A Zebrafish JAK1-A634D Gain-of-Function Model Provides New Insights into the Pathogenesis of Familial Hypereosinophilia

Introduction: JAK/STAT signaling is of major importance for hematopoiesis and immune function. A heterozygous JAK1 c.1901C>A de novo point mutation was identified in three members of a family (a mother and two male sons). These patients presented with a unique clinical phenotype of severe atopic dermatitis, markedly elevated peripheral blood eosinophil counts with eosinophilic infiltration of the liver and gastrointestinal tract, hepatosplenomegaly, and failure to thrive. Treatment with ruxolitinib, a JAK1/2 inhibitor, resulted in reduced eosinophilia and improved growth (Del Bel et al. JACI 2017). However, the role of this JAK1 mutation in hematopoiesis remains unclear.

Methods: ubi-JAK1-WT or A634D-P2A-sfGFP-pA zebrafish transgenic lines were generated incorporating the human genes. Both of these constructs were injected into embryos, but the survival rate for the JAK1-A634D embryos was very low. Six hour post-fertilization (hpf) embryos expressing the mutant transgene were treated with 1 µM ruxolitinib for 24h, which dramatically increased survival. Whole-mount in situ hybridization (WISH) was used to assess changes in hematopoietic stem cells/blood cell populations due to wild-type or JAK1-A634D expression at 24, 28, 36 and 48 hpf using a battery of blood lineage specific probes. O-dianisidine staining was employed to evaluate hemoglobin levels. ImageJ cell counter pluggin was used for cell quantification. Gata1 and o-dianisidine staining intensity was analyzed using Ilastik and CellProfiler. Six groups of 30-50 embryos were analyzed by RNA sequencing (RNAseq): control, JAK1-WT and JAK1-A634D at 28 and 36 hpf. RNA was extracted using Trizol. RNAseq data was analyzed with edgeR and enrichment analysis was performed using DAVID 6.8.

Results: JAK1-A634D was toxic to zebrafish embryos but they survived to sexual maturity post-ruxolitinib therapy. F1 embryos from JAK1-A634D founders with high expression of the transgene exhibited abnormal development (70% of embryos) which could be reduced to 12% of embryos following 24h exposure to 1 µM ruxolitinib. This rescue suggests that JAK1-A634D is highly active in zebrafish and its activity needs to be under tight regulation for normal development. Both JAK1-WT and A634D mutant fish demonstrated increased HSPCs (c-myb/runx1 expression) at 36 hpf (p<0.0001) through 48 hpf (A634D>WT>controls; p=0.0308 and p=0.0045, respectively). pu.1/spi1 early myeloid expression showed a similar expression pattern at 28 hpf, A634D>WT>controls; p<0.0001 for both). At 24 and 28 hpf, granulocytes expressing myeloperoxidase (mpx) were increased in both JAK1-WT and JAK1-A634D mutants, but particularly high in JAK1-A634D embryos at 48 hpf (p<0.0001). Macrophages marked by l-plastin (lcp) expression, showed similar cell numbers in JAK1-WT and controls at 24 and 28 hpf, but a significant decrease in JAK1-A634D embryos (p<0.0001). An increase of carboxypeptidase A5 positive mast cells was observed in both JAK1-WT and JAK1-A634D embryos at 48 hpf compared with controls (p<0.0001). We observed an increase of gata1a expression at 24 hpf in A634D embryos (p=0.0058, p=0.0065), but o-dianisidine staining at 48 hpf revealed a decrease in hemoglobin in both JAK1 lines (p<0.0001). The latter data is supported by RNAseq results revealing decreased expression of hemoglobin-associated genes at 36 hpf in both transgenic lines. RNAseq revealed a total of 365 genes differentially expressed, with enrichment for genes coding for proteosomal and ribosomal proteins. In the JAK-1WT we found 79 genes upregulated and 115 downregulated; in the JAK1-A634D, 94 genes were upregulated and 77 were downregulated. JAK-STAT pathway genes were upregulated in the A634D mutant (p<0.05), as well as a trend for genes associated with hematopoiesis (p=0.0633) and negative regulation of cell proliferation (p=0.0658).

Conclusions: JAK1-A634D plays an important role in zebrafish development and specifically impacts the myeloid lineage, particularly the granulocyte branch, from which eosinophils arise. Ruxolitinib can revert the arrested development phenotype, demonstrating the specificity of this targeted therapy. Our model provides the first in vivo evidence of the effects of the JAK1-A634D mutation on hematopoiesis and will provide further insight in understanding the cause of familial hypereosinophilia.