INTRODUCTION
Cell cycle control has significant roles in multiple physiological processes, such as stem cell pluripotency, cell proliferation and differentiation, and embryonic morphogenesis (
Vousden and Lu, 2002;
Ornitz, 2005). Mitotic spindle assembly and chromosome segregation are the pivotal events for regulating the cell cycle, especially for mitosis phase progression. A number of scaffold proteins have been identified that associate with the mitotic spindle and play important roles in spindle assembly and cell cycle control. Examples of these proteins include γ-tubulin, NuMA, TPX2, dynactin and nucleolar spindle-associated protein (NuSAP) (
Merdes et al., 1996;
Hetzer et al., 2000). NuSAP is a well-known microtubule-associated protein. The homologues of human NuSAP have been reported in orangutan, dog, mouse and frog. The N-terminal box and C-terminal KEN box share the most protein sequence identity between amphibians and primates. Carmeliet’s group found that NuSAP has altered expression during cell cycle progression at both mRNA and protein level (
Raemaekers et al., 2003). Recent reports showed that NuSAP can be degraded by anaphase promoting complex/cyclosome formation (APC/C) via the ubiquitin/proteasome system (
Li et al., 2007). The KEN box is recognized by APC/C-cdh1 or APC/C-cdc20, which controls the protein level of NuSAP during cell cycle progression. However, the biological significance of the degradation or knockdown of NuSAP still remains unclear.
The GTPase Ran is essential for regulating nucleocytoplasmic transport in interphase cells and for organizing the spindle apparatus during mitosis (
Yudin and Fainzilber, 2009). NuSAP is targeted by Ran GTPase to stabilize microtubule (
Ribbeck et al., 2006). During mitosis, dissociated NuSAP is immobilized on chromatin and produces a high concentration of microtubules around chromatin (
Ribbeck et al., 2007). Moreover, overexpression or knock-down of NuSAP results in mitotic spindle deficiency (
Raemaekers et al., 2003). In spite of that these
in vitro studies which support the involvement of NuSAP in linking microtubule with mitotic chromosomes during cell cycle regulation, the
in vivo roles of NuSAP on cell fate and cell behaviors are largely unknown.
Here, we identified a novel homologue of NuSAP in zebrafish, which shares 36% amino acid sequence identity with human NuSAP. The transcripts were restrictively expressed in the brain, neural crest and blood island as shown by in situ hybridization. We used antisense oligonucleotide morpholino to knockdown nusap1 mRNA in zebrafish and found that the nusap1 morphants display impaired morphogenesis and defective cell migration in the trunk and yolk extension. These phenotypes can be rescued by co-injection of nusap1 mRNA. Interestingly, injection of nusap1 mRNA alone causes apoptosis in the eyes and hindbrain. These original phenotypes confirm the previous in vitro observations about NuSAP in regulating microtubule stabilization and mitosis, and further suggest the in vivo significance of Nusap1 in cell migration and cell cycle progression.
RESULTS
Cloning and identification of zebrafish Nusap1 gene
By screening FGF-responsive genes in a zebrafish cDNA microarray, we classified 32 candidate ORF fragments into strong-, moderate- and weak-response groups to Fgf8 (
Huang et al., 2007). Although
in situ hybridization data suggested that Fgf8 increased
nusap1 transcription only slightly (Data not shown), we still focused on this gene for the interest in its functional relationship with human NuSAP. Sequencing of full length
nusap1 showed that its product was the same as that from a locus in zebrafish contig 13 (XP_695834.1) with the open reading frame (ORF) between position 121 and 1461. To determine the homology of NuSAP, protein sequences from different species were aligned using ClustalW software, and the evolutionary tree indicated that the primary sequence of Nusap1 is highly conserved among vertebrates. The identity of NuSAP between zebrafish and human is 36% (Fig. 1A), and zebrafish Nusap1 has highest homology to its frog orthologue (Fig. 1B).
Although phylogenetic analysis reveals that zebrafish Nusap1 branching is basal to other species and the mammalian Nusap1s are clustered into a separate group, several conserved markers surrounding zebrafish and human orthologues are arranged in the same orientation, including Ndufaf1 (NADH dehydrogenase-ubiquinone 1 alpha subcomplex, assembly factor 1) and Rtf1 (Paf1/RNA polymerase II complex component). A detailed scheme of gene positions surrounding Nusap1 is shown in Fig. 2, indicating the authentic orthologous relationship of zebrafish Nusap1 on chromosome 13 with human NuSAP on chromo-some 15.
Expression profile of nusap1
To confirm that Nusap1 is expressed in zebrafish, its expression pattern was probed by whole mount
in situ hybridization. As shown in Fig. 3,
nusap1 was present in single-cell embryos, suggesting that it is a maternal gene. The zygotic expression was ubiquitous detected from the 1024 cell stage to mid-gastrulation stage (Fig. 3B–D and 3D’), and the pattern in a salt and pepper manner implied its functions in cell cycle progression and proliferation by several experiences from large scale
in situ screening work in other institutes (
Herzog et al., 2003). After the somite formation,
nusap1 transcripts were detected in forebrain, hindbrain, neural crest, blood island and proliferative cells (Fig. 3E–F and 3F’). The same stages of embryo development were analyzed by a sense-control probe in parallel, and analysis of a 30% epiboly embryo is shown in panel H.
nusap1 expression gradually weakened after 24 hpf, and cannot be detected by the same method after 3 dpf (Fig. 3G). Consistently, RT-PCR also detected the expression of
nusap1 transcripts post fertilization in a gradually weakened manner (Fig. 3I).
Knockdown of nusap1 causes cell migration defect in neural crest
Although
in vitro evidence supports the roles of NuSAP in mitosis and microtubule stabilization (
Ribbeck et al., 2006), the
in vivo significance of Nusap1 function is still largely unknown. The unique temporal and spatial expression pattern of
nusap1 in zebrafish suggests that Nusap1 may have important roles
in vivo. The cellular and genetic function of Nusap1 was further tested by morpholino knockdown. To specifically knockdown
nusap1 at the mRNA level, two ATG morpholinos were designed against its 5’-UTR and start codon regions respectively. As shown in Fig. 4A–C, micro-injection of plasmid with N-terminal Nusap1 and GFP fusion resulted in highly mosaic GFP expression, while co-injection of Nusap1-GFP plasmid with
nusap1 MO1 efficiently inhibited the transient GFP expression. Consistently, semi-quantitative Western blot showed reduced GFP expression in
nusap1 morphants (Fig. 4D). These results confirmed that
nusap1 MO1 specifically inhibited Nusap1 gene translation in a dose-dependent manner.
In comparison to the ctrMO injected embryos, the trunk of nusap1 morphants were rougher and shorter, with a smaller and curving tail (Fig. 5A and 5B). TUNEL assay did not detect apparent apoptosis, and angiogenesis and myogenesis appeared normal in nusap1 morphants (Supplemental Fig. 1). Consistent with the above phenotypes, in situ hybridization revealed that nusap1 morphants have impaired expression patterns of crestin and sox9b, which are two markers for neural crest (Fig. 5C–F, 5C’ and 5D’). Staining of neural crest cells and precursor of pigment cells with these two markers indicated a reduced distance of cell migration. However, the nusap1 morphants and control embryos showed a similar mRNA expression level of these markers in whole embryos as measured by quantitative real-time PCR (Fig. 5G and 5H). These results implicate that Nusap1 is an important regulator in neural crest development, and the differences in the crestin and sox9b expression patterns are possibly due to abnormal cell behaviors, such as cell migration, rather than the altered cell fate.
Overexpression of nusap1 results in embryonic apoptosis
To further test the specificity and efficiency of MO, mRNA encoding the nusap1 full length open reading frame was co-injected with MO (Fig. 6A–C). This overexpression can abrogate nusap1 morphants, and interestingly, mRNA of nusap1 alone caused drastic apoptosis (Fig. 6A). A whole mount TUNEL assay showed that apoptosis was concentrated in the retina and hindbrain (Fig. 6E and 6F). Statistics data revealed that the small tail caused by nusap1 MO injection was also rescued by its overexpression (Fig. 6D). These observations imply that nusap1 MO1 is able to rescue the apoptosis phenotype caused by nusap1 over-expression.
DISCUSSION
Nusap1 is a novel microtubule-associated protein identified several years ago, and both its expression and distribution are regulated by cell cycle progression. The human NuSAP expression profile shows that it is mainly expressed in immune organs, including the thymus, bone marrow and fetal liver. Sequence analysis indicates that Nusap1 is conserved among vertebrates. So far, all characterized Nusap1s have been identified in vertebrates, from human to puffer fish. Only a similar member was found in Ciona intestinalis (XP_002127601). Thus, whether invertebrates contain Nusap1 or not is still unknown. Our on-going progress includes characterizing the expression of Nusap1 in sea squirts and amphioxuses, as well as, investigating its significance in CNS development and the evolution of Nusap1.
In this report, we used zebrafish as an animal model to investigate the expression and biological functions of Nusap1. As revealed by in situ hybridization analysis, zebrafish nusap1 is a maternal transcript, and its zygotic expression can be detected in the salt and pepper pattern and decreases after two days, which is similar to the expression of human NuSAP. The temporal and spatial expression profile of nusap1 suggests its potential roles in neural crest development and apoptosis regulation.
When the
nusap1 expression was up- or down-regulated by microinjection of mRNA or morpholino, respectively, development of zebrafish embryos was disrupted. This is consistent with the functions of human NuSAP—deletion or overexpression of which destroys the normal cell cycle progression (
Raemaekers et al., 2003). Microtubule-associated proteins control the assembly of the mitotic spindle and chromosome segregation, and thus, their expressions are strictly and precisely regulated. The TUNEL assay we showed here had moderate signal treated by sub-lethal dose of UV activity. Recent paper investigated the relationship between cell cycle arrest and cell death by different environmental stress and different p53 phosphorylation mechanism (
Li et al., 2009). It is possible that Nusap1 is involved in this mechanism.
The
nusap1 overexpression induced apoptosis, which often involves p53 activation. However co-injection of morpholino against zebrafish
p53 with
nusap1 mRNA failed to rescue the above phenotype (data not shown), suggesting that the apoptosis phenotype may be not caused by p53 activation (
Robu et al., 2007). Nevertheless, due to the pivotal roles of p53 and its complicated post-translation modifications, we could not exclude that the functions of Nusap1 are independent of p53 signaling. Studies from clinical cases indicate that methionine stress induces increased toxicity in NuSAP and p53 double minus condition (
Kokkinakis et al., 2005), implying the potential relationship between NuSAP with p53 and suggesting a further direction of Nusap1 as a tumor suppressor.
MATERIALS AND METHODS
Zebrafish maintenance and plasmid construction
The AB zebrafish strain was raised and kept under standard conditions at ~28.5°C. Embryos were staged according to Zebrafish Book. The coding sequence of zebrafish nusap1 was amplified with a pair of specific primers as zNSF (5’-CCGCTCGAGATAACATGGATTTAGACTC-3’) and zNSR (5’-CGCGGATCCGGTCAGAATAGTTTTTCTTC-3’) from a 24 hpf zebra-fish cDNA, and cloned into pBLUE-scriptKS+ and pXT7 expression vector that contains the UTR from Xenopus β-globin. All PCR products were verified by sequencing.
Synthesis of nusap1 mRNA
Capped mRNA was synthesized in vitro from linearized plasmid DNA using T7 or SP6 RNA polymerase (Cap-Scribe, Roche 11581066001). After removal of DNA by treatment with RNase-free DNase, the mRNA was purified with the RNeasy Mini Kit (QIAGEN) and dissolved in RNsae-free water.
Morpholinos design and microinjection
Two morpholinos were designed to knockdown endogenous
nusap1 with a standard control: 5’-TTGAAGAGGCACAAGTA-TAATA-3’ (
nusap1-MO1, −48 to −25 from ATG), 5’-TATTGCCTATAAAATCGACGTAAG-3’ (
nusap1-MO2, −26 to −3 from ATG) and 5’-CCTCTTACCTCAGTTACAATTTATA-3’ (ctrMO). To verify the knockdown efficiency of the morpholinos, the a Nusap1-GFP construct was generated by subcloning a 1016-bp fragment of
nusap1 that contains a 95-bp 5’-UTR and its adjacent coding sequence for the first 307 amino acids in-frame into pEGFP-N1 vector, and Reverse Transcription PCR for testing endogenous
nusap1 mRNA level in Fig. 3 was performed using same pair of primers as RT-F (5’-ACAAACCCAAACTAACCGATC-3’) and RT-R (5’-ACGAGCTGGAGTCTTCACCAG-3’). The total embryo lysates were probed by Western blot using an anti-GFP and anti-β-actin antibody. The capped mRNA was diluted to 300 ng/μL concentration before microinjection. The mRNA or morpholino was microinjected into the yolk or cytoplasm between one- and two-cell stages as previously described (
Meng et al., 1999).
Whole-mount in situ hybridization and real-time PCR
Digoxigenin-UTP-labeled antisense RNA probe was generated by
in vitro transcription and used for
in situ hybridization. Whole-mount
in situ hybridization followed standard procedure with minor modifications (Thisse et al. 2001). Quantitative RT-PCR was performed as previously described (
Xiong et al., 2006). Specific pairs of primers were referred to followings: 5’-AGACTGCTGTAAGGAGTGTCCTC-3’ (chd forward), 5’-CCATGAAGTCCTCTATGCATTCCG-3’ (chd reverse), 5’-CAGAGCTCACTTAGGGAAAGGCTC-3’ (bmp2 forward), 5’-CCAATAGTCTAGTGATGGGCTCCTG-3’ (bmp2 reverse), 5’-CCGGTCTGCTCAGTCCAGACC-3’ (gata1 forward), 5’-GGAAAGGGCTACTGGACCAGAC-3’ (gata1 reverse), 5’-GGA-CAGCCTCCTCCCTAAGGC-3’ (ctn forward), 5’-CAGTTCTAGCT-GGTTGATGCGG-3’(ctn reverse), 5’-ATGGATGATGAAATTGCCG-CAC-3’(β-actin forward), 5’-ACCATCACCAGAGTCCATCACG-3’ (β-actin reverse), 5’-TCAGACGAGAAGACGGAACA-3’(myod forward), 5’-CACGATGCTGGACAGACAAT-3’ (myod reverse), 5’-GGGAC-CATTGTGGTCGACAG-3’ (shha forward), 5’-GCTTGAGTTTACT-GACATCCC-3’ (shha reverse), 5’-CCCGCGCGGAGCCGCC-GCTGC-3’ (sox9b forward) and 5’-GCAGGTGCGGGTACTGG-TCCGC-3’(sox9b reverse) (
Maves et al., 2007;
Warga et al., 2009).
TUNEL assay
Embryos at 24 hpf were fixed with 4% paraformaldehyde overnight and subjected to the TUNEL assay using the fluorescein cell death detection kit (Roche 11684795910) as previously described (
Chen et al., 2009).
Higher Education Press and Springer-Verlag Berlin Heidelberg 2010
PDF
(477KB)
