Abstract

I
Zebrafish, our genomically tractable little fish, could not be better suited to the task of precision medicine-based drug discovery. To begin a precision medicine-based screen, an investigator needs to start with a disease of interest, followed by acquiring a zebrafish model of the disease. Some models may be already available through stock resources (https://zebrafish.org/home/guide.php; www.sanger.ac.uk/science/collaboration/zebrafish-mutation-project) or by collaborative sharing with colleagues. However, there are limited number of models available for human diseases, and thus, a novel model may be engineered using any of the currently available genomic tools such as CRISPR-CAS9, TALENs, or gene traps.2–4 In some situations, morpholinos may be the best choice, especially when the zebrafish phenotype is faithful to the human condition. Once the model is determined to phenotypically mirror the human condition and a suitable assay for phenotypic change developed, the search for small molecules that can “treat” the phenotype can begin, a precision medicine dream.
A PubMed search using the terms “zebrafish” and “drug discovery” conducted on September 6, 2016, revealed 239 primary research articles and 98 reviews, demonstrating the acceptance and power of zebrafish in drug discovery. Here are three examples from the past 12 months, in which zebrafish models of human diseases are used to identify potential novel therapeutic agents.
Novel Treatments for Epilepsy
Epilepsy, defined as two or more unprovoked seizures in an individual, is a common condition estimated to occur in 1.8% of adults and 1% of children. It is estimated that there are 5.1 million Americans living with epilepsy or a seizure disorder (www.cdc.gov/epilepsy/basics/fast-facts.htm; accessed September 9, 2016). The Institute of Medicine recently issued a report, Epilepsy Across the Spectrum: Promoting Health and Understanding (www.nationalacademies.org/hmd/Reports/2012/Epilepsy-Across-the-Spectrum.aspx; accessed September 9, 2016) that can be downloaded and is targeted to a broad audience to enhance understanding of the types of epilepsy as well as the multiple causes and treatments for epilepsy.
For an individual patient, pharmacological treatment choices to prevent seizures have been largely focused on the type of seizure, age of onset, response to previous drugs, and limitation of side effects. Treatment failures occur in up to a third of all patients suffering from epilepsy; thus, there is an unmet need to develop new antiepileptics.5,6 Drug screens in zebrafish models of human epilepsy provide a personalized medicine approach to development of novel anticonvulsants to treat patients who have failed available treatment options.
Dravet syndrome is a devastating form of epilepsy refractory to currently available anticonvulsant treatments. Caused by pathogenic variants in the gene encoding the neuron voltage-gated sodium channel, SCN1A, children with Dravet syndrome suffer frequent seizures and manifest cognitive deficits. Children with Dravet syndrome are at a risk of sudden unexpected death in epilepsy (SUDEP). 7
To identify new compounds that suppress seizure activity in Dravet syndrome, Dinday and Baraban 8 evaluated zebrafish with homozygous mutations in scn1, the zebrafish ortholog of SCN1A. Homozygous mutant zebrafish exhibit unprovoked seizure activity that can be measured by increased movement velocity compared with wild-type zebrafish. To identify drugs that normalize swimming velocity and reduce seizure activity, individual zebrafish embryos at 5 to 6 dpf were placed in flat-bottomed 96-well plates, each well incubated with one element of a library of 1012 repurposed drugs and screened for changes in swimming velocity. A 44% reduction in movement velocity was considered as a “positive” effect. Through this method, 20 drugs were identified that inhibited spontaneous seizures in the scn1Lab homozygous mutants of which 6 were chosen for a second assay. The second assay measured brain electrical activity, essentially, a zebrafish equivalent of an electroencephalogram (EEG); two drugs were effective in suppressing electrical seizure activity. In addition to identifying effective drugs for further study, the investigators documented drugs in the library that were toxic, causing death or near death in zebrafish embryos. These “negative” data are a secondary important contribution as they will be helpful for other scientists who use the same drug library for other conditions.
Prevention of Aminoglycoside Toxicity
Hearing loss is the most common human sensory condition. Hearing loss can be identified in 1.5 per 1000 newborns 9 and up to 15% of adults older than the age of 18. Although the majority of individuals with hearing loss may have an underlying genetic cause of hearing loss, 10 environmental exposures, particularly to aminoglycoside antibiotics, have a significant impact on hearing.
Aminoglycoside antibiotics are used quite frequently in hospital settings. Despite their known auditory toxicity based on dosage and serum level, these drugs are commonly used to combat potentially life-threatening infections, and thus, the risk of hearing loss may not outweigh the benefit of fighting infection. Aminoglycoside toxicity can occur in any individual treated with this group of antibiotics, but there are patients who are genetically susceptible to aminoglycoside toxicity by virtue of specific mitochondrial sequence variants in the gene encoding mitochondrial 12S ribosomal RNA (MT-RNR1). 11
Aminoglycoside toxicity results in hair cell loss through various cell death pathways. 12 Zebrafish are ideal to study hair cells as they have a ready supply of beautiful hair cells situated externally in the lateral line. Accessibility, orthologous function, and conserved gene expression profiles make the lateral line a robust model for the study of inner ear hair cells of the human ear. 13 Aminoglycosides are toxic to the hair cells of the lateral line in an orthologous manner to mammalian inner ear hair cells. 12
In an effort to identify drugs that could block lateral line hair cell death from aminoglycoside exposure, Kruger et al. 14 tested 502 compounds in 5 dpf embryos individually plated and then assayed for hair cell death loss by imaging patterns and counting the number of surviving fluorescently labeled hair cells. Four compounds of the bisbenzylisoquinoline class were chosen for further study based on reproducibility of results. These compounds were found to protect hair cells from incubation with neomycin and gentamicin, both commonly used aminoglycoside antibiotics. The investigators then sought to determine the mechanism of protection.
First, they performed a cell proliferation assay and determined that the drugs actually protected hair cells rather than enhanced hair cell regeneration from the pool of nearby cells with the potential for redifferentiation into hair cells. Then, they focused on aminoglycoside uptake by hair cells. It is known that aminoglycoside antibiotics enter hair cells through the mechanoelectrical transduction (MET) channels on the apical surface of the hair cell stereocilia bundle. It is also known that compounds containing a quinolone ring block MET channel uptake; thus, the investigators hypothesized that the protective drugs would block entry of aminoglycosides through the MET channel.
To test this hypothesis, the investigators first incubated the embryos in gentamicin conjugated to Texas Red. Of the three drugs tested, all inhibited uptake of the labeled gentamicin into hair cells compared with controls. Furthermore, all three drugs inhibited uptake of FM1-43, a vital dye that is MET channel selective. The investigators also determined that each drug did not interfere with the normal function of the MET channel by evaluating microphonic potential, a measure of signal transduction by the hair cells. Finally, the investigators determined that the aminoglycoside protective drugs did not interfere with the intended use of the aminoglycoside antibiotics; to combat bacterial pathogens!
The investigators also evaluated the effect of one of their compounds on cisplatinum toxicity. As platinum is a commonly used chemotherapeutic agent used to treat various cancers, which is proposed to damage hair cells after entry into the MET channel. They tested one of the drugs in the group and found that at high dosages, hair cells were protected from cisplatinum as well. These studies have important implications for protecting hearing from drugs that can enter through the MET channels of hair cells.
Sideroblastic Anemia Treatment with Glycine and Folate
Sideroblastic anemia describes a group of conditions by which red blood cells are formed, but lack adequate hemoglobin to effectively transport oxygen. Infants and children with congenital sideroblastic anemia are transfusion dependent and require chelation therapy to prevent iron overload, a major side effect of frequent transfusions. Sideroblastic anemias can be caused by pathogenic variants in a number of genes; however, pathogenic variants in SLC25A38 are associated with a specific autosomal recessive form of the disease. 15 Infants with this condition suffer significant morbidity, and hence, the unmet need for this condition is to identify potential therapies based on the known function of SLC25A38.
At the start of their research, Fernandez-Murray et al. 16 had some evidence to show that SLC25A38 was an amino acid transporter and its possible cargo was serine or glycine. The investigators identified the yeast ortholog of SLC25A38, Hem25, and found that if Hem25 was deleted, heme synthesis was impaired. Treatment of yeast with glycine and a downstream intermediate of heme synthesis, 5-aminolevulinic acid, rescued heme synthesis in Hem25 null yeast, supporting the role of glycine transport by SLC25A38 as a critical first step in heme synthesis.
To test the hypothesis that glycine could rescue sideroblastic anemia in a zebrafish model, the investigators used a very well-controlled morpholino experiment by which both zebrafish orthologs of SLC25A38, slc25a38a, and slc25a38b were knocked down, and zebrafish embryos showed reduced hemoglobin synthesis compared with wild type. Glycine treatment of slc25a38a and slc25a38b knockdown zebrafish was unsuccessful in rescuing the phenotype. The investigators then noted that glycine is also critical for folate synthesis. Folate is an essential nutrient in zebrafish and mammals, meaning that folate must be obtained through dietary sources. However, yeast can synthesize folate from glycine. Treating the zebrafish with both glycine and folate rescued heme synthesis. A precision medicine human patient trial of folate and glycine to treat sideroblastic anemia caused by autosomal recessive pathogenic variants in SLC25A38 is being planned.
As our understanding of the genomic basis of human diseases increases, it is our anticipation that precision medicine-based treatments will become standard of care. Zebrafish are poised to be the organism of choice to test precision medicine hypotheses in an effort to accelerate the discovery of novel treatments for human diseases.
Footnotes
Disclosure Statement
No competing financial interests exist.
