Isolation and Profiling of Human Primary Mesenteric Arterial Endothelial Cells at the Transcriptome Level

Summary

The protocol describes the isolation, culture, and profiling of endothelial cells from human mesenteric artery. Additionally, a method is provided to prepare human artery for spatial transcriptomics. Proteomics, transcriptomics, and functional assays can be performed on isolated cells. This protocol can be repurposed for any medium- or large-size artery.

Abstract

Endothelial cells (ECs) are crucial for vascular and whole-body function through their dynamic response to environmental cues. Elucidating the transcriptome and epigenome of ECs is paramount to understanding their roles in development, health, and disease, but is limited in the availability of isolated primary cells. Recent technologies have enabled the high-throughput profiling of EC transcriptome and epigenome, leading to the identification of previously unknown EC cell subpopulations and developmental trajectories. While EC cultures are a useful tool in the exploration of EC function and dysfunction, the culture conditions and multiple passages can introduce external variables that alter the properties of native EC, including morphology, epigenetic state, and gene expression program. To overcome this limitation, the present paper demonstrates a method of isolating human primary ECs from donor mesenteric arteries aiming to capture their native state. ECs in the intimal layer are dissociated mechanically and biochemically with the use of particular enzymes. The resultant cells can be directly used for bulk RNA or single-cell RNA-sequencing or plated for culture. In addition, a workflow is described for the preparation of human arterial tissue for spatial transcriptomics, specifically for a commercially available platform, although this method is also suitable for other spatial transcriptome profiling techniques. This methodology can be applied to different vessels collected from a variety of donors in health or disease states to gain insights into EC transcriptional and epigenetic regulation, a pivotal aspect of endothelial cell biology.

Introduction

Lining the lumen of blood vessels, endothelial cells (ECs) are crucial regulators of vascular tone and tissue perfusion. ECs are remarkable in their ability to react to the extracellular environment and adapt to changes in the dynamics and composition of blood flow. These dynamic responses are mediated through a network of intracellular signaling events, including transcriptional and post-transcriptional modulations with spatio-temporal resolution. The dysregulation of these responses is implicated in many pathologies, including but not limited to cardiovascular disease, diabetes, and cancer^1,2.

A large proportion of studies make use of cell lines or animal models to interrogate EC transcriptome. The former is a useful tool, given the relative ease of use and inexpensiveness. However, serial culturing can introduce phenotypic alterations to ECs, such as fibroblastic features and a lack of polarization, disconnecting them from their in vivo state³. The primary cells, e.g., human umbilical vein EC (HUVEC) have been a popular choice since the 1980s but are derived from a developmental vascular bed that does not exist in adults, thus unlikely to fully represent mature ECs. Animals, especially mouse models, better represent the physiological or pathophysiological environment of ECs and allow interrogation of transcriptomes as a result of genetic perturbation. Murine ECs can be isolated from various tissues, including the aorta, lungs, and adipose tissues using enzyme-based procedures^4,5,6,7. However, the isolated cells cannot be used for multiple passages unless transformed⁶ and are often limited in numbers, which requires pooling from multiple animals^5,8,9.

The advent of new technologies exploring the vessel architecture at a transcriptomic level, particularly with single-cell resolution, has enabled a new era of endothelial biology by revealing novel functions and properties of ECs^{5, 10,11,12,13,14}. A rich resource built by Tabula Muris investigators collected single-cell transcriptomic profiles of 100,000 cells including ECs across 20 different murine organs¹⁵, which revealed both common EC marker genes and unique transcriptomic signatures with inter-and intra-tissue differences^5,13. Nevertheless, there are clear differences between mouse and human in genome, epigenome, and transcriptome, especially in the non-coding regions^16,17,18. These aforementioned drawbacks emphasize the importance of analysis of ECs using human samples in order to gain a faithful profile of ECs in their native state in health and disease.

Most EC isolation methods rely on physical dissociation through homogenization, finely cutting, and mincing the tissue before incubation with proteolytic enzymes for differing times. The enzymes and conditions also vary considerably between tissue types, from trypsin to collagenase, used alone or in combination^19,20,21. Further antibody-based enrichment or purification are often included to increase the purity of ECs. Typically, antibodies against EC membrane markers, e.g., CD144 and CD31 are conjugated to magnetic beads and added to the cell suspension^22,23. Such a strategy can be generally adapted for EC isolation from multiple human and mouse tissues, including the techniques introduced in this protocol.

In their native state, ECs interact with multiple cell types and may exist in vascular niches where cell proximity is crucial for function. While single-cell and single-nuclear RNA-sequencing (scRNA and snRNA-seq) studies have been paramount to the recent breakthroughs in describing EC heterogeneity, the dissociation process disrupts tissue context and cell-cell contact, which are also important to understand EC biology. Developed in 2012 and named Method of the Year in 2020²⁴, spatial transcriptome profiling has been utilized to profile global gene expression while retaining the spatial features in various tissues including brain²⁵, tumor²⁶, and adipose tissue²⁷. The technologies can be targeted, using specialized probes specific for particular RNA sequences attached to affinity reagents or fluorescent tags, thus detecting select genes at subcellular resolution^28,29,30,31. They can also be untargeted^32,33, typically using spatially barcoded oligonucleotides to capture RNA, which together get converted to cDNA for subsequent seq library preparation and hence has the advantage of deducing whole tissue gene expression in an unbiased manner. However, spatial resolution is not currently achieved at a single cellular level with commercially available technologies. This can be overcome to some extent with data integration with scRNA-seq data, ultimately allowing the mapping of single-cell transcriptome in a complex tissue context while retaining its original spatial information³⁴.

Herein, a workflow is described to profile EC transcriptome using human superior mesenteric artery, a peripheral artery that has been used to study vasodilation, vascular remodeling, oxidative stress, and inflammation^35,36,37. Two techniques are described: 1) to isolate and enrich ECs from the intima of blood vessels combining mechanical dissociation and enzymatic digestion suitable for single-cell transcriptome sequencing or subsequent in vitro culture; 2) to prepare arterial sections for spatial transcriptome profiling (Figure 1). These two techniques can be performed independently or complementarily to profile ECs and their surrounding cells. Furthermore, this workflow can be adapted for use on any medium or large artery.

Protocol

Human tissue studies were conducted on deidentified specimens obtained from the Southern California Islet Cell Resource Center at City of Hope. The research consents for the use of postmortem human tissues were obtained from the donors' next of kin, and ethical approval for this study was granted by the Institutional Review Board of City of Hope (IRB No. 01046).

1. Physical dissociation (estimated time: 1-2 h)

1. Place a fresh artery on a 10 cm dish and wash with sterile Dulbecco's phosphate-buffered saline (D-PBS).
2. With sterile forceps and dissection scissors, remove the fat and outer connective tissues until the vessel is clean. Measure the length of the vessel with a ruler placed outside of the dish and take pictures for records (Figure 1A).
3. Pre-warm appropriate digestion buffer to 37 °C: for scRNA-seq use cell-dissociation enzyme; for cell culture assays use 1-6 mg/mL of Collagenase D, 3 mg/mL of bacteria-derived protease, 100 mM of HEPES, 120 mM of NaCl, 50 mM of KCl, 5 mM of glucose, with 1 mM of CaCl₂^6,38 (pH 7.0, does not require adjustment) added 5 min prior to use.
4. Take the mesenteric artery (typically with a diameter of 6-8 mm) and cut lengthwise with scissors to open the vessel lumen vertically.
5. Using needles, attach the vessel onto black wax on all four corners, leaving the intima exposed (Figure 1B).
6. Add 1 mL of pre-warmed digestion buffer to intima. Take a sterile scalpel and scrape the lumen of the vessel gently twice.
   NOTE: The purpose of this procedure is to dissociate the intimal layer, which may need practice. Enough force needs to be exerted to remove endothelium, but not so much that the deeper layers of tissue are also removed, which will reduce the EC representation.
7. Transfer the digestion buffer into a 5 mL tube. Add 1 mL of digestion buffer to the intima and pipette up and down carefully to collect the remaining cells and add to the 5 mL tube. Incubate cells at 37 °C, rotating at 150 rpm for 5 min.
8. Add 2 mL of M199 medium (or D-PBS if proceeding with scRNA-seq) to the cell suspension to quench the enzymatic reaction. Mix gently and centrifuge at 4 °C, 600 x g for 5 min.
9. Remove the supernatant and store the supernatant separately to culture as a control to observe if all cells are captured in the pellet in step 1.8. Resuspend the cell pellet in 1 mL of M199 medium (or D-PBS if proceeding with scRNA-seq).
10. Assess the cell viability by mixing 10 µL of trypan blue with 10 µL of cell stock. Observe the cell morphology and count the cells using a hemocytometer.
    ​NOTE: At this point, cells can be used for protein or RNA quantification or cultured in vitro using EC culture media following standard protocols³⁹. Alternatively, they can be prepared for sequencing as described in the next section.
11. For the culture, coat two wells of a 6-well plate by pipetting 500 µL of attachment reagent into wells for 30 min at room temperature. After removal of attachment reagent and washing with sterile D-PBS, dispense the full cell stock into one well, and the supernatant kept as a control into the second well.

2. Preparation for scRNA-seq studies (estimated time: 3-4 h)

1. Centrifuge the cell stock from step 1.11 at 4 °C, 600 x g for 5 min. Remove the supernatant and resuspend the pellet gently with a P1000 pipette and wide-bore tip in 1 mL of 0.04% bovine serum albumin (BSA) in D-PBS. Mix well to ensure a single-cell suspension.
2. Pass the solution through a 40 µm strainer to remove cell debris. If debris remains, pass through a second strainer into 4 mL of 0.04% BSA in D-PBS.
3. Centrifuge at 4 °C, 600 x g for 5 min. Remove the supernatant and resuspend the pellet in 500 µL of 0.04% BSA in D-PBS using a P1000 pipette with a wide-bore pipette tip. Mix well to ensure a single-cell suspension.
4. Assess the cell viability by mixing 10 µL of cell stock with 10 µL of trypan blue. Using a hemocytometer, observe the morphology, determine whether there is a single-cell suspension without clusters, check for tissue debris, and calculate the number of living and dead cells.
5. If cells are clustered or debris remains, repeat washing with 0.04% BSA in D-PBS or pass through 40 µm strainer again.
   NOTE: While this will reduce the yield, it is important that cells exist in a clean single-cell suspension for optimal sequencing.
6. The single-cell suspension can be used for scRNA-seq.

3. Spatial transcriptome profiling (estimated time: 3-4 h)

1. Add isopentane (2-methylbutane) to a metal canister and chill in liquid nitrogen (LN₂) or dry ice (LN₂ is preferred as it will bring the temperature lower than dry ice). Pour optimal cutting temperature compound (OCT) in wells of labeled plastic cryomolds taking care to not create bubbles and remove those that appear. Leave cryomold on dry ice to chill.
2. Cut at least two coronal sections, 1 cm in length of the vessel. Using long (12") forceps, submerge the tissue in isopentane until frozen. Quickly submerge the tissue in OCT at orientation with the lumen visible in the center. Take care to remove any bubbles, especially those next to the tissue.
3. Using long (12") forceps, grasp the cryomolds, and hold them in the metal canister. While the base of the molds should be in the liquid, it should not be too deep to allow isopentane to run over the tissue. Observe the freezing, the OCT will become white progressively from the outside in 1-2 min.
4. Store these sections in a sealed container at -80 °C for up to 6 months.
   NOTE: Maintain samples on dry ice at all times and do not allow more than one freeze-thaw cycle. This tissue can be used for histological analysis, RNA/DNA fluorescence in situ hybridization (FISH), or spatial transcriptomics, which will be described now.
5. Set up the cryostat temperature to -20 °C for the chamber and -10 °C for the specimen head. Equilibrate OCT embedded vessel sections, knives, brushes, and slides to -20 °C in the cryostat for approximately 30 min. While equilibrating, clean the machine and all equipment that may touch the sections, including the blade, with 70% ethanol followed by RNase decontamination solution.
6. Attach the sample by dispensing a small amount of OCT onto the circular cryostat block and placing the sample on top before it freezes. Place the block into the center of the specimen head and screw the block in place using a tall black handle on the left. Cut away any excess OCT surrounding the vessel.
7. Setting the cutting thickness to 10 µm on the cryostat, cut approximately 60 sections and place them in a pre-cooled 1.5 mL tube. These sections will be used to assess the RNA quality. Upon removal from cryostat, immediately add 1 mL of RNA extraction reagent and vortex until the sections are completely dissolved.
8. Add 200 µL of chloroform and mix until the solution resembles a strawberry milkshake. Centrifuge at 11,200 x g for 10 min at 4 °C.
9. Collect the aqueous phase (clear phase on top of white and pink layers) and add 500 µL of isopropanol to it. Mix by inverting the tubes over 10 times and place at -80 °C for at least 20 min.
10. Centrifuge at 11,200 x g for 10 min at 4 °C. Dissolve the purified RNA pellet in 5 µL of RNase-free water. Examine the RNA Integrity Number (RIN).
    NOTE: Samples with RIN ≥ 7 are preferred, but RIN ≥ 6 is acceptable. The number of tissue sections and volume of solution for RNA dissolving may require optimization depending on the system used to ensure the final RNA concentration is within the RIN function range to allow for accurate RIN evaluation.
11. Practice cutting and placing sections properly within the fiducial frames using plain glass slides before proceeding to commercially available tissue optimization or gene expression slides. This can be achieved by drawing 6.5 mm x 6.5 mm squares on a plain glass slide, cooling to -20 °C in the cryostat, and placing sections in this capture area.
12. Cut 10 µm sections with anti-roll plate in place, flip and carefully flatten by gently touching the section through the surrounding OCT.
13. Using RNase-free cryostat brushes place the tissue section within the square, using only the surrounding OCT. Immediately place one finger (in gloves) on the backside of the capture area to melt the section to the slide.
14. Once the section has adhered, place the slide onto the cryobar to allow the section to freeze. Care must be taken to avoid tissue folding and tissue overlying the demarcated edges of the fiducial frame. If necessary, cut the tissue into halves or quarters along the lumen using a blade to ensure the vessel section fits within the 6.5 mm x 6.5 mm frame.
15. Cut 10 µm sections of tissue as per 3.12-3.14 onto tissue optimization or gene expression slides. Transfer the slide to a slide mailer placed on dry ice. Store slides at -80 °C for up to 4 weeks before proceeding with published spatial protocols^33,34.

4. Sequencing data analysis (estimated time: up to 1 week depending on familiarity with software)

NOTE: For spatial transcriptomic analysis only, skip to step 4.8. scRNA-seq data is processed using the standardized pipeline aligned to human hg38 reference transcriptome. The R package Seurat (v3.2.2) is used to analyze scRNA-seq data following published guidelines⁴⁰.

1. Filter using well-established quality control metrics: rare cells with very high numbers of genes (potentially multiplets) and cells with high mitochondrial percentages (low-quality or dying cells often present mitochondrial contamination) are removed.
2. Normalize data using "sctransform", a method to improve sample integration compared to log-normalization. These normalized data are used for dimensionality reduction and clustering, while log-normalized expression levels are used for analysis based on gene expression levels, such as cell-type classification⁴¹.
3. Select marker genes per cell type, for example, platelet endothelial cell adhesion molecule-1 (PECAM1) and von Willebrand factor (VWF) for ECs, lumican (LUM) and procollagen C-Endopeptidase Enhancer (PCOLCE) for fibroblasts, B-cell translocation gene 1 (BTG1) and CD52 for macrophages, C1QA and C1QB for monocytes, and myosin heavy chain 11 (MYH11) and transgelin (TAGLN) for vascular smooth muscle cells³⁶.
4. Compute the average expression level across single cells between each pair of markers in order to have a single marker with an average expression level associated with each cell type⁴².
5. Apply a Gaussian Mixture Model (GMM) with two components to the expression data of each marker across single cells in order to separate cells into two sets, namely, highly expressing the marker and lowly expressing the marker. In this way, for each marker, each single cell is assigned to one of the two components⁴².
6. Use statistical enrichment for the set of marker genes, a Fisher's exact test, to assign a cell type to each cluster. By doing so, a set of p-values are obtained per cluster, each corresponding to a cell type. The cell type with the lowest p-value is assigned to that cluster.
7. Process the spatial transcriptomic data using the Space Ranger pipelines resulting in spatial de-barcoding and generation of quality control (QC) metrics, including total aligned reads to hg38 human transcriptome, the median number of Unique Molecular Identifier (UMI) or genes per spot³⁵.
   NOTE: The R package Seurat (v3.2.2) is used to analyze the Space Ranger-processed data following published guidelines⁴⁰.
8. Normalize data using "sctransform", a method which is demonstrated to improve the downstream analysis compared to log-normalization given the heterogeneity of the tissue and the very high variance in counts across the spots.

Results

The analysis of ECs from mesenteric artery using a combination of mechanical and enzymatical dissociation or cryopreservation for use in various downstream assays is depicted here (Figure 1). ECs can be profiled in mesenteric arteries using the following steps: A) mechanical dissociation from the intima coupled with collagenase digestion to culture cells; B) generation of single-cell suspension for scRNA-seq; or C) cross-sections of the artery can be embedded in OCT to be cryosectioned to pr...

Discussion

The presented workflow details a set of techniques to profile ECs from a single piece of human artery with single-cell and spatial resolution. There are several critical steps and limiting factors in the protocol. One key to transcriptome profiling is the freshness of the tissue and RNA integrity. It is important to maintain tissues on ice as much as possible prior to processing to minimize RNA degradation. Typically, the post-mortem tissues are processed between 8-14 h after the time of death. However, beginning the iso...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1.5 mL micro-centrifuge tube	USA Scientific	1615-5500
10 cm dish	Genesee Scientific	25-202
23G needles	BD	305145
2-methylbutane	Thermo Fisher	AC327270010
40 µm strainer	Fisher	14100150
4200 TapeStation System	Agilent Technologies	G2991BA
5 mL tube	Thermo Fisher	14282300
6-well plate	Greiner Bio-One	07-000-208
Attachment factor	Cell Applications	123-500	Attachment reagent in the protocol
Black wax			Any commercial black wax can be used
Bovine serum albumin heat shock treated	Fisher	BP1600-100
CaCl2	Fisher	BP510
Centrifuge	Eppendorf
Chloroform	Fisher	C607
Collagenase D	Roche	11088866001
Cryostat	Leica
Cryostat brushes
D-Glucose	Fisher	D16-1
Dimethyl sulfoxide	Fisher	MT25950CQC
Dispase II	Roche	4942078001	Bacteria-derived protease in the protocol
Disposable Safety Scalpels	Myco Instrumentation	6008TR-10
D-PBS	Thermo Fisher	14080055
Ethanol	Fisher	BP2818-4
Fetal bovine serum	Fisher	10437028
Hemocytometer	Fisher	267110
HEPES	Sigma Aldrich	H3375-100g
High sensitivity D1000 sample buffer	Agilent Technologies	5067-5603
High sensitivity D1000 screen tape	Agilent Technologies	5067-5584
Incubator			Kept at 37 °C 5% CO2
Isopropanol	Fisher	BP26324
KCl	Fisher	P217-3
Liquid nitrogen
Medium 199	Sigma Aldrich	M2520-10X
Metal cannister
Microscope	Leica		To assess cell morphology
Microvascular endothelial culture medium	Cell Applications	111-500
NaCl	Fisher	S271-1
New Brunswick Innova 44/44R Orbital shaker	Eppendorf
Optimal Cutting Temperature compound	Fisher	4585
Plastic cryomolds	Fisher	22363553
RNA screen tape	Agilent Technologies	5067-5576
RNA screen Tape sample buffer	Agilent Technologies	5067-5577
RNase ZAP	Thermo Fisher	AM9780
RNase-free water	Takara	RR036B	RNase-free water (2) in kit
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
Superfrost PLUS Gold Slides	Fisher	1518848
TRIzol reagent	Fisher	15596018
Trypan Blue	Corning	MT25900CI
TrypLE Express Enzyme (1X) phenol red	Thermo Fisher	12605010	Cell-dissociation enzyme in the protocol
Visium Accessory Kit	10X Genomics	PN-1000215
Visium Gateway Package, 2rxns	10X Genomics	PN-1000316
Visium Spatial Gene Expression Slide & Reagent Kit, 4 rxns	10X Genomics	PN-1000184