ALBERT

Description: Whole exome sequencing analyses are increasingly performed on patients presenting with suspected inherited disease but lacking classical mutations linked to presented phenotypes. Using whole-exome sequencing in SBDS-negative Shwachman-Diamond Syndrome (SDS) families, we recently identified three independent patients, each of whom carried a heterozygous de novo missense variant of SRP54 (encoding signal recognition particle 54 kDa). The SRP54 protein is a key component of the ribonucleoprotein complex that mediates the co-translational targeting of secretory and membrane proteins to the endoplasmic reticulum (ER). Whilst two of the identified patients were carrying nucleotide transversion in SRP54 (p.T115A and p.G226E), which manifested in typical SDS features like neutropenia and exocrine pancreatic insufficiency, the third patient was carrying a nucleotide deletion (p.T117Δ), which only manifested in mild neutropenia without additional SDS features (Carapito et al. 2017, JCI). Here, we describe a zebrafish knock-out (KO) mutant as the very first transgenic in vivo model of SRP54 deficiency, translate our previous findings into living organisms and propose disease-driving mechanisms. We show that homozygous srp54 mutant zebrafish are suffering not only from severe neutropenia as shown by flow cytometry and Whole-Mount-In-Situ Hybridization (WISH), but also from gross developmental defects leading to early embryonic lethality. In fact, srp54-/- zebrafish did not survive more than 72 hours post fertilization, indicating that complete loss of Srp54 is not compatible with life. Injection with wild-type human SRP54 mRNA induced transient restoration of SRP54 protein expression and slightly enhanced the survival of the homozygous mutants. However, long-term viability could not be restored, revealing that srp54 is not only critically required during early embryogenesis but also at later stages of development. Heterozygous siblings on the other hand are viable and display only mild neutropenia but no pancreas defects. Interestingly however, injection of mutant mRNAs of human SRP54 (p.T115A, p.T117Δ, p.226E) into heterozygous srp54 KO mutants aggravated the phenotype inducing more profound neutropenia and pancreas changes similar to those observed in classical SDS patients. Of note, these effects were more severe for the transversions p.T115A and p.G226E. Mutation p.T117Δ only caused a minor reduction in the number of neutrophils, without affecting the pancreas. To further investigate SRP54 driven neutrophil defects, we used lentiviral transduction to exogenously express human SRP54 mutant variants in promyelocytic HL-60 cells. When stimulating these cells to differentiate by ATRA treatment, we found significantly impaired morphologic differentiation and CD11b surface induction compared to control cells. The severity of these effects was again specific to the three different identified mutations, with p.T115A and p.G226E being more severe than p.T117Δ. These findings confirm the type-specific effects of SRP54 mutations and indicate that SRP54 defects interfere with neutrophil differentiation and thus ultimately lead to neutropenia. Collectively, we here describe a novel zebrafish disease model of SDS and congenital neutropenia founding on SRP54 as molecular driver. Our model demonstrates that at least one healthy allele of srp54 is pivotal for survival, which is in line with the findings in humans, where homozygous mutations in SRP54 have never been detected. We reveal that the phenotypic manifestation of heterozygous SRP54 mutations strongly depends on the type of mutation: while mutations likely causing a simple SRP54 loss of function (e.g. p.T117Δ) induce a rather mild phenotype characterized by moderate neutropenia only (analogous to the heterozygous fish mutant), more severe SDS-like phenotypes involve SRP54 mutations that exert dominant negative effects (e.g. p.T115A and p.G226E). Ultimately, we make use of the promyelocytic cell line HL-60 to propose neutrophil differentiation defects as the underlying cause of SRP54 driven neutropenia. At the time being, RNA sequencing and protein expression analyses are performed in our laboratory, which will add to the understanding of the mechanistical background of the neutrophilic differentiation blockage and eventually uncover novel treatment strategies for SRP54 deficiency. Disclosures No relevant conflicts of interest to declare.

Description: Mouse xenografts are routinely used for in vivo studies on human acute myeloid leukemia (AML) cells. However, a significant proportion of primary AML samples (~50-60% of cases) are not able to repopulate mice, or respectively require a particularly long time (up to 9 months) to show detectable engraftment in such models. Here we report that changing the transplantation time-point from day (5:00 pm) to night (4:00 am) strongly improved the engraftment of human AML cells in NOD/SCID/IL2Rgnull (NSG) mice. Sublethally irradiated gender-matched NSG mice of 6-8 weeks of age (n=40) were transplanted with primary AML cells (n=5 patients). For each sample, equal numbers of freshly thawed AML cells were injected via the tail vein at both time-points; transplanted mice were afterwards monitored for human leukemic engraftment in peripheral blood (PB) and bone marrow (BM) using multicolor flow cytometry and immunohistochemistry. In all n=5 analyzed AML patient samples, transplantation at night strongly promoted leukemogenesis, with 2/5 samples showing engraftment only in the night transplant condition and 3/5 samples showing engraftment in both but significantly higher leukemic burden in night vs. day transplanted mice (Fig. 1a), although similar leukemia-specific mutations were retrieved in mice of the two conditions by next generation sequencing analysis (Fig. 1b). Limiting dilution experiments (n=2 AML samples) further confirmed these findings and showed that in fact lower numbers of AML cells were required to initiate leukemia when cells were transplanted at night versus day times. Given that the immediate process that happens after intravenous injection of AML cells into mice involves homing to BM niches, we hypothesized that the time-point of transplantation influences engraftment by modulating homing. Therefore, we performed homing assays as described for healthy hematopoietic stem/progenitor cells (HSPCs); we injected equally treated human CFSE labeled AML cells via tail vein injection at day versus night time points and analyzed murine BMs 8 hours later for the presence of human leukemic cells. Indeed, significantly higher BM colonization by leukemic cells was observed after night vs. day time injections (0.31±0.09 vs. 0.13±0.18, p

Description: Sequencing analyses are increasingly performed on patients presenting with suspected congenital disease but lacking classical mutations linked to the presented phenotype. However, in case a novel mutation is found, its causal contribution to the patients' clinical symptoms is yet unclear and requires further exploration in functional studies. Here we use the zebrafish model to analyze the functional relevance of heterozygous SRP54 gene mutations identified in patients with unexplained neutropenia (T117del) or Shwachman-Diamond like disease involving severe neutropenia and exocrine pancreas insufficiency (T115A, G226E) (Carapito et al, 2017). The function of wildtype SRP54 protein was explored in hematopoiesis and pancreas development using two different antisense morpholino oligonucleotides (MO) and a zebrafish srp54 mutant. Reduced neutrophil numbers were observed in MO versus control injected fish when analyzed by WISH for mpx or using Tg(lyz:DsRed) and Tg(mpx:eGFP) lines. Morphants displayed not only quantitatively but also qualitatively impaired neutrophils, which migrated less to injury sites in tail fin injury assays. Furthermore, exocrine but not endocrine pancreas impairment was observed in MO versus control injected fish. Both neutrophiles and exocrine pancreas development could be rescued by co-injection with wildtype SRP54 mRNA. Interestingly, in the corresponding Sanger mutant, homozygous srp54-/- fish display not only most severe neutropenia but also gross developmental defects which cause early lethality at 2 days post-fertilization and thus impede assessment of pancreas development. Heterozygous srp54+/- fish are viable, show less severe neutropenia than homozygous fish and, interstingly, have no exocrine pancreas defect. The delineated experiments explore thus the effects of different dosages of residual functional SRP54 protein (wildtype 〉 srp54+/- 〉 morpholino treated 〉 srp54-/-) on phenotypic manifestation. The obtained results indicate that the dose of residual SRP54 protein that can functionally act within the SRP may dictate disease phenotype in patients with SRP54 mutations (neutropenia only 〉 more severe neutropenia and exocrine pancreas defects 〉 most severe neutropenia, early lethality). We thus hypothesize that the different mutations observed in patients differentially affect the functionality of the residual wildtype SRP54 protein and thus associate with isolated neutropenia (T117del) or more severe disease with Shwachman-Diamond like features (T115A, G226E). To mimic the heterozygote mutation status observed in patients, we injected the different mutated mRNAs in srp54+/- zebrafish. None of the injected mRNAs was able to rescue the neutropenia phenotype, indicating that all mutated proteins are functionally impaired. Confirming our hypothesis, injection of T115A and G226E in fact further aggravated neutropenia and additionally induced exocrine pancreas defects, while T117del did not. Taken together, these analyses support the notion that full deletion of SRP54 is not compatible with life and therefore homozygous SRP54 are not observed in patients. Instead, heterozygous SRP54 mutations may cause only neutropenia or alternatively the more severe Shwachman-Diamond like syndrom, depending on the nature of the underlying mutation (either functional null or dominant negative). Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2530 Poster Board II-507 A decade of research on human embryonic stem cells (ESC) has paved the way for the discovery of alternative approaches to generate pluripotent stem cells. Combinatorial overexpression of a limited number of proteins linked to pluripotency in ESC was recently found to reprogram differentiated somatic cells back to a pluripotent state, enabling the derivation of isogenic (patient-specific) human pluripotent stem cell lines (Park et al, 2008). Current research is focusing on improving reprogramming protocols (e.g. circumventing the use of retroviral technology and oncoproteins) and methods for differentiation into transplantable tissues of interest. In mouse ESC, we have previously shown that the embryonic morphogens BMP4 and Wnt3a direct blood formation via activation of Cdx and Hox genes. Ectopic expression of Cdx4 and HoxB4 enables the generation of mouse ESC-derived hematopoietic stem cells (HSC) capable of multilineage reconstitution of lethally irradiated adult mice. We have asked whether these signaling pathways patterning blood fate are conserved during hematopoietic development from human induced pluripotent stem (iPS) cells generated in our laboratory. Our data showed robust differentiation of iPS cells to mesoderm and to blood lineages, comparable to reports on differentiation of human ESC in this system. We detected robust formation of CD34+ (28.9±12), CD45+ (26.8±13.4) and CD34+CD45+ (16.1±13.7) cells, and a high incidence of CFU-initiating cells in functional colony assays, predominantly displaying myeloid but also some mixed CFU-GEMM activity. Similar to our findings in mouse ESC, mesodermal and hematopoietic genes were expressed in waves, and expression was augmented by supplementation of cultures with BMP4. Mesodermal markers (e.g. BRACHYURY ) were induced at day 2, and declined after day 9, when hematopoietic markers (SCL) appeared, indicating conversion of mesoderm to progenitors of the blood lineage. Expression of all three human CDX genes (CDX1, CDX2 and CDX4) peaked at day 6, suggesting that the function of CDX genes to pattern preformed mesoderm to blood fate may be conserved in human embryogenesis. Ongoing experiments in our laboratory focus on genetic modification of human iPS cells to study effects of specific genes during human emrbyonic hematopoiesis. Furthermore we have succeeded in transducing iPS cells with lentiviruses that allow GFP expression and puromycin selection, thus indicating feasibility for genetic modification. Taken together, our results show robust hematopoietic differentiation of human iPS cells and suggest that genetically modified in vitro differentiating iPS cells can be used to study human developmental hematopoiesis. Characterizing genetic pathways governing human embryonic blood formation will direct differentiation of induced pluripotent stem cells into repopulating hematopoietic stem cells, enabling generation of isogenic cell replacement therapies. Moreover, this experimental approach enables modeling of hematologic diseases, opening up a novel platform for gradual studies of genetic mechanisms during disease pathogenesis. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction The caudal-type homeobox (Cdx) gene family has been mainly studied during early development for its role in axial elongation and antero-posterior patterning. More recently, Cdx genes were shown to regulate embryonic hematopoiesis by interactions with the canonical Wnt pathway and Hox genes. The role of Cdx genes in adult hematopoiesis remains poorly understood. Adult hematopoietic stem and progenitor cells derived from healthy murine bone marrow (BM) express low levels of Cdx1 and Cdx4 but not Cdx2. However, the majority (〉80%) of human acute myeloid (AML) and lymphoid leukemias (ALL) were shown to express the human homologue CDX2, and ectopic induction of Cdx2 expression was sufficient to robustly induce myeloid leukemia in murine bone marrow cells. On the molecular level, the leukemogenic activity of Cdx2 was associated with modulation of Hox and Klf4 gene expression (Faber et al, 2013). The current study further explores the role of CDX2 in leukemogenesis by analyzing the effects of CDX2 expression induction or repression on human healthy and malignant hematopoietic cells and its molecular effects on the Wnt signaling pathway known to regulate Cdx genes during embryonic development. Methods Human bone marrow or mobilized peripheral blood derived CD34+ cells as well as the human leukemic cell lines SKM-1, NOMO-1, EOL-1 and NALM16 were exposed to lentiviruses containing CDX2 overexpression, shRNAs against CDX2 or control constructs. Efficient modulation of CDX2 expression was verified on gene expression level by qRT-PCR and on protein level by immunoblot analysis. CDX2 modified and control cells were subjected to growth, colony forming (CFU), cell cycle, flow cytometry and qRT-PCR gene expression analysis assays and analyzed in vivo upon xenotransplantation in NOD/SCID/IL2Rγnull (NSG) mice. To explore the effect of Dickkopf-1 (DKK1), recombinant human DKK1 protein or carrier was supplemented to the methylcellulose in CFU assays. Results shRNA-mediated knockdown of CDX2 in leukemic cell lines lead to reduced growth (SKM-1, NALM16) and CFU formation (SKM-1 cells). Consistently, CDX2 knockdown SKM-1 cells showed lower ability to repopulate NSG mice and, upon subcutaneous injection in the flank, gave rise to much smaller tumors when compared to control cells, supporting the notion that CDX2 plays roles in human leukemogenesis. In contrast to the data published in mice, healthy human CD34+ cells transduced to overexpress CDX2 were unable to induce leukemia upon transplantation in NSG mice within an observation period of 5 months. On the molecular level, CDX2 modified cells showed differential expression of Klf4 and Hox but also Wnt pathway associated genes. Notably, robust induction of the canonical Wnt-inhibitory molecule DKK1 was observed in both healthy CD34+ stem/progenitor and leukemic cells upon CDX2 induction, while CDX2 suppression showed opposite effects. Analysis of the DKK1 promotor region revealed an interspecies conserved putative binding site for CDX2 as well as multiple HOX gene binding sites, suggesting that CDX2 can modulate DKK1 expression directly but also via its downstream HOX genes. Importantly, CFU assays performed on CDX2-knockdown cells showed a rescue of colony formation upon stimulation with DKK1 protein as compared to treatment with carrier only, demonstrating that the observed molecular interaction is functionally relevant in human leukemic cells. In contrast, control leukemic cells treated with DKK1 showed reduced CFU formation, indicating that CDX2 might act through DKK1 activation to fine-tune Wnt signal activation to the dosage that best promotes leukemogenesis and leukemic cell growth and survival. Conclusion Taken together, our data indicate that CDX2 employs DKK1 activation to modulate the Wnt signaling pathway and thereby growth, clonogenic capacity as well as in vivo tumorigenicity of human leukemia cells. In contrast to murine cells, CDX2 activation requires cooperative molecular events in order to induce leukemia in human healthy stem and progenitor cells. Disclosures: No relevant conflicts of interest to declare.

Description: Background: Over the last years, the zebrafish has emerged as a versatile novel experimental model for hematopoietic studies. The major genetic pathways have proven conserved between fish and mammalian hematopoiesis. Several oncogenes involved in human leukemia have been successfully overexpressed in a transient fashion in zebrafish embryos. However, despite first encouraging results these experimental models often failed to fully recapitulate human myeloid malignancy, perhaps due to early lethality caused by off-target expression or lack of secondary events necessary for full malignant transformation. Material and Methods: Here we took advantage of the Gal4/UAS binary system and of existing transgenic lines to overexpress the human oncogenic HRAS gene in zebrafish hematopoietic cells under control of specific promoters (fli.1, pu.1, runx.1). HRAS-transgenic lines were generated and fish followed by microscopy from early development until, if possible, sexual maturity. Hematopoietic cell development was studied at embryonic (fli.1, pu.1), larval (pu.1, runx1) and adult stages (runx1) by in situ hybridization and real-time PCR analysis of hematopoietic gene expression, flow cytometry, immunohistochemistry and/or blood smear morphological assessment. Results: All HRAS transgenic lines showed hematopoietic abnormalities. However, different phenotypes were observed depending on the promoter driving the oncogene expression. HRAS induction via the early hematopoietic promoter fli.1 affected primitive hematopoiesis inducing a myelo-erythroid proliferation characterized by the expansion of the caudal hematopoietic tissue, enhanced expression of myelo-erythroid genes and delayed erythrocyte maturation (Alghisi et al. 2013). Surprisingly, no obvious effects were noted on the emerging hematopoietic stem cells (HSCs) in the aorta-gonado-mesonephros (AGM) region and studies at later stages were hampered by early lethality of the fish due to vascular defects and cardiac edema. The lethality at early stages was also observed using the myeloid promoter pu.1 that induced the expansion of primitive myeloid cells along with severe developmental defects. In contrast, HRAS expression driven by runx1, a known HSC marker, did not affect primitive hematopoiesis and allowed studies at later developmental stages. Supporting the results obtained with the fli.1 promoter, no alteration was noted in the AGM of runx1-HRAS induced fish. However, at 1 month post-fertilization, flow cytometry analyses revealed a prominent cellular expansion of the hematopoietic stem/progenitor cells in the kidney marrow, the zebrafish definitive hematopoietic compartment. Kidney marrow cytospin preparation and flow cytometry analysis confirmed high numbers of undifferentiated cells indicating that HRAS -overexpression in this model induced stem/progenitor cell proliferation. Interestingly, first analyses suggested that the numbers of differentiated cells might be reduced in these fish, implying that HRAS-induced stem/progenitors have impaired differentiation capacity. Outlook and conclusion: We are currently further investigating the effects of runx1-driven HRAS on the hematopoietic compartment and generate tools to explore potential cooperation of HRAS with other oncogenes during leukemogenesis. Successfully established zebrafish leukemia models shall be eventually used for identification of therapeutically active compounds in small molecule screens. Disclosures No relevant conflicts of interest to declare.

Description: The Evi1 locus was originally identified as a common site of retroviral integration in murine myeloid tumors. Several reports associate Evi1 expression with aggressiveness in myeloid leukemia. Since developmental pathways often reactivate in cancer, we hypothesized that Evi1 also plays critical roles during developmental hematopoiesis. Here, we employ the zebrafish model to study how evi1 modulates early blood development. We find that indeed zebrafish evi1 co-localizes with the hematopoietic markers scl, gata1, pu.1 and gata2 in the posterior lateral mesoderm and the rostral blood islands, indicating involvement in primitive hematopoiesis. Knockdown of evi1 via three independent Morpholino Oligonucleotides impairs embryonic myelopoiesis as shown by reduced pu.1, mpo and l-plastin staining, while not affecting hemangioblast formation and primitive erythropoiesis. Additionally, we observe reduced levels of cd41 expression upon evi1 knockdown, indicating that megakaryopoiesis is also impaired. Notably, at later time-points, evi1 is also expressed in the AGM region and evi1 morphants show strong reduction of runx1/c-myb expression in this region, demonstrating an additional role in hematopoietic stem cell (HSC) formation. Consistently, evi1 morphants lack ikaros+ lymphocyte precursor cells and rag1+ T-lymphocytes, and less circulating globin+, lyz+ and cd41+ cells are detected in transgenic fish analyzed by flow cytometry at 5 days post fertilization. To dissect the mechanisms by which evi1 regulates HSC formation, we furthermore analyzed genes specifically expressed in the dorsal aorta, where HSCs emerge from the hemogenic endothelium. We hypothesize that evi1 is indeed regulating HSC specification from hemogenic endothelial cells, since detailed analyses show defective aortic expression of efnb2a and dlc. How this effect is mediated is currently being investigated. Furthermore, TUNEL, anti-activated Caspase-3, anti-phospho Histone H3 and BrdU assays show both increased apoptosis and decreased proliferation in the AGM region of evi1 morphants as compared to control injected fish, suggesting important additional effects of evi1 on HSC biology beyond the specification step. Similar effects are seen in the rostral blood islands and are possibly responsible for the observed reduction of primitive myeloid progenitors. We also examined potential downstream targets of evi1. Previous reports in adult murine hematopoietic cells suggest that Evi1 affects HSC proliferation through regulation of Gata2. Indeed, gata2 co-injection enhances hematopoietic cell viability also in zebrafish embryos, and co-injection of gata2 mRNA is able to fully rescue both primitive myelopoiesis and HSC formation. We are currently investigating whether gata2 expression can compensate evi1 requirement during HSC specification or, alternatively, might rescue HSC numbers by amplifying residual HSCs that escape evi1 inhibition. Taken together, our data demonstrate that evi1 plays multiple roles during hematopoietic development, critically regulating the biology of primitive myeloid progenitors and pre-formed HSCs, as well as HSC specification from the hemogenic endothelium. Currently, we analyze the molecular mechanisms that mediate evi1 effects during these different regulatory steps and use the CRISPR/Cas9 system to generate evi1 mutant fish for further validation of our findings. Disclosures No relevant conflicts of interest to declare.

Description: Patients with acute myeloid leukaemia (AML) often achieve remission but subsequently die of relapse driven by chemotherapy resistant leukemic stem cells (LSCs). To initiate and maintain cancer, LSCs must also escape immunosurveillance. However, in vivo studies on human LSCs largely disregard lymphocyte mediated anti-tumor immunity due to the use of immunocompromised mice. Here we investigate the immunosurveillance mediated by NKG2D, a danger detector expressed by cytotoxic lymphocytes such as natural killer (NK) cells that recognizes stress-induced ligands (NKG2DL) of the MIC and ULBP protein families on AML cells. Staining of n=175 de novo AML with antibodies against MICA, MICB and ULB2/5/6 or an NKG2D-Fc chimeric protein recognizing pan-NKG2DL expression revealed NKG2DL to heterogeneously express among leukemic cells of the same patient (Fig. 1a). As expected, NKG2DLpos AML cells were efficiently cleared by natural killer (NK) cells, while NKG2DLneg leukemic cells escaped NK cell lysis. Interestingly, these NKG2DLneg AML cells also showed immature morphology, enhanced in vitro clonogenicity (39±47 colonies vs. 1±4, p