hiPSC Differentiation to Brain Microvascular Endothelial Cells (BMEC)

ByMolly McCloskey

INTRODUCTION

The blood-brain barrier (BBB) is a highly selective endothelial barrier that serves a crucial role in maintaining tissue homeostasis in the brain. Many current models, however, suffer from low TEER and permeability values far greater than physiological levels. Recently, human induced pluripotent stem cells (hiPSC) have been differentiated into BMECS with very high TEER and good expression of tight junction proteins. We sought to replicate this protocol in our lab with non-patient specific iPSCs in order to develop a BBB model in our microchip devices.

METHODS

Differentiation 

Before differentiation (D-3):

hiPSCs (WiCell, [iPSC(IMR90)-4]) were singularized with Accutase and plated onto the vitronectin-coated plates at a surface density of 1.7 x 10^5, 2.7 x 10^5, or 3.7 x 10^5 cells/mL (equivalent to 3.7 x 10^4, 5.8 x 10^4, and 7.9 x 10^4 cells/cm^2 based on email from Britta’s employee) in mTeSR1 supplemented with 10 µM ROCK inhibitor Y-27632 in a 24-well plate. hPSCs were expanded in mTeSR1 (no ROCK inhibitor) for 3 days.

Experiment 1: A separate set of cells were plated at 50 x 10^3 cells/cm^2 and expanded as described above and processed for immunochemistry to check for markers of pluripotency, Oct-4 and Nanog (Day 0 of differentiation, parallel cells).

To initiate differentiation at day 0 (D0):

Experiment 2: Cells were treated with 0, 6, 8, and 10 µM CHIR99021 in DeSR1 (DMEM/Ham’s F12, 1xMEM-NEAA, 0.5xGlutaMAX, and 0.1 mM beta-mercaptoethanol).

Experiment 3: Cells were treated with 10 µM CHIR99021 in DeSR1 (DMEM/Ham’s F12, 1xMEM-NEAA, 0.5xGlutaMAX, and 0.1 mM beta-mercaptoethanol) and maintained in this medium for 4 days.

Primitive Streak (D1):

The medium was changed to DeSR2: DeSR1 plus 1x B27 every day for another 4 days.

Experiment 2: Cells were processed for immunochemistry to check for a primitive streak marker, Brachyury.

Experiment 3: A separate set of cells were treated as described above and processed for immunochemistry to check for a primitive streak marker, Brachyury.

Mesodermal Intermediate (D4):

Experiment 3: A separate set of cells were treated as described above and processed for immunochemistry to check for a mesodermal intermediate marker, PAX2.

Endothelial Cell Progenitors (D5):

Experiment 3: A separate set of cells were treated as described above and processed for immunochemistry to check for an endothelial cell progenitor marker, VEGFR2.

To drive BMEC specification (D6):

The medium was changed to hECSR1 (hESFM, 20 ng bFGF, 10mM RA, and B27) for 2 days.

Replate cells in 96 well plate, Transwells, and devices (D8):

Cells were singularized with Accutase and plated onto Collagen IV, Fibronectin-coated plates at a surface density of 1 x 10^6 cells/cm^2 in hECSR1 and maintained in this medium for one day.

BMEC maintenance (D9):

The medium was changed to hECSR2: hECSR1 lacking bFGF or RA for one day (multiple days for Transwells).

BMEC characterization (D10):

Cells were processed for immunochemistry to check for tight junction markers, ZO-1, Occludin, and Claudin-5.

Immunochemistry

Cells were rinsed with ice-cold phosphate-buffered saline (PBS) once, fixed in 4% paraformaldehyde (PFA) for 15 min. Cells were then blocked with 10% goat serum in PBS containing 0.3% Triton X-100 for 30 min (10% PBSGT). Modifications to this protocol were made for Brachyury staining and ZO-1/Occludin/Claudin staining based on the published methods paper.

Primary antibodies:

Antibodies were diluted in 10% PBSGT (except Brachyury), and cells were incubated in the antibody solutions at 4°C overnight.

Secondary antibodies:

After three PBS washes, cells were incubated with secondary antibodies in 10% PBSGT (goat anti-mouse Alexa Fluor 488, except Brachyury and ZO-1) for 1 hour at room temperature. Cells were then washed with PBS three times and were visualized.

RESULTS

Immunochemistry

D0 Parallel, Markers of Pluripotency

Fig. 1. Expanded and plated cells maintain markers of pluripotent prior to inducing differentiation.

Experiment 2: D1, Marker of Primitive Streak (optimizing CHIR and seeding densities)

Fig. 2. Increased levels of our lot of CHIR are necessary to induce differentiation, with comparable results between seeding densities. Note nuclear localization of transcription factor with increased CHIR concentrations. Images are taken with a 20X objective.

Experiment 3: D1, Marker of Primitive Streak (optimizing seeding densities)

Fig. 3. Comparable results to previous experiment, with good induction of differentiation using 10 µM CHIR and 3 different seeding densities. Images are taken with a 4X objective.
Fig. 4. Comparable results to previous experiment, with good induction of differentiation using 10 µM CHIR and 3 different seeding densities. Images are taken with 4X and 20X objectives.

Experiment 3: D4, Marker of Mesodermal Intermediate (optimizing seeding densities)

Fig. 5. Cells have successfully transitioned into mesodermal intermediates, but staining is somewhat heterogenous throughout samples. Images are taken with 4X and 20X objectives.

Experiment 3: D5, Marker of Endothelial Cell Progenitors (optimizing seeding densities)

Fig. 6. Cells have successfully transitioned into endothelial progenitors but staining is even more heterogenous throughout samples. Note some localization at the periphery of cells, where an active receptor should be found. Images are taken with 4X and 20X objectives.
Fig. 7. Further examples demonstrating cells have successfully transitioned into endothelial progenitors but staining is heterogenous throughout samples. Images are taken with a 20X objective.

Experiment 3: D10, Tight Junction Markers to characterize BMECs (optimizing seeding densities)

96 Well Plate Staining:

Fig. 8. Heterogenous staining for Claudin and ZO-1 in 96 well plate, implying partial yield of successfully differentiated BMECs. There may be slightly stronger expression for cells seeded at 3.7 x 105cells/ml. Images are taken with 4X and 20X objectives.
Fig. 9. Heterogenous staining for Occludin and ZO-1 in 96 well plate, implying partial yield of successfully differentiated BMECs. There may be slightly stronger expression for cells seeded at 3.7 x 105cells/ml. Staining for Occludin is weaker than for Claudin. Images are taken with 4X and 20X objectives.

Microchip Device Staining:

Fig. 10. Heterogenous staining for Claudin and ZO-1 in microchip devices seeded at 3.7 x 105cells/ml, implying partial yield of successfully differentiated BMECs. Images are taken with 4X and 20X objectives.
Fig. 11. Heterogenous staining for Occludin and ZO-1 in microchip devices seeded at 3.7 x 105cells/ml, implying partial yield of successfully differentiated BMECs. Staining for Occludin is weaker than for Claudin. Images are taken with 4X and 20X objectives.

I did not include images from other devices and seeding densities, but the results were comparable to what we saw in the 96 well plate.

TEER

Need new EndOhm cup. Chopstick method was done quickly (had to run to class), with measurements same as background. Unclear if this is due to gaps in monolayer, poor pipetting, poor acquisition, etc… Fixed cells and will check monolayer using cresyl violet.

DISCUSSION

We have successfully cultured hiPSCs, with moderately efficient differentiation into endothelial progenitor cells. We are currently expanding these cells to freeze and try differentiation again in hopes to save time in future experiments. It was necessary to optimize CHIR concentrations for our lot, increasing the published 6 µM to 10 µM. We had poor differentiation to endothelial progenitors our first time through (not included in post), which we believe was due to suboptimal active concentrations of CHIR. In addition, we found comparable staining at each check point, but different morphologies, for cells seeded at 1.7, 2.7 and 3.7 x 10^5 cells/ml. In our final product BMECs, while we could successfully grow cells on both platforms, staining was very heterogenous in both devices and 96 well plates, likely due to suboptimal differentiation efficiency.

Future Plans:

We will try seeding endothelial progenitors at different concentrations, as we appeared to have multiple layers of cells instead of monolayer, and will likely try cell sorting the endothelial progenitors prior to reseeding. We will optimize TEER protocols using hCMEC/D3s, purchasing a new EndOhm chamber to improve accuracy of reading, before measuring TEER on our hiPSC-derived BMECs.

UPDATE

METHODS (1)

Differentiation (Changes only)

Replate cells in 96 well plate, Transwells, and devices (D8):

Cells were singularized with Accutase (NEW METHOD: sat in Accutase 30 min+ minutes until dissociated and in single cell) and plated onto Collagen IV, Fibronectin-coated plates at varying surface densities: 1 x 10^6 cells/cm^2,  0.75 x 10^6 cells/cm^2,  0.5 x 10^6 cells/cm^2, or  0.25 x 10^6 cells/cm^2, in hECSR1 and maintained in this medium for one day.

RESULTS

Immunochemistry

D10, Tight Junction Markers to characterize BMECs (optimizing reseeding method and densities)

24 Well Plate Staining:

Fig. 12. Mostly heterogenous staining for Occludin or Claudin-5 and ZO-1 reseeded in a 24 well plate at 1 x 10^6 or 0.5 x 10^6 cells/cm^2. Images are taken with a 10X objective using the confocal microscope.

96 Well Plate Staining:

Fig. 13. Mostly heterogenous staining for Claudin-5 and ZO-1 reseeded in a 96 well plate at 1 x 10^6 or 0.5 x 10^6 cells/cm^2. Images are taken with a 10X objective using the confocal microscope. Note the cells seeded at a lower density are less compact, but both appear to form good monolayers.

METHODS (2)

Differentiation (Changes only)

Replate cells in Transwells (D8):

Cells were plated onto Matrigel-coated 12-well and 24-well Transwell plates in hECSR1 and maintained in this medium for 4 days. TEER was taken each day using both EndOhm Chamber and STX methods.

RESULTS

TEER peaked around 20 ohm-cm^2 at Day 10.

DISCUSSION

In the first update experiment, we saw great junctional staining and monolayer formation. In the following experiment, cell morphology prior to replating was unusual, which may explain low TEER. TEER trends are the same (peak at D10) and size of Transwell does not appear to matter when replating.

PROTOCOL IS CHANGING. Our collaborators have highlighted issues with this protocol, and cells appear to be more epithelial than endothelial. They have been working on a novel protocol that is in final stages before submitting for publication. The cells they get with this protocol appear to be more endothelial and are representative of post-capillary venules (lower TEER, site of transmigration). I will put iPSC work on pause until I am trained on the novel protocol in Bern.

Similar Posts

Leave a Reply