Role of microRNAs and their downstream target transcription factors in zebrafish thrombopoiesis.

Previous studies have shown that human platelets and megakaryocytes carry microRNAs suggesting their role in platelet function and megakaryocyte development, respectively. However, a comprehensive study on the microRNAs and their targets has not been undertaken. Zebrafish thrombocytes could be used as a model to study their role in megakaryocyte maturation and platelet function because thrombocytes have both megakaryocyte features and platelet properties. In our laboratory, we identified 15 microRNAs in thrombocytes using single-cell RNA sequencing. We knocked down each of these 15 microRNAs by the piggyback method and found knockdown of three microRNAs, mir-7148, let-7b, and mir-223 in adult zebrafish led to an increase in the percentage of thrombocytes. Functional thrombocyte analysis using plate tilt assay showed no modulatory effect of the three microRNAs on thrombocyte aggregation/agglutination. We also found enhanced thrombosis using arterial laser thrombosis assay in a group of zebrafish larvae after mir-7148, let-7b, and mir-223 knockdowns. These results suggested mir-7148, let-7b, and mir-223 are repressors for thrombocyte production. We then explored miRWalk database for let-7b downstream targets and then selected those that are expressed in thrombocytes, and from this list based on their role in differentiation selected 14 genes, rorca, tgif1, rfx1a, deaf1, zbtb18, mafba, cebpa, spi1a, spi1b, fhl3b, ikzf1, irf5, irf8, and lbx1b that encode transcriptional regulators. The qRT-PCR analysis of expression levels of the above genes following let-7b knockdown showed changes in the expression of 13 targets. We then studied the effect of the 13 targets on thrombocyte production and identified 5 genes, irf5, tgif1, irf8, cebpa, and rorca that showed thrombocytosis and one gene, ikzf1 that showed thrombocytopenia. Furthermore, we tested whether mir-223 regulates any of the above 13 transcription factors after mir-223 knockdown using qRT-PCR. Six of the 13 genes showed similar gene expression as observed with let-7b knockdown and 7 genes showed opposing results. Thus, our results suggested a possible regulatory network in common with both let-7b and mir-223. We also identified that tgif1, cebpa, ikzf1, irf5, irf8, and ikzf1 play a role in thrombopoiesis. Since the ikzf1 gene showed a differential expression profile in let-7b and mir-223 knockdowns but resulted in thrombocytopenia in ikzf1 knockdown in both adults and larvae we also studied an ikzf1 mutant and showed the mutant had thrombocytopenia. Taken together, these studies showed that thrombopoiesis is controlled by a network of transcription regulators that are regulated by multiple microRNAs in both positive and negative manner resulting in overall inhibition of thrombopoiesis.

Role of MicroRNAs and their Downstream Target Transcription Factors in Zebrafish Thrombopoiesis.

Previous studies have shown that human platelets and megakaryocytes carry microRNAs suggesting their role in platelet function and megakaryocyte development, respectively. However, a comprehensive study on the microRNAs and their targets has not been undertaken. Zebrafish thrombocytes could be used as a model to study their role in megakaryocyte maturation and platelet function because thrombocytes have both megakaryocyte features and platelet properties. In our laboratory, we identified 15 microRNAs in thrombocytes using single-cell RNA sequencing. We knocked down each of these 15 microRNAs by the piggyback method and found knockdown of three microRNAs, mir-7148, let-7b , and mir-223 in adult zebrafish led to an increase in the percentage of thrombocytes. Functional thrombocyte analysis using plate tilt assay showed no modulatory effect of the three microRNAs on thrombocyte aggregation/agglutination. We also found enhanced thrombosis using arterial laser thrombosis assay in a group of zebrafish larvae after mir-7148, let-7b , and mir-223 knockdowns. These results suggested mir-7148, let-7b , and mir-223 are repressors for thrombocyte production. We then explored miRWalk database for let-7b downstream targets and then selected those that are expressed in thrombocytes, and from this list based on their role in differentiation selected 14 genes, rorca, tgif1, rfx1a, deaf1, zbtb18, mafba, cebpa, spi1a, spi1b, fhl3b, ikzf1, irf5, irf8 , and lbx1b that encode transcriptional regulators. The qRT-PCR analysis of expression levels of the above genes following let-7b knockdown showed changes in the expression of 13 targets. We then studied the effect of the 13 targets on thrombocyte production and identified 5 genes, irf5, tgif1, irf8, cebpa , and rorca that showed thrombocytosis and one gene, ikzf1 that showed thrombocytopenia. Furthermore, we tested whether mir-223 regulates any of the above 13 transcription factors after mir-223 knockdown using qRT-PCR. Six of the 13 genes showed similar gene expression as observed with let-7b knockdown and 7 genes showed opposing results. Thus, our results suggested a possible regulatory network in common with both let-7b and mir-223 . We also identified that tgif1, cebpa, ikzf1, irf5 , irf8 , and ikzf1 play a role in thrombopoiesis. Since the ikzf1 gene showed a differential expression profile in let-7b and mir-223 knockdowns but resulted in thrombocytopenia in ikzf1 knockdown in both adults and larvae we also studied an ikzf1 mutant and showed the mutant had thrombocytopenia. Taken together, these studies showed that thrombopoiesis is controlled by a network of transcription regulators that are regulated by multiple microRNAs in both positive and negative manner resulting in overall inhibition of thrombopoiesis.

Pancytopenia and thrombosis defects in zebrafish mutants of Fanconi anemia genes.

Progressive pancytopenia is a common feature observed in DNA crosslink repair deficiency disorder, Fanconi anemia (FA). However, this phenotype has not been recapitulated in single FA gene knockout animal models. In this study, we analyzed hematological characteristics in zebrafish null mutants for two FA genes, fanca and fanco. In adult mutants, we demonstrate age-associated reduction in blood cell counts for all lineages, resembling progressive pancytopenia in FA patients. In larval mutants, we demonstrate vascular injury-induced thrombosis defects, particularly upon treatment with crosslinking agent diepoxybutane (DEB), indicating DNA damage induced inefficiency of thrombocytes.

Splenectomy in zebrafish: a new model for immune thrombocytopenia.

In humans, splenectomy is performed to treat many clinical disorders, including immune thrombocytopenia. However, the incidence of splenectomies for immune thrombocytopenia as a therapeutic has significantly declined over the past decade due to the availability of new therapies. Infection and sepsis as a result of splenectomies are well documented, but other long-term effects are not well characterized. Evidence suggests that persons who have had a prior splenectomy may be at an increased risk of vascular conditions. Also, elevated levels of cell-derived microparticles appear to contribute to an increased risk of thrombosis and cardiovascular disease. However, in vivo studies on the increased levels of microparticles following splenectomy are limited. In order to understand the effects of splenectomies, we developed a protocol for splenectomy in adult zebrafish. After anesthesia, the spleen was removed under a stereomicroscope after making an incision on the ventral side of the fish. The spleen was removed by pulling with forceps. The incision was closed by Vetbond tissue glue. Blood collected from both splenectomized zebrafish and those that underwent sham surgeries was immunolabeled with polyclonal antisera against αIIb, followed by flow cytometry. We observed elevated levels of thrombocytes and their microparticles in splenectomized zebrafish. Finally, by injecting αIIb antibody intravenously into zebrafish, we found the thrombocyte counts decreased, suggesting the fish developed immune thrombocytopenia like conditions, which were then reversed by splenectomy. In summary, the model developed here should be useful to study molecular changes due to splenectomy. Also, the zebrafish will be useful in modeling treatment of immune thrombocytopenia like conditions.

Knockdown and Knockout of Tissue Factor Pathway Inhibitor in Zebrafish.

Tissue factor pathway inhibitor (TFPI) is an anticoagulant that inhibits factor VIIa and Xa in the blood coagulation pathways. TFPI contains three Kunitz domains, K1, K2, and K3. K1 and K2 inhibit factor VIIa and Xa, respectively. However, the regulation of TFPI is poorly studied. Since zebrafish has become an alternate model to discover novel actors in hemostasis, we hypothesized that TFPI regulation could be studied using this model. As a first step, we confirmed the presence of tfpia in zebrafish using reverse transcription polymerase chain reaction. We then performed piggyback knockdowns of tfpia and found increased coagulation activity in tfpia knockdown. We then created a deletion mutation in tfpia locus using the CRISPR/Cas9 method. The tfpia homozygous deletion mutants showed increased coagulation activities similar to that found in tfpia knockdown. Taken together, our data suggest that tfpia is a negative regulator for zebrafish coagulation, and silencing it leads to thrombotic phenotype. Also, the zebrafish tfpia knockout model could be used for reversing this thrombotic phenotype to identify antithrombotic novel factors by the genome-wide piggyback knockdown method.

Trypsin induces an aversive response in zebrafish by PAR2 activation in keratinocytes.

Previously we have shown that trypsin, a protein typically involved in digestion, is released from gills of both fresh and saltwater fishes into surrounding water under stress or injury. We have also shown that each species produces trypsin with different specific activities. In this report, using zebrafish as a model, we identified that trypsin induces an aversive response in zebrafish larvae and adult zebrafish. Since Protease-Activated Receptor 2 (PAR2) responds to trypsin, we tested whether the aversive response is dependent on the activation of PAR2 located on the zebrafish skin cells. Zebrafish larvae treated separately with neomycin and zinc sulfate also showed aversive response indicating neuromast, and olfactory cells are not involved in this aversion. Cultured keratinocytes from zebrafish showed a response to trypsin. Zebrafish larvae subjected to knockdown of par2a also exhibited reduced escape response. Similarly, par2a-deficient mutant larvae displayed no response to trypsin. Since it has been shown that stress activates PAR2 and sends signals to the brain as shown by the increased c-fos expression, we tested c-fos expression in adult zebrafish brains after trypsin treatment of adults and found enhanced c-fos expression by qRT-PCR. Taken together, our results show that the trypsin activates PAR2 on keratinocytes signaling the brain, and this pathway of trypsin-induced escape response will provide a unique communication mechanism in zebrafish. Furthermore, since PAR2 activation also occurs in pain/pruritus sensing, this model might be useful in elucidating components of signaling pathways in pain/pruritus.

Role of ribosomal RNA released from red cells in blood coagulation in zebrafish and humans.

Hemolytic disorders are characterized by hemolysis and are prone to thrombosis. It has previously been shown that the RNA released from damaged blood cells activates clotting. However, the nature of the RNA released from hemolysis is still elusive. We found that after hemolysis, red blood cells from both zebrafish and humans released RNA that contained mostly 5.8S ribosomal RNA (5.8S rRNA), This RNA activated coagulation in zebrafish and human plasmas. By using both natural and synthetic 5.8S rRNA and its truncated fragments, we found that the 3'-end 26-nucleotide-long RNA (3'-26 RNA) and its stem-loop secondary structure were necessary and sufficient for clotting activity. Corn trypsin inhibitor (CTI), a coagulation factor XII (FXII) inhibitor, blocked 3'-26 RNA-mediated coagulation activation in the plasma of both zebrafish and humans. CTI also inhibited zebrafish coagulation in vivo. 5.8S rRNA monoclonal antibody inhibited both 5.8S rRNA- and 3'-26 RNA-mediated zebrafish coagulation activity. Both 5.8S rRNA and 3'-26 RNA activated normal human plasma but did not activate FXII-deficient human plasma. Taken together, these results suggested that the activation of zebrafish plasma is via an FXII-like protein. Because zebrafish have no FXII and because hepatocyte growth factor activator (Hgfac) has sequence similarities to FXII, we knocked down the hgfac in adult zebrafish. We found that plasma from this knockdown fish does not respond to 3'-26 RNA. To summarize, we identified that an rRNA released in hemolysis activates clotting in human and zebrafish plasma. Furthermore, we showed that fish Hgfac plays a role in rRNA-mediated activation of coagulation.

Knockdown screening of chromatin binding and regulatory proteins in zebrafish identified Suz12b as a regulator of tfpia and an antithrombotic drug target.

Tissue factor pathway inhibitor (TFPI) is an anticoagulant protein that inhibits factor VIIa and Xa in the coagulation cascade. It has been shown that forkhead box P3 protein is a TFPI transcriptional repressor. However, there are no studies on chromatin remodeling that control TFPI expression. We hypothesized that the genome-wide knockdowns of the chromatin binding and regulatory proteins (CBRPs) in zebrafish could identify novel tfpia gene regulators. As an initial step, we selected 69 CBRP genes from the list of zebrafish thrombocyte-expressed genes. We then performed a 3-gene piggyback knockdown screen of these 69 genes, followed by quantification of tfpia mRNA levels. The results revealed that knockdown of brd7, ing2, ing3, ing4, and suz12b increased tfpia mRNA levels. The simultaneous knockdown of these 5 genes also increased tfpia mRNA levels. We also performed individual gene and simultaneous 5-gene knockdowns on the 5 genes in zebrafish larvae. We found that after laser injury, it took a longer time for the formation of the thrombus to occlude the caudal vessel compared to the control larvae. We then treated the larvae and adults with a chemical UNC6852 known to proteolytically degrade polycomb repressor complex 2, where SUZ12 is a member, and observed prolongation of time to occlude (TTO) the caudal vein after laser injury and increased tfpia mRNA levels in larvae and adults, respectively. In summary, our results have identified novel epigenetic regulators for tfpia and exploited this information to discover a drug that enhances tfpia mRNA levels and prolongation of TTO. This discovery provides the basis for testing whether UNC6852 could be used as an antithrombotic drug. This approach could be used to study the regulation of other plasma proteins, including coagulant and anticoagulant factors.

RNaseH-mediated simultaneous piggyback knockdown of multiple genes in adult zebrafish.

We recently developed a piggyback knockdown method that was used to knockdown genes in adult zebrafish. In this method, a vivo morpholino (VMO) piggybacks an antisense deoxyoligonucleotide (dO) into the somatic cells and reduces the cognate mRNA levels. In this paper, we tested whether we can piggyback more than one dO with one VMO. We designed various hybrids that had more than one dO that could be piggybacked with one VMO. We chose f7, f8, and αIIb genes and tested their knockdown by the appropriate assays. The knockdown with piggybacking either two or three dOs by one VMO yielded > 85% knockdown efficiency. We also performed knockdown of argonautes and rnaseh separately along with f7. We found the knockdown of f7 occurs when knockdown of argonautes happens and not when rnaseh knockdown was performed, suggesting that RNaseH is involved in mRNA degradation. In conclusion, we developed a method where we could knockdown three genes at one time, and by increasing the concentration of VMO by twofold, we could knockdown six genes simultaneously. These multiple gene knockdowns will not only increase the efficiency of the method in whole genome-wide knockdowns but will also be useful to study multifactorial disorders.

Effect of MS222 on Hemostasis in Zebrafish.

MS222 is a compound used in anesthetizing vertebrates, including fish and frogs. Several side effects of this anesthetic have been reported, but its effect on hemostasis has not been studied. In our laboratory, we have used zebrafish for more than 2 decades as a model system to study hemostasis. During this period, we have had trouble in collecting blood from anesthetized zebrafish and observed more rapid blood clotting than in nonanesthetized counterparts. However, no systematic studies regarding the effect of MS222 on zebrafish hemostasis are available. In this study, we performed various assays such as gill bleeding, measurement of Hct, total blood cell counts, thrombocyte counts, thrombocyte aggregation, and coagula- tion and measured the amount of blood collected. We found that Hct values, the amount of blood collected, bleeding, and coagulation differed significantly between anesthetized and nonanesthetized fish. Our results suggest that blood collected after MS222 anesthesia of zebrafish has altered hemostasis.