Type IV collagen is a major structural component of all basement membranes in vertebrates.1–3 Type IV collagen consists of 6 distinct α-chains (α1–α6).4–13 All α-chains are composed of 3 domains, the N-terminal 7S domain, the middle triple helical domain and the C-terminal globular noncollagenous (NC1) domain.14 The α-chains self-assemble into helical trimers via their NC1 and 7S domains and further form a 3D scaffold.15, 16 This scaffold interacts with other basement membrane constituents, such as laminins and nidogens, to form a network. This network is essential for tissue function by providing structural and endothelial cell support, serving as ligands for cell receptors and regulating angiogenesis.6, 17–25 While the α1 and α2 chains are ubiquitously found in all human basement membranes,3, 5 the other 4 chains are expressed in a tissue and organ specific distribution.1, 3, 26
Because of limited diffusion of oxygen in tissues, almost all cells in the body are found within a maximum radius of 100–200 μm around a blood vessel.27–29 Without neovascularization, tumors usually cannot exceed a size of 1 mm3 and therefore cannot develop an invasive phenotype.27, 29 Angiogenesis, the formation of new capillaries from preexisting blood vessels, is critical to many normal physiological processes.30, 31 Mammal angiogenesis is highly regulated, and it occurs during development, reproduction or wound healing processes.29, 32 Angiogenesis is essential for the progression of various pathological disorders including diabetic retinopathy, rheumatoid arthritis, as well as tumor growth and metastasis.27–31 The switch to an angiogenic phenotype requires both the upregulation of angiogenic stimulators and the downregulation of angiogenesis inhibitors.31, 33, 34 The precise molecular mechanisms leading to the “angiogenic switch” are not yet completely understood.27–29, 35, 36 In this regard, more than 20 stimulators of angiogenesis and more than 30 angiogenesis inhibitors have been identified.27, 34
An angiogenic phenotype of tumors requires upregulation of angiogenesis stimulators, and downregulation of angiogenesis inhibitors.31, 33 VEGFs, Angiopoëtins and Ephrins effect primarily the vascular system.28 Vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) are among the major inducers of angiogenesis.27, 28, 35, 37 To date, a number of endogenous angiogenesis inhibitors have been identified, and certain factors such as Angiostatin,38 Endostatin,39 Canstatin,18 Arresten17 and Tumstatin,6, 19, 40 are proteolyzed fragments of larger proteins and are likely generated in vivo. In cancer and several other diseases, uncontrolled capillary growth can have dramatic consequences, and our understanding of the role of these endogenous angiogenesis inhibitors in this setting is still evolving.
Several studies have shown that inhibitors of collagen metabolism have anti-angiogenic properties; this supports the notion that basement membrane collagen synthesis and deposition is crucial for blood vessel formation and survival.6, 17 Additionally, NC1 domains of type IV collagen have been identified as novel integrin ligands and regulators of endothelial proliferation, migration and angiogenesis.4, 26, 34, 41, 42 Here, we study the function of the NC1 domain of the α6 chain of type IV collagen (α6NC1) in the regulation of angiogenesis and tumor growth.43
Material and methods
Wild type C57Bl/6 mice were purchased from The Jackson Laboratories (Bar Harbor, ME: http://www.jax.org). Transgenic Rip1Tag2 mice were obtained from NCI mouse models of human cancers consortium (MMHCC: http://www.cancer.gov). The generation and phenotypic characterization of Rip1Tag2 mice have been described previously.44 Mice were maintained at the Beth Israel Deaconess Medical Center animal facility under standard conditions. The standard chow of all β-tumor-bearing mice was supplemented with glucose pellets (Harlan Teklad) beginning at 8 weeks of age, to prevent hypoglycemia because of hyperinsulinemia from developing insulinoma. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center.
Recombinant production of α6NC1 in E. coli
Human α6NC1 was produced in E. coli as a fusion protein with a N-terminal poly-histidine tag. The sequence encoding α6NC1 was amplified by PCR from a commercial cDNA library (CLONTECH, Palo Alto, CA) using the following primers: 5′-GCT AGC GAG CAT GAG AGT GGG CTA CAC G-3′ as forward primer and 5′-GGG CCC GTG GCA GGT GCC ACC CTA CAG GC-3′ as reverse primer. The primers for the α6(IV) NC1 domain were designed based on published sequences for human type IV collagen chains.45, 46 The resulting cDNA fragment was inserted into pET-28b(+) vector (Novagen, Madison, WI) after the N-terminal poly-histidine tag. Plasmid constructs encoding α6NC1 were first transformed into NovaBlue competent cells (Novagen, Madison, WI) and then transformed into BL21 (DE3) for protein expression (Novagen). For protein expression, an overnight bacterial miniculture was used to inoculate a 1,000 ml maxiculture Luria-Bertani (LB) medium containing Kanamycin (30 μg/ml). This culture was grown for ∼4 hr until the cells reached an OD600 of 0.6. Protein expression was induced by adding 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG). After about 2 hr of induction, cells were harvested by centrifugation at 10,000g and lyzed by resuspension in 6 M guanidine, 0.1 M NaH2PO4, 0.01 M TRIS-HCl (pH 8.0) and 1.5% TritonX-100. Resuspended cells were incubated for 1 hr at +4°C on a shaker, then briefly sonicated and centrifuged at 12,000g for 30 min at +4°C. The supernatant was passed over a 2 ml Ni-nitrilotriacetic acid-agarose column (TALON metal affinity resin, BD Biosciences) 4–6 times at a speed of 2 ml/min. Nonspecifically bound protein was removed by washing with both 6 M guanidine and 8 M urea in 0.1 M NaH2PO4, 0.01 M TRIS-HCl, 1.5% TritonX-100 (pH 8.0). α6NC1 protein was eluted from the column with 3 concentrations of imidazole (50, 125 and 250 mM) in 8 M Urea, 0.1 M NaH2PO4, 0.01 M TRIS-HCl (pH 8.0). The eluted protein was dialyzed twice against PBS at 4°C, a part of the total protein precipitated during dialysis. Dialyzed protein was collected and centrifuged at 5,000g; pellet and supernatant were collected for further analysis. Protein concentration in each fraction was determined using the BCA assay kit (Pierce) according to the manufacturer's recommendations. The purity of the protein was determined by SDS-PAGE and Western blot. The total yield from 1-l culture was ∼600 μg of soluble, and ∼3,200 μg of insoluble α6NC1, respectively. Endotoxin levels were determined using the QCL-1000® Chromogenic Limulus Amebocyte Lysate (LAL) Endpoint Assay (Cambrex, East Rutherford, NJ) according to the manufacturer's recommendations. Contamination with bacterial endotoxin in all of the protein preparations was <13 endotoxin units (EU)/mg protein (equivalent to <0.39 EU/ml). It has been shown that addition of exogenous endotoxin at a concentration of 0.5 EU/ml (1 EU/ml equals 0.1 ng/ml lipopolysaccharide) to the cell culture media does not inhibit endothelial cell proliferation nor affect cell viability in a 48-hr incubation period. Polymyxin B (5 μg/ml) was used to inactivate endotoxin, and no difference of the inhibitory effect could be detected in experiments lacking Polymyxin B.
HPLC reversed-phase chromatography
HPLC reversed-phase chromatography was performed as previously described.47 In brief, 5 μg of Ni-NTA purified recombinant human α6NC1 protein expressed in E. coli (see above) was loaded onto a NovaPak 3.9-mm C-18 column (Waters, Milford, MA) and developed with a linear gradient of 10–80% acetonitrile during 45 min. Two-milliliter fractions were collected, lyophylized, resuspended in SDS sample buffer and analyzed by immunoblotting.
The LLC cell line (Lewis lung carcinoma cells) was purchased from the American Type Culture Collection (Rockville, MD), and cells were maintained in Dulbecco's Modified Eagle Medium (DMEM, Gibco), supplemented with 10% fetal calf serum (Gibco) and 100 μg Plasmocin (InvivoGen, San Diego, CA) at 37°C in a 5% CO2 atmosphere.
In vivo tumor studies
Lewis lung carcinoma tumor burden study
Lewis lung carcinoma cells were harvested from cell culture and 1 × 106 LLC cells (in sterile PBS) were injected subcutaneously into 8- to 10-week-old male wild type C57Bl/6 mice. The mice were divided into 2 groups as a tumor volume of ∼100 mm3 was observed. Mice in each experimental group received a daily intravenous injection of α6NC1 (1 mg/kg BW [30 μg/ml]) in a total volume of 100 μl 10% DMSO in PBS. The control group received equal volumes of 10% DMSO in PBS without α6NC1. Mice were sacrificed after 10 days or when a tumor volume of 3,000 mm3 was observed. Tumor volume was measured every other day starting on Day 0 (beginning of treatment). Tumor length and width were measured using a digital Vernier caliper, and tumor volume was calculated using the standard formula for approximating the volume of a spheroid (length × width2 × 0.52).17, 39 Volume ± SEM was plotted over the treatment period. Additionally, on the day of sacrifice the tumors were weighed, and tumor weight ± SEM was plotted.
Regression trial in Rip1Tag2 transgenic mice
The mice in these studies were males and females of the RIP1Tag2 transgenic mouse lineage (C57Bl/6J background). Mice were treated from 12 to 16 weeks.48 Mice in the experimental group received a daily intravenous injection of 1 mg/kg BW α6NC1 [30 μg/ml] in a total volume of 100 μl 10% DMSO in PBS. The control group daily received equal volumes of 10% DMSO in PBS.
Assessment of tumor burden in RIP1Tag2 transgenic mice
Mice (6 per group) were sacrificed at the end of the trial and tumors were microdissected from freshly excised pancreas. Tumor length and width were measured using a Vernier caliper, and tumor volume was calculated using the standard formula for approximating the volume of a spheroid (length × width2 × 0.52).39, 48
Matrigel plug assay
In vivo Matrigel plug assay was performed as previously described.49 Matrigel™ (BD Biociences, Franklin Lakes, NJ) was thawed overnight at 4°C. Before injection Matrigel was mixed with 20 U/ml Heparin (Pierce, Rockford, IL), 150 ng/ml basic fibroblast growth factor (bFGF; R&D, Minneapolis, MN) and 30 μg/ml α6NC1 (1 mg/kg BW). Control groups received no angiogenesis inhibitor. The Matrigel mixture was injected subcutaneously into the back of 10- to 12-week-old male C57BL/6 mice (each group n = 10) using a 21-gauge needle. Ten days after Matrigel injection, mice were sacrificed and the Matrigel plugs removed. After fixation in 10% formalin, the Matrigel plugs were embedded in paraffin, sectioned and stained with hematoxylin and eosin (HE). Sections were examined by light microscopy, and the number of blood vessels from 10 different high-power fields (HPFs) was counted and averaged. Photomicrographs were taken using a Zeiss Axioscop 2plus fluorescence microscope and Axiovision digital imaging software (Zeiss, Oberkochen, Germany).
Cell proliferation and viability assay
Cell proliferation was determined using the colorimetric WST-1 cell proliferation assay (Roche Applied Systems GmbH, Mannheim, Germany) according to the manufacturer's recommendations. Calf pulmonary artery endothelial cells (CPAE) were obtained from American Type Culture Collection (ATCC, Rockville, MD), and were maintained in Dulbecco's modified Eagle medium (DMEM, Gibco), supplemented with 10% fetal calf serum (Gibco) and 100 μg Plasmocin (InvivoGen, San Diego, CA) at 37°C in a 5% CO2 atmosphere. CPAE cells were seeded at a concentration of 4 × 103 cells/well into a 96-well plate in 100 μl DMEM containing different concentrations of α6NC1. Polymyxin B (Sigma, St. Louis, MO) at a concentration of 5 μg/ml was used to inactivate endotoxin.50 After incubation for 48 hr at 37°C and 5% CO2, 10 μl of cell proliferation reagent WST-1 was added into each well and incubated for 4 hr at 37°C in a 5% CO2 atmosphere. The absorbance was measured at 450 nm against a blank background control using a microplate reader. All experiments were performed in quadruplicate.
Pooled human umbilical vein endothelial cells (HUVEC) were obtained from Cambrex (East Rutherford, NJ) and were maintained in EBM media (Cambrex) supplemented with the EGM-2 BulletKit® (Cambrex) at 37°C in a 5% CO2 atmosphere. Cells of passages 2 through 6 were used for all experimental procedures. HUVECs were seeded at a concentration of 4 × 103 cells/well into a 96-well plate in 100 μl EGM-2 containing different concentrations of α6NC1. Polymyxin B (Sigma) at a concentration of 5 μg/ml was used to inactivate endotoxin.50 After incubation for 48 hr at 37°C and 5% CO2, 10 μl of cell proliferation reagent WST-1 was added into each well and incubated for 4 hr at 37°C in a 5% CO2 atmosphere. The absorbance was measured at 450 nm against a blank background control using a microplate reader. All experiments were performed in quadruplicate.
Gel electrophoresis and Western blotting
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed as previously described.47, 51 For reducing SDS-PAGE, 30 μg protein were diluted with SDS sample buffer, boiled for 5 min and separated on 10% SDS polyacrylamide gels. Gels were either stained with Coomassie Brilliant Blue, or proteins were transferred to Immobilon P membranes (Millipore, Billerica, MA) by semidry blotting. Unspecific binding sites were blocked with blocking solution (5% milk in TBST) for 30 min at room temperature (RT). Primary antibodies (polyclonal rabbit anti-α6NC1antibody; monoclonal anti-His-Tag antibody) diluted in blocking solution) were used at a 1:2,000 dilution, and horseradish peroxidase-conjugated secondary antibodies (Promega, Madison, WI) were used at a 1:10,000 dilution (in blocking solution for 1 hr at RT. Immunoreaction was visualized by enhanced chemiluminescence (ECL; Amersham, Freiburg, Germany) and film exposure (Denville, Metuchen, NJ).
Immunofluorescence labeling was performed as described previously.6, 47, 51 In brief, frozen sections of LLC tumor xenografts were analyzed for intratumoral microvessel density (MVD). Rat anti-CD31 antibody (Pharmingen, San Diego, CA) was used at a 1:50 dilution (in blocking solution 1% BSA) and applied for 1 hr at room temperature, followed by Rhodamine-conjugated anti-rat IgG (1:200 in blocking solution) for 1 hr at room temperature. CD31 positive blood vessels in 10 randomly selected high-power fields (HPF) were counted and averaged. Photomicrographs were taken using a Zeiss Axioscop 2plus fluorescence microscope and Axiovision digital imaging software (Zeiss, Oberkochen, Germany).
All values are expressed as mean ± SEM. Analysis of variance (ANOVA) was used to determine statistical differences between groups using SPSS software. Bonferroni Post Hoc analysis was performed, when equal variances were assumed. Dunnett's T3 Post Hoc analysis was performed, when equal variances were not assumed. A level of p < 0.05 was considered statistically significant.
Human α6NC1 was produced in E. coli as a fusion protein with a N-terminal poly-histidine tag using pET-28b(+) bacterial expression plasmid and was purified by Ni-NTA agarose column chromatography. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the purified protein revealed a monomeric band of about 25 kDa (Fig. 1a). The α6NC1 protein was analyzed by Western blot using a rabbit polyclonal antibody raised against the NC1 domain of the α6 chain of type IV collagen (α6(IV) NC1) and a monoclonal anti-polyhistidine tag antibody (Fig. 1b). HPLC reversed-phase chromatography of the purified protein revealed 1 major peak at a retention time (RT) of 33.2–35.2 min using a linear acetonitrile gradient (51.6–54.75% acetonitrile) (Figs. 1c and 1d). The recombinant human α6NC1 protein had endotoxin levels <13 EU/mg protein (equivalent to <0.39 EU/ml) in all of the protein preparations, as analyzed by the QCL-1000® LAL Endpoint Assay. These levels are considered below the threshold of any activity against eukaryotic cells.
During the process of angiogenesis, endothelial cell proliferation is required to form new blood vessels. Therefore, we assessed the inhibitory effect of α6NC1 on endothelial cell proliferation. In proliferation assays, a dose-dependent inhibition of 10% serum-stimulated calf pulmonary artery endothelial cells (CPAE) was detected with an ED50 (Effective dose, 50%) of 5 μg/ml (Fig. 2a) using soluble protein produced in E. coli. Furthermore, human umbilical vein endothelial cell (HUVEC) proliferation was also inhibited by α6NC1 in a dose-dependent fashion (Fig. 2b). Polymyxin B (5 μg/ml) was used to inactivate endotoxin, and a difference in the inhibitory effect was not detected (Figs. 2a and 2b).
To test the in vivo activity of α6NC1, we performed Matrigel plug assays52 in normal C57Bl/6 mice. Matrigel™ was injected in the presence of bFGF with or without α6NC1. A 50% reduction in the number of blood vessels was observed at a concentration of 1 mg/kg of α6NC1 (Fig. 3).
Next, 2 established mouse tumor models were used to test the antitumor activity of α6NC1. α6NC1 inhibits the growth of LLC tumors when compared to placebo-treated mice (Fig. 4). α6NC1was also used for a tumor regression trial using Rip1Tag2 transgenic mice, which develop spontaneous pancreatic carcinoma. α6NC1 significantly suppresses growth of pancreatic carcinoma in Rip1Tag2 mice when compared to placebo-treated mice (Fig. 5). The inhibition of tumor growth was associated with a significant decrease in the microvessel density in the tumor tissue (Fig. 6).
Recombinant human α6NC1 protein inhibits endothelial cell proliferation and in vivo neovascularization. α6NC1 also suppresses the tumor growth in 2 different mouse models of cancer.
To date, a number of matrix derived endogenous angiogenesis inhibitors have been identified, such as Endostatin,39 Canstatin,18 Arresten,17 Tumstatin6, 19, 40 and Endorepellin.53, 54 Endostatin was identified as the NC1 domain of the α1 chain of type XVIII collagen; Canstatin, Arresten and Tumstatin are derived from type IV collagen chains.6, 17–19, 40 The α1, α2 and α3 NC1 domains of type IV collagen were identified as inhibitors of angiogenesis and tumor growth, whereas the NC1 domains of the α4 and α5 chain did not reveal this anti-angiogenic property. In this study we demonstrate that the α6NC1 domain exhibits an anti-angiogenic property and inhibits tumor growth. Several type IV collagen NC1 domains exhibit anti-angiogenic property and have been given unique nomenclature to highlight such novel function. Therefore, we propose that the α6NC1 domain be designated Hexastatin, to highlight the fact that this NC1 domain is derived from the α6 chain of type IV collagen and causes “stasis” of angiogenesis and tumor growth.
Arresten binds to α1β1 integrin on proliferating endothelial cells, and its antiangiogenic activity is mediated by this integrin.17, 55 Canstatin binds to ανβ3 and α3β1 integrin on proliferating endothelial cells,18, 43, 56, 57 and Tumstatin binds to endothelial cells via ανβ3 integrin and α6β1 integrin,6, 58, 59 respectively. Thus, it is likely that α6NC1 also binds to an endothelial cell integrin, and that its activity maybe mediated by an integrin. Future studies will likely shed more light on the mechanism of α6NC1-mediated inhibition of angiogenesis.
This work was supported by grants from the US National Institutes of Health DK55001 (R.K.), DK62987 (R.K.), DK61688 (R.K.), AA53194 (R.K.), a research fund from the Division of Matrix Biology at Beth Israel Deaconess Medical Center, the Stop and Shop Pediatric Tumor Foundation (H.S.). T.M.M. is supported by the German Research Foundation DFG (MU 2298/2-2).