Báo cáo khoa học: The C-terminus of viral vascular endothelial growth factor-E partially blocks binding to VEGF receptor-1

The C-terminus of viral vascular endothelial growth factor-E partially blocks binding to VEGF receptor-1 Marie K. Inder, Lyn M. Wise, Stephen B. Fleming and Andrew A. Mercer

Department of Microbiology and Immunology, University of Otago, Dunedin, New Zealand

Keywords Orf virus; Parapoxvirus; vascular endothelial growth factor; VEGF-E; vascular endothelial growth factor receptor

Correspondence L. M. Wise, Virus Research Unit, Department of Microbiology and Immunology, University of Otago, PO Box 56, Dunedin, New Zealand Fax: +64 3 479 7744 Tel: +64 3 479 7723 E-mail: lyn.wise@stonebow.otago.ac.nz

(Received 12 August 2007, revised 24 October 2007, accepted 12 November 2007)

doi:10.1111/j.1742-4658.2007.06189.x

Vascular endothelial growth factor (VEGF) family members play impor- tant roles in embryonic development and angiogenesis during wound healing and in pathological conditions such as tumor formation. Parapox- viruses express a new member of the VEGF family which is a functional mitogen that specifically activates VEGF receptor (VEGFR)-2 but not VEGFR-1. In this study, we show that deletion from the viral VEGF of a unique C-terminal region increases both VEGFR-1 binding and VEGFR- 1-mediated monocyte migration. Enzymatic removal of O-linked glycosyla- tion from the C-terminus also increased VEGFR-1 binding and migration of THP-1 monocytes indicating that both the C-terminal residues and O-linked sugars contribute to blocking viral VEGF binding to VEGFR-1. The data suggest that conservation of the C-terminal residues throughout the viral VEGF subfamily may represent a means of reducing the immuno- stimulatory activities associated with VEGFR-1 activation while maintain- ing the ability to induce angiogenesis via VEGFR-2.

angiogenesis and vascular permeability. VEGFR-3 regulates formation of the lymphatic vasculature via VEGF-C and VEGF-D.

We, and others [7–16], have characterized a group of viral-derived VEGFs, collectively designated VEGF-E, which are encoded by members of the genus Parapox- virus, namely Orf virus (ORFV), Pseudocowpoxvirus (PCPV), Parapoxvirus of red deer in New Zealand (PVNZ) and bovine papular stomatitis virus (BPSV). These viruses infect specific ungulates and can readily infect humans [17,18], and the resulting lesions in the skin demonstrate extensive vascular dilation, dermal edema and endothelial cell proliferation [17,19–21].

Abbreviations BPSV, bovine papular stomatitis virus; ORFV, Orf virus; PCPV, Pseudocowpoxvirus; PVNZ, Parapoxvirus of red deer in New Zealand; SA-HRP, streptavidin-peroxidase; VEGF, vascular endothelial growth factor; VEGFR, VEGF receptor.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

207

Members of the vascular endothelial growth factor (VEGF) family of molecules have emerged as major regulators of new blood vessel formation during vascu- logenesis and angiogenesis [1–3]. These proteins have critical roles during embryogenesis and in normal adult tissues, during wound healing and in pathological con- ditions such as tumor formation. Currently, the mam- malian VEGF family includes VEGF-A, VEGF-B, VEGF-C, VEGF-D and placenta growth factor. VEGF family members exert their biological activity via a family of tyrosine kinase receptors, VEGF recep- tor (VEGFR)-1, VEGFR-2 and VEGFR-3 [4–6]. VEGFR-1 is bound by VEGF-A, VEGF-B and pla- centa growth factor and is primarily expressed on endothelial and hematopoietic cells and may have a role in monocyte recruitment and pro-inflammatory gene expression. VEGFR-2 is bound by VEGF-A, VEGF-C and VEGF-D and is the primary signaling receptor of VEGF-induced endothelial cell mitogenesis, Viral VEGF proteins differ from mammalian VEGF family members in that they specifically bind and acti- vate VEGFR-2 and in general show little or no affinity for VEGFR-1 [8,9,11–16]. The absence of VEGFR-1 binding is surprising given that the structural predic- tions for viral VEGFs are very similar to that of

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

A

B

PCPVVR364VEGF, ORFVNZ7VEGF

VEGF-A [10,22], and that viral VEGFs conserve many residues shown to be vital for VEGF-A binding to VEGFR-1 [23,24]. Although domain-exchange studies with a viral VEGF identified loop regions that are essential for VEGFR-2 binding [25], a structural basis for the inability of viral VEGF to bind VEGFR-1 has not yet been determined. There are, however, a number of structural features that may influence the including a C-terminal specificity of viral VEGFs, Pro ⁄ Thr-rich sequence, encoding potential O-linked glycosylation sites, that is unique to and highly con- served among the viral VEGFs [8,10,13–16]. In this study, we examined the role of the C-terminal region of the VEGF encoded by ORFV strain NZ2 (ORFVNZ2VEGF) in the receptor-binding specificity and biological activities of the viral VEGFs.

Results

Amino acid sequence comparisons

the viral VEGFs

Fig. 1. Sequence comparison of the parapoxviral VEGF C-termini and schematic representation of ORFVNZ2VEGF, VEGF-A and (A) The C-terminal amino acid sequences of ORFVNZ2- mutants. and VEGF, BPSVV660VEGF, PVNZRD86VEGF are shown [8,10,13,14]. Predicted O-linked glyco- sylation sites are shaded gray [37]. (B) Schematic representation of ORFVNZ2VEGF, VEGF-A and mutants. ORFVNZ2VEGF-DC is a mutant of ORFVNZ2VEGF in which the 16 C-terminal (C-term) amino acids have been deleted. VEGF-A-NZ2C is a mutant of VEGF-A in which the heparin-binding domain (HBD) has been replaced with the 16 C-terminal amino acids from ORFVNZ2VEGF. All proteins are FLAG (F)-tagged at the C-terminus for detection and purification. The locations of the conserved N-linked glycosylation site (N) and the variable loop regions (L1–3) are indicated by dashes and light gray boxes, respectively.

and PVNZ

that is not

Comparison of the predicted amino acid sequences of from ORFV strains NZ2 (ORFVNZ2VEGF) and NZ7 (ORFVNZ7VEGF), BPSV strain V660 (BPSVV660VEGF), PCPV strain VR634 strain RD86 (PCPVVR634VEGF) (PVNZRD86VEGF) revealed a Thr ⁄ Pro rich C-termi- nus, containing putative O-linked glycosylation sites (Fig. 1A) found in VEGF-A or other VEGF family members [10]. To examine the role of this conserved C-terminus in the unique receptor speci- ficity and biological activities of viral VEGFs, we constructed a mutant of ORFVNZ2VEGF in which the 16 C-terminal residues were deleted, designated ORFVNZ2VEGF-DC (Fig. 1B). We also constructed a mutant in which the heparin-binding domain of VEGF-A was replaced with the 16 C-terminal residues designated VEGF-A-NZ2C of ORFVNZ2VEGF, (Fig. 1B).

indicating that each contained N-linked glycosylation. Further treatment with sialidase and O-glycosidase reduced VEGF-A-NZ2C and ORFVNZ2VEGF by another 2–4 kDa, although no size shift was observed for VEGF-A and ORFVNZ2VEGF-DC (Fig. 2). The absence of O-linked glycosylation on ORFVNZ2VEGF- the prediction that ORFVNZ2VEGF DC supports contains an O-linked glycosylation site in the 16 C-terminal residues (Fig. 1). The observation that the addition of the 16 C-terminal residues to VEGF-A- NZ2C are associated with the gain of O-linked glyco- sylation also supports this prediction (Fig. 1).

Production and glycosylation state of VEGF mutants

and ORFVNZ2VEGF-DC Deletion of the C-terminus of ORFVNZ2VEGF increases its affinity for VEGFR-1 and VEGFR-2

To investigate the role of the ORFVNZ2VEGF C-ter- minus in receptor specificity, we tested the ability of the VEGF mutants to bind immobilized dimeric Ig fusion proteins containing the extracellular domains of human VEGFR-1 or VEGFR-2 using a receptor-bind- ing ELISA.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

208

VEGF-A and VEGF-A-NZ2C demonstrated signifi- cant levels of binding to VEGFR-1 compared with VEGF-A-NZ2C were expressed with a FLAG octapeptide at the C-terminus and then purified. SDS ⁄ PAGE under reducing condi- tions followed by silver staining revealed bands at (cid:2) 26–28 and 19–20 kDa, for VEGF-A-NZ2C and ORFVNZ2VEGF-DC, respectively (Fig. 2). Deglycosy- lation treatment with N-glycosidase reduced the monomeric size of VEGF-A-NZ2C and ORFVNZ2- VEGF-DC by 2–4 kDa (Fig. 2), which was similar seen for VEGF-A and ORFVNZ2VEGF, to that

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

did

1.1 lgÆmL)1 Preincubation of soluble VEGF-A or VEGF-A- NZ2C with VEGFR-1 significantly inhibited the binding of the receptor to immobilized VEGF-A at tested (‡ 0.4 lgÆmL)1, P £ 0.05), all concentrations significantly not whereas ORFVNZ2VEGF inhibit VEGFR-1 binding to VEGF-A at any con- centration tested (Fig. 3C). ORFVNZ2VEGF-DC did, inhibit binding of VEGFR-1 to VEGF-A however, (P £ 0.05; concentration of from a Fig. 3C).

Fig. 2. The C-terminus of ORFVNZ2VEGF has O-linked glycosyla- tion. Purified VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF and ORFVNZ2- VEGF-DC were analyzed before and after enzymatic removal of N- and O-linked sugars. The proteins were resolved under reducing conditions following treatment with the indicated combinations of enzymes as described in Experimental procedures. Proteins were visualized by western blotting using anti-FLAG Ig. Molecular mass markers are indicated.

Preincubation of soluble VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF or ORFVNZ2VEGF-DC significantly inhibited the binding of VEGFR-2 to immobilized VEGF-A at all concentrations tested (‡ 0.2 lgÆmL)1, P £ 0.05; Fig. 3D). ORFVNZ2VEGF-DC was, however, significantly more potent than ORFVNZ2VEGF from a concentration of 0.6 lgÆmL)1, whereas no significant differences were observed between VEGF-A-NZ2C and VEGF-A at any of the concentrations tested (P £ 0.05; Fig. 3D).

The interactions of the VEGF mutants with VEG- FR-1 and VEGFR-2 were further tested in bioassays that detect receptor binding and cross-linking at the cell surface. These assays made use of BaF3 cell lines expressing chimeric receptors consisting of the extracel- lular domain of either human VEGFR-1 or murine VEGFR-2 and the transmembrane and cytoplasmic domains of the erythropoietin receptor [26,27]. Binding and cross-linking of the chimeric receptors induce cell proliferation.

the concentrations mock-purified protein at all of tested (‡ 0.1 lgÆmL)1, P £ 0.05; Fig. 3A). As reported previously [15,16], ORFVNZ2VEGF did not bind VEGFR-1 at any concentration tested (Fig. 3A). By contrast, ORFVNZ2VEGF-DC showed significant levels of binding to VEGFR-1 from a concentration of 3.3 lgÆmL)1 (P £ 0.05; Fig. 3A). concentration of 11 ngÆmL)1 VEGF-A and VEGF-A-NZ2C stimulated significant proliferation of cells expressing VEGFR-1 from the lowest concentration tested (‡ 1.2 ngÆmL)1, P £ 0.05), although ORFVNZ2VEGF did not induce cellular pro- liferation (Fig. 3E). ORFVNZ2VEGF-DC stimulated significant proliferation of cells expressing VEGFR-1 (P £ 0.05; from a Fig. 3E).

however, potent

tested VEGF-A and VEGF-A-NZ2C demonstrated signifi- cant levels of binding to VEGFR-2 compared with mock-purified protein at all of the concentrations tested (‡ 0.1 lgÆmL)1, P £ 0.05; Fig. 3B). ORFVNZ2VEGF and ORFVNZ2VEGF-DC also showed significant bind- ing to VEGFR-2 from a concentration of 0.4 lgÆmL)1 (P £ 0.05; Fig. 3B). No significant differences in binding to VEGFR-2 were observed at any concen- tration between VEGF-A and VEGF-A-NZ2C or ORFVNZ2VEGF and ORFVNZ2VEGF-DC (P £ 0.05; Fig. 3B). VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF and ORFVNZ2VEGF-DC were each able to stimulate sig- nificant proliferation of cells expressing VEGFR-2, in from a concentration of the presence of heparin, 0.8 ngÆmL)1 (P £ 0.05; Fig. 3F). ORFVNZ2VEGF-DC was, than significantly more ORFVNZ2VEGF, VEGF-A and VEGF-A-NZ2C at all (0.8–22 ngÆmL)1, P £ 0.05; concentrations Fig. 3F).

increase in VEGFR-1 binding

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

209

To further examine the receptor specificity of the VEGF mutants, we tested their ability to inhibit VEGF-A bind the dimerized Ig fusion proteins con- taining the extracellular domains of human VEGFR-1 or VEGFR-2 under soluble binding conditions using a competitive displacement ELISA [8,13]. In summary, ORFVNZ2VEGF-DC showed a signi- ficant compared with ORFVNZ2VEGF in the three different assays (Fig. 3A,C,E). In addition, a consistent, but not signifi- cant, decrease was observed in VEGFR-1 binding by VEGF-A-NZ2C compared with VEGF-A. No

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

Enzymatic removal of the O-linked glycosylation of ORFVNZ2VEGF increases its affinity for VEGFR-1 and VEGFR-2

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

210

To examine the role O-linked glycosylation might play in the receptor specificity and biological activities of consistent differences were noted in VEGFR-2 binding between VEGF-A-NZ2C and VEGF-A (Fig. 3B,D,F). ORFVNZ2VEGF-DC did, however, show a small, but in VEGFR-2 binding, compared significant increase, with ORFVNZ2VEGF, in two of the three assay sys- tems (Fig. 3D,F).

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

Fig. 3. Deletion of the C-terminus of ORFVNZ2VEGF increases binding to VEGFR-1 and VEGFR-2. Immobilized VEGFR-1–Ig (A) or VEGFR-2– Ig (B) fusion protein was incubated for 2 h with increasing concentrations of soluble VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF, ORFVNZ2VEGF- DC or with medium alone. Bound VEGF protein was detected with horseradish peroxidase-conjugated M2 anti-FLAG Ig. Values are expressed as a binding index, defined as the mean increase in absorbance ± SEM at 450 nm over the background (n = 2) and are represen- tative of three separate experiments. Soluble VEGFR-1–Ig (C) or VEGFR-2–Ig (D) fusion protein was incubated for 2 h with increasing concentrations of VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF and ORFVNZ2VEGF-DC or with medium alone. The mixture was then added to VEGF-A-coated wells to capture free VEGFR–Ig, which was detected with a biotinylated sheep anti-human Ig SA-HRP conjugate. The results are presented as the percentage of the maximal absorbance of VEGFR–Ig bound. Values are expressed as mean ± SEM (n = 2) and are rep- resentative of three separate experiments. The abilities of the VEGF mutants to bind and cross-link VEGFR-1 (E) or VEGFR-2 (F) were tested using specific bioassay cell lines. Bioassay cells were washed and resuspended in dilutions of VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF and ORFVNZ2VEGF-DC, or with medium alone for 48 h at 37 (cid:3)C. DNA synthesis was quantified by [3H]thymidine incorporation and b-counting. Values were expressed as a proliferation index, defined as the mean increase in cell proliferation ± SEM over the background (n = 2) and are representative of two separate experiments. An asterisk indicates a significant (P £ 0.05) difference from the background level recorded when no growth factor was added. A cross indicates a significant (P £ 0.05) difference in receptor binding between ORFVNZ2VEGF-DC and the equivalent concentration of ORFVNZ2VEGF.

Removal of the C-terminal residues or O-linked glycosylation of ORFVNZ2VEGF increases its recruitment of THP-1 monocytes

Previous studies have shown that mammalian VEGF family members induce monocyte chemotaxis via their interaction with VEGFR-1 [28–30]. Thus, using a Trans- well(cid:2) assay (Corning Costar, Corning, NY, USA), we examined the chemotactic response of THP-1 mono- cytes to treatment with the VEGF mutants. The human THP-1 monocytic cell line has previously been shown to respond to VEGF family members in a manner similar to human peripheral blood monocytes [29]. the viral VEGFs, ORFVNZ2VEGF was treated with sialidase and O-glycosidase to remove the O-linked glycosylation. The enzyme-treated protein, ORFVNZ2- VEGF-DOglyc was then repurified and analyzed by SDS ⁄ PAGE and western blotting which revealed bands at (cid:2) 38–42 and 19–20 kDa, under nonreducing and reducing conditions, respectively (Fig. 4A). N-Gly- cosidase treatment reduced the monomeric size of ORFVNZ2VEGF-DOglyc by 2–3 kDa (Fig. 4A), but further treatment with sialidase and O-glycosidase did not result in a size shift. These results confirm that ORFVNZ2VEGF-DOglyc has retained its N-linked glycosylation and the majority of the O-linked glyco- sylation has been removed. To investigate the role of

antibody VEGF-A and VEGF-A-NZ2C were able to induce significant migration of cells from a concentration of 4 ngÆmL)1 (P £ 0.05), whereas ORFVNZ2VEGF did not induce cell migration (Fig. 5). ORFVNZ2VEGF-DC and ORFVNZ2VEGF-DOglyc were also able to induce significant migration of cells from a concentration of 20 ngÆmL)1 (P £ 0.05; Fig. 5). Migration of THP-1 monocytes induced by VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF-DC and ORFVNZ2VEGF-DOglyc was significantly inhibited by preincubation of the cells with against VEGFR-1 neutralizing (P £ 0.05) (Fig. 5).

that ORFVNZ2VEGF O-linked glycosylation plays in receptor specificity we tested the ability of ORFVNZ2VEGF-DOglyc to bind and cross-link VEGFR-1 and VEGFR-2 in the BaF3 bioassays. Consistent with our previous results, VEGF-A and ORFVNZ2VEGF-DC were able to stimu- late significant proliferation of cells expressing VEG- FR-1 from 1.2 and 11 ngÆmL)1, respectively (P £ 0.05), whereas ORFVNZ2VEGF did not induce cellular pro- liferation (Fig. 4B). ORFVNZ2VEGF-DOglyc was less than ORFVNZ2VEGF-DC but did stimulate potent significant proliferation of cells expressing VEGFR-1 concentration tested (100 ngÆmL)1, the highest at P £ 0.05; Fig. 4B).

(Fig. 3). By contrast,

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

211

potent than the increase The increase in monocyte migration induced by ORFVNZ2VEGF-DC, compared with ORFVNZ2VEGF, was consistent with its increase in VEGFR-1 binding in the receptor-binding assays (Figs 3 and 4). In addi- tion, the slight decrease in monocyte migration by VEGF-A-NZ2C, compared with VEGF-A, was consis- tent with its decrease in VEGFR-1 binding in the receptor-binding assays the increase in monocyte migration induced by ORFVNZ2- VEGF-DOglyc, compared with ORFVNZ2VEGF, was greater in VEGFR-1 binding observed in the receptor-binding assay (Fig. 4). VEGF-A, ORFVNZ2VEGF, ORFVNZ2VEGF-DC and ORFVNZ2VEGF-DOglyc were each able to sti- mulate significant proliferation of cells expressing VEGFR-2, in the presence of heparin, from a concen- tration of 0.3 ngÆmL)1 (P£ 0.05; Fig. 4C). ORFVNZ2- VEGF-DC and ORFVNZ2VEGF-DOglyc were both than VEGF-A and significantly more ORFVNZ2VEGF at all concentrations tested (0.3– 7.4 ngÆmL)1, P £ 0.05; Fig. 4C).

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

the C-terminal O-linked glycosylation from Fig. 4. Removal of ORFVNZ2VEGF increases binding to VEGFR-1 and VEGFR-2. (A) Purified ORFVNZ2VEGF and ORFVNZ2VEGF-DOglyc were analyzed before and after enzymatic removal of N- and O-linked sugars. The proteins were resolved under reducing conditions following treat- ment with the indicated combinations of enzymes as described in Experimental procedures. Proteins were visualized by silver staining. Molecular mass markers are indicated. The ability of ORFVNZ2VEGF-DOglyc to bind and cross-link VEGFR-1 (B) or VEGFR-2 (C) was tested using specific bioassay cell lines. Bioassay cells were washed and resuspended in dilutions of VEGF-A, ORFVNZ2VEGF, ORFVNZ2VEGF-DC and ORFVNZ2VEGF-DOglyc, or with medium alone for 48 h at 37 (cid:3)C. DNA synthesis was quanti- fied by [3H]thymidine incorporation and b-counting. Values were expressed as a proliferation index, defined as the mean increase in cell proliferation ± SEM over the background (n = 2) and are repre- sentative of two separate experiments. An asterisk indicates a sig- nificant (P £ 0.05) difference from the background level recorded when no growth factor was added. A cross indicates a significant (P £ 0.05) difference in receptor binding between ORFVNZ2VEGF- DC or ORFVNZ2VEGF-DOglyc and the equivalent concentration of ORFVNZ2VEGF.

ORFVNZ2VEGF recognition of VEGFR-1 and that the O-linked sugars associated with the C-terminus play a contributing role. Interestingly, a previous study using Escherichia coli-expressed viral VEGF (strain D1701), which would not have contained O-linked glycosylation, reported minimal VEGFR-1 binding [11,12]. The assay employed in this study directly mea- sures binding and cross-linking of chimeric VEGFR-1 and may be more sensitive than the endothelial cell- based assays used in the previous study.

Discussion

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

212

Removal of the C-terminal residues from ORFVNZ2- VEGF also appeared to slightly enhance signaling via VEGFR-2, as measured in the BaF3 cell proliferation assay (Fig. 3F). This suggests that the C-terminal region of viral VEGF interferes with both VEGFR-1 and VEGFR-2 binding. This is consistent with previ- ous studies, which have shown that the binding sites for VEGFR-1 and VEGFR-2 overlap [23,31,32]. A slight increase in VEGFR-2 binding by ORFVNZ2- VEGF-DC was also seen in the competitive receptor- binding ELISA (Fig. 3D). However, no difference in VEGFR-2 binding was observed in the direct receptor- binding ELISA (Fig. 3B). This latter assay relies on immunodetection of the FLAG peptide fused to the VEGF protein. Because the precise location of the FLAG peptide, in relation to the VEGF homology domain, varies between proteins, the accessibility of the FLAG peptide within the VEGF ⁄ VEGFR complex may also differ. This may influence the sensitivity of the assay and mask the slight differences in VEGFR-2 observed in other assays. Interestingly, removal of the In this study, we demonstrated a role for the highly conserved O-glycosylated C-terminus in determining the unique receptor-recognition profile and biological activities of viral VEGFs. Complete removal of the C-terminus of ORFVNZ2VEGF increased its binding to VEGFR-1. In addition, enzymatic removal of the O-linked glycosylation within the C-terminal region improved the ability of ORFVNZ2VEGF to activate VEGFR-1, but not to the same level as that of the deletion mutant. These results indicate that the C-ter- minal residues play a dominant role in preventing

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

Fig. 5. Removal of the C-terminal residues or O-linked glycosylation increases ORFVNZ2VEGF-induced VEGFR-1-dependent chemotaxis of THP-1 monocytes. THP-1 monocytes (1 · 105 cells) were added to the upper chamber of a Transwell(cid:2) insert. The indicated concentrations of VEGF-A, VEGF-A-NZ2C, ORFVNZ2VEGF, ORFVNZ2VEGF-DC and ORFVNZ2VEGF-DOglyc, or medium alone were added to the lower com- partments, the inserts were incubated for 6 h at 37 (cid:3)C and migrated cells that remained attached to the insert membrane were stained and counted as described in Experimental procedures. Where indicated, THP-1 monocytes were preincubated with a neutralizing antibody against VEGFR-1 for 16 h and then assayed as described. Results were expressed as a migration index, defined as the mean increase in cell migration, ± SEM, over background (n = 8) and are representative of three experiments. Migration indexes that were significantly above that of medium only are indicated by an asterisk (P £ 0.05). A cross indicates a significant (P £ 0.05) difference between cell migration induced in the presence and absence of antibody.

the C-terminal

O-linked glycosylation from ORFVNZ2VEGF also slightly increased the response seen in the VEGFR-2 BaF3 assay, to the level seen with the deletion mutant (Fig. 4C). This indicates that steric hindrance by the O-linked sugars, and not the C-terminal residues, may reduce the viral VEGF’s affinity for VEGFR-2.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

213

inhibitory effects of residues on VEGFR binding may be dependent on their interac- tions with other structural features of the viral VEGF. Removal of the C-terminal region from ORFVNZ2- VEGF did not elevate VEGFR binding to the level seen for VEGF-A, which suggests that other structural elements play a role in determining VEGFR specificity of the viral VEGFs. It has been hypothesized that the inability of viral VEGF to bind VEGFR-1 may be due to the absence of a functional groove on the receptor- binding face of the homodimer [10,22]. The lack of this groove may prevent viral VEGFs effectively binding the domain 2–3 linker region of VEGFR-1. In support of this model, changing Arg46 to Iso within loop 1 of the groove region of ORFVNZ2VEGF increased its affinity for both VEGF receptors [22]. In addition, replacement of both loop 1 and loop 3 of ORFVNZ2- VEGF with those of VEGF-A partially restored bind- ing to VEGFR-1 [22]. Neither mutation, however, completely restored VEGFR-1 binding to the levels seen for VEGF-A [22]. These findings in conjunction with the results of our study suggest that the groove region of the receptor binding face and the C-terminal residues make separate but additive contributions to the inability of viral VEGF to bind VEGFR-1. It would therefore be interesting to construct a viral Recently the crystal structure of ORFVNZ2VEGF was solved revealing a high similarity to the known structures of other VEGF family members [22,25]. The C-terminal region, however, appeared to be highly flex- ible and could not be resolved by crystallography. Its location adjacent to the receptor-binding face, as illus- trated in Fig. 6, suggests it is well placed to influence the receptor-recognition profile of viral VEGF. The C-terminal residues may therefore form a direct physi- cal block, preventing access of the VEGFRs to the receptor-binding face. Curiously, replacement of the heparin-binding domain of VEGF-A with the C-termi- nal residues of the viral VEGF did not significantly alter the affinity of VEGF-A for VEGFR-1 or VEG- FR-2. A simple explanation for this finding would be that the C-terminal residues of the VEGF-A mutant, due to the influence of adjacent residues or other struc- tural features of VEGF-A, are not orientated towards the receptor-binding face and are therefore unable the to influence receptor recognition. Alternatively,

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

A

the C-terminus influence the receptor-recognition pro- file of viral VEGFs. It was also proposed that BPSVV660VEGF residues adjacent to the receptor-bind- ing face influence the orientation of loop 3 and the width of the groove, thereby affecting its ability to bind the VEGFRs [8]. The orientation of the receptor-bind- ing face of BPSVV660VEGF may therefore be such that the C-terminal residues have less interaction with the loop regions. Alternatively, this region of the other viral VEGFs, which differ from BPSVV660VEGF, maybe act as a site of interaction with the C-terminal residues.

B

(PDB identifier 1QTY)

VEGFR-1 activation has been shown to mediate VEGF-induced immunostimulatory activities such as dendritic cell activation, monocyte migration, inflam- matory cytokine induction and other important anti- viral immune defenses [29,30,33,34]. Removal of either the C-terminal residues or O-linked glycosylation from the viral VEGF resulted in an increase in VEGFR-1- mediated migration of THP-1 monocytes. Conserva- tion of the C-terminal residues throughout the viral VEGF family may therefore represent a means of viral immune evasion. The viral VEGFs, however, retain sufficient binding of VEGFR-2 to potently induce vas- cular dilation, dermal edema and proliferation of endothelial cells, thereby contributing to the prolifera- tive and highly vascularized nature of parapoxvirus lesions [15,35,36].

Experimental procedures

derived

and ORFVNZ2VEGF were

Fig. 6. Structural elements of ORFVNZ2VEGF involved in VEGFR binding. Ribbon representation of the structure of (A) VEGF-A:Flt- 1-D2 (VEGFR-1 domain 2) [32] and (B) ORFVNZ2VEGF (PDB identifier 2GNN) [22]. The VEGF dimers are shaded gray with one monomer in a darker shade, and with loops 1, 2 and 3, shaded blue, green and red, respectively. VEGFR- 1 domain 2 is shaded purple. Residues Ser94–Asn99 within loop 3 of one of the ORFVNZ2VEGF monomers are missing, as they were disordered due to their intrinsic flexibility. The residues of VEGF-A that form the groove implicated in VEGFR-1 binding, and the resi- dues of ORFVNZ2VEGF in the equivalent positions, are shaded black. The flexible C-terminus of each ORFVNZ2VEGF monomer is labeled and shaded orange, with the disordered residues Thr120– Arg133 drawn schematically as dashes. The putative O-linked glycosylation sites within the C-termini are shaded yellow.

Expression vectors

for VEGF-A (murine VEGF iso- Expression vectors from form 164) pAPEX-3 and have been described previously [15]. A DNA fragment containing nucleotides 4–368 of ORFVNZ2VEGF was amplified by PCR from viral DNA with the following (5¢-AGCGCCCGGCGCGCCAGA primers: NZ2-DC 5¢ AGTTGCTCGTCGGCATAC-3¢, AscI site underlined) (5¢-ACTCGAACGCGTTCGTGGTCTA and NZ2-DC 3¢ CAATCGCA-3¢, MluI site underlined). A DNA fragment containing nucleotides 4–458 of VEGF-A and 55 nucleo- tides from the C-terminus of ORFVNZ2VEGF was ampli- fied from pAPEX–mVEGF-A [15] with the following (5¢-CATGGCGCGCCTGATGAAC primers: mV-for 5¢ TTTCTGCTGTCTTGG-3¢) and mV-NZ2C 3¢ (5¢-CTGAC GCGTGCGGCGTCTTCTGGGCGGCCTTGTGGTCGT CGGTGGCGTGGTTGTGAACTTTGGTCTGCATTCA CATCGGCT-3¢). The PCR products were digested with AscI and MluI and ligated to pAPEX–mVEGF-A [15], from which the DNA sequence encoding VEGF-A but not the FLAG octapeptide (IBI ⁄ Kodak, Rochester, NY, USA) had been removed by digestion with AscI.

VEGF mutant in which, loops 1 and 3 are replaced with those from VEGF-A, and the C-terminal residues are deleted, to ascertain whether these regions are the determinants of VEGFR recognition or if additional structural features are involved.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

214

The C-terminal residues are conserved among all members of the viral VEGF family. Recently, a variant viral VEGF from BPSV has been reported that shows greater recognition of VEGFR-1 and VEGFR-2 than the other viral VEGFs [8]. BPSVV660VEGF varies from the other viral VEGFs in its C-terminal residues, yet conserves the O-linked glycosylation sites. This sup- ports our finding that the specific amino acids within

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

protein. The

Protein synthesis from the expression vectors described above gave rise to secreted polypeptides tagged with the FLAG octapeptide at their C-termini. The C-terminal deletion mutant of ORFVNZ2VEGF was designated ORFVNZ2VEGF-DC, and the domain exchange mutant of VEGF-A with the ORFVNZ2VEGF O-glycosylated C-ter- minus was designated VEGF-A-NZ2C.

1 h. The mixture was then transferred to plates coated with VEGF-A and incubated at 25 (cid:3)C for 1 h to capture the unbound VEGFR–Ig captured fusion VEGFR–Ig fusion protein was detected by biotinylated anti-human Ig (Dako, Glostrup, Denmark) and streptavi- din-peroxidase (SA-HRP; Sigma) and tetramethylbenzidine substrate reagent and quantified by measuring the absor- bance at 450 nm.

Recombinant protein production

Recombinant FLAG-tagged proteins were expressed in 293-EBNA cells, purified and quantitated as previously described [15]. A mock elution sample was obtained from conditioned medium of pAPEX-3-transfected 293-EBNA cells that underwent the same purification process.

and the

transmembrane

Bioassays for the binding and cross-linking of the extracellular domains of VEGFR-1 and VEGFR-2

Bioassays for monitoring the binding and cross-linking of lines VEGFR-1 and VEGFR-2, using BaF3-derived cell expressing chimeric receptors consisting of the extracellular, ligand-binding domains of human VEGFR-1 or mouse VEGFR-2 and cytoplasmic domains of the erythropoietin receptor, were carried out as previously described [26,27]. Briefly, the bioassay cell lines were incubated with various concentrations of purified growth factors for 48 h at 37 (cid:3)C. DNA synthesis was quan- tified by measuring [3H]thymidine incorporation during a further 16 h incubation.

Proteins were treated with N-glycosidase F (Roche, Mann- heim, Germany), sialidase (neuraminidase, Roche), and O-glycosidase (Roche), and resolved by SDS ⁄ PAGE and visualized by silver staining or western blotting, as previ- ously described [14].

Protein deglycosylation

Chemotaxis assays using THP-1 monocytes were carried out in 24-well plates containing Transwell(cid:2) inserts of 5 lm pore size (Corning Costar, Corning, NY, USA), as previ- ously described [8]. Briefly, monocytes were loaded into inserts with various concentrations of purified growth fac- tor in the bottom compartment and then incubated for 6 h at 37 (cid:3)C. Nonmigrated cells were removed from the upper side of the filter membrane and the adherent cells on the lower side were fixed in gluteraldehyde then stained using Gill’s hemotoxylin. For a quantitative assessment of migrated cells, a total of four fields of ·40 magnification from two different wells was counted.

immunoplates (Nunc, Roskilde, Den- Maxisorp 96-well mark) were coated with 500 ngÆmL)1 VEGFR-1–Ig or VEGFR-2–Ig fusion proteins (R&D Systems, Minneapolis, in coating buffer (15 mm Na2CO3, 35 mm MN, USA) NaHCO3, pH 9.6) at 4 (cid:3)C for 16 h and blocked with 0.5% BSA and 0.02% Tween 20 at room temperature for 1 h. Plates were washed between steps with wash buffer (NaCl ⁄ Pi and 0.02% Tween 20). Immobilized VEGFR–Igs were then incubated with a titration of purified VEGFs at room temperature for 2 h. Captured VEGF was detected by horseradish peroxidase-conjugated M2 anti-FLAG Ig (Sigma, St Louis, MO, USA) and developed with tetra- methylbenzidine substrate reagent (B&D Biosciences, San Diego, CA, USA) and quantified by measuring absorbance at 450 nm.

Chemotaxis assay ELISA receptor-binding assay

Enzymatic removal of O-linked glycosylation from ORFVNZ2VEGF

Purified protein (150 lg) was diluted in 0.05 m sodium phosphate buffer (pH 7) containing 0.1% SDS. Twenty mU of sialidase and 25 mU of O-glycosidase was added and the mixture was incubated at 37 (cid:3)C for 3 h. The protein was then re-purified and quantitated, as previously described [15].

ELISA competitive displacement receptor binding assay

Maxisorp 96-well incubated with immunoplates were 400 ngÆmL)1 VEGF-A in coating buffer at 4 (cid:3)C for 16 h and blocked with 1% BSA and 0.02% Tween 20 at 37 (cid:3)C for 45 min. Plates were washed between steps with wash buffer. Samples of purified growth factors, serially diluted (NaCl ⁄ Pi with 0.4% BSA, 0.02% in binding buffer Tween 20 and 2 lgÆmL)1 heparin in VEGFR-2 assays only), were incubated with 300 ngÆmL)1 human VEGFR-1– Ig or VEGFR-2–Ig in non-absorbent plates at 25 (cid:3)C for

Statistical analysis was performed using analysis of variance (single factor anova) with significant points of difference (P £ 0.05) determined using Tukey’s test.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

215

Statistical analysis

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

Acknowledgements

12 Ogawa S, Oku A, Sawano A, Yamaguchi S, Yazaki Y & Shibuya M (1998) A novel type of vascular endothe- lial growth factor, VEGF-E (NZ-7 VEGF), preferen- tially utilizes KDR ⁄ Flk-1 receptor and carries a potent mitotic activity without heparin-binding domain. J Biol Chem 273, 31273–31282.

13 Ueda N, Inder MK, Wise LM, Fleming SB & Mercer AA (2007) Parapoxvirus of red deer in New Zealand encodes a variant of viral vascular endothelial growth factor. Virus Res 124, 50–58.

This study was partially supported by the Health Research Council of New Zealand. Lyn Wise was sup- ported in part by the University of Otago Health Sciences Career Development Program Postdoctoral Fellowship Award. We thank Nicola Real for expert technical assistance.

References

1 Carmeliet P (2005) VEGF as a key mediator of angio-

genesis in cancer. Oncology 69(Suppl. 3), 4–10.

14 Ueda N, Wise LM, Stacker SA, Fleming SB & Mercer AA (2003) Pseudocowpox virus encodes a homolog of vascular endothelial growth factor. Virology 305, 298– 309.

2 Ferrara N (2004) Vascular endothelial growth factor: basic science and clinical progress. Endocrin Rev 25, 581–611.

15 Wise LM, Ueda N, Dryden NH, Fleming SB, Caesar C, Roufail S, Achen MG, Stacker SA & Mercer AA (2003) Viral vascular endothelial growth factors vary extensively in amino acid sequence, receptor-binding specificities and the ability to induce vascular permeabil- ity yet are uniformLy active mitogens. J Biol Chem 278, 38004–38024.

16 Wise LM, Veikkola T, Mercer AA, Savory LJ, Fleming

3 McColl BK, Stacker SA & Achen MG (2004) Molecu- lar regulation of the VEGF family – inducers of angio- genesis and lymphangiogenesis. APMIS 112, 463–480. 4 Olsson AK, Dimberg A, Kreuger J & Claesson-Welsh L (2006) VEGF receptor signalling – in control of vascu- lar function. Nat Rev Mol Cell Biol 7, 359–371.

5 Shibuya M & Claesson-Welsh L (2006) Signal transduc- tion by VEGF receptors in regulation of angiogenesis and lymphangiogenesis. Exp Cell Res 312, 549–560. 6 Zachary I (2003) VEGF signalling: integration and

SB, Caesar C, Vitali A, Makinen T, Alitalo K & Stacker SA (1999) Vascular endothelial growth factor (VEGF)-like protein from orf virus NZ2 binds to VEGFR2 and neuropilin-1. Proc Natl Acad Sci USA 96, 3071–3076.

17 Haig DM & Mercer AA (1998) Ovine diseases. Orf. Vet

multi-tasking in endothelial cell biology. Biochem Soc Trans 31, 1171–1177.

Res 29, 311–326.

7 Delhon G, Tulman ER, Afonso CL, Lu Z, delaConcha- Bermejillo A, Lehmkuhl HD, Piccone ME, Kutish GF & Rock DL (2004) Genomes of the parapoxvirus orf virus and bovine papular stomatitis virus. J Virol 78, 168–177.

18 Mercer A, Fleming S, Robinson A, Nettleton P & Reid H (1997) Molecular genetic analyses of parapoxviruses pathogenic for humans. Arch Virol Suppl 13, 25–34. 19 Groves RW, Wilson-Jones E & MacDonald DM (1991) Human orf and milkers’ nodule: a clinicopathologic study. J Am Acad Dermatol 25, 706–711.

20 Horner GW, Robinson AJ, Hunter R, Cox BT & Smith

8 Inder MK, Ueda N, Mercer AA, Fleming SB & Wise LM (2007) Bovine papular stomatitis virus encodes a functionally distinct VEGF that binds both VEGFR-1 and VEGFR-2. J Gen Virol 88, 781–791.

R (1987) Parapoxvirus infections in New Zealand farmed red deer (Cervus elaphus). NZ Vet J 35, 41–45.

9 Lyttle DJ, Fraser KM, Fleming SB, Mercer AA & Rob-

21 Nagington J, Lauder IM & Smith JS (1967) Bovine

papular stomatitis, pseudocowpox and milker’s nodules. Vet Rec 81, 306–313.

inson AJ (1994) Homologs of vascular endothelial growth factor are encoded by the poxvirus orf virus. J Virol 68, 84–92.

10 Mercer AA, Wise LM, Scagliarini A, McInnes CJ,

22 Pieren M, Prota A, Ruch C, Kostrewa D, Wagner A, Biedermann K, Winkler F & Ballmer-Hofer K (2006) Crystal structure of the Orf virus NZ2 variant of vascu- lar endothelial growth factor-E. J Biol Chem 281, 19578–19587.

Buettner M, Rhiza HJ, McCaughan CA, Fleming SB, Ueda N & Nettleton PF (2002) Vascular endothelial growth factors encoded by Orf virus show surprising sequence variation but have a conserved, functionally relevant structure. J Gen Virol 83, 2845–2855.

23 Keyt B, Nguyen H, Berleau L, Duarte C, Park J, Chen H & Ferrara N (1996) Identification of vascular endo- thelial growth factor determinants for binding KDR and Flt-1 receptors. J Biol Chem 271, 5638–5646. 24 Li B, Fuh G, Meng G, Xin X, Gerritsen M, Cunning-

ham B & deVos A (2000) Receptor-selective variants of human vascular endothelial growth factor. J Biol Chem 275, 29823–29828.

11 Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Buttner M, Rziha HJ & Dehio C (1999) A novel vascular endothelial growth factor encoded by Orf virus, VEGF-E, mediates angiogenesis via signalling through VEGFR-2 (KDR) but not VEGFR-1 (Flt-1) receptor tyrosine kinases. EMBO J 18, 363–374.

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

216

M. K. Inder et al.

Role of viral VEGF’s C-terminus in VEGFR binding

growth factor: crystal structure and functional mapping of the kinase domain receptor binding site. Proc Natl Acad Sci USA 94, 7192–7197.

25 Kiba A, Yabana N & Shibuya M (2003) A set of Loop- 1 and -3 structures in the novel vascular endothelial growth factor (VEGF) family member, VEGF-E NZ7, is essential for the activation of VEGFR-2 signalling. J Biol Chem 278, 13453–13461.

32 Wiesmann C, Fuh G, Christinger HW, Eigenbrot C, Wells JA & de Vos AM (1997) Crystal structure at 17 A˚ resolution of VEGF in complex with domain 2 of the Flt-1 receptor. Cell 91, 695–704.

33 Autiero M, Luttun A, Tjwa M & Carmeliet P (2003)

26 Makinen T, Veikkola T, Mustjoki S, Karpanen T, Cati- mel B, Nice EC, Wise LM, Mercer A, Kowalski J, Ker- ajaschki D et al. (2001) Isolated, lymphatic endothelial cells transduce growth, survival and migratory signals via the VEGF-C ⁄ D receptor VEGFR-3. EMBO J 20, 4762–4773.

Placental growth factor and its receptor, vascular endo- thelial growth factor receptor-1: novel targets for stimu- lation of ischemic tissue revascularization and inhibition of angiogenic and inflammatory disorders. J Thromb Haemost 1, 1356–1370.

27 Stacker SA, Vitali A, Caesar C, Domagala T, Groenen LC, Nice E, Achen MG & Wilks AF (1999) A mutant form of vascular endothelial growth factor (VEGF) that lacks VEGF receptor-2 activation retains the ability to induce vascular permeability. J Biol Chem 274, 34884– 34892.

28 Clauss M, Weich H, Breier G, Knies U, Rockl W,

34 Gabrilovich D, Ishida T, Oyama T, Ran S, Kravtsov V, Nadaf S & Carbone DP (1998) Vascular endothelial growth factor inhibits the development of dendritic cells and dramatically affects the differentiation of multiple hematopoietic lineages in vivo. Blood 92, 4150–4166. 35 Kiba A, Sagara H, Hara T & Shibuya M (2003) VEG- FR-2-specific ligand VEGF-E induces non-edematous hyper-vascularization in mice. Biochem Biophys Res Commun 301, 371–377.

Waltenberger J & Risau W (1996) The vascular endo- thelial growth factor receptor Flt-1 mediates biological activities. Implications for a functional role of placenta growth factor in monocyte activation and chemotaxis. J Biol Chem 271, 17629–17634.

36 Savory LJ, Stacker SA, Fleming SB, Niven BE & Mer- cer AA (2000) Viral vascular endothelial growth factor plays a critical role in orf virus infection. J Virol 74, 10699–10706.

29 Selvaraj SK, Giri RK, Perelman N, Johnson C, Malik P & Kalra VK (2003) Mechanism of monocyte activa- tion and expression of proinflammatory cytochemokines by placenta growth factor. Blood 102, 1515–1524.

30 Shibuya M (2001) Structure and dual function of vascu- lar endothelial growth factor receptor-1 (Flt-1). Int J Biochem Cell Biol 33, 409–420.

37 Julenius K, Molgaard A, Gupta R & Brunak S (2005) Prediction, conservation analysis and structural charac- terization of mammalian mucin-type O-glycosylation sites. Glycobiology 15, 153–164.

31 Muller YA, Li B, Christinger HW, Wells JA, Cunning- ham BC & de Vos AM (1997) Vascular endothelial

FEBS Journal 275 (2008) 207–217 ª 2007 The Authors Journal compilation ª 2007 FEBS

217