Folate-mediated one-carbon metabolism. (A) Structure of folate derivatives. (B) FOCM consists of multiple reactions and contributes to amino acid metabolism, nucleotide synthesis and SAM formation. The enzymes involved in FOCM include: FPGS, folylpolyglutamate synthase; GGH: γ-glutamylhydrolase; DHFR, dihydrofolate reductase, SHMT1, cytosolic serine hydroxymethyltransferase; SHMT2, mitochondrial SHMT; MTHFD1, cytosolic tri-functional enzyme, C1-THF-synthetase; MTHFD2/2L, mitochondrial bi-functional enzyme (5,10-methylene-THF dehydrogenase/5,10-methenyl-THF cyclohydrolase); MTHFD1L, mitochondrial 10-formy-THF synthetase; FDH, 10-formyltetrahydrofolate dehydrogenase; MTHFS, 5-formyltetrahydrofolate cyclo-ligase; MTR, 5-methyltetrahydrofolate-homocysteine methyltransferase; MAT, methionine adenosyltransferase; AHCY, adenosylhomocysteinase; MTHFR, 5,10-methylene-THF dehydrogenase; TS, thymidylate synthetase; GAR/AICAR TF, glycinamide ribonucleotide/aminoimidazole carboxamide ribonucleotide transformylase. T, folate transporters; MFT, mitochondrial folate transporter; the green area indicates the folate cycle, and yellow area indicates the methionine cycle.

Morphological and functional characterization on the eyes of FD larvae. (A,B) Control and FD larvae at indicated stages with/without folate supplementation were imaged laterally for the head region and quantified for the eye area (dotted circle) with Image J (An on-line software for scientific image analysis.). The larvae in folate supplemented groups were exposed to the indicated folate adducts at 25 hpf and until observation. (C) The diameter of lens was estimated from the images of H&E stained cryo-sections prepared from the larvae at 5 dpf. (D) The experimental flow of optomotor response analysis (OMR). (E) Control and FD larvae with/without folate supplementation were analyzed for OMR at 7 dpf individually and quantified. WT, wild-type; C (control), heat-shocked wild-type larvae; M, FD transgenic larvae with mild FD. S, FD transgenic larvae with severe FD. L, length of lens; ON, optic nerve; RPE, retinal pigmented epithelium; IS/OS, inner segment/outer segment junction; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer; FA, folic acid (1mM); 5-CHO THF, 5-formyltetrahydrofolate (0.5 mM); 5-CH₃ THF, 5-methyltetrahydrofolate (1 mM). Statistical data are shown in mean ± SEM. ***p-value < 0.001; **p-value < 0.01; *p-value < 0.05.

Developmental timing-specific analysis for the cell proliferation and cell cycle in developing larvae. The larvae of control and FD groups at 26, 48, and 72 hpf were subjected to immunostaining with anti-PH3 antibodies (M-phase marker) (A), as well as TUNEL assay for apoptosis (B). The number of PH3+ cells and apoptotic cells within eye area were quantified by counting the cells with fluorescent signal using the on-line software Image J (C). Larvae at 26 hpf (D) and 72 hpf (E) were dispersed, stained with propidium iodide (PI) and analyzed for cell cycle with flow-cytometry and quantified. White arrows indicate the signals of PH3-positive cells; yellow arrows represent the signals of apoptosis; red dotted lines circle the eye area. Statistical data are shown in mean ± SEM. **p-value < 0.01; *p-value < 0.05.

Larval folate composition and response to rescuing agents. Both control and FD embryos continuously exposed to rescuing agents starting from 25 hpf were examined for their intracellular folate composition (A) at indicated stages with HPLC as described in materials and methods, eye size (B–F) at 3 dpf and OMR (G) at 7 dpf. The rescuing agents include deoxy-ribonucleoside triphosphate (dNTP) (B), Deoxythymidine triphosphate (dTTP) (C), hypoxanthine (Hx) (D), formate (NaF, sodium formate) (E) and N-acetyl-cysteine (NAC) (F). Statistical results are shown in mean ± SEM for (B–F) and percentage in a group for (G). ***p < 0.001; **p < 0.01; *p < 0.05.

The mRNA levels of enzymes involved in FOCM in larvae. (A) FD larvae harvested at 5-dpf were subjected to mRNA extraction and RT-PCR analysis for the expression of folate enzymes. (B) Larvae displaying FD induced by heat-shock at 1, 3, and 5 dpf were harvested at 7-dpf and examined for the expression of folate enzymes with RT-PCR. (C) FD larvae collected at the indicated stages were examined for the developmental timing-specific expression of mthfd1L during embryogenesis with real-time PCR. Ctl, control; FD, folate deficiency; aldh1l1, zebrafish 10-formyltetrahydrofolate dehydrogenase; shmt1, cytosolic serine hydroxymethyltransferase; shmt2, mitochondrial serine hydroxymethyltransferase; dhfr, dihydrofolate reductase; mthfr, methylenetetrahydrofolate reductase. ***p-value < 0.001; **p-value < 0.01; *p-value < 0.05.

In vitro and in vivo characteristics of zebrafish mthfd1L. (A) The complete recombinant zebrafish Mthfd1L, encompassing the N-terminal signal peptide, was co-localized with mitochondria. Huh-7 cells transfected with zmthfd1L/pcDNA 3.1 myc-His (zMthfd1L) and pcDNA 3.1 myc-His (Mock) (left panel) and primary cultured zebrafish embryonic cells from embryos injected with mthfd1L cRNA (right panel) were stained with DAPI for nucleus, Mitotracker Deep Red 633 for mitochondria, and anti-His for zebrafish Mthfd1L. (B) The cytosolic and mitochondrial fractions of 293 cells transfected with zmthfd1L/pcDNA 3.1 myc-His and Mock analyzed with Western blotting confirmed the mitochondrial localization of zebrafish Mthfd1L. (C) The neural tube defect (NTD)-like phenotype was found in mthfd1L morphant (MO), but not in wild-type larvae (Ctl) at 3 dpf. (D,E) The eye size of mthfd1L morphants with (MO+cRNA)/without (MO) co-injecting synthesized mthfd1L capped-mRNA (cRNA) for rescue were examined and quantified at 3 dpf. Embryos were injected with translational MO (tMO) or splicing site MO (sMO) at 1–4-cell stage. (F) 1 dpf zebrafish embryos injected with mthfd1L cRNA were collected for embryonic lysates and subjected to Western blotting analysis. (G–I) Mthfd1L morphants at 3 dpf were analyzed with HPLC for folate composition. c, cytosolic fraction; m, mitochondrial fraction; α-tub, cytosolic marker; and Cox4, mitochondrial marker. MO, wild-type embryos injected with 10 ng mthfd1L MO; MO+cRNA, wild-type embryos co-injected with 10 ng mthfd1L MO and 1 ng mthfd1L synthesized RNA simultaneously. ***p < 0.001; **p < 0.01; *p < 0.05.

Effects of altering mthfd1L expression on eye development in folate-deficient embryos. (A) FD mthfd1L morphants had their eye size measured at 3 dpf. (B) FD embryos were injected with 0–250 pg zebrafish mthfd1L cRNA and their eye size was measured at 3 dpf. (C)mthfd1L-cRNA-injected control and FD embryos had their OMR estimated at 7 dpf. (D–F) 30 hpf control and FD embryos with or without 250 pg mthfd1L cRNA injection were subjected to HPLC analysis for their content of 10-CHO-THF (D), 5-CH₃-THF (E), and THF/5,10-CH₂-THF (F). Ctl, control; mFD, mild folate deficiency; sFD, severe folate deficiency; *p < 0.05; **p < 0.01; ***p < 0.001.

The interplay between Mthfr and Mthfd1L. (A) FD larvae collected at the indicated stages were examined for the developmental timing-specific expression of mthfr during embryogenesis. (B) Wild-type embryos injected with mthfd1L cRNA or MO were examined for the expression of mthfr at 1 dpf. (C) FD embryos with/without formate supplementation were examined for mthfr mRNA levels at 2 dpf. (D) FD embryos with/without formate supplementation were examined for mthfd1L mRNA levels at 5 dpf. (E) Control and FD embryos supplemented with or without 5 mM formate were measured for their total folate content at 2 dpf. ***p-value < 0.001; **p-value < 0.01; *p-value < 0.05.

Analysis of ocular cell differentiation in FD larvae and response to folate derivatives. (A,C) Control and FD larvae at 48 and 72 hpf were characterized with WISH for the spatial distribution of foxe3 (ocular lens marker) and quantified for the percentage of larvae displaying abnormal expression patterns. (B,D) The aberrant distribution of rhodopsin observed in the eyes of FD larvae revealed by WISH at 3 dpf were categorized into four scoring categories, based on the severity of anomalies displayed, with “3” to be normal and “1” to be complete lack of signal. The severity of abnormality was quantified by the averaged score of an anomaly for each larva observed for each experimental group, including those with/without folate supplementation. Data were obtained from at least three independent trials with the total sample numbers ranging from 7 to 80 for each group. C, control; M, mild folate deficiency; S, severe folate deficiency; FA, folic acid; 5-F, 5-CHO-THF; 5-M, 5-CH₃-THF. Statistical data are shown in percentage of a group for (C) and mean ± SEM for (D). ***p < 0.001; **p < 0.01.

Ocular development in FD larvae and response to retinoic acid. Eye size (A) and OMR (B) of FD larvae with/without RA treatment were estimated at 5 and 7 dpf, respectively. The distributions of rhodopsin transcripts in larvae with/without RA supplementation were characterized with WISH at 3 dpf (C) and quantified (E). The distribution of aldh1a2 and aldh1a3 transcripts in larvae with/without RA supplementation were characterized with WISH at 30 hpf (D) and quantified (F). C, control; M, mFD; S, sFD. Statistical data are shown in mean ± SEM for (A,E) and percentage in a group for (B,F). ***p < 0.001; *p < 0.05.

Proposed mechanism of FD-induced ocular defects. Solid lines indicate the experiments that were demonstrated in the current studies; dotted lines indicate the rationales supported by the listed references ([1] Henry et al., 2017; [2] James et al., 1994; [3] Leung et al., 2017; [4] Irwin et al., 2016; [5] Arasaradnam et al., 2008; [6] Bryant et al., 2018; [7] Duan et al., 2016).