Anemia in apoc2 mutant zebrafish. a Wright–Giemsa (Hema) and o-dianisine staining of peripheral blood cells from adult (18–20-month-old) male wild-type (WT) and apoc2 mutant zebrafish. b Quantitative results of peripheral blood cell count (n = 11 for WT and n = 9 for apoc2 mutant groups). c Representative bright field images and quantitative results of blood cell (yellow arrows) count in the caudal vein (outlined with white dashed lines) of WT and apoc2 mutants at 52 hpf (n = 8 in WT and n = 9 in apoc2 mutant groups) and 6.3 dpf (n = 11 for WT and n = 13 for apoc2 mutant groups). See also Supplementary Movies 1 and 2. d Wright–Giemsa staining of blood smears from WT and apoc2 mutants at 52 hpf and 6.3 dpf. Immature erythrocytes containing larger and less condensed nuclei are indicated with red arrows in a and d. Immature erythrocytes with weaker hemoglobin staining are indicated with green arrows in a. e o-Dianisine staining of 52 hpf and 6.3 dpf WT and apoc2 mutant embryos. f Effect of lomitapide: WT and apoc2 mutants were treated with 5 μM lomitapide starting from 2 dpf until embryos were fixed at 6 dpf for ORO staining. Black arrows point to intestinal lipid accumulation and green arrows to circulating lipids. g Blood cell counts in the caudal vein of WT, apoc2 mutants and apoc2 mutants treated with lomitapide at 6.3 dpf (n = 5 in each group). h Blood cell counts in the caudal vein of 14 dpf WT, apoc2 mutants fed with normal diet and apoc2 mutants fed with low-fat diet (LFD) starting at 5 dpf (n = 4 in WT and apoc2 mut groups each; n = 5 in apoc2 + LFD group). Scale bars, 20 μm in a, c, and d, 100 μm in e, and 200 μm in f. Mean ± SEM; ***P < 0.001 (Student’s t test)

Hematopoietic defects in lpl mutants. a In situ hybridization with lpl antisense and sence probes in WT embryos at 2 dpf. b Diagram of lpl CRISPR target sites and the predicted truncated protein caused by the 2 nt deletion, which results in a codon shift and premature translation termination. c qPCR results of lpl and apoc2 mRNA expression in WT, apoc2 and lpl mutants at 5 dpf (n = 3 in each group). d Plasma TG levels in adult (9–15-month-old) male WT, apoc2 and lpl mutants (n = 5 in each group). e Peripheral blood cell count in adult (15-month-old) male WT, apoc2 and lpl mutants (n = 5 in each group). f Wright–Giemsa staining of blood smears from 6.3 dpf WT, apoc2 and lpl mutants. g, h Representative images and quantitative results of blood cell (yellow arrows) count in the caudal vein (outlined with white dashed lines) of WT, apoc2 and lpl mutants at 6.3 dpf (n = 5 in WT and apoc2 mut groups each; n = 6 in lpl mut group). i o-Dianisine staining of 52 hpf and 6.3 dpf WT, apoc2 and lpl mutant embryos. Scale bars, 200 μm in a; 20 μm in f, g, and 100 μm in i. Quantitative results are mean ± SEM; *P < 0.05 and ***P < 0.001 (Student’s t test)

Hematopoietic defects in apoc2 and lpl mutants occur during HSPC expansion. a In situ hybridization with cmyb/runx1 and β-globin probes in WT, apoc2 and lpl mutants at 26 hpf, 30 hpf and 50 hpf. b In situ hybridization with cmyb/runx1, β-globin, rag1 (green arrows) and foxn1 (black arrows) probes in WT, apoc2 and lpl mutants at 80 hpf. foxn1 is a thymus development marker, used as a control. c, d Representative images and numbers of GFP^low cells (HSPCs, white arrows) in the CHT region of cd41:EGFP transgenic WT and apoc2 mutants at 54 and 80 hpf (n = 10 in WT and n = 8 in apoc2 mutant groups at 54 hpf; n = 6 in each group at 80 hpf). e, f Representative images and numbers of GFP-positive cells in the thymus region at 54 hpf (n = 9 in WT and n = 8 in apoc2 mutant groups) and at 4 dpf (n = 8 in each group). Scale bars, 200 μm in a and b; 50 μm in c–f. Mean ± SEM; *P < 0.05 and ***P < 0.001 (Student’s t test)

APOC2 mimetic peptide and VLDL rescue anemia in apoc2 mutants. a, b Representative images of BODIPY staining and quantitative results of BODIPY fluorescence intensity in WT, apoc2 mutants and the apoc2 mutants injected with APOC2 mimetic peptides (CII-a, active; CII-i, inactive) at 6.3 dpf (n = 5 in each group). c, d Representative images of blood cells in the caudal vein region and quantitative results of blood cell count in WT, apoc2 mutants and the apoc2 mutants injected with APOC2 mimetic peptides at 6.3 dpf (n = 6 in WT and mut + CII-i goups each; n = 8 in apoc2 mut group; n = 7 in mut + CII-a group). e In situ hybridization with cmyb/runx1 and β-globin probes in WT, apoc2 mutants and the apoc2 mutants injected with APOC2 mimetic peptides at 80 hpf. Embryos were injected with peptides at 2 dpf. f In situ hybridization with cmyb/runx1 and β-globin probes in WT, apoc2 and lpl mutants, including those injected with VLDL or LDL at 2 dpf. Embryos were analyzed at 80 hpf. Scale bars, 50 μm in a and c; 200 μm in e and f. Mean ± SEM; ***P < 0.001 (Student’s t test)

Parabiosis of apoc2 and lpl mutants rescues defective hematopoiesis. a The CHT region of gata2:EGFP, sdf1a:mCherry double-positive embryos at 2.5 dpf. b Flow cytometry of gata2:EGFP and sdf1a:mCherry positive cells isolated from the CHT region. c RT-qPCR analysis of FACS-sorted gata2:EGFP and sdf1a:mCherry positive cells, using lpl, runx1 and sdf1α primers. Mean ± SD of two independent experiments. d, e Parabiosis of a sdf1α:mCherry or a ahmc:EGFP with a WT embryo. Right-hand panels are enlarged segments showed in white quadrangles in left-hand panels. Yellow dashed lines trace WT embryos’ boundaries. f, g Rescue of hyperlipidemia in apoc2 or lpl mutants by parabiosis with WT embryos. Upper panels: two separated embryos (WT and apoc2 or lpl). Lower panels: larva 1 and larva 2 from a parabiosis pair (WT with apoc2 or WT with lpl). Black arrows point to ORO staining in the lumen of blood vessels. h In situ hybridization with runx1/cmyb probe in WT, individual apoc2 and lpl mutants, and parabiotic apoc2 and lpl mutants at 80 hpf. Scale bars, 50 μm in f, g, and 200 μm in a, d, e and h. i Diagram of lipoprotein metabolism in parabiotic lpl and apoc2 mutants. In the lpl mutant, no Lpl is expressed. However, VLDL secreted by the lpl mutant (orange) delivers Apoc2 through the shared circulation to the apoc2 mutant, in which Lpl is expressed but its own VLDL (yellow) contains no Apoc2. VLDL from the lpl mutant compensates lack of Apoc2 in the apoc2 mutant and the reconstituted Apoc2/Lpl catalyzes hydrolysis of TG to release FFAs into the shared circulation, which in turn rescue the hematopoiesis defect in both mutants

ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.

DHA rescues hematopoiesis in apoc2 mutants. a–c apoc2 mutant embryos were injected with free fatty acid docosahexaenoic acid (DHA), oleic acid (OA), or eicosapentaenoic acid (EPA) at 48 hpf. a o-Dianisine staining of 6.3 dpf larvae. b Representative bright field images and quantitative results of blood cell (yellow arrows) count in the caudal vein (outlined with white dashed lines) at 6.3 dpf. Mean ± SEM; n = 10 (WT), n = 8 (apoc2 mut, apoc2 mut + OA, and apoc2 mut + EPA), and n = 9 (apoc2 mut + DHA). ***P < 0.001 (Student’s t test). c In situ hybridization with cmyb/runx1 and β-globin probes. Scale bars, 100 μm in a; 50 μm in b; and 200 μm in c

FFA-DHA but not TG-DHA rescues hematopoiesis in apoc2 and lpl mutants. a, b In situ hybridization with runx1/cmyb and β-globin probes in WT, apoc2 and lpl mutants injected with FFA-DHA (a) or TG-DHA (b) at 2 dpf; embryos were fixed at 80 hpf for in situ hybridization. Scale bars, 200 μm. c Schematic representation of the working hypothesis. VLDL delivers both the TG substrate and Apoc2, an obligatory activator of Lpl to the hematopoietic niche. Lpl, expressed on stromal and/or endothelial cells, catalyzes hydrolysis of TG to produce FFAs. Among FFAs released by the Apoc2/Lpl catalysis, the essential fatty acid DHA supports normal hematopoiesis. apoc2 and lpl mutant zebrafish in which TG hydrolysis is blocked, have a defect in HSPC maintenance and differentiation. Administration of DHA as a free fatty acid, but not DHA esterified into a TG, rescues the hematopoiesis defect in apoc2 and lpl mutant zebrafish

In situ hybridization with apoc2 antisense and sense probes. In 2 dpf zebrafish, apoc2 is mainly expressed in the intestine and the yolk.

lpl mutants develop hyperlipidemia. (A) Primers, flanking exon 4, the site of CRISPR target, are used to amplify lpl cDNA. (B) Single lpl cDNA band of the same size in WT and lpl mutant embryos indicates that no alternative splicing occurs in the mutant. (C) Sequence results indicate a 2nt deletion in cDNA from lpl mutants. (D) ORO staining of WT, apoc2 and lpl mutants at 6 dpf. (E) Representative images and quantitative results of BODIPY neutral lipid staining of WT, apoc2 and lpl mutants at 6 dpf. Results are mean±SEM; n=5 (WT and apoc2 mut) and n=6 (lpl mut); ***P<0.001 (Student’s t-test). Scale bars, 100 μm in D, and 50 μm in E.

Expression of hematopoietic markers in apoc2 mutants at different developmental stages. (A) In situ hybridization with gata1, β-globin and pu.1 probes in WT and apoc2 mutants at 20 hpf. (B) In situ hybridization with runx1 at 26 hpf (top) and cmyb at 52 hpf in WT and apoc2 mutants’ VDA (middle images) and CHT (bottom) regions. (C) In situ hybridization with cmyb/runx1, β-globin and rag1 probes in WT and apoc2 mutants at 80 hpf. rag1 signals in the thymus are indicated with green arrows. Scale bars, 200 μm.

Hematopoietic defect in apoc2 mutants is not due to delayed angiogenesis. (A) Angiogenesis in fli1:eGFP WT, apoc2 mutant, the apoc2 mutant synchronized with WT, and the apoc2 mutant injected with DHA, all imaged at 80 hpf. apoc2 mutant synchronization was achieved by maintaining the embryos at 30°C from 72 to 80 hpf, while WT embryos were kept at 28°C. DHA was injected at 48 hpf. Yellow arrows point to sub-intestinal vessels. Red arrows point to the yolk. (B) In situ hybridization with runx1/cmyb primers in the same groups as in panel A, at 80 hpf. The CHT regions traced with red dashed quadrangles are enlarged in right-hand panels. Scale bars, 200 μm.

Lomitapide does not rescue the hematopoietic defect in apoc2 mutants. In situ hybridization with cmyb/runx1 and β-globin probes in WT, apoc2 mutants and the apoc2 mutants treated with 5 μM lomitapide. Lomitapide treatment was started at 2 dpf and continued until the embryos were fixed at 80 hpf for in situ hybridization. Scale bar, 200 μm.

apoc2 mutants show no increase in apoptosis in the CHT region. TUNEL staining in the CHT region of 3.3 dpf WT and apoc2 mutant embryos. Embryos incubated with Dnase I for 15 min at room temperature were used as a positive control. Embryos incubated with a labeling solution only, without terminal transferase, were used as a negative control. Scale bar, 50 μm.

DHA but not oleic or eicosapentaenoic acid rescues hypochromia in apoc2 mutants. (A) apoc2 mutant embryos were injected with free fatty acid DHA, OA or EPA at 48 hpf. Wright-Giemsa staining of peripheral blood cells in 6.3 dpf zebrafish. Red arrows point to immature erythrocytes containing larger and less condensed nuclei. Scale bar, 10 μm.

TG-DHA but not the vehicle increases hematopoietic markers in WT zebrafish. In situ hybridization with cmyb/runx1 and β-globin probes in WT and WT injected with POPC or TG-DHA:POPC micelles. Embryos were injected at 2 dpf and fixed at 80 hpf for in situ hybridization. Left- and right-hand side images are the same as in Fug. 8B. Images in the middle were obtained from the same experiment as those shown in Fig. 8B. Scale bar, 200 μm.