The nearly simultaneous convergence of human genetics and advanced molecular technologies has led to an improved understanding of human diseases. At the same time, the demand for drug screening and gene function identification has also increased, albeit time- and labor-intensive. However, bridging the gap between in vitro evidence from cell lines and in vivo evidence, the lower vertebrate zebrafish possesses many advantages over higher vertebrates, such as low maintenance, high fecundity, light-induced spawning, transparent embryos, short generation interval, rapid embryonic development, fully sequenced genome, and some phenotypes similar to human diseases. Such merits have popularized the zebrafish as a model system for biomedical and pharmaceutical studies, including drug screening. Here, we reviewed the various ways in which zebrafish serve as an in vivo platform to perform drug and protein screening in the fields of rare human diseases, social behavior and cancer studies. Since zebrafish mutations faithfully phenocopy many human disorders, many compounds identified from zebrafish screening systems have advanced to early clinical trials, such as those for Adenoid cystic carcinoma, Dravet syndrome and Diamond–Blackfan anemia. We also reviewed and described how zebrafish are used to carry out environmental pollutant detection and assessment of nanoparticle biosafety and QT prolongation.

After spinal cord injury (SCI) in mammals, neuronal regeneration is limited; in contrast, such regeneration occurs quickly in zebrafish. Member A of the acidic nuclear phosphoprotein 32 (ANP32a) family is involved in neuronal development, but its function is controversial, and its involvement in zebrafish SCI remains unknown. To determine the role of zebrafish ANP32a in the neuronal regeneration of SCI embryos, we microinjected ANP32a mRNA into embryos from zebrafish transgenic line Tg(mnx1:GFP) prior to SCI. Compared to control SCI embryos, the results showed that the regeneration of spinal cord and resumption of swimming capability were promoted by the overexpression of ANP32a mRNA but reduced by its knockdown. We next combined fluorescence-activated cell sorting with immunochemical staining of anti-GFAP and immunofluorescence staining against anti-PH3 on Tg(gfap:GFP) SCI embryos. The results showed that ANP32a promoted the proliferation and cell number of radial glial cells at the injury epicenter at 24 h post-injury (hpi). Moreover, when we applied BrdU labeling to SCI embryos derived from crossing the Tg(gfap:GFP) and Tg(mnx1:TagRFP) lines, we found that both radial glial cells and motor neurons had proliferated, along with their increased cell numbers in Anp32a-overexpression SCI-embryos. On this basis, we conclude that ANP32a plays a positive role in the regeneration of zebrafish SCI embryos.

Neuron-specific Enolase 2 (Eno2) is an isozyme primarily distributed in the central and peripheral nervous systems and neuroendocrine cells. It promotes neuronal survival, differentiation, and axonal regeneration. Recent studies have shown that Eno2 localized on the cell membrane of motor neurons acts as a receptor for extracellular phosphoglycerate kinase 1 (ePgk1), which is secreted by muscle cells and promotes the neurite outgrowth of motor neurons (NOMN). However, interaction between Eno1, another isozyme of Enolase, and ePgk1 failed to return the same result. To account for the difference, we constructed seven point-mutations of Eno2, corresponding to those of Eno1, and verified their effects on NOMN. Among the seven Eno2 mutants, eno2-siRNA-knockdown NSC34 cells transfected with plasmid encoding the 419th aspartic acid mutated into serine (Eno2-[D419S]) or Eno2-[E420K] showed a significant reduction in neurite length. Moreover, the Eno2-ePgk1-interacted synergic effect on NOMN driven by Eno2-[D419S] was more profoundly reduced than that driven by Eno2-[E420K], suggesting that D419 was the more essential residue involved in NOMN mediated by Eno2-ePgk1 interaction. Eno2-ePgk1-mediated NOMN appeared to increase the level of p-Cofilin, a growth cone collapse marker, in NSC34 cells transfected with Eno2-[D419S] and incubated with ePgk1, thereby inhibiting NOMN. Furthermore, we conducted in vivo experiments using zebrafish transgenic line Tg(mnx1:GFP), in which GFP is tagged in motor neurons. In the presence of ePgk1, the retarded growth of axons in embryos injected with eno2-specific antisense morpholino oligonucleotides (MO) could be rescued by wobble-eno2-mRNA. However, despite the addition of ePgk1, the decreased defective axons and the increased branched neurons were not significantly improved in the eno2-[D419S]-mRNA-injected embryos. Collectively, these results lead us to suggest that the 419th aspartic acid of mouse Eno2 is likely a crucial site affecting motor neuron development mediated by Eno2-ePgk1 interaction, and, hence, mutations result in a significant reduction in the degree of NOMN in vitro and axonal growth in vivo.

ENDOU-1 encodes an endoribonuclease that overcomes the inhibitory upstream open reading frame (uORF)-trap at 5′-untranslated region (UTR) of the CHOP transcript, allowing the downstream coding sequence of CHOP be translated during endoplasmic reticulum (ER) stress. However, transcriptional control of ENDOU-1 remains enigmatic. To address this, we cloned an upstream 2.1 kb (−2055~+77 bp) of human ENDOU-1 (pE2.1p) fused with reporter luciferase (luc) cDNA. The promoter strength driven by pE2.1p was significantly upregulated in both pE2.1p-transfected cells and pE2.1p-injected zebrafish embryos treated with stress inducers. Comparing the luc activities driven by pE2.1p and −1125~+77 (pE1.2p) segments, we revealed that cis-elements located at the −2055~−1125 segment might play a critical role in ENDOU-1 upregulation during ER stress. Since bioinformatics analysis predicted many cis-elements clustered at the −1850~−1250, we further deconstructed this segment to generate pE2.1p-based derivatives lacking −1850~−1750, −1749~−1650, −1649~−1486, −1485~−1350 or −1350~−1250 segments. Quantification of promoter activities driven by these five internal deletion plasmids suggested a repressor binding element within the −1649~−1486 and an activator binding element within the −1350~−1250. Since luc activities driven by the −1649~−1486 were not significantly different between normal and stress conditions, we herein propose that the stress-inducible activator bound at the −1350~−1250 segment makes a major contribution to the increased expression of human ENDOU-1 upon ER stresses.