S. A. Henley and A. Sandra, Am. J. Med., 1988, 85, 59–70.
12. M. Presta, P. Dell'Era, S. Mitola, E. Moroni, R. Ronca and M. Rusnati, Cytokine Growth Factor Rev., 2005, 16, 159–178.
13. C. Suri, P. F. Jones, S. Patan, S. Bartunkova, P. C. Maisonpierre, S. Davis, T. N. Sato and G. D. Yancopoulos, Cell, 1996, 87, 1171–1180.
14. F. Ciardiello and G. Tortora, Clin. Cancer Res., 2001, 7, 2958–2970.
15. Y. Mizushina, I. Watanabe, K. Ohta, M. Takemura, H. Sahara, N. Takahashi, S. Gasa, F. Sugawara, A. Matsukage, S. Yoshida and K. Sakaguchi, Biochem. Pharmacol., 1998, 55, 537–541.
16. K. Ohta, Y. Mizushina, N. Hirata, M. Takemura, F. Sugawara, A. Matsukage, S. Yoshida and K. Sakaguchi, Chem. Pharm. Bull. (Tokyo), 1998, 46, 684–686.
17. C. Benning, Annu. Rev. Plant Physiol. Plant Mol. Biol., 1998, 49, 53–75.
18. S. Hanashima, Y. Mizushina, T. Yamazaki, K. Ohta, S. Takahashi, H. Sahara, K. Sakaguchi and F. Sugawar, Bioorg. Med. Chem., 2001, 9, 367–376.
19. H. Sahara, S. Hanashima, T. Yamazaki, S. Takahashi, F. Sugawara, S. Ohtani, M. Ishikawa, Y. Mizushina, K. Ohta, K. Shimozawa, S. Gasa, K. Jimbow, K. Sakaguchi, N. Sato and N. Takahashi, Jpn. J. Cancer Res., 2002, 93, 85–92.
20. K. Ohta, Y. Mizushina, T. Yamazaki, S. Hanashima, F. Sugawara and K. Sakaguchi, Biochem. Biophys. Res. Commun., 2001, 288, 893–900.
21. Y. Mori, H. Sahara, K. Matsumoto, N. Takahashi, T. Yamazaki, K. Ohta, S. Aoki, M. Miura, F. Sugawara, K. Sakaguchi and N. Sato, Cancer Sci., 2008, 99, 1063–1070.
22. M. Miura, I. Sakimoto, K. Ohta, F. Sugawara and K. Sakaguchi, Anti-Cancer Drugs, 2007, 18, 1–5.
23. I. Sakimoto, K. Ohta, T. Yamazaki, S. Ohtani, H. Sahara, F. Sugawara, K. Sakaguchi and M. Miura, Cancer Res., 2006, 66, 2287–2295.
24. J. Homola, S. S. Yee and G. Gauglitz, Sens. Actuators, B, 1999, 54, 3–15.
25. C. K. O'Sullivan and G. G. Guilbault, Biosens. Bioelectron., 1999, 14, 663–670.
26. K. Andersson, D. Areskoug and E. Hardenborg, J. Mol. Recognit., 1999, 12, 310–315.
27. H. Matsuno, H. Furusawa and Y. Okahata, Chem.–Eur. J., 2004, 10, 6172–6178.
28. Y. Takakusagi, K. Kuramochi, M. Takagi, T. Kusayanagi, D. Manita, H. Ozawa, K. Iwakiri, K. Takakusagi, Y. Miyano, A. Nakazaki, S. Kobayashi, F. Sugawara and K. Sakaguchi, Bioorg. Med. Chem., 2008, 16, 9837–9846.
29. Y. Takakusagi, K. Takakusagi, K. Kuramochi, S. Kobayashi, F. Sugawara and K. Sakaguchi, Bioorg. Med. Chem., 2007, 15, 7590–7598.
30. Y. Takakusagi, K. Takakusagi, F. Sugawara and K. Sakaguchi, Expert Opin. Drug Discovery, 2010, 5, 361–389.
31. Y. Takakusagi, K. Takakusagi, F. Sugawara and K. Sakaguchi, Tanpakushitsu Kakusan Koso, 2009, 54, 1203–1209.
32. J. Brieger, P. Schroeder, J. Gosepath and W. J. Mann, Int. J. Mol. Med., 2005, 15, 145–151.
33. D. H. Gorski, M. A. Beckett, N. T. Jaskowiak, D. P. Calvin, H. J. Mauceri, R. M. Salloum, S. Seetharam, A. Koons, D. M. Hari, D. W. Kufe and R. R. Weichselbaum, Cancer Res., 1999, 59, 3374–3378.
34. N. Ferrara, K. J. Hillan and W. Novotny, Biochem. Biophys. Res. Commun., 2005, 333, 328–335.
35. L. G. Presta, H. Chen, S. J. O'Connor, V. Chisholm, Y. G. Meng, L. Krummen, M. Winkler and N. Ferrara, Cancer Res., 1997, 57, 4593–4599.
36. K. J. Kim, B. Li, K. Houck, J. Winer and N. Ferrara, Growth Factors, 1992, 7, 53–64.
37. E. W. Ng and A. P. Adamis, Can. J. Ophthalmol., 2005, 40, 352–368.
38. S. L. Fine, D. F. Martin and P. Kirkpatrick, Nat. Rev. Drug Discovery, 2005, 4, 187–188.
39. E. S. Gragoudas, A. P. Adamis, E. T. Cunningham, Jr, M. Feinsod and D. R. Guyer, N. Engl. J. Med., 2004, 351, 2805–2816.
40. C. Bell, E. Lynam, D. J. Landfair, N. Janjic and M. E. Wiles, In Vitro Cell. Dev. Biol.: Anim., 1999, 35, 533–542.
41. J. Ruckman, L. S. Green, J. Beeson, S. Waugh, W. L. Gillette, D. D. Henninger, L. Claesson-Welsh and N. Janjic, J. Biol. Chem., 1998, 273, 20556–20567.
42. R. S. Warren, H. Yuan, M. R. Matli, N. A. Gillett and N. Ferrara, J. Clin. Invest., 1995, 95, 1789–1797.
43. Y. A. Muller, Y. Chen, H. W. Christinger, B. Li, B. C. Cunningham, H. B. Lowman and A. M. de Vos, Structure, 1998, 6, 1153–1167.
44. J. H. Lee, M. D. Canny, A. De Erkenez, D. Krilleke, Y. S. Ng, D. T. Shima, A. Pardi and F. Jucker, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 18902–18907.
45. W. A. Henne, D. D. Doorneweerd, J. Lee, P. S. Low and C. Savran, Anal. Chem., 2006, 78, 4880–4884.
46. X. Mao, L. Yang, X. L. Su and Y. Li, Biosens. Bioelectron., 2006, 21, 1178–1185.
47. M. A. Starovasnik, H. W. Christinger, C. Wiesmann, M. A. Champe, A. M. de Vos and N. J. Skelton, J. Mol. Biol., 1999, 293, 531–544.
48. C. Wiesmann, G. Fuh, H. W. Christinger, C. Eigenbrot, J. A. Wells and A. M. de Vos, Cell, 1997, 91, 695–704.
49. B. A. Keyt, H. V. Nguyen, L. T. Berleau, C. M. Duarte, J. Park, H. Chen and N. Ferrara, J. Biol. Chem., 1996, 271, 5638–5646.
50. Y. Yamada, Y. Oike, H. Ogawa, Y. Ito, H. Fujisawa, T. Suda and N. Takakura, Blood, 2003, 101, 1801–1809.
51. L. M. Ellis, Mol. Cancer Ther., 2006, 5, 1099–1107.
52. H. Q. Miao, P. Lee, H. Lin, S. Soker and M. Klagsbrun, FASEB J., 2000, 14, 2532–2539.
53. S. Soker, S. Takashima, H. Q. Miao, G. Neufeld and M. Klagsbrun, Cell, 1998, 92, 735–745.
54. B. A. Appleton, P. Wu, J. Maloney, J. Yin, W. C. Liang, S. Stawicki, K. Mortara, K. K. Bowman, J. M. Elliott, W. Desmarais, J. F. Bazan, A. Bagri, M. Tessier-Lavigne, A. W. Koch, Y. Wu, R. J. Watts and C. Wiesmann, EMBO J., 2007, 26, 4902–4912.
55. R. Mamluk, Z. Gechtman, M. E. Kutcher, N. Gasiunas, J. Gallagher and M. Klagsbrun, J. Biol. Chem., 2002, 277, 24818–24825.
56. S. Soker, H. Fidder, G. Neufeld and M. Klagsbrun, J. Biol. Chem., 1996, 271, 5761–5767.
57. S. Soker, H. Q. Miao, M. Nomi, S. Takashima and M. Klagsbrun, J. Cell. Biochem., 2002, 85, 357–368.
58. H. Jia, A. Bagherzadeh, B. Hartzoulakis, A. Jarvis, M. Lohr, S. Shaikh, R. Aqil, L. Cheng, M. Tickner, D. Esposito, R. Harris, P. C. Driscoll, D. L. Selwood and I. C. Zachary, J. Biol. Chem., 2006, 281, 13493–13502.
59. A. Jarvis, C. K. Allerston, H. Jia, B. Herzog, A. Garza-Garcia, N. Winfield, K. Ellard, R. Aqil, R. Lynch, C. Chapman, B. Hartzoulakis, J. Nally, M. Stewart, L. Cheng, M. Menon, M. Tickner, S. Djordjevic, P. C. Driscoll, I. Zachary and D. L. Selwood, J. Med. Chem., 2010, 53, 2215–2226.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental procedures, additional SPR sensorgrams, and detailed data of T7 phage display selection. See DOI: 10.1039/c1md00180a

‡ Current affiliation: National Cancer Institute (NCI), National Institutes of Health (NIH), Bethesda, MD, USA.

---