Protocol for efficient generation of human artery and vein endothelial cells from pluripotent stem cells

Summary

Blood vessels permeate all organs and execute myriad roles in health and disease. Here, we present a protocol to efficiently generate human artery and vein endothelial cells (ECs) from pluripotent stem cells within 3–4 days of differentiation. We delineate how to seed human pluripotent stem cells and sequentially differentiate them into primitive streak, lateral mesoderm, and either artery or vein ECs. We differentiate stem cells in defined, serum-free culture media in monolayers, without feeder cells or genetic manipulations.

For complete details on the use and execution of this protocol, please refer to Ang et al. ¹

Graphical abstract

Highlights

• •Guidance on maintaining undifferentiated human pluripotent stem cells
• •Instructions to convert human pluripotent stem cells into primitive streak
• •Instructions to convert primitive streak into lateral mesoderm
• •Instructions to convert lateral mesoderm into either artery or vein endothelial cells

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Blood vessels permeate all organs and execute myriad roles in health and disease. Here, we present a protocol to efficiently generate human artery and vein endothelial cells (ECs) from pluripotent stem cells within 3–4 days of differentiation. We delineate how to seed human pluripotent stem cells and sequentially differentiate them into primitive streak, lateral mesoderm, and either artery or vein ECs. We differentiate stem cells in defined, serum-free culture media in monolayers, without feeder cells or genetic manipulations.

Before you begin

Here we describe a step-by-step method to differentiate human pluripotent stem cells (hPSCs, including embryonic and induced pluripotent stem cells) into artery and vein endothelial cells (ECs) with high speed and purity.¹ ECs constitute the inner layer of blood vessels, and are fundamental to human health and disease.^2,3,4 This differentiation protocol builds on preceding pioneering efforts to convert hPSCs into ECs,^{5,6,7,8,9,10,11,12,13,14,15,16,17,18} as well as salient studies of how artery vs. vein ECs arise in the developing embryo.^{2,19,20,21,22,23,24}

We differentiate hPSCs into artery and vein ECs in monolayer cultures through the sequential addition of media containing various activators and inhibitors of extracellular signaling pathways. We employ serum-free and defined culture media. No feeder cells or genetic manipulations are needed to effect differentiation.

Institutional permissions

At Stanford University, hPSC differentiation experiments require regulatory approval from the Stem Cell Research Oversight (SCRO) committee. Readers carrying out this experimental protocol must likewise secure permission from their respective institutions.

Prepare stocks of Geltrex basement membrane

Timing: Around 12 h (to thaw Geltrex overnight)

hPSCs cannot adhere to cell culture plastic alone. Instead, cell culture plastics must first be coated with basement membrane proteins to provide a permissive substrate for hPSC attachment and growth. Various options for basement membrane proteins exist. For cost and ease of use, we use Geltrex.

CRITICAL: Geltrex is a temperature sensitive hydrogel that polymerizes at room temperature. Always keep Geltrex cold (at −20°C or 4°C) and avoid prolonged exposure to room temperature. Geltrex exists in a liquid form only at approximately 4°C. When exposed to room temperature for prolonged amounts of time, it will solidify into a gel and become unusable. When working with Geltrex at room temperature, work fast.

• 1.When received from the supplier, the 5 mL vial of Geltrex should be stored in a −20°C freezer.
• 2.Before aliquoting Geltrex, thaw it in a 4°C refrigerator overnight.
• 3.
  Within a sterile cell culture hood, use a pipette to mix Geltrex and aliquot it into sterile 1.5 mL microcentrifuge tubes.

  • a.
    Pre-cool microcentrifuge tubes at −20°C to avoid exposing Geltrex stocks to room-temperature tubes.
  • b.
    Gently mix the thawed Geltrex aliquot with a p1000 pipette, taking care not to introduce air bubbles.
  • c.
    Aliquot 500 μL of Geltrex in 1.5 mL microcentrifuge tubes without prolonged exposure to room temperature. If prolonged exposure to room temperature is expected, aliquots can be prepared on ice or a cooling block.
  • d.
    Store these stocks immediately in the −20°C freezer. Frozen undiluted Geltrex stocks can be stored at −20°C for at least one year.
• 4.
  Create a working stock of diluted Geltrex by diluting it 1:100 with cold DMEM/F12. This 1:100 diluted Geltrex stock can be stored in the 4°C refrigerator for at least one month.

  • a.
    Place a frozen 500 μL aliquot of Geltrex within a sterile cell culture hood, and use a p1000 pipette to add 500 μL of cold DMEM/F12 to the 1.5 mL microcentrifuge tube.
  • b.
    Immediately place the 1.5 mL microcentrifuge tube in the 4°C refrigerator or on ice, for approximately 5 min. The frozen Geltrex will begin dissolving in the DMEM/F12 medium.
  • c.
    In parallel, while the Geltrex dissolves in the 1.5 mL microcentrifuge tube, separately add 49 mL of cold DMEM/F12 to a new 50 mL conical tube.
  • d.
    Use a p1000 pipette to triturate the Geltrex solution in the 1.5 mL microcentrifuge tube several times and ensure it is fully suspended.
  • e.
    Transferring the Geltrex solution to a new 50 mL conical tube containing 49 mL of cold DMEM/F12.
  • f.
    Triturate the 1:100 diluted Geltrex stock in the 50 mL conical tube to ensure that the Geltrex is evenly distributed throughout.
  • g.
    This 1:100 diluted Geltrex stock is now ready to use to coat cell culture plastics.

Pause point: Once diluted in DMEM/F12, Geltrex can be stored in the 4°C refrigerator for up to 2 weeks. If kept at 4°C, it will remain liquid.

Coating cell culture plates with Geltrex basement membrane

Timing: Around 1 h

• 5.Remove the liquid working stock of 1:100 diluted Geltrex from the 4°C refrigerator.
• 6.
  Within a sterile cell-culture hood, coat cell culture wells or dishes with Geltrex by pipetting diluted 1:100 Geltrex into the center of a well or dish.

  • a.
    Add a volume of Geltrex that is one-half the volume of media typically used to grow cells in that format.

Note: For example, add 0.25 mL Geltrex per well of a 24-well plate, 0.5 mL Geltrex per well of a 12-well plate, 1 mL Geltrex per well of a 6-well plate, and 5 mL Geltrex per 10-cm dish.

• 7.Quickly return the liquid working stock of diluted 1:100 Geltrex to the 4°C refrigerator.
• 8.Gently swirl the cell culture plate or dish to ensure that the layer of Geltrex is evenly distributed across the surface of the well or dish.

Note: If there are visible gaps in the Geltrex layer, the basement membrane matrix will not adsorb to that part of the dish, and hPSCs will not be able to attach at those locations.

Note: If swirling cannot fill these gaps, additional Geltrex can be added until the entire plate surface is covered.

• 9.Place the Geltrex-coated cell culture dishes or plates in the 37°C incubator for at least 30 min, to allow Geltrex to polymerize and adsorb onto the surface of the dish or plate.

Note: Geltrex can also be coated for up to two days. We avoid Geltrex coating beyond two days because eventually, the Geltrex solution begins to evaporate in the 37°C incubator.

CRITICAL: Add Geltrex to the cell culture plate for a minimum of 30 min to allow it to polymerize on the bottom of the plate. If prematurely removed, the Geltrex will not adequately coat the plate and hPSC adhesion will be compromised.

Pause point: Geltrex-coated cell culture plates can be stored in the 37°C incubator for up to 2 days.

Reconstitute stocks of growth factors and small molecules

Timing: Around 3 h

Pause point: This PBS + 0.1% BSA solution (Table 1) can be stored at 4°C for up to 1 year.

• 10.
  Prepare a filtered solution of PBS + 0.1% BSA, which will be used to subsequently resuspend recombinant growth factor proteins.

  • a.
    Add 666 μL of BSA Fraction V to 50 mL of PBS, and then sterilely filter it through a 50-mL-size 0.22 μm filter unit in the cell culture hood.
• 11.
  In a sterile cell culture hood, use PBS + 0.1% BSA solution to resuspend recombinant growth factor proteins.

  • a.
    Resuspend Activin A powder to 50 μg/mL using PBS + 0.1% BSA. Activin is a TGF-β pathway activator.
  • b.
    Resuspend BMP4 powder to 50 μg/mL using PBS + 0.1% BSA. BMP4 is a BMP pathway activator.
  • c.
    Resuspend FGF2 powder to 100 μg/mL using PBS + 0.1% BSA. FGF2 is an FGF pathway activator.
  • d.
    Resuspend VEGF-A powder to 100 μg/mL using PBS + 0.1% BSA. VEGF-A is a VEGF pathway activator.
  • e.
    Aliquot growth factors into 1.5 mL microcentrifuge tubes.

Note: Aliquots should be prepared with reasonable working volumes so repeated freezing and thawing are avoided. Once resuspended to 50–100 μg/mL, these growth factors can be stored in the −20°C freezer for at least 1 year.

Note: It is important to store recombinant growth factors at high concentrations (50–100 μg/mL), as proteins are more stable at high concentrations. Always check manufacturer’s notes for any lot-specific recommendations.

• 12.In a sterile cell culture hood, resuspend 100 mg of insulin powder, which was lyophilized from a hydrochloric acid solution, in 10 mL of sterile water to generate a 10 mg/mL insulin stock.

Note: Aliquots of this 10 mg/mL insulin stock should be stored at −20°C, where they are stable for at least 1 year.

• 13.
  In a sterile cell culture hood, resuspend small molecule chemical compounds in recommended diluent.

  Note: Most small molecules can be resuspended in DMSO; however, manufacturer’s notes for lot-specific solubility should always be checked before resuspending small molecules.

  Note: Allow time for small molecules to fully dissolve. Ensure that residual small molecule powder in the cap of the original vial from the manufacturer is also dissolved.

  • a.
    Resuspend Thiazovivin powder to 20 mM using DMSO. Thiazovivin is a ROCK inhibitor.²⁵
  • b.
    Resuspend CHIR99021 powder to 10 mM using DMSO. CHIR99021 is a GSK3 inhibitor,²⁶ and therefore activates WNT signaling.
  • c.
    Resuspend XAV939 powder to 10 mM using DMSO. XAV939 is a tankyrase inhibitor, and therefore inhibits WNT signaling.²⁷
  • d.
    Resuspend Forskolin powder to 10 mM using DMSO. Forskolin is an adenylate cyclase activator, and therefore elevates intracellular cAMP levels.²⁸
  • e.
    Resuspend SB505124 powder to 10 mM using DMSO. SB505124 is a TGF-β pathway inhibitor.²⁹
  • f.
    Resuspend GDC0941 powder to 10 mM using DMSO. GDC0941 is a PI3K inhibitor.³⁰
  • g.
    Resuspend RO4929097 powder to 10 mM using DMSO. RO4929097 is a γ-secretase inhibitor, and therefore inhibits NOTCH signaling.³¹
  • h.
    Resuspend PD0325901 powder to 10 mM using DMSO. PD0325901 is a MAPK/ERK pathway inhibitor.³²
  • i.
    Resuspend DMH1 powder to 10 mM using DMSO. DMH1 is a BMP pathway inhibitor.³³
  • j.
    Once resuspended to 10–20 mM in DMSO, the small molecules can be stored in the −20°C freezer for several years.

    Note: Aliquots should be prepared with reasonable working volumes to avoid repeated freezing and thawing, which can cause small molecules to precipitate out of solution. When removing small molecules from the freezer for use, aliquots should be visually inspected to ensure that small molecules are still in suspension.

  • k.
    Starting from the 10 mM DMH1 stock, dilute it ten-fold in DMSO to generate a 1 mM working stock.

    Note: As described below, DMH1 is used at very low concentrations in artery EC induction media, and therefore this 1 mM working stock provides greater experimental convenience when preparing media.

    Note: This 1 mM DMH1 stock can be stored in the −20°C freezer, but we would recommend freeze-thawing it only several times, as it is a lower concentration stock whose stability is uncertain. From time to time, create a new 1 mM DMH1 working stock from the 10 mM DMH1 master stock.

• 14.
  Resuspend 2-phospho-ascorbic acid (“AA2P”) in sterile H₂O to 100 mg/mL. AA2P is a stabilized form of vitamin C (ascorbic acid).

  • a.
    Bring a sterile beaker to a weighing scale and tare it. Add AA2P powder to the beaker.

    Note: Weighing AA2P can be performed on a non-sterile weighing scale outside of the cell culture hood. Subsequently, AA2P will be sterilized by filtering through a 0.22 μm filter.

  • b.
    Bring the beaker to the cell culture hood and add an appropriate amount of sterile H₂O to resuspend the AA2P powder to 100 mg/mL.
  • c.
    In the cell culture hood, sterilely filter the AA2P solution using a 50-mL-size 0.22 μm filter unit.

    Note: This AA2P solution can be stored in the −20°C freezer for at least 1 year.

    CRITICAL: Resuspend the growth factors and small molecules specifically at the concentrations tabulated in Table 2. When concentrated, growth factors and small molecules tend to be more stable during long-term storage. We have found that storing some of these differentiation factors as more dilute stocks degrades their activity.

    Pause point: Once resuspended, growth factors and small molecules can be stored in the −20°C freezer for at least 1 year.

Table 1.

50 mL of PBS + 0.1% BSA

Reagent	Final concentration	Amount
PBS	1×	50 mL
BSA Fraction V	0.1% v/v	666 μL

Storage: PBS + 0.1% BSA can be stored at 4°C for at least 1 year (v/v, volume by volume).

Table 2.

Resuspension of lyophilized growth factors and small molecules needed for differentiation

Reagent	Stock concentration	Diluent
Activin A	50 μg/mL	PBS + 0.1% BSA
BMP4	50 μg/mL	PBS + 0.1% BSA
FGF2	100 μg/mL	PBS + 0.1% BSA
VEGF-A	100 μg/mL	PBS + 0.1% BSA
Insulin	10 mg/mL	PBS
Thiazovivin	20 mM	DMSO
CHIR99021	10 mM	DMSO
XAV939	10 mM	DMSO
Forskolin	10 mM	DMSO
SB505124	10 mM	DMSO
GDC0941	10 mM	DMSO
RO4929097	10 mM	DMSO
PD0325901	10 mM	DMSO
DMH1	10 mM (∗Subsequently re-dilute and create 1 mM working stock for convenience)	DMSO
2-phospho-ascorbic acid (AA2P)	100 mg/mL	H2O

Preparing human pluripotent stem cell medium

Timing: Around 12 h (to thaw mTeSR Plus supplement overnight)

• 15.To prepare 500 mL of media, thaw the 100 mL bottle of 5× supplement in the 4°C refrigerator overnight.

Note: Optionally, the 5× supplement can be thawed by brief submersion in a 37°C water bath. However, leaving the supplement in warm water for too long may compromise heat-sensitive media components. Check the water bath regularly to assess when the 5× supplement is thawed; do not overheat it. You may want to set a 10-min timer to remind yourself to check the water bath periodically.

Note: Each 500 mL unit of mTeSR Plus comprises two separate components: 400 mL of liquid basal medium (which should be stored in the 4°C refrigerator upon receipt from the manufacturer) and 100 mL of a frozen 5× supplement (which should be stored in the −20°C freezer upon receipt from the manufacturer).

• 16.Add 100 mL of the thawed 5× mTeSR Plus supplement + 5 mL of 100× penicillin/streptomycin into the 400 mL of mTeSR Plus basal medium and mix thoroughly (Table 3).

Note: The 500 mL of fully supplemented mTeSR Plus is ready for use to maintain undifferentiated hPSCs.

Note: This fully supplemented mTeSR Plus medium (mTeSR Plus supplement + mTeSR Plus basal medium + penicillin/streptomycin, hereafter referred to as “mTeSR Plus”) can be stored in the 4°C refrigerator for at least 1 month.

Note: Penicillin/streptomycin is optional but is recommended as a prophylactic against bacterial contamination. We have not found that it impairs hPSC maintenance or differentiation.

CRITICAL: mTeSR Plus medium is thermally sensitive to some extent. Do not leave mTeSR Plus medium in the warm water bath or at room temperature for prolonged amounts of time. A critical component of mTeSR is FGF2,^34,35 which is thermally unstable and has a half-life of ∼4–6 h at 37°C.³⁶ Unnecessarily warming mTeSR could degrade FGF2. According to the manufacturer, mTeSR Plus contains a thermally-stabilized form of FGF2 (https://www.stemcell.com/why-mtesrplus), but we still recommend the same precaution to avoid unnecessarily warming mTeSR Plus medium.

Pause point: mTeSR Plus alone, as well mTeSR Plus + 1 μM Thiazovivin, are stable for at least 1 month in the 4°C refrigerator.

• 17.
  Prepare a 50 mL stock of mTeSR Plus+ 1 μM Thiazovivin (Table 4), which will be used when dissociating and passaging hPSCs as single cells.

  Note: Thiazovivin is a small-molecule ROCK inhibitor, which is critical for cell survival after dissociating and passaging hPSCs as single cells.²⁵

  • a.
    Thaw the 20 mM Thiazovivin stock, and then add 2.5 μL Thiazovivin to 50 mL of mTeSR Plus medium (mTeSR Plus + 1% penicillin/streptomycin, as described above) to prepare 50 mL of mTeSR Plus + 1 μM Thiazovivin medium for future use (Table 4).
  • b.
    Triturate the mTeSR Plus + Thiazovivin stock well to ensure that the Thiazovivin is evenly suspended throughout the media.

    Note: We use Thiazovivin, which is more potent than the widely-used ROCK inhibitor Y-27632. Y-27632 also non-specifically inhibits other kinases besides ROCK, such as PRK2, RSK2 and AMPK.³⁷

Table 3.

500 mL of mTeSR Plus + penicillin/streptomycin (abbreviated “mTeSR Plus” for brevity)

Reagent	Final concentration	Amount
mTeSR Plus basal medium	1×	400 mL
mTeSR Plus, 5× supplement	1×	100 mL
Penicillin/streptomycin	1% v/v	5 mL

Storage: mTeSR Plus can be stored at 4°C for at least 1 month (v/v = volume by volume).

Table 4.

50 mL of mTeSR Plus + 1 μM Thiazovivin

Reagent	Final concentration	Amount
mTeSR Plus + penicillin/streptomycin	1×	50 mL
Thiazovivin, 20 mM stock	1 μM	2.5 μL

Storage: mTeSR Plus + 1 μM Thiazovivin can be stored at 4°C for at least 1 month.

Thawing human pluripotent stem cells

Timing: 1 h

• 18.
  Before thawing hPSCs, prepare 1) a Geltrex-coated plate, 2) a 50 mL conical tube containing DMEM/F12, and 3) a box of dry ice.

  • a.
    Before thawing hPSCs, prepare Geltrex-coated wells to allow for cell seeding as described above in Section 2 (“coating cell culture plates with Geltrex basement membrane”).

    Note: We generally coat 1 well in a 6-well plate with Geltrex to thaw 1 cryovial of hPSCs. However, this is only a general guideline. If a small number of hPSCs were frozen in the cryovial, it might be preferable to thaw them into a smaller size well. The opposite applies to cryovials containing a larger number of frozen hPSCs.

  • b.
    Before thawing hPSCs, pipette 30 mL of DMEM/F12 medium into a 50 mL conical tube. Later, the hPSCs will be thawed into this 50 mL conical tube to dilute out the otherwise-toxic freezing media.
  • c.
    Fill a box with dry ice. This dry ice box will be used to keep hPSCs frozen while transporting the cryovial to the 37°C water bath.
• 19.Take a frozen cryovial of hPSCs from the liquid nitrogen tank, and immediately place the cryovial on dry ice before transporting the dry ice box to the 37°C water bath.

Note: Before thawing the cells, it is critical to keep the cryovial frozen on dry ice so that they are rapidly thawed when transferred to the water bath, instead of slowly thawing on their own at room temperature or on standard ice.

• 20.Take the frozen cryovial out of dry ice and place it in the 37°C water bath to thaw.
• 21.Gently swirl the cryovial within the warm water in a circular motion. Only submerge the bottom part of the cryovial in the water bath.

Note: Ensure that the warm water does not touch the mouth or the cap of the cryovial, from where the cells will be removed.

• 22.
  Every 15–20 s, briefly check the cryovial to check the thawing progress.

  • a.
    When only a small amount of ice remains, take out the cryovial and briskly walk to the cell culture hood—in the intervening time, the remaining small shard of ice will thaw.

Note: First, the water in the water bath is not sterile. Do not let any of the water bath’s warm water contact the upper part of the cryovial, which might lead to microbial contamination. Second, ensure that the water bath is set at 37°C. Excessively warm water baths may compromise cell viability.

• 23.Spray down the cryovial with 70% ethanol to decontaminate it. Dry the cryovial to remove traces of lingering water from the water bath. Then, place the cryovial into the cell culture hood.
• 24.Gently add ∼1 mL of DMEM/F12 to the cryovial, adding the DMEM/F12 drop by drop to avoid shocking the cells. There should be ∼1 mL of frozen hPSCs in freezing medium (Table 5) in the cryovial.
• 25.Then, remove the entire ∼2 mL volume and transfer it into the pre-prepared 50 mL conical tube of 30 mL DMEM/F12 to dilute out the freezing media.
• 26.
  Add 1 mL of DMEM/F12 to the empty cryovial to recover any remaining hPSCs.

  • a.
    Transfer the last remaining mL to the 50 mL conical tube, which contains the hPSCs suspended in DMEM/F12.
• 27.Centrifuge the conical tube at 500 × g for 5 min at 4°C to pellet hPSCs.
• 28.While the hPSCs are being pelleted, prepare the Geltrex-coated wells by aspirating off the Geltrex and adding the appropriate volume of mTeSR Plus + 1 μM Thiazovivin (i.e., 1 mL for one well of a 6-well plate).

CRITICAL: Do not omit the ROCK inhibitor Thiazovivin. It is crucial for the survival of single hPSCs. When dissociated from colonies, single hPSCs experience actin-myosin contractions that lead to membrane blebbing, and consequently, death.^38,39 ROCK inhibition prevents actin-myosin contractions, enabling the survival of single hPSCs.^38,39 However, once hPSCs have re-aggregated to form small colonies, ROCK inhibitor is no longer needed.

• 29.After pelleting hPSCs in the 50 mL conical tube, aspirate or decant the supernatant taking care not to disturb the pellet.

Note: It is acceptable to leave a very small trace of supernatant, as accidentally removing any of the pelleted hPSCs would be inadvisable.

• 30.After removing the supernatant, lightly tap the 50 mL conical tube on the surface of the cell culture hood several times to gently break up the pellet.
• 31.Add 1 mL of fresh mTeSR Plus + 1 μM Thiazovivin to the 50 mL conical tube and triturate several times to resuspend the pellet. Then, add the resuspended hPSCs to the destination well.

CRITICAL: Do not excessively triturate when resuspending the pelleted hPSCs, as this will compromise cell viability, especially during thawing.

• 32.Wash the 50 mL conical tube one more time with 1 mL of fresh mTeSR Plus + 1 μM Thiazovivin and triturate several times to recover any remaining hPSCs.
• 33.Add the resuspended hPSCs to the destination well.
• 34.Gently rock the plate in a cross pattern (left, then right; up, then down) to distribute hPSCs evenly across the well.

Note: Avoid swirling the plate, which can promote cell clumping at the center of the well. Confirm that cells are evenly distributed by checking the well under the microscope.

• 35.Return the cell-culture plate containing the plated hPSCs to the 37°C incubator to allow the hPSCs to adhere and grow.

Note: As a general rule of thumb, while hPSCs are being thawed and are recovering, it is better not to excessively disturb the cells. However, if thawing is successful, it will often be possible to see hPSC clumps adhering to the well within several hours.

• 36.
  The following day, examine hPSCs under the microscope.

  Note: Dead cells in suspension are typical, as freeze-thawing is a stressful process for hPSCs. Based on the degree of hPSC adhesion, add either fresh mTeSR Plus medium (without ROCK inhibitor) or mTeSR Plus + 1 μM Thiazovivin, as described below.

  • a.
    If many hPSC colonies have adhered, aspirate the media, and add fresh mTeSR Plus medium (without ROCK inhibitor).

    Note: ROCK inhibitor is only required for the survival of hPSCs when they have adhered as single cells. After they divide and form colonies, ROCK inhibitor is dispensable for their continued survival.

  • b.
    Continue to maintain hPSCs as described in the below section “Culturing human pluripotent stem cells in an undifferentiated state.”
  • c.
    If very few hPSCs have adhered, add mTeSR Plus + 1 μM Thiazovivin each day.

    CRITICAL: Do not culture thawed hPSCs in mTeSR Plus alone (i.e., without Thiazovivin) until it appears that hPSCs are forming small colonies and are on their way to recovery.

    Note: If there were very few hPSCs in the cryovial, it can take several days for surviving cells to proliferate and reform small colonies.

Table 5.

50 mL of freezing medium

Reagent	Final concentration	Amount
FBS	90% v/v	45 mL
DMSO	10% v/v	5 mL

Storage: freezing medium can be stored at 4°C for 2 weeks.

Culturing human pluripotent stem cells in an undifferentiated state

Timing: 5 min

We culture undifferentiated hPSCs on Geltrex-coated plates in mTeSR Plus medium supplemented with penicillin/streptomycin. However, there are multiple other alternative cell culture media (e.g., mTeSR1, Essential 8 [E8], and others) and basement membrane matrices (e.g., Matrigel, recombinant human vitronectin protein, and others) that can be used to propagate undifferentiated hPSCs.⁴⁰

hPSCs should always be grown using the medium and basement membrane to which they have been adapted. Abrupt switching to a different pluripotent stem cell culture medium and/or basement membrane may be detrimental to hPSC survival and can lead to spontaneous differentiation of hPSCs. As such, switching culture conditions for undifferentiated hPSCs (e.g., between mTeSR vs. E8) should be done gradually and carefully. We strongly recommend banking hPSCs grown in their original culture conditions before switching cells to different culture conditions. Once a frozen bank has been established, cells can be conditioned gradually to a different medium and/or basement membrane.

• 37.Every day, take hPSCs out from the incubator and visually examine them under a phase-contrast microscope for the confluence and visual appearance of hPSCs.

Note: Ensure hPSCs are morphologically uniform with an undifferentiated morphology (Figure 1, left image). Undifferentiated hPSCs form colonies with well-defined borders and cells are morphologically uniform throughout the colony⁴¹ (Figure 1, left image). It is important to look for spontaneously differentiated cells, which have a different morphology from the surrounding hPSCs (Figure 1, middle and right images). Differentiated cells sometimes arise at the colony periphery, and other times at the colony center. Spontaneous differentiation is often evident when hPSCs are overly confluent, and colonies are merging, or alternatively, when the centers of colonies become too dense.

Note: If they arise, spontaneously-differentiated cells in hPSC cultures will often persist and will subsequently confound experiments wherein hPSCs are subjected to differentiation protocols.

Note: Small pockets of morphologically-differentiated cells can be removed by aspirating them. To do so, observe the culture on a phase contrast microscope, and use a marker to label the underside of the plate where morphologically-differentiated cells are located. Then, transfer the plate to a cell culture hood and use an aspirator to gently touch the bottom of the plate at the locations indicated by the marker. After aspirating differentiated cells, add fresh mTeSR Plus medium and observe the plate again under the microscope to ensure that only undifferentiated hPSCs remain. However, it may be preferable to simply discard cultures with morphologically-differentiated cells and to thaw a new stock of undifferentiated hPSCs, as complete elimination of spontaneously-differentiated cells is often arduous or impossible.

• 38.To maintain hPSCs in an undifferentiated state, use a sterile pipette to aspirate media from the cell culture dish at the edge of the well and then add fresh mTeSR Plus. mTeSR Plus medium can be added every day, or alternatively, every other day. Add 1 mL of mTeSR Plus per well of a 12-well plate, 2 mL of mTeSR Plus per well of a 6-well plate, and 10 mL of mTeSR Plus per 10-cm dish. After aspirating the media, do not leave the hPSCs to dry for too long before adding fresh media.

Note: According to the manufacturer’s instructions, mTeSR Plus can be added every other day. However, in some cases (e.g., high cell confluence), it may be preferred to add mTeSR Plus daily. Additionally, it is optional to warm mTeSR Plus to room temperature before adding it to hPSCs. If mTeSR Plus is warmed in a water bath, prolonged heating should be avoided, as it may degrade heat-sensitive components in mTeSR Plus.

• 39.Change media of hPSCs every day, or alternatively, every other day until the cells are ready to be passaged.

Note: When hPSC colonies occupy ∼70% of the total available surface area of the well and/or when colony borders are merging with one another, it is time to passage hPSCs, as described in the below section “passaging human pluripotent stem cells for undifferentiated maintenance”.

CRITICAL: It is extremely important to passage hPSCs in a timely fashion and to prevent them from being overconfluent, which often leads to spontaneous differentiation. If it is borderline whether cells are ready to be passaged on a given day or can be passaged the next day, one should likely split the cells on the current day, as hPSCs can rapidly overgrow, which can lead to spontaneous differentiation.

Figure 1.

Left: A colony of undifferentiated H1 hPSCs

Note the well-defined colony border and the uniform morphology of cells. Also note the absence of spontaneously differentiated cells. Middle and right: H1 hPSCs, with some spontaneously differentiated cells (see arrows), whose morphology is overtly distinct from the undifferentiated cells. Scale bar: 250 μm.

Passaging human pluripotent stem cells for undifferentiated maintenance

Timing: 30 min

We use two separate methods to passage hPSCs. For long-term maintenance of undifferentiated hPSCs, we passage hPSCs as clumps using EDTA,⁴² as described in this section. To seed hPSCs for differentiation, we instead dissociate hPSCs into single cells with Accutase, as described in a subsequent section (“dissociate hPSCs into single cells using Accutase, and seed them for differentiation”). However, repeated dissociation of hPSCs into single cells leads to karyotypic abnormalities,^43,44 presumably because epithelial cells require cell-to-cell contact with neighboring cells to correctly segregate their chromosomes.⁴⁵

Thus, for long-term maintenance of undifferentiated hPSCs, we passage them as clumps using EDTA to preserve a normal karyotype.⁴² EDTA is a Ca²⁺ chelator which inactivates E-cadherin, a cell-surface protein that allows hPSCs to adhere to one another. Brief treatment with EDTA, as described here, loosens hPSC colonies such that they can be easily broken into smaller clumps upon mechanical scraping or trituration. During EDTA-based passaging, no ROCK inhibitor is needed for cell survival because these hPSC clumps can survive on their own.⁴² For this purpose, we employ a commercially-available EDTA solution, known as Versene.

• 40.Before passaging hPSCs, ensure that the recipient plate has already been coated with Geltrex, as described above.
• 41.Aspirate mTeSR Plus from a largely-confluent hPSC culture.
• 42.
  Add Versene and incubate the cells for 7 min at room temperature.

  • a.
    Add a volume of Versene that is one-half the volume of media that would typically be used to grow cells in that format (e.g., add 0.25 mL Versene per well of a 24-well plate, 0.5 mL Versene per well of a 12-well plate, 1 mL Versene per well of a 6-well plate, and 5 mL Versene per 10-cm dish).
  • b.
    Set a timer for 7 min to ensure that cells are not exposed to Versene for an excessive length of time.

CRITICAL: First, do not treat hPSCs with Versene for longer than 7 min. The goal of this 7 min Versene treatment is to loosen hPSCs but not to bring them into suspension. Prolonged EDTA incubation can completely dissociate hPSCs into single cells and is eventually toxic. Second, conduct this 7-min Versene dissociation at room temperature. Higher temperatures, such as 37°C, seem to accelerate cell dissociation and can thus be toxic.

Note: Versene can be stored at either room temperature or at 4°C.

• 43.While hPSCs are being treated with Versene, aspirate Geltrex from the destination plates and add mTeSR Plus medium (without ROCK inhibitor).

Note: Perform this step in parallel while hPSCs are being incubated in Versene, because after Versene dissociation, all steps will proceed rapidly and the destination plates should already be filled with mTeSR Plus before seeding.

Note: As hPSCs are being incubated in Versene at room temperature, observe cells under the microscope. After 7 min of Versene treatment, hPSCs should appear loosened (bright and more rounded) but the majority of hPSCs should still be adhered to the plate. Rare single cells may already appear in suspension.

• 44.
  After 7 min of Versene incubation, completely aspirate the Versene, and add mTeSR Plus medium.

  • a.
    Add a volume of mTeSR Plus that is one-half the volume of media that would typically be used to grow cells in that format (e.g., add 0.25 mL mTeSR Plus per well of a 24-well plate, 0.5 mL mTeSR Plus per well of a 12-well plate, 1 mL mTeSR Plus per well of a 6-well plate, and 5 mL mTeSR Plus per 10-cm dish).
• 45.Gently scrape the hPSCs with a scraper, making sure to scrape all parts of the well.

Note: Upon scraping, hPSCs should immediately detach from the well, which will be visually evident. When holding the scraper, position one’s gloved hands so that they are not hovering above the well, to avoid contamination.

Note: After scraping off cells, one can check the plate under the microscope to ensure that hPSCs have fully detached. However, even without a microscope, it should be visually apparent whether any hPSCs are still attached to the plate, as adhered sheets of hPSCs appear as a static beige film visible to the naked eye.

• 46.After scraping off hPSCs into mTeSR Plus, triturate cells once or twice within the well using a 5 mL serological pipette, and then transfer the mTeSR Plus medium (containing the hPSC clumps) to a 50 mL conical tube.
• 47.Add fresh mTeSR Plus again to the well, triturate the medium once or twice within the well to collect any remaining hPSC clumps, and then transfer the mTeSR Plus medium (containing the hPSC clumps) to the same 50 mL conical tube.
• 48.At this point, the 50 mL conical tube will contain hPSC clumps suspended in mTeSR Plus medium. Add additional mTeSR Plus medium to adjust the final volume of suspended hPSCs as needed for your experiment.
• 49.Triturate the suspension several times to bring all hPSCs evenly into suspension but do not triturate excessively, or else hPSCs may be dissociated into single cells and be killed.

Note: We recommend performing trituration with a 5 mL serological pipette to preserve medium-sized colonies, as trituration with a p1000 pipette can break hPSCs into very small colonies that take longer to grow. However, in some cases, smaller colonies may be desirable; thus, the decision of how much to triturate should be decided based on the specific experiments that are planned.

CRITICAL: Do not over-triturate the hPSC clumps, which will dissociate them into single cells and kill them, because ROCK inhibitor is not used during Versene-based passaging. For instance, triturating more than 10 times would generally be excessive at this stage.

Note: At this step, calculate the amount of mTeSR Plus medium needed to passage hPSCs at the desired ratio. We generally passage hPSCs at a 1:6 ratio, after which they will typically become confluent again in 3–4 days. For example, one can start with 1 well of a 6-well plate containing hPSCs and passage it 1:6. In that case, after removing Versene, one would add 1 mL of mTeSR Plus, and scrape. One would then wash again with 1 mL of mTeSR Plus to collect any remaining hPSC clumps. Consequently, 2 mL of mTeSR Plus medium (containing hPSC clumps) would be transferred to the 50 mL conical tube. One would then add 4 mL of mTeSR Plus medium to bring the total volume to 6 mL (to achieve a 1:6 splitting ratio). Then, this 6 mL solution could be dispensed into 6 wells of a 6-well plate, as described below.

• 50.Gently pipette mix hPSC clumps to ensure they are evenly suspended in mTeSR Plus medium. Then, pipette the mixed hPSCs into recipient Geltrex-coated plates.

Note: As described above, these Geltrex-coated plates should already contain mTeSR Plus medium lacking ROCK inhibitor.

• 51.Immediately after seeding cells, evenly shake the plate in a cross pattern (left, then right; up, then down) several times to make sure clumps are evenly distributed across the plate/well.

Note: Do not swirl the plate in a circular motion, as this will make clumps settle in the center of the plate/well.

• 52.Quickly check under a microscope that cells are evenly distributed throughout the well. If not, continue shaking the plate.

CRITICAL: Immediately shake the plate after adding the cell suspension. hPSCs can begin to adhere to Geltrex rapidly within minutes, and once unevenly plated it is impossible to dislodge them and re-distribute them.

• 53.Return the plate to the 37°C incubator, to allow hPSCs to adhere to the plate.

Note: If stacking multiple plates of hPSCs on top of one another in the incubator, ensure that each plate is flat such that hPSCs will seed evenly across the plate. If unevenly stacked, hPSCs will seed asymmetrically across the plate, thereby forming colonies of uneven size.

• 54.The following morning, use a microscope to observe the plated cells. They should have re-formed hPSC colonies with an undifferentiated morphology, as shown in Figure 2.
• 55.Replace media with fresh mTeSR Plus medium every day, or alternatively, every other day, until colonies grow bigger and are ready to be passaged again (see section “culturing human pluripotent stem cells in an undifferentiated state”).

Figure 2.

H9 hPSCs were dissociated into clumps with Versene (EDTA solution) and seeded 1:8 onto new plates without ROCK inhibitor

24 h later, the hPSCs survived and formed small colonies, as shown in this photograph. Scale bar: 1 mm.

Cryopreservation of human pluripotent stem cells

Timing: 1 h

When cryopreserving undifferentiated hPSCs, we recommend starting from a largely – but not completely — confluent culture of hPSCs. Completely confluent hPSC cultures may contain traces of spontaneous differentiated cells that are difficult to identify morphologically, owing to high confluence. Dissociate hPSCs with EDTA (as described in section “passaging human pluripotent stem cells for undifferentiated maintenance”) and then freeze them in 90% FBS + 10% DMSO, as described below.

Alternatively, serum-free commercial freezing media such as CryoStor CS10 can be used. Before using commercial freezing media, manufacturer recommendations for product-specific recommendations should be consulted.

Generally, one largely confluent 10-cm dish of undifferentiated hPSCs can be dissociated and distributed into 10 cryovials. This will generate sufficient hPSC numbers such that, at a future point, each cryovial can be thawed into 1 well of a 6-well plate.

• 56.
  Before dissociating hPSCs, prepare the requisite reagents necessary for cryopreservation.

  • a.
    Label cryovials with the name of the hPSC line, its passage number, the name of the experimenter freezing the cells, and the date. We recommend labeling cryovials using printed stickers from a label maker. Alternatively, one can also use an alcohol-resistant cryogenic marker.
  • b.
    Prepare freezing medium (90% FBS + 10% DMSO).

Note: We recommend preparing 45 mL FBS + 5 mL DMSO, and then filtering it through a 50 mL volume, 0.22 μm sterile filter unit.

Note: This freezing medium can be stored at 4°C for 2 weeks.

• 57.Aspirate mTeSR Plus from a largely-confluent hPSC culture and add Versene (EDTA solution) for 7 min at room temperature, as described in the above section (“passaging human pluripotent stem cells for undifferentiated maintenance”).

CRITICAL: Do not dissociate hPSCs into single cells using Accutase before cryopreservation. As described elsewhere, continued rounds of single cell dissociation followed by replating can lead to karyotypic abnormalities in hPSCs.

• 58.After 7 min of Versene incubation, completely aspirate the Versene, and add DMEM/F12.
• 59.Use a scraper to scape hPSCs into DMEM/F12.
• 60.After scraping, transfer clumps into a 50 mL conical tube.
• 61.Then, add fresh DMEM/F12 to wash the well to collect any remaining hPSCs, and transfer this fresh DMEM/F12 to the 50 mL conical tube.

Note: Add extra DMEM/F12 to the conical tube so there is at least 20 mL of medium.

Note: At this stage, scrape hPSCs into DMEM/F12 (as opposed to mTeSR Plus). Later, these hPSCs will be centrifuged and the supernatant will be aspirated. Using mTeSR Plus at this stage will waste it because it will be simply aspirated.

• 62.Centrifuge the 50 mL conical tube at 500 × g for 5 min at 4°C to pellet hPSCs.
• 63.While the 50 mL conical tube is being centrifuged, carefully remove the lids of the cryovials in the sterile cell culture hood.
• 64.Carefully aspirate or decant the supernatant, taking care not to disturb the pellet.

Note: It is acceptable to leave a small trace of DMEM/F12.

• 65.Gently tap the 50 mL conical tube on the surface of the cell culture hood several times to loosen the pellet.
• 66.
  Prepare hPSC clump suspension for cryopreservation.

  • a.
    Gently resuspend the pellet in freezing medium and triturate several times to ensure that hPSC clumps are evenly suspended.

    Note: Use a volume of freezing medium such that each final cryovial will receive 1 mL. For instance, when freezing hPSCs in 10 cryovials, resuspend the pellet in 10 mL of freezing medium.

    CRITICAL: Do not over-triturate hPSCs at this stage. Dissociation and cryopreservation are stressful for hPSCs, and over-trituration will kill cells.

  • b.
    Pipette 1 mL of hPSCs suspended in freezing medium into each cryovial.
  • c.
    Once complete, screw on the lid for each cryovial, and transfer the cryovials to a freezing canister.
  • d.
    Transfer the freezing canister to the −80°C freezer for at least 1 day.
  • e.
    After the freezing canister has stayed in the −80 °C freezer for at least 1 day, transfer the cryovials to the liquid nitrogen tank for long-term storage.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-CD144 (VE-CADHERIN) Alexa Fluor 647 antibody (clone 55-7H1)	BD Biosciences	561567 (1:20)
Anti-CD144 (VE-CADHERIN) FITC antibody (clone 55-7H1)	BD Biosciences	560411 (1:20)
Anti-CD144 (VE-CADHERIN) PE antibody (clone 55-7H1)	BD Biosciences	561714 (1:20)
Chemicals, peptides, and recombinant proteins
mTeSR Plus medium	STEMCELL Technologies	100-0276
Essential 8 medium	Gibco	A1517001
Penicillin/streptomycin	Thermo Fisher Scientific	15-140-122
Geltrex LDEV-free, hESC-qualified, reduced growth factor basement membrane matrix	Thermo Fisher Scientific	A1413302
Versene solution (EDTA dissociation buffer)	Thermo Fisher Scientific	15040066
Accutase - enzyme cell detachment medium	Thermo Fisher Scientific	00-4555-56
TrypLE express enzyme (1×)	Thermo Fisher Scientific	12604013
DMEM/F12 + GlutaMAX	Thermo Fisher Scientific	10565042
IMDM + GlutaMAX	Thermo Fisher Scientific	31980-097
Ham’s F12 + GlutaMAX	Thermo Fisher Scientific	31765-092
Polyvinyl alcohol	Sigma	P8136-250G
Chemically defined lipid concentrate	Thermo Fisher Scientific	11905-031
1-thioglycerol (monothioglycerol)	Sigma	M6145-100ML
Recombinant human insulin	Sigma	11376497001
Recombinant human VEGF165	R&D Systems	293-VE-0500
Recombinant human FGF2	R&D Systems	233-FB-01M
Human transferrin	Sigma	10652202001
Recombinant human BMP4	R&D Systems	314-BP-050
Penicillin/streptomycin	Thermo Fisher Scientific	15070-063
Recombinant Activin	R&D Systems	338-AC-500/CF
GDC-0941	Cellagen Technology	C4321-25
Forskolin	Tocris	1099
Thiazovivin	Tocris	3845
XAV939	Tocris	3748
Ascorbic acid-2-phosphate (AA2P)	Sigma	49752-10G
DMH1	Tocris	4126
SB505124	Tocris	3263
RO4929097	Cellagen Technology	C7649-10
CHIR99021	Tocris	4423
PD0325901	Tocris	4192
DMSO (dimethyl sulfoxide)	Sigma	D2650
Critical commercial assays
Stericup 500 mL volume filter, 0.22 μm pore size	Millipore	S2GVU05RE
Steriflip 50 mL volume filter, 0.22 μm pore size	Thermo Fisher Scientific	SE1M179M6
Fisherbrand cell scrapers	Thermo Fisher Scientific	08-100-241
Bovine albumin fraction V (7.5% solution)	Thermo Fisher Scientific	15260037
UltraPure DNase/RNase-free distilled water	Thermo Fisher Scientific	10977023
PBS, pH 7.4	Thermo Fisher Scientific	10010049
Fetal bovine serum (FBS)	R&D Systems (formerly Atlanta Biologicals)	S11550
Cell freezing canister	BioCision	BCS-4050
Cryogenic cell freezing vials	Thermo Fisher Scientific	377267
Experimental models: Cell lines
H1 hESCs	WiCell	WA01
H7 hESCs	WiCell	WA07
H9 hESCs	WiCell	WA09
WTC11 hiPSCs	Coriell Institute for Medical Research	GM25256: Kreitzer et al.46
SUN004.1.9 CAG-GFP hiPSCs	Hiro Nakauchi’s laboratory, Stanford University	Ang et al.1
SUN004.2 CAG-tdTomato hiPSCs	Hiro Nakauchi’s laboratory, Stanford University	Ang et al.1

Materials and equipment

This section describes how to prepare the CDM2 basal medium used for all steps of hPSC differentiation into artery and vein ECs. This CDM2 basal medium is supplemented with specific recombinant growth factor proteins and small molecule chemical compounds to create the respective differentiation media used for each differentiation step.

Prepare CDM2 basal medium for differentiation

Timing: 2 h

Chemically Defined Medium 2 (CDM2) is used as the basal medium for all differentiation steps in this protocol, and its composition has been described previously.^1,47,48,49 CDM2 is itself a variant of the original “chemically defined medium” used for mouse embryonic stem cell differentiation.⁵⁰ Given that the composition of CDM2 is completely defined, one can supplement the media with agonists and antagonists of various signaling pathways and assay the effects of these signaling perturbations on hPSC differentiation in a well-controlled and serum-free culture system.^1,47,48,49

CDM2 consists of 50% IMDM and 50% F12, both of which are nutrient-rich basal media, which are supplemented with concentrated lipids (which serve as a cellular energy source), monothioglycerol (an antioxidant), insulin and transferrin (which support cell proliferation), polyvinyl alcohol (PVA; a synthetic polymer that might stabilize proteins in the cell-culture medium), and finally penicillin/streptomycin (as prophylactic antibiotics) (Table 6).

Table 6.

1000 mL of CDM2 basal medium

Reagent	Final concentration	Amount
IMDM +GlutaMAX +HEPES +Sodium Bicarbonate	50% v/v	500 mL
F12 + GlutaMAX	50% v/v	500 mL
Polyvinyl alcohol	1 mg/mL	1 g
Concentrated lipids	1% v/v	10 mL
Monothioglycerol/1-thioglycerol, 11.5 M stock	450 μM	39.13 μL
Insulin, 10 mg/mL stock	0.7 μg/mL	70 μL
Transferrin, 30 mg/mL stock	15 μg/mL	500 μL
Penicillin/streptomycin	1% v/v	10 mL

Storage: CDM2 basal medium can be stored at 4°C for up to 2 months.

CDM2 contains a lower concentration of insulin (0.7 μg/mL) relative to the original CDM.⁵⁰ We decreased the insulin concentration because the first step of hPSC differentiation into primitive streak is blocked by PI3K signaling,⁵¹ and insulin is known to activate PI3K signaling.

When preparing CDM2, it is important to note that one of its key components—polyvinyl alcohol—must be brought into solution by gentle warming and magnetic stirring.

Pause point: CDM2 basal medium is stable for up to 2 months in the 4°C refrigerator.

• •Bring a 500 mL sterile beaker to a weighing scale and tare it. Add 1 g of polyvinyl alcohol powder to the beaker.

Note: Weighing polyvinyl alcohol can be performed on a non-sterile weighing scale outside of the cell culture hood. Subsequently, CDM2 will be filtered through a 0.22 μm filter to sterilize it.

• •Add 200 mL of IMDM to the beaker containing polyvinyl alcohol. Then, place a sterile magnetic stir bar into the beaker. Place the beaker on a combined hot plate and magnetic stirrer platform, and apply magnetic stirring and gentle warming.

Note: It is important to use magnetic stirring and gentle warming to bring polyvinyl alcohol into suspension. Periodically monitor the solution to ensure that polyvinyl alcohol has been brought into suspension, and that the media does not become overheated and boil.

• •
  Add all CDM2 media components to a sterile, 1-L-size 0.22 μm filter unit in the cell culture hood, and then filter the medium.

  • ○
    After polyvinyl alcohol has been dissolved in 200 mL of IMDM, allow the solution to cool at room temperature before adding it to the filter unit.
  • ○
    Add the remainder of the 300 mL of IMDM and all other CDM2 media components to the filter unit.
  • ○
    The bottle of concentrated lipids may appear cloudy; mix thoroughly with a 10 mL serological pipette before withdrawing 10 mL and adding it to the filter unit.
  • ○
    Once all CDM2 media components have been added to the filter unit, apply the vacuum to filter the medium.

Prepare mid primitive streak differentiation medium (for day 1 of hPSC differentiation)

Timing: 10 min

During gastrulation, the first developmental lineage decision encountered by pluripotent cells is to differentiate into either the primitive streak (which gives rise to endoderm and mesoderm) or ectoderm.⁵² Additionally, there are multiple regions of the primitive streak, including the anterior, middle, or posterior primitive streak.⁵² To differentiate hPSCs into middle primitive streak within 24 h, we simultaneously provide BMP, FGF, TGF-β, and WNT pathway activators for 24 h in the defined, serum-free CDM2 basal medium.^1,49 This in vitro differentiation protocol is inspired by how BMP, FGF, TGF-β, and WNT specify primitive streak in vivo.^{53,54,55,56,57,58,59}

Pause point: Mid primitive streak differentiation medium can be stored in the 4°C refrigerator for up to 3 days.

• •Thaw frozen aliquots of Activin, BMP4, CHIR99021, and FGF2.
• •Aliquot 10 mL of CDM2 basal medium into a 50 mL conical tube.
• •Pipette the indicated amounts of the above differentiation factors into the CDM2 basal medium.

Note: We recommend using differentiation media within 3 days. We have not rigorously tested its stability for more than 3 days in the 4°C refrigerator, nor have we tested whether it can be frozen and thawed.

CRITICAL: Before adding each differentiation factor, ensure that its stock concentration is the same as is listed in Table 7, which in turn corresponds to the stock concentrations listed in Table 2.

Table 7.

10 mL of mid primitive streak differentiation medium (for day 1 of hPSC differentiation)

Reagent	Mechanism	Stock concentration	Final concentration	Amount
Activin A	TGF-β activator	50 μg/mL	30 ng/mL	6 μL
BMP4	BMP activator	50 μg/mL	40 ng/mL	8 μL
CHIR99021	WNT activator	10 mM	6 μM	6 μL
FGF2	FGF activator	100 μg/mL	20 ng/mL	2 μL
CDM2 basal medium (generated as described above)	Basal medium	1×	1×	10 mL

Storage: mid primitive streak differentiation medium can be stored at 4°C for up to 3 days.

Prepare lateral mesoderm differentiation medium (for day 2 of hPSC differentiation)

Timing: 10 min

The second differentiation step entails the transition from primitive streak into the lateral mesoderm. The lateral mesoderm serves as a multipotent progenitor population that gives rise to a diverse array of lineages, including heart, blood, limb, and—of relevance to this protocol—ECs.^24,60 To differentiate hPSC-derived middle primitive streak cells into lateral mesoderm cells within 24 h, we simultaneously provide BMP, VEGF, and cAMP pathway activators, alongside WNT, TGF-β, and PI3K inhibitors, with 2-phospho-ascorbic acid ("AA2P"; a stabilized vitamin C analog) for 24 h in the defined, serum-free CDM2 basal medium.¹

This in vitro differentiation protocol is inspired by multiple discoveries in developmental and stem cell biology. First, activation of the BMP pathway differentiates primitive streak into lateral mesoderm.⁴⁹ Second, simultaneous inhibition of WNT and TGF-β suppresses differentiation of primitive streak into unwanted paraxial mesoderm and endoderm, respectively.^1,48,49 Third, VEGF and cAMP activate the expression of vascular transcription factors, including Etv2.^{61,62,63,64,65} Finally, vitamin C enhances EC differentiation in vitro,^1,10 although the underlying mechanisms are incompletely understood.

Pause point: Lateral mesoderm differentiation medium can be stored in the 4°C refrigerator for up to 3 days.

• •Thaw frozen aliquots of VEGF, XAV939, Forskolin, SB505124, AA2P, GDC0941, and BMP4.
• •Aliquot 10 mL of CDM2 basal medium into a 50 mL conical tube.
• •Pipette the indicated amounts of the above differentiation factors into the CDM2 basal medium.

Note: We recommend using differentiation media within 3 days. We have not rigorously tested its stability for over more than 3 days in the 4°C refrigerator, nor have we tested whether it can be frozen and thawed.

CRITICAL: Do not substitute the listed differentiation factors for other factors. For instance, replacing the TGF-β pathway inhibitor SB505124 with the alternative TGF-β pathway inhibitor A8301 will compromise differentiation. A8301 inhibits both the VEGF and TGF-β receptors,⁶⁶ and VEGF signaling is required for EC differentiation.

CRITICAL: Before adding each differentiation factor, ensure that its stock concentration is the same as is listed in Table 8, which in turn corresponds to the stock concentrations listed in Table 2.

Table 8.

10 mL of lateral mesoderm differentiation medium (for day 2 of hPSC differentiation)

Reagent	Mechanism	Stock concentration	Final concentration	Amount
VEGF	VEGF activator	100 μg/mL	100 ng/mL	10 μL
XAV939	WNT inhibitor	10 mM	1 μM	1 μL
Forskolin	Elevates cAMP	10 mM	10 μM	10 μL
SB505124	TGF-β inhibitor	10 mM	2 μM	2 μL
AA2P	Stabilized vitamin C	100 mg/mL	200 μg/mL	20 μL
GDC0941	PI3K inhibitor	10 mM	2.5 μM	2.5 μL
BMP4	BMP activator	50 μg/mL	40 ng/mL	8 μL
CDM2 basal medium (generated as described above)	Basal medium	1×	1×	10 mL

Storage: lateral mesoderm differentiation medium can be stored at 4°C for up to 3 days.

Prepare artery EC differentiation medium (for day 3 of hPSC differentiation)

At the third differentiation step, lateral mesoderm bifurcates into either artery or vein ECs. To differentiate hPSC-derived lateral mesoderm cells into artery ECs within 24 h, we simultaneously provide VEGF and TGF-β pathway activators, alongside WNT, BMP, and PI3K inhibitors, and 2-phospho-ascorbic acid ("AA2P"; a stabilized vitamin C analog) for 24 h in the defined, serum-free CDM2 basal medium.¹

This in vitro differentiation protocol is inspired by multiple discoveries in developmental and stem cell biology. First, activation of VEGF^62,67 and TGF-β^1,68 specifies artery ECs. Second, inhibition of WNT and BMP suppresses the formation of unwanted non-EC lineages, such as heart progenitors.¹ Third, inhibition of PI3K represses the formation of vein ECs.⁶⁹ Finally, vitamin C enhances EC differentiation in vitro,^1,10 although the underlying mechanisms are incompletely understood.

Pause point: Artery differentiation media can be stored in the 4°C refrigerator for up to 3 days.

CRITICAL: Do not substitute the listed differentiation factors for other factors. For instance, replacing the BMP pathway inhibitor DMH1 with the alternative BMP pathway inhibitor LDN193189 will compromise differentiation. LDN193189 inhibits both the VEGF and TGF-β receptors,⁶⁶ and VEGF signaling is required for EC differentiation.

CRITICAL: Before adding each differentiation factor, ensure that its stock concentration is the same as is listed in Table 9, which in turn corresponds to the stock concentrations listed in Table 2. In particular, the calculations in Table 9 assume that you are starting from a 1 mM working stock of DMH1 (not the 10 mM master stock).

Table 9.

10 mL of artery EC differentiation medium (for day 1 of hPSC differentiation)

Reagent	Mechanism	Stock concentration	Final concentration	Amount
VEGF	VEGF activator	100 μg/mL	100 ng/mL	10 μL
XAV939	WNT inhibitor	10 mM	1 μM	1 μL
DMH1	BMP inhibitor	1 mM	250 nM	2.5 μL
AA2P	Stabilized vitamin C	100 mg/mL	200 μg/mL	20 μL
GDC0941	PI3K inhibitor	10 mM	2.5 μM	2.5 μL
Activin	TGF-β activator	50 μg/mL	15 ng/mL	3 μL
CDM2 basal medium (generated as described above)	Basal medium	1×	1×	10 mL

Storage: artery EC differentiation medium can be stored at 4°C for up to 3 days.

Prepare vein EC differentiation media (for days 3 and 4 of hPSC differentiation)

At the third differentiation step, lateral mesoderm converts into either artery or vein ECs. Differentiation of hPSC-derived lateral mesoderm cells into vein ECs occurs over 48 h, and two different types of differentiation media are employed for the first vs. second 24-h intervals.

In the first 24-h interval, we provide a VEGF pathway activator, alongside WNT, BMP, TGF-β, and NOTCH inhibitors, with 2-phospho-ascorbic acid ("AA2P"; a stabilized vitamin C analog) for 24 h.¹ In the second 24-h interval, we provide a WNT pathway activator, alongside MAPK/ERK, TGF-β, and NOTCH inhibitors, with 2-phospho-ascorbic acid ("AA2P"; a stabilized vitamin C analog) for 24 h.¹ The defined, serum-free CDM2 basal medium is used for both intervals.¹

This in vitro differentiation protocol is inspired by multiple discoveries in developmental and stem cell biology. The first regards the temporally dynamic modulation of the VEGF pathway. Temporally-dynamic activation, followed by inhibition, of the VEGF pathway is required for vein EC development in vivo.^62,67,70 In vivo, VEGF specifies incipient ECs, but subsequently, VEGF must be inhibited to prevent ECs from adopting an arterial fate, and to direct them towards a venous identity.^62,67,70 Accordingly, we found that VEGF activation for 24 h, followed by MAPK/ERK pathway inhibition for 24 h, was crucial for vein EC differentiation in vitro.¹ Second, at the first 24-h interval, we inhibit WNT and BMP to suppress the formation of unwanted non-EC lineages, such as heart progenitors.¹ Third, for the entire 48-h period, we inhibit the artery-specifying NOTCH and TGF-β pathways to suppress artery EC formation and to consolidate venous identity.^1,71 Finally, vitamin C enhances EC differentiation in vitro,^1,10 although the underlying mechanisms are incompletely understood.

Pause point: Vein differentiation media can be stored in the 4°C refrigerator for up to 3 days.

CRITICAL: Do not substitute the listed differentiation factors for other factors. For instance, replacing the BMP pathway inhibitor DMH1 with the alternative BMP pathway inhibitor LDN193189 will compromise differentiation. LDN193189 inhibits both the VEGF and TGF-β receptors,⁶⁶ and VEGF signaling is required for EC differentiation.

CRITICAL: Before adding each differentiation factor, ensure that its stock concentration is the same as is listed in Table 10, Table 11, which in turn corresponds to the stock concentrations listed in Table 2. In particular, the calculations in Table 10 assume you are starting from a 1 mM working stock of DMH1 (not the 10 mM master stock).

Table 10.

10 mL of vein EC differentiation medium A (for day 3 of hPSC differentiation)

Reagent	Mechanism	Stock concentration	Final concentration	Amount
VEGF	VEGF activator	100 μg/mL	100 ng/mL	10 μL
XAV939	WNT inhibitor	10 mM	1 μM	1 μL
DMH1	BMP inhibitor	1 mM	250 nM	2.5 μL
AA2P	Stabilized vitamin C	100 mg/mL	200 μg/mL	20 μL
SB505124	TGF-β inhibitor	10 mM	2 μM	2 μL
RO4929097	NOTCH inhibitor	50 μg/mL	2 μM	2 μL
CDM2 basal medium (generated as described above)	Basal medium	1×	1×	10 mL

Storage: vein EC differentiation medium A can be stored at 4°C for up to 3 days.

Table 11.

10 mL of vein EC differentiation medium B (for day 4 of hPSC differentiation)

Reagent	Mechanism	Stock concentration	Final concentration	Amount
CHIR99021	WNT activator	10 mM	1 μM	1 μL
PD0325901	MAPK/ERK inhibitor	10 mM	500 nM	0.5 μL
AA2P	Stabilized vitamin C	100 mg/mL	200 μg/mL	20 μL
SB505124	TGF-β inhibitor	10 mM	2 μM	2 μL
RO4929097	NOTCH inhibitor	10 mM	2 μM	2 μL
CDM2 basal medium (generated as described above)	Basal medium	1×	1×	10 mL

Storage: vein EC differentiation medium B can be stored at 4°C for up to 3 days.

Step-by-step method details

Dissociate hPSCs into single cells using Accutase, and seed them for differentiation

Timing: 1 h

When seeding hPSCs for differentiation, we dissociate them into single cells using Accutase and plate them onto Geltrex-coated plates. As described above, when dissociated into single cells, hPSCs must be plated in mTeSR Plus supplemented with a ROCK inhibitor (Thiazovivin)²⁵ to survive (Troubleshooting 1).

Seeding cells sparsely is of paramount importance for efficient differentiation. Certain developmental signals (e.g., BMP) do not act evenly throughout a large hPSC colony, but rather only activate signaling in peripheral cells.^72,73 This has been attributed to the localization of BMP receptor to the basolateral edges of an hPSC,⁷³ such that a hPSC colony will only receive signals emanating from the “side” (as opposed from the culture medium “above” it). Seeding hPSCs sparsely at the start of differentiation enables more uniform cellular responses to these developmental signals. However, with regard to seeding density, there is a balance: seeding hPSCs overly sparsely will lead to complete cell death during differentiation.

Seeding hPSCs at the correct density is extremely important to subsequently achieve highly efficient differentiation of hPSCs into artery and vein ECs. We suggest that, for the first experiment, a new experimenter should test a wide range of cell seeding densities in a 12-well plate. For mTeSR Plus-grown hPSCs, we recommend seeding 25,000–50,000 hPSCs/cm² (i.e., ∼95,000–190,000 hPSCs/well of a 12-well plate). An experimenter can also try to seed cells above and below this recommended range to gauge the effects of under- and over-confluence.

We also recommend that one seeds at least two extra wells of hPSCs than is needed for differentiation. These “control wells” can be maintained in mTeSR Plus for several days while other parallel wells in the same plate are being differentiated.

A step-by-step protocol for dissociating hPSCs with Accutase is provided here.

• 1.Aspirate mTeSR Plus from a largely-confluent hPSC culture.
• 2.
  Add Accutase to dissociate the cells.

  • a.
    Add half the volume of Accutase that you would typically add of cell culture media suited for that type of cell culture well or dish.

Note: For instance, add at least 0.5 mL of Accutase per well of a 12-well plate, 1 mL of Accutase per well of a 6-well plate or 4 mL per 10-cm dish to fully cover the surface.

• 3.
  Incubate the hPSCs at 37°C for ∼5 min to enable Accutase to dissociate the cultures.

  • a.
    Continue observing the well/plate during Accutase incubation.

    Note: It should be visually obvious when cell clusters or sheets begin detaching from the bottom of the well/plate, even without the aid of microscopy.

  • b.
    After several minutes or after some colonies begin detaching (whichever happens first), gently tap the bottom of the well/plate several times.

    Note: Most hPSC colonies should freely come into suspension. If not all colonies detach, wait several more minutes, after which most colonies should detach.

• 4.
  Add DMEM/F12 to the plate to wash off colonies.

  Note: For instance, for 1 well in a 6-well plate to which 1 mL of Accutase was added, add 1 mL of DMEM/F12 to rinse off the cells and transfer the resuspended cells to a 50 mL conical tube. Triturate once or twice to ensure that all hPSCs have entered suspension.

  • a.
    Add another 1 mL of DMEM/F12 to rinse off any remaining cells, then add that 1 mL to the same 50 mL conical tube.
• 5.To the 50 mL conical tube, add sufficient DMEM/F12 to dilute the original volume of Accutase by at least 1:5.
• 6.Afterwards, centrifuge at 500 × g for 5 min at 4°C to pellet hPSCs.
• 7.While the dissociated hPSCs are pelleting, in the meantime, prepare recipient Geltrex-coated cell culture plates to receive hPSCs.

Note: As described above, ensure that cell culture plates have been coated with Geltrex for a minimum of 30 min.

• 8.Afterwards, aspirate away the Geltrex; a very thin translucent film will remain adsorbed onto the plastic.
• 9.Then, add half a volume of mTeSR Plus + 1 μM Thiazovivin stock to the well or dish.

Note: For instance, when about to seed one well in a 6-well plate with hPSCs, add 1 mL of mTeSR Plus + 1 μM Thiazovivin and when about to seed a 10-cm dish with hPSCs, add 5 mL of mTeSR Plus+ 1 μM Thiazovivin.

CRITICAL: Thiazovivin is critical for the survival of hPSCs that have been dissociated into single cells. Do not omit Thiazovivin.

• 10.After centrifugation is complete, carefully retrieve the 50 mL conical tube from the centrifuge and briefly hold it up to the light to see if a cell pellet is visible.

Note: When passaging small numbers of cells, it may not be possible to visualize the pellet.

• 11.Carefully aspirate or decant the supernatant, taking care not to disturb the cell pellet.
• 12.Once the supernatant has been aspirated, gently flick the bottom of the 50 mL conical tube to slightly dislodge the cell pellet.
• 13.Resuspend the cell pellet in mTeSR Plus + 1 μM Thiazovivin, preferably using a 5 mL or a 10 mL pipette tip.

Note: It should be visually apparent when the dissociated cells are resuspended, as the mTeSR Plus should become ‘cloudy’ with cells. After triturating, no extremely large clumps of cells should be visibly anymore. It is important at this stage to evenly resuspend hPSCs in mTeSR Plus, to ensure an accurate cell count (see below step) and to ensure that equal numbers of hPSCs are seeded per recipient well.

CRITICAL: Do not over-triturate hPSCs when resuspending them, as this can kill them.

• 14.After resuspending hPSCs in mTeSR Plus + 1 μM Thiazovivin, take an aliquot of cells to count on a hemocytometer or cell counter.

Note: Generally speaking, a largely-confluent well of hPSCs in a 6-well plate yields ∼2–4 million cells.

• 15.After cell counting is complete, adjust the volume of mTeSR Plus to seed hPSCs at the desired density.

Note: It is imperative to ensure that hPSCs are evenly resuspended and that equal numbers of hPSCs are seeded per recipient well.

Note: For instance, if an experimenter wishes to seed 150,000 hPSCs per well of a 12-well plate, one should resuspend hPSCs at a density of 300,000 cells/mL. Subsequently, one would seed 0.5 mL of cell suspension per well of a 12-well plate; each well would already contain 0.5 mL of mTeSR Plus + 1 μM Thiazovivin (as described above). In this scenario, each final well of a 12-well plate would contain 150,000 cells in 1 mL of media.

• 16.
  Pipette the hPSC suspension from the 50 mL conical tube into recipient wells: dispense 0.5 mL per well of a 12-well plate; 1 mL per well of a 6-well plate; or 5 mL per 10-cm dish.

  Note: The recipient dishes should already have been pre-coated with an equal volume of mTeSR Plus + 1 μM Thiazovivin.

  • a.
    When adding the cell suspension to recipient dishes, do so briskly.

    Note: If pipetted too slowly, the cells will very slowly clump at the bottom of the 5 mL or 10 mL pipette tip, leading to different wells having different numbers of cells.

  • b.
    Before dispensing cells into recipient plates, ensure that recipient wells already contain enough mTeSR Plus + 1 μM Thiazovivin (as described above).

    Note: It is important to ensure that recipient wells already contain some media. Otherwise, adding cell suspension directly to an empty well or plate can lead to uneven cell plating.

• 17.Immediately after seeding cells, evenly shake the plate in a cross pattern (left, then right; up, then down) several times to ensure clumps are evenly distributed across the plate/well.

Note: Do not swirl the plate in a circular motion, as this will make clumps settle in the center of the plate/well. Quickly check under a microscope that cells are evenly distributed throughout the well. If not, continue shaking the plate (Troubleshooting 2).

CRITICAL: To ensure even cell seeding, immediately shake the plate after adding the cell suspension. hPSCs will begin to adhere to Geltrex rapidly. Once unevenly plated it is impossible to dislodge cells and re-distribute them.

• 18.Return the cells to the 37°C incubator.

Note: If stacking multiple plates of hPSCs on top of one another in the incubator, ensure that each plate or dish is flat so that hPSCs will seed evenly across the plate. If unevenly stacked, hPSCs will seed asymmetrically across the plate, thereby forming colonies of uneven size.

• 19.The following morning, use a microscope to observe the plated cells.

Note: There may be some degree of cell death, but in the presence of 1 μM Thiazovivin, there should be substantial numbers of surviving cells. If passaged properly, within 12–24 h post-seeding, hPSCs should have reformed small clumps or even branched “webs” of cells, depending on their density (Figures 3 and 4A). These hPSCs are ready for differentiation, as described below.

Figure 3.

H1 hPSCs were dissociated into single cells with Accutase and seeded onto a new 12-well plate at a density of 150,000 cells/well in the presence of mTeSR Plus + 1 μM Thiazovivin (ROCK inhibitor)

hPSCs cultured in mTeSR Plus form colonies with rounded borders. However, after plating in the presence of Thiazovivin, hPSCs can form webs, or alternatively, colonies with jagged borders. ROCK inhibitors reversibly disrupt hPSC morphology, owing to how ROCK regulates the actin-myosin cytoskeleton.^38,39 However, withdrawing ROCK inhibitor leads to a rapid re-normalization of hPSC colony morphology (not shown). Scale bar: 250 μm.

Figure 4.

Morphology of H1 hPSCs being sequentially differentiated into artery or vein ECs

At each stage of differentiation, cultured cells undergo morphological changes to adopt shapes reminiscent of their morphology within the embryo, thus allowing one to visually track differentiation progress.

(A) Photograph of undifferentiated hPSCs, which were dissociated with Accutase into single cells and then seeded in mTeSR Plus + Thiazovivin for 24 h. Note that colonies comprise tightly-packed pluripotent cells but the colonies feature jagged borders, as expected given ROCK inhibition.

(B) Primitive streak cells are larger, and are less densely packed together, consistent with how pluripotent cells undergo an epithelial-to-mesenchymal transition in the primitive streak in vivo.

(C) Lateral mesoderm cells are usually triangular and more dispersed from one another.

(D–F) ECs harbor characteristic elongated or cobblestone endothelial morphology. Non-ECs within the culture are larger, mesenchyme-like in morphology, and sometimes appear bright and raised, as evident in the vein EC photograph.

(G) Illustration of differentiation strategy. Scale bar: 400 μm.

Differentiate hPSCs into artery and vein ECs

Timing: 3 days (for artery ECs) or 4 days (for vein ECs)

After dissociating hPSCs into single cells using Accutase and seeding them in mTeSR Plus + Thiazovivin, as described above (section “dissociate hPSCs into single cells using Accutase, and seed them for differentiation”), sequentially differentiate them into artery or vein ECs (Figure 4). This entails treatment of hPSCs with mid primitive streak induction medium (for 24 h) to differentiate them into day 1 mid primitive streak cells, which are then treated with lateral mesoderm induction medium (for 24 h) to differentiate them into day 2 lateral mesoderm cells. Subsequently, the differentiation paths bifurcate. hPSC-derived lateral mesoderm cells can be treated with artery EC induction medium (for 24 h) to yield day 3 artery ECs. Alternatively, lateral mesoderm cells can be treated with vein EC induction medium A (for 24 h), followed by vein EC induction medium B (for 24 h), to yield day 4 vein ECs (Troubleshooting 3).

• 20.Warm day 1 mid primitive streak differentiation medium (Table 7) and DMEM/F12 medium to room temperature.
• 21.After allowing Accutase-dissociated hPSCs to plate for ∼24 h (as described in the above section “dissociate hPSCs into single cells using Accutase, and seed them for differentiation”), aspirate mTeSR Plus + Thiazovivin.
• 22.Briefly wash cells with DMEM/F12 to dilute any traces of the prior medium: add DMEM/F12, gently swirl the plate, and then immediately aspirate it.
• 23.Add mid primitive streak induction medium for 24 h.

Note: After 24 h of mid primitive streak differentiation, there should be a dramatic morphological change. While undifferentiated hPSCs are small epithelial cells that form colonies with bright borders, primitive streak differentiation is accompanied by an epithelial-to-mesenchymal transition. This is analogous to the epithelial-to-mesenchymal transition that accompanies the differentiation of pluripotent epiblast cells into primitive streak cells in vivo.⁷⁴In vitro, individual primitive streak cells are larger, less tightly packed, and may begin to spread out from the colony border (Figure 4B).

Note: Some degree of cell death is expected during primitive streak differentiation. However, if there are no surviving cells, this may indicate that hPSCs were seeded too sparsely (Troubleshooting 3).

• 24.Warm day 2 lateral mesoderm differentiation medium (Table 8) and DMEM/F12 medium to room temperature.
• 25.Briefly wash cells with DMEM/F12 to dilute any traces of the prior medium: add DMEM/F12, gently swirl the plate, and then immediately aspirate it.
• 26.Add lateral mesoderm induction medium for 24 h.

Note: After 24 h of lateral mesoderm differentiation, most cells will morphologically appear triangular and they will migrate away from their colonies, becoming more dispersed (Figure 4C).

Note: Some degree of cell death is expected during primitive streak differentiation. However, if there are no surviving cells, this may indicate that hPSCs were seeded too sparsely (Troubleshooting 3).

• 27.Warm day 3 artery EC differentiation medium (Table 9), day 3 vein EC differentiation medium A (Table 10), and DMEM/F12 medium to room temperature.
• 28.Briefly wash cells with DMEM/F12 to dilute any traces of the prior medium: add DMEM/F12, and then immediately aspirate it.
• 29.
  Add day 3 differentiation media for 24 h.

  • a.
    For artery EC differentiation, add artery EC differentiation medium for 24 h.
  • b.
    For vein EC differentiation, add vein EC differentiation medium A for 24 h.

Note: After 24 h of differentiation, most cells will morphologically appear like ECs, with distinctive elongated morphology: representative photos of artery ECs are shown in Figure 4D and photos of cells undergoing the first stage of vein differentiation are shown in Figure 4E. However, EC differentiation usually is not 100% efficient. Larger cells with mesenchyme-like morphology can also appear, whose exact identity is unknown; however, they express mesenchymal genes, as assessed by single-cell RNA sequencing.¹

Note: At this stage, artery ECs have completed differentiation (3 days of hPSC differentiation). They are ready to be harvested for downstream assays, such as flow cytometry, immunostaining, qPCR, or RNA sequencing.

Note: However, as noted above, lateral mesoderm cells require 48 h of differentiation to be converted into vein ECs, entailing 24-h treatment with vein EC differentiation medium A, followed by 24-h treatment with vein EC differentiation medium B (Figure 4G). The continuation of vein EC differentiation is described in the next step.

• 30.Warm day 4 vein EC differentiation medium B (Table 11) and DMEM/F12 medium to room temperature.
• 31.Briefly wash cells with DMEM/F12 to dilute any traces of the prior medium: add DMEM/F12, and then immediately aspirate it.
• 32.Add vein EC differentiation medium B for 24 h.

Note: At this stage, vein ECs have completed differentiation (4 days of hPSC differentiation). They are ready to be harvested for downstream assays, such as flow cytometry, immunostaining, qPCR, or RNA sequencing. As shown in Figure 4F, day 4 vein ECs have an endothelial cobblestone-like morphology. As noted above, EC differentiation often is not 100% efficient and one can often observe cells with large, mesenchyme-like morphology coexisting alongside ECs (Figure 4F).

Expected outcomes

Differentiation purity

This differentiation protocol typically generates CD144⁺ artery ECs with 90%–100% purity, and CD144⁺ vein ECs with 70%–100% purity (Figure 5) (Troubleshooting 4). We quantify differentiation efficiency using flow cytometry to assess expression of CD144 (VE-CADHERIN), which is an endothelial cell-surface marker expressed on both artery and vein ECs.¹ Our flow cytometry protocol is described elsewhere¹; in brief, it entails dissociating ECs with TrypLE Express and staining them for 15–45 min at 4°C with an anti-CD144 antibody conjugated to either Alexa Fluor 647, FITC, or PE, followed by analysis on a flow cytometer. Samples should be kept cold and protected from light during staining. Flow cytometric analysis of CD144 expression assesses the purity of differentiated ECs, but it does not measure their arteriovenous identity. Notably, the purity of vein ECs tends to be lower than that of artery ECs (Figure 5), as further described in the “limitations” section.

Figure 5.

Flow cytometry of multiple hPSC lines differentiated into artery or vein ECs using the aforementioned protocol

H1, H7, and H9 hESCs were grown in mTeSR Plus medium, whereas WTC11 hiPSCs were grown in E8 medium. As negative controls, undifferentiated hPSCs were analyzed.

Arteriovenous marker expression

Immunostaining, quantitative PCR (qPCR), and RNA sequencing can be used to assess the expression of arteriovenous markers in hPSC-derived artery and vein ECs; these methods are described elsewhere.¹ For instance, qPCR of artery and vein ECs generated from three independent lines (H1 hESCs, SUN004.1.9 hiPSCs, and SUN004.2 hiPSCs) reveals remarkably consistent patterns of arteriovenous marker gene expression (Figure 6). While both artery and vein ECs express pan-EC markers (e.g., CD144 and CD31), artery ECs predominately express arterial markers (e.g., SOX17, CXCR4, and DLL4), whereas vein ECs exclusively express venous markers (NR2F2, APLNR, NT5E, and FLRT2) (Figure 6).

Figure 6.

qPCR analysis of mRNA expression in artery and vein ECs generated from three independent hPSC lines (H1 hESCs, SUN004.1.9 hiPSCs, and SUN004.2 hiPSCs)

As a negative control, undifferentiated hPSCs were analyzed. There are modest variations in the expression levels of individual genes across these three hPSC lines, perhaps reflecting inter-individual human genetic variation or minor variability across hPSC lines. In this experiment, hPSC-derived day 3 artery ECs and day 4 vein ECs were expanded for 3–6 days in their respective expansion media, as described elsewhere.¹ However, similar results can be obtained upon qPCR analysis of hPSC-derived day 3 artery ECs and day 4 vein ECs that were directly analyzed without further in vitro expansion. Data are represented as mean ± SEM.

Limitations

This protocol to differentiate hPSCs into artery and vein ECs has been robustly reproduced by multiple laboratories across a range of hPSC lines (Figures 5 and 6). However, the following limitations should be considered.

Presence of non-endothelial cells

This differentiation protocol generates ECs with high efficiency, but nevertheless some non-ECs arise contemporaneously. For instance, in the accompanying manuscript,¹ we demonstrated that this differentiation protocol generated on average 97.9 ± 2.4% pure artery ECs across 4 hPSC lines (H1, H7, H9, and SUN004.1.9). However, the remaining cells in the culture are non-ECs, which are reminiscent of mesenchyme-like cells but whose precise identity remains to be determined.¹ As aforementioned, these mesenchyme-like cells are larger and are thus morphologically apparent; Figure 4F shows that vein ECs with characteristic endothelial morphology co-exist alongside larger mesenchymal cells. This limitation underscores that we do not fully understand the signals that control the bifurcation of ECs vs. mesenchyme-like cells during mesoderm differentiation.

For applications requiring homogeneous ECs, for instance bulk-population transcriptional or chromatin assays, we recommend purifying ECs to eradicate non-ECs. One option is fluorescence-activated cell sorting (FACS), but this can be time consuming to scale to large numbers of cells.

Another option to purify hPSC-derived ECs to homogeneity is magnetically-activated cell sorting (MACS). An advantage of MACS is that it can be scaled to tens or hundreds of millions of hPSC-derived ECs, thus enabling en masse cell production and purification. The accompanying manuscript also describes using anti-CD34 magnetic microbeads to purify ECs to homogeneity via MACS.¹ CD34 is a pan-endothelial cell-surface marker expressed by both hPSC-derived artery and vein ECs.¹ To assess MACS efficacy, we typically take a small aliquot of cells to perform flow cytometry analysis of CD144 expression both pre- and post-MACS. In our general experience, CD34 MACS enrichment is accompanied by a ∼50% cell loss: for instance, if vein EC differentiation yields 10 million cells, of which only 80% are CD144+ ECs (as assessed by flow cytometry, i.e., there are 8 million ECs), after MACS there will be ∼4 million ECs remaining but they will be ∼100% CD144+. In the accompanying manuscript, we provide multiple examples of hPSC-derived EC purification via MACS.¹

Vein EC differentiation is less efficient, and slower, than artery EC differentiation

In this protocol, vein EC differentiation takes one day longer than artery EC differentiation (4 vs. 3 days, respectively; Figure 4).¹ Generally, vein EC differentiation tends to be less efficient (generating 70%–100% pure CD144⁺ vein ECs) relative to artery EC differentiation (which often yields CD144⁺ artery ECs with 90%–100% purity) (Figure 5). At present, we do not fully understand why vein EC differentiation is less efficient and more prolonged than artery ECs. This may relate to unresolved mysteries surrounding venous development²⁴ that must be resolved in vivo to devise more effective in vitro differentiation protocols. Further optimization of developmental signaling perturbations, or physical parameters including cell density, may eventually improve the efficiency and rapidity of vein EC differentiation.

EC differentiation is only efficient across a specific range of cell seeding densities

Seeding cells at the correct density is paramount to achieving efficient differentiation into artery and vein ECs. To obtain highly pure artery and vein ECs, optimization of initial (day 0) hPSC seeding density should be optimized by each individual experimenter, especially for each new hPSC line (Figure 7).

Figure 7.

We recommend that the experimenter test a range of hPSC seeding densities on day 0 of differentiation

Results from three independent experiments are shown here.

(A and B) Examples of appropriate cell seeding density.

(C) Cells seeded slightly too densely. Primitive streak cells look uniform, without an overt pocket of undifferentiated hPSCs in the colony center. However, lateral mesoderm cells and artery ECs are quite dense. Interspersed among the vein ECs are raised mesenchymal cells (i.e., non-ECs) (see white arrows). Scale bar: 400 μm.

As aforementioned, for mTeSR Plus-grown hPSCs, we strongly recommend testing a range of seeding densities between 25,000–50,000 hPSCs/cm² (i.e., ∼95,000–190,000 hPSCs/well of a 12-well plate) on day 0 of differentiation (Figure 7). If hPSCs are seeded too sparsely, extensive cell death will occur during differentiation. If hPSCs are seeded too densely, differentiation density will decrease because some hPSCs will fail to differentiate in the center of excessively large colonies. This is attributed to the fact that some developmental signals (e.g., BMP) do not act evenly throughout a hPSC colony, but rather only act on peripheral cells,^72,73 as described above.

Different hPSC lines may grow at different rates, underscoring the importance of testing a range of cell seeding densities at day 0 of differentiation. While this protocol describes the use of mTeSR Plus-propagated hPSCs, we have empirically found that hPSCs grown in E8 medium⁴⁰ tend to proliferate more rapidly. We have successfully differentiated E8-grown WTC11 hPSCs into ECs with high efficiency (Figure 5), but they must be seeded even more sparsely at day 0 of differentiation. mTeSR1 and E8 are similar and both contain TGF-β1 (0.6 ng/mL) and FGF2 (100 ng/mL); however E8 is a simplified version of mTeSR1 without certain extraneous additives (e.g., pipecolic acid, GABA, LiCl, albumin or β-mercaptoethanol).⁴⁰ GABA, a component of mTeSR1,^34,35 may slow hPSC growth,⁷⁵ perhaps contributing to faster hPSC proliferation in E8. For E8-grown WTC11 hPSCs, on day 0 of differentiation, we seed 50,000 and 75,000 cells per well of a 12-well plate for artery vs. vein EC differentiation, respectively. This is significantly sparser than what we use for mTeSR Plus-grown hPSCs (95,000–190,000 cells per well of a 12-well plate).

Additionally, one should test multiple cell seeding densities when switching between different plate formats (e.g., 6-well plate versus 10-cm dish). For instance, when transitioning to 6-well plates, we screened a range of cell seeding densities, and found that seeding 400,000–600,000 hPSCs per well of a 6-well plate on day 0 of differentiation (∼40,000–60,000 hPSCs/cm²) was optimal for mTeSR Plus-grown H1 hPSCs. In our experience, the optimal densities for different plate formats do not always scale perfectly with the expected surface area of a well/dish described by the manufacturer.

Considerations regarding the choice of hPSC line and methods for hPSC maintenance

In the authors’ collective experience, all wild-type hPSC lines tested thus far — including hESC and hiPSC lines — have efficiently differentiated into artery and vein ECs, as shown in Figures 5 and 6 and our accompanying manuscript.¹ Our differentiation method is also effective on hPSC lines bearing constitutively expressed fluorescent proteins, or carrying fluorescent proteins knocked into various endogenous genes of interest, as described in our accompanying manuscript.¹

While we have not encountered this scenario thus far, it is formally possible that certain hPSC lines will differentiate less efficiently. In this case, we recommend that the experimenter test a range of cell seeding densities on day 0 of differentiation, as described in the below “Troubleshooting” section. Additionally, hPSC lines bearing mutations in genes required for endothelial development will also be expected to differentiate less efficiently or not at all.

As aforementioned, this protocol employs hPSCs grown in mTeSR Plus medium on Geltrex-coated plates. However, there are multiple commercially available media (e.g., E8) and basement membrane matrices (e.g., vitronectin or laminin-511) available to grow undifferentiated hPSCs.⁴⁰ For hPSC lines cultured in other conditions, we strongly recommend adapting them to mTeSR Plus and Geltrex for several passages before commencing differentiation as described in this protocol.

Similarity to artery and vein ECs in vivo

Cells in vivo and in vitro are not identical, and this proviso must be taken under consideration when planning to use hPSC-derived artery and vein ECs to model vascular biology or disease in vitro. hPSC-derived artery and vein ECs express a suite of artery- or vein-specific markers, as described in Figure 6 and our accompanying manuscript,¹ but they are not identical to human artery and vein ECs in vivo. hPSC-derived artery and vein ECs respectively show 74% and 83% transcriptional similarity to their in vivo counterparts within the 4-week-old Carnegie Stage 13 human fetus.^1,76 Likewise, multiple papers have demonstrated that ECs freshly isolated from human beings change their gene expression upon culture.^77,78 The top genes that distinguish hPSC-derived artery and vein ECs in vitro vs. their in vivo counterparts largely relate to protein translation and metabolism.¹ We do not fully understand the mechanistic bases for these in vivo vs. in vitro transcriptional differences. However, differing oxygen levels, nutrient availability, extracellular matrix, and mechanical forces²⁴ encountered by ECs in vivo and in vitro likely contribute to these divergences.

Long-term maintenance of differentiated hPSC-derived artery and vein ECs

This protocol focuses on the differentiation of hPSC-derived day 3 artery ECs and day 4 vein ECs. In the accompanying manuscript,¹ we describe how to cryopreserve these artery and vein ECs, after which they can be thawed and maintained for up to 6 days while retaining respective expression of artery- or vein-specific markers. However, we have not extensively tested whether these hPSC-derived artery and vein ECs can be reiteratively passaged and expanded in vitro in long-term cultures.

Troubleshooting

Problem 1: Poor survival of hPSCs after dissociation and passaging

Ample numbers of undifferentiated hPSCs should survive after dissociation, either as clumps (EDTA/Versene-based maintenance of undifferentiated hPSCs) or as single cells (Accutase-based dissociation of hPSCs into single cells in preparation for differentiation, Figure 8A). While some degree of cell death is expected after dissociation and passaging, abnormally poor cell survival should be addressed (Figures 8B and 8C). If hPSCs are exhibiting poor cell survival in the undifferentiated state, this is an indication of a problem and will be significantly exacerbated upon differentiation (Figure 8D).

Figure 8.

H1 hPSCs were dissociated with Accutase and plated for 24 h in mTeSR Plus medium + Thiazovivin

(A) Example of appropriate cell seeding density and survival. One can proceed to add mid primitive streak differentiation medium.

(B and C) Examples of excessive cell death or sparse hPSC seeding density.

(D) If hPSCs seed too sparsely on day 0 of differentiation, the addition of primitive streak differentiation medium will lead to severe cell death. Scale bar: 1 mm.

Potential solution

• •Shorter dissociation times. Prolonged exposure to either Versene or Accutase is toxic to hPSCs. Use the times recommended above. In particular, ensure that hPSCs do not exceed 7 min in Versene. Set a timer to remind oneself when to terminate dissociation.
• •Testing for mycoplasma. Mycoplasma contamination can cause extreme cell death. Ensure that undifferentiated hPSCs are mycoplasma negative. Multiple commercially available mycoplasma test kits are available, but we recommend the MycoStrip 100 because it entails a simple test strip without having to run an agarose gel.
• •Inclusion of ROCK inhibitor (Thiazovivin) to permit survival after Accutase-based passaging. Ordinarily, hPSCs do not survive single cell dissociation.^38,39 After dissociating hPSCs into single cells using Accutase, inhibition of ROCK is critical for the survival of singularized hPSCs.^38,39 Remember to always seed freshly-thawed hPSCs, or Accutase-dissociated hPSCs, in Thiazovivin-containing medium.
• •Avoid excessive trituration after Accutase-based passaging. After dissociating hPSCs with Accutase, the experimenter must resuspend single cells and count them before seeding them for differentiation. During this process, excessive trituration will lead to excessive cell death. Resuspend cells with a 5 or 10 mL pipette tip, which exerts less force than a p1000 pipette tip. Count the number of times that one triturates the cell suspension, and avoid excessive trituration.

Problem 2: Uneven hPSC cell seeding after passaging, leading to hPSC colonies of grossly unequal sizes

After passaging, it is important to ensure that hPSCs are evenly seeded across the well or dish. If hPSCs are unevenly seeded – i.e., they are concentrated in one part of the well or dish, or that some wells have more hPSCs than other wells – they will form colonies of grossly unequal sizes, in turn compromising differentiation. Efficient differentiation requires the even seeding of hPSCs at a sparse density across the well or dish. If colonies are too large at one part of the plate, they will not efficiently differentiate, whereas extremely sparse hPSCs at other parts of the plate will die.

Potential solution

• •Seed hPSCs more rapidly, to avoid hPSCs settling in the pipette tip. When seeding hPSCs with a pipette tip (e.g., a 5 or 10 mL pipette tip) into a new plate, dispense them briskly. If one waits too long, the hPSCs will begin settling in the pipette tip owing to the influence of gravity, such that the bottom of the pipette tip will have more hPSCs than the top of the pipette.
• •Shake the plate in a cross pattern immediately after seeding hPSCs. After seeding hPSCs onto a new plate, it is critical to rock the plate in a cross pattern (left, then right; up, then down) to distribute hPSCs evenly across the well. Do not swirl the plate, which promotes cell clumping at the center of the well. Do not wait too long to shake the plate after seeding hPSCs, as cells can adhere to the plate within minutes; once they adhere, they cannot be readily dislodged. Immediately after shaking the plate, check that cells are evenly distributed by checking the well under the microscope before placing plates in the incubator.
• •Ensure that plates are evenly arrayed in the incubator. After seeding hPSCs on a new plate and returning it to the incubator, ensure that they are laid flat on the incubator shelf. Tools are available to check that the incubator shelf is indeed flat and is not slightly angled. Additionally, if one stacks multiple plates on top of one another, ensure that they are evenly stacked on top of one another, or consider laying each plate flat on the incubator shelf without another plate beneath it.

Problem 3: Cell death during differentiation

If executed correctly, differentiation should yield ample numbers of artery and vein ECs (Figures 4 and 7). Near-complete or complete cell death during differentiation is unusual (Figure 8D), and should be addressed. Nevertheless, it is important to emphasize that some degree of cell death is expected during each differentiation step. Some cells die during day 1 of differentiation into the mid primitive streak. Additionally, PI3K signaling drives cell proliferation, but we add a PI3K inhibitor (GDC0941) as part of day 2 lateral mesoderm differentiation medium and day 3 artery EC differentiation medium, likely contributing to some degree of cell loss seen at those differentiation steps.

Potential solution

• •Seeding more hPSCs. In our experience, the leading cause for excessive cell death during differentiation is seeding too few hPSCs (Figure 8B and 8C). As mentioned previously, we strongly recommend that each new experimenter should test a range of cell seeding densities for mTeSR Plus-grown hPSCs — 25,000–50,000 hPSCs/cm² (i.e., ∼95,000–190,000 hPSCs/well of a 12-well plate) — on day 0 of differentiation. Typically, seeding fewer hPSCs will lead to near-complete cell death upon differentiation (Figure 8D). Each hPSC line can grow at a slightly different rate, and in turn, we strongly recommend that the experimenter test a range of hPSC seeding densities.
• •Avoid over-triturating hPSCs during the cell seeding process. When seeding hPSCs, avoid over-trituration, as described above in “problem 1: Poor survival of hPSCs after dissociation”. Excessive trituration will lead to poor hPSC survival upon plating, in turn precipitating excessive death upon differentiation.
• •Using fresh CDM2 basal medium. Typically, CDM2 basal medium can be stored for several months at 4°C. However, it is possible that prolonged storage is deleterious — certain CDM2 components, including insulin, are required for cell survival — and therefore one can prepare fresh CDM2 medium.
• •Ensure small molecules thaw correctly. Whenever thawing small molecules, check to ensure that they are evenly suspended and that they are fully thawed before preparing differentiation medium. This is especially important after repeated freeze-thaw cycles. If small molecules have precipitated, they will form crystalline shards that will kill cells, and which will be visible under the microscope. If small molecules have precipitated, order fresh stocks or use a different aliquot prepared from the same master stock.
• •Avoid keeping cells dry for too long after aspiration. Differentiating cells are somewhat fragile. When aspirating old differentiation media and adding either DMEM/F12 (to wash cells) or fresh differentiation medium, do not keep cells dry for too long.
• •Avoid exposing differentiating cells to Ca²⁺/Mg²⁺-free PBS. When washing cells between each differentiation step, use DMEM/F12. If cells are washed with Ca²⁺/Mg²⁺-free PBS for prolonged amounts of time, they will detach and die. As mentioned previously, Ca²⁺ is required for cadherin-mediated cell adhesion.

Problem 4: Poor differentiation efficiency

We have demonstrated that multiple hPSC lines can be efficiently differentiated into artery and vein ECs, and this is reproducible across multiple independent experimenters and laboratories (see Figure 5 and our accompanying paper¹). Vein EC differentiation (which typically yields 70%–100% pure CD144⁺ ECs) is generally less efficient than artery EC differentiation (which often produces 90%–100% pure CD144⁺ ECs), but significantly lower differentiation efficiencies are abnormal and should be rectified.

Potential solution

• •Seeding fewer hPSCs. In our experience, the leading cause for poor differentiation efficiency is seeding excessive numbers of hPSCs on day 0 of differentiation. As emphasized previously, each new experimenter should test a range of cell seeding densities for mTeSR Plus-grown hPSCs — 25,000–50,000 hPSCs/cm² (i.e., ∼95,000–190,000 hPSCs/well of a 12-well plate) — on day 0 of differentiation. Generally speaking, seeding too many hPSCs will lead to poor differentiation efficiency, ascribed to uneven action of developmental signals (e.g., BMP) across an excessively-wide hPSC colony.^72,73 If hPSCs were seeded too densely, after the first day of differentiation it will be visually apparent that the colony border has differentiated into larger primitive streak-like cells, while the colony center appears to comprise small and tightly-packed undifferentiated hPSCs.
• •Use of CDM2 basal medium. Differentiation should be conducted with the CDM2 basal medium described here. Do not substitute CDM2 with different basal medium. For instance, CDM2 contains lower concentrations of insulin (0.7 μg/mL), which in turn is important to ensure efficient primitive streak generation.
• •Double check small molecule and growth factor reconstitution and storage. Resuspend lyophilized small molecules and growth factors with the diluents, and at the concentrations, indicated in Table 2. Improper reconstitution of differentiation factors can compromise differentiation. As noted above, small molecules and growth factors are more stable when stored at higher concentrations.
• •Double check small molecule identity. Use the exact small molecules for differentiation that are specified in Table 2, and do not substitute them for alternative substitutes. Each small molecule has its own set of “off-target” effects. As mentioned above, we found that (1) replacing the BMP pathway inhibitor DMH1 with the alternative BMP pathway inhibitor LDN193189 or (2) replacing the TGF-β pathway inhibitor SB505124 with the alternative TGF-β pathway inhibitor A8301 will compromise EC differentiation. A8301 and LDN193189 both share an important “off-target” effect: inhibiting VEGF receptor,⁶⁶ and VEGF signaling is crucial for EC differentiation.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Lay Teng Ang (layteng@stanford.edu).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Sherry Li Zheng (slzheng@stanford.edu).

Materials availability

The above protocol does not entail the use of any new, unique reagents.

Data and code availability

The above protocol does not entail the use of any new datasets or codes.

Acknowledgments

We thank Liying Ou, Linda Heneghan, Aaron McCarty, Catherine Carswell-Crumpton, Courtney Scheuch, Polly Leung, Valerie Hyun-Hee Park, Diane Qin, Heather Gentner, Mike Alvarez, Laura Dunkin-Hubby, the Stanford Institute for Stem Cell Biology & Regenerative Medicine, and the Stanford Diabetes Research Center (NIH P30DK116074) for infrastructure support. This work was supported by NIH Director’s Early Independence Award DP5OD024558 (K.M.L.), NIH R01HL128503 (K.R.-H.), NIH R01HL169787 (E.S.), JDRF Northern California Center of Excellence (K.M.L.), Additional Ventures Innovation Fund (K.M.L.), Additional Ventures Expansion Award (K.M.L. and L.T.A.), Stanford Ludwig Center for Cancer Stem Cell Research and Medicine (K.M.L.), Stanford Beckman Center for Molecular and Genetic Medicine (K.M.L.), ARPA-H HEART 1AY1AX000002-01 (K.M.L.), Stanford Maternal and Child Health Research Institute Transdisciplinary Initiatives Program (K.M.L., L.T.A., and K.R.-H.), HHT Foundation (E.S.), Singapore Ministry of Education Academic Research Fund T2EP30122-0008 (C.C.), Lee Kong Chian School of Medicine Vascular Research Initiative (C.C.), California Institute for Regenerative Medicine TB1-01195 (A.T.N. and A.C.), Stanford School of Medicine Dean’s Postdoctoral Fellowship (Q.Y.), NIH T32GM007276 (Z.A.A.-U.), Stanford Sarafan ChEM-H Chemistry/Biology Interface Training Program (Z.A.A.-U.), and the anonymous, Fickel, and Weintz families (K.M.L.). S.L.Z. was supported by NSF and Stanford Graduate Fellowships and NIH T32GM007790. K.R.-H. is an HHMI Investigator. L.T.A. is a Siebel Investigator and Additional Ventures Catalyst to Independence Fellow. K.M.L. is a Human Frontier Science Program Young Investigator (RGY0069/2019), Packard Foundation Fellow, Pew Scholar, Baxter Foundation Faculty Scholar, and The Anthony DiGenova Endowed Faculty Scholar.

Author contributions

K.M.L., A.T.N., A.C., and L.T.A. developed the method to differentiate hPSCs into artery and vein ECs. S.L.Z., K.J.L., Q.Y., Z.A.A.-U., F.-M.N., S.K.J., Y.Q., V.K.W., C.C., E.S., and K.R.-H. tested the robustness and reproducibility of the hPSC differentiation method. K.M.L. and L.T.A. drafted the manuscript, which was edited by S.L.Z. and K.J.L., with input from all other authors.

Declaration of interests

Stanford University has filed patent applications related to endothelial differentiation, with K.M.L., L.T.A., A.C., and A.T.N. listed as inventors. A.C. is presently at Orca Bio and A.T.N. is presently at the University of California, Davis, but A.C. and A.T.N. contributed to this work when they were at Stanford University.