Whole-Mount In Situ Hybridization in Zebrafish Embryos and Tube Formation Assay in iPSC-ECs to Study the Role of Endoglin in Vascular Development

A single mutation on eng (CD105) has been reported in patients with HHT. The mutation leads to increased EC proliferation and reduced flow-mediated EC elongation^1,2. It has also been previously reported that ENG deficiency reduces endothelial markers expression (i.e., kdrl, cdh5, hey2) in zebrafish³. Endoglin, mainly expressed in endothelial cells, is a transmembrane glycoprotein and functions as a co-receptor for transforming growth factor β (TGF-β) family members. It directs BMP9 binding on the cell surface to regulate downstream gene, including Id1 expression, to induce stem cell differentiation toward ECs⁴. Thus, the eng gene plays important roles in vasculogenesis and human vascular disease^5,6. We have previously examined the effects of endoglin knockdown on vessel formation in zebrafish embryos, followed by analysis of iPSCs-derived ECs acquired from an HHT patient bearing an eng mutation⁷. This protocol demonstrates the effects of ENG deficiency on endothelial progenitor marker expression and tube formation, which is a quantifiable method for measuring in vitro angiogenesis.

To study eng spatial and temporal expression, WISH is employed to detect gene expression in vivo⁸. In situ hybridization (ISH) is a method of using labeled probes with complement sequences of target nucleic acids (DNA or mRNA) to detect and visualize target nucleic acid hybrids in a fixed specimen. The process amplifies gene expression signals in vivo and is used to detect the expression of genes by microscopy. WISH has been widely used in various model animals, especially in zebrafish⁹. It is also used to acquire the following data: 1) gene spatial/temporal expression patterns, which provide information about gene function and classification; and 2) specifically expressed gene markers that are used in high-throughput drug or mutant screening¹⁰.

Chromogenic probes are easily degraded with traditional chromosome in situ hybridization (CISH), which results in high background noise and nonspecific signals^11,12. The WISH method uses two independent double Z probes, which are designed to hybridize to target RNA sequences. Each probe contains an 18-25 sequence complementary to the target RNA and a 14 base tail sequence (conceptualized as Z). The target probes are used in a pair (double Z). The two tail sequences together form a 28 base hybridization site for the preamplifier, which contains 20 binding sites for the amplifier. The amplifier, in turn, contains 20 binding sites for the label probe and can theoretically yield up to 8,000 labels for each target RNA sequence.

This advanced technology facilitates simultaneous signal amplification and background suppression to achieve single-molecule visualization while preserving tissue morphology¹³. Further modification of the WISH methods is based on previous research¹⁴, including extra fixation and DAB staining. Provided here is an improved WISH protocol that can work even if the target RNA is partially decreased or degraded. Advantages include that it can be completed in 24 h without RNase-free conditions. Signals can also be simultaneously detected through multiple channels from multiple targets, and the results are consistent and compatible with results from different high-throughput automation platforms.

Results from animal models do not necessary reflect the same phenomenon that occurs in humans. ENG contains two pairs of conserved cysteines, C30-C207 and C53-C182, which form disulfide bridges in orphan regions. To further study the role of eng in HHT patients, tube formation assays with iPSCs derived from HHT patients have been carried out in cells without/with eng mutations (Cys30Arg, C30R)¹⁵. After Kubota et al. first reported the tube formation experiment¹⁶, the assay has been developed in several ways. It has been used to identify angiogenic or antiangiogenic factors, define the signaling pathways in angiogenesis, and identify genes regulating angiogenesis¹⁷.

Prior to the availability of patient-derived iPSC-ECs, researchers used primarily cultured ECs to study angiogensis¹⁶. However, for endothelial cells, it is a technical challenge to transduce exogenous genes with a virus, because of the limited passage number that ECs can undergo. This is because there is hardly enough cellular material to be collected from human vessels either from surgery or matched approved donors. With the invention of the iPSC generation technique by Shinya Yamanaka, human ECs derived from iPSCs can be used reliably in in vitro experiments, as reported previously¹⁸.

Using virally transduced ECs with limited numbers and passages may be sufficient for signaling studies, but for functional studies, it is better to generate mutant pluripotent stem cell lines, (either iPSCs or CRISPER/Cas9-targeted hESCs), then differentiate them into ECs for angiogenesis studies that use tube formation assays¹⁹. Tube formation can be used to evaluate the function of endothelial cells bearing mutations. This protocol also describes tube formation on an µ-slide angiogenesis plate, which is an easy, cost-effective, and reproducible method.

The protocol below provides a reliable and systemic method for studying the role of specific genes in vascular formation, along with details for in vivo expression pattern and in vitro functional quantification for modeling human disease.

All animal experiments described were approved by the Research Ethics Committee of Zhejiang University school of medicine.

1. Whole-mount in situ hybridization

1. Zebrafish line husbandry and reproduction
   1. Feed and raise all adult zebrafish at 26-28 °C in a recirculating aquaculture system with 14 h light/10 h dark cycle for each day. Use AB (wild-type) zebrafish lines for the following procedures.
   2. Put a layer of pebbles in the bottom of the breeding box as a shelter for eggs. Group parental fish in the ratio of 2:1 (i.e., two females and one male) into each box. Let the female fish spawn for 1 day. At the end of spawning, remove the parent fish immediately to prevent them from eating the eggs.
   3. Inject the 2 ng of morpholinos and 500 pg of mRNA into one-cell stage embryos using the Femto Jet injection system under a microscope (endoglin-MOs sequence: 5'-GATGAACTCAACACTCGTGTCTGAT-3'; 5-mispair control MOs sequence: 5'-AAACAGACCACATCCTCTTCATCTC-3').
   4. Incubate the zygotes in acidic sea water at 27 °C for approximately 48 h.
2. Collection and fixation of zebrafish embryos
   1. Use a plastic transfer pipette to collect 20-40 zebrafish embryos from the circulating system water (pH = 5.0) in a 1.5 mL tube when they are dechorionated and reach a specific period of development. Add 1 mL of freshly prepared 4% PFA solution at room temperature (RT).
      NOTE: The period at which the embryo chosen is based on the purpose of the experiment. Table 1 shows the fixation time required for different embryonic periods.
   2. Remove the fixation solution, wash embryos with phosphate-buffered solution with Tween-20 (PBST, dilute 1 mL of Tween into 1000 mL of PBS solution) 3x for 5 min each at RT.
   3. Dehydrate the embryos through incubation in a series of 25%, 50% and 75% methanol (diluted in PBST) for 5 min each. Transfer embryos to 100% methanol for 5 min and replace with fresh methanol. Store the embryos at -20 °C for overnight or longer.
3. Hydration and digestion
   1. Hydrate embryos that were preserved in 100% methanol using a sieve with nylon mesh at the bottom to sequentially wash embryos in the wells of 12 well plates with a series of 75%, 50%, and 25% methanol (diluted in PBST) for 5 min each. Wash the embryos with PBST 3x for 5 min each at RT.
   2. Add 50 µL of proteinase K (10 mg/mL) into the tube with embryos and follow the digestion time in the Table 2 at RT.
   3. Remove the proteinase K and wash embryos with PBST 3x for 5 min each at RT.
4. Probe hybridization and post-fixation
   1. Put the probes designed according to each gene's mRNA sequences in a 40 °C hybridizatin system until the precipitate is dissolved, which should take ~10 min. Mix the target probes of eg (XM_007116.7) and an additional two important genes involve in the early mesoderm endotheial progenitor formation(here, aplnr [NM_001075105.1] and nrp1a [NM_001040326.1]) in a 1.5 mL tube at a 50:1:1 ratio.
   2. Add 50-100 µL of mixed target probes into each tube with embryos, then incubate overnight at 40 °C.
      NOTE: Pre-mixed probes must be pre-warmed to 40 °C and cooled to RT before use.
   3. Prepare a 0.2x saline sodium citrate solution with Tween-20 (SSCT): dissolve 175.3 g of NaCl and 88 g of sodium citrate into 1 L of dH₂O to obtain a 20x SSC solution. Then, dilute the solution 1:100 in dH₂O to obtain a 0.2 x SSCT solution. Dilute Tween-20 1:1000 in the 0.2x SSCT solution.
   4. Transfer the recycled probes to a new tube. Wash the embryos in 0.2x SSCT solution 3x for 15 min each at RT. The recycled probes can be reused 5x-10x.
   5. Use 4% paraformaldehyde (PFA) to fix the embryos for 30 min at RT. Wash the embryos in 0.2x SSCT solution 3x for 15 min each at RT.
      NOTE: The work of Gross-Thebing et al. suggests a post-fixation period of 10 min¹⁴, but the actual amount must be adjusted accordingly. Here, we tested 10 min post-fixation 3x, and the sensitivity of DAB dying is significantly influenced. After investigating, ensure that 0.5-1.0 h of fixation is the optimal condition (a shorter fixation period cannot produce a clear signal from the target RNA, and the appropriate extension of fixation time helps strengthen the signal and provide less background noise).
5. Sequential amplifier and label probe hybridization
   1. Remove the SSCT solution and replace with 50 µL of Amp 1. Allow the embryos to settle in Amp 1 at 40 °C for 30 min. Then, wash the embryos in 0.2x SSCT solution 3x for 15 min each at RT.
   2. Remove the SSCT solution and add 50 µL of Amp 2. Allow the embryos to settle at 40 °C for 15 min. Then, wash the embryos in 0.2x SSCT solution 3x for 15 min each at RT.
   3. Remove the SSCT solution, add 50 µL of Amp 3 and tap the tube mildly. Then, incubate the embryos at 40 °C for 30 min. Wash the embryos in 0.2x SSCT solution 3x for 15 min each at RT.
   4. Remove the SSCT solution, add 50 µL of Amp 4 dropwise, and carefully tap the tube. Then, incubate the embryos at 40 °C for 15 min. Wash the embryos in 0.2x SSCT 3x for 15 min each at RT.
6. Counterstaining and microscopy
   1. Remove the SSCT solution and add 50 µL of DAPI to each tube. Incubate the embryos overnight at 4 °C.
   2. Wash the embryos with PBST and prepare them for imaging using a 1% low melting point agarose (LMP) solution (1 g of LMP dissolved in 100 mL of dH₂O) in a Petri dish filled with PBST.
   3. Use a DAB peroxidase substrate kit to stain the specimen. Add color reagents A, B, and C (50 µL each) to 1 mL of distilled water and mix well to obtain a complete DAB working fluid. Add the fluid to the specimen and cover for 10 min.
   4. Wash the specimen thoroughly with PBST after staining.
   5. Use an optical microscope with a photographic function for imaging the samples.
      NOTE: If taking images later, tubes should be wrapped in aluminum foil and stored at 4 °C.

2. Tube formation assay

NOTE: The eng mutant and wildtype ECs (control) were differentiated from iPSCs derived from an HHT patient (here, a 62 year-old female patient with recurrent epistaxis since 22 years old, gastrointestinal bleeding, pulmonary arteriovenous malformations, carrying a missense eng mutation in position c.88T>C of exon 2) and a healthy donor (without eng mutation), which were provided by Peking Union Medical College Hospital with approval from the college research ethics committee.

1. Control and HHT iPSC cell cultures
   1. Add a certain volume of basement membrane matrix (e.g., Matrigel) to a 6 well cell culture plate to cover the bottom of the well and incubate the plates for 1 h at 37 °C.
      NOTE: When first using the basement membrane matrix stored at -20 °C, incubate on ice or in a frost-free 4 °C refrigerator until thawed. Then, dilute at 1:100 in DMEM/F12, then distribute the diluent into 200 µL tubes containing 100 µL each. Store the tubes at -20 °C and avoid repeated freezing and thawing. When adding the diluted basement membrane matrix, do not create bubbles. After adding the diluted basement membrane matrix, shake the 6 well plate gently by hand so that it evenly covers the bottom.
   2. After coating the plate, remove the used basement membrane matrix solution from the wells. Then, add 2 mL of mTeSR1 medium into each well, and plate the iPSCs harvested from the last passage at approximately 1 x 10⁶ per well.
2. Generation and expansion of ECs from iPSCs
   1. After approximately 4 days of iPSC culture, exchange the culture medium with BEL medium supplemented with activin A (25 ng/mL), BMP4 (30 ng/mL), VEGF165 (50 ng/mL), and CHIR99021 (1.5 µM). Grow the cells for 3 days to generate mesoderm cells.
   2. Replace the medium described in step 2.2.1 with BEL medium supplemented with VEGF (50 ng/mL) and SB431542 (10 µM) for 4 days to expand vascular ECs. Treat the cells with the same medium and culture for another 3-4 days.
   3. Use CD31-dynabeads to purify the mature vascular ECs: refer to the instructions in the CD31-dynabeads kit, which links human CD31 antibody with magnetic beads to combine CD31 molecules expressed on ECs, then collect the CD31-positive cells by elution buffer. Maintain and expand the purified ECs in EC-SFM medium containing VEGF165 (30 ng/mL), bFGF (30 ng/mL), and FBS (1%).
3. Plating endothelial cells on Matrigel-coated plates and microscopy
   1. Add 10 µL of basement membrane matrix per well to coat the angiogenesis plates (see Table of Materials) and incubate at 30 min at 37 °C.
   2. Harvest endothelial cells after checking endothelial markers CD31 and VE-cadherin with immunofluorescence staining and resuspend them in endothelial growth medium 2 by pipetting repeatedly. Add 50 µL of cell suspension (2 x 10⁵ cells/mL) per well onto the solidified matrix and incubate the cells for 3-5 h at 37 °C.
      NOTE: Before the tube formation experiment, endothelial markers should be checked to make sure the proper phenotype of functional endothelial cells. 1 x 10⁴ cell per well is an ideal number for tube formation. Too many or too few cells are not conducive to tube formation. Add cells from the side of the dish and do not touch the basement membrane matrix. Use light microscope in high magnification field to check the cells and take photos.
4. Quantitative results of tube formation
   1. Observe the results with a microscope with high resolution and take pictures of at least 10 areas for each group to obtain reliable data statistics.
   2. Examine endothelial tube formation: estimate the extent of tube formation by inspecting the overall tube length, tube number, and branch points. Assess and count the number and length of branches using ImageJ software.

Whole mount in situ hybridization is based on a principle similar to fluorescence resonance energy transfer. It is designed to improve both sensitivity and specificity in zebrafish ISH as well as amplify target-specific signals without affecting the background signal.

In 24 hpf zebrafish embryos, endoglin is highly expressed in the posterior cardinal vein (PCV), intersegmental vessels (ISVs), and blood islands³. It is hard to control the staining time in traditional chromogenic in situ hybridization (CISH), with non-positive signals occurring in some regions, such as yolk sac (Figure 1A). To investigate the role of eng in early vascular development, a hemogenic endothelial marker was used (aplnra) as well as another endothelial progenitor marker (nrp1a) to examine which type of endothelium was affected by eng silencing²⁰. The expression of aplnra and nrp1a was weak in 24 hpf embryos. Compared to CISH, the weak signal from the two genes was clearly demonstrated in the tail region after eng knockdown in WISH (Figure 1B). The knockdown of eng RNA caused reduced expression of two types of endothelial cells marker genes expression in an HHT animal model.

Before conducting experiments on differentiated iPSCs, the cells were identified and confirmed as ECs. Morphologically, they appeared as cobblestone shape. Expression of endothelial cell surface molecular markers CD31, CD146, VE-cadherin, and vWF was confirmed with IF staining²¹. After magnetic-based isolation using CD31, the functional assay was conducted as described below.

Here, the tube formation assay was performed using the eng mutant and control iPSC-derived endothelial cells. Statistical analysis was performed based on the number, branches, and lengths of tubes. A novel parameter was also introduced based on previous work³, points of angiogenesis, in order to reflect tube formation of endothelial cells. Eventually, it was found that eng mutant endothelial cells formed fewer branches than control endothelial cells, and that branches in eng mutant endothelial cells significantly increased after stimulation with vascular endothelial growth factor (VEGF) (Figure 2A,B). The mutation in eng resulted in defective vessel tube formation.

Figure 1: Endoglin knockdown decreased the expression of endothelial markers. (A) CISH and WISH were performed to determine the expression of endoglin in 24 hpf embryos (n > 30). The red box represents the enlarged region. Scale bars = 200 µm. (B) aplnra and nrp1a expression by WISH in 24 hpf zebrafish embryos of the control-MO and eng-MO group. The red arrow indicates regions where the expression of these genes significantly decreased. Please click here to view a larger version of this figure.

Figure 2: Tube formation of eng mutant and control iPSC-ECs. (A) The tube formation by the four groups, including the control group, the control + VEGF (control endothelial cells + 30 ng/mL VEGF), the eng mutant (eng mutant endothelial cells) group, and the eng mutant + VEGF (eng mutant endothelial cells + 30 ng/mL VEGF) group. Tube formation was assessed and photographed after 3 h. Scale bars = 100 µm. (B) Quantitative results of tube formation are shown. At the covered area the number of branches and their length of tubes were calculated by Image J. Error bars represent the SD of the mean values from three independent experiments. A value of P was considered statistically significant when *p < 0.05. Please click here to view a larger version of this figure.

Table 1: Fixation time required for different embryonic stages.

Table 2: Digestion time required for different embryonic stages.

This protocol applied an improved whole-mount in situ RNA analysis platform for zebrafish and tube formation assays on iPSC-ECs derived from an HHT patient. The traditional ISH method requires a longer experimental cycle with extra experimental steps. The protocol has some important improvements, the use of independent double Z probes and iPSCs derived from a patient with HHT that were applied in WISH assays and tube formation, respectively. These refinements are crucial for enhancing the detection sensitivity compared with what is observed with traditional ISH. The uniquely designed probes and target signal amplification allow the detection of rarely expressed transcripts (aplnra is a poorly expressed gene in 24 hpf embryos). One modification is the extra fixation after probe hybridization. The extra fixation can enhance the combination of probes and transcripts. DAB staining is also applied rather than fluorescence staining, because it was found that it was difficult to perform fluorescence staining in certain low-abundance genes. Then, iPSCs were used for angiogenesis assays, which increases the significance and clinic relevance of this work. Generation of iPSCs requires strict culture conditions and represents a limitation of the protocol.

Some critical steps must be highlighted. In WISH analysis, the digestion time of zebrafish embryos should be followed in strict accordance with the text. A shorter digestion time cannot guarantee that the probe combines with transcripts successfully. In contrast, extended digestion time will destroy the integrity of embryos. In the tube formation experiments, the imaging timing should be well-controlled. Generally, the endothelial cells will become tubes within 12 h but the tubular structure tends to disintegrate after 24 h. The cell number is also important for tube formation, as a cell density that is too high or too low can lead to failed tube formation. If it is difficult to induce endothelial cells to form tubes, it is helpful to include adding growth factors such as VEGF to the culture medium as a positive control to test the ability of ECs to form tubular structures.

In summary, the improved WISH method has been considered an advantageous technique for laboratory workflows. Recently, our research group has started to test this assay in clinical samples. Considering higher sensitivity and more efficient detection than observed in traditional methods, this improved WISH method holds significant promise for developing and implementing RNA-based molecular diagnostics²².

Performance of tube formation assay using patient iPSC-ECs allows confirmation of results from animal studies as well as identification of disease-causing mutations from single nucleotide polymorphism in the eng gene. The combination of these methods clarifies the role of eng in vascular formation and hold potential in gene correction of patient iPSC-ECs for cardiovascular cell therapy²³.