Endothelial identity not found – Beyond passage 38, commercial cardiac microvascular endothelial cells do not express CD31 and VE-cadherin

Abstract

Few immortalized cardiac microvascular endothelial cell (CMEC) lines are available, particularly mouse lines. We purchased the CLU510 mCMEC line (Cedarlane), isolated by fluorescence-activated cell sorting for CD31 and VE-cadherin. The cell line has been used in previous studies, although none report CD31 or VE-cadherin expression. We analyzed endothelial profile of two vials of passage 38 cells. CD31 and VE-cadherin mRNA were hardly expressed in mCMECs compared to primary mouse lung ECs. CD31 and VE-cadherin protein levels were also negligible compared to multiple EC lines. Thus, CLU510 mCMECs beyond P38 do not harbor an endothelial phenotype. Caution should be warranted when using commercial cells and journals should carefully consider the validity of results when essential characterization of cell lines is omitted.

1. Introduction

For many applications, commercial cell lines are a convenient option. Usually, quality control experiments are provided as part of the datasheet and as such, validation is not performed by scientists. If commercial cell lines differentiate into other cell types, or undergo other unwanted changes, and cell lines are not properly validated, reproducibility of the data obtained from these experiments will be low. Endothelial cell (EC) lines are readily available, with some highly reliable established lines such as human umbilical vein ECs (HUVECs). However, few commercial cardiac microvascular EC (CMEC) lines are available, in particular mouse lines (mCMECs).

The CLU510 mCMEC line, known in literature as MCEC, was generated by Barbieri and Weksler and published in 2007 [[1], [2], [3]] and commercialized by Cedarlane. This line was isolated by performing fluorescence-activated cell sorting (FACS) for CD31, followed by a second round of FACS for VE-cadherin (CD144). The authors also provide representative staining images for CD31 and VE-cadherin, as well as β-catenin and von Willebrand factor [1]. The CLU510 cell line has been utilized in various lines of research (Supplementary Table S2). None of the studies that utilized this commercial cell line showed, in absolute values, that the cells were CD31 or VE-cadherin positive, which are the markers on which these cells were originally isolated.

Initially obtained to perform studies with CLU510 mCMECs, we noted that our mCMECs did not appear positive for endothelial markers. Thus, we assessed the expression of EC markers on which these cells were isolated at mRNA and protein level. Relative gene expression and protein levels of CD31 and VE-cadherin were near undetectable in CLU510 mCMECs. Thus, the commercial line beyond P38 cannot be considered endothelial.

2. Methods

2.1. Cell culture CLU510 mCMECs

Two biological replicates (vials) of CLU510 mCMECs (G111114 P38, nov 11 2014 TM and B111114 P38, nov 10 2114 TM) were thawed according to the manufacturer's instructions. Both vials were received at passage 38 (P38), which for naming purposes is P1 in this paper. As per the manufacturer's instructions, culture medium consisted of Dulbecco's Modified Eagle Medium (DMEM, Sigma, containing 24,000 mg/L glucose, 0.584 g/L glutamine, 3.7 g/L sodium bicarbonate, and without sodium pyruvate), supplemented with 5 % v/v heat-inactivated fetal bovine serum (FBS) (Gibco), 10 mmol/L HEPES (Sigma), and 10 mmol/L penicillin-streptomycin (P/S) (Thermo Fisher Scientific). Frozen cells were thawed according to the manufacturer's instructions and seeded on 0.2 % gelatin coated wells or in T75 flasks. Passing and freezing of cells was also performed per the manufacturer's instructions. Cells were passed up to P5. Cells were incubated at 37 °C and 5 % CO₂.

2.2. Cell culture positive controls

We obtained primary mouse lung ECs (mLECs), HUVECs, and human CMECs (hCMECs) to serve as positive controls. All cells were passaged 3 times. The frozen mLECs were thawed and cultured on 0.2 % gelatin coated plates in endothelial cell medium (ECM, ScienCell) with the included supplements (5 % FBS, endothelial cell growth supplement (ECGS), and P/S). HUVECs (CC-2519, Lonza) were cultured in the same medium. hCMECs (CC-7030, Lonza) were thawed and cultured on 0.2 % gelatin coated plates in endothelial growth medium (EGM-2MV, ScienCell) with the included supplements (5 % FBS, hydrocortisone, hFGF-B, VEGF, R3-IGF-1, ascorbic acid, hEGF, and P/S). Cells were incubated at 37 °C and 5 % CO₂. Medium was refreshed every other day until cells reached full confluence to proceed with RNA isolation and protein lysates.

2.3. RNA isolation and RT-qPCR

Of the two biological mCMEC replicates, three technical replicates (n = 3) were obtained for RT-qPCR to enhance validity of the results. For G111114, cells were P3–5, and for B111114 cells were P2–4. Positive control mLECs were P2–4 (n = 3). Cells were lysed with TRIzol™ reagent (Ambion). RNA was isolated with the RNA MiniPrep protocol (Zymo Research), according to the manufacturer's instructions. 1 μg of RNA was reverse transcribed to complementary deoxyribonucleic acid (cDNA) with the iScript™ cDNA synthesis kit using a T100™ thermal cycler (Bio-Rad Laboratories Inc.). Real-time quantitative polymerase chain reaction (RT-qPCR) was performed with 12.5 ng cDNA and detected by addition of iQ™ SYBR Green using a C1000™ Thermal cycler (Bio-Rad Laboratories Inc.). We determined relative expression levels of EC markers CD31 and VE-cadherin, as well as RLP0, which served as the housekeeping gene (Supplementary Table S1). All samples were loaded as triplicates. Relative gene expression was determined based on the 2^-dCt method.

2.4. Western blot

Of the two biological mCMEC replicates, three technical replicates (n = 3) of G111114, and B111114 were obtained. Cells were P3–5, and P2–4, respectively. Wells were washed 1× with phosphate buffered saline (PBS, Gibco) and lysed by addition of lysis buffer (4× NuPage sample buffer, 1 M DTT, MilliQ). Lysates were heated to 99 °C and stored at −80 °C until use. Prior to loading the lysates, they were heated to 95 °C and briefly spun down. Cell lysates were loaded on 4–15 % precast Criterion™ gradient gels (Bio-Rad Laboratories Inc.). HUVEC (P3–5), hCMEC (P4–6), and mLEC (P2–4) lysates were loaded in triplicate as positive controls. Proteins were separated by electrophoresis in sodium dodecyl sulfate (SDS) running buffer run at 100 V until the dye front reached the bottom of the gel. Wet tank membrane transfer with PVDF membranes ran at 0.3 A for 120 min. Membranes were blocked in 3 % (w/v) bovine serum albumin (BSA) in tris-buffered saline with 0.1 % (v/v) tween (TBS-T). Primary antibodies were incubated 1:1000 in 3 % BSA-TBS-T overnight at 4 °C, and were as follows: Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH, 2118S, Cell Signaling), CD31 (ab28364, Abcam), VE-cadherin (ab33168, Abcam), eNOS (9572S Cell Signaling), and VCAM-1 (13662S, Cell Signaling). After washing the membranes in TBS-T, secondary antibody (anti-rabbit immunoglobulin G-horseradish peroxidase (P0448, Dako)) was incubated 1:5000 in 3 % BSA in TBS-T for 1 h at room temperature. Membranes were imaged on the Amersham Imager 600 (GE Healthcare Bio-Sciences AB). Protein levels relative to GAPDH were analyzed with Image Studio Lite (LI-COR Biosciences).

2.5. Immunofluorescence

mCMECs (G111114 and B111114, P2–3) and HUVECs (P3) were seeded in 24 wells black μ-plates (82,426, Ibidi) and cultured as described above. After reaching confluence, mCMECs and HUVECs were washed with PBS and fixed in 4 % (w/v) PFA for 10 min followed by 3 PBS washes and stored at 4 °C until processed. Cells were permeabilized in 0.25 % (v/v) Triton X-100 in TBS at room temperature for 15 min whilst shaking. Blocking was done in 1 % (w/v) BSA in TBS for 1 h at room temperature. Primary antibodies for VE-cadherin 1:200 (ab33168, Abcam) and CD31 1:200 (ab28364 Abcam, BBA7 R&D Systems, 250,590 Abbiotec, and 1506 Santa Cruz, Supplementary Fig. S1) were incubated overnight at 4 °C in 1 % BSA in TBS. After three washes with PBS, secondary antibodies (AF488 donkey anti-Rabbit IgG (A21206), AF488 donkey anti-mouse IgG (A21202), Thermo Fisher Scientific), phalloidin (1:250, A22287, Thermo Fisher Scientific), and DAPI (1:400, D1306, Thermo Fisher Scientific) were incubated in 1 % BSA in TBS for 1 h, at room temperature and protected from light. Cells were washed three times in PBS and subsequently stored in PBS at 4 °C, protected from light until image acquisition. Images were obtained at 40× or 63× oil-immersion magnification on the Zeiss Axiovert 200 M (Zeiss Group) equipped with X-light V3 spinning disc (CrestOptics S.p.A.) and Prime BSI Express sCMOS camera (Teledyne Photometrics).

2.6. Statistical analyses

Data are presented as mean ± standard error of the mean (SEM). Relative gene expression of CD31, VE-cadherin, and RLP0 in mCMECs compared to mLECs was determined with Student's t-test. Comparisons between more than two groups (protein levels between mCMECs, hCMECs, and HUVECs) were analyzed with One-way ANOVA and Tukey's post hoc multiple testing correction. Analyses were performed with GraphPad Prism 9.1.0 (GraphPad Software). A p-value <0.05 was considered statistically significant.

3. Results

3.1. Absence of EC markers CD31 and VE-cadherin at RNA and protein levels

RT-qPCR of two biological replicates of CLU510, G111114 and B111114 (n = 3 each), revealed low expression of CD31 and VE-cadherin (Fig. 1A,B), the two EC-specific markers that were originally used for cell isolation and selection. Relative expression of CD31 and VE-cadherin was similar for the two biological replicates. Compared to primary mLECs (n = 3), relative CD31 expression was >170-fold lower for both mCMEC vials (p = 0.0003) and VE-cadherin expression was at least 164-fold lower in mCMECs than in primary mLECs (p = 0.0481 B111114 compared to mLECs, p = 0.472 G111114 compared to mLECs).

Fig. 1.

Expression of CD31 and VE-cadherin in CLU510 mCMECs. Relative mRNA expression of (A) CD31 and (B) VE-cadherin. Assessed in two biological replicates (CLU510 G111114, P3–5, yellow/black squares and B111114, P2–4, orange/white squares, n = 3 passages), and in control mLECs P2–4 (n = 3, black). Values are relative to housekeeping gene RLP0. Mean ± SEM. Student's t-test. *p < 0.05, ***p < 0.001 G111114 and B111114 compared to mLECs. (C) CD31 and (D) VE-cadherin protein levels relative to GAPDH, normalized to hCMECs. Assessed in two CLU510 mCMEC vials (G111114, P3–5, yellow/black squares and B111114, P2–4, orange/white squares, n = 3 passages). Positive controls are mLECs P2 (for antibody specificity only), hCMECs P4–6 (n = 3, grey) and HUVECs P2–4 (n = 3, white). Mean ± SEM. One-way ANOVA. **p < 0.01, ***p < 0.001 G111114 and B111114 compared to hCMECs. ^##p < 0.01 G111114 and B111114 compared to HUVECs. ^††p < 0.01 HUVECs compared to hCMECs. (E) Morphology of CLU510 mCMECs. Immunofluorescent images of G111114 mCMECs (P2) and HUVECs (P3). VE-cadherin (green, ab33168, Abcam), f-actin (magenta, A22287, Thermo Fisher Scientific), and DAPI (cyan, D1306, Thermo Fisher Scientific). Scale bar represents 50 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

As transcriptional expression of the two markers used to isolate immortalized CLU510 mCMECs was very low, absence of protein was also expected. Indeed, Western blots revealed little to no detectable CD31 (p < 0.001) and VE-cadherin (p < 0.01) in two CLU510 vials (B111114 and G111114, n = 3 each) compared to commercial hCMECs and HUVECs (n = 3 each) also isolated based on CD31 and VE-cadherin sorting (Fig. 1C,D). Under basal conditions mCMECs also did not express eNOS and negligible VCAM-1 protein (Supplementary Fig. S1). To illustrate cellular morphology, mCMECs were imaged for VE-cadherin, nuclear DNA, and f-actin (Fig. 1E), and CD31 (Supplementary Fig. S2).

4. Discussion

The CLU510 mCMECs were isolated by performing FACS for CD31, followed by a second round of FACS for VE-cadherin. The authors also provide representative staining images for CD31 and VE-cadherin to show presence at protein level, although this was not quantified [1]. Our validation for mRNA of these markers detected 519-fold lower CD31 in G111114 vial mCMECs and 170-fold lower CD31 in B111114 vial mCMECs, compared to primary mLECs. Similarly, relative VE-cadherin mRNA expression was 1036-fold and 164-fold lower in G111114 and B111114 CLU510 mCMECs, respectively, compared to primary mLECs. In line with this, protein levels of CD31 and VE-cadherin were negligible in comparison to commercial hCMECs and HUVECs.

Unfortunately, the studies that utilized the CLU510 cell line have not published findings (or lack thereof) of CD31 or VE-cadherin (Supplementary Table S2) on which the mCMECs were originally isolated [1]. A 2008 publication by the original authors soon after the cell line was characterized, showed expression of VE-cadherin with immunofluorescence [2]. Apart from the original studies, Wilhelmi et al. (2020) quantified relative CD31 expression by RT-qPCR but normalized treated mCMECs to untreated CLU510 mCMECs [4]. Similarly, Mu et al. (2021), quantified relative VE-cadherin expression of treated CLU510 mCMECs by normalizing untreated mCMECs VE-cadherin levels by RT-qPCR [5]. Due to the normalization the basal CD31 and VE-cadherin mRNA levels remain unknown. Perhaps the absolute values were low, indicating negligible CD31 and VE-cadherin mRNA levels similar to our findings.

Thus, with time and high passaging it is possible that the cells have lost their endothelial phenotype. Overall, we urge authors and reviewers to remain critical of commercial cell lines, and ensure that cell lines were validated before performing experimental studies.

Funding

This work is supported by Netherlands Cardiovascular Research (CVON) and Dutch CardioVascular Alliance (DCVA) initiatives of the Netherlands Heart Foundation (2020B005 DCVA-DOUBLE-DOSE).

CRediT authorship contribution statement

Sarah Hilderink: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Jolanda van der Velden: Funding acquisition, Project administration, Supervision, Writing – review & editing. Diederik W.D. Kuster: Conceptualization, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material