A vaccine targeting angiomotin induces an antibody response which alters tumor vessel permeability and hampers the growth of established tumors
Angiomotin (Amot) is one of several identified angiostatin receptors expressed by the endothelia of angiogenic tissues. We have shown that a DNA vaccine targeting Amot overcome immune tolerance and induce an antibody response that hampers the progression of incipient tumors. Following our observation of increased Amot expression on tumor endothelia concomitant with the progression from pre-neoplastic lesions to full-fledged carcinoma, we evaluated the effect of anti-Amot vaccination on clinically evident tumors. Electroporation of plasmid coding for the human Amot (pAmot) significantly delayed the progression both of autochthonous tumors in cancer prone BALB-neuT and PyMT genetically engineered mice and transplantable TUBO tumor in wild-type BALB/c mice. The intensity of the inhibition directly correlated with the titer of anti-Amot antibodies induced by the vaccine. Tumor inhibition was associated with an increase of vessels diameter with the formation of lacunar spaces, increase in vessel permeability, massive tumor perivascular necrosis and an effective epitope spreading that induces an immune response against other tumor associated antigens. Greater tumor vessel permeability also markedly enhances the antitumor effect of doxorubicin. These data provide a rationale for the development of novel anticancer treatments based on anti-Amot vaccination in conjunction with chemotherapy regimens.
KeywordsAngiomotin DNA vaccination Vessel permeability Antibodies Chemotherapy
Oncoantigens are self-molecules expressed at the tumor site that play a significant role in promoting tumor growth and can be the targets of anti-tumor vaccines . Cellular  and DNA  vaccines may overcome immune tolerance and trigger a protective immune response against oncoantigens overexpressed by tumor cells [4, 5] or in tumor microenvironment . Following vaccine-elicited antibody and cell-mediated immune attack, oncoantigens cannot be easily down-modulated nor negatively immunoedited as a consequence of their cancer-driving role . However, while oncoantigen vaccines effectively and persistently hamper the expansion of incipient tumors, their efficacy fades away when they are administered to mice bearing advanced tumors. Their protective potential is thus restricted to tumor prevention [5, 7].
In sharp contrast to these common findings, here we show that a DNA vaccine against an oncoantigen expressed by normal, but overexpressed by endothelial cells of tumor vessels  also inhibits the growth of large established clinically evident tumors.
As angiogenesis drives tumor growth, inhibition of its underlying signalling pathways through vaccine-induced immune response could provide a new way of intervening with the progression of established tumors.
Induction of tumor angiogenesis is regulated by numerous pro- and anti-angiogenic factors . Angiostatin specifically hampers endothelial cell migration and tumor vascularisation in mouse tumor models . Its anti-migratory effects are mediated by angiomotin (Amot), one of angiostatin receptors . Amot is a membrane-associated protein present on the endothelial cell surface of angiogenic tissues  characterized by conserved coiled-coil and carboxy termini-PDZ domains . A shorter angiomotin isoform (p80) confers a hyper-migratory and invasive phenotype in transfected cells  and induces endothelial cell migration during angiogenesis . The longer (p130) isoform localizes to tight junctions, regulates cell shape and appears to play a role in the later phase of angiogenesis .
We have shown that a DNA vaccine targeting Amot can overcome immune tolerance and induce a significant antibody response that mimic the effect of angiostatin. These antibodies inhibit endothelial cell migration, block tumor cell- and basic fibroblast growth factor-induced angiogenesis in the matrigel plug assay and prevent growth of transplanted tumors without impairing normal stromal or retina vessels . We now show that antibodies elicited alter tumor vessel permeability and structure. These multifaceted effects of vaccine-induce anti-Amot antibodies lead to inhibition of established clinically evident mammary tumors, massive tumor perivascular necrosis, and an effective tumor antigen presentation resulting in a form of epitope spreading that induces an immune response against other oncoantigens overexpressed by tumor cells. Greater tumor vessel permeability also boosts the local accumulation of drugs and enhances their antitumor effect. These data provide a rationale for the development of fresh anticancer treatments based on anti-Amot vaccination in conjunction with chemotherapy regimens.
Materials and methods
Inbred female BALB/c mice, either wild type or overexpressing the transforming activated rat HER-2/neu oncogene under control of the MMTV promoter (BALB-neuT mice) , were bred under specific pathogen-free conditions by Biogem (Ariano Irpino, Italy) or at the Molecular Biotechnology Center (Turin, Italy). Mice overexpressing the polyoma virus middle T under control of the MMTV promoter (PyMT) were purchased from Jackson Laboratories (Bar Harbor, Maine). Mice were treated in conformity with European Guidelines and policies as approved by the University of Turin Ethical Committee.
Culture cell lines
TUBO cells, a cloned rat Her2/neu+ cell line established from a lobular carcinoma of a BALB-neuT mouse , were cultured in Dubecco’s Modified Eagle Medium supplemented with GlutaMAX™ I, d-glucose, HEPES buffer (DMEM; Gibco, Rockville, MD), and 20% fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO). Mouse aortic endothelial cells (MAE) transfected with human p80 Amot (MAE Amot) or empty vector (MAE vector)  were maintained in DMEM (Gibco) containing 10% FBS, 1% Penicillin–Streptomycin (Sigma-Aldrich) in the presence of 5 μg/ml Puromycin (Gibco). HMEC-1 (American Type Culture Collection, Manassas, VA) were cultured onto EC attachment factor (Gibco) coated tissue culture plates in EndoGRO™ medium (Millipore, Billerica, MA) supplemented with 5% FBS (Sigma) and 10 mM l-glutamine, 5 ng/ml epidermal growth factor, 1 μg/ml hydrocortisone, 0.75 U/ml heparin sulfate, 50 μg/ml ascorbic acid (all from MiIlipore) and 1% Penicillin–Streptomycin (Sigma-Aldrich).
Wild type BALB/c mice were challenged subcutaneously in the inguinal region with the lethal dose of 1 × 105 TUBO cells. The progression of autochthonous mammary carcinomas in the mammary glands of both BALB-neuT and PyMT mice and of subcutaneous TUBO tumors was monitored weekly. Progressively growing masses with a mean diameter of >1 mm were regarded as tumors, measured with a caliper in the two perpendicular diameters, and the mean diameter was recorded. Progression of tumors was monitored until tumor masses were evident in all ten mammary glands (BALB-neuT and PyMT mice) or until a tumor exceeded a mean diameter of 10–15 mm, when mice were sacrificed for humane reasons. In a few experiments mammary glands with tumor of progressive stages and subcutaneous tumors of different sizes were collected, stored in RNA later (Sigma-Aldrich) at 4°C for 24 h, and then snap-frozen in liquid nitrogen and stored at −80°C until they were processed for morphological analysis.
RNA extraction and quantitative real time PCR (qPCR)
Total RNA was isolated from specimens using an IKA-Ultra-Turrax® T8 homogenizer (IKA®-Werke, Staufen, Germany) and TRIzol® reagent (Invitrogen, San Diego, CA). Genomic DNA contaminations were removed from total RNA with the DNA-free kit (Ambion, Austin, TX). RNA quality was estimated by the Agilent 2100 Bioanalyzer (Agilent Technologies, Milan, Italy) and RNA was quantified with a NanoVue Plus Spectrophotometer (GE Healthcare, Milan, Italy). Total RNA was divided into aliquots and stored at −80°C till use. To detect Amot mRNA, 1 μg of DNAse-treated RNA (DNA-free™ kit) was retrotranscribed with RETROscript™ reagents (Ambion) and qRT-PCRs were carried out using gene-specific primers (QuantiTect Primer Assay; Qiagen, Chatsworth, CA), SYBR green and a 7900HT Fast Real Time PCR System (Applied Biosystems, Milan, Italy). Quantitative normalization was performed on the expression of beta-actin and the between-sample relative expression levels were calculated using the comparative delta Ct (threshold cycle number) method (2−ΔΔCt) with a control sample as the reference point.
Protein preparation and immunoblotting
Total protein extracts were obtained by using a boiling buffer containing 0.125 M Tris/HCl, pH 6.8 and 2.5% sodium dodecyl sulphate (SDS). 50 μg proteins were separated by SDS-PAGE and electroblotted onto polyvinylidene fluoride membranes (BioRad, Hercules, CA). Membranes were blocked in 5% Blotto non-fat milk (Santa Cruz, CA) Tris buffered saline (TBS)-Tween buffer (137 mM NaCl, 20 mM Tris/HCl, pH 7.6, 0.1% Tween-20) for 1 h at 37°C, then incubated with appropriate primary and secondary antibodies in 1% milk TBS-Tween buffer, overnight at 4°C and for 1 h at room temperature respectively, and visualized by enhanced chemiluminescence (ECL®, Amersham Biosciences, Piscataway, NJ). The anti-Amot rabbit polyclonal antibody (TLE) was raised against a peptide corresponding to the 24 C-terminal amino acids of Amot, as previously described ; anti-Vinculin, goat anti-mouse IgG HRP-conjugated, goat anti-rabbit IgG HRP-conjugated monoclonal antibodies were all from Santa Cruz.
In vivo treatments
Empty pcDNA3 plasmid and plasmid coding for human p80 Amot (pAmot) were generated as previously described in details . Fifty μg of plasmid in 20 μl of 0.9% NaCl were injected in the quadriceps muscle of anesthetized mice. Immediately after the injection, two 25-ms trans-cutaneous electric low voltage pulses with an amplitude of 150 V and a 300-ms interval were administered at the injection site via a multiple needle electrode connected to an electroporator (Cliniporator™, IGEA s.r.l., Carpi, Italy). BALB-neuT and PyMT transgenic mice were vaccinated twice at 16th–18th and 6th–8th weeks of age, respectively. BALB/c mice injected with a lethal dose of TUBO cells were vaccinated when tumor mass reached 4 mm mean diameter and 7 days after. Few mice so treated were also injected intravenously (i.v.) once, 2 days after the second vaccination, with the maximum tolerated dose of doxorubicin (Sigma-Aldrich) (10 mg/kg of body weight).
To evaluate the presence of anti-Amot and anti-Her2/neu (anti-neu) antibodies, sera from mice were collected 1 week after the last vaccination and tested by ELISA. 96 well/plates (Costar®, Sigma-Aldrich) were coated with 100 ng/well of recombinant human-Amot (Origene, Rockville, MD) or recombinant rat Her2/neu protein (Genway, San Diego, CA), overnight at 4°C. Coated plates were then blocked with 3% non-fat milk (Santa Cruz) in TBS-Tween buffer for 2 h at 37°C. Plates were incubate with sera diluted in blocking buffer (1:100 or 1:1,000) overnight at 4°C. To remove any non-specific antibody or excess serum protein, plates were washed 3 times with TBS-Tween buffer. HRP-conjugated anti-mouse IgG antibody (Sigma-Aldrich) was diluted 1:2,000 in blocking buffer and incubated for 1 h at 37°C. Plates were washed as described above for 6 times, followed by the addition of chromogenic 3,3′,5,5′-Tetramethylbenzidine substrate (Sigma-Aldrich). Reaction was stopped by addition of HCl 2 N and the optical density measured at 450 nm with a microplate reader (680XR, BioRad). Anti-Amot and anti-neu IgG isotype titration was performed by ELISA assay as described above using rat biotin-conjugated anti-mouse IgG1, IgG2a, IgG2b, IgG3 (BD Pharmingen, San Diego, CA) as secondary antibodies. Plates were then incubated for 30 min with streptavidin-HRP (R&D Systems, Minneapolis, MN) diluted 1:200 in TBS-Tween buffer and then reactions were carried forward as described above.
HMEC-1 cells were plated in 96-well plates at a density of 5 × 103 cells/well in EndoGRO™ medium (Millipore) with 2% FBS. After 18 h medium was removed and cell were fixed with 2.5% glutaraldehyde and stained with 0.1% crystal violet to determine t0. Cell proliferation was then stimulated adding fresh medium with 5% FBS. At the same time purified IgG from sera of pAmot or pcDNA3 vaccinated mice were added at the concentration of 20 μg/ml. Total IgG were purified using Melon™ Gel IgG Spin Purification kit (Thermo Scientific, Milan, Italy) according to manufacturer’s instructions. After 48 h incubation, cells were fixed with 2.5% glutaraldehyde and stained with 0.2% crystal violet. The dye was solubilized using 10% acetic acid, and optical density was measured with a Microplate Reader 680 XR (BioRad) at 570 nm wavelength. Data are presented as mean ± SEM of three replicates.
Dynamic contrast enhanced magnetic resonance imaging (DCE–MRI)
For the study of Amot expression in tumors, tumor cryosections (10 μm) were prepared from frozen samples using a Microm HM 560 cryostat. Sections were dried for 1 h and fixed in 4% paraformaldehyde for 10 min at room temperature and subsequently permeabilized with PBS-0.1% Triton X-100 (Sigma-Aldrich) for 5 min. Tumor sections were blocked in 5% horse serum (Sigma-Aldrich) and incubated with appropriate primary antibody for 1 h at room temperature. After 1 h incubation with secondary antibodies the sections were mounded with flouromount with DAPI (Sigma-Aldrich) and visualized using the Zeiss 700 laser scanning microscope or the Zeiss Axioplan 2 fluorescence microscope. TLE was the primary anti-Amot antibody . All secondary antibodies were from Molecular Probes (Alexa Fluor, Invitrogen).
Vessel permeability was investigated by injecting 50 nm polymer microspheres (Duke Scientific, Palo Alto, CA) diluted in 0.9% NaCl (Sigma-Aldrich) to a volume of 100 μl into the tail vein 6 h before sacrifice. Tumors were then fixed in paraformaldehyde 1% for 1 h at 4°C, rinsed several times with PBS, infiltrated overnight with 30% sucrose in PBS at 4°C, embedded in Optimum Cutting Temperature (OCT; Bio Optica, Milan, Italy) compound and then frozen at −80°C. Tumor vasculature and necrosis were evaluated on tumors fixed in 10% neutralized formaldehyde solution and embedded in paraffin, or fixed in pyridoxal phosphate and embedded in OCT. 4 μm thick sections were stained with anti-CD31 antibody (Santa Cruz). Other sections were stained with hematoxylin and eosin as previously described in detail .
Statistical differences were evaluated through the GraphPad software 5.0 (GraphPad Inc. San Diego, CA) by using Mantel-Cox log-rank test for the incidence of autochthonous tumors in transgenic mice; Yates’s χ2 test for the regression of transplantable tumors. All other statistical differences were determinate with Student’s t test.
Amot expression increases at later stages of cancer progression
Amot expression levels was analyzed in in vitro cultured TUBO cells as well as in TUBO tumors grown in BALB/c mice (Fig. 1b, c). Even if Amot transcript was present (Fig. S1b), Western blot analysis showed that Amot protein was undetectable on cultured TUBO cells (Fig. 1b) while it was evident in established TUBO tumors (Fig. 1c). Immunofluorescence analysis on cryosections of established TUBO tumors (Fig. 1d) and autochthonous carcinomas of BALB-neuT (Fig. 1e) and PyMT mice (Fig. S2) disclosed Amot expression on endothelial cells of tumor vessels.
Anti-Amot vaccination hampers the growth of autochthonous mammary carcinomas in BALB-neuT and PyMT mice
Anti-Amot vaccination hampers the growth of established transplantable tumors
Anti-Amot vaccination increases tumor vessel permeability
Studies with fluorescent microspheres administered i.v. 7 days after the second electroporation (Fig. 4e, f) shows a significant microsphere extravasation and trapping in the payload of perivascular areas of the tumors from pAmot immunized mice (Fig. 4f). Almost no extravasation was found in pcDNA3 electroporated mice (Fig. 4e).
Vessel alteration following anti-Amot vaccination results in a major epitope spreading
TUBO tumors overexpress the rat Her2/neu receptor . Even so, their growth in BALB/c mice does not elicit an immune recognition of the tumor associated antigens nor a reactive immune response [16, 27]. However, ELISA assay showed that anti-neu antibodies were present in sera from pAmot vaccinated mice (Fig. 3b). Interestingly, both slow and fast progressors developed a significant anti-neu antibody response, indicating that pAmot vaccination was sufficient to induce tumor antigen recognition, independently from the titer of anti-Amot antibodies induced and the effect on tumor growth. Also anti-neu antibodies were mainly IgG2a and IgG2b (Fig. S3).
Vessel permeability following anti-Amot vaccination improves chemotherapy
The data reported here show that a DNA vaccination targeting Amot in tumor vasculature impairs the progression of established experimental mammary tumors.
This inhibition was observed in two distinct strains of cancer-prone genetically engineered mice whose females undergo mammary carcinogenesis. The aggressive progression of mammary lesions in BALB-neuT mice is driven by the activated rat HER-2/neu oncogene . Atypical hyperplasia already evident in puberal BALB-neuT mice progresses into multiple foci of invasive metastatic cancer by week 16, when pAmot DNA vaccination was started . Even quicker evolution of mammary lesions is driven by polyoma middle T in PyMT mice. The hyperplastic lesion evident in the mammary glands at 4 weeks of age already progresses to the stage of early carcinoma by week 8 , when these mice received the second pAmot vaccination. The markedly delayed onset of palpable tumors following pAmot vaccination acquires a special significance considering that anti-neu DNA vaccination is only able to elicit a marked protection when administered to BALB-neuT mice at the stage of multiple in situ carcinomas, whereas its protective ability is already marginal when it is administered at week 16 . This fading of the protection is a common finding with anti-oncoantigen vaccines. The elicited response is effective until the incipient tumor is formed by a limited number of cells, while it loses its efficacy against a large number of proliferating and genetically instable tumor cells . On the other hand, the growth of an established tumor relies on tumor ability to induce neovascularization and blood supply. The initial stages of a mammary tumorigenesis take advantage from the vascular tree of mammary glands, whereas as the carcinoma enlarges its dependence from newly formed Amot-positive capillary sprouts increases. Moreover, reactive lymphocytes and antibodies more readily reach their target on endothelial cells than on the cells of an established tumor .
Also in PyMT mice pAmot vaccination started at week 6 and 8 of age markedly extends the time of appearance of a palpable tumor and the overall survival. As comparison, in PyMT mice successful vaccination against alpha-lactoalbumin administered at the 6th week of age no more than slightly reduces the size of growing tumors .
pAmot vaccination performed on BALB/c mice challenged with TUBO cells and bearing a fast growing, clinically evident 4 mm mean diameter tumor was also still able to markedly impair its growth in a way directly proportional to the titer of vaccine-induced anti-Amot antibodies.
This unique ability of anti-Amot vaccination to hamper the progression of established autochthonous and transplantable tumors classifies Amot as a target oncoantigen of special interest because of its restricted spatiotemporal expression .
Amot overexpression was found in vessels of transplantable and autochthonous experimental mammary carcinomas , in vessels of human breast tumors where it correlates with the metastatic spreading  and in vessels associated with Kaposi’s sarcoma . Systemic and local treatment with a anti-Amot monoclonal antibody prevents pathological vessel formation associated with tumor growth . In BALB-neuT mice Amot expression in the mammary glands becomes prominent only following the vigorous angiogenic switch that accompanies the progression of preneoplastic lesions towards invasive cancer, characterized by burgeoning capillary sprouts [15, 21]. In the same way, Amot expression is not detectable in in vitro cultured TUBO cells. However, following a TUBO cell challenge, its expression increases at the tumor site with the expansion of tumor-induced neovascularisation. This restricted overexpression of Amot accounts for the absence of autoimmune reactions affecting the vascularisation of normal mice . pAmot vaccination does not impair the fertility of immunized female mice nor their ability to deliver and feed newborn mice fully normal in number and size (not shown).
We have previously shown that the immune response that follows pAmot DNA vaccination mostly, if not completely, rests on the induction of anti-Amot antibodies . This finding fits in well with the direct correlation between the titer of anti-Amot antibodies elicited by the vaccine and the intensity of the inhibition of tumor progression we have now observed. To tease apart the main features associated with the inhibition of tumor progression we exploited mice challenged with TUBO cells. First, histological analysis showed that following pAmot vaccination tumor vessels become larger and give rise to saccular and lacunar spaces. This acquired vessel pattern was associated with numerous areas of tumor perivascular necrosis. To get an assessment of how vessel modification following pAmot vaccination leads to perivascular necrosis we exploited DCE–MRI, a noninvasive technique measuring a combination of tumor perfusion and vessel permeability  that is currently used in clinical practice to detect early changes in the tumor induced by antiangiogenic therapies [34, 35]. We exploited the B22956/1 CA , a Gd-chelate with high affinity for human serum albumin in order to exploit the contrast efficiency showed by this system when combined with 1T MRI scanner  and the increased specificity for the detection of vascular permeability in comparison to small molecular weight Gd-complexes .
In pAmot vaccinated mice B22956/1 accumulates in the tumor area more than in control mice, showing that tumor vessel permeability and tumor perfusion are markedly changed. The higher DCE–MRI values of the slope of the curve to reach the peak enhancement and the higher peak itself show that B22956/1 perfuses the tumor area of pAmot vaccinated mice better and with faster kinetics. These findings suggesting that pAmot vaccination increases number and dimension of tumor vessel fenestrations are endorsed by the higher spread of fluorescent microspheres of dimensions similar to B22956/1 when bound to the human serum albumin.
The enhanced permeability attained by tumor vessels leads to a selective microsphere extravasation at the tumor site and the entrapment and persistence of the CA used in DCE–MRI in the perivascular interstitial spaces of tumor area. Histological observation shows that these features acquired by tumor vessels following pAmot vaccination correlate with multiple areas of perivascular necrosis of the tumor. Capture and processing of the dying tumor cells by antigen presenting cells recruited in the spread areas of perivascular necrosis account for the onset of a significant antibody response against neu, a major oncoantigen overexpressed by TUBO cells .
An almost direct correlation among the titer of anti-Amot antibodies induced by pAmot vaccination and the intensity of tumor growth inhibition was evident. However, the induction of anti-neu antibodies suggests that the inhibition of neu+ TUBO tumors results from a synergistic action of anti-Amot antibodies targeting tumor vasculature and anti-neu antibodies that directly inhibit TUBO cell proliferation [38, 39]. An improved antitumor effect may stem from the combined reaction against endothelia and tumor antigens [8, 40]. Nevertheless, the lack of a correlation between anti-neu antibodies and the protective effect of anti-Amot vaccination suggests a main role of anti-Amot antibodies in the protection observed. These are mainly IgG2a and IgG2b and thus, besides their direct effect on endothelial cell proliferation, they can also act by recruiting both complement and FcγRIV expressing cells  to damage tumor vasculature.
The increased permeability acquired by tumor vessels following pAmot vaccination also allowed an otherwise ineffective single dose of doxorubicin to induce a delay of tumor growth and a complete rejection of established TUBO tumors in 4 out of 7 mice. It is well known that antiangiogenic drugs may improve the efficacy of chemotherapy. Following the treatment with a humanized monoclonal antibody neutralizing vascular endothelial growth factor (bevacizumab) human tumors showed vessel maturation and stabilization associated with enhanced tumor perfusion, cytotoxic drug delivery and often prolonged patients survival . Phase III clinical trials with bevacizumab in combination with chemotherapy have shown significant improvements in progression-free and overall survival when compared with chemotherapy alone .
In conclusion, present data extend to clinical evident tumors our previous observation on the ability of anti-Amot vaccination to hamper tumor onset . The impaired expansion of established autochthonous and transplantable tumors, the lacunar structure and increased permeability acquired by tumor vessels, perivascular tumor necrosis and the epitope spreading leading to induction of an immune response to tumor associated antigens are multiple features triggered by the vaccination against Amot. These not only concur to impair the progression of established tumors, but also make them susceptible to an otherwise poorly effective chemotherapy. These findings prospect anti-Amot vaccination alone and in combination with conventional chemotherapy as an attractive fresh strategy for tumor therapy.
We thank Prof. Guido Forni for contributing to the discussion and organization of this manuscript, and Prof. John Iliffe for its revision and editing. This project was funded under the auspices of EU Consortium of Anticancer Antibody Development (EUCAAD) 200755. The project EUCAAD has received research funding from the EU Community’s Seventh Framework Programme. Prof. Lars Holmgren is supported by grants from the Swedish Research Council, Swedish Cancer Society, Cancer Society Stockholm and FP7 EUCAAD.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.