FDA-Approved Drug Screening for Compounds That Facilitate Hematopoietic Stem and Progenitor Cells (HSPCs) Expansion in Zebrafish

Hematopoietic stem cells (HSCs) are a specialized subset of cells with self-renewal and multilineage differentiation potency, which are essential for their function in bone marrow or umbilical cord blood transplantation to treat blood disorders. Expanding the hematopoietic stem and progenitor cells (HSPCs) ex vivo is essential to understand the HSPCs-based therapies potency. Here, we established a screening system in zebrafish by adopting an FDA-approved drug library to identify candidates that could facilitate HSPC expansion. To date, we have screened 171 drugs of 7 categories, including antibacterial, antineoplastic, glucocorticoid, NSAIDS, vitamins, antidepressant, and antipsychotic drugs. We found 21 drugs that contributed to HSPCs expansion, 32 drugs’ administration caused HSPCs diminishment and 118 drugs’ treatment elicited no effect on HSPCs amplification. Among these drugs, we further investigated the vitamin drugs ergocalciferol and panthenol, taking advantage of their acceptability, limited side-effects, and easy delivery. These two drugs, in particular, efficiently expanded the HSPCs pool in a dose-dependent manner. Their application even mitigated the compromised hematopoiesis in an ikzf1−/− mutant. Taken together, our study implied that the larval zebrafish is a suitable model for drug repurposing of effective molecules (especially those already approved for clinical use) that can facilitate HSPCs expansion.


Introduction
Hematopoietic stem and progenitor cells (HSPCs) transplantation has been a major stem cell-based curative therapy in the treatment of hematologic diseases, including leukemia, immune deficiencies, hemoglobinopathies, and metabolism-based disorders, since the late 1950s, due to their capacity of reconstructing blood system [1]. Accessibility of bone marrow transplantation for patients is restricted by the short availability of immune-matched donors [2]. Umbilical cord blood (UCB) has become an increasingly popular source of transplantable HSPCs because of its rapid availability with less-stringent immune-matching requirements [3]. Therefore, the ability to expand sufficient HSPCs prior to transplantation has great clinical significance.
harbor compromised proliferation of HSPCs, they rescued the HSPCs expansion-defective phenotypes. These results indicated that ergocalciferol and panthenol had the potential for clinical application in HSPCs expansion and enrichment.

Zebrafish Maintenance
The wild type zebrafish (Danio rerio) line was purchased from China Zebrafish Resource Center (CZRC, China). The Tg(CD41:GFP) transgenic line [27] was used to label HSPCs and thrombocytes. They were raised and maintained according to a standard procedure. During the experimental period, pH volume ranged from 7.8 to 7.9. Dissolved oxygen ranged from 6.95 to 7.23 mg/L. Water temperature ranged from 26.5 to 28 • C. Concentrations of ammonia-N and nitrite nitrogen were maintained lower than 0.2 and 0.005 mg/L, respectively. Salinity of water was 0.2 ppt. Zebrafish were maintained in a 12:12 h light-dark cycle. Embryos were collected from natural spawning and raised at 28.5 • C in egg water with 0.003% 1-phenyl-2-thiourea (PTU) at 12 hpf. The maintenance procedures and experiments of zebrafish were complied with guidelines approved by the Ethics Committee of the College of Life Science, Southwest University (Chongqing, China) with Approval ID: 2,018,092,308. This guidance ensured a clean and disease-free comfortable living environment for the animals.

Drug Treatment
The FDA-approved drug library was purchased from MicroSource Discovery System (CT 06755-1500). Cholecalciferol (S4063), Calcitriol (S1466), and Calcifediol (S1469) were purchased from Selleck Chemicals (Houston, TX, USA). D-Pantothenicacid (B2002) was purchased from Apexbio (Boston, TX, USA). All these drugs were prepared as stock solutions by dissolving it in dimethyl sulfoxide (DMSO). For the treatments, the stock solution was diluted in egg water until reaching the working concentrations (5-20 µM). N-Phenylthiourea (PTU) was ordered from Sigma-Aldrich (St. Louis, MO, USA) and dissolved in water as a stock solution. Then, the PTU stock solution was diluted with egg water until reaching the desired working concentration (0.2 mM).

Whole-Mount In Situ Hybridization (WISH) and Quantification
A standard protocol [28] was followed for preparing antisense RNA probes. The following antisense probes labeled with digoxigenin were used: cmyb, lyz, and gata1. The embryos were fixed in 4% paraformaldehyde (PFA) at RT for 4 h. The signals were observed under a SteREO Discovery.V20 microscope (Carl Zeiss, Oberkochen, Baden-Wurttemberg, Germany). WISH signals were measured as previously described [29]. For cmyb + , lyz + , and gata1 + signals' quantification, we selected the CHT region to estimate the signals areas (pixels) by using ImageJ (Rawak Software Inc., Stuttgart, Baden-Wurttemberg, Germany), as described previously [30].

TUNEL Assay and EdU Incorporation
For TUNEL (Terminal deoxynucleotidyl transferase dUTP nick end labeling assay) assay and EdU cell proliferation labeling, the larvae were fixed in 4% paraformaldehyde, then stored in PBS solution at 4 • C overnight. As part of the TUNEL assay, detection and quantification of cell death were checked using In Situ Cell Death Detection Kit, TMR Red (12156792910; Roche, Basel, Switzerland), which was performed according to the manufacturer's instructions. For EdU labeling, we injected the EdU (1 nL, 10 mM) into the heart of larvae [32] and fixed the sample after 2 h; the subsequent experiments followed the protocol of Click-iT EdU Kit (C10340; Life Technologies, Carlsbad, CA, USA) to label the proliferation cells at S phase.

Fluorescence-Activated Cell Sorting (FACS) Analysis
Single-cell suspensions of zebrafish CHT (caudal hematopoietic tissue) were fulfilled as reported previously [33]. In brief, 25 Tg(CD41:GFP) larvae CHT region were digested with 0.25% trypsin at 28.5 • C for 40 min. Then, the cell suspensions were obtained by pipetting and filtration of a 40 µm cell strainer. All FACS analyses were performed by MoFlo XDP (Beckman Coulter, Brea, CA, USA) following the manufacturer's instructions and reported previously [34].

Quantification and Statistics
Statistical analyses were performed by GraphPad Prism6.0 (GraphPad Software Inc., San Diego, CA, USA). The positive signals areas in larval CHT (caudal hematopoietic tissue) were manually scored and double-confirmed blindly. All quantified data (Mean ± SEM) were analyzed by two-tailed Student's t test. The significant difference was indicated by a p-value < 0.05 statistically.

A Wide-Range Drug Screen for HSPCs Expansion in Zebrafish Using an FDA-Approved Library
During zebrafish hematopoiesis, the caudal hematopoietic tissue (CHT) region is functionally similar to the mammalian fetal liver, which is an HSPCs expansion site [35]. Therefore, we focused on this region and characterized the HSPCs proliferation signatures in different stages. We selected the Tg(CD41:GFP) transgenic line, in which GFP high cells mark thrombocytes, while GFP low cells label HSPCs [27]. Then, we used phosphorylated histone H3 (PH3) immunofluorescent staining to indicate proliferative cells in the G2/M phase ( Figure 1A). The quantification data ( Figure 1B) showed that PH3 + signals in CD41-GFP low populations increased from 2.33% ± 0.55% (2.5 dpf, days post fertilization) to 9.96% ± 0.90% (4 dpf) and then decreased to 6.78% ± 0.89% (5 dpf).
Based on the statistical analysis, we concluded that HSPCs have high proliferation capacity from 3 to 4 dpf during zebrafish embryogenesis. Therefore, we paid attention to this time frame to design a preliminary drug screening system. Firstly, we collected 3 dpf wild-type embryos into 12-well plates prior to adding 1 mL egg water with 10 µM drug. After 24 h of treatment, we fixed these embryos at 4 dpf to detect the HSPCs by examining cmyb (a HSPCs marker) [36] signals using whole-mount in situ hybridization ( Figure 1C). In subsequent experiments, we used this model to conduct large-scale drug screening to find candidates in promoting HSPCs expansion (results are summarized in Table A1 in Appendix A). 3.5 dpf, 6.26 ± 0.43; 4 dpf, 9.96 ± 0.90; 6.78 ± 0.90). (C) An overview of the experimental design in this study for drug screening by using zebrafish. A total of 20 wild-type embryos (3 dpf) were transferred to each well in a 12-well plate format. Then, embryos were administrated with one of 171 FDA-approved drugs for 24 h and screened for quantitative increases or decreases of signals in the CHT (caudal hematopoietic tissue) region at 4 dpf. The red box indicates the CHT region, and blue arrowheads indicate cmyb + signals. Mean ± SEM, n = 7; Scale bar, 50 µm.

Identification and Characterization of Drugs in Controlling HSPCs Homeostasis
Based on the initial screening data, we divided the treated samples into normal (standard), increased, and decreased groups, according to the quantified areas of cmyb + signals ( Figure 2A and Table A1). Ultimately, we screened 171 FDA-approved drugs. Among them, a high percentage of compounds (69%, 118 of 171) failed to alter HSPC homeostasis, which was set as normal or standard. However, 21 (12%) and 32 (19%) drugs led to increased or decreased pools of cmyb + HSPCs, respectively ( Figure 2B). These drugs were classified to 7 groups, including antibacterial, antineoplastic, glucocorticoid, NSAIDS, vitamin, antidepressant, and antipsychotic drugs ( Figure 2C).

The Contribution of Vitamin Drugs to HSPCs Expansion and Mitigation of the Hematopoietic Phenotypes in Ikzf1 −/− Mutants by Ergocalciferol and Panthenol
Among the drugs that led to HSPCs expansion, we attempted to find suitable molecules for clinical application with limited side-effects and easy delivery. After deliberating, we determined that vitamin drugs met our requirement. Consistently, there is research demonstrating that vitamin D receptor (VDR) signaling is essential in HSPCs production and differentiation [37][38][39] and that vitamin A-retinoic acid signaling regulates HSC dormancy [40], suggesting that vitamin drugs may play an important role in HSPCs maintenance.
In order to validate the effects of vitamin drugs on HSPCs expansion, we set out to seek a HSPCs proliferation-defective mutant. Ikzf1 is a Krüppel-like zinc-finger transcription factor that plays a crucial role in the development of T and B cells. Additionally, loss of Ikzf1 leads to compromised HSPCs expansion [41]. Therefore, we used these six vitamin drugs to treat ikzf1 −/− mutants. The cmyb in situ hybridization results presented that only ergocalciferol and panthenol treatment enlarged the HSPCs population in ikzf1 −/− mutants ( Figure 3E). The statistical results indicated that the rescue efficiency on the ikzf1 −/− mutants blood defect phenotypes was 50% (6/12) by ergocalciferol and 71.4% (10/14) by panthenol ( Figure 3E). Consistently, we also adopted flow cytometry to analyze the proportion of CD41-GFP low cells within the whole CHT cells after treating ergocalciferol and panthenol ( Figure 3F). The quantification results indicated that, compared to the DMSO group (0.44% ± 0.041%), the proportion increased markedly to 0.78% ± 0.015% by ergocalciferol and 0.84% ± 0.019% by panthenol ( Figure 3G). This data supports the drastic effects of ergocalciferol and panthenol on the HSPCs expansion.

The Dose-Dependent Effects of Ergocalciferol and Panthenol on HSPCs Expansion
Because of their impressive effect on HSPCs expansion and ikzf1 −/− mutants phenotypes mitigation, we selected ergocalciferol and panthenol for further study. Due to their parallel drug impact on HSPCs expansion, we were curious about whether the two molecules shared the similar structures. We referred to the structural formula of ergocalciferol ( Figure S1A) and panthenol ( Figure S1B). Ergocalciferol (C 28 H 44 O) and panthenol (C 9 H 19 NO 4 ) belong to the vitamin D or vitamin B family, and there is an enormous range in molecular weight (396.65 to 205.25). Those data showed that ergocalciferol and panthenol structures were quite distinct and mechanisms on HSPCs expansion may be different. Furthermore, in order to elicit the impact of the two drugs on cell apoptosis, we counted TUNEL + signals in the CHT region. The confocal images and statistical analysis indicated that ergocalciferol (5 ± 1) and panthenol (3 ± 1) failed to affect HSPCs apoptosis, compared to the DMSO group (3 ± 1) ( Figure S1C,D).

Discussion
In the FDA approval process, 73-82% of projects remained active in Phase II; however, 57-60% of the projects failed because of poor efficacy due to insufficient of human data [44]. Therefore, screening FDA-approved 'old' drugs is advantageous because of their established safety testing in humans, which saves time and cost. To this end, we performed high-efficiency screening of FDA-approved compounds and attempted to identify candidates that promote HSPCs expansion-in total, we screened 171 drugs and obtained 21 drugs. However, our screening also identified 32 drugs that cause diminishment of HSPCs, which might be useful in the alleviation of malignant blood diseases like leukemia. Meanwhile, we discovered 118 drugs whose administration had no effects on HSPCs expansion, most likely because the concentration (10 µM) was not sufficient for drug efficiency display in our preliminary screening, which probably resulted in some useful drugs missing out. However, we had chosen the following concentration that was typically used for compound screening assays (1-10 µM) [45]. Nevertheless, we plan to increase drug concentration to screen these drugs again in future studies.
Among the 21 drugs, we focused on vitamin drugs because of their low side-effects. The conventional view is that vitamins are organic compounds that people need in small quantities from foods. Uptake deficiency leads to hypovitaminosis, such as nyctalopia (vitamin A deficiency) and rickets (vitamin D deficiency). Additionally, vitamins and their derivatives have been applied to clinical therapeutics, such as acute promyeloid leukemia (APL) [46,47]. From our results, we found six vitamins, including biotin, α-tocopheryl acetate, ergocalciferol, panthenol, ascorbic acid, and retinol, that contribute to HSPCs expansion to a large degree. Interestingly, ergocalciferol and panthenol were able to ameliorate the HSPCs expansion deficiency phenotypes in ikzf1 −/− mutants. Consistently, the impact on HSPC expansion was dose-dependent, indicating that the effect of the drugs is specific as it has a dose-dependent action.
As a steroid hormone, vitamin D plays a role in regulating the metabolism of calcium and phosphate. Ergocalciferol (vitamin D2) belongs to the vitamin D family and is derived from the plant sterol ergosterol [48]. To date, no study has uncovered its role in hematopoiesis. Its analog, 1,25(OH)D3, an active form of vitamin D3, has been reported to stimulate HSPCs production via vitamin D receptor (VDR)-induced transcription, which can activate the expression of inflammatory cytokine CXCL8. However, when we used this drug (calcitriol, 20 µM) to treat embryos at 3 dpf, it was found to be lethal to larval zebrafish. The probable reason was that zebrafish embryos were unbearable at this concentration. Therefore, we will attempt to find the appropriate concentrations for larval zebrafish, to investigate the effects in the future. Based on our results, other analogs made no contribution to HSPCs expansion. Therefore, we hypothesized that the modulation mechanism of ergocalciferol may depend on an undiscovered pathway. Panthenol (provitamin B5) is a precursor of pantothenic acid (vitamin B5), which is an essential part of coenzyme A. This enzyme plays a significant role in the metabolism of cells, including the transfer of the acyl group during fatty acid biosynthesis and gluconeogenesis. It also promotes fibroblast proliferation and therefore promotes wound healing [49,50]. Nonetheless, the role of panthenol in HSPCs expansion is unclear.
Although our emphasis was directed toward ergocalciferol and panthenol, it did not indicate that other drugs with promotion or inhibition activities were dispensable. For instance, drugs with inhibition activities may be used to treat diseases related to abnormal proliferation of blood cells. For a substantial portion of these drugs, the detailed mechanisms remain unclear. In the future, understanding and characterizing the specific cellular targets of these drugs will be important and interesting. In addition, we used a drug concentration of 10 µM in the preliminary screening to save time and improve efficiency. However, this concentration might be insufficient to achieve a medicinal effect for certain types of drugs due to the different efficacy possessed by each drug [51,52], which is highly related to its properties, binding receptor, mechanism, and signaling pathway.
In summary, we adopted zebrafish as an in vivo model system to screen and evaluate FDA-approved drugs. Our aim was to identify novel molecules that influence HSPCs proliferation and provide a basis to begin to explore possible drugs to facilitate HSPCs expansion. Ultimately, we validated the effectiveness of ergocalciferol and panthenol. In future studies, we will investigate the mechanism of drug action and explore the possibility of clinical trials.

Conclusions
Collectively, we conducted a wide-range screening of FDA-approved drugs and uncovered a series of compounds that stimulate HSPCs expansion in zebrafish embryos, especially ergocalciferol and panthenol. Our study demonstrates that these drugs have potential for clinical application. In the future, more studies are required to characterize these drug targets and determine their utility and efficacy in mammalian or human disease models.

Supplementary Materials:
The following supporting information is available online at https://www. mdpi.com/article/10.3390/cells10082149/s1. Figure S1: The structural formulas of ergocalciferol and panthenol and their effects on cell apoptosis.