Experimental approaches to studying the nature and impact of splicing variation in zebrafish.

From a fixed number of genes carried in all cells, organisms create considerable diversity in cellular phenotype through differential regulation of gene expression. One prevalent source of transcriptome diversity is alternative pre-mRNA splicing, which is manifested in many different forms. Zebrafish models of splicing dysfunction due to mutated spliceosome components provide opportunity to link biochemical analyses of spliceosome structure and function with whole organism phenotypic outcomes. Drawing from experience with two zebrafish mutants: cephalophonus (a prpf8 mutant, isolated for defects in granulopoiesis) and caliban (a rnpc3 mutant, isolated for defects in digestive organ development), we describe the use of glycerol gradient sedimentation and native gel electrophoresis to resolve components of aberrant splicing complexes. We also describe how RNAseq can be employed to examine relatively rare alternative splicing events including intron retention. Such experimental approaches in zebrafish can promote understanding of how splicing variation and dysfunction contribute to phenotypic diversity and disease pathogenesis.

Hydrogen peroxide in inflammation: messenger, guide, and assassin.

Starting as a model for developmental genetics, embryology, and organogenesis, the zebrafish has become increasingly popular as a model organism for numerous areas of biology and biomedicine over the last decades. Within haematology, this includes studies on blood cell development and function and the intricate regulatory mechanisms within vertebrate immunity. Here, we review recent studies on the immediate mechanisms mounting an inflammatory response by in vivo analyses using the zebrafish. These recently revealed novel roles of the reactive oxygen species hydrogen peroxide that have changed our view on the initiation of a granulocytic inflammatory response.

Morphologic and functional characterization of granulocytes and macrophages in embryonic and adult zebrafish.

The zebrafish is a useful model organism for developmental and genetic studies. The morphology and function of zebrafish myeloid cells were characterized. Adult zebrafish contain 2 distinct granulocytes, a heterophil and a rarer eosinophil, both of which circulate and are generated in the kidney, the adult hematopoietic organ. Heterophils show strong histochemical myeloperoxidasic activity, although weaker peroxidase activity was observed under some conditions in eosinophils and erythrocytes. Embryonic zebrafish have circulating immature heterophils by 48 hours after fertilization (hpf). A zebrafish myeloperoxidase homologue (myeloid-specific peroxidase; mpx) was isolated. Phylogenetic analysis suggested it represented a gene ancestral to the mammalian myeloperoxidase gene family. It was expressed in adult granulocytes and in embryos from 18 hpf, first diffusely in the axial intermediate cell mass and then discretely in a dispersed cell population. Comparison of hemoglobinized cell distribution, mpx gene expression, and myeloperoxidase histochemistry in wild-type and mutant embryos confirmed that the latter reliably identified a population of myeloid cells. Studies in embryos after tail transection demonstrated that mpx- and peroxidase-expressing cells were mobile and localized to a site of inflammation, indicating functional capability of these embryonic granulocytes. Embryonic macrophages removed carbon particles from the circulation by phagocytosis. Collectively, these observations have demonstrated the early onset of zebrafish granulopoiesis, have proved that granulocytes circulate by 48 hpf, and have demonstrated the functional activity of embryonic granulocytes and macrophages. These observations will facilitate the application of this genetically tractable organism to the study of myelopoiesis.

Zebrafish--an emerging genetic model for the study of cytokines and hematopoiesis in the era of functional genomics.

Now that whole genomes are sequenced, the identification of gene function rather than gene discovery is a major challenge. Saturation mutagenesis and screening for mutant phenotypes are methods that allow sampling of the genome for lesions in genes critical for particular physiological processes. This approach promises to provide new insights into gene function, even for molecularly well-characterized processes such as hematopoiesis and cytokine signaling. Animal models for such genetic approaches have traditionally included Drosophila and the mouse. Recently, the zebrafish (Danio rerio) has emerged as a flexible and informative vertebrate for genetic studies. Zebrafish hematopoiesis has a morphological and molecular complexity closer to that of mammals than does Drosophila, providing scope for recognizing mutant zebrafish phenotypes representing finely tuned lesions in these processes. Compared to mice, zebrafish represent an economical, flexible, and genetically tractable animal model for mutagenesis studies. The structure of the teleost genome creates several phylogenetic issues in assessing zebrafish and piscine orthologues and paralogues of known mammalian genes, here exemplified by a cytokine ligand (interleukin-1beta), kinase receptors (c-kit and c-fins), and a family of intracellular signaling molecules (JAK kinases). Several anemic zebrafish mutants are now genetically characterized, and others present hematopoietic phenotypes that promise novel insights into the regulation of hematopoiesis.

Comparison of effects of the tyrosine kinase inhibitors AG957, AG490, and STI571 on BCR-ABL--expressing cells, demonstrating synergy between AG490 and STI571.

STI571 (formerly CGP57148) and AG957 are small molecule inhibitors of the protein tyrosine kinase (PTK) p145(abl) and its oncogenic derivative p210(bcr-abl). AG490 is an inhibitor of the PTK Janus kinase 2 (JAK2). No direct comparison of these inhibitors has previously been reported, so this study compared their effects on factor-dependent FDC-P1, 32D, and MO7e cells and their p210(bcr-abl)-expressing factor-independent derivatives. STI571 was a more potent inhibitor of (3)H-thymidine incorporation in p210(bcr-abl)-expressing cells than was AG957, and it showed superior discrimination between inhibitory effects on parental cell lines and effects on their p210(bcr-abl)-expressing derivatives. Assays performed with and without growth factor demonstrated that STI571 but not AG957 reversed the p210(bcr-abl)-driven factor independence of cell lines. p210(bcr-abl)-expressing cells were less sensitive to AG490 than to AG957 or STI571. However, for p210(bcr-abl)-expressing clones from all 3 cell lines, synergistic inhibition was demonstrated between STI571 and concentrations of AG490 with no independent inhibitory effect. Inhibition of nucleic acid synthesis with AG957 treatment was associated with reduced cell numbers, reduced viability, and small pyknotic apoptotic cells. At concentrations of STI571 that reversed the p210(bcr-abl) factor-independent phenotype, STI571 treatment and growth factor deprivation together were sufficient to induce apoptosis. This study concludes that, for the cell lines studied, (1) STI571 is a more potent and more selective inhibitor of a p210(bcr-abl)-dependent phenotype than AG957; (2) AG490 synergizes with STI571 to enhance its inhibitory effect on p210(bcr-abl)-driven proliferation; and (3) the combination of p210(bcr-abl)-tyrosine kinase inhibition and growth factor signal withdrawal can be sufficient to induce apoptotic death of transformed cells. (Blood. 2001;97:2008-2015)

Hereditary spherocytosis in zebrafish riesling illustrates evolution of erythroid beta-spectrin structure, and function in red cell morphogenesis and membrane stability.

Spectrins are key cytoskeleton proteins with roles in membrane integrity, cell morphology, organelle transport and cell polarity of varied cell types during development. Defects in erythroid spectrins in humans result in congenital hemolytic anemias with altered red cell morphology. Although well characterized in mammals and invertebrates, analysis of the structure and function of non-mammalian vertebrate spectrins has been lacking. The zebrafish riesling (ris) suffers from profound anemia, where the developing red cells fail to assume terminally differentiated erythroid morphology. Using comparative genomics, erythroid beta-spectrin (sptb) was identified as the gene mutated in ris. Zebrafish Sptb shares 62.3% overall identity with the human ortholog and phylogenetic comparisons suggest intragenic duplication and divergence during evolution. Unlike the human and murine orthologs, the pleckstrin homology domain of zebrafish Sptb is not removed in red cells by alternative splicing. In addition, apoptosis and abnormal microtubule marginal band aggregation contribute to hemolysis of mutant erythrocytes, which are features not present in mammalian red cells with sptb defects. This study presents the first genetic characterization of a non-mammalian vertebrate sptb and demonstrates novel features of red cell hemolysis in non-mammalian red cells. Further, we propose that the distinct mammalian erythroid morphology may have evolved from specific modifications of Sptb structure and function.

T lymphocytes from granulocyte colony-stimulating factor-/- mice produce large quantities of interferon-gamma in a chronic infection model.

Little is known about the role of granulocyte colony-stimulating factor (G-CSF) in the response to chronic bacterial infections. To address this we infected G-CSF knock out (G-CSF-/-) mice with Mycobacterium avium. Infection was not exacerbated in G-CSF-/- mice despite a deficiency in the total bone marrow cells, colony-forming haemopoietic cells, granulocytes and monocyte precursors in the bone marrow. Peritoneal cells from G-CSF-/- produced less nitric oxide (NO) upon culture in vitro with antigen than did wild-type (WT) cells. Unexpectedly, T cells from infected G-CSF-/- mice were able to produce significantly more interferon-gamma (IFN-gamma) than the wild type (WT) controls. T cells from G-CSF-/- mice still produced more IFN-gamma even when in vitro NO production was inhibited, while enzyme-linked immunospot assay (ELISPOT) assays showed more IFN-gamma-producing cells in the G-CSF-/- mice. This was confirmed by intracellular cytokine staining (ICCS), which showed that there were more IFN-gamma producing T cells in vivo in the G-CSF-/- than the WT controls following M. avium infection. It is possible that a deficit of NO in vivo allows T cells to develop a higher IFN-gamma-producing phenotype. Thus we show a novel relationship between G-CSF and IFN-gamma production by T cells revealed in this chronic bacterial infection model.

Functional deficiencies of peritoneal cells from gene-targeted mice lacking G-CSF or GM-CSF.

Gene-targeted mice lacking the hemopoietic growth factors, granulocyte colony-stimulating factor (G-CSF) or granulocyte-macrophage (GM)-CSF, show increased susceptibility to infection with the facultative intracellular bacterium, Listeria monocytogenes. The resident peritoneal cell populations from G-CSF(-/-) and GM-CSF(-/-) mice showed reduced production of the bactericidal molecule nitric oxide. Macrophage-mediated tumoricidal activity and phagocytosis of Listeria were reduced in G-CSF(-/-), but not in GM-CSF(-/-), mice. In G-CSF(-/-) mice, there was an unexpected expansion (from 18% in WT to 38%) of a population of cells with morphology intermediate between typical macrophages and typical lymphocytes. These cells had some of the features of poorly differentiated macrophages, being adherent to plastic but poorly phagocytic, nonspecific esterase positive but myeloperoxidase negative. They were largely negative for the macrophage marker F4/80 and for Thy1, B220, and Gr1. Their disproportionate presence, and the corresponding deficiency in typical macrophages, possibly accounts for some of the functional deficiencies observed in G-CSF(-/-) mice.

Essential roles for granulocyte-macrophage colony-stimulating factor (GM-CSF) and G-CSF in the sustained hematopoietic response of Listeria monocytogenes-infected mice.

The in vivo roles of granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte (G)-CSF were studied in factor-deficient gene-targeted knockout mice infected with the facultative intracellular bacterium Listeria monocytogenes. Previous results showed that G-CSF-/- mice had an underlying selective deficiency in granulopoiesis, but GM-CSF-/- mice had little disturbance in resting hematopoiesis. Nevertheless, in this study it is revealed that 3 days after intraperitoneal infection with 2 x 10(5) Listeria, GM-CSF-/- mice harbored 50-fold more organisms in their spleen and liver than similarly infected wild-type mice. This was accompanied by a severe depletion of bone marrow hematopoietic cells and a deficient inflammatory response in their peritoneal cavity. Thus, GM-CSF is essential for emergency, but not resting, hematopoiesis. In contrast, G-CSF-/- mice were markedly susceptible to low doses (2 x 10(4)) of Listeria intraperitoneally. After infection, the acute (1 day) granulocyte infiltration to the peritoneal cavity was normal compared with wild type, but the more prolonged monocyte response was deficient, reflecting a continued decrease in bone marrow cellularity and hematopoiesis over 3 days, which was not observed in infected wild-type mice. It is thus apparent that G-CSF deficiency affects monocytopoiesis as well as granulopoiesis during infection.

Mice lacking both granulocyte colony-stimulating factor (CSF) and granulocyte-macrophage CSF have impaired reproductive capacity, perturbed neonatal granulopoiesis, lung disease, amyloidosis, and reduced long-term survival.

Mice lacking granulocyte colony-stimulating factor (G-CSF) are neutropenic with reduced hematopoietic progenitors in the bone marrow and spleen, whereas those lacking granulocyte-macrophage colony-stimulating factor (GM-CSF) have impaired pulmonary homeostasis and increased splenic hematopoietic progenitors, but unimpaired steady-state hematopoiesis. These contrasting phenotypes establish unique roles for these factors in vivo, but do not exclude the existence of additional redundant functions. To investigate this issue, we generated animals lacking both G-CSF and GM-CSF. In the process of characterizing the phenotype of these animals, we further analyzed G-CSF- and GM-CSF-deficient mice, expanding the recognized spectrum of defects in both. G-CSF-deficient animals have a marked predisposition to spontaneous infections, a reduced long-term survival, and a high incidence of reactive type AA amyloidosis. GM-CSF-deficient mice have a modest impairment of reproductive capacity, a propensity to develop lung and soft-tissue infections, and a similarly reduced survival as in G-CSF-deficient animals. The phenotype of mice lacking both G-CSF and GM-CSF was additive to the features of the constituent genotypes, with three novel additional features: a greater degree of neutropenia among newborn mice than in those lacking G-CSF alone, an increased neonatal mortality rate, and a dominant influence of the lack of G-CSF on splenic hematopoiesis resulting in significantly reduced numbers of splenic progenitors. In contrast to newborn animals, adult mice lacking both G-CSF and GM-CSF exhibited similar neutrophil levels as G-CSF-deficient animals. These findings demonstrate that the additional lack of GM-CSF in G-CSF-deficient animals further impairs steady-state granulopoiesis in vivo selectively during the early postnatal period, expand the recognized roles of both G-CSF and GM-CSF in vivo, and emphasize the utility of studying multiply deficient mouse strains in the investigation of functional redundancy.

Increased tolerance to endotoxin by granulocyte-macrophage colony-stimulating factor-deficient mice.

The contribution of granulocyte-macrophage CSF (GM-CSF) to endotoxin-mediated septic shock has been assessed by treating GM-CSF-deficient mice with LPS. Hypothermia and loss in body weight were markedly attenuated in LPS-treated GM-CSF-deficient mice compared with similarly treated control mice; moreover, the levels of circulating IFN-gamma, IL-1alpha, and IL-6 were lower in LPS-treated GM-CSF-deficient mice than LPS-treated control mice. Intriguingly, the peak levels of TNF-alpha in response to LPS treatment were the same in the serum of GM-CSF-deficient mice and control mice, although in GM-CSF-deficient mice, TNF-alpha persisted longer. Activation of macrophages by LPS, resulting in expression of cytokines including TNF-alpha and IL-1, is thought to underlie endotoxin-mediated effects. Accordingly, the response of peritoneal macrophages from GM-CSF-deficient mice to LPS was studied in vitro. LPS-stimulated peritoneal macrophages from GM-CSF-deficient mice produced significantly less IL-1alpha and nitric oxide than macrophages from wild-type mice, although there was no difference in TNF-alpha production. Collectively, these observations indicate that GM-CSF contributes to cytokine production in LPS-mediated septic shock, and that the attenuated production of these secondary cytokines (IFN-gamma, IL-1alpha, and IL-6) may contribute to the endotoxin-resistant phenotype of GM-CSF-deficient mice.

Bioactive murine and human interleukin-12 fusion proteins which retain antitumor activity in vivo.

Interleukin-12 (IL-12) is unique amongst cytokines in being a disulfide-linked heterodimer of two separately encoded subunits (p35 and p40). We expressed single chain IL-12 proteins from retroviral constructs in which the two IL-12 subunits were linked by a 6-15 amino acid polypeptide linker, with deletion of the 22 amino acid leader sequence of the trailing subunit. The murine fusion protein IL-12.p40.L.delta p35 containing a (Gly4Ser)3 linker was stably expressed, bioactive in vitro, and had an apparent specific activity comparable to that of native and recombinant IL-12. Western blotting confirmed that murine IL-12.p40.L.delta p35 retained the linking polypeptide sequences. The analogous human IL-12.p40.L.delta p35 fusion protein containing a Gly6Ser linker was bioactive with an apparent specific activity comparable to recombinant human IL-12. In a preexisting CMS-5 tumor model, CMS-5 cells secreting either native or fusion protein forms of IL-12 prolonged survival and led to complete tumor regression.

CSF-deficient mice--what have they taught us?

Haemopoietic growth factor-deficient mice have been particularly instructive for defining the usual physiological role of these factors. Mice now exist lacking the granulopoietic factors G-CSF, GM-CSF, M-CSF (CSF-1), SCF, several other factors influencing haemopoiesis (including erythropoietin, interleukins 5 and 6), combinations of these factors (GM- & M-CSF; G- & GM-CSF; G- & GM- & M-CSF) and several CSF receptor components. Most of these mice were generated by targeted gene disruption, others are spontaneously arising mutants. The phenotypes of these mice indicate that the granulopoietic factors have both unique and redundant roles in vivo. Some factors are uniquely important in baseline myelopoiesis. Experimental infection of CSF-deficient mice indicates unique roles for some factors in emergency 'overdrive' haemopoiesis. Recovery from myeloablation evaluates the role of CSFs in emergency 'restoring normality' haemopoiesis. Redundancy also exists in the capacity of CSFs to support complete granulocyte development in vivo. Some factors are not involved in all the in vivo roles suggested by the range of their actions demonstrable in vitro. Some CSFs have indispensable roles in non-haemopoietic tissues. Some factors have in vivo roles not anticipated from previous studies. Mice deficient in several factors have identified compensating roles for factors by revealing exacerbated and additional phenotypic features, and may unmask additional in vivo roles.

The influence of granulocyte/macrophage colony-stimulating factor on dendritic cell levels in mouse lymphoid organs.

To ascertain whether the development of dendritic cells (DC) in mouse lymphoid organs is dependent on granulocyte/macrophage colony-stimulating factor (GM-CSF), we determined the number of DC in the thymus, spleen and lymph nodes of normal mice, of mice with the genes coding for GM-CSF or its receptor inactivated, and of transgenic mice with excessive levels of GM-CSE DC were extracted from the tissues and enriched prior to flow cytometric analysis. The total DC level and the incidence of DC expressing lymphoid-related markers (CD8(hi) CD11b(lo)) and myeloid-related markers (CD8(lo) CD11b(hi)) were monitored. Both in GM-CSF null mice, and GM-CSF receptor null mice, DC of all surface phenotypes were present in all lymphoid organs; only small decreases in DC levels were recorded, except for the lymph nodes of GM-CSF receptor null mice which showed a more pronounced (threefold) decrease in DC numbers. Since the GM-CSF receptor null mice lacked the beta chain common to the GM-CSF, interleukin (IL)-3 and IL-5 receptors, the development of DC in the absence of GM-CSF was not due to common beta chain mediated developmental signals elicited by IL-3 or IL-5. In GM-CSF transgenic mice, there was only a 50 % increase in DC numbers in thymus and spleen, paralleling an increase in overall cellularity, but a more pronounced (threefold) increase in DC numbers in lymph nodes. There was no evidence that GM-CSF had a selective effect on any particular DC subpopulation defined by CD8 or CD11b expression. We conclude that the development of most lymphoid tissue DC can proceed in the absence of GM-CSF, although this cytokine can produce some elevation of DC levels. It is not clear whether the enhancing effect of GM-CSF is direct or an indirect effect mediated by other cytokines.

Granulocyte-macrophage colony-stimulating factor is not responsible for the correction of hematopoietic deficiencies in the maturing op/op mouse.

Osteopetrotic (op/op) mice are characterized by an autosomal recessive inactivating mutation resulting in the absence of biologically active colony-stimulating factor-1 (CSF-1). Consequently, young op/op mice have a severe deficiency of macrophages and osteoclasts resulting in excessive bone formation, occlusion of the marrow cavity, and reduced marrow hematopoietic activity. Recently, we showed that the osteopetrosis and hematopoietic deficiencies evident in young op/op mice are not permanent but are progressively corrected with age. There are increases in osteoclast activity; bone resorption; femoral marrow space; and marrow hematopoietic activity, cellularity, and macrophage content. In the present study we show that CSF-1-/- granulocyte-macrophage colony-stimulating factor (GM-CSF)(-/-)-deficient mice also undergo the same pattern of hematopoietic correction as the op/op mouse. Also, like the op/op mouse, the peritoneal cellularity and macrophage content of CSF-1/GM-CSF-deficient mice remains severely reduced. Our data show that the "knockout" of GM-CSF does not change the op/op phenotype, and that GM-CSF is not essential for the correction of the hematopoietic deficiencies in the op/op mouse. Importantly, the data also show that neither GM-CSF nor CSF-1 is an absolute requirement for the commitment of primitive hematopoietic stem cells to the macrophage lineage or for the differentiation of at least some classes of macrophages. This finding suggests that an alternate regulatory factor can be involved in macrophage and osteoclast commitment, differentiation, and function in vivo.

Development of functional macrophages from embryonal stem cells in vitro.

When cultured under appropriate in vitro conditions, embryonal stem cells (ESCs) form embryoid bodies (EBs) that contain mature hematopoietic cells, including cells of the monocyte-macrophage lineage. A two-step in vitro culture system for generation of ESC-derived macrophages has been developed and optimized. Maximum numbers of macrophage-containing colonies developed in secondary hematopoietic cultures of cells from disrupted EBs after 9 to 12 days of differentiation when interleukin-3 (IL-3) and macrophage colony-stimulating factor (M-CSF) were included in both primary and secondary cultures. Over 10(5) viable, phagocytically active macrophages were generated from cultures initiated by 7500 ESCs. The inclusion of stem cell factor (SCF) in primary cultures not only increased the frequency of progenitor cells but also the cellular heterogeneity of colonies. SCF in secondary cultures increased the cellularity, but not the frequency, of macrophage-containing colonies; although cellular heterogeneity was also increased, there was still an overall increase in yield of macrophages.