A cell culture model of resistance arteries is described, allowing for the dissection of signaling pathways in endothelium, smooth muscle, or between endothelium and smooth muscle (the myoendothelial junction). The selective application of agonists or protein isolation, electron microscopy, or immunofluorescence can be utilized using this cell culture model.
The myoendothelial junction (MEJ), a unique signaling microdomain in small diameter resistance arteries, exhibits localization of specific proteins and signaling processes that can control vascular tone and blood pressure. As it is a projection from either the endothelial or smooth muscle cell, and due to its small size (on average, an area of ~1 µm2), the MEJ is difficult to study in isolation. However, we have developed a cell culture model called the vascular cell co-culture (VCCC) that allows for in vitro MEJ formation, endothelial cell polarization, and dissection of signaling proteins and processes in the vascular wall of resistance arteries. The VCCC has a multitude of applications and can be adapted to suit different cell types. The model consists of two cell types grown on opposite sides of a filter with 0.4 µm pores in which the in vitro MEJs can form. Here we describe how to create the VCCC via plating of cells and isolation of endothelial, MEJ, and smooth muscle fractions, which can then be used for protein isolation or activity assays. The filter with intact cell layers can be fixed, embedded, and sectioned for immunofluorescent analysis. Importantly, many of the discoveries from this model have been confirmed using intact resistance arteries, underscoring its physiological relevance.
In small diameter resistance arteries, a thin internal elastic lamina (IEL) separates endothelial (EC) and smooth muscle cells (SMC). The cells can project through holes in this elastic matrix and make direct cytoplasmic connections via gap junction channels1,2. This unique structure is termed the myoendothelial junction (MEJ). The MEJ is a small (approximately 0.5 µm x 0.5 µm, depending on the vascular bed), cellular projection composed predominantly of endothelial cells but may originate from smooth muscle as well3,4. A number of investigators have demonstrated the immense complexity of signaling networks that occur selectively at the MEJ, rendering it an especially important location for facilitating bi-directional signaling between endothelium and smooth muscle5,6,7,8,9,10.
However, a mechanistic dissection of signaling pathways at the MEJ is difficult in an intact artery. Because the MEJ is a cellular projection, it is not currently possible to isolate an in vivo MEJ from the vascular wall. For this reason, the VCCC model1 was developed. Importantly, the VCCC replicates physiological endothelial morphology11 and polarization of signaling between the apical and MEJ portions of the cell12. It was this unique model that facilitated the discovery that alpha hemoglobin is in the endothelial cell, polarized to the MEJ. This is in contrast to conventionally cultured endothelial cells which do not express alpha hemoglobin13. For researchers interested in microvascular endothelium, it may be more appropriate to use endothelial cells that have been cultured in the VCCC, particularly if the intent is to dissect signaling pathways that occur in small diameter resistance arteries.
Using a sturdy plastic insert containing a filter with small diameter pores (0.4 µm in diameter) to co-culture two distinct cell types prevents the cells from migrating between layers. It results in a 10 µm distance between the cells, which is significantly longer than in vivo, but still replicates many of the in vivo characteristics of MEJs, including protein localization and second messenger signaling1,14. In addition, the VCCC allows for targeting of cell type-specific signaling via the addition of an agonist or antagonist to the specific cellular compartment. For example, loading the EC with BAPTA-AM to chelate calcium, and stimulating the SMC with phenylephrine14. In contrast to other descriptions of co-culture models15,16,17,18,19,20,21, this provides instructions on isolating the distinct fractions for mRNA and protein, including the distinct MEJ fraction within the filter pores. The addition of this technique to the VCCC allows for specific investigation of changes in mRNA localization or transcription22, protein phosphorylation12,23, and protein activity12. This article will describe plating of EC on top of the filter and SMC on the bottom, although it is possible to culture the two cell types in different conformations11,18,24.
1. Plating Cells for the VCCC
2. Harvesting Fractions from the VCCC
3. Harvesting the VCCC for En Face Immunofluorescence
4. Harvesting the VCCC for Transverse Immunofluorescence
5. Embedding the VCCC for Transverse Immunofluorescence
6. Sectioning and Staining the VCCC for Transverse Immunofluorescence
The filter inserts used for the VCCC have small, 0.4 µm holes. Figure 1A, an en face view of the VCCC filter inset, shows the pores in which the MEJ can form in vitro. The schematic depicts the smooth muscle cells plated on the lower side of the filter one day before the endothelial cells are plated on the opposite side (Figure 1B). Once the cells are plated, the EC, MEJ, and SMC fractions can be isolated three days later. An example of this is shown in Figure 2A, where plasminogen activator inhibitor 1 (PAI-1) is highly expressed in the MEJ fraction relative to the rest of the EC, which is also seen at in vivo MEJs25. These data demonstrate the utility of the VCCC in recapitulating the intact arteriolar blood vessel wall.
Next the VCCC in immunofluorescence and electron microscopy is demonstrated. The VCCC is left intact and fixed with PFA for transverse immunofluorescence. Figure 3A shows the smooth muscle specific alpha-1 adrenergic receptor and the endothelial specific bradykinin receptor. Phalloidin staining of F-actin shows the actin bridges in the MEJ (white arrow) between the two cells, which replicates the expression seen in an intact artery26. In Figure 3B, a cross-section of the VCCC as viewed by transmission electron microscopy is demonstrated, which is rendered in 3D in Figure 3C. These views of the VCCC demonstrate the ability to specifically localize proteins, receptors, and ultrastructure to in vitro MEJs.
Figure 1: Plating of the Vascular Cell Co-Culture (VCCC). (A) En face light micrograph of the pores in the insert filter. Black arrows denote individual holes. (B) Schematic of plating the VCCC. Day 1 shows inversion of the filter insert and plating SMC on the bottom. On day 2, the EC are plated on the opposite side of the filter and the VCCC remains in this conformation in a 6-well plate until harvest. White arrows indicate the side of the filter where the cells are plated and will grow. Scale bar = 1 mm. Please click here to view a larger version of this figure.
Figure 2: Protein enrichment in the MEJ. (A) Immuno-transmission electron microscope image of mouse artery with expression of alpha globin at the MEJ (gold beads which appear as black dots). (B) Western blot showing plasminogen activator inhibitor 1 (PAI-1) expression is enriched in the MEJ fraction versus the EC or SMC fractions. Plastic EC indicates EC grown on a plastic dish, which do not form MEJ. Scale bar = 1 µm. Please click here to view a larger version of this figure.
Figure 3: Transverse immunofluorescence and electron microscopy in transverse sections of the VCCC. (A) Staining for smooth muscle and endothelial specific proteins. F-actin staining is throughout both cell types and the in vitro MEJ (white arrow). An example of electron microscopy used to study ultrastructure of the in vitro MEJs both in 2D (B) and 3D (C). Scale bar = 10 µm (A); 2.5 µm (B); and approximately 2.5 µm vertical and 0.5 µm horizontal (C). Please click here to view a larger version of this figure.
The VCCC has a number of advantages in terms of recapitulating MEJs in vitro, but there are points of discussion when determining if the VCCC can be utilized for specific application. For example, if different cell types are to be used with this model, it will need to be optimized. We have used primary human cells but it is possible to isolate cells from bovine aortas24, or wildtype or knockout mice, and use them in the VCCC1,5,14.
If using the VCCC for MEJ isolation, the most important technique to master is thorough scraping of the filter to ensure that the purity of the MEJ fraction is not altered by excess endothelial or smooth muscle cells. We have shown via Western blotting that alpha hemoglobin is the best marker of a "pure" MEJ isolation13, as it should be strongly expressed in the MEJ fraction versus the endothelial cell fraction. Another is PAI-1, which regulates the formation of MEJ (Figure 2)25. We feel these may be especially good markers for a robust MEJ isolation. For extra confidence in the harvest technique, the filters may be kept after scraping for harvest, and stained for EC/SMC markers25.
Imaging protein expression in the VCCC can be performed in two main ways: en face (Figure 1A) or transverse (Figure 3). Both preparations allow for the visualization of proteins within the holes of the filter from two different perspectives. The en face technique involves arteries that are cut open longitudinally and then pinned out flat, with the EC facing up, to visualize both the EC monolayer and the holes in the IEL, the only place where MEJ can form. A fluorescent signal within the holes of the IEL indicates localization of the protein within the MEJ10,27. This same principle can be translated to the VCCC by imaging the EC monolayer from above, without needing to embed it in paraffin or make sections. The transverse method is more complex to master but is critical for distinct visualization of proteins in the EC versus the SMC monolayers (Figure 3).
The lack of flow or stretch in the system is a potential limitation, but it is possible to add flow in an endothelial-smooth muscle cell co-culture16,19. It appears that the static conditions still allow for physiologically accurate protein expression and activation, so it may be that the heterocellular contact is the most important factor in dissection of MEJ signaling pathways.
There are multiple applications of this technique beyond resistance artery signaling as the VCCC is not necessarily limited to specific cell types. Other co-culture models have utilized outgrowth endothelial cells and osteoblasts17, or modeled the blood brain barrier with endothelial and pericyte/astrocyte co-cultures21. Metastasis of tumor cells through the endothelium can also be quantified using this model21. It is also possible to grow cells on the filter and culture another cell type in the bottom of the 6-well dish to test paracrine signaling, or grow endothelial cells on the upper part of the membrane to look at polarization of the cell (our unpublished observations). Leukocytes or other circulating cells can be added to the EC media and the adhesion of cells can be quantified12. In addition, the VCCC can be used to investigate pathologies such as smooth muscle migration24 and proliferation in response to endothelial injury15,18.
In conclusion, we have presented a method to isolate MEJ from an in vitro cell culture system, which allows for the investigation of signaling processes between two coupled cell types. Importantly, the co-culture system is not limited to the study of the MEJ and can be valuable to answer a number of questions, especially those involving heterocellular communication.
The authors have nothing to disclose.
This work was supported by National Heart, Lung and Blood Institute grants R01088554, T32HL007284 and American Heart Association predoctoral fellowship 14PRE20420024. We especially thank the St. Jude Cellular Imaging Shared Research Core, including Randall Wakefield and Linda Horner for the electron microscopy images, Adam Straub for assistance with representative immunofluorescence images, and Pooneh Bagher for her valuable feedback on the manuscript protocol.
24mm diameter, 0.4µm Transwell Polyester membrane filter insert | Corning | 3450 | This is one 6 well plate of inserts |
225cm2 flask | Corning | 431082 | |
6 well plates (without filter inserts) | Corning | 3506 | |
Primary human coronary artery endothelial cells | Lonza | CC-2585 | |
Primary human coronary artery smooth muscle cells | Lonza | CC-2583 | |
EBM-2 EC Basal media | Lonza | CC-3156 | |
EC growth factors (EGM2 MV Single quots) | Lonza | CC-4147 | Add to basal media before use |
SmBM SMC basal media | Lonza | CC-3181 | |
SMC growth factors (SmGM-2 SingleQuot) | Lonza | CC-4149 | Add to basal media before use |
0.5% Trypsin-EDTA 10X | Gibco | 15400-054 | Dilute to 2X with sterile dPBS |
Hemocytometer | VWR | 15170-208 | |
large petri dishes for SMC Plating (150mm) | Corning | 353025 | |
Bovine gelatin | Sigma | G-9382 | |
Human fibronectin | Corning | 356008 | 5µg/ml of fibronectin and store at -20 |
10X Hanks' Buffered salt solution (HBSS) | Gibco | 14065-056 | Dilute to 1x with sterile water |
10X Dulbecco's Phosphate buffered saline (dPBS) | Gibco | 14200-075 | Dilute to 1x with sterile water |
Disposable Curved Scalpel (30mm cutting edge) with handle | Amazon | MDS15210 | |
Forceps | Fisher | 17-467-201 | |
Cell lifter | Corning | 3008 | |
Cell scraper | BD Falcon | 353085 | |
50mL conical tube | Corning | 352070 | |
10mm petri dishes | Falcon | 35-1008 | |
Rneasy mini kit | Qiagen | 74104 | |
RNA lysis reagent | Qiagen | 79306 | |
Paraformaldehyde | Sigma | 158127 | Make 4% solution with dPBS |
70% ethanol | Fisher | 04-355-305 | |
Cavicide disinfectant | Fisher | 131000 | |
RIPA protein lysis buffer | Sigma | R0278 | |
Sonic dismembranator | Fisher | Model 50 | |
1X PBS for harvesting EC/SMC and immunocytochemistry | Gibco | 10010023 | does not need to be sterile |
Name | Company | Catalog Number | Comments |
Embedding/Sectioning/Staining | |||
Paraffin | Fisher | P31-500 | |
Automated tissue processer | Excelsior | ES | |
Embedding station + cold plate | Leica | HistoCore Arcadia | |
Tissue Tek Accu-Edge Microtome blade | VWR | 25608-961 or 25608-964 | |
Microscope slides | Thermo Fisher | 22-230-900 | |
Coverslips | Thermo Fisher | 12-548-5E | |
Prolong Gold mounting medium | Thermo Fisher | P36931 | |
Name | Company | Catalog Number | Comments |
Other Solutions | **Perform all steps in sterile conditions if using the solution on live cells | ||
Sterile dPBS | Autoclave water to sterilize. Using the sterile water make DPBS (450mL sterile water: 50mL 10X DPBS), mix and filter sterilize (0.2µm filter). Final concentration: 1X |
||
sterile HBSS | Autoclave water to sterilize. Using the sterile water make HBSS (450mL sterile water: 50mL 10X HBSS), mix and filter sterilize (0.2µm filter). Final concentration: 1X |
||
Fibronectin stock solution | Use 50mL of sterile water to fibronectin to make stock of 1mL aliquots, then freeze until needed Final concentration: 100µg/mL |
||
Fibronectin to coat SMC side of VCCC | thaw one aliquot and add 1mL of fibronectin (100µg/mL) to 19mLs of 1X HBSS to reach a final concentration of 5µg/mL. Final concentration: 5µg/mL |
||
2x Trypsin EDTA solution | Dilute the 10X 0.5% Trypsin-EDTA to 2X with 1X dPBS (40mL DPBS: 10mL 10X Trypsin) . Final concentration: 2X |
||
Blocking/antibody solution | (9mL 1X PBS, 25mL Triton X100, 50mg bovine serum albumin, 500mL normal serum) | ||
Bovine gelatin to coat EC VCCC | Mix 0.5g bovine gelatin into 100mL distilled water. Autoclave prior to first use and store at room temperature. Keep in sterile conditions. Final concentration: 0.50% |
||
1X PBS for harvesting cells | does not need to be sterile |