Identification of Basp1 as a novel angiogenesis-regulating gene by multi-model system studies

Abstract

We have previously used the genetic diversity available in common inbred mouse strains to identify quantitative trait loci (QTLs) responsible for the differences in angiogenic response using the corneal micropocket neovascularization (CoNV) assay. Employing a mouse genome-wide association study (GWAS) approach, the region on chromosome 15 containing Basp1 was identified as being significantly associated with angiogenesis in inbred strains. Here, we developed a unique strategy to determine and verify the role of BASP1 in angiogenic pathways. Basp1 expression in cornea had a strong correlation with a haplotype shared by mouse strains with varied angiogenic phenotypes. In addition, inhibition of BASP1 demonstrated a dosage-dependent effect in both primary mouse brain endothelial and human microvascular endothelial cell (HMVEC) migration. To investigate its role in vivo, we knocked out basp1 in transgenic kdrl:zsGreen zebrafish embryos using a widely adopted CRISPR-Cas9 system. These embryos had severely disrupted vessel formation compared to control siblings. We further show that basp1 promotes angiogenesis by upregulating β-catenin gene and the Dll4/Notch1 signaling pathway. These results, to the best of our knowledge, provide the first in vivo evidence to indicate the role of Basp1 as an angiogenesis-regulating gene and opens the potential therapeutic avenues for a wide variety of systemic angiogenesis-dependent diseases.

1 |. INTRODUCTION

Angiogenesis is a process of forming new blood vessels from existing vessels and plays a key role during development and regeneration. Unregulated angiogenesis is seen in the leading causes of many human diseases such as cancer, rheumatoid arthritis, cardiovascular disease, and age-related macular degeneration (AMD). Genetic variability in genes that regulate angiogenesis can influence the susceptibility to and progression of angiogenesis-dependent diseases where angiogenesis has been shown to directly affect their pathophysiology or etiology.^1–3 For instance, polymorphisms in a number of angiogenic regulators such as vascular endothelial growth factor (VEGF) or basic fibroblast growth factor (bFGF) have been shown to be the contributors to susceptibility and progression in a number of cancers and also ocular diseases such as AMD.^1,4–6 The identification of genetic factors that influence angiogenesis will provide valuable information regarding the pathophysiology of angiogenic-dependent diseases and has the potential to serve as prognostic tests in various of these disorders.

We previously used the genetic diversity available in common inbred mouse strains to identify multiple quantitative trait loci (QTLs) responsible for differences in angiogenic response induced by bFGF in the corneal micropocket eye assay (CoNV).⁷ We performed the CoNV assay on 42 different age-matched males of inbred mouse strains and observed a diverse range of measured new vessel area among the tested strains.⁷ Given the recent sequencing of most common inbred mouse strains,⁸ we performed a genome-wide association study (GWAS)⁷ to identify novel QTLs responsible for the variation in angiogenic response to bFGF. To correct for population structure and genetic relatedness among inbred strains, we applied the efficient mixed model association (EMMA) method in our GWAS to assess the association of potential variants to angiogenic response.⁹ We then identified five novel loci with genome-wide significance (P < .05) on multiple chromosomes.⁷ Here, we further refined the mapping of the area with genome-wide significance on chromosomes 15 and used both expression analyses and in vivo functional modeling in zebrafish. We successfully identified Basp1, brain abundant membrane attached signal protein 1, as a major contributing gene in the regulation of angiogenesis among the strains that were phenotyped in our assay.⁷

BASP1 (also known as NAP22) was originally identified in brain extracts and it is particularly abundant in nerve terminals during brain development and it is known to be involved in axon regeneration.^10,11 BASP1 is also expressed in various other tissues,^12,13 but its precise functions are still unknown. The characterization of BASP1 and its associated pathways could lead to new avenues for its therapeutic use against a broad spectrum of angiogenesis-dependent diseases.

2 |. MATERIALS AND METHODS

2.1 |. Total RNA/protein extraction and expression analysis

Total RNA was extracted from corneas as previously described.⁷ Briefly, corneas from five mice per group were harvested and immediately stored in RNAlater RNA Stabilization reagent (Qiagen). Each cornea was trimmed excluding the limbus and was homogenized using PRO200 homogenizer (PRO Scientific). The total RNA was isolated from dissected corneas using RNeasy Kit (Qiagen). Total RNA of 5 μg for each sample was reverse transcribed with the Superscript First-strand cDNA synthesis kit (Invitrogen) using a random primer. Quantitative real-time PCR (qRT-PCR) for each gene (listed in Supplemental notes) was performed in triplicate using TaqMan Gene Expression assays (Life Technologies). The relative level of each RNA sample is calculated using the ΔΔCt method normalized to the corresponding housekeeping gene Gapdh and 18S rRNA levels. Student’s t-test comparing C57BL/6J as reference control and different strains were performed. Statistical significance was defined as P < .05. For protein extraction, cornea samples were trimmed including the limbus and then, appropriate amount of T-PER® Tissue Protein Extraction Reagent (Thermo Fisher Scientific) containing Thermo Fischer Scientific Halt Protease Inhibitor Cocktail, EDTA-Free (Product No. 87785) was added. After homogenization, total sample proteins were determined using Bicinchoninic Acid Assay (BCA). A total of 5 μg of protein samples were mixed with 6X, SDS sample buffer (Boston BioProducts), heated to 100°C for 5 minutes, and then, loaded on a 4%−20% gradient SDS PAGE gel (Bio-Rad) It was then resolved and electro-blotted onto a PVDF membrane. Blots were blocked with 5% nonfat milk and incubated overnight at 4°C with the appropriate antibody (primary antibodies and dilutions are listed in supplemental notes). The blots were then washed, incubated with a 1:10 000 dilution of corresponding HRP-conjugated secondary antibody, and labeled proteins were detected using enhanced chemiluminescence. Bands were detected and quantified using a ChemiDOC MP gel imaging system (Bio-Rad).

2.2 |. Transient gene silencing and in vitro functional assays

C57BL/6J Mouse brain primary endothelial cells (Cell Biologics) were cultured in endothelial cell medium with growth factors (Cell Biologics) according to the vendor’s instructions and used only at early passages (before passage 5). Endothelial cells were characterized prior to any functional assays for the expression of endothelial-specific markers using antibodies VE-cadherin (CD144) and CD31 (BD BioScience) by immunofluorescence staining. Human microvascular endothelial cells-dermal (HMVEC-d) (Lonza) were cultured in EGM™−2MV BulletKit (Lonza) according to the vendors’ instructions and used only at early passages (before passage 5; cells were characterized prior to any functional assays by Western blot using CD31 (BD BioScience) and VEGFR2 (Cell Signaling) antibodies, specific markers to endothelial cells). Endothelial cells were seeded in 60 mm plates and transfected with high-grade quality FlexiTube Genesolution siRNA for mouse and human BASP1 (PMID: NM 022300 and NM_001271606, respectively; Qiagen) using Lipofectamine RNAiMAX (Life Technologies) according to the manufacturer’s protocol. As negative control, we used non-targeting control siRNA, which has no known target mRNA and is shown to have no effect on the cells (Negative Control siRNA, Qiagen). Protein knockdown was assessed by western blot for 72 hours after transfection. For migration assay, polycarbonate transwell inserts, 6.5 mm diameter with 8.0 μm pores, were coated with 50 μL of fibronectin (20 μg/mL) (BD BioScience). Cells were harvested and resuspended in EBM-2 (Lonza) containing 0.1% of BSA (Bovine Serum Albumin) (Thermo Fisher Scientific). Each well was plated with 20 000 cells containing full serum media or EBM-2 containing 0.1% of BSA as negative control. Cells were allowed to migrate for 4 hours. Cells on the top of the membrane were removed using cotton-tipped applicators and membranes were processed using Diff-Quick (Dade Diagnostics). Following staining, membranes were rinsed in PBS and removed from the insert using a scalpel. Membranes were then mounted on slides, and the number of cells in a microscopic field was counted manually. For proliferation assay, both mouse primary endothelial and HMVEC-d cells were plated separately at 2000 cells/mL into 96-well culture plates containing cell culture media. Cells were incubated at different time points (24, 48, and 72 hours). Cell proliferation was evaluated using either Cell counting kit-8 (Dojindo) or the CyQUANT® Cell Proliferation Assay Kit (Molecular Probes). Fluorescence measurements were made using a microplate reader with excitation at 485 ± 10 nm and emission detection at 530 ± 12.5 nm. Student’s t-tests comparing wild-type controls and siRNA-treated cells were performed. Statistical significance was defined as P < .05. For tube formation assay, 10 000 primary endothelial cells were seeded onto cultured growth factor-reduced Matrigel in 1 μ-Slide Angiogenesis (ibidi, Gräfelfing, Germany) at a density of 2 × 10⁵ cells/mL according to the vendor’s instructions. Tube formation was examined under a phase-contrast microscope after 18 hours incubation in cell culture chamber. To quantify the tube length, four different regions within the well were imaged (5× objective), and the length of each tube was measured using ImageJ software.

2.3 |. Zebrafish maintenance, gene editing, and caudal fin angiogenesis assay

Zebrafish were maintained and handled according to our vertebrate animal protocol that has been approved by Boston Children’s Hospital Animal Care Committee and includes detailed experimental procedure for all in vivo experiments described in this paper. Zebrafish embryos were cultured in “egg water” consisting of 0.03% sea salt and 0.002% methylene blue. For generating basp1^−/−, we used CRISPR-Cas9 assay in transgenic kdrl:zsGreen zebrafish. The gRNA sequences used in this experiment are listed in Supplemental notes. Briefly, Alt-R(R) crRNA and tracrRNA were purchased from IDT (Coralville, IA) and resuspended to 100 μmol/L final concentration. Pooled gRNA solution was created with a final concentration of 3 μmol/L crRNAs and 3 μmol/L tracrRNA in IDT duplex buffer and annealed by heating to 95°C for 5 minutes and allowed to cool to room temperature. Cas9 2.0 protein (Thermo Fisher Scientific, Rockford, lL), final concentration of 385 ng/μL, was incubated with crRNA:tracrRNA duplexes at 37°C for 15 minutes. Three nanoliters of this mix was injected directly into the cell of one-cell stage embryos. For each round of injections, on average two clutches were injected with 100 embryos injected per clutch. Average survival rate of embryos from 0 days post fertilization (dpf) to 2 dpf was 75%. For protein extraction, 10 age-matched zebrafish wild type and mutant were separately deyolked at 72-hour post fertilization as previously described.¹⁴ Then, appropriate amount of T-PER® Tissue Protein Extraction Reagent (Thermo Fisher Scientific) containing Thermo Fischer Scientific Halt Protease Inhibitor Cocktail, EDTA-Free (Product No. 87785) was added to each sample prior to homogenization. For caudal fin angiogenesis assay, tail fins were amputated as previously described.¹⁵ Briefly, fish were anesthetized in 0.6 mmol/L of Tricaine and amputated using razor blades. For quantification, the first week of post amputation (wpa) full caudal fin regenerated area were measured using ImageJ software. The regenerated area was normalized to the original size of the total fin prior to the amputation.

2.4 |. Statistical analyses

For Western blot experiments, the highest signal detected for the β-actin on the blot was selected; the value of this band was used to normalize the rest of the β-actin bands on the blot to calculate the normalization factor for each lane. To calculate the normalized signal of each sample on the blot, the observed signal intensities of protein of interest was divided by its corresponding β-actin lane normalization factor. Statistical comparisons between two groups were conducted by unpaired, two-tailed t-test. Data are represented as mean ± SD. A value of P < .05 was considered to indicate statistical significance.

3 |. RESULTS

3.1 |. Characterization of genome-wide significant peaks on chromosome 15

We previously identified two different genome-wide significant association peaks on Chromosome 15, one is located at 24 682 428 bp (SNP rs33886061) and another at 96 431 470 bp (SNP rs50093952).⁷ Within each association peak, we selected candidate genes that either were coding nonsynonymous SNPs with potential functional consequences or had SNPs that were present within the promoter or untranslated regions that could alter the gene expression. We previously identified, Slc38A1 (Solute Carrier Family 38, Member 1, PMID: 608 490) that bears a coding-nonsynonymous SNP (rs50093952) in an amino acid residue that is conserved in all mammalian species.^7,16 This gene has already been implicated in the regulation of angiogenesis,^16–18 serving as validation of our approach. To investigate the other genome-wide significant association peak on Chromosome 15, we generated the gene expression profiles from unstimulated corneas (no pellet) and stimulated corneas with 20 ng bFGF pellet (10 eyes, excluding the limbus, pooled from 5 mice) of 4 different strains with varying angiogenic responses (SM/J, C57BL/6J, AKR/J, and 129S1/SvImJ). We then examined the mRNA expression levels of eight unique reference sequences within this region by quantitative RT-PCR. Based on the expression data and proximity to the highest-scoring SNP (rs33886061) in the GWAS, we focused on Basp1 as a possible candidate as a novel angiogenesis-regulating gene. Haplotype analysis of baseline unstimulated corneas demonstrated that Basp1 baseline expression was lower in SM/J and 129S1/SvImJ mouse strains with haplotype blocks containing the rs33886061 SNP (Figure 1A). Additionally, when bFGF stimulated corneas of AKR/J mice were compared to 129S1/SvImJ, there was a greater increase in protein expression of Basp1 in the AKR/J corneas (Figure 1B,C).

FIGURE 1.

Gene expression of Basp1 within the statistically significant region on Chromosome 15 identified by EMMA. A, Expression data of Basp1 show a strong correlation with haplotypes among inbred mice. Data represent mean ± standard error of the mean (SEM) mRNA expression levels from 10 age-matched corneas. B, Western blot of unstimulated cornea from 129S1/SvImJ and AKR/J verifying the RT-PCR experiments. Notably, a greater increase in protein expression of BASP1 in bFGF stimulated corneas. C, Quantification of BASP1 protein levels. Each sample was normalized to its β-actin. * and **P < .05 and P < .02, respectively; two-sided Student’s t-test. Bars indicate mean ± SD

3.2 |. BASP1 inhibition alters the endothelial cell activity in vitro

In order to determine whether BASP1 expression levels alter the endothelial cell function, we assessed the effect of transient BASP1 knockdown by siRNA on proliferation, migration, and Matrigel-based tube formation of C57BL/6J mouse primary brain microvascular endothelial cells. We observed a significant decrease in the ability of primary endothelial cells to migrate to full serum-containing media in Basp1 siRNA-transfected cells compared to scrambled control siRNA-transfected cells (Figure 2A). In addition, we observed a lack of tube formation and significant decrease in capillary lengths in Basp1 siRNA-transfected cells compared to scrambled-control siRNA-transfected endothelial cells (Figure 2B). However, we did not observe a difference in endothelial cells proliferation between the two experimental groups (Figure 2C). We confirmed the knockdown of Basp1 by Western blot (Figure 2D). In order to show the potential for this gene to be involved in angiogenesis in humans, we assessed the effect of transient BASP1 knockdown by siRNA on both proliferation and migration of human microvascular endothelial cells (HMVECs). We observed a similar result in the HMVEC as in mouse primary endothelial cell functional studies (Figure 2E,F) suggesting the conservation of BASP1 function between human and mouse, as well as the specificity of BASP1 function in endothelial cells.

FIGURE 2.

In vitro endothelial cell functional assay. A, C57BL/6J primary brain microvascular endothelial cells migration in response to full serum media (FSM) was significantly decreased when cells were transfected with Basp1-specific siRNA compared to scramble-siRNA control (representative figure from three independent experiments). Initially 20 000 cells were seeded per well and basal medium containing 0.1% BSA was used as negative control. *P < .01. B, Lack of tube formation and a significantly decrease in capillary lengths in Basp1 knockdown in primary endothelial cells. A lumen-like structure was observed in scramble siRNA-transfected endothelial cells. C, Primary endothelial cell proliferation did not change significantly when cells were transfected with Basp1-specific siRNA compared to scramble siRNA control. Started with seeded 10 000 cells per well using Cell counting kit-8 (Dojindo). P < .05; two-sided Student’s t-test. Bars indicate mean ± SD. D, Western blot analyses show a significant knockdown of Basp1 in primary endothelial cells using siRNA. E, HMVEC migration in response to FSM was significantly decreased when cells were transfected with BASP1-specific siRNA compared to scramble siRNA control (representative figure from three independent experiments: Figure S1). F, HMVEC proliferation did not change significantly when cells were transfected with BASP1-specific siRNA compared to scramble siRNA control (CyQUANT® Cell Proliferation Assay Kit; Molecular Probes). Western blot of HMVECs after siRNA transfection show a significant knockdown of BASP1 in HMVECs (Figure S1)

3.3 |. A basp1 haploinsufficiency impairs angiogenesis in developing zebrafish

Using a widely adopted CRISPR-Cas9 system, we generated basp1 null mutants in kdrl:zsGreen zebrafish. We used the online tool CHOPCHOP (http://chopchop.cbu.uib.no/index.php)¹⁹ for selecting specific target sites and designed gRNAs targeting exon 1 and 2. Custom guide RNAs together with Cas9 protein were injected into zebrafish embryos at one-cell stage as previously described in 20. Injected embryos were raised to adults and genotyped by PCR to determine if the CRISPR targeting caused site-specific mutations. Animals with mutations were then crossed to WT fish and the progenies were analyzed to determine if the lesion had been transmitted through the germ line. We confirmed germ line transmission and recovered multiple stable basp1 mutant fish lines. Significant vascular defects and gaps in the formation of intersegmental vessels (ISVs) (Figure 3A,B) were observed at 72-hours post fertilization (hpf) in offspring from in-crosses of a basp1 mutant line containing a small deletion in exon 1. We then confirmed that mutants with the phenotypes at 3 dpf were homozygous for the basp1 haploinsufficiency by both RT-PCR analysis of gene expression and Sanger sequencing of the amplicons of the genomic DNA (Figure 3C,D). In order to further confirm the genotype-phenotype correlations, off-springs from mating heterozygotes of the basp1 mutant zebrafish line were genotyped at 72 hpf; in total, 607 fish were analyzed. All confirmed wild-type and heterozygous mutant zebrafish did not display any vascular defects; however, all confirmed homozygous basp1^−/− had the vascular defects. Based on the Mendelian ratio, homozygous mutant fish were underrepresented in clutches of all heterozygous in-crosses in the genotyped fish live at 72 hpf (Figure 3E). Our data suggests survival of only ~51% of the homozygous embryos at 72 hpf when compared to the number of wild-type siblings in heterozygous in-crossings.

FIGURE 3.

basp1 haploinsufficiency causes vascular defects at 72 hpf. A, Off-springs from mating heterozygotes show strong phenotype-genotype correlation. Wild-type and heterozygous mutant zebrafish did not show any vascular defects. Significant intersegmental vascular (ISV) defects in basp1^−/− compared to wild-type offspring control siblings (representative figures from three crosses). Note the gaps present in the intersegmental vessels suggesting missing or abnormal formation. Also note vessel absence in the head (arrows). B, Bar graph of normal and defective ISVs in wild-type and basp1^−/− mutants, showing average ± SD from four independent crossings. Missing and abnormal formation of ISVs are considered as defective vessels. (* and **P < .001 and P < .0001, respectively). C, Sanger sequencing of basp1 mutants show 11 base pair deletion resulting in a premature termination codon, causing a basp1 knockout. D, RT-PCR of 72 hpf basp1^−/− show a significantly lower amount of basp1 transcripts. E, Reduced survival of basp1 homozygous mutant zebrafish assayed at 72 hpf stage. Generation of adult zebrafish from mating of heterozygous adults was genotyped for exon 1 deletion in basp1 gene at 72 hpf. The percentage of fish for each basp1 genotype, wild-type (+/+), heterozygous (+/−), and homozygous mutant fish (−/−) is represented. The total number of live animals genotyped from each mating is shown to the right of each bar

3.4 |. Regenerative capacity is significantly lower in basp1^−/− zebrafish

To assess the effects of basp1 on vascular regeneration, we used a caudal fin regeneration assay in adult basp1^−/− and wild-type zebrafish. We measured the size of regenerated tissue after 7 days post amputation since a substantial part of angiogenesis takes part within this time frame.²¹ We found a significant decrease in overall fin regeneration in basp1^−/− zebrafish accompanied with possibly a disrupted vascular network as evident by hemorrhage (Figure 4A). Impaired vascular regeneration was only seen in basp1^−/− adults, whereas wild-type and basp^+/− zebrafish had a normal blood vessel network within a larger regenerated tissue area (Figure 4B).

FIGURE 4.

Caudal Fin angiogenesis assay. A, Significant decrease in tissue regeneration in basp1^−/− zebrafish compared to that in age-matched wild-type zebrafish. Amputated sites are marked by the dotted lines. Arrows indicate bleeding sites and reduced regenerated areas. B, The areas of regenerated tissue were quantified and normalized based on the size of resected original tissue as described in the Materials and Methods section (*P < .05 vs control, n = 5). C, Total and active β-catenin is significantly downregulated accompanied by a decrease in notch signaling-related genes including Notch ligand dll4 in caudal fins of basp1^−/− zebrafish compared to wild type. D, Quantification of total and active β-catenin and Notch ligand dll4. Each sample was normalized to its β-actin. * and **P < .05 and P < .01, respectively; two-sided Student’s t-test. Bars indicate mean ± SD

3.5 |. Reduced β-catenin and Notch signaling in basp1^−/− zebrafish

Multiple signaling pathways such as Wnt/β-catenin, Fibroblast growth factor (fgf), and Hedgehog (hh) signaling are known to be activated and control different phases of caudal fin regeneration upon amputation.²² In order to identify which pathways basp1 might be affecting, we isolated and extracted proteins from the caudal fins at day 0 and day 7 of both regenerated and non-regenerated tissue of basp1^−/− and age-matched wild-type kdrl:zsGreen zebrafish (n = 5 from each group). We observed a significant decrease in total and active β-catenin in basp1 null mutants compared to wild type (Figure 4C). This was also accompanied with significant decrease in Notch signaling-related gene, such as Notch ligand dll4 (Figure 4C) that is predominantly expressed in the endothelium. We next evaluated the Wnt/β-catenin pathway and Notch signaling in regenerated tissues at 7 days post amputation time and similarly observed a significant decrease in total and active β-catenin and Notch signaling targeted genes in basp1^−/− mutants compared to wild type (Figure S2A,B). We next assessed the protein levels of β-catenin in HMVECs after BASP1 knockdown by siRNA (Figure 5). We also observed a significant decrease in total and active β-catenin levels compared to the control scrambled siRNA-treated HMVECs (Figure 5A,B). This was accompanied by the downregulation of Notch target gene, Hes1, suggesting the downregulation of Notch signaling pathway (Figure 5C). Our data suggests that basp1 plays a role in regulating β-catenin levels in a cell autonomous way in endothelial cells that could affect the downstream regulation of Notch signaling pathways.

FIGURE 5.

Total and active β-catenin protein levels in BASP1 knockdown in HMVECs. We observed a significant decrease in total (A), active β-catenin (B) and Notch target hes1 (C) protein levels when BASP1 is knockdown using siRNA in HMVECs. D, Quantification of total and active β-catenin, and Notch target hes1. Each sample was normalized to its β-actin. * and **P < .05 and P < .01, respectively; two-sided Student’s t-test. Bars indicate mean ± SD

We also evaluated the effect of basp1 haploinsufficiency on β-catenin protein level during early zebrafish development. Off-springs of heterozygous basp1^+/− zebrafish parents were genotyped and pooled based on phenotypes at 72 hpf for protein extraction. Embryos with basp1 haploinsufficiency had significantly reduced β-catenin levels and activity compared to age-matched wild type (Figure 6A). This was accompanied by downregulation of active notch and Notch signaling-related genes, including Notch target hes1 (Figure 6B). However, we found a slightly increase in sonic hh (shh) levels in basp1^−/− compared to wild type (Figure 6C). This upregulation may be a result of compensation for the significant decrease in other key signaling pathways.

FIGURE 6.

Notch and Wnt/β-catenin signaling pathway are suppressed in basp1^−/− zebrafish embryos. A, A significant decrease in active notch1 protein in basp1^−/− zebrafish compared to wild type. B, Total and active β-catenin is significantly downregulated accompanied with decrease in notch signaling-related genes including Notch target hes1 in basp1^−/− zebrafish compared to wild type. C, A slight increase in shh protein expression in basp1^−/− zebrafish. D, Quantification of total and active β-catenin, and Notch target hes1. Each sample was normalized to its β-actin. *P < .05; two-sided Student’s t-test. Bars indicate mean ± SD

3.6 |. Strong correlation between Basp1 expression and total β-catenin level in mouse cornea

We further assessed if Wnt/β-catenin pathway was differentially regulated when Basp1 protein expression levels were significantly different in strains such as 129S1/SvImJ compared to AKR/J strain (Figure 1B). Interestingly, we found a strong correlation between Basp1 protein levels and the protein level and activity of β-catenin in the cornea of these two mouse strains. The total protein level and activity of β-catenin were significantly lower in 129S1/SvImJ strain compared to AKR/J, where Basp1 protein level was significantly higher (Figure 7). This was accompanied by the downregulation of active Notch and Notch signaling-related genes, including ligand DII4 and Notch target hey1 (Figure 7).

FIGURE 7.

Upregulation of Notch and Wnt/β-catenin signaling in AKR/J strain with high Basp1 expression level. AKR/J strain has a significantly higher level in total, active β-catenin and active notch1 protein that is accompanied with increase in expression of notch signaling-related genes including Notch target Hey1 compared to lower expressing strain 129S1/SvImJ (7A,B). C, Quantification of total and active β-catenin, and Notch target hes1. Each sample was normalized to its β-actin. *P < .05; two-sided Student’s t-test. Bars indicate mean ± SD

3.7 |. Upregulation of Wnt5a in basp1^−/− zebrafish

Almost all the offspring of in-crosses of the surviving homozygous basp1 mutant adults displayed major vascular defects at 72 hpf while notably the majority of the offspring survived after 7 dpf and to adulthood (Figure S3). In order to show whether these basp1^−/− adults survived through an alternative β-catenin independent Wnt signaling pathway (non-canonical), we assessed the level of Wnt5a and found a significantly higher level of Wnt5a in basp1^−/− adult zebrafish compared to age-matched wild types (Figure 8A,B). We also observed a similar association in mouse cornea where the 129S1/SvImJ strain, which had a lower level of Basp1, had a significantly higher levels of Wnt5a compared to AKR/J strain (Figure 8C,D). These data suggest that Wnt5a can compensate for the downregulation of canonical Wnt signaling in adult zebrafish and mouse cornea. Notably, we did not observe a significant upregulation of Wnt5a in neither BASP1 siRNA-transfected HMVECs nor the regenerated tissue from adult basp1^−/− zebrafish in caudal fin amputation assay when compared to scramble control siRNA-transfected cells and wild-type zebrafish, respectively (Figure S4).

FIGURE 8.

Wnt5a levels in basp1^−/−. A, Homozygous basp1 zebrafish mutant embryos show a higher level of Wnt5a compared to their wild-type siblings. C, We observed a similar correlation in mouse cornea where a lower expression of Basp1 in 129S1/SvImJ strain results in higher protein levels of Wnt5a compared to high Basp1-expressing AKR/J strain. Quantification of Wnt5a in zebrafish (8B) and mouse cornea (8D). Each sample was normalized to its β-actin. *P < .05; two-sided Student’s t-test. Bars indicate mean ± SD

3.8 |. basp1 activates Wnt/β-catenin signaling by inhibiting cxxc4

Cxxc4 is a negative regulator of the Wnt signaling pathway by directly binding to the PDZ domain of Dishevelled (DVL).²³ Cxxc4 interaction with DVL competes with DVL-AXIN binding where it is critical for β-catenin stability and activation of Wnt signaling.²³ Here, we assessed the level of cxxc4 in basp1^−/− embryos and found a significant increase in cxxc4 protein level in basp1 null mutants compared to age-matched wild type (Figure 9A,B). Our data suggests that cxxc4, an epigenetic silencer of the β-catenin gene, is suppressed in the presence of basp1 that results in the upregulation of Dll4/Notch1 signaling pathway.

FIGURE 9.

Basp1 promotes angiogenesis through downregulating CXXC4. A, Upregulation of cxxc4 in homozygous basp1 mutant. Our data suggests that in the absence of basp1, angiogenesis is downregulated by cxxc4 suppression, an epigenetic silencer of the β-catenin gene, and thus downregulating Dll4/Notch1 signaling pathway. B, Quantification of cxxc4 protein level. Each sample was normalized to its β-actin. *P < .05; two-sided Student’s t-test. Bars indicate mean ± SD

4 |. DISCUSSION

Identification of genes responsible for the genetic heterogeneity of angiogenic responsiveness will advance our understanding and ability to treat angiogenesis-dependent diseases by furthering the basic knowledge of the role of each gene in angiogenic pathways. Further these genetic factors that influence human angiogenesis could serve as diagnostic tests or predictive biomarkers to antiangiogenic therapies in various human disorders. This information could promote earlier intervention and treatment of angiogenic diseases before the occurrence of irreversible changes.

We previously showed that the corneal neovascularization model in common inbred mice strains provides a highly reliable, accurate, and reproducible phenotype that can be used for GWAS-based genetic approaches in order to identify the genetic loci regulating angiogenesis.⁷ Here, we developed a unique strategy of using multi-model systems to confirm the regulatory role of Basp1, identified through our previous GWAS screen,⁷ in angiogenesis. We used the expression analysis in mice, in vitro functional studies in both mouse primary brain microvascular endothelial cells and HMVECs and in vivo functional modeling in zebrafish to successfully establish the role of Basp1 in corneal angiogenesis and notably its critical role in angiogenesis during early zebrafish development.

BASP1 protein has different functions and is expressed particularly highly in the brain. It is found in many cell types within multiple areas of the brain including nerve terminals, glial cells, and endothelial cells.^9,10 BASP1 plays a key role during brain development and is known to be involved in axon regeneration. Basp1 knockout mice have severe abnormalities in the development of the nervous system and early postnatal lethality.²⁴ BASP1 has also been identified in various other tissues and cell types throughout the body¹³ and has been shown to have cell-specific activities and functions.^12,25–29 Furthermore, BASP1 has been identified as a tumor suppressor of a broad range of cancer types.^25–29 In a recent study, it was shown BASP1 is a co-repressor of estrogen receptor α (ERα) in breast cancer patients.²⁸ High levels of BASP1 has been shown to be a poor prognostic factor for cervical cancer and promotes tumor growth.³⁰ BASP1 is also known to bind to Wilms’ tumor (WT1) suppressor and converts it from a transcriptional activator to a repressor where it could affect the expression level of WT1 target genes that are involved in tissue development, growth, and differentiation.¹² WT1 is expressed in multiple tissues in developing embryos, including heart, lung, kidney, and central nervous system.³¹ Notably, BASP1 is co-expressed with WT1 in similar spatial and temporal patterns in many of these tissues, implicating it as a co-factor of WT1 during development.¹² In endothelial cells, WT1 expression is maintained during angiogenesis and it is necessary for vascular formation.³²

WT1 is shown to negatively regulate Wnt/β-catenin pathway during normal testis development³³ Wnt/ β-catenin is an evolutionary conserved signaling system that has strong links to the Notch signaling during vascular development.³¹ Both in vitro and in vivo studies have shown that β-catenin upregulates Dll4 transcription in endothelial cells and strongly increases Notch signaling.³⁴ Notch is activated in endothelial cells via cleavage of Notch intracellular domain (NICD) by γ-secretase, causing the translocation of NICD to the nucleus where it regulates transcription of multiple target genes, such as Hes1 and Hey1.³⁵ Notch signaling plays a key role in angiogenesis and is regulated by several Notch ligands Delta-like (Dll1, Dll3, and Dll4), or Jagged1and Jagged2.^35,36 Dll4 is known to be upregulated in response to VEGF in endothelial cells at the tip region of angiogenic vasculature.³⁷ High expression of Dll4 defines the tip cell phenotype and it is thought to activate Notch and suppress the tip phenotype in neighboring endothelial cells (stalk cells).³⁷ Moreover, Jagged1 promotes proliferation and sprouting phenotype by inhibiting Dll4/Notch signaling in endothelial cells.³⁷ The unique balance and specific selection of ligand binding to Notch receptors among endothelial cells is still not clear.³⁷ The fact that Dll4 and Jagged1 have opposing effects on sprouting angiogenesis,³⁸ may be the result of upstream signals controlling the expression of one or the other ligand. The results from our study suggest that in the absence of Basp1, the levels of Cxxc4 (a negative regulator of the Wnt signaling pathway) are significantly increased where it results in downregulation of total and active β-catenin level in endothelial cells and thus a decrease in Dll4/Notch1 signaling pathway in these cells.

A recent study has shown that BASP1 regulates the Wnt signaling pathway in taste cells,³⁹ confirming our observations. However, our work, to the best of our knowledge, is the first to investigate the role of Basp1 in regulating angiogenesis via Wnt/β-catenin signaling pathway. In this study, we initially found that among common inbred mouse strains Basp1 expression strongly correlates with a cis-genomic variation within the same haplotype. We next evaluated its role in angiogenesis by in vitro functional assays in both primary mouse brain microvascular endothelial cells and HMVECs. Knockdown of BASP1 by siRNA in both mouse primary endothelial cells and HMVECs led to significant decreases in migration and angiogenic activity. In HMVECs, this is accompanied by a significant decrease in total, active β-catenin, active Notch, and Notch target gene HES1 levels compared to ones transfected with scrambled siRNA. Our in vivo functional modeling in zebrafish also supported the role of basp1 in angiogenesis since basp1 deficiency in zebrafish caused significant vascular defects in early development.

The caudal fin in zebrafish is an ideal tissue to study angiogenic response due to its accessibility and fast regeneration.²¹ Angiogenesis plays a key role in this process and is required for regeneration.²¹ Our results demonstrate the critical role of basp1 in zebrafish caudal fin regeneration. We observed significantly impaired vascular regeneration in basp1^−/− adults compared to age-matched wild-type zebrafish. We also observed disrupted revascularization as evident by the leaky vasculature within the amputated area of basp1^−/− adults suggesting defective blood vessel formation that might inhibit caudal fin regeneration. Wnt/β-catenin pathway is known to play a key role in different phases of caudal fin regeneration. Wnt/β-catenin signaling interacts with the fgf pathway to promote proliferation and regenerative cell proliferation in the zebrafish lateral line neuromast.²² Disruption in any of these pathways could adversely alter the regenerative capacity of zebrafish caudal fin.^40–42 Based on our data, we propose that basp1 regulates revascularization through the Wnt/β-catenin pathway and accompanied pathways such as notch signaling. Further studies to elucidate how BASP1 modulates CXXC4, an epigenetic silencer of the β-catenin gene, expression will provide valuable new information in understanding the direct role of BASP1 in angiogenic responsiveness.

Notably, for those homozygous basp1 null mutants who were able to survive to adulthood, we observed an upregulation of non-canonical Wnt pathway that is β-catenin independent. Interestingly, we found a similar pattern of Basp1 protein level and β-catenin activation in mice cornea. In AKR/J mice, Basp1 upregulation is associated with higher levels of total and active β-catenin in the cornea. The high levels of β-catenin are accompanied by higher active notch and notch signaling-related genes such as Notch target gene Hey1. Collectively, these data suggest that Basp1 regulates angiogenesis via regulating the expression of the β-catenin gene in endothelial cells and thus downstream Dll4/Notch1 signaling pathway in angiogenic tip endothelial cells that make up the leading front of vascular sprout.

In summary, our results show that lack of Basp1 expression impairs angiogenesis during early zebrafish development and is associated with declines in Wnt/β-catenin target genes. Basp1 activates the Notch signaling pathway by increasing total and active β-catenin levels. Thus, Basp1 could be a suitable target for antiangiogenic therapy and for the treatment of other disorders that involve aberrant regulation of Wnt/β-catenin pathway.

Abbreviations:

AMD

age-related macular degeneration

BASP1

brain abundant membrane attached signal protein 1

bFGF

basic fibroblast growth factor

CoNV

corneal micropocket neovascularization

DVL

Dishevelled

DLL

Delta-like ligands

EMMA

efficient mixed model association

GWAS

genome-wide association

HMVEC

human microvascular endothelial cell

NICD

Notch intracellular domain

QTL

quantitative trait loci

VEGF

vascular endothelial growth factor

WT1

Wilms’ tumor