Abstract

Background

   B-cell CLL/lymphoma 6 member B (BCL6B) operates as a sequence-specific
   transcriptional repressor within the nucleus, playing crucial roles in
   various biological functions, including tumor suppression, immune
   response, stem cell self-renew, and vascular angiogenesis. However,
   whether BCL6B is involved in endothelial cell (EC) development has
   remained largely unknown. ETS variant transcription factor 2 (ETV2) is
   well known to facilitate EC differentiation. This study aims to
   determine the important role of BCL6B in EC differentiation and its
   potential mechanisms.

Methods

   Doxycycline-inducible human induced pluripotent stem cell (hiPSC) lines
   with BCL6B overexpression or BCL6B knockdown were established and
   subjected to differentiate into ECs and vessel organoids (VOs). RNA
   sequencing analysis was performed to identify potential signal pathways
   regulated by BCL6B during EC differentiation from hiPSCs. Quantitative
   real-time PCR (qRT-PCR) was used to detect the expression of
   pluripotency and vascular-specific marker genes expression. EC
   differentiation efficiency was determined by Flow cytometry analysis.
   The performance of EC was evaluated by in vitro Tube formation assay.
   The protein expression and the vessel-like structures were assessed
   using immunofluorescence analysis or western blot. Luciferase reporter
   gene assay and chromatin immunoprecipitation (ChIP)-PCR analysis were
   used to determine the regulatory relationship between BCL6B and ETV2.

Results

   Functional ECs and VOs were successfully generated from hiPSCs.
   Notably, overexpression of BCL6B suppressed while knockdown of BCL6B
   improved EC differentiation from hiPSCs. Additionally, the
   overexpression of BCL6B attenuated the capacity of derived hiPSC-ECs to
   form a tubular structure. Furthermore, compared to the control VOs,
   BCL6B overexpression repressed the growth of VOs, whereas BCL6B
   knockdown had little effect on the size of VOs. RNA sequencing analysis
   confirmed that our differentiation protocol induced landscape changes
   for cell/tissue/system developmental process, particularly vascular
   development and tube morphogenesis, which were significantly modulated
   by BCL6B. Subsequent experiments confirmed the inhibitory effect of
   BCL6B is facilitated by the binding of BCL6B to the promoter region of
   ETV2, led to the suppression of ETV2's transcriptional activity.
   Importantly, the inhibitory effect of BCL6B overexpression on EC
   differentiation from hiPSCs could be rescued by ETV2 overexpression.

Conclusions

   BCL6B inhibits EC differentiation and hinders VO development by
   repressing the transcriptional activity of ETV2.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s13287-024-03832-y.

   Keywords: B-cell CLL/lymphoma 6 member B, Human induced pluripotent
   stem cell, Endothelial cell, Vessel organoids, ETS variant
   transcription factor 2

Background

   B-cell CLL/lymphoma 6 member B (BCL6B), also referred to as BAZF, ZNF62
   and ZBTB28, belongs to the B-cell lymphoma 6 (BCL6) gene family
   [[47]1]. Members of this family encode zinc finger proteins with the
   Broad-Complex, Tramtrack and Bric a brac/POxvirus and Zinc finger
   (BTB/POZ) domain, functioning as sequence-specific transcriptional
   repressors. The zinc finger motifs and BTB/POZ domain of BCL6B are 94%
   and 65% identical to those of BCL6 at the amino acid level [[48]1].
   Although binding to similar target DNA sequences to act as
   transcriptional repressors, the tissue expression pattern and
   pathological function of BCL6B differ from that of BCL6. Some studies
   have indicated that BCL6B is involved in diverse biological functions,
   such as facilitating spermatogonial stem cell self-renewal [[49]2],
   stimulating secondary response of memory CD8 + T lymphocytes [[50]3],
   and repressing hepatocellular carcinoma [[51]4, [52]5], colorectal
   carcinoma [[53]6, [54]7], and breast cancer [[55]8]. A recent study by
   Miruto Tanaka et al. showed that BCL6B contributes to ocular vascular
   diseases via Notch signal silencing [[56]9]. BCL6B is enriched in
   endothelial cells (ECs) [[57]10], however, as a transcriptional
   repressor, the function and mechanisms through which BCL6B regulates EC
   development remain poorly understood.

   The two-dimensional (2D) cell culture system has been successfully
   applied in biomedical fields over the past 100 years which has
   significantly help us to improve our knowledge of cellular signaling
   pathways, develop guidelines for the design of candidate compounds,
   identify potential drug targets, and clarify the underlying
   pathological mechanisms of diseases. However, such a system comes with
   many disadvantages such as lack of hierarchical structure,
   dimensionality, cellular diversity, and cell–cell or cell–matrix
   interactions, all of these properties are crucial for maintaining the
   in vivo cellular functions and tissues homeostasis. Such a limitation
   can be circumvented by the newly established human organoids. 3D human
   organoids derived from primary tissues or stem cells can self-renew,
   and self-organize, and exhibit multicellularity and functionality
   similar to those of in vivo organs. Particularly, human induced
   pluripotent stem cells (hiPSCs) have the capability to differentiate
   into various cell types [[58]11], which allows for studying human
   development, cell differentiation and the role of critical
   transcriptional factors in cell fate decision [[59]12]. Importantly,
   recent advances in hiPSC-derived organoids enable a better
   recapitulation of the three dimensional (3D) microstructures and
   functions of human tissues [[60]13]. HiPSC-derived organoids provide a
   more sophisticated tool and model system for exploring the critical
   signal pathways underlying human embryonic development and cell
   differentiation, modeling human diseases, uncovering disease aetiology,
   and screening drug candidates [[61]14].

   ETS variant transcription factor 2 (ETV2), also known as ER71, a member
   of the E26 transformation-specific transcription factor family, has
   been identified as a master transcription factor for the
   hematoendothelial lineage development [[62]15, [63]16]. ETV2
   overexpression in mouse embryonic stem cells directly regulates Flk1
   expression and initiates the de novo generation of ECs [[64]17,
   [65]18], and ETV2 alone is sufficient to directly convert human
   fibroblasts into ECs [[66]19, [67]20]. Although much effort has been
   made to study the importance of ETV2 in EC differentiation, little is
   known about its upstream regulators. In zebrafish, transcription factor
   FOXC1a/b serves as a direct upstream regulator of Etsrp (orthologous to
   human ETV2) and involves in mesoderm specification into angioblasts
   [[68]21]. However, the upstream regulators of ETV2 in mammal still
   remains unclear.

   In this study, utilizing both hiPSC-derived 2D ECs and 3D vessel
   organoids (VOs), we uncovered that BCL6B hampers EC differentiation and
   VO development from hiPSCs by repressing the transcriptional activity
   of ETV2.

Methods

Cell culture and EC differentiation

   hiPSCs (passage 25–40) were maintained under feeder-free conditions in
   defined E8 media (Gibco, A1516901) on tissue culture plates coated with
   hESC-qualified Matrigel (Corning, 354277) at 37 °C in 5% CO2. The
   culture medium was exchanged daily. Routine passaging of hiPSCs at 1:4,
   and single-cell dissociation were carried out using Accutase solution
   (Sigma, A6964). The hiPSCs were cryopreservation with CryStor CS10
   (Stemcell, 100–1061). In addition, we performed a mycoplasma
   contamination test every two weeks to ensure the health of all the cell
   lines (data not shown).

   The EC differentiation protocol was adapted from a published protocol
   [[69]22]. In brief, hiPSCs were cultured in Essential 8 medium
   reached ~ 80% confluency, then were dissociated into single cells, and
   seeded in 6-well plates with 10 μM ROCK inhibitor Y27632 (Selleck,
   S1049) on D0 at a density of 10^5 cells/well. On D1, hiPSCs were
   treated with 8 μM CHIR99021 (Selleck, S2924) in Chemically Defined
   Medium (CDM) for 3 days. On D4, differentiated cells were treated with
   50 ng/ml VEGF (PeproTech, 100–20) and 10 mM SB431542 (Selleck, S1067)
   until cells reached ~ 100% confluent on D6. For cell dissociation,
   TrypLE Select Enzyme (10 ×) (Gibco, A1217701) was used for 10 min until
   most cells became singlets. The cells were then passed through a 40-μm
   strainer. Cell dissociation was neutralized by adding 2 ml of
   Endothelial Cell Medium (ECM, ScienCell, 1001), followed by magnetic
   associated cell sorting (MACS) to purify hiPSC-derived ECs with CD144
   (VE-Cadherin) MicroBeads (Miltenyi, 130-097-857) or subjected to flow
   cytometry analysis. The CDM consists of IMDM (50%); Ham's F12 Nutrient
   Mix (50%); BSA (0.25%); Lipid concentrate (1X); ITS (0.1%); Ascorbic
   acid (50 µg/ml); Monothioglycerol (450 µM); and Glutamax (1X).

   Human umbilical vein endothelial cells (HUVECs) were purchased from
   OriCell® (Cat. HUVEC-20001), and cultured in Endothelial Cell Medium
   (ECM; ScienCell, 1001) supplemented with 5% heat-inactivated fetal
   bovine serum (ScienCell, 0025), and 1% endothelial cell growth
   supplement (ECGS; ScienCell, 1052). All HUVECs in this paper were used
   by passage 5. HEK293T cells were maintained in our lab with DMEM medium
   (Gibco, 11965092) supplemented with 10% fetal bovine serum (PAN,
   ST30-3302). All cells were cultured at 37 °C and 5% CO[2].

Generation of human blood VOs

   A step-by-step protocol detailing the differentiation of the human VOs
   can be found in Nature Protocol [[70]23]. Briefly, hiPSCs were
   dissociated into single cells with Accutase. Subsequently, the cells
   were resuspended in the E8 medium containing 10 μM Y27632 and seeded at
   1000 cells/well in round bottom ultra-low attachment 96-well plates
   (Costar, 7007) on D-1, with a volume of 100 μl per well. The plate was
   then centrifuged at 100 g for 3 min to assemble hiPSCs into embryoid
   bodies (EBs) and placed in an incubator at 37 °C with 5% CO2. After
   24 h (D0), the culture medium was replaced by the N2B27 medium with
   12 μM CHIR99021 and 30 ng/ml BMP4 (PeproTech, 120–05). On D3, the N2B27
   medium was renewed with 100 ng/ml VEGF and 2 μM forskolin (HY-15371).
   Starting from D6, the resulting cell aggregates were embedded in
   Matrigel:Collagen I (1:1) gels and overlaid with ECM containing 15%
   FBS, 100 ng/mL VEGF and 100 ng/ml FGF-2 (PeproTech, 100-18B). This
   medium was changed every two days. Around D10-12 vascular networks were
   established, and either directly analyzed or networks from individual
   cell aggregates were extracted from gels and further cultured in round
   bottom ultra-low attachment 96-well plates. These vascular networks
   self-assembled into vascular organoids, and the medium was changed
   every three days until organoids were ready for analysis.

Establishment of BCL6B overexpression and knockdown hiPSC lines

   The human BCL6B coding regions, fused with Flag tags, was amplified
   through polymerase chain reaction (PCR) using hiPSC cDNA as template.
   Subsequently, these templates were cloned into adapted FUW-tetO-MCS
   vector (MiaoLingBio, [71]P48786). Concurrently, shRNA constructs were
   generated using the Tet-pLKO-puro vector (MiaoLingBio, P0171),
   following previously described methods [[72]24]. For viral production,
   10 µg of target vectors (BCL6B-FLAG, BCL6B shRNA, and FUW-M2rtTA
   (MiaoLingBio, P0521), along with lentiviral envelope and the packaging
   plasmids psPAX2 (MiaoLingBio, P0261) and pMD2.G (MiaoLingBio, P0262),
   were transfected into 293 T cells using lipofectamine 3000 (Invitrogen,
   L3000015). After 72 h, lentivirus supernatants were collected and
   filtered through a 0.45 µm syringe filter. hiPSCs were coinfected with
   BCL6B-FLAG and FUW-M2rtTA or infected solely with BCL6B shRNA
   lentivirus using a spin infection method, as previously described
   [[73]25]. Two days post-infection, puromycin-resistant clones positive
   clones were selected using 1 μg/ml puromycin (MCE, HY-B1743A) for
   4 days to establish BCL6B overexpression (BCL6B OE) or BCL6B knockdown
   (BCL6B KD) hiPSC lines, respectively. The successful establishment of
   hiPSC lines was confirmed by western blotting and then applied for
   later differentiation.

Immunofluorescence analysis

   For hiPSCs-derived ECs, the purified ECs were seeded in a 24-well plate
   with coverslips and cultured until reaching 100% confluence. The cells
   were fixed in 4% paraformaldehyde (PFA) for 15 min, permeabilizated
   with 0.5% Triton X-100 for 10 min, followed by blocking with 10% goat
   serum for 30 min. Subsequently, the cells were incubated with mouse
   anti-CD31 (Abcam, ab9498) and rabbit anti-VE Cadherin (ab33168) at 4 °C
   overnight. Secondary antibodies used were Alexa Fluor 488-conjugated
   goat anti-rabbit (Abcam, ab150077) and Alexa Fluor 647-conjugated goat
   anti-Mouse (Abcam, ab150115). DAPI (Sigma, D9542) was used to mark cell
   nuclei. Stained coverslips were mounted with mounting medium and stored
   at 4 °C before imaging. All images were acquired by confocal imaging
   systems.

   For whole-mount staining of organoids, the samples were fixed in 4% PFA
   for 1 h at room temperature (RT), followed by three washes with PBS.
   Subsequently, the organoids were incubated in 0.5% TritonX-100 at RT
   for 1 h. After blocking with 5% BSA in 0.1% TritonX-100 at RT for 1 h,
   organoids were incubated with primary antibodies at 4 °C for 48 h,
   washed with PBS, and then incubated with secondary antibodies at 4 °C
   for 48 h. The stained organoids underwent three washes with PBS before
   being mounted on glass microscope slides using Mounting Medium
   (Invitrogen, [74]P36961). To preserve the 3D structure of the organoids
   before confocal imaging, Polybead Microspheres were placed between the
   slide and the coverslip.

Flow cytometry analysis

   HiPSC-ECs (D6) were dissociated into single-cell suspensions using
   TrypLE Select Enzyme for 8–10 min. After resuspension in staining
   buffer (Biolegend, 420201), approximately 0.5 × 10^6 single cells for
   each group were incubated with PE anti-human CD144 antibody (Biolegend,
   348506) and FITC anti-human CD31 antibody (Biolegend, 303104) for
   30 min. The results were analyzed using the BD LSR FortessaX-20
   cytometer and FlowJo software.

   The vascular organoids (D15) were mechanically disrupted and
   disaggregated using 3U/mL Dispase (Gibco, 17105041), 2U/mL Liberase
   (Sigma, 5401020001) and 100U DNAse (Stemcell, 07900) in PBS for 30 min
   at 37 °C while rotating. Subsequently, single cells were stained with
   the following antibodies for 30 min: PE anti-human CD144 antibody
   (Biolegend, 348506), FITC anti-human CD31 antibody (Biolegend, 303104)
   and APC anti-human PDGFRβ antibody (Biolegend, 323608), followed by
   FACS analysis.

RNA extraction, cDNA synthesis, and qRT-PCR

   The total RNA of hiPSCs, hiPSC-ECs and 10 VOs were extracted using
   Trizol (Invitrogen, 15596026), followed by reverse transcription to
   generate cDNA using the HiScript II 1st Strand cDNA Synthesis Kit
   (Vazyme, R212-01). qRT-PCR was performed using the Applied Biosystems
   QuantStudio 5 qPCR system with the ChamQ SYBR qPCR Master Mix (Low ROX
   Premixed) (Vazyme, Q331-02). Relative mRNA expression was determined by
   the delta cycle time with human GAPDH serving as the internal control
   for data normalization. The primer sequences were as follows:

   GAPDH: forward, 5′-TCGGAGTCAACGGATTTGGT-3′, reverse,
   5′-TTCCCGTTCTCAGCCTTGAC-3′;

   NANOG: forward, 5′-CAATGGTGTGACGCAGAAGG-3′, reverse,
   5′-TGCACCAGGTCTGAGTGTTC-3′;

   OCT4: forward, 5′-CTCGAGAAGGATGTGGTCCG-3′, reverse,
   5′-TGACGGAGACAGGGGGAAAG-3′;

   PECAM1: forward, 5′-AGACGTGCAGTACACGGAAG-3′, reverse,
   5′-TTTCCACGGCATCAGGGAC-3′;

   VE-Cadherin: forward, 5′-CGCAATAGACAAGGACATAACAC-3′, reverse,
   5′-GGTCAAACTGCCCATACTTG-3′;

   VWF: forward, 5′-CCCGAAAGGCCAGGTGTA-3′, reverse,
   5′-AGCAAGCTTCCGGGGACT-3′;

   VEGFR2: forward, 5′-GAGGGGAACTGAAGACAGGC-3′, reverse,
   5′-GGCCAAGAGGCTTACCTAGC-3′;

   PDGFRβ: forward, 5′-ATCAGCAGCAAGGCGAGC-3′, reverse,
   5′-CAGGTCAGAACGAAGGTGCT-3′.

Western blotting

   Cell pellets were lysed using RIPA buffer containing 1% PMSF. Protein
   concentration was measure by the BCA method using a Pierce BCA Protein
   Assay Kit (23225, Thermo Fisher Scientific). Samples were resolved on
   10% sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE), followed by
   transfer to a PVDF membrane at 100v for 120 min. Membranes were blocked
   with 5% non-fat dry milk in 1xTBST at RT for 1 h and then incubated
   with primary antibody at 4 °C overnight. The following antibodies were
   used: mouse anti-CD31 (Abcam, ab9498); rabbit anti-VE Cadherin
   (ab33168); rabbit anti-VWF (ab6994); rabbit anti-BCL6B (Origene,
   TA369826); rabbit anti-FLAG (Abcam, ab205606); and mouse anti-beta
   actin (Abcam, ab8226). After appropriate washing with 1xTBST, the
   membranes were incubated with Goat Anti-Rabbit IgG H&L (HRP) (Abcam,
   ab6721) or Goat Anti-Mouse IgG H&L (HRP) (Abcam, ab6789) at RT for 1 h.
   A chemiluminescent assay was performed using ECL substrates (Abcam,
   ab133406), and signals were detected with a FUSION Solo S
   chemiluminescence imaging system (5200 Muti, Tanon).

Tube formation assay

   hiPSC-ECs were seeded onto 96-well plates coated with 50 µl matrigel
   (BD, 356231) at a density of 2 × 10^4 cells per well and then incubated
   for 4 h. Tubular structures were photographed under a light microscope
   (Zesis) and analyzed using Image J software (NIH, Bethesda, MD).

Luciferase reporter assay

   Luciferase reporter transfection and dual-luciferase assays were
   performed as follows. Briefly, HEK293T cells were seeded at 5 × 10^4
   cells/well in 24-well plates and transfected using Lipofectamine 2000
   (Invitrogen, 11668019) with 0.8 μg reporter vector carrying firefly
   luciferase (Promega, E1751) and the indicated target sequences. The
   PRL-TK vector (carrying Renilla luciferase; Promega, E2241) was
   co-transfected as an internal control. At 24 h after transfection,
   cells were lysed using lysis buffer (Vazyme, DL101-01) and subjected to
   a dual-luciferase assay using the Dual-Luciferase® Reporter (DLR™)
   Assay System (Promega, E1910) according to the manufacturer’s protocol.

RNA sequencing (RNA-Seq) analysis

   A total amount of 1 µg RNA was used for the RNA sample preparations.
   Sequencing libraries were generated using NEB-Next® Ultra™ RNA Library
   Prep Kit for Illumina® (NEB, USA). The clustering of the index-coded
   samples was performed on a cBot Cluster Generation System using TruSeq
   PE Cluster Kit v3-cBot-HS. After cluster generation, the libraries were
   sequenced on an Illumina Novaseq platform. Differential expression
   analysis of two conditions/groups (three biological replicates per
   condition) was performed using the R package DESeq2 (1.16.1). Genes
   with an adjusted P-value < 0.05 found by DESeq2 were assigned as
   differentially expressed. Gene Ontology (GO) enrichment analysis of
   differentially expressed genes (DEGs) was applied using the R package
   clusterProfiler. GO terms of DEGs with corrected P value less than 0.05
   were significantly enriched. We used the R package clusterProfiler to
   assess the enrichment of DEGs in Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathways. The RNA sequencing data in this study were
   archived in supplementary table.

Chromatin immunoprecipitation (ChIP) assay

   After the cells were cross-linked, the ChIP assay was performed using a
   ChIP kit (CST, 9003S). Specifically, for each immunoprecipitation
   reaction, the nuclear fraction from 2 × 10^7 cells and 3 μg of HA-tag
   antibody (CST, 3742S) or control IgG were used. Purified DNA fragments
   were utilized for subsequent PCR amplification with the following
   specific primer sets: forward, 5′-GCAAGTGGATCACTTGAGGT-3′, reverse,
   5′-TGCGATCTCAGCTCACTGCAA-3′. The resultant PCR products (expected
   amplified size is 190 bp) were analyzed by performing 2% (w/v) agarose
   gel electrophoresis.

Statistical analysis

   Results are presented as mean ± SD from a minimum of three independent
   experiments. Group comparisons were conducted using the non-paired
   two-tailed Student's t-test. In cases of non-normal distribution, the
   two-tailed F test was employed. For comparisons involving more than two
   groups, ANOVA was applied. A p value of < 0.05 were considered
   statistically significant.

Results

Generation and characterization of ECs from hiPSCs

   To investigate whether BCL6B plays a functional role in endothelial
   cell (EC) development, a feeder-free, and chemically defined protocol
   (Supplementary figure: Fig. S1A) was adapted in this study to generate
   ECs from human induced pluripotent stem cells (hiPSCs). Throughout the
   differentiation process, cell morphology was gradually changed
   (Supplementary figure: Fig. S1B). Differentiated hiPSC-derived ECs were
   purified using a CD144 antibody before being re-plated for further
   characterization on D7 (Supplementary figure: Fig. S1C). The purified
   cells displayed a typical EC morphology (Supplementary figure: Fig.
   S1C) and were positive for canonical EC markers, VE-Cadherin and CD31
   (Supplementary figure: Fig. S1C). We further confirmed that
   hiPSC-derived ECs shared a comparable capillary-like tube formation
   capacity with HUVECs (Supplementary figure: Fig. S1D). Quantitative
   data revealed similar junction number and tube length between HUVECs
   and hiPSC-derived ECs (Supplementary figure: Fig. S1E), suggesting
   functionally similarities between primary and hiPSC-derived ECs.

   The differentiation of hiPSCs into ECs involved two stages: mesoderm
   induction and EC fate specification. Using quantitative real-time PCR
   (qRT-PCR), we observed a gradual increase in the gene expression levels
   of PECAM1, CDH5, VWF, and BCL6B during the hiPSC-EC differentiation
   process from D0 to D6. (Supplementary figure: Fig. S1F). The protein
   expression of VWF and VE-cadherin exhibited the same increasing pattern
   as their transcriptional levels throughout the hiPSC differentiation
   process. Surprisingly, the protein expression levels of BCL6B followed
   an opposite trend, becoming barely detectable between D5 and D6 of EC
   differentiation when EC markers peaked (Supplementary figure: Fig.
   S1G). This suggests that lower expression of BCL6B at the protein level
   is required for EC differentiation, whereas high expression of BCL6B at
   the protein level could affect EC generation.

BCL6B overexpression suppressed EC differentiation from hiPSCs

   To investigate the involvement of BCL6B in EC differentiation, we
   developed a doxycycline inducible BCL6B-flag expression vector
   (Fig. [75]1A), and established an inducible hiPSC line to control the
   temporal expression of BCL6B upon the addition of doxycycline
   (Fig. [76]1B). Next, we treated the inducible hiPSC line with
   doxycycline to induce BCL6B overexpression throughout the EC
   differentiation process (Fig. [77]1C). The differentiation efficiency,
   assessed by flow cytometry analysis of CD144^+ cells on D6, revealed
   that compared to the differentiation efficiency of control hiPSCs at
   8.55% ± 0.13%, overexpression of BCL6B in hiPSCs resulted in a decrease
   in the differentiation of hiPSCs into ECs, with efficiencies of
   5.54% ± 0.25%. In contrast, the group treated with doxycycline alone
   displayed no significant differences (Fig. [78]1D and E).

Fig. 1.

   [79]Fig. 1
   [80]Open in a new tab

   BCL6B overexpression suppressed ECs differentiation from hiPSCs. A
   Schematic illustration of the doxycycline-induced BCL6B overexpression
   vector using the tetO-on system. B Verification of doxycycline-induced
   BCL6B protein expression in the established inducible
   BCL6B-overexpression hiPSC line (BCL6B OE). C Schematic illustration of
   doxycycline addition to induce BCL6B overexpression. D Representative
   flow cytometry results of CD144^+ cells in hiPSC-ECs from different
   groups. E Statistical analysis of the CD144^+ cells in hiPSC-ECs from
   different groups. **, P < 0.01. F Schematic illustration of doxycycline
   addition at different time points. G The protein expression of VWF,
   CD31, VE-cadherin and BCL6B in D6 ECs derived from BCL6B OE hiPSC
   started to be treated with doxycycline at different time points. H
   Representative flow cytometry results of CD144^+ cells in hiPSC-ECs
   from different groups. I Statistical analysis of the CD144^+ cells in
   hiPSC-ECs from different groups. *, P < 0.05; **, P < 0.01. J
   Representative brightfield images of tube-like structures formed by
   hiPSC-ECs and BCL6B OE hiPSC-ECs (left panel), and quantitative
   analysis of tube length and junction number (right panel). OE,
   overexpression. **, P < 0.01

   Given that the differentiation of EC from hiPSCs involves two stages
   and mesodermal cells express high levels of BCL6B
   (Supplementary figure: Fig. S1G), we further asked if EC
   differentiation efficiency could be temporally controlled by BCL6B
   activation. To test this, we treated the inducible hiPSCs with
   doxycycline at different time points until D6 of EC differentiation
   (Fig. [81]1F). Our western blot analysis confirmed that two days of
   doxycycline treatment were sufficient to achieve high expression levels
   of BCL6B (Fig. [82]1G). Consequently, we observed that overexpression
   of BCL6B significantly decreased the protein expression of VWF,
   VE-cadherin, and CD31 in hiPSC-ECs (Fig. [83]1G). Interestingly,
   regardless of the timing of doxycycline treatment, all groups exhibited
   decreased EC differentiation, although overexpression of BCL6B
   initiated in the early stages tended to show more severe inhibitory
   effects on EC differentiation (Fig. [84]1H and I). Furthermore, the
   overexpression of BCL6B during differentiation significantly attenuated
   the capacity of derived hiPSC-ECs to form a tubular structure,
   suggesting compromised EC function of BCL6B-overexpressing
   hiPSC-derives ECs (Fig. [85]1J). The above results clearly show that
   overexpression of BCL6B inhibits hiPSCs differentiation into ECs.

Downregulation of BCL6B improved EC differentiation from hiPSCs

   To firmly establish a causal role for increased BCL6B activity in
   reduced EC differentiation efficiency and functional properties, we
   further investigated whether inhibiting the expression of BCL6B in
   hiPSCs would enhance their capacity to generate ECs. To achieve this,
   we constructed doxycycline-inducible vectors for effective silencing of
   endogenous BCL6B in hiPSCs upon doxycycline treatment (Fig. [86]2A).
   Subsequently, we initiated EC differentiation in the presence of
   doxycycline at six time points (D0, 1, 2, 3, 4, and 5) during the
   differentiation process (Fig. [87]2B). Our findings revealed that
   knockdown of BCL6B significantly increased EC differentiation, with
   efficiencies increasing from 9.4% ± 0.11% to 15.8% ± 0.23% (Fig. [88]2C
   and D). Furthermore, the highest differentiation efficiency was
   observed when BCL6B inhibition was initiated earlier than D1 of
   differentiation (Fig. [89]2C and D), which was consistent with the
   observation that a more profound inhibitory effect of BCL6B
   overexpression on EC differentiation was achieved at D0 of
   differentiation (Fig. [90]1I). Moreover, knocking down of BCL6B
   significantly increased the protein expression of VWF, VE-cadherin, and
   CD31 in hiPSC-ECs (Fig. [91]2E). However, the derived hiPSC-ECs with
   BCL6B knockdown did not demonstrate enhanced tubular structure
   formation capacity when compared to those without BCL6B inhibition
   during the differentiation process (Fig. [92]2F).

Fig. 2.

   [93]Fig. 2
   [94]Open in a new tab

   Downregulation of BCL6B improved ECs differentiation from hiPSCs. A
   Schematic illustration of the doxycycline-induced BCL6B shRNA vector
   using the tetO-on system. B Verification of doxycycline-induced BCL6B
   protein knock-down in the established inducible BCL6B-knock down hiPSC
   line (BCL6B KD). C Schematic illustration of doxycycline addition to
   induce BCL6B knockdown. D Representative flow cytometry results of
   CD144^+ cells in hiPSC-ECs from different groups. E Statistical
   analysis of the CD144^+ cells in hiPSC-ECs from different groups. *,
   P < 0.05; **, P < 0.01. F Protein levels of VWF, VE-cadherin, CD31, and
   BCL6B in D6 ECs derived from BCL6B KD hiPSC started to be treated with
   doxycycline at different time points. G Representative brightfield
   images of tube-like structures formed by hiPSC-EC and BCL6B KD
   hiPSC-ECs (left panel), and quantitative analysis of tube length and
   junction number (right panel)

BCL6B impaired blood vessel organoids generation

   To further validate the critical role of BCL6B in vessel development
   within a more complex context, we generated vessel organoids (VOs) from
   hiPSCs. Briefly, hiPSCs were assembled into embryoid bodies (EBs) via
   centrifugation in ultra-low attachment 96-well plates, followed by a
   24-h incubation prior to mesoderm induction using CHIR99021 and BMP4
   (Supplementary figure: Fig. S2A). After 3 days, the differentiated EBs
   were treated with VEGF and forsklin for 2 additional days for further
   vascular lineage development. Subsequently, the EBs were embedded in
   Matrigel:Collagen I (1:1) gels to promote EC sprouting and structure
   maturation. Brightfield imaging of VOs showed a significant increase in
   the size of VOs throughout the differentiation process
   (Supplementary figure: Fig. S2B). Confocal imaging revealed the
   formation of complex, interconnected vascular networks by CD31^+ ECs
   (Supplementary figure: Fig. S2C). Moreover, these self-organizing 3D
   VOs exhibited proper localization of pericytes, defined by the
   molecular markers ⍺-SMA and PDGFRβ. Additionally, the vessel-like
   structures were enveloped by a basement membrane and mesenchymal
   fibroblasts, as confirmed by immunostaining for Collagen IV (Col IV)
   and Vimentin, respectively (Supplementary figure: Fig. S2D).

   We further performed qRT-PCR to determine the expression of
   pluripotency and vascular-specific marker genes in VOs. At D15 of
   differentiation, hiPSC-VOs were negligible for OCT4, NANOG, and SOX2,
   and showed drastic upregulation of EC markers (PECAM1, CDH5, VWF, and
   VEGFR2) and a pericyte marker (PDGFRβ) (Supplementary figure: Fig.
   S2E). In line with this, our flow cytometry results revealed that
   approximately 33.4% of total cell population in the hiPSC-VOs were
   CD144^+ ECs at D15 of differentiation (Supplementary figure: Fig. S2F).
   To determine whether hiPSC-VOs respond to proinflammatory cytokines
   with a pro-adhesive phenotype, we challenged the hiPSC-VOs with
   lipopolysaccharides (LPS). Immunofluorescence analysis showed a
   significant increase in intracellular adhesion molecule-1 (ICAM1) level
   upon LPS treatment (Supplementary figure: Fig. S2G). Thus, we have
   successfully established a fully structured and functional human VO
   model.

   With the successful establishment of blood VOs, we proceeded to
   generate BCL6B gene-perturbed blood VOs using the previously developed
   inducible hiPSC lines. Remarkably, the control VOs showed a gradual
   increase in the size from 219.9 ± 3.5 µm at D0 to 1067.3 ± 24.6 µm at
   D5 (Fig. [95]3A). In contrast, BCL6B overexpression, particularly after
   D2 of differentiation, significantly repressed the growth of VOs (from
   212.6 ± 3.8 µm at D0 to 717.3 ± 24.2 µm at D5), whereas BCL6B knockdown
   had little effect on the size of VOs (from 220.9 ± 4.5 µm at D0 to
   1073.4 ± 29.2 µm at D5) (Fig. [96]3A and B). BCL6B knockdown
   significantly upregulated the expression levels of CD31, VWF and
   VEGFR2, while BCL6B overexpression downregulated the expression levels
   of VEGFR2 and PDGFRβ in D15 hiPSC-VOs (Fig. [97]3C). Importantly,
   Immunofluorescence staining showed that both Col IV and PDGFRβ proteins
   were significantly downregulated upon BCL6B overexpression in day 15
   hiPSC-VOs (Fig. [98]3D). Collectively, these data strongly suggest that
   BCL6B is a negative regulator in EC differentiation and vascular
   network formation in hiPSC-derived VOs.

Fig. 3.

   [99]Fig. 3
   [100]Open in a new tab

   Overexpression of BCL6B impaired VOs development. A Brightfield images
   illustrating the development of VOs from hiPSCs with or without BCL6B
   gene manipulation over a 6-day differentiation period. B Statistical
   analysis of the diameter of VOs derived from control, BCL6B OE, and
   BCL6B KD hiPSCs. N = 16; ***, P < 0.001. C mRNA expression of vessel
   markers (VWF, CDH5, CD31, VEGFR2 and PDGFRβ) and BCL6B in D15 VOs
   (n = 10). Data normalized to GAPDH expression. D Immunostaining and
   quantification of PDGFRβ and Col IV in D15 VOs. Col IV, Collagen IV. *,
   P < 0.05; **, P < 0.01

BCL6B inhibited EC differentiation from hiPSCs by repressing ETV2
transcription

   To identify the potential downstream targets regulated by BCL6B during
   EC differentiation, we conducted RNA-seq analysis in control (Ctrl) and
   BCL6B overexpression (OE) differentiating hiPSCs at different
   differentiation stages (day 2 and 4) (Fig. [101]4A). RNA sequencing
   analysis showed that compared to day 2 differentiating hiPSCs, there
   were 1191 up-regulated genes and 1818 down-regulated genes,
   respectively, in day 4 differentiating hiPSCs (Fig. [102]4B). Gene
   Ontology (GO) enrichment analysis of differently expressed genes (DEGs)
   between day 4 and day 2 differentiating hiPSCs revealed significant
   enrichments of multiple cell/tissue/system developmental processes and
   cell differentiation (Supplementary figure: Fig. S3A), further
   confirming that our differentiation protocol effectively and
   specifically drives vascular development and EC differentiation.
   Moreover, KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway
   enrichment analysis showed multiple important cellular development and
   differentiation signaling pathways (Rap1, Hippo, stem cell
   pluripotency, Wnt, and Calcium) were activated during EC
   differentiation from hiPSCs (Supplementary figure: Fig. S3B). As
   expected, hundreds of genes were modulated by BCL6B overexpression in
   differentiating hiPSCs, peaking at day 4 (Fig. [103]4B), and
   importantly, a variety of cell/tissue/system developmental and
   differentiation processes (circulatory system and tube development) as
   well as multiple KEGG pathways (MAPK, Rap1, Ras, Calcium, Relaxin, Wnt,
   PI3K-Akt, cAMP, and stem cell pluripotency) were dramatically modified
   by BCL6B overexpression (Supplementary figure: Fig. S3C-3F), supporting
   an important regulatory role for BCL6B in EC differentiation.

Fig. 4.

   [104]Fig. 4
   [105]Open in a new tab

   RNA-seq analysis showing ETV2 inhibition by BCL6B during EC
   differentiation from hiPSCs. A Schematic illustration of doxycycline
   addition during EC differentiation and sample collection for analysis.
   B Bar graphs show upregulated and downregulated gene numbers in control
   (Ctrl) and BCL6B overexpression (OE) differentiating hiPSCs at day 2
   and 4 respectively. C Heatmap showing expression levels of the
   ETS-domain transcription factors in differentiating hiPSCs with
   indicated treatments. D mRNA expression of BCL6B and ETV2 during EC
   differentiation from control or BCL6B OE hiPSCs upon doxycycline
   treatment as shown in A. Data normalized to GAPDH expression. *,
   P < 0.05; ***, P < 0.001. E mRNA expression of BCL6B and ETV2 during EC
   differentiation from control or BCL6B KD hiPSCs upon doxycycline
   treatment as illustrated in A. Data normalized to GAPDH expression. *,
   P < 0.05; **, P < 0.01; ***, P < 0.001

   Several transcription factors, particularly the ETS-domain
   transcription factors, have critical roles in vascular EC
   differentiation or vascular development [[106]26, [107]27]. These
   factors include ETS proto-oncogene 1 (ETS1), ETS proto-oncogene 2
   (ETS2), ETS transcription factor ERG, ETS transcription factor ELK3,
   Fil1, ETS variant transcription factor 2 (ETV2) [[108]18, [109]28] and
   ETS variant transcription factor 6 (ETV6). It therefore plausible for
   us to speculate that BCL6B mediates EC differentiation and vascular
   development by regulating ETS-domain transcription factors. Indeed,
   heatmap showed that while majority of ETS-domain transcription factors
   (ELK3, ERG, ETS1, ETS2, ETV1, ETV2, ETV5, ETV6, and FLI1) were
   significantly increased during EC differentiation from hiPSCs, two of
   them (ELK4 and ETV4) were decreased, with the most significant changes
   at day 4 (Fig. [110]4C). Crucially, we observed an opposite trend for
   most of these genes upon BCL6B overexpression (Fig. [111]4C), inferring
   a negative regulatory role for BCL6B in these genes during EC
   differentiation from hiPSCs. Then we performed qRT-PCR analysis to
   confirm above observation. We verified that the induction of BCL6B
   overexpression in hiPSCs through doxycycline significantly decreased
   the mRNA expression of ETV2 during the EC specification stage
   (Fig. [112]4D), which was consistent with the result of RNA-seq.
   Previous study showed that ETV2 activation facilitates EC
   differentiation from hiPSCs [[113]29]. We found that the abrogation of
   BCL6B expression significantly enhanced the transcription of ETV2
   during EC differentiation (Fig. [114]4E).

   It has been well-known that BCL6B is a transcription repressor. To
   further investigate whether BCL6B is capable of binding to ETV2
   promoter and repress its transcription, we first analyzed the ETV2
   promoter using JASPAR website ([115]https://jaspar.elixir.no/), and
   found that a putative binding site of BCL6B upstream of the ETV2
   promoter region between − 1145 bp and − 1129 bp. We then generated
   three different ETV2 promoter-luciferase constructs, one carrying a
   2-kb ETV2 promoter (pGL3-ETV2, Fig. [116]5A), another one carrying a
   1.1-kb truncated promoter by deleting the BCL6B binding site
   (pGL3-ETV2-del, Fig. [117]5B), and the third one carrying a 2.0 kb
   promoter but in which the BCL6B binding site is mutated (pGL3-ETV2-mut,
   Fig. [118]5C). We next examined the effect of BCL6B on the ETV2
   luciferase reporter in 293 T cells, and found that BCL6B overexpression
   dramatically decreased the full-length ETV2 gene promoter activity. In
   response to BCL6B overexpression, both constructs carrying either a
   truncated (pGL3-ETV2-del) or a mutated (pGL3-ETV2-mut) ETV2 gene
   promoter exhibited a much higher level of promoter activity, compared
   to pGL3-ETV2 (Fig. [119]5D). Importantly, a much lower ETV2 gene
   promoter activity was observed in both the truncated (pGL3-ETV2-del)
   and the mutated (pGL3-ETV2-mut) ETV2 gene promoter vector in absence of
   BCL6B overexpression, suggesting that BCL6B binding element within ETV2
   gene promoter is important for ETV2 gene transcription at base level.
   To further confirm the binding of BCL6B to ETV2 promoter, we
   constructed BCL6B overexpressing vector carrying a HA-tag and infected
   differentiating hiPSCs with this vector for 4 days, followed by
   ChIP-PCR analysis using an anti-HA antibody (Fig. [120]5E and F). The
   expected size (190 bp) of ETV2 promoter amplicon containing the BCL6B
   binding element was detected in the formaldehyde-mediated DNA–protein
   complexes immunoprecipitated with anti-HA antibody (Fig. [121]5G). On
   the other hand, no signal was detected when normal rabbit IgG was added
   (Fig. [122]5G). Our results show that BCL6B is able to bind to ETV2
   promoter and suppresses ETV2 gene expression.

Fig. 5.

   [123]Fig. 5
   [124]Open in a new tab

   BCL6B transcriptionally inhibited ETV2 expression. A–C Schematic
   illustration of three luciferase reporter vectors carrying the
   full-length (A), truncated (B) or mutated (C) ETV2 gene promoters. D
   Luciferase activity in 293 T cells after transfection with luciferase
   reporter vectors, with or without the BCL6B overexpression vector for
   24 h. The renilla luciferase vector (pRL-TK) was co-transfected as a
   control. **, P < 0.01; ***, P < 0.001. E Schematic illustration of
   BCL6B with HA-tag overexpression vector. F Verification of BCL6B
   protein expression in hiPSCs as determined by western blot. G ChIP-PCR
   assay was performed to determine the binding of BCL6B to ETV2 promoter

   Next, to further identify the role of BCL6B-ETV2 signal axis in EC
   differentiation from hiPSCs, we firstly conducted immunofluorescence
   staining of BCL6B and ETV2 in hiPSC-derived ECs. The result showed both
   BCL6B and ETV2 were mainly expressed in nuclei in hiPSC-derived ECs
   (Fig. [125]6A) and HUVECs (Supplementary figure: Fig. S4A). To further
   confirm the co-expression of BCL6B and ETV2 in hiPSC-derived ECs, we
   generated a lentiviral vector with BCL6B overexpression along with EGFP
   and transfected it into hiPSCs, followed by EC differentiation
   (Supplementary figure: Fig. S4B). EGFP^+ differentiated ECs
   (BCL6B^+/CD144^+) were sorted from differentiating hiPSCs at day 6
   (Supplementary figure: Fig. S4C) and subjected to qRT-PCR analysis
   (Supplementary figure: Fig. S4D) and immunofluorescence staining assay
   (Supplementary figure: Fig. S4E), respectively. Data from qRT-PCR
   analysis showed that ETV2 gene expression was significantly
   downregulated upon BCL6B overexpression in sorted EGFP^+ differentiated
   ECs (Supplementary figure: Fig. S4, C and D). Immunofluorescence
   staining also confirmed co-expression of BCL6B and ETV2 in
   hiPSC-derived ECs (Supplementary figure: Fig. S4E).

Fig. 6.

   [126]Fig. 6
   [127]Open in a new tab

   The inhibitory effect of BCL6B overexpression on EC differentiation was
   rescued by ETV2 overexpression. A Immunostaining of BCL6B and ETV2 in
   hiPSC-derived ECs. B Protein expression levels of BCL6B and ETV2
   throughout EC differentiation from D0 to D6, as determined by western
   blot. C Schematic illustration of ETV2 overexpression vector. D
   Verification of ETV2 protein expression in hiPSCs as determined by
   western blot. E Representative flow cytometry results of CD144^+ cells
   in hiPSC-ECs from different groups. F Statistical analysis of the
   CD144^+ cells in hiPSC-ECs from different groups. **, P < 0.01; ***,
   P < 0.001

   Additionally, while BCL6B expression was gradually decreased during EC
   differentiation from hiPSCs, we observed a gradual increasing protein
   expression of ETV2 during same period (Fig. [128]6B). Subsequently, we
   constructed lentiviral vector with ETV2 overexpression (Fig. [129]6C,
   [130]D), and conducted a co-gene transduction during EC differentiation
   from hiPSCs. As expected, while BCL6B and ETV2 overexpression alone
   significantly reduced and increased EC differentiation from hiPSCs, the
   inhibitory effect of BCL6B overexpression on EC differentiation was
   largely rescued by ETV2 overexpression (Fig. [131]6E, [132]F).
   Collectively, these data strongly suggest that BCL6B suppresses EC
   differentiation through inhibiting ETV2.

Discussion

   ECs derived from hiPSCs with their unlimited expansion potential, are
   considered an alternative cell source for experimental studies, and
   especially, therapeutic approaches [[133]30]. Generally, there are 3
   primary methods for EC derived from differentiating embryonic stem
   cells (ESCs) or iPSCs [[134]31]. First, ESC and iPSC can be
   differentiated into 3D EBs, in which spontaneously undergo
   differentiation into the three germ layers: ectoderm, endoderm, and
   mesoderm, with emergence of numerous lineages of cells. Addition of a
   variety of growth factors promotes EC differentiation within the EB. In
   a second method, coculture of differentiating ESC or iPSC with other
   stromal cells to promote EC lineage differentiation from the mesoderm.
   Finally, EC differentiation performed in 2D culture on cell culture
   plates coated with Matrigel or other proteins using specific culture
   mediums with sequentially added recombinant growth factors. Several
   independent laboratories have successfully generated ECs from hiPSCs or
   human pluripotent stem cells (hPSCs). Amy Cochrane. et al. [[135]32]
   reported that the efficiency of EC differentiation from hiPSCs was 62%.
   An approximately 15% CD31 + cells was observed in differentiating EBs
   derived from mouse ESCs [[136]33]. Moreover, about 18–33%
   differentiation efficiency was achieved with some hiPSC lines
   [[137]34–[138]36]. In our study using the adapted protocol, we observed
   a basal differentiation efficiency of approximately 10%, which was much
   lower than that (61.8–88%) in original protocol [[139]22], while Yingyi
   Quan et al. [[140]37] reported a 43.5% differentiation efficiency in
   their study when followed the original protocol. Although defined
   conditions have been described, a very low efficiencies of EC
   differentiation (< 1.5%) was also observed in the early protocols
   [[141]38]. Multiple factors, for example, fetal bovine serum (FBS),
   bovine serum albumin (BSA), the quality and source of stem cell lines,
   and growth factors, may attribute to the different EC differentiation
   efficiency reported in various studies.

   BCL6B is reported to be involved in a diversity of biological
   functions, including tumor suppression[[142]4, [143]6, [144]8], immune
   response[[145]3], stem cell self-renew[[146]2], and vascular
   angiogenesis[[147]10], all of which are related to its function as a
   transcriptional repressor. Interestingly, it has been reported that
   BCL6B-deficient mice are viable and no gross anatomical abnormalities
   were observed after birth, with altered T cell activation [[148]39] and
   hematopoiesis [[149]40]. However, later studies showed that BCL6B gene
   disruption in vivo markedly impaired angiogenesis in the mouse retina
   and wounded skin [[150]10], as well as choroidal neovascularization
   lesions and ocular edema in the neuroretina [[151]9], indicating an
   essential role for BCL6B in ocular neovascularization and edema. These
   observations suggest that while BCL6B is critical for
   neovascularization under pathological conditions, it may be redundant
   during embryonic development and under physiological conditions due to
   gene duplication or complementary effects from other paralogous genes.

   Given its role as a transcriptional repressor, we anticipated a broader
   functional impact with BCL6B. Online database indicated a high mRNA
   expression level of BCL6B in human ECs [[152]10], which is consistent
   with our qRT-PCR data. However, at the protein level, the BCL6B
   expression is notably low. This discrepancy adds intrigue to unveiling
   its role in EC development and function. We constructed inducible
   overexpression and knockdown vectors to manipulate BCL6B expression
   during EC differentiation from hiPSCs and VO generation from hiPSCs.
   The data clearly demonstrated that BCL6B inhibits EC differentiation.
   Although we clearly observed that knockdown of BCL6B in hiPSCs
   significantly increased EC differentiation, impaired BCL6B expression
   in hiPSC-derived ECs exhibited negligible effect on tube formation.
   However, Ohnuki et al. reported that BCL6B knockdown abrogated EC
   network formation in HUVECs [[153]10], indicating that BCL6B may play a
   distinct role in primary and hiPSC-derived ECs. Similar to tube
   formation, we also found that BCL6B overexpression impaired the size of
   VOs during the first 5 days of differentiation, whereas knockdown of
   BCL6B had little effect on VO size.

   Mechanistically, we uncovered that BCL6B inhibits ETV2 transcription,
   consequently inhibiting EC differentiation. It has been widely known
   that ETV2 is a master transcriptional regulator required for the
   development of both endothelial and erythropoietic lineages [[154]41].
   Of note, a recent study [[155]42] has shown a differential ETV2
   threshold requirement for endothelial and erythropoietic development.
   Specifically, a low-to-moderate level of ETV2 expression is sufficient
   to induce and sustain EC differentiation, while a higher level of ETV2
   expression is required for erythropoiesis rather than for
   vasculogenesis. Therefore, this interesting finding together with our
   data indicate that fine-tuning regulation of ETV2 gene expression by
   BCL6B is essential for EC differentiation from hiPSCs or maintaining EC
   functions. Moreover, as a transcriptional repressor, BCL6B is expected
   to have numerous binding sites in the genome, and other potential
   targets of BCL6B may contribute to the inhibition of EC differentiation
   from hiPSCs. A comprehensive approach to interrogate the potential
   targets of BCL6B would involve using ChIP-seq technology. Additionally,
   the inconsistency between BCL6B mRNA and protein expression during EC
   differentiation suggests a negative-feedback regulation in BCL6B gene
   regulation during EC differentiation from hiPSCs, which merit further
   investigation.

   ECs line the interior of all blood and lymphatic vessels, playing key
   roles in delivering oxygen and nutrients, regulating blood flow,
   modulating immune cell trafficking and maintaining tissue homeostasis
   [[156]43]. Vascular EC dysfunction is central to the progression of
   most chronic conditions, including ischemic heart disease [[157]44] and
   stroke [[158]45], the top two global causes of mortality. A published
   study has demonstrated that BCL6B contributes to ocular vascular
   diseases via notch signal silencing [[159]9].

Conclusion

   In conclusion, this study is the first time to demonstrate that BCL6B
   can modulate EC differentiation as well as blood VO development, and
   that BCL6B exerts its modulatory effect on EC differentiation
   programmed by fine-tuning regulation of ETV2 gene transcription. Our
   study reveals a new role of BCL6B in EC development and function,
   making it a potential target for precision therapy of vascular
   diseases.

Supplementary Information

   [160]Supplementary Material 1.^ (11.6MB, xlsx)
   [161]Supplementary Material 2.^ (2.2MB, pdf)

Acknowledgements