香京julia种子在线播放

A total of 420 unrelated sporadic HSCR patients were studied, which included 322 males and 98 females with the male:female ratio of 3.29:1. Diagnosis of HSCR was based on histopathological criteria for HSCR: (1) absence of enteric plexuses with histological evaluation of the aganglionic tract and (2) increased acetylcholinesterase immunohistochemical staining in the nerve fibers. All patients were sporadic cases and had HSCR as the sole phenotype. Patients were classified into three subgroups based on the segment length of aganglionosis: 323 S-HSCR, 58 L-HSCR, and 39 total colonic aganglionosis (TCA). We randomly selected 1,665 gender- and ethnicity-matched healthy individuals who visited Xinhua Hospital for routine health checkup, as controls including 1,281 males and 384 females (the male:female ratio of 3.34:1). HSCR patients and unrelated controls were all Han Chinese and recruited in Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. The study was performed according to principles of the Declaration of Helsinki, and the study protocol was approved by the institutional review board of Xinhua Hospital (IRB: XHEC-WSJSW-2018-029). All participants or their parents signed an informed consent form. Genomic DNA was extracted from peripheral blood leukocytes using the QIAamp DNA Blood Mini Kit according to the manufacturer’s protocol (Qiagen, Hilden, Germany).

Two DSCAM SNP (rs2837770 and rs8134673) implicated to be associated with HSCR in previous studies were selected for replication. Tag SNPs were selected using The Genome Variation Server¹ based on the HapMap CHB (Han Chinese in Beijing) data. We selected 112 tag SNPs including rs2837770 and rs8134673 with the criteria of minor allele frequency (MAF) ≥ 0.01 and r² ≥ 0.8 to cover DSCAM region. A total of 21 SNPs were selected to cover BACE2 gene region with the criteria of MAF ≥ 0.01 and r² ≥ 0.8. We genotyped the 133 tag SNPs to investigate the associations of DSCAM and BACE2 with HSCR susceptibility. SNPs were genotyped using a Fluidigm platform (Fluidigm Corp., CA). Allele-specific fluorescent (FAM or VIC) primers and common reverse primers were employed for genotyping, and EP1 SNP Genotyping Analysis software was used to analyze the data. We placed one duplicate sample to each 96-well sample plate to assess genotyping accuracy.

Association analysis was performed using PLINK 1.09 with additive model (Purcell et al., 2007). The genotype distribution of each SNP was tested for Hardy–Weinberg equilibrium (HWE) in both case and control population. Linkage disequilibrium (LD) structure was examined by Haploview 4.2 program. The functional consequences of the associated SNP were investigated by checking HaploRegv4.1 database (Ward and Kellis, 2011). The study-level significance was P < 0.00038 (0.05/133).

All zebrafish experiments were performed on AB zebrafish in accordance with protocols approved by the Animal care and Use Committee of Xinhua Hospital. To prevent pigment formation, embryos were treated with 0.003% 1-phenyl-2-thiourea (PTU) (Sigma-Aldrich) at 24 h post-fertilization (hpf) (Qiu et al., 2020).

Total RNA was isolated from 15 wild-type zebrafish embryos of different development stages using TRIzol reagent (TaKaRa, Japan). cDNA was synthesized by using the RevertAid Fist Strand cDNA Synthesis kit (Thermo Fisher Scientific, United States). qRT-PCR was performed using SYBR Green (TaKaRa, Otsu, Japan) on a QuantStudio Dx Real-Time PCR Instrument with QuantStudio Dx Software (Applied Biosystems, Foster City, CA, United States). The 18s ribosomal RNA (18-s) gene was chosen as the reference gene (McCurley and Callard, 2008). The relative expression levels of each sample were calculated using the RQ formula (RQ = 2^{–Δ Δ Cq}) (Bustin et al., 2009); and these assays were carried out in three independent triplicates, with final calculations based on the means of triplicate wells. The sequences of primers are shown in Supplementary Table 1.

We constructed the antisense RNA probe against dscama, dscamb, and bace2. Total RNA was extracted from zebrafish embryos at 48 hpf, and cDNA was obtained by RT-PCR. Target fragments of dscama, dscamb, and bace2 were amplified using cDNA as template. The primers are shown in Supplementary Table 2. PCR products were run and isolated on 1.2% agarose gel and purified using a QIAquick Gel Extraction Kit (QIAGEN, Germany). The purified products were cloned into pGEM-T Easy Vector (Promega, United States) and used to synthesize antisense RNA probes labeled by DIG RNA labeling mix (Roche, Penzberg, Germany) (Galicia et al., 2018).

Whole-mount in situ hybridization (WISH) was performed as previously described (Cunningham and Monk, 2018). Zebrafish embryos were fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) overnight at 4°C, then PFA was removed, and 100% methanol (MeOH) was used to store embryos at −20°C for at least 2 h. Embryos were digested with proteinase K for an appropriate time to allow a better tissue penetration. The embryos were prehybridized for 2–4 h in prehybridized solution without probe at 60°C and then hybridized with Hyb (+) containing 50–200 ng of antisense RNA probes overnight at 60°C. After the probes were removed and strictly washed, embryos were blocked for 1–3 h at room temperature and incubated with Anti-Digoxigenin-AP Fab fragments at 4°C overnight. BM purple AP substrate as developing solution was used to detect the hybridization signals by a microscope (SMZ25, Nikon, Chiyoda, Japan).

The pCS2DEST vector containing the human full-length open reading frame sequences of DSCAM was purchased from Addgene (Edie et al., 2018). The full length of zebrafish bace2 was amplified from total RNA, which was obtained from zebrafish embryos at 48 hpf, and the segment was cloned into pGEM-T Easy Vector and verified by sequencing. The primer sequences for bace2 amplification are as follows: the forward primer 5′-ATGCGGCTCTACGGGCTACTGCTACT-3′ and the reverse primer 5′-TCATGGGACAATCCTGACCTGTGGGA-3′. After linearization of plasmids, the mMessage mMachine kit (Ambion, TX, United States) was used for transcription at 37°C for 2 h, and lithium chloride was used for further precipitation and purification of mRNA.

Embryonic microinjection was performed into one- to four-cell embryos via a gas-driven apparatus; then the embryos were incubated in 28.5°C up to 5 dpf; and during this period, egg water was changed twice a day, and dead eggs and shed shells were removed. In overexpression experiments, 100 pg of human DSCAM mRNA and 100 pg of bace2 mRNA were injected. Splice-blocking morpholino oligonucleotides (SBMOs) were designed to knockdown the expression of dscama, dscamb, and bace2 (Supplementary Table 3). The target MO and a standard control MO were purchased from Gene Tools, LLC (OR, United States). MOs were diluted to working concentrations (0.125, 0.25, and 0.375 mM) in sterile double-distilled water, and approximately 2 nl/embryo MO was injected into the blastomeres. In morpholino rescue experiments, 100 pg of DSCAM mRNA/embryo and 4 ng of dscama-MO/embryo were co-injected, 50 pg of DSCAM mRNA/embryo and 2 ng of dscamb-MO/embryo were co-injected, and 100 pg of bace2 mRNA/embryo and 4 ng of bace2-MO/embryo were co-injected.

To verify the effectiveness of the splice blocking MOs, gel electrophoresis of PCR products and qRT-PCR were carried out. Total RNA was extracted from SBMO-injected and control MO-injected embryos at 48 hpf (n = 15) using RNeasy Mini Kit (Qiagen, Hilden, Germany). cDNA was synthesized using RevertAid Fist Strand cDNA Synthesis kit (Thermo Fisher Scientific, United States). We amplified cDNA sequences across the MO target site by PCR, and the PCR products were visualized by gel electrophoresis (Supplementary Table 4). For qRT-PCR analysis, total RNA was isolated from the 48-hpf embryos (n = 40) injected with the control MO, the splicing MOs, and mRNA synthesized in vitro. The qRT-PCR analysis was performed on a QuantStudio Dx Real-Time PCR Instrument with QuantStudio Dx Software (Applied Biosystems), and the expression of the target genes was normalized to the housekeeping gene 18s (Supplementary Tables 5,6). The relative expression of the target genes in the SBMO-injected or mRNA-injected embryos to control MO-injected embryos was determined using methods reported previously (Bustin et al., 2009).

Embryos were raised to 5 dpf at 28.5°C and then fixed overnight by 4% PFA. After a series of washing, decolorization, permeability, and fixation, embryos were blocked with blocking solution for 3 h at 4°C. The HuC/D antibody (Thermo Fisher Scientific, United States) was stained for ENS neurons as previously reported (Sribudiani et al., 2018). Alexa Fluor 488 AffiniPure Goat Anti-Mouse IgG (H + L) antibody (Yeasen, China) was used to incubate with the embryos, and images were acquired by a fluorescence microscope (SMZ25, Nikon, Chiyoda, Japan). The absence or reduction of enteric neurons in the distal intestine was defined as an HSCR-like phenotype (Jiang et al., 2015; Gui et al., 2017). We counted the number of enteric neurons in the most distal six somite lengths from the anal pore.

To explore the potential correlations of DSCAM and BACE2 with members from the RET activation pathway, the EDNRB signaling pathway, transcription factors during ENS formation, and other known disease genes, we construct a network utilized a web-based interface GeneMANIA² (Warde-Farley et al., 2010). The key genes of RET activation pathway (RET, GDNF, GFRA1, GFRA2, and NRTN), EDNRB signaling pathway (EDNRB, EDN3, and ECE1), two transcription factors (SOX10 and PHOX2B), and three genome-wide association study (GWAS)-reported genes (NRG1, SEMA3C, and SEMA3D), together with DSCAM and BACE2, were analyzed using GeneMANIA.

Statistical analyses were performed with GraphPad Prism 8 (GraphPad Software). All results are expressed as the mean ± standard error. A two-sided unpaired Student t test was carried out for statistical comparison between two groups. P < 0.05 was considered statistically significant.

Among the 133 SNPs genotyped, nine were excluded due to poor genotyping success rate (<5%). The remaining 124 SNPs had a genotyping success rate >99% and conformed to HWE (P > 0.01). The genotyping success rate of all individuals was above 98%. The distributions of the allele frequencies of these SNPs in HSCR patients and controls are shown in Supplementary Table 7. None of the 124 SNPs passed quality control and reached the study-level significance (P < 0.00038). Twelve intronic SNPs of DSCAM showed suggestive association, with a P < 0.05. None of the 19 BACE2 SNPs met suggestive significance level (P < 0.05; Table 1 and Supplementary Table 7).

Two SNPs (rs2837770 and rs8134673) that were reported to be associated with isolated HSCR showed no significant association in the current sample (P > 0.05). These two SNPs were in high LD (r² = 0.90, Figure 1). The frequencies of associated allele rs2837770 G and rs8134673 G were both 0.54 in Chinese control samples of a previous study, which was similar to the frequencies in the current controls (rs2837770 G, 0.560; rs8134673 G, 0.567; Table 1). The frequencies of rs2837770 G (0.575) and rs8134673 G (0.581) were higher in the current cases than in controls, which showed the same effect direction as a previous study.

Rs430255 (P_Addtive = 0.0052, OR = 1.36, 95% CI: 1.10–1.68) and rs2837756 (P_Addtive = 0.0091, OR = 1.23, 95% CI: 1.05–1.43) showed the most remarkable association with a P < 0.01. Rs430255 and rs2837756 are located in intron 19 and intron 3 of DSCAM. Functional annotation revealed that rs430255, rs2837756, and SNPs in high LD with rs2837756 all altered the sequences of multiple transcription factor binding motifs (Supplementary Tables 8,9).

We also performed association analysis stratified by the segment length of aganglionosis. Both rs430255 (P_Addtive = 0.0088, OR = 1.38, 95% CI: 1.08–1.76) and rs2837756 (P_Addtive = 0.0021, OR = 1.31, 95% CI: 1.10–1.55) showed a slightly stronger association with S-HSCR than with the total cases. However, these two SNPs were not associated with susceptibility to L-HSCR and TCA. The two reported SNPs (rs2837770 and rs8134673) showed no association in the stratification analysis (Table 2 and Supplementary Table 10).

Our association analysis suggested a role for DSCAM in the susceptibility of HSCR. However, we could not exclude BACE2 as a strong susceptibility gene for three reasons. First, the associated SNPs in DSCAM region might have a long-range regulation effect on BACE2, which is adjacent to DSCAM. Second, rare variants were associated with HSCR risk, which was not assessed in our study. Last, BACE2 has a role in enteric neuron function (Fattahi et al., 2016; Tang et al., 2018). Therefore, we further explored the possible role of both DSCAM and BACE2 in the development of ENS using zebrafish.

We investigated the expression levels of the zebrafish orthologs during early development. A search in the Ensemble database revealed that there were two orthologs of DSCAM, namely, dscama and dscamb, which showed high sequence similarity with the human orthologs (84.9 and 79.5% homology, respectively). Similarly, the sequence of bace2 has 63.1% sequence homology to human BACE2. The qRT-PCR results showed that the expression patterns of dscama and dscamb were similar in the early stages during zebrafish development. The lowest expression levels of the two genes were observed 24 hpf, which increased gradually from 24 to 96 hpf. The expression level of bace2 was lower than that of dscama and dscamb in the embryos of zebrafish (Figure 2A).

We further determined the spatiotemporal expression patterns of the three genes with WISH analysis using antisense nucleic acid probes. The results showed that dscama mRNA was strongly expressed in the midbrain, telencephalon, and diencephalon. At 24 and 48 hpf, widespread dscama expression was observed in the central nervous system (CNS), the hindbrain and spinal cord (Figures 2B,C). The expression pattern of dscamb mRNA was similar to that of dscama (Figures 2F,G). It was known that neural precursors differentiated and began to extend axons and dendrites in these regions (Yimlamai et al., 2005). Of note, dscama and dscamb began to be expressed in the gut tube in zebrafish larvae at 72 hpf (Figures 2D,E,H,I). These results suggested that dscams might play a role in the development of CNS and ENS. We observed that bace2 began to be expressed in the neural crest cells and pigment epithelial cells of the retina from 48 hpf (Figures 2J,K). An expression signal for bace2 was detected in the intestinal primordium at 72 hpf, and the signal was significantly enhanced at 96 hpf (Figures 2L,M). The results indicated that bace2 might have a more significant effect in the intestine than dscams.

DS HSCR is caused by the presence of three copies of total or partial chromosome 21. It is assumed that chr 21 transcripts are overexpressed by about 50% in cells with an extra copy of this chromosome. Recent studies have demonstrated an increased transcript level of the three-copy genes for a subset of them (Lockstone et al., 2007; Prandini et al., 2007; Letourneau et al., 2014; Olmos-Serrano et al., 2016). We, therefore, injected hDSCAM and bace2 mRNA into the embryos of zebrafish to test the effect of gene overexpression on the development of ENS. The qRT-PCR analysis showed that the expression of target genes was markedly increased (Figure 3A). Embryos injected with both hDSCAM and bace2 mRNA at 5 dpf exhibited no gross morphological defects and no reduction of enteric neurons in the distal intestine as compared with control embryos (P > 0.05; Figures 3B,C).

DSCAM and BACE2 were not found to be up-regulated in tissues of DS patients or mouse model in previous studies (Lockstone et al., 2007; Prandini et al., 2007; Letourneau et al., 2014; Olmos-Serrano et al., 2016). Additionally, the deficiency of other known HSCR genes accounted for the disease risk (Emison et al., 2005; Amiel et al., 2008; Garcia-Barcelo et al., 2009; Jiang et al., 2015; Gui et al., 2017; Porokuokka et al., 2019). Therefore, we used splicing MO to interfere with the expression of dscama, dscamb, and bace2 in zebrafish embryos to explore the effect of loss of function of the two genes in ENS development (Figure 4). The RT-PCR results confirmed the knockdown of target genes by SBMO, and the expression levels of target genes were significantly reduced in the splicing-MO embryos (Figures 4A,B). Both dscama and dscamb MO injection caused obvious morphological defects in a proportion of embryos, usually manifested as the overall shortening of the embryos and multiple tissue disorders, the most common of which was the shrinking brain and/or curling tail (Supplementary Figure 1). When the two MOs were co-injected, the proportion of abnormal embryos increased and the morphological defects became more serious. Importantly, immunostaining with antibodies against a neuronal marker, HuC/D, showed that the density of enteric neurons was significantly reduced in the distal intestine of dscama MO, dscamb MO, and co-injection morphants at 5 dpf as compared with control MO embryos.

Since increasing the amount of morpholino injection led to a higher mortality rate of larvae, for further enteric neurons counting, we chose a dose that could reduce the number of enteric neurons but cause the minimum death rate. The abnormality rate was 13.1% (n = 84) for 0.25 mM of dscama MO injection, while it was 60.0% (n = 60) for 0.25 mM of dscamb MO injection. We decreased the concentration of dscamb MO to 0.125 mM, and the dysmorphic rate became 13.41% (n = 82) but was still effective. Immunostaining results showed that the average number of enteric neurons was 89.53 ± 25.41 in the last six somite lengths of 0.25 mM of dscama MO morphants and 92.83 ± 40.20 in 0.125 mM of dscamb MO morphants. Both were significantly lower than that in embryos injected with control MO (125.4 ± 20.19 for 0.25 mM of control MO, P < 0.0001; 126.90 ± 17.18 for 0.125 mM of control MO, P = 0.0001, Figures 4C–H). When 4.2 ng/embryo dscama MO and 0.125 mM of dscamb MO were co-injected, the reduction of enteric neuron was more significant (85.53 ± 16.02; n = 95) as compared with embryos injected with 0.375 mM of control MO (123.1 ± 24.28; P < 0.0001, Figures 4I–K).

The bace2 morphants injected with 0.25 mM of MO appeared morphologically normal. However, immunostaining analysis showed that the number of enteric neurons was significantly reduced in the distal intestine of bace2 MO morphants 5 dpf as compared with control MO embryos (93.00 ± 17.70 vs. 126.6 ± 26.60; P < 0.0001, Figures 4L–N).

A protein–protein interaction (PPI) network for the DSCAM and BACE2 with critical signaling pathway genes was constructed, and the correlations were evaluated using the GeneMANIA database (Figure 5 and Supplementary Table 11). DSCAM was co-expressed with GFRA1, NRTN, ERBB4, SOX10, and PAX3 and was correlated with EDNRB, GFRA1, and SEMA3C in terms of genetic interactions. BACE2 was co-expressed with NRG1, ERBB3, and PAX3 and had genetic interactions with EDNRB, GFRA1, EDN2, EDN3, ERBB4, PAX3, and SEMA3C. Further functional prediction revealed that these proteins showed correlations with neural crest cell development [false discovery rate (FDR) = 2.60 × 10^–6), neural crest cell differentiation (FDR = 4.07 × 10^–6), and neural crest cell migration (FDR = 6.12 × 10^–5).