Chapter 5. The molecular pathogenesis of CALR mutations using zebrafish animal
3. Materials and Methods
Wild-type AB strain of zebrafish (Danio rerio) and the transgenic lines Tg(cd41:GFP)117 and Tg(fli1:EGFP)118 were maintained and manipulated with standard measure as previously described.119 The stages of embryonic development were determined based on Kimmel et al.120 Pigmentation was blocked by using 0.003% 1-phenyl-2-thiourea in some experiments. For pharmacologic inhibition, embryos were incubated with ruxolitinib (Abmole Bioscience, Houston, TX, USA) or fedratinib (Abmole Bioscience) from 1-2 cells stage to 5 days post fertilization (dpf) with or without microinjection of CALR mRNA. The zebrafish experiments were approved by the MacKay Memorial Hospital Animal Care and Use Committee.
3.2 Identification of zebrafish ortholog of human CALR
Human genes located in 19p13.11-13.2 were identified using the National Center for Biotechnology Information (NCBI) Map Viewer.121 Genes surrounding the 3 zebrafish calr genomic regions were identified using Ensembl122 and Synteny database.123 Human CALR protein sequence was used to BLASTP against zebrafish GRCz10 using the Ensembl platform (Ensembl release 82).122 Alignment and comparative analysis between protein sequences was performed using the Clustal Omega algorithm124 and edited by GeneDoc.125
3.3 Human and zebrafish CALR cDNAs cloning and mRNA synthesis
Full-length wild-type CALR cDNA was cloned from K562 cells into T&A™ Cloning Vector (Yeastern Biotech Co., Taipei, Taiwan) according to the manufacturer’s protocol (Forward primer: 5'-GATCCTCGAGATGCTGCTATCCGTGCCGCTGC-3'; reverse primer: 5'-GATCGAATTCCTACAGCTCGTCCTTGGCCTGGC-3'; restriction enzyme sites in bold letters: XhoI-EcoRI). Human CALR type 1 (CALR-del52) and type 2 (CALR-ins5) mutated cDNAs were obtained by custom gene synthesis. Full-length CALR cDNAs were subcloned in the pCS2+ vector and into a bicistronic pSYC-102 T2A vector (a gift from Dr. Seok-Yong Choi) replacing the mCherry-CAAX reporter gene using the In-Fusion Cloning Kit (Clontech, Mountain View, CA, USA) (Figure 13).126 All vector sequences were verified by sequencing. The mMessage mMachine SP6 kit (Ambion, Austin, TX, USA) was used for in vitro transcription of capped mRNAs from vectors according to the manufacturer’s protocol. mRNAs from the bicistronic pSYC-102-CALR vectors were only used to express EGFP and CALR concurrently in wild-type zebrafish embryos and only embryos expressing green fluorescence were collected under fluorescence microscope for use in the reverse-transcription and real-time polymerase
chain reaction.
3.4 Morpholinos and microinjection
Morpholinos (MOs) blocking splicing of mpl and epor, and translation (ATG/5'UTR) of csf3r were purchased from Gene Tools (Philomath, OR, USA) (MO sequences are listed in Table 12).117,127,128
Standard control MO was used as negative control. Embryos at the 1-2 cells stage were injected with MO (1 ng) or mRNAs (100 pg). Co-injection of each MO and CALR mutant mRNA was performed in a subset of embryos.
3.5 Reverse-transcription and real-time polymerase chain reaction
Total RNA was extracted from embryos using miTotal Miniprep System (Viogene, New Taipei City, Taiwan) and reverse transcribed using a High Capacity cDNA Reverse Transcription Kits (Applied Biosystems, Foster City, CA). Primer sequences are listed in Table 13. Fast SYBR® Green Master Mix (Applied Biosystems) was used for real-time quantitative polymerase chain reaction according to the manufacturer's instructions.
3.6 Western blotting
Total proteins were extracted from zebrafish embryos at 24 hours post fertilization (hpf).
Equal amounts of protein were denatured and electrophoresed. Membranes were immunoblotted with the following primary antibodies: CALR (Abcam, Cambridge, UK;
recognizing N-terminal sequences of both human and zebrafish wild-type CALR proteins), gapdh and customized mutant CALR (GeneTex, Hsinchu City, Taiwan; specifically recognizing human CALR exon 9 indel mutant protein sequence), STAT5 (Santa Cruz, Dallas, TX, USA), and phospho-STAT5 (Cell Signaling, MA, USA).
3.7 Imaging
Live embryos were imaged using a Leica MSV269 fluorescence stereomicroscope and photographed using a Leica DFC425 C digital camera and Leica Application Suite software. GraphPad Prism 7 software and ImageJ (National Institutes of Health) were used to process images.
3.8 Statistical analysis
The Student t test or ANOVA test were used. Data were expressed as mean ± standard error of the mean (SEM). Significance was determined at P<0.05(*), P<0.01(**) and P<0.001(***).
4. Results
4.1 Zebrafish ortholog of human CALR
To search for the zebrafish ortholog of human CALR gene, human CALR protein sequence was used to BLASTP against zebrafish GRCz10 (Ensembl release 82). We identified 3 annotated zebrafish orthologs of the human CALR gene (ENSG000001792), calr (ENSDARG00000076290 at chromosome 8), calr3a (ENSDARG00000103979 at chromosome 22) and calr3b (ENSDARG00000102808 at chromosome 2). After comparative analysis using the Clustal Omega algorithm, the amino acid sequence of zebrafish calr, calr3a, and calr3b proteins shares an overall 75%, 71% and 70% identity to human CALR protein sequence, respectively. The 3 functional domains in CALR are conserved in all 3 zebrafish calr proteins, including the KDEL ER retention signal at the C-terminus (Figure 14A). In addition, the genomic loci surrounding human chromosome 19p13.2 containing the CALR gene are syntenic with the regions of zebrafish calr on chromosome 8, calr3a on chromosome 22 and calr3b on chromosome 2 based on the
search in NCBI Map Viewer, Ensembl database and Synteny database (Figure 14B). These results indicated that the 3 zebrafish calr genes are likely true orthologs of human CALR.
4.2 Effects of mutant CALR expression on thrombopoiesis and angiogenesis in zebrafish
For the expression of mRNA in zebrafish embryo, we first performed a dose-finding study ranging from 50 to 200 pg CALR mRNA. Phenotype could be observed at dose of 100 pg mRNA per embryo which was compatible with normal development for most embryos. All CALR proteins were adequately expressed at comparable amount at dose of 100 pg (Figure 15A middle panel). The expression of CALR-del52 and CALR-ins5 mutant proteins was also confirmed by mutant CALR specific antibody (Figure 15A top panel). Therefore, 100 pg mRNA was injected throughout the study. To determine whether mutant CALR had an effect on hematopoietic stem and progenitor cells (HSPCs) in zebrafish, we injected the 3 mRNAs encoding CALR wild-type (CALR-wt), CALR-del52, and CALR-ins5 into 1-2 cells stage embryos of the cd41:GFP line, and the numbers of cd41+ cells in the caudal hematopoietic tissue (CHT) at 3 dpf indicating the HSPCs were counted.117 Expression of both CALR-del52 and CALR-ins5 mutant mRNA significantly increased the numbers of HSPCs in the CHT when compared with CALR-wt mRNA (Figure 15B). However, the numbers of HSPCs did not have statistically significant difference between CALR-del52 and CALR-ins5 mutant groups at this developmental stage. To ascertain that the increase of HSPCs was not affected by the change in angiogenesis during early development, mRNAs encoding CALR-wt, CALR-del52, and CALR-ins5 were injected into 1-2 cells stage embryos of the fli1:EGFP line. No obvious changes in the angiogenesis were visualized in CALR-wt and mutant CALR expressing embryos at 3 dpf when compared with uninjected control (Figure 15C). To determine whether CALR had an effect on mature thrombocyte,
the numbers of cd41+ thrombocytes in the cd41:GFP line were counted at 5 dpf after injection. Mutant CALR-del52 significantly increased the number of cd41+ thrombocytes (mean 162.5±4.1 per embryo) when compared to CALR-wt (mean 117.1±3.1 per embryo, P<0.001), mutant CALR-ins5 (mean 128.3±6.1 per embryo, P<0.001) and uninjected control (mean 136.7±3.0 per embryo, P<0.001) (Figure 15D). Although mutant CALR-ins5 slightly increased the number of cd41+ thrombocytes when compared to CALR-wt, there was no statistically significant difference. Together, our data demonstrated that the effect of mutant CALR on thrombopoiesis in zebrafish is dependent on the presence of the novel C-terminus and is also related to specific CALR mutant protein sequences.
4.3 Mutant CALR requires mpl to cause thrombocytosis in zebrafish
To test whether cytokine receptors are involved in the pathogenesis of thrombocytosis caused by mutant CALR in zebrafish, mpl, epor and csf3r MOs (1 ng) were injected in 1-2 cells stage embryos of cd41:GFP line and assayed for their effects on the number of cd41+ thrombocytes at 5 dpf. Co-injection of CALR-del52 mutant mRNA (100 pg) with each MO was also performed in a subset of embryos. At 5 dpf, the number of cd41+ thrombocytes significantly decreased upon mpl knockdown (mean 43.6±4.9 per embryo) when compared to the control MO group (mean 123.5±5.9 per embryo, P<0.001) and the mutant CALR-del52 group (P<0.001) (Figure 16). Importantly, co-injection of CALR-del52 mutant mRNA (mean 73.7±5.1 per embryo) can only partially reverse the knockdown effect of mpl MO. In contrast, the numbers of cd41+ thrombocytes did not decrease significantly upon epor MO (mean 110.6±5.5 per embryo) or csf3r MO (mean 116.6±5.6 per embryo) knocked-down compared with the control MO group. When CALR-del52 mutant mRNA was co-injected with epor (mean 151.7±9.2 per embryo) or csf3r (mean 153.6±7.2 per embryo) MOs, the numbers of cd41+ thrombocytes were comparable to those of
CALR-del52–injected embryos (both P=0.3). Collectively, these findings indicated that the expression of mutant CALR causes thrombocytosis through an mpl-dependent mechanism in zebrafish.
4.4 Effects of CALR mutants on lineage-specific and cytokine gene expression
The increase in thrombopoiesis upon expression of mutant CALR prompted us to evaluate their effects on hematopoietic lineage-specific, thrombopoiesis,129 cytokine and cytokine receptor gene expression in zebrafish embryos at 3 dpf. The expression of HSC gene runx1 was significantly upregulated in CALR-ins5 group but was modestly downregulated in CALR-del52 group (Table 14). Also, the expression of c-myb and scl was only downregulated in CALR-del52 group. Although gata1 was modestly downregulated in mutant CALR groups, the expression of α-eHb and β-eHb was not affected by both CALR mutants. The expression of early (spi1b) and late myeloid (mpo: granulocytic; l-plastin:macrophage) lineage genes, epo and epor showed no significant changes. However, the expression of lymphoid lineage genes (rag1, rag2 and lck) was modestly downregulated in mutant CALR groups. While the expression of mpl was significantly downregulated in both mutant CALR groups, both tpo and csf3r expressions were only downregulated in CALR-del52 group. In the group of genes related to thrombopoiesis, only the expression of nbeal2 was significantly downregulated in CALR-del52 group.
4.5 Effects of mutant CALR on jak-stat signaling in zebrafish
We then investigated whether the expression of mutant CALR can activate the jak-stat signaling in zebrafish. The expression of CALR-del52 mRNA significantly increased stat5 phosphorylation (Figure 17A, lane 2). Furthermore, treatment with ruxolitinib and fedratinib significantly ameliorated the enhanced stat5 phosphorylation induced by
CALR-del52 mRNA (Figure 17A, lane 3, 4). In addition, treatment with ruxolitinib significantly decreased the numbers of cd41+ thrombocytes in uninjected control as well as CALR-del52–injected embryos in a dose-dependent manner (Figure 17B). Whereas, treatment with fedratinib only had minimal inhibitory effect on the number of cd41+ thrombocytes in uninjected control embryos, and had a modest and significant dose-independent inhibitory effect on mutant CALR induced thrombocytosis (Figure 17C).
Our results demonstrated that mutant CALR mediated pathogenic thrombopoiesis involves jak-stat activation that can be blocked by JAK inhibitors.
5. Discussion
In this study, we have used the zebrafish animal model to examine the pathogenesis of mutant CALR in MPNs. We first identified 3 zebrafish orthologs for human CALR gene.
We have shown that expression of the CALR-del52 mutant disturb thrombopoiesis and increase the number of HSPCs in the CHT followed by significant thrombocytosis in the zebrafish embryo. These findings are consistent with the myeloproliferative phenotype in retroviral mouse bone marrow transplantation models elicited by mutant CALR expression characterized by thrombocytosis and megakaryocytic hyperplasia recapitulating those seen in patients with ET and myelofibrosis.40,42
The highly conserve protein sequences between human CALR and the 3 zebrafish calr genes suggested functional conservation between human and zebrafish CALR. Ma et al.
recently reported that MO knockdown of calr perturbs myeloid and HSCs lineages during zebrafish embryonic development including a decrease in the expression of genes associated with myeloid lineages at 24 hpf and an increase in the expression of cmyb at 48, 72 and 96 hpf.130 We have also shown that the expression of genes involved in
lineage-specific hematopoiesis, thrombopoiesis, cytokines and cytokine receptors could be perturbed by the expression of mutant CALR in zebrafish during early development. These data suggested that zebrafish calr genes play an important role in the regulation of vertebrae hematopoiesis. In addition, our data suggested that mutant CALR does not promote thrombopoiesis through the upregulation of mpl and tpo levels. Rather, the downregulation of mpl and tpo might represent a negative-feedback mechanism related to increased thrombopoiesis due to mutant CALR expression.
Based on the data from the murine and zebrafish animal models, the causative relationship between CALR mutations and thrombocytosis can be confirmed, and CALR mutations have been established as one of the driver mutations in MPNs. Furthermore, we demonstrated that expression of CALR-del52 (type 1 mutation) causes higher thrombocyte count than CALR-ins5 (type 2 mutation) at 5 dpf in zebrafish embryo. Similar finding has been reported in murine model that marked thrombocytosis was rapidly induced in CALR-del52-expressing mice and then progressed to myelofibrosis, and CALR-ins5-expressing mice only developed modest thrombocytosis resembling mild ET phenotype.40 This is consistent with the clinical finding that CALR-del52 mutation is more frequently detected in PMF than in ET,18 and also confirms the differential effects of CALR variants on thrombopoiesis and clinical phenotypes.90,131,132
To further elucidate the molecular pathogenesis of mutant in our zebrafish model, we have used MO knockdown experiments to show that only the mpl MO can significantly attenuate the effect of mutant CALR on thrombopoiesis. Both epor and csf3r MOs were not able to inhibit the effect of mutant CALR. These findings indicated that mpl has an essential and specific role required by mutant CALR to cause thrombocytosis in zebrafish.
Because CALR is physiologically functioning as a chaperone for MPL, it is reasonable to speculate that mutant CALR may interact directly with mpl to cause thrombocytosis in zebrafish. Our data are consistent with those recently reported by several groups of researchers using in vitro cell line models.39-43 In these studies, both the novel C-terminus of CALR mutants and the direct interaction of mutant CALR with MPL receptor are required to activate MPL and the downstream JAK-STAT signaling which in turn is responsible for cytokine-independent growth of Ba/F3-MPL and UT-7/TPO cell lines.
Chachoua et al. reported that the specific activation of MPL receptor by mutant CALR required both the presence of extracellular N-glycosylation residues of MPL and the glycan-binding site at the novel C-terminus of mutant CALR.41 In addition, Elf et al.
reported that the physical interaction between mutant CALR and MPL is dependent on the positive electrostatic charge of the C-terminus of the mutant CALR but not dependent on specific novel C-terminal sequence.42 Recently, Balligand et al. reported similar finding that highly similar but not identical murine Calr exon 9 frameshift mutants also require Mpl interaction to activate the JAK/STAT signaling.133 Moreover, the positive charge predominant novel C-terminus of the mutant CALR results in different calcium binding capacity which may alter calcium homeostasis and signaling processes in mutant cells. All these structural differences and changes will likely contribute to the different clinical phenotypes seen in different CALR variants.90,131,132,134
We have also demonstrated that the expression of human CALR mutant is able to activate jak-stat signaling in zebrafish. In addition, jak-stat signaling in zebrafish can also be inhibited by JAK2 inhibitors used in clinical trials illustrating that the conserved signaling machinery in human and zebrafish.135-137 Our data showed that ruxolitinib treatment results in a dose-dependent inhibitory effect on both normal thrombopoiesis and thrombocytosis
caused by mutant CALR in zebrafish. By contrast, JAK2-selective inhibitor fedratinib has only minimal inhibitory effects on normal thrombopoiesis but has modest and dose-independent inhibitory effect on thrombocytosis caused by mutant CALR. Our data suggested that fedratinib can normalize the thrombocytosis caused by the expression of mutant CALR and does not cause significant thrombocytopenia in zebrafish model. These observations are comparable with the findings that both ruxolitinib and fedratinib have been demonstrated to have clinical responses in MPN patients harboring CALR mutations.138-140 However, fedratinib has less hematological toxicities than ruxolitinib especially thrombocytopenia which is a dose-limiting toxicity of ruxolitinib.135-137 Despite both JAK inhibitors are effective in the reduction of splenomegaly and the relief of clinical symptoms, they are not likely to substantially modify the natural history of the BCR-ABL-negative classic MPNs including CALR-mutated PMF. Importantly, these JAK inhibitors are not specifically designed for JAK2V617F mutation. However, the unique pathogenic mechanism of mutant CALR in MPNs has led to the possibility of new therapeutic approach targeting the interaction and binding between mutant CALR and MPL. In this regard, our results highlight the advantage and support the use of zebrafish as a relevant in vivo whole organism model for the testing and screening of therapeutic compounds targeting mutant CALR.115
In conclusions, we have used the zebrafish model to show that mutant CALR promotes the activation of jak-stat signaling through an mpl-dependent mechanism to mediate pathogenic thrombopoiesis during zebrafish early hematopoiesis. These findings are consistent with those observed in in vitro cell line and mouse models and illustrated that the signaling machinery related to mutant CALR tumorigenesis are conserved between human and zebrafish. Zebrafish has also been shown to be a relevant in vivo model for the
development of novel therapeutic compounds targeting mutant CALR. Future studies using stable mutant CALR transgenic or knock-in zebrafish models for this purpose will be warranted.
Chapter 6
Conclusions and future work
In this dissertation, we have examined the roles of CALR mutations in MPNs through clinical studies and basic research using the zebrafish animal models. CALR mutations have now been recognized as an important clonal marker in MPNs. Therefore, we developed a rapid and sensitive HRMA with the CFX Connect real-time system to detect CALR exon 9 mutations in ET patients. We have also evaluated the clinical and prognostic significance of CALR mutations and JAK2/CALR co-mutations in Taiwanese ET patients.
We confirmed that CALR mutations were associated with younger age, higher platelet count and lower hemoglobin level in ET patients. Interestingly, we also detected various CALR exon 9 alterations in 22% JAK2-mutated ET patients, and JAK2-mutated ET patients with concomitant CALR alterations were associated with oldest age, higher thrombotic events after diagnosis, higher major arterial thrombotic events after diagnosis and more patients being high risk group for thrombo-hemorrhagic complications. Our data suggested that JAK2-mutated ET patients with concomitant CALR alterations probably define a specific subgroup of patients with increased risk of thrombotic events. After the evaluation of B cell immune profiles in a cohort of 54 adult Taiwanese ET patients, we confirmed that increased activated B cells were universally present in JAK2-mutated, CALR-mutated and triple-negative ET patients when compared to healthy adults. Finally, we used the zebrafish animal models to investigate the molecular pathogenesis of CALR mutations. Our findings showed that mutant CALR activates jak-stat signaling through an mpl-dependent mechanism to mediate pathogenic thrombopoiesis in zebrafish, and illustrated that the signaling machinery related to mutant CALR tumorigenesis are conserved between human and zebrafish.
Although our results have shed lights on the roles of CALR mutations in MPNs, several experiments may still need to confirm and validate our findings in the future. For example, transgenic zebrafish neutrophil reporter line driven by mpx promoter can be used to investigate the role of BAFF expression in neutrophils and B cells activation. In addition, the generation of stable mutant CALR zebrafish lines using transgenic or knock-in methods will be need to validate our findings using MO knock-down zebrafish models. Moreover, stable mutant CALR zebrafish lines can also be used to conduct high throughput CALR mutant specific drug screening. It is also noteworthy that the role of JAK1-mediated signaling associated with mutant CALR can be studied because the JAK1/JAK2 inhibitor ruxolitinib had significantly higher inhibitory effect on mutant CALR-mediated thrombopoeisis in zebrafish than that of the JAK2-selective inhibitor fedratinib. We believe that the results of these translational researches will further advance our understanding on the roles of CALR mutations in MPNs and may also help to develop novel therapeutic compounds targeting mutant CALR in the future.