Table 1.
Sequences of the oligos used in this study.
Fig 1.
Comparative in silico analysis of soat1 and soat2.
(A-B) Genomic organizations of soat1 (A) and soat2 (B) in zebrafish, human, mouse, rat, Xenopus and chicken showed similar pattern of exonal distributions. In the regions coding for MBOAT domains, the exon/intron structures are almost indentical between soat1 and soat2 and among various species. The sizes of the exons are shown in the box representing exons, the grey area in the exon boxes are untranslated regions, and the black regions represent MBOAT domains. (C) Phylogenetic trees were constructed using neighbor joining method with percentage identity distances with the software Jalview2, and showed that soat1 and soat2 genes are grouped together respectively, with yeast ARE1 and ARE2 as outgroups. The reference genes used in this comparative analysis include: SOAT1 of human (ENST00000367619), mouse (ENSMUST00000051396), rat (ENSRNOT00000005677), chicken (ENSGALT00000006691), Xenoupus (ENSXETT00000050224), SOAT2 of human (ENST00000301466), mouse (ENSMUST00000023806), rat (ENSRNOT00000015368), xenopus (ENSXETT00000003832), and yeast are1 (YCR048W), and are2 (YNR019W). The sequences of zebrafish soat1 and soat2 were obtained from the cloned sequences.
Fig 2.
Sequence similarity of Soat1 among various animal species.
Zebrafish Soat1 amino acid sequences are aligned with the SOAT1 orthologs from human (ENSP00000356591), rat (ENSRNOP00000005677), mouse (ENSMUSP00000058344), chicken (ENSGALP00000006691) and Xenopus (ENSXETP00000050224). The amino acid sequence of zebrafish Soat1 was predicted according to the cloned sequence. The MBOAT domain is predicted between 168 to 517 residue (arrow under sequences). The acyl-CoA binding domain (FYXDWWN, orange box above the sequences) and the postulated catalytical residue for esterification (His457, red asterike above the sequences) are conservative through species. The identical amino acids are shown in the black box, while the less conservative regions are enclosed by white box.
Fig 3.
Sequence similarity of Soat2 among various animal species.
Zebrafish Soat2 amino acid sequences are aligned with the SOAT2 orthologs from human (ENSP00000301466), rat (ENSRNOP00000015368), mouse (ENSMUSP00000023806), and Xenopus (ENSXETP00000003832). The amino acid sequence of zebrafish Soat2 was predicted according to the cloned sequence. The MBOAT domain is predicted between 140 to 505 residue (arrow under sequences). The acyl-CoA binding domain (FYXDWWN, orange box above the sequences) and the postulated catalytical residue for esterification (His445, red asterike above the sequences) are conservative through species. The identical amino acids are shown in the black box, while the less conservative regions are enclosed by white box.
Fig 4.
Distinct expression profiles of soat1 and soat2.
RT-PCR analysis using cDNA collected from zebrafish adult tissue (A) and embryos (B) indicated that soat1 is expressed ubiquitously in every tissue tested, including the eye (ey), brain (br), skin (sk), intestine (in), ovary (ov), liver (li), heart (he), kidney (ki), testis (te) and gill (gi), and could be detected at freshly laid eggs (0 hpf) through 48 hpf. Soat2 is expressed only in the brain, liver, intestine and testis in the adult zebrafish, and could not be detected until 12 hpf in the embryos. β-actin served as loading control while the reaction without template (-) served as negative control.
Fig 5.
Expression patterns of zebrafish soat1 during embryogenesis.
Whole-mount in situ hybridization shows that soat1 can be detected in all blastomeres at 1 cell stage (A), 3 hpf (B), and 6 hpf (C). The expression of soat1 was prominent at yolk sac at 12 hpf (D) and 24 hpf (E). By 48 hpf (F, G), soat1 was detected in the brain, retina, pectoral fin, hatching gland and pericardium. By 72 hpf (H, I), soat1 was also observed prominently in the liver primordia and intestine.
Fig 6.
Expression patterns of zebrafish soat2 during embryogenesis.
Whole-mount in situ hybridization shows that soat2 can be detected at yolk sac at 12 hpf (A) and 24 hpf (B). By 48 hpf (C, D), soat2 was detected in the brain, retina, yolk sac, hatching gland, pericardium and intersomitic regions. By 72 hpf (E, F), soat2 was also observed in the liver primordial and intestine.
Fig 7.
Increased intracellular accumulation of neutral lipid and CEs as a result of esterification of cholesterol to fatty acyl-CoA.
(A) Oil Red O (ORO) staining of eGFP stable expressing (Ctrl), zebrafish saot1 stable expressing (Soat1), and zebrafish soat2 stable expressing (Soat2) HEK293 cells after incubation with low (15 μM oleic acids and 1 μg/mL cholesterol) or high (150 μM oleic acids and 10 μg/mL cholesterol) level of substrate at room temperature for 6 hrs. The expression of zebrafish Soat2 significantly increased the intracellular accumulation of neutral lipid indicated its enzymatic activity of cholesterol esterification, and this accumulation of neutral lipid is significantly increased with the increased levels of substrates. Note the significantly increased neutral lipid in WT-high compared to WT-low reflected the activity of endogenous human SOAT1 expressed by HEK293 cells. The letters above each column indicate the statistical groups, and the data sharing the same letters indicates no significant difference. (B) The cell lysate of HEK293 cells with zebrafish soat2 overexpression contained significantly higher CE content than either control group or soat1 overexpression group. Consistent with the ORO staining, the CE content in soat1 group was comparable to the control group. (C) RT-PCR confirmed the expression of both zebrafish soat1 and eGFP in Soat1-overexpressing HEK293 cells, zebrafish soat2 and eGFP in Soat2-overexpressing HEK293 cells and only eGFP in control group in which the HEK293 cells were transfect with empty vectors. (D) The HEK293 cells overexpressing zebrafish Soat2 (Soat2-DsRed) contained more and larger intracellular lipid droplets containing NBD-cholesterol as compared to the HEK293 cells overexpressing DsRed only (DsRed). (E) Avasimibe (AVA) significantly reduced the intracellular accumulation of neutral lipid, but the cells with zebrafish soat2 overexpression still showed a significantly more ORO staining than other AVA-treated cells. (F) Pyripyropene A (PPPA) could significantly reduce the intracellular accumulation of neutral lipid by inhibiting the enzymatic activity of both endogenous human SOAT1 and zebrafish Soat2 when compared to the vehicle control (DMSO). The letters above each column indicate the statistical groups, and the data sharing the same letters indicates no significant difference.
Fig 8.
Delayed yolk consumption by yolk injection of PPPA.
After the injection of PPPA, AVA or DMSO into the yolk at 3 hpf, embryos with grossly normal appearance at 24 hpf were subjected to yolk area measurement to estimate the yolk consumption. Yolk injection of PPPA resulted in significant larger yolk size at 72 and 84 hpf when compared to the DMSO group suggesting a delayed consumption of yolk content. *: P < 0.05, PPPA vs. DMSO; ***: P < 0.001, PPPA vs. DMSO; #: P < 0.05, PPPA vs. AVA
Fig 9.
The yolk size phenotype in yolk-soat2 knockdown zebrafish embryos.
(A) The soat2 specific morpholino (s2-MO) was designed to target on the boundary of exon 4 and intron 4 (green concaved arrowhead). Specific primers for amplifying targeted sequence flank exon 3 through exon 6 (concaved arrows). (B) The effect of s2-MO was confirmed by reverse-transcription polymerase chain reaction (RT-PCR) and a shorter product (soat2*) was observed with the expected normal product (soat2) when s2-MO was injected to the embryos. The sequence of soat2* was confirmed to lack of the entire exon 4 and predictively to give rise to a truncated protein with only 99 amino acids. (C) The embryos with yolk injection of s2-MO at 3 hpf showed significantly larger yolk size when compared to the embryos received control morpholino (Ctrl-MO).
Fig 10.
A schematic illustration of the molecular mechanism for yolk cholesterol trafficking in zebrafish embryo.
Before the formation of YSL, free cholesterol diffuses from yolk to blastomeres directly. After YSL is formed at about 4 hpf, free cholesterol are transported from yolk to embryo through Npc1l1. The blood circulation in zebrafish embryos begins at about 24 hpf, and the importance of lipoproteins in yolk lipids absorption could be observed at 48 hpf when Mttp or Apo C-II was defective. The mRNA of soat2 is expressed after 12 hpf, but its enzyme does not evidently contribute to the yolk cholesterol trafficking until the assembly of lipoproteins become prominent after 48 hpf. The level of CEs surged at 72 hpf and the loss of Soat2 activity resulted in delayed yolk consumption most prominently at 72 hpf indicated that the activity of Soat2 peaked at 72 hpf.