Isolation of Mouse Pancreatic Endothelial Cells

Abstract

The pancreas is a vital organ for maintaining metabolic balance within the body, in part due to its production of metabolic hormones such as insulin and glucagon, as well as digestive enzymes. The pancreas is also a highly vascularized organ, a feature facilitated by the intricate network of pancreatic capillaries. This extensive capillary network is made up of highly fenestrated endothelial cells (ECs) important for pancreas development and function. Accordingly, the dysfunction of ECs can contribute to that of the pancreas in diseases like diabetes and cancer. Thus, researching the function of pancreatic ECs (pECs) is important not only for understanding pancreas biology but also for developing its pathologies. Mouse models are valuable tools to study metabolic and cardiovascular diseases. However, there has not been an established protocol with sufficient details described for the isolation of mouse pECs due to the relatively small population of ECs and the abundant digestive enzymes potentially released from the acinar tissue that can lead to cell damage and, thus, low yield. To address these challenges, we devised a protocol to enrich and recover mouse pECs, combining gentle physical and chemical dissociation and antibody-mediated selection. The protocol presented here provides a robust method to extract intact and viable ECs from the whole mouse pancreas. This protocol is suitable for multiple downstream assays and may be applied to various mouse models.

Introduction

The pancreas, key to metabolic control and homeostasis, is a highly vascularized organ. The pancreas has both endocrine and exocrine functions, controlling the regulation of blood glucose and digestive enzymes, respectively. These two compartments are linked together by the extensive network of pancreatic blood vessels, facilitating the exchange and transport of oxygen, hormones, and enzymes. Critically, this dense capillary network penetrates the Islet of Langerhans, a cluster of hormone-regulating cells within the pancreas responsible for its endocrine function, consisting of the glucagon-secreting alpha (α) cells, the insulin-secreting beta (β) cells and the somatostatin-secreting delta (δ) cells^1,2. Although the islets only make up 1-2% of the pancreatic mass, they receive 20% of total blood flow³, highlighting the importance of islet vasculature. The pancreatic capillaries are primarily made up of highly fenestrated endothelial cells (ECs), which are surrounded by mural pericytes. These capillary ECs play a vital role in the islet development, maturation, and (dys)function and form intimate crosstalks with various endo- and exocrine cells⁴ (Figure 1).

Endothelial dysfunction has been observed in both Type 1 and Type 2 diabetes, the most common conditions caused by pancreatic islet dysfunction^5,6. Both islet microvascular density and morphology can be altered in diabetes⁷. Moreover, pancreatic cancer, a highly aggressive tumor that can also be manifested as diabetes, is characterized by high microvascular density with poor perfusion⁸. Given the pivotal structural and functional roles of ECs in both normal and diseased pancreatic tissue, there is a pertinent need to study the pECs in development, physiology, and pathology to unveil the mechanisms that drive health or diseases.

Numerous protocols have been developed for the isolation of ECs from different murine (e.g., brain^9,10, lung¹¹, heart¹², liver¹³, skeletal muscles¹⁴, and adipose tissues¹⁵) and human (e.g., brain¹⁶, visceral adipose tissue^17,18, peripheral nerves¹⁹, lung^20,21,22, and mesenteric artery²³) tissues. These protocols typically involve the use of enzymatic digestions (e.g., by collagenase, trypsin²⁴, dispase^24,25, and liberase²⁶), followed by an antibody-based enrichment step. Moreover, these protocols tend to rely on extended durations of digestions in high concentrations of enzymes with vigorous agitation at 37 °C (Table 1). Due to the unique features of the pancreas, including that it houses a plethora of endogenous digestive enzymes, these existing protocols cannot be directly applied to isolate pECs. First, the extracellular matrix (ECM) composition of the pancreas is different from other tissues. While collagenase is commonly used for EC isolation, there are multiple subtypes with different tissue-specific dissociation capabilities, thus requiring optimization. Second, and crucial to pEC isolation, the release and activation of pancreatic endogenous enzymes can significantly hinder the isolation process. To this end, caution needs to be taken to minimize the rupture of the exocrine acinar cells (the primary source of zymogens, proteases, and RNase²⁷), which can induce further cell damage and result in low cell viability and overall affect recovery^27,28,29.

To address these challenges, we have adapted methods from existing EC isolation protocols and established a new protocol suitable for EC isolation from mouse pancreases. Specifically, we describe here a workflow (Figure 2) using collagenase Type I (typically implemented for lung EC isolation), lower digestion temperatures and no agitation (to prevent activation of pancreatic zymogens), and DNase^30,31,32 supplementation (to prevent DNA-induced apoptosis and improve cell viability, and an antibody for CD31³³ (PECAM1, a pan-EC marker). The described protocol produces EC populations isolated from mouse pancreas that can be used for gene expression profiling and protein assays.

Protocol

Tissue isolation was performed under the approved study protocol #17010 by the Institutional Animal Care and Use Committee (IACUC) of Beckman Research Institute, City of Hope (Duarte, California, USA). Here, we use Tie-2CreERT2;Rosa26-TdTomato mouse line in C57BL/6 background at 8 - 12 weeks of age. In this line, ECs are labeled with TdTomato when induced with tamoxifen as previously described³⁴. However, this protocol can be adapted for all ages of adult mice with different genotypes and genetic backgrounds.

1. Tissue collection (estimated time: 1-2 h)

NOTE: The time estimate is 15 min/animal, recommended maximum 3 animals pooled per dissociation samples.

1. Euthanize animals by carbon dioxide (CO₂) overdose, followed by secondary confirmation of death. Before dissection, spray the whole body with 70% ethanol (EtOH), completely disinfecting the operating area.
2. Stretch out and secure limbs using pins. Using surgical scissors, make a small midline incision through the skin and peritoneum starting from the lower abdominal area and extending toward the thoracic region.
3. Expose the ribcage by cutting through the diaphragm and lift the ribcage to expose the heart.
4. Insert a needle (25G-30G) connected to a syringe containing ice-cold sterile PBS, into the left ventricle of the heart.
5. Start the perfusion by injecting the PBS through the heart at a rate of 5 - 10 mL/min. Stop after injecting 10 mL, or until the liver and kidneys become discolored.
6. After perfusion, locate the pancreas (below the stomach and attached to the duodenum)³⁵ (Figure 3A) and carefully remove it using dissecting scissors and forceps.
7. Once isolated, transfer pancreas to a 50 mL tube containing 10 mL of ice-cold PBS + 0.2 mg/mL trypsin inhibitor, keep on ice (Figure 3B).
   NOTE: Pancreas from a maximum of 3 mice can be pooled at this step and transferred into 1 tube. Samples can be kept on ice for a maximum time of 2 h.

2. Collagenase preparation (estimated time: 15-30 min)

1. Prepare solutions and buffers as described below.
   1. Dissociation solution: Add collagenase Type I into 50 mL tube to a final concentration of 1 mg/mL in Dulbecco's phosphate buffered saline (DPBS 1x (Ca²⁺, Mg²⁺; see Table of Materials) and 0.001 U/mL DNase I + 0.2 mg/mL Trypsin inhibitor.
   2. Dissociation stop solution: Add phosphate buffered saline (PBS 1x, without Ca²⁺, Mg²⁺; same for all the rest PBS in this protocol; see Table of Materials), 5% Bovine Serum Albumin (BSA), 0.001 U/mL DNase I or DMEM + 10% FBS + 0.001 MU/mL DNase I.
   3. Wash buffer: Add PBS 1x, 0.5% Bovine Serum Albumin (BSA), 0.001 U/mL DNase I + 0.05 mg/mL Trypsin Inhibitor.
2. Dissolve buffers completely and filter through a 0.22 µm filter. Set aside and keep on ice.

3. Digestion (estimated time: 40 min-1 h)

1. Remove pancreas from ice-cold PBS and place on Petri dish on top of ice. Carefully remove excess tissue (e.g., spleen, fat, debris) with surgical scissors and tweezers (Figure 3C).
   NOTE: It is crucial to remove fat, as excess of lipids can interfere with tissue digestion.
2. Transfer the trimmed mouse pancreas to a 5 mL tube containing 1 mL of dissociation solution.
3. Keep tube on ice and mince the pancreatic tissues with dissecting scissors into fine pipettable pieces, e.g. 0.5 mm or 1 mm. Transfer lysate into a 50 mL tube and keep on ice.
4. Add additional pancreas to the 5 mL tube used to initially mince the pancreatic tissue and add additional 1 mL of collagenase. Repeat as necessary for multiple pancreas preparations.
5. Add 2 mL of collagenase solution to the 5 mL tube to remove any residual pancreas tissue and transfer to 50 mL tube. Total collagenase volume should be 5 mL for 3 pancreas or 3 mL for a single pancreas.
6. Once all samples have undergone mechanical dissociation, transfer the 50 mL tubes containing the homogenized tissue lysates to a 37 °C water bath and incubate for 10 min.
   NOTE: Do not vigorously agitate or vortex the tissue mixture during this process as this can cause rupture of acinar tissues and the release of abundant endogenous enzymes and DNA.
7. Remove the tubes from water bath and gently swirl to mix well, and leave in water bath for another 10 min. (Total time: 20 min).
   NOTE: Do not exceed 30 min in 37 °C water bath to avoid over-digestion.
8. Place tubes on ice, and let the homogenate settle by gravity. Once settled, transfer supernatant through a 70 µm filter, and add 5 mL of dissociation stop solution.
9. Add an additional 1 mL of dissociation solution to the remaining pellet and triturate with a 18G needle.
10. Pass the remaining homogenate through the filter and wash with an additional 5 mL of dissociation stop solution.
11. Spin cell suspension at 300 x g for 10 min, at 4 °C. Remove supernatant completely by aspiration and keep cell pellet on ice.
12. Resuspend cell pellet with 1 mL of wash buffer and transfer to a 1.5 mL microcentrifuge tube.
13. Take a small aliquot from the resuspended cell pellet (10 µL) and mix with trypan blue at a 1:1 ratio to count cells using a cell counter and measure viability.

4. Endothelial cell (EC) enrichment (estimated time: 2-3 h)

1. Add 1 µL of anti-CD31-biotin antibody for every 1 x 10⁷ cells counted and incubate for 30 min at 4 °C on a rotator (about 5 µL per 3 pancreas).
2. Add 20 µL of anti-biotin microbeads and incubate for 40 min at 4 °C on a tube rotator.
3. Set up a magnetic stand with the column separator holder and apply the column. Equilibrate column with 3 mL of wash buffer.
4. Add sample to the column and wash with a total of 9 mL wash buffer.
   NOTE: If cell concentration is high, dilute cells with an additional 2 mL of wash buffer in a separate tube. Add 1 mL of sample at a time into the column to prevent the column from clogging.
5. Remove column from column holder and place in a 15 mL tube. Add 5 mL of wash buffer, to the column. Use plunger to push down cells in the column and collect elution (which contains enriched EC).
6. Spin down flow-through and elution/EC fraction at 300 x g for 10 min at 4 °C. Count cells and measure viability as in step 3.13.
7. Validate EC-enrichment via qPCR of flow-through and EC-fractions. Use NK6 Homeobox 1 (Nkx6.1) as a general marker for endocrine cells and platelet and endothelial cell adhesion molecule (Pecam1) as a marker for EC³³. Use 36B4 (Rplp0, acidic ribosomal phosphoprotein p0) as an internal control for gene expression normalization.
   NOTE: Collected samples can be used for protein or RNA analyses at this step.

5. Culturing pECs

1. Prepare M199 medium. Supplement M199 medium to a final concentration of 20% fetal bovine serum (FBS) and 0.1% Penicillin-Streptomycin.
2. Coat a 60 mm cell culture plate with collagen Type I (1 mg/mL in 0.1 M acetic acid) and allow to sit for 30 min at room temperature.
3. Remove collagen Type I and rinse the plate with 1X PBS, twice. Add 2 mL of prepared M199 Media and place cell culture plate in standard cell culture incubator (5% CO₂, 37 °C) for 5 min.
4. Once cells have been isolated, add isolated cells to pre-warmed M199 cell culture plate. Keep cells in culture conditions for approximately 14 days and change media every 2-3 days.

Representative Results

Following this protocol, approximately 2 x 10⁶ live cells can be obtained when pooling 3 mouse pancreases, and 750,000 cells from a single mouse pancreas. To validate the enrichment of EC, we performed the following analyses: 1) quantitative PCR: compared to the flow-through (FT) samples (i.e., the non-CD31 antibody-bound fractions), the EC fractions had significantly higher levels of Pecam1 (encoding CD31) and Kdr (encoding VEGFR2), two EC marker genes³³, and lower levels of NK6 homeobox 1 (Nkx6.1), a transcription factor in the pancreatic endocrine cells (Figure 4A); 2) western blotting: EC-enriched samples from three preparations (pEC-1-3) showed strong signals of CD31 at the expected size (~130 kDa). In comparison, the input and FT samples showed minimal CD31 signal. As a positive control, lysates of MS1 cells, a pEC cell line, also showed strong CD31 signal (Figure 4B); 3) trypan blue staining: we found the viability of isolated cells was around 50% (Figure 4C); 4) flow cytometry following a published protocol¹⁵ : through gating on the live cell population (Figure 4D), we found the CD31+ cells were at 88% and the TdTomato+ cells were at 76%. Of note, the majority of isolated cells co-express CD31 and TdTomato (Figure 4E). As negative controls, the unstained MS1 cells and MS1 incubated with IgG showed no live cells as marked by CD31. As the positive control, i.e., MS1 cells stained for CD31 showed close to 100% live CD31+ cells (Figure 4F). We also cultured the isolated cells, performed immunofluorescence staining of PECAM1 following a published protocol³⁶, and imaged TdTomato, which showed robust signals of both markers in the isolated cells, suggesting the presence of ECs (Figure 4G).

Figure 1: Illustration of vascularization of pancreas and pancreatic islets. A pictorial depiction of the pancreatic endothelial cells (ECs) and intra-islet ECs. Please click here to view a larger version of this figure.

Figure 2: Overview of workflow. 1) Pancreas dissection, 2) mechanical digestion, 3) enzyme incubation, 4) centrifugation, 5a) supernatant removal and 5b) secondary Collagenase digestion of the cell pellets with trituration for filtering in 6), 7), and 8) pellet reconstitution and antibody binding, 9) microbeads incubation, 10) MACS column enrichment, 11) cell counting for viability, and 12) downstream analysis. Please click here to view a larger version of this figure.

Figure 3: Tissue processing steps. Images of (A) the abdominal cavity, (B) the isolated mouse pancreas, and (C) the trimmed mouse pancreas. Please click here to view a larger version of this figure.

Figure 4: Validation of isolated EC fractions from mouse pancreas. (A) qPCR of FT samples and enriched EC fraction with CD31 for enrichment. Pecam1 (CD31), Kdr (VEGFR2), and Nkx6.1 (pancreatic β-cell specific transcription factor)³³. * denotes p <0.05 compared to FT based on t tests. n = 3-4 pooled samples, consisting of 3 mice each. (B) Western blotting of CD31. pEC-1-3 indicate enriched ECs from 3 separate pooled mouse samples. Cell lysates of 30 mg was loaded onto each lane, followed by Ponceau S staining. (C) Cell counting and trypan blue staining of enriched pECs for viability assay. (D) Flow cytometry gating strategy for mouse pECs stained by 4', 6-diamidino-2-phenylindole (DAPI) and CD31 antibody. (E) Flow cytometry histograms of unstained MS1 (top left), MS1 stained with AlexaFluor-488 secondary antibody (IgG control; top right), MS1 stained with CD31 (bottom left), and the isolated mouse pECs stained with CD31 (bottom right). (F) Percentage of live cells quantified using flow cytometry. (G) Immunofluorescence of CD31 expression (green) with DAPI (blue) counterstain and endogenous TdTomato fluorescent reporter (red). Scale bar = 100 µm. Please click here to view a larger version of this figure.

Figure 5: CD31 and CD144 specificity in mouse pancreas. (A) Immunofluorescence of mouse pancreas with CD31 (Magenta) and CD144 (Cyan), with DAPI counterstain. Image captured using a 10x lens under a widefield fluorescent inverted microscope. Scale bar = 100 µm. Dashed lines demarcate islet area. (B) Uniform Manifold Approximation and Projection (UMAP) showing Cdh5 and Pecam1 expression in the islet ECs from C57BL6 mice detected by scRNA-seq (GSE128565)³⁷. Please click here to view a larger version of this figure.

Enzyme	Use	Temperature and Time	Reference
Collagenase I	Single cell dissociation, whole pancreas EC enrichment	37 °C, 20 min	This protocol
Collagenase IV	Islet Isolation and and non-enzymatic dissociation	37 °C, 20 min	46
Collagenase + Trypsin	Examining ductal cells in mouse pancreas	37 °C, 30 min	47
Collagenase B	Single cell suspension of pancreas	37 °C, 30 min	48
Collagenase NBI	Islet Isolation	37 °C, 30 min	49
Collagenase P	Acinar viable suspension	37 °C, 15 min	45
Collagenase XI	Islet Isolation	37 °C, 15 min	35
Dispase	Single cell suspension	37 °C, 45 min	24
Liberase	Single cell suspension	37 °C, 10 min	25, 48
Trypsin	Single cell dissociation, whole pancreas	4 °C, O/N	28

Table 1: Summary of various protocols used to process mouse pancreas samples.

Discussion

In this article, we present a protocol for enrichment and isolation of the pECs. Similar to previous EC isolation protocols from other tissues or organs, this protocol consists of three major processes, namely, physical dissociation, enzymatic digestion, and antibody-based EC enrichment. To address the unique challenges in processing the pancreas, we introduced several key adaptations and critical steps within our protocol: 1) a gentle one-step collagenase digestion with a short incubation time, 2) supplementation of higher concentration of DNase and a trypsin inhibitor to minimize DNA contamination and endogenous enzyme digestion, which can interfere with antibody binding and thus diminish EC enrichment and viability and 3) cell counting after the digestion step to adjust how much antibody is added to the sample to optimize EC enrichment. In the process of optimizing this protocol, we tested other enzymes used in previous studies, e.g., trypsin²⁸, collagenase Type IV³¹, and other digestion temperatures, e.g. 4°C, as well as different incubation times (Table 1). We found collagenase Type I, 20 min, at 37 °C to be the optimal condition for the mouse pEC isolation and used this here. While the addition of an anticoagulant during the perfusion may improve overall cell viability, typical anticoagulating agent, e.g., EDTA or Heparin may interfere with the activity of certain enzymes, e.g., collagenase for cell dissociation and reverse transcriptase used in qPCR. To prevent the unexpected/undesirable effect of these anticoagulants, PBS (without Ca²⁺ and Mg²⁺) for perfusion is used.

We focused on developing a protocol for enriching EC population from whole mouse pancreases, which would include both intra-islet ECs and ECs in the exocrine tissues, resulting in a higher yield. In the current protocol, we pooled together 3 animals during the dissociation step and on average achieved a single-cell suspension with approximately 2.0 x 10⁶ viable cells. We have also tested samples from single mouse pancreas and about 750,000 cells can be obtained. It is worth noting that when adjusting to process a single animal, less dissociation solution (3 mL dissociation per pancreas) should be used. A crucial step in the protocol is to completely trim and remove any residual fat that is attached to the pancreas (Figure 3B). Adipose tissue contamination in the collagenase digestion will interfere with the pancreatic single cell dissociation step diminishing cell viability in the subsequent enrichment steps³⁸. Moreover, to enrich pECs we used a pan-EC surface marker, PECAM1/CD31. Using PECAM1/CD31 antibody, we were able to significantly enrich EC fractions, evident by the gene expression analysis. In the development of this protocol, we had also tested the utility of CD144 antibody, which has also been commonly used to isolate ECs¹³. However, we were not able to enrich pECs (data not shown). Intriguingly, immunofluorescent staining revealed that whereas CD31 is expressed in all ECs in the pancreas tissue, CD144 antibody appears to stain only intra-islet ECs (Figure 5A). Additionally, we analyzed publicly available single cell RNA-seq data from islets isolated from C57BL/6 mice (GSE128565)³⁷ and found that approximately 50% of the islet ECs show expression of both Cdh5 (VE-Cad) and Pecam1, but the rest of islet ECs express only Cdh5 or Pecam1 (Figure 5B). Thus, we speculate that CD31 antibodies is more appropriate for pEC isolation and the utility of CD144 antibody for this purpose warrants future investigation. Moreover, there are commercially available anti-CD31 conjugated with magnetic beads which can simplify the two-step antibody binding step to a singular step.

As demonstrated, we performed qPCR, immunoblotting, flow cytometry, and immunofluorescent staining to validate the isolated EC fraction. It is conceivable that these fractions can be subjected to other assays, such as (sc)RNA-seq and mass spectrometry. A limitation of this protocol is that we have not been able to achieve robust subculture and perform functional assays. Given that subculture of isolated mouse ECs has been shown to be challenging, this is an area warranting future efforts. One way to address this may be the introduction of large T antigen, which has been used to immortalize primary cells³⁹.

The recent advances and application of scRNA sequencing have enabled high-resolution profiling of pancreatic cells⁴⁰ and especially islet cells⁴¹. However, most of the existing studies have focused on endocrine cell populations and used tissue dissociation protocols that does not efficiently retain ECs. As a result, EC fractions often end up scarce, constituting only 3%-5% in the entire captured cell populations^42,43,44. A recent protocol uses two-step enzymatic digestion and deactivation and a fluorescently activated cell sorting step and retrieves ECs at 12% of the entire pancreatic cell population⁴⁵. Complementary to this protocol, our protocol provides a simplified workflow, which may be more cost-effective and less technically demanding should ECs be the focus of the study. Although this protocol has only been tested in mouse pancreas, it can be scaled up and adapted to isolate ECs from other organisms, including human pancreas.

Materials

Name	Company	Catalog Number	Comments
1.5 mL eppendorf	USA Scientific	1615-5500
10 cm dish	Genesee Scientific	25-202
25G needles	BD	305145
2X Taq Pro Universal SYBR qPCR Master Mix	Vazyme	Q712-03-AA
5 mL eppendorf	Thermo Fisher	14282300
6-well plate	Greiner Bio-One	07-000-208
70 µm strainer	Fisher	22-363-548
Anti-CD31-biotin	Miltenyi Biotech	REA784
Bovine serum albumin heat shock treated	Fisher	BP1600-100
CaCl2	Fisher	BP510
Centrifuge	Eppendorf
Collagen Type 1, from calf skin	Sigma Aldrich	C9791	Attachment reagent in the protocol
Collagenase Type 1	Worthington Bio	LS004197
Countess Automatic Cell Counter	Thermo Fisher
DAPI	Thermo Fisher	D1306	immunofluorescence
Disposable Safety Scalpels	Myco Instrumentation	6008TR-10
DNAse I	Roche	260913
D-PBS (Ca2+,Mg2+)	Thermo Fisher	14080055
Ethanol	Fisher	BP2818-4
Fetal bovine serum	Fisher	10437028
Incubator			Kept at 37 °C 5% CO2
LS Columns	Miltenyi Biotech	130-042-401
M199	Sigma	M2520-1L
MACS MultiStand with the QuadroMACS Separator	Miltenyi Biotech	130-042-303
Medium 199	Sigma Aldrich	M2520-10X
Microbeads anti-biotin	Miltenyi Biotech	130-090-485
Microscope	Leica		To assess cell morphology
Molecular Grade Water	Corning	46-000-CM
NaCl	Fisher	S271-1
New Brunswick Innova 44/44R Orbital shaker	Eppendorf
PECAM1 (CD31) Antibody	Abcam	ab56299	immunofluorescence
PECAM1 (CD31) Antibody	R&D Systems	AF3628
Phosphate Buffered Saline (10X) (no Ca2+,no Mg2+)	Genesee Scientific	25-507-XB
Primer 36B4 Forward mouse	IDT		AGATTCGGGATATGCTGTTGGC
Primer 36B4 Revese mouse	IDT		TCGGGTCCTAGACCAGTGTTC
Primer Kdr Forward mouse	IDT		TCCAGAATCCTCTTCCATGC
Primer Kdr Reverse mouse	IDT		AAACCTCCTGCAAGCAAATG
Primer Nkx6.1 Reverse mouse	IDT		CACGGCGGACTCTGCATCACTC
Primer Nxk6.1 Forward mouse	IDT		CTCTACTTTAGCCCCAGCG
Primer PECAM1 Forward mouse	IDT		ACGCTGGTGCTCTATGCAAG
Primer PECAM1 Reverse mouse	IDT		TCAGTTGCTGCCCATTCATCA
RNase ZAP	Thermo Fisher	AM9780
RNase-free water	Takara	RR036B
Sterile 12" long forceps	F.S.T	91100-16
Sterile fine forceps	F.S.T	11050-10
Sterile fine scissors	F.S.T	14061-11
Tissue Culture Dishes 2cm	Genesee Scientific	25-260
TRIzol reagent	Fisher	15596018
Trypan Blue	Corning	MT25900CI
Trypsin Inhibitor	Roche	10109886001
Tween-20
VE-Cadherin Antibody	Abcam	ab33168	immunofluorescence
Waterbath