| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Institute of Bioengineering, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland;
Tumor Biology ImClone Systems, New York, New York, USA; and
Experimental Therapeutics, ImClone Systems, New York, New York, USA
2Correspondence: Institute of Bioengineering, School of Life Sciences, LMBM, Station 15, École Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland. E-mail: melody.swartz{at}epfl.ch
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: VEGF-C • mouse • in vitro wound healing • vasculogenesis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
. In tumor models, neutralization of VEGF-C with soluble VEGFR-3 reduced lymphangiogenesis and lymphatic function at the periphery of VEGF-C-overexpressing tumors (5)
. Signaling through VEGFR-3 has therefore been implicated in tumor lymph node metastases (6
7
8)
. Indeed, this importance has been demonstrated through the inhibition of VEGFR-3, which prevented prostate and melanoma (9)
and breast cancer metastases (10)
by suppressing tumor-associated lymphangiogenesis.
However, lymphangiogenesis is a complex process comprising multiple events, including lymphatic endothelial cell (LEC) proliferation, migration, organization, and vessel maturation. Although VEGFR-3 signaling has been shown to be unnecessary for lymphatic maintenance and function of existing lymphatic vessels (5)
, its roles in each of the various steps of lymphatic development remain unclear. Proteolytically processed (mature) VEGF-C and -D can also bind to VEGFR-2 (11
12
13)
, and the activation of VEGFR-2 can also induce or guide lymphangiogenesis (14
, 15)
. Furthermore, it has been reported that VEGFR-2 can form heterodimers with VEGFR-3 (16
, 17)
, but it is unclear whether heterodimer phosphorylation signals different functional responses than homodimer phosphorylation. It is also not known how these receptors interact from a functional perspective and how their different signaling patterns may affect different stages of lymphatic development.
Here, we explore the individual and combined roles of VEGFR-2 and VEGFR-3 signaling in three distinct components of adult lymphangiogenesis—LEC proliferation, migration, and vessel organization—using complementary in vivo and in vitro models. We use a model of adult skin regeneration in the mouse tail that we developed previously to investigate the role of interstitial flow in the organization of lymphangiogenesis during dermal wound healing. In that study, we showed that lymphangiogenesis proceeds by unidirectional migration of LECs in the direction of interstitial fluid flow along fluid channels, followed by subsequent organization into a functional network of lymphatic vessels in a manner reminiscent of vasculogenesis (18
, 19)
. Thus, in this model, adult lymphangiogenesis can be clearly divided temporally into three discrete phases—migration (determined by the distal-to-proximal distribution of cells in the regenerating region), proliferation (determined by total LEC number), and capillary organization (determined by immunostaining thick sections and via microlymphangiography)—allowing us to investigate the importance of VEGFR-2 and VEGFR-3 signaling at each phase.
Using this model, we previously demonstrated that endogenous VEGF-C expression is increased during lymphatic migration and proliferation but decreased during later stages of lymphatic organization (18
, 19)
and that delivery of exogenous VEGF-C induced lymphatic hyperplasia without improving lymphatic organization or function beyond control levels (20)
. Additionally, in this same model, blockage of VEGFR-3 with the monoclonal antibody (mAb) mF4-31C1 prevented regeneration of functional lymphatic vessels in the mouse skin without affecting preexisting vessels (21)
. These findings suggest that while VEGF-C/VEGFR-3 signaling is necessary for lymphangiogenesis, it may not serve an important role in the functional or organizational evolution of lymphangiogenesis. On the basis of these results, we hypothesized that VEGFR-2 and VEGFR-3 signaling are important primarily during the early stages of lymphangiogenesis, which require LEC migration and proliferation but are less important during the subsequent organization of LECs into functional lymphatic capillaries. To this end, antagonist mAbs against mouse VEGFR-2 and -3 were delivered at different stages of physiological lymphatic regeneration in adult mouse skin to determine the specific roles of these receptors in lymphangiogenesis. To support our in vivo results, we also examined the importance of these receptors using in vitro models of human LEC migration, proliferation, and 3-D tubulogenesis in the presence or absence of antagonist mAbs to human VEGFR-2 and VEGFR-3.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and/or anti-mouse VEGFR-3 (mF4–31C1) (21)
were injected intraperitoneally (i.p.) every two days. For in vitro studies, anti-human VEGFR-2 (IMC-1121a, 20 µg/mL) and anti-human VEGFR-3 (hF4–3C5, 10 µg/mL) (23)
were used.
In vivo studies
Lymphangiogenesis model
For all studies, 6- to 8-wk-old, female BALB/c mice (Charles River Laboratories, L’Arbresle, France) were used; at least three mice were used for each condition at each time point examined. Mice were anesthetized with an i.p. injection of ketamine (65 mg/kg), xylazine (13 mg/kg), and acepromazine (2 mg/kg). An analgesic, butorphanol (0.05 mg/kg), was administered subcutaneously (s.c.) twice daily for three days following the procedure. All protocols were approved by the Veterinary Authorities of the Canton Vaud according to Swiss law (protocol number #1687).
The regenerating region of skin was created as described previously (18)
. Briefly, a 2-mm wide circumferential band of dermal tissue (in which the lymphatic network in the tail skin is contained) was excised midway up the tail, leaving the underlying bone, muscle, major blood vessels, and tendons intact. The area was then covered with a close-fitting, gas-permeable silicone sleeve and filled with type I rat tail collagen (BD Pharmingen, San Diego, CA, USA). The collagen scaffold provides a preexisting matrix, in which epithelial and subepithelial tissues readily regenerate. LECs later observed within this region were thus the result of de novo cell migration, proliferation, and organization.
The neutralizing antibodies were then administered as described above for durations that varied by experimental group (Table 1
).
|
Detection of functional lymphatics via microlymphangiography
To visualize lymph flow patterns in situ, a 2 mg/mL solution of fluorescein-conjugated dextran of 70 kDa (Molecular Probes, Carlsbad, CA, USA) was injected intradermally into the tail tip at a constant pressure where it was taken up and transported by the lymphatics in the proximal direction, revealing functional lymphatic vessels and fluid channels. Fluorescence images were captured with a Zeiss MRm camera on a Zeiss Axiovert 200 M fluorescence microscope.
Immunofluorescence and immunohistochemistry
Tail specimens were cut into either 12- or 60-µm-thick longitudinal cryosections and immunostained. To detect LECs, a rabbit polyclonal antibody (pAb) against the lymphatic-specific hyaluronan receptor LYVE-1 (Upstate, Charlottesville, VA, USA) was used along with an Alexa Fluor 488 conjugated goat anti-rabbit secondary antibody (Ab) (Molecular Probes). To detect blood endothelial cells, a rat polyclonal CD31 Ab (BD Pharmingen) was used along with an Alexa Fluor 546-conjugated goat antirat secondary Ab (Molecular Probes). Cell nuclei were labeled with 4',6'-diamidino-2-phenylidole (DAPI) (Vector Laboratories, Burlingame, CA, USA). Thin-section images of the regenerating region were imaged (as above) and assembled into complete montages in Photoshop (Adobe Systems, San Jose, CA, USA). Thick sections were imaged with confocal microscopy on a Zeiss LSM 510 META confocal microscope; maximum projections are presented. LECs were defined as cells with a nucleus completely surrounded by LYVE-1 labeling. After defining the borders of the regenerating region, the number of LECs within each half were counted and summed across three random 12-µm sections from each animal.
In vitro studies
Lymphatic endothelial cell culture
Human dermal lymphatic endothelial cells (LECs) were isolated from neonatal foreskins, as described previously (24)
and maintained in EBM basal media (Cambrex, Walkersville, MD, USA) supplemented with 20% FBS, 1% penicillin-streptomycin-amphotericin B (both GIBCO, Carlsbad, CA, USA), 1 µg/mL hydrocortisone, and 50 µmol/l DBcAMP (both Sigma, St. Louis, MO, USA) in collagen type I-coated flasks (50 µg/mL). They were used until passage 8.
In vitro proliferation assay
The effects of VEGFR-2 and VEGFR-3 signaling on LEC proliferation were evaluated using a colorimetric bromodeoxyuridine (BrdU) kit (Calbiochem, San Diego, CA, USA). LECs were serum-starved for 2 h, seeded onto a collagen-coated, 96-well plate (3x104 cells/well), and allowed to adhere for an additional 2 h. One-hundred microliters of full medium were supplemented with 100 ng/mL recombinant human wild-type (WT) VEGF-C (rhVEGF-C, 2179-VC-025, R&D Systems, Minneapolis, MN), along with either 1) 10 µg/mL Hf4–3C5, 2) 20 µg/mL IMC-1121a, 3) a combination of both antibodies, or 4) no antibodies. Proliferation (proportional to BrdU incorporation) was measured after 24, 48, and 72 h of receptor blockage. Samples were fixed and permeabilized; then BrdU incorporation was detected according to the manufacturer’s instructions.
In vitro migration assay
Polycarbonate transwell inserts (8 µm pore, Millipore, Billerica, MA, USA) were coated with 50 µg/mL type I collagen (BD Biosciences). LECs were seeded at a density of 105 cells per insert in basal growth medium. The lower chamber consisted of either basal media alone or full-growth media supplemented with 100 ng/mL rhVEGF-C and either 1) 10 µg/mL Hf4–3C5, 2) 20 µg/mL IMC-1121a, 3) a combination of both antibodies, or 4) no antibodies. Appropriate neutralizing antibodies were also added to the top chamber for the duration of the experiment. Chambers were incubated at 37°C for 24 h before the samples were washed with PBS and fixed in methanol at 4°C. Nonmigrated cells were removed with a cotton swab. The membranes were then removed and mounted with Vectashield containing DAPI (Vector). Samples were visualized with a x40 objective, and the numbers of migrated cells within 8 random fields of view were counted. Migration per insert was then calculated and expressed as a fold increase over control.
In vitro tubulogenesis assay
A gel suspension assay was used to assess LEC organization, independent of proliferation and migration, into capillary tubules within a 3D environment. LECs were seeded within Growth Factor Reduced Matrigel (BD Biosciences) at densities of either 0.5 x 106 (low density) or 1.5 x 106 cells per mL (high density). The solution was plated into 8-well coverslip chamber slides (Lab-TEK Nalge Nunc, Naperville, IL, USA), and the gel was allowed to polymerize for 2 h before treatment. Gels were subsequently maintained at 37°C/5% CO2 in either basal media or full media supplemented with 100 ng/mL rhVEGF-C and either 1) 10 µg/mL hF4–3C5, 2) 20 µg/mL IMC-1121a, 3) a combination of both antibodies, or 4) no antibodies. After 6 days, gels were fixed in 2% PFA, and cell structures were visualized with Alexafluor 488-conjugated phalloidin (Molecular Probes) and counterstained with DAPI. Samples were visualized using confocal microscopy (as above). Tube length per unit area was calculated using Image J software (NIH, Bethesda, MD, USA).
Immunoprecipitation and immunoblot for receptor phosphorylation
LECs were grown to 95% confluence in 10-cm culture dishes. Cells were washed in PBS and serum starved for 48 h and then incubated in the presence of either 10 µg/mL hf4–3C5, 10 µg/mL IMC-1121a, a combination of both, or neither for 30 min. Samples were then treated with 100 ng/mL WT rhVEGF-C for 20 min, lysed with modified radio-immunoprecipitation assay (RIPA) buffer supplemented with a protease inhibitor cocktail (% PMSF; Pierce, Rockford, IL, USA) and phosphatase inhibitor cocktails (Sigma). Cleared lysate supernatants were then incubated with either 2 µg/ml anti-human VEGFR-3 (sc-321, Santa Cruz Biotechnologies, Santa Cruz, CA, USA) or 2 µg/ml anti-human VEGFR-2 (AF357, R&D Systems) and 30 µl Protein A/G agarose beads (Pierce) overnight at 4°C. The pellets were then washed 3 times in inhibitor-supplemented lysis buffer, washed once in PBS, and boiled in laemmli sample buffer. Protein samples were then separated by SDS-PAGE, transferred onto nitrocellulose membrane, and phosphorylated VEGFR-2 or VEGFR-3 was detected with mouse anti-human PY20 (1.0 µg/mL, Upstate) and an HRP-conjugated secondary Ab (Amersham Biosciences, Uppsala, Sweden). Total receptor was detected with either anti-VEGFR-3 (sc-321) or anti-VEGFR-2 (sc-6251; Santa Cruz).
Statistical methods
For determination of LEC numbers in the regenerating region, at least three sections were counted per specimen. Data are presented as means ± one SD. All P values were calculated using a two-sided Student’s t test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A) demonstrated that unorganized LECs began to migrate into and proliferate within the distal half of the regenerating region 10 days postsurgery (Fig. 1B
). At day 17, LECs were present in both distal and proximal halves but were largely unorganized. At day 60, LECs in the control mice were highly organized into lymphatic vessels, indicating a regenerated lymphatic capillary network. All LYVE-1-positive cells (and structures) were confirmed to be LECs in this model (Supplemental Fig. 1
).
|
Microlymphangiography, in which a fluorescent tracer is injected intradermally into the tip of the tail and taken up by the functional distal lymphatics, confirmed that the newly formed functional lymphatic vessels in the regenerating region did not sprout from preexisting vessels but instead organized after migration (Fig. 1C
). At both days 10 and 17, prior to organization of LECs into regenerating lymphatic capillaries, upstream lymph collected at the distal boundary and slowly diffused across the regenerating region. At day 60, when LECs were organized into vessels, upstream lymph was transported through the region inside regenerated hexagonal lymphatic vessels.
Effects of VEGFR-3 neutralization on lymphatic regeneration in vivo
To determine the involvement of VEGFR-3 in different stages of lymphangiogenesis, VEGFR-3 was inhibited during different stages of lymphatic regeneration. Consistent with our previous observations (21)
, we found that inhibition of VEGFR-3 from day 0 to day 60 completely prevented lymphangiogenesis, as demonstrated by both immunohistochemical and functional analyses (Fig. 1D-E
). When VEGFR-3 blockade was imposed subsequent to the initiation of lymphangiogenesis but prior to substantial lymphatic migration (days 10–60 of regeneration), LECs failed to migrate into the regenerating region and no functional vessels were formed. This finding confirmed that VEGFR-3 signaling is necessary for LEC migration into regenerating tissue.
At day 17 of normal regeneration, we found large numbers of unorganized LEC clusters, but no tubular capillary-like structures (Fig. 1F
); thus, we also inhibited VEGFR-3 from day 17 to day 60. Surprisingly, LEC clusters efficiently organized into functional vessels despite the absence of VEGFR-3 signaling and were indistinguishable from regenerated lymphatics in control mice (Fig. 1D, E
). Thus, VEGFR-3 is crucial for the migration and colonization of the wound during regenerative lymphangiogenesis but not for their organization into functional vessels.
LECs appear inside the regenerating region as a result of their proliferation, migration, or both, since the regenerating region is initially acellular. Furthermore, it has been established that LECs migrate and organize from the distal end to the proximal end in this model, consistent with the direction of fluid flow (18
, 19)
. Thus, LECs found in the regenerating region result primarily from cell migration from the distal host dermis. The total number of LECs and their distal-to-proximal distribution in the regenerating region are indicative of the degree of LEC migration and proliferation, respectively, over a given period of time. In untreated mice, quantitative measurements of LECs (LYVE-1-labeled cells in thin cryosections) in the distal vs. proximal halves of the regenerating region confirmed their unidirectional migration, with cells initially entering the distal region at day 10 and gradually migrating until, by day 60, the LECs are uniformly distributed throughout the entire region (Fig. 1F
and G). Blocking VEGFR-3 activation from day 10 to 60 or from day 17 to 60 significantly decreased the total number of LECs as compared to untreated controls (Fig. 1F
). Blocking of VEGFR-3 also prevented LEC migration into the proximal portion of the regenerating region (Fig. 1G
), resulting in the skewing of the LEC distribution toward the distal region. Specifically, we observed a significant difference between the percentage of LECs in the distal vs. proximal portions of the regenerating regions in mice treated with blocking mAbs to VEGFR-3 during days 10–60 and days 17–60. These distributions were remarkably similar to the distributions seen at days 10 and 17 in the wounds of control animals (Fig. 1G
), lending further credence to arrested migration. Their overall numbers were similar to those when blocking was begun as well, attesting to arrested proliferation.
Blocking VEGFR-3 from day 17 to 60 prevented further LEC migration and proliferation in regenerating tissue but did not hinder the organization of LECs into functional capillaries within the regenerating region. Therefore, these findings strongly suggest that VEGFR-3 activation is critical for LEC migration and proliferation but not their functional organization into lymphatic vessels.
Effects of VEGFR-2 neutralization on adult lymphangiogenesis
The effects of blocking VEGFR-2 activation on lymphangiogenesis were similar to those resulting from blocking VEGFR-3 (Fig. 2
A). First, VEGFR-2 blocking from day 0–60 completely blocked any lymphangiogenic response, consistent with other reports that VEGFR-2 is essential to lymphangiogenesis (14
, 25
, 26)
. VEGFR-2 blocking from day 10–60 also resulted in the prevention of lymphatic migration, while blocking from day 17–60 allowed existing LECs in the regenerating region to organize into lymphatic vessels. These data suggest that the activation of VEGFR-2 is also necessary for LEC proliferation and migration but not for LEC organization into functional vessels. However, while analysis of total LEC numbers in the regenerating region (Fig. 2C
) when VEGFR-3 was blocked between days 17 and 60 demonstrated that the number of LECs remained constant at the level present at the initiation of treatment (day 17), LEC numbers following VEGFR-2 blocking for the same period were reduced. This may reflect a small role of VEGFR-2 in LEC survival during later stages of lymphangiogenesis.
|
Effects of combined VEGFR-2 and VEGFR-3 neutralization on adult lymphangiogenesis
Since the neutralization of either VEGFR-2 or VEGFR-3 signaling alone inhibited LEC migration and proliferation but did not prevent their subsequent organization into lymphatic capillaries, we hypothesized that VEGFR-3 may be redundant with VEGFR-2 for lymphatic organization. To explore this possibility, mice were treated with both antagonist antibodies to VEGFR-2 and VEGFR-3 during various periods of skin regeneration (blocking between day 0 and 60, between day 10 and 60, and between day 17 and 60). Lymphatic regeneration was completely arrested in all cases (Fig. 2B-D
), confirming the hypothesis and demonstrating the necessity for signaling of at least one of these receptors for all stages of lymphangiogenesis.
Roles of VEGFR-2 and VEGFR-3 signaling in LEC proliferation in vitro
The processes necessary for lymphatic regeneration, LEC proliferation, migration, and capillary organization, can be more readily assessed using in vitro studies that allow the roles of VEGFR-2 and VEGFR-3 signaling on each of these processes to be investigated in greater detail. Unlike the multitude of potentially contributing factors in the regeneration region that may cloud the in vivo assessments, these LEC-only in vitro studies permit a transparent examination of the specific impacts receptor blocking has on LEC behavior. To assess the roles of VEGFR-2 and VEGFR-3 on proliferation, migration, and organization in vitro, we first confirmed that our cultured LECs possessed functional receptors by immunoprecipitation (Fig. 3
A). Stimulation with recombinant human WT VEGF-C-induced receptor signaling via tyrosine phosphorylation of both VEGFR-2 (top two panels) and VEGFR-3 (lower two panels). Furthermore, the efficiency of both human blocking antibodies was clarified. Neutralization of VEGFR-3 specifically prevented VEGFR-3 phosphorylation but showed little to no effect on VEGFR-2 signaling. Likewise, neutralization of VEGFR-2 blocked its phosphorylation down to baseline levels (seen when no VEGF-C was added to the medium) but allowed VEGFR-3 to phosphorylate. We also saw the presence of VEGFR-2/-3 heterodimers, as seen in each of the immunoprecipitation experiments (i.e., a lower MW band (corresponding to VEGFR-3) was phosphorylated along with the 200-kDa band (corresponding to VEGFR-2) that was coimmunoprecipitated (top).
|
Complete medium supplemented with 100 ng/mL VEGF-C promoted LEC proliferation (7.2±0.3-fold increase over matching control after 72 h). Blocking the activation of either VEGFR-2 or VEGFR-3 significantly (but not completely) inhibited LEC proliferation (to 5.0±0.3-fold and 3.7±0.4-fold, respectively; Fig. 3B
). Combined inhibition of both receptors further reduced proliferation (to 2.2±0.3-fold).
Role of VEGFR-2 and VEGFR-3 signaling in LEC migration in vitro
The relative importance of VEGFR-2 and VEGFR-3 signaling in LEC migratory responses in vitro mirrored those seen in vivo (Fig. 3C
). Full media supplemented with 100 ng/mL recombinant human WT VEGF-C (untreated) induced a 14 ± 3-fold increase in migration over basal levels. Inhibition of either VEGFR-2 or VEGFR-3 significantly decreased migration by 58% and 75%, respectively. Furthermore, combined inhibition of both receptors completely abolished LEC migration stimulated by VEGF-C. Thus, the activation of both receptors in vitro was necessary for LEC migration in an apparently additive manner.
Role of VEGFR-2 and VEGFR-3 signaling in LEC organization in vitro
Finally, we examined the roles of VEGFR-2 and VEGFR-3 in capillary organization using in vitro tubulogenesis experiments under conditions that do not require LEC proliferation.
When seeded at 1.5 x 106 cells/mL, LECs readily formed tubular networks with numerous filopodia between cells of neighboring structures (Fig. 4
A). Similar multicellular structures were also seen in cultures with VEGFR-3 blocking. Organization was also seen when VEGFR-2 was blocked, although this was hindered in comparison to control and VEGFR-3 blockage. When the receptors were blocked together, almost no tubular structures were observed. To quantify these observations, the organizational capacity of LECs within a 3D matrix for each culture condition was measured as total tube length per unit area (Fig. 4B
). Although capillary structures were clearly visible, VEGFR-2 blockage resulted in reduced tube length compared with full media supplemented with VEGF-C alone. In contrast, blocking VEGFR-3 had no effect on the ability of LECs to form tube-like networks. Low density (0.5x106 cells/mL) seeded LECs failed to organize with or without blocking (data not shown), suggesting that LEC organization in 3D gels requires a minimum seeding density or maximum LEC-LEC distance for cell signaling.
|
These observations corroborate our in vivo data and demonstrate that while lymphatic capillary organization does not require VEGFR-3 signaling when VEGFR-2 signaling is functional, one or the other is needed, and VEGFR-2 appears to be more important than VEGFR-3 in LEC organization.
VEGF-C receptor ligation was required for VEGFR-3 phosphorylation and signaling events to occur. Neutralization of VEGFR-3 before exposure to VEGF-C, eliminated VEGFR-3 phosphorylation but blockage of VEGFR-2 did not effect VEGFR-3 signaling capabilities.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although less physiologically relevant, the in vitro studies allowed further confirmation of receptor roles specifically in LEC proliferation, migration, and organization in the absence of other cells or confounding cellular events that might also be affected by VEGFR-2 and VEGFR-3. Consistent with our in vivo data, VEGFR-3 was completely redundant with VEGFR-2 signaling for LEC tubulogenesis. In order for tube formation to show receptor redundancy, high seeding density was necessary to minimize the amount of migration required for vessel formation, since lower seeding densities did not promote tubulogenesis under any blocking condition. Again, this redundancy was consistent with our in vivo data, where the VEGFR-3-independent organization of new lymphatic capillaries occurred only if sufficient time were allowed for LECs to proliferate and migrate into the region of regeneration. Likewise, failures in organization, specifically in the proximal half of the regenerating region, are thus likely the result of a sparse LEC population at the start of receptor blocking. Thus, the strong corroboration between our in vivo and in vitro data indicates that the effects seen in vivo are most likely due to receptor activation on LECs and not on other cells that might express VEGFR-3, such as macrophages (25
, 27
, 28)
.
It has been previously reported that heterodimerization with VEGFR-2 may be required for VEGFR-3 phosphorylation in aortic endothelial cells (16)
, but LECs are known to express higher levels of VEGFR-3 than blood endothelial cells (12
, 24)
, particularly those from large vessels, and it has also been shown that both VEGFR-3 homodimers and VEGFR-3/-2 heterodimers can form in human LECs (17)
. Our data support the concept that both homodimer and heterodimer forms play different functional roles, and indeed, we see that both forms are phosphorylated with VEGF-C (Fig. 3A
). However, we were not able to conclusively determine which dimer pairs (or combinations thereof) are required for the different stages of LEC proliferation, migration, and capillary organization.
Our findings have important implications for VEGF-C therapies designed to induce lymphangiogenesis and help corroborate seemingly conflicting reports on its effects. If VEGFR-3 activation is largely redundant for lymphatic organization, but necessary for LEC migration and proliferation, as our data show, then VEGF-C therapy would only be potentially useful in augmenting lymphangiogenesis in areas where lymphatics are not present or are present in suboptimal densities. It may also have the effect of making normally regenerating lymphatics more hyperplastic rather than creating a denser lymphatic capillary network, as VEGF-C overexpression has been reported to lead to increased lymphatic diameter (hyperplasia) without increased lymphatic density (20
, 29
, 30)
. Furthermore, reports that VEGF-C delivery results in increased lymphangiogenesis and/or improved lymphatic function come largely from models in which lymphatic function was disrupted by wounding (31
32
33)
; in this case, VEGF-C could act by restoring lymphatic proliferation and migration into and across the wound, where lymph or interstitial flow would then help drive lymphatic organization. Finally, VEGF-C delivery in tissues that are normally devoid of lymphatics, such as the cornea would certainly be required to initiate lymphatic endothelial cell migration and proliferation (25
, 34
, 35)
without necessarily being required for lymphatic organization. Taken together, these data suggest that VEGF-C therapy may be successful in cases in which lymphangiogenesis needs to be initiated (as in the case of congenital lymphedema) but not in cases where lymphatic capillaries are intact but poorly functional (e.g., if the tissue matrix were badly damaged, such as following radiation therapy) or blocked downstream (e.g., following lymph node resection).
In summary, our results demonstrate that VEGFR-3 cooperates with VEGFR-2 in early stages of lymphangiogenesis by inducing LEC migration and proliferation but serves redundant functions in later stages of lymphatic capillary organization. Importantly, we show that VEGFR-3 signaling is required neither for the organization of lymphatic capillaries nor for establishing or maintaining lymphatic function, which may rely instead on the activation of other lymphatic receptors such as Tie2 (35
, 36)
and neuropillin-2 (37)
, as well as on functional cues such as interstitial fluid flow (18
, 19)
. These results also have important implications for antilymphangiogenesis therapy and corroborate recent evidence (10)
demonstrating that combined inhibition of VEGFR-2 and VEGFR-3 may more effectively reduce tumor lymphangiogenesis and consequent lymphatic metastasis than the inhibition of either receptor alone.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication June 15, 2006. Accepted for publication November 9, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Uzarski, M. B. Drelles, S. E. Gibbs, E. L. Ongstad, J. C. Goral, K. K. McKeown, A. M. Raehl, M. A. Roberts, B. Pytowski, M. R. Smith, et al. The resolution of lymphedema by interstitial flow in the mouse tail skin Am J Physiol Heart Circ Physiol, March 1, 2008; 294(3): H1326 - H1334. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
200-kDa and 110-kDa bands, which can be inhibited when VEGFR-3 is neutralized. The second blot shows total VEGFR-3. IC indicates where an isotype-matched control antibody against PDGFR-alpha, IMC-3G3, was used in place of blocking antibodies in the medium. B) After 72 h of treatment, blockade of either VEGFR-2 or VEGFR-3 reduced VEGF-C-induced LEC proliferation. This was further inhibited when both receptors were blocked simultaneously. C) Blocking either VEGFR-2 or VEGFR-3 substantially decreased VEGF-C-induced LEC migration, while combined blockade completely abolished this migratory response. **P < 0.01 vs. untreated, #P < 0.05 vs. other conditions as indicated.