We present a detailed protocol to generate a murine xenograft model of venous malformation. This model is based on the subcutaneous injection of patient-derived endothelial cells containing hyper-activating TIE2 and/or PIK3CA gene mutations. Xenograft lesions closely recapitulate the histopathological features of VM patient tissue.
Venous malformation (VM) is a vascular anomaly that arises from impaired development of the venous network resulting in dilated and often dysfunctional veins. The purpose of this article is to carefully describe the establishment of a murine xenograft model that mimics human VM and is able to reflect patient heterogeneity. Hyper-activating non-inherited (somatic) TEK (TIE2) and PIK3CA mutations in endothelial cells (EC) have been identified as the main drivers of pathological vessel enlargement in VM. The following protocol describes the isolation, purification and expansion of patient-derived EC expressing mutant TIE2 and/or PIK3CA. These EC are injected subcutaneously into the back of immunodeficient athymic mice to generate ectatic vascular channels. Lesions generated with TIE2 or PIK3CA-mutant EC are visibly vascularized within 7‒9 days of injection and recapitulate histopathological features of VM patient tissue. This VM xenograft model provides a reliable platform to investigate the cellular and molecular mechanisms driving VM formation and expansion. In addition, this model will be instrumental for translational studies testing the efficacy of novel drug candidates in preventing the abnormal vessel enlargement seen in human VM.
Defects in the development of the vasculature are the underlying cause of many diseases including venous malformation (VM). VM is a congenital disease characterized by abnormal morphogenesis and expansion of veins1. Important studies on VM tissue and endothelial cells (EC) have identified gain-of-function mutations in two genes: TEK, which encodes the tyrosine kinase receptor TIE2, and PIK3CA, which encodes the p110α (catalytic subunit) isoform of PI3-kinase (PI3K)2,3,4,5. These somatic mutations result in ligand-independent hyper-activation of key angiogenic/growth signaling pathways, including PI3K/AKT, thereby resulting in dilated ectatic veins3. Despite these important genetic discoveries, the subsequent cellular and molecular mechanisms triggering abnormal angiogenesis and the formation of enlarged vascular channels are still not fully understood.
During normal and pathological angiogenesis, new vessels sprout from a pre-existing vascular network and EC undergo a sequence of important cellular processes including proliferation, migration, extracellular matrix (ECM) remodeling and lumen formation6. Two- and three- dimensional (2D/3D) in vitro cultures of EC are important tools to investigate each of these cellular properties individually. Nevertheless, there is a clear demand for a mouse model recapitulating pathological vessel enlargement within the host microenvironment while providing an efficient platform for preclinical evaluation of targeted drugs for translational research.
Up to date, a transgenic murine model of VM associated with TIE2 gain-of-function mutations has not been reported. Current transgenic VM mouse models rely on the ubiquitous or tissue-restricted expression of the activating mutation PIK3CA p.H1047R3,5. These transgenic animals provide significant insight into whole-body or tissue-specific effects of this hotspot PIK3CA mutation. The limitation of these models is the formation of a highly pathological vascular network resulting in early lethality. Thus, these mouse models do not fully reflect the sporadic occurrence of mutational events and localized nature of VM pathology.
On the contrary, patient-derived xenograft models are based on the transplantation or injection of pathological tissue or cells derived from patients into immunodeficient mice7. Xenograft models are a powerful tool to broaden knowledge about disease development and discovery of novel therapeutic agents8. In addition, using patient-derived cells allows scientists to recapitulate mutation heterogeneity to study the spectrum of patient phenotypes.
Here, we describe a protocol where patient-derived VM EC which express a mutant constitutively-active form of TIE2 and/or PIK3CA are injected subcutaneously in the back of athymic nude mice. Injected vascular cells are suspended in an ECM framework in order to promote angiogenesis as described in previous vascular xenograft models9,10,11. These VM EC undergo significant morphogenesis and generate enlarged, perfused pathological vessels in the absence of supporting cells. The described xenograft model of VM provides an efficient platform for preclinical evaluation of targeted drugs for their ability to inhibit uncontrolled lumen expansion.
Patient tissue samples were obtained from participants after informed consent from the Collection and Repository of Tissue Samples and Data from Patients with Tumors and Vascular Anomalies under an approved Institutional Review Board (IRB) per institutional policies at Cincinnati Children’s Hospital Medical Center (CCHMC), Cancer and Blood Disease Institute and with approval of the Committee on Clinical Investigation. All animal procedures described below have been reviewed and approved by the CCHMC Institutional Animal Care and Use Committee.
1. Preparation of materials and stock solutions
2. Isolation of endothelial cells from VM patient tissue
3. Endothelial cell selection and expansion
4. VM patient-derived xenograft protocol
NOTE: In this protocol we use 5‒6 week old, male immunodeficient, athymic nude Foxn1nu mice.
All animal procedures must be approved by the Institutional Animal Care and Use Committee (IACUC).
5. Tissue collection and processing
6. Lesion sectioning
7. Hematoxylin and Eosin (H&E)
8. Immunohistochemistry
9. Analysis of human-derived Vascular Channels
NOTE: Vascularity of VM lesions is quantified by measuring vascular area and vascular density. Only UEA-I positive, human-derived vascular channels are considered for quantification.
This protocol describes the process of generating a murine xenograft model of VM based on the subcutaneous injection of patient-derived EC into the back of immunodeficient nude mice. Endothelial cell colonies can be harvested within 4 weeks after initial cell isolation from VM tissue or lesional blood (Figure 1A,B). The day after injection, the xenograft lesion plug covers a surface area of approximately 80‒100 mm2. In our hands, lesion plugs with TIE2/PIK3CA-mutant EC are visibly vascularized and perfused within 7‒9 days from injection 14,15 (Figure 1C-E). However, the extent of lesion growth is variable and reflects on patient and sample heterogeneity.
Lesion plugs closely recapitulate the histopathological features of human VM tissue: enlarged vascular channels lined by a thin layer of endothelial cells (Figure 1F‒H). These vascular structures typically contain erythrocytes, confirming functional anastomoses with the host mouse vasculature (Figure 1F‒H). Immunohistochemical staining using the human specific lectin UEA-I can confirm that cells lining vascular lesions are derived from human implanted cells rather than mouse vasculature (Figure 1H). A scheme summarizing the steps from VM-EC isolation to dissection of lesion plug is presented in Figure 2.
Figure 1: Representative results.
(A) Representative image of primary mixed cell culture three weeks after isolation from VM tissue before EC selection. Typical endothelial cell colony (EC) and contaminating fibroblast (FB). (B) Image of a purified (CD31 bead-selection) endothelial cell culture from VM patient-derived tissue. Scale bar = 200 µm. (C) The lesion will form a spherical structure. Vascularization is visible due to blueish color through the skin of nude mice. (D) Dashed lines show how lesion size is recoded by measuring the length (L) and the width (W) using a caliper. (E) Photo of visibly vascularized, xenograft lesion explant at day 9. Scale bar = 1 cm. (F) Representative image of a lesion plug section. The x-plane pattern in which five high power field images are taken for quantification are indicated by white dashed boxes. Scale bar = 1000 µm. (G‒H) Representative images of VM lesion plug sections. (G) Hematoxylin and Eosin staining and (H) immunohistochemistry of human specific lectin UEA-I. Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 2: Schematic of the workflow to generate a patient-derived xenograft of VM.
(A) Endothelial cells isolated from patient VM lesion solid tissue or lesional blood are plated and, when 80% confluency is reached, are selected by anti CD31-conjugated immunomagnetic beads and expanded. (B) For subcutaneous injection of EC, on day 0, skin on the backside of the mouse is pinched using forefinger and thumb to create a tent-like structure. Lesions are measured at day 1 and then every other day (red arrows) using a caliper through experimental day 9. Lesions are dissected and processed for histological analysis. Please click here to view a larger version of this figure.
Here, we describe a method to generate a patient-derived xenograft model of VM. This murine model presents an excellent system that allows researchers to gain a deeper understanding of pathological lumen enlargement and will be instrumental in developing more effective and targeted therapies for the treatment of VM. This can be easily adapted to investigate other types of vascular anomalies such as capillary lymphatic venous malformation16. There are several steps that are crucial for the successful generation of reproducible vascular lesions. First, the patient-derived endothelial cells must be pure (without the presence of other cell types) and growing exponentially at the time of injection. Contaminating fibroblast or other mesenchymal non-EC can be easily recognized by elongated morphology. On rare occasion, it is possible that even after purification using anti-CD31 antibody conjugated magnetic beads, a small number of non-EC remain in the culture. These cultures require further purification with endothelial specific cell surface markers. As an alternative approach, single cell clonal expansion of endothelial cells is possible. This would reinsure the homogeneity of mutant-EC as all of the cells within one culture would derive from one single cell. However, this approach is not recommended for VM-derived EC as cells tend to top their proliferation capabilities and convert into a senescent phenotype after 9‒10 passages. It is critical to use cells between passages 3‒8 for xenograft experiments and to not passage cells the day before the injection.
The xenograft model can be modified to investigate other vascular anomalies carrying different activating mutations. Moreover, as patient tissue samples are difficult to access for some laboratories, the xenograft model can be adapted by using EC, such as human umbilical cord blood endothelial cells (HUVEC), genetically engineered to express the mutation/s known to cause dysfunctional vascular growth15,17.
The number of cells recommended for the injection in the xenograft is 2.5 x 106 cells/200 µL of BMEM. However, if the cell number is insufficient it is possible to either reduce the number of injections per animal to one or to reduce the injection volume to a minimum of 100 µL. For the latter, it is however important to maintain the cell density ratio e.g., 1.25 x 106 cells/100 µL BMEM. When working with BMEM, all the steps must be performed on ice to avoid solidification of the cell suspension before injection. During injection, it is important and that the needle is inserted at an angle of 45° directly under the skin and away from the muscle tissue, as injecting into muscle impedes lesion reproducibility and makes the lesion dissection difficult. A total of two injections can be performed on each mouse—one on the right and one on the left side of each animal. The second injection in the same mouse can serve as a technical replicate. More injections on the back are not recommended as lesions grow over time and might interfere with each other. For statistical analysis in pre-clinical studies comparing xenograft plugs of treated versus untreated (vehicle only) mice, we recommend the use of a minimum of 5 animals (10 xenograft plugs) per study group. If available, the second injection could alternatively be used as a ‘internal control’ using non-mutant EC. We have used primary non-mutant EC, such as HUVEC, as a control and have shown that these cells formed a negligible number of small channels14,15. Furthermore, in these HUVEC control lesion plugs, we have noticed infiltration of murine-derived vascular channels into the plug after day 9. If the experimental design requires longer incubation times, these infiltrating channels can be easily excluded from analysis by staining for a human-specific marker such as human-specific CD31 antibody or Ulex europaeus agglutinin I (UEA-I) that does not cross react with mouse.
To ensure that the lesion does not become a burden to animal health and wellbeing it is important to observe lesion size, record mouse weight daily, and pay attention to any side complications such as bleeding and bruising. If the lesion volume exceeds 500 mm3, the experiment has to be terminated.
When vascular lesions are enlarged and perfused, extreme attention must be paid during dissection to avoid rupturing the lesion. It is important to avoid touching the lesion plug with dissection tools and leave excessive surrounding tissue (such as skin) attached to the plug. This prevents collapse of the vascular structures within the xenograft plug which would interfere with accurate analysis.
Finally, to maintain consistency, it is important that the initial histological analysis begins in the center of the plug (about 50‒70 µm into the tissue) rather than the border regions where anastomosing mouse vasculature might be present. It is highly recommended to stain the tissue sections with a human-specific EC marker, such as UEA-I or an alternative human-specific antibody which will not cross-react with mouse, in order to confirm that vascular structures are formed by human-derived EC rather than invading mouse EC.
The authors have nothing to disclose.
The authors would like to thank Nora Lakes for proofreading. Research reported in this manuscript was supported by the National Heart, Lung, and Blood Institute, under Award Number R01 HL117952 (E.B.), part of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Athymic nude mice, (Foxn1-nu); 5-6 weeks, males | Envigo | 069(nu)/070(nu/+) | Subcutaneous injection |
Biotinylated Ulex europeaus Agglutinin-I (UEA-I) | Vector Laboratories | B-1065 | Histological anlaysis |
Bottle top filter (500 ml; 0.2 µM) | Thermo Fisher | 974106 | Cell culture |
Bovine Serum Albumin (BSA) | BSA | A7906-50MG | Cell culture; Histological analysis |
Calcium cloride dihydrate (CaCl2.2H2O) | Sigma | C7902-500G | Cell culture |
Caliper | Electron Microscopy Sciences | 50996491 | Lesion plug measurment |
CD31-conjugated magnetic beads (Dynabeads) | Life Technologies | 11155D | EC separation |
Cell strainer (100 μM) | Greiner | 542000 | Cell culture |
Collagenase A | Roche | 10103578001 | Cell culture |
Conical Tube; polypropylene (15 mL) | Greiner | 07 000 241 | Cell culture |
Conical Tube; polypropylene (50 mL) | Greiner | 07 000 239 | Cell culture |
Coplin staining jar | Ted Pella | 21029 | Histological anlaysis |
Coverglass (50 X 22 mm) | Fisher Scientific | 12545E | Histological anlaysis |
DAB: 3,3'Diaminobenzidine Reagent (ImmPACT DAB) | Vector Laboratories | SK-4105 | Histological anlaysis |
Dulbecco's Modification of Eagle's Medium (DMEM) | Corning | 10-027-CV | Cell culture |
DynaMag-2 | Life Technologies | 12321D | EC separation |
Ear punch | VWR | 10806-286 | Subcutaneous injection |
EDTA (0.5M, pH 8.0) | Life Technologies | 15575-020 | Histological anlaysis |
Endothelial Cell Growth Medium-2 (EGM2) Bulletkit (basal medium and supplements) | Lonza | CC-3162 | Cell culture |
Eosin Y (alcohol-based) | Thermo Scientific | 71211 | Histological anlaysis |
Ethanol | Decon Labs | 2716 | Histological anlaysis |
Fetal Bovine Serum (FBS) , HyClone | GE Healthcare | SH30910.03 | Cell culture |
Filter tip 1,250 μL | MidSci | AV1250-H | Multiple steps |
Filter tip 20 μL | VWR | 10017-064 | Multiple steps |
Filter tip 200 μL | VWR | 10017-068 | Multiple steps |
Formalin buffered solution (10%) | Sigma | F04586 | Lesion plug dissection |
Hemacytometer (INCYTO; Disposable) | SKC FILMS | DHCN015 | Cell culture |
Hematoxylin | Vector Hematoxylin | H-3401 | Histological anlaysis |
Human plasma fibronectin purified protein (1mg/mL) | Sigma | FC010-10MG | Cell culture |
Hydrogen Peroxide solution (30% w/w) | Sigma | H1009 | Histological anlaysis |
ImageJ Software | Analysis | ||
Isoflurane, USP | Akorn Animal Health | 59399-106-01 | Subcutaneous injection |
magnesium sulfate heptahydrate (MgSO4.7H2O) | Sigma | M1880-500G | Cell culture |
Basement Membrane Matrix (Phenol Red-Free; LDEV-free) | Corning | 356237 | Subcutaneous injection |
Microcentrifuge tube (1.5 mL) | VWR | 87003-294 | EC separation |
Microscope Slide Superfrost (75mm X 25mm) | Fisher Scientific | 1255015-CS | Histological anlaysis |
Needles, 26G x 5/8 inch Sub-Q sterile needles | Becton Dickinson (BD) | BD305115 | Subcutaneous injection |
Normal horse serum | Vector Laboratories | S-2000 | Histological anlaysis |
Penicillin-Streptomycin-L-Glutamine (100X) | Corning | 30-009-CI | Cell culture |
Permanent mounting medium (VectaMount) | Vector Laboratories | H-5000 | Histological anlaysis |
Pestle Size C, Plain | Thomas Scientific | 3431F55 | EC isolation |
Phosphate Buffered Saline (PBS) | Fisher Scientific | BP3994 | Cell culture |
Scale | VWR | 65500-202 | Subcutaneous injection |
Serological pipettes (10 ml) | VWR | 89130-898 | Cell culture |
Serological pipettes (5ml) | VWR | 89130-896 | Cell culture |
Sodium carbonate (Na2CO3) | Sigma | 223530 | Cell culture |
Streptavidin, Horseradish Peroxidase, Concentrate, for IHC | Vector Laboratories | SA-5004 | Cell culture |
Syringe (60ml) | BD Biosciences | 309653 | Cel culture |
SYRINGE FILTER (0.2 µM) | Corning | 431219 | Cell culture |
Syringes (1 mL with Luer Lock) | Becton Dickinson (BD) | BD-309628 | Subcutaneous injection |
Tissue culture-treated plate (100 X 20 mm) | Greiner | 664160 | Cell culture |
Tissue culture-treated plate (145X20 mm) | Greiner | 639160 | Cell culture |
Tissue culture-treated plates (60 X 15) mm | Eppendorf | 30701119 | Cell culture |
Tris-base (Trizma base) | Sigma | T6066 | Histological anlaysis |
Trypan Blue Solution (0.4 %) | Life Technologies | 15250061 | Cell culture |
Trypsin EDTA, 1X (0.05% Trypsin/0.53mM EDTA) | Corning | 25-052-Cl | Cell culture |
Tween-20 | Biorad | 170-6531 | Histological anlaysis |
Wheaton bottle | VWR | 16159-798 | Cell culture |
Xylenes | Fisher Scientific | X3P-1GAL | Histological anlaysis |