Xenopus NM23-X4 regulates retinal gliogenesis through interaction with p27Xic1
- Toshiaki Mochizuki†1,
- Aikaterini Bilitou†1,
- Caroline T Waters1,
- Kamran Hussain1,
- Massimo Zollo2 and
- Shin-ichi Ohnuma1, 3Email author
© Mochizuki et al.; licensee BioMed Central Ltd. 2008
Received: 26 October 2007
Accepted: 05 January 2009
Published: 05 January 2009
In Xenopus retinogenesis, p27Xic1, a Xenopus cyclin dependent kinase inhibitor, functions as a cell fate determinant in both gliogenesis and neurogenesis in a context dependent manner. This activity is essential for co-ordination of determination and cell cycle regulation. However, very little is known about the mechanism regulating the context dependent choice between gliogenesis versus neurogenesis.
We have identified NM23-X4, a NM23 family member, as a binding partner of p27Xic1. NM23-X4 is expressed at the periphery of the ciliary marginal zone of the Xenopus retina and the expression overlaps with p27Xic1 at the central side. Our in vivo functional analysis in Xenopus retina has shown that knockdown of NM23-X4 activates gliogenesis. Furthermore, co-overexpression of NM23-X4 with p27Xic1 results in the inhibition of p27Xic1-mediated gliogenesis, through direct interaction of NM23-X4 with the amino-terminal side of p27Xic1. This inhibitory effect on gliogenesis requires serine-150 and histidine-148, which correspond to the important residues for the kinase activities of NM23 family members.
This study demonstrates that NM23-X4 functions as an inhibitor of p27Xic1-mediated gliogenesis in Xenopus retina and suggests that this activity contributes to the proper spatio-temporal regulation of gliogenesis.
Co-ordination of cell cycle regulation and developmental signaling is essential for proper development of the nervous system [1, 2]. Recently, it has become clear that developmental signals can act on cell cycle regulation of neural progenitors, while cell cycle components can influence developmental signals. However, little is known about the detailed mechanisms by which this co-ordination occurs. The retina is an ideal system to analyze the communication between cell cycle regulation and cell fate determination. It consists of six major types of neurons: retinal ganglion cells – horizontal, amacrine, bipolar, cone and rod photoreceptor cells; and one glial cell type – Müller glial cells. All neurons and glial cells are generally produced from a multipotent pool of retinal stem cells in a sequentially determined order [3, 4]. In the Xenopus retina, retinal ganglion cells are the first cell type to be determined, at around stage 27/28, while bipolar and Müller glial cells are the last to be determined at around stage 37/38.
In the retina, differentiation of neurons and glial cells is preceded by cell cycle exit of the progenitors during early G1 phase. The fate of each cell type appears to be influenced during the final cell cycle [5–8]. Insufficient accumulation of a cell fate determinant in the last cell cycle allows cell cycle progression without determination . G1-phase cell cycle components, such as cyclin Ds, have important roles in the choice between continuous proliferation and cell fate determination . Cells arrested in early G1 are more likely to be influenced by determinants than cycling cells, while cell cycle activation tends to inhibit the activity of determinants . In addition, many determinants have the ability to induce cell cycle exit in G1 [2, 11, 12], suggesting bi-directional communication between cell cycle components and determinants.
p27Xic1 is a Xenopus member of the Cip/Kip cyclin dependent kinase (CDK) inhibitor (CDKI) family that promotes Müller glial cell fate determination in the retina . The Cip/Kip CDKI family has a conserved CDK/cyclin binding domain in its amino terminus and member-specific sequences in the carboxyl terminus. p27Xic1 activates gliogenesis through its amino terminus but in a cell cycle regulation-independent manner. However, in the presence of proneural genes, such as Xath5, p27Xic1 potentiates the activities of proneural genes . A similar neurogenic activity of p27Xic1 has been demonstrated in Xenopus primary neurogenesis [14, 15] and is also mediated through the amino terminus of the protein. Also, p57Kip2 influences specification of amacrine cell subtypes in the mouse retina . All these observations indicate that the Cip/Kip CDKIs regulate both gliogenesis and neurogenesis through their amino termini in a context-dependent manner. Recently, this dual role has been demonstrated in the mammalian system as well. Mouse retinal progenitors lacking Math5, a determinant of retinal ganglion cells, upregulate p27Kip1 and differentiate into Müller glial cells . Furthermore, CDKIs regulate differentiation processes through regulation of actin and microtubule structure [18–22]. These activities are mediated by direct interaction between the carboxyl termini of CDKIs and the regulators of cytoskeletal structures. These observations indicate that CDKIs have three independent activities: cell cycle regulation, cell fate determination, and regulation of cytoskeletal structures. Previous work has shown that the amino terminus of p27Xic1 promotes neurogenesis by stabilizing neurogenin , but this mechanism cannot explain how p27Xic1 promotes glial rather than neuronal fate in the retina. To address this conflict, we have screened for proteins that interact with the amino terminus of p27Xic1 and regulate its gliogenic activity in the Xenopus retina. Here, we identify NM23-X4 as a binding partner of p27Xic1 and show that NM23-X4 regulates gliogenic activity of p27Xic1 in Xenopus retinogenesis.
Identification of NM23-X4 as a novel binding protein of p27Xic1
Interaction between NM23 and CDKIs
Next, we wanted to know if p27Xic1 interacts with other NM23 family members. As shown in Figure 2B, NM23-X3 also binds strongly to p27Xic1, while all other NM23s tested interacted weakly. The human NM23-H1 and H4 interacted with p27Xic1 similarly to their Xenopus orthologs (Figure 2C). Taken together, these results suggest that NM23-X4 and/or X3 are likely to be biologically significant binding partners of p27Xic1.
Mammals have three Cip/Kip CDKIs (p21Cip1, p27Kip1, and p57Kip2). Interestingly, NM23-X4 interacted with p21Cip1 and p57Kip2, but not with the human p27Kip1 (Figure 2D). We identified two other Xenopus CDKIs, p16Xic2 and p17Xic3, which also have gliogenic functions, similar to that of p27Xic1, although their expression levels in the developing retina are lower . Interestingly, we found that p16Xic2 does not interact with NM23-X4 (unpublished data). Sequence alignment revealed that p27Xic1 is likely to be an orthologue of p57Kip2 . Since NM23-X4 is a binding partner of p27Xic1/p57Kip2, we reasoned that NM23-X4 is likely to work upstream of, or at the same level with, p27Xic1 as an effector in retinogenesis.
NM23-X4 is a negative regulator of p27Xic1-mediated gliogenesis
NM23-X4 inhibits p27Xic1-mediated gliogenesis through their direct interaction
In order to gain more insight into the NM23-X4 action, we examined the putative NM23-X4 interaction domain of p27Xic1. Mutational analysis showed that NM23-X4 binds to two amino-terminal constructs of p27Xic1, NT (1–96) and 1–91, but not to the carboxy-terminal one (Figure 6B and data not shown). Interestingly, NM23-X4 did not interact with the 1–30 construct or the 31–96 construct, which is sufficient for both cyclin/CDK binding and Müller cell inductive activity similarly to NT and 1–91 constructs , suggesting that NM23-X4 binding requires an area covering both parts of 1–30 and 31–91. Also, the realization that the immunostained bands of NT and 1–91 are always stronger than that of the wild type suggested that the carboxy-terminal region of p27Xic1 might negatively influence the binding between NM23-X4 and p27Xic1. We then analyzed the effect of NM23-X4 and -X1 on gliogenesis mediated by p27Xic1 derivatives (Figure 6C). Interestingly, both NM23-X4 and -X1 effectively inhibited the glial-promoting effect of full length and NT constructs, while the effects of the 31–96 construct were hardly inhibited by these isoforms. These results imply that the inhibitory function of NM23 proteins requires their interaction with p27Xic1 and that NM23-X4 does not inhibit p27Xic1 activity downstream of p27Xic1. Consistent with this, NM23-X4 does not inhibit gliogenesis mediated by p16Xic2 (Figure 6D), with which it does not bind. Next, the effect of co-introduction of shX4-B and -B on gliogenesis was analyzed. Figure 6E shows that shX4-B does not activate gliogenesis when p27Xic1 is downregulated (in the presence of shXic1-B), indicating that activation of gliogenesis by shX4-B occurs in the presence of p27Xic1.
Serine-150 and histidine-148 are required for the NM23-X4 activity
We then tried to identify the residue(s) or domain(s) of the NM23 proteins important for the regulation of p27Xic1 activity. Some critical residues and domains of NM23-H1 have been identified previously (Figure 1B). Histidine-118 of NM23-H1 (histidine-148 of NM23-X4) is required for the NDPK and histidine-dependent kinase activities . Serine-120 of NM23-H1 (serine-150 of NM23-X4) is also important for histidine dependent kinase activity . The KPN loop forms part of the active center important for the kinase activity and mediates hexamer formation . Therefore, three mutants of NM23-X4 (NM23-X4 H148C, S150G, and ΔKPN) were constructed and their interaction with p27Xic1 and effects on p27Xic1-mediated gliogenesis were analyzed. As shown in Figure 6F, all three mutants interacted with p27Xic1. Interestingly, NM23-X4 H148C and S150G mutants did not inhibit p27Xic1-mediated gliogenesis, while ΔKPN did (Figure 6G), suggesting that NDPK and/or histidine dependent protein kinase activities are required for regulation of p27Xic1 activity. This might also suggest that deletion of the KPN loop in the ΔKPN mutant does not structurally abrogate the active center formation as is reported for the case of NM23-H1 .
Overexpression of NM23 family members promotes gliogenesis
Effect of NM23-X4 on cell cycle regulation
Developmental roles of the NM23 family members in retinogenesis
NM23 family members possess NDPK activity, which is required for production of ATP or GTP. In addition, several unique activities of NM23 family members have been reported, such as protein kinase, exonuclease and DNA repair activities [24, 25]. Although NM23 family members are expressed in the central nervous system, very little was known about their roles in neural development. We report that all Xenopus homologs are expressed in the retina, that NM23-X3 and -X4 are specifically expressed in retinal precursor cells in the CMZ and that NM23-X4 has an essential role in the distribution of retinal cell types through interaction with p27Xic1.
The observation that inhibition of NM23-X4 function by shRNA constructs increased the proportion of Müller cells (Figure 5) indicates that endogenous NM23-X4 functions as a regulator of gliogenesis in the retina. This function is associated with its ability to inhibit p27Xic1-mediated gliogenesis (Figure 6A). Our previous work showed that p27Xic1 plays an important role in both neurogenesis and gliogenesis in a context-dependent manner [8, 13, 14]. This means that CDKIs regulate the decision between neurogenesis versus gliogenesis depending on the presence of neurogenic stimuli in the progenitors . In the Xenopus retina, glial cells are formed after neurogenesis has taken place. The expression of p27Xic1 in the CMZ as retinogenesis progresses is consistent with its role in cell fate determination . As shown in Figure 4D–F, at stage 39, NM23-X4 is expressed at the peripheral side of the CMZ. The expression domain of NM23-X4 overlaps with that of p27Xic1 in the central region of the CMZ (Figure 4F,G). From the expression patterns of NM23-X4 and p27Xic1, their interaction and functions, we propose a model, schematically represented in Figure 8D. According to this, p27Xic1 acts at the central part of the CMZ to induce Müller glial cells. NM23-X4 is responsive to suppress this gliogenic activity of p27Xic1 at the peripheral part of the CMZ, where their two expression domains coincide. This suppression results in inhibition of Müller glial cell production and maintenance of the neurogenic potential of the early progenitors in the retinal cell lineage.
Although it was previously reported that NM23-H4 is largely located in mitochondria , our analysis using a deletion construct of NM23-X4 lacking its mitochondria-sorting signal showed that mitochondrial localization is not required for its gliogenic activity (data not shown). This is supported by the observation that all other NM23 members tested showed similar gliogenic activities (Figure 7Q) and p27Xic1 binds to both wild-type NM23-X4 and its amino-terminal processed form (data not shown). It is more likely that the NM23 family regulates the activity of CDKIs in the cytosol because Cip/Kip CDKIs localize in the cytosol and shuttle to the nucleus depending on the cellular context. This notion is also supported by the cytosolic localization observed when we are staining against tagged forms of exogenous NM23-X4 in embryos in addition to the mitochondrial staining (data not shown).
How does NM23-X4 inhibit the gliogenic activity of CDKIs? We showed that direct interaction with the amino-terminal half of p27Xic1 and a specific NM23-X4 activity, probably as a NDPK or protein kinase, are required for the inhibition of p27Xic1. Although several residues of CDKIs are phosphorylated, a majority of the phosphorylation sites are located at the carboxy-terminal half. In the amino-terminal half, only threonine-57 of p21Cip1, serine-10 and tyrosine-88 of p27Kip1 have been reported as phosphorylation sites [40–42]. However, the threonine-57 and serine-10 sites are not conserved in p27Xic1 and, moreover, NM23 family members are not known to possess tyrosine kinase activity, arguing against these sites on p27Xic1 being the ones phosphorylated by NM23-X4. Previously, NM23-H1 was reported to phosphorylate the kinase suppressor of Ras in a histidine dependent manner . Also, the aspartic acid at 319 of aldolase C is phosphorylated by NM23-H1 . In bacteria and plants, histidine kinase activity has very important roles through consequent phosphorylation of aspartic acid and glutamic acid [44, 45]. Interestingly, vertebrate CDKIs have three conserved aspartic acids and glutamic acid in the CDK/cyclin binding domain at the amino termini. Our preliminary work has shown that mutation of these residues abrogates the NM23-X4 effect (data not shown). Ongoing work will verify if CDKIs are direct targets of NM23-X4 action.
In addition to the inhibitory role of NM23-X4 on gliogenesis, we have shown that a large gain of NM23-X4 activates gliogenesis. We have provided evidence that, at the endogenous level, NM23-X4 works as a negative regulator of p27Xic1-mediated gliogenesis through direct protein interaction as shown by the knock-down analysis. Data from overexpression assays suggest that NM23-X4 works as an activator of gliogenesis in a mechanism largely independent of p27Xic1. Further work will be required to determine how NM23-X4 may act as an activator of gliogenesis. It is evident that it affects cell fate determination. Recent work has shown that purine-mediated signaling has a major role in eye development . It is of great interest that molecules involved in ATP/ADP enzymatic steps are also part of a regulatory network, along with transcription factors and other partners, to affect eye development. The exact mechanisms are yet to be elucidated in future work.
NM23-X4-mediated regulation of cell type distribution versus cell cycle regulation
Distribution of retinal cell types can be influenced by modulation of cell cycle regulation and cell death in addition to cell fate determination. Our analysis shows that endogenous NM23-X4 functions as an activator of the cell cycle (Figure 8A–C). This is in accordance with the finding that NM23-X4 binds to the CDK/cyclin domain at the amino-terminal region of p27Xic1. The binding may be antagonistic and compete with the binding of other CDK/cyclin partners, abolishing the formation of a complex. Is this cell cycle related activity responsible for the regulation of the ratio of Müller glial cells in the retina? p27Xic1 has been shown to activate Müller gliogenesis along with its cell cycle inhibitory action. The cell fate determination activity is separable from the cell cycle regulatory one, although both activities are mediated by the amino-terminal region. Our observations regarding binding of NM23-X4 with the amino-terminal region of p27Xic1 and functional assays suggest that NM23-X4 also regulates both the cell fate determination and cell cycle related activity of p27Xic1. Further analysis is required to exclude the possibility of direct cell cycle involvement in the change of cell type distribution.
Diverse roles of CDKIs and NM23s in biological processes
As mentioned, the amino termini of CDKIs regulate both cell cycle and cell fate determination. Recent papers show that the carboxyl termini of CDKIs directly bind with regulators of cytoskeletal proteins. For example, p21Cip1, p27Kip1, and p57Kip2 directly bind to Rho kinase/ROCK, RhoA and Stathmin, and LIM kinase, respectively [18–22], to influence neurite formation and cell migration. Knockout mice of CDKIs show normal cell division, suggesting that they are not essential for the fundamental cell cycle regulation. Furthermore, differentiation factors are known to induce expression of CDKIs. Taken together, these observations indicate that the Cip/Kip CDKIs have three independent functions: cell cycle inhibition, cell fate determination, and differentiation. It is demonstrated that, upon induction by secreted differentiation factors, the primary role of CDKIs lies in their ability to regulate the process of neuro/gliogenesis by co-coordinating cell cycle arrest, cell fate determination, and differentiation. Our results also demonstrate that NM23 members can work in a CDKI-dependent context and participate in this regulatory network. NM23-H1 was originally identified as an inhibitor of cancer metastasis . However, still little is known about the mechanism of this action. Recently, some papers showed that the Cip/Kip CDKIs are major regulators of cancer metastasis . Our observations overall propose the attractive hypothesis that CDKIs, regulated by NM23s, may have a major role in controlling both neural development and cancer metastasis.
We have identified NM23-X4 as a binding partner of p27Xic1 and showed that its expression in the CMZ overlaps with that of p27Xic1. Our functional analysis demonstrates that endogenous NM23-X4 negatively regulates p27Xic1-mediated gliogenesis and potentiates neurogenesis through protein-protein interactions. The glia-inhibitory function of NM23-X4 is likely to require its NDPK or histidine dependent kinase activity. Our results suggest that NM23-X4 functions in a temporal and spatial manner to regulate the process of gliogenesis during sequential retinal cell fate determination.
Materials and methods
Bacterial two-hybrid screening and plasmid construction
The bacterial two-hybrid screening was performed by BacterioMatch™ (Stratagene Texas, USA) using a bait construct, pBT-Xic1, that has the amino terminus (1-96 amino acids) of p27Xic1 and the BacterioMatch Xenopus embryo library (Stratagene). A positive clone including NM23-X4 was identified from several positive clones after confirmation steps. NM23-X1, -X2, and -X3 expression plasmids were kindly provided by Dr A Mazabraud . Through a Xenopus EST database search and blast search we identified NM23-X4 [GenBank:BU900096], -X5 [GenBank:CK805921], -X6 [GenBank:BX845338], -X7 [GenBank:BG555939] and -X8 [GenBank:BE679366]. All clones were obtained and subcloned in a pCS2FLAG vector. Mutants H148C, S150G, ΔKPN loop, and Δ(1-33) of NM23X4 were made with a site-directed mutagenesis or deletion strategy in pCS2FLAG according to standard procedures. The p27Xic1, its mutants and the p16Xic2 constructs were described previously [13, 30]. HA-tagged constructs were made using HA-pcDNA3 vector.
Design of shRNA constructs for p27Xic1 and NM23-X4
shRNA target sequences were designed using the web site siDirect . Two shRNA constructs against different target sites were constructed using pSUPER (Oligoengine Seattle, Washington, USA)). The designed oligonucleotides for p27Xic1 and NM23-X4 are as follows: shXic1-A forward, gatccccgcggtgccagggagttaaa ttcaagagatttaactccctggcaccgc tttttggaaa; shXic1-A reverse, agcttttccaaaaagcggtgccagggagttaaa tctcttgaatttaactccctggcaccgc ggg; shXic1-B forward, gatccccccgtggagactaaagatgt ttcaagagaacatctttagtctcccacgg tttttggaaa; shXic1-B reverse, agcttttccaaaaaccgtggagactaaagatgt tctcttgaaacatctttagtctccacg gggg; shX4-A forward, gatcccccgactggtgggagaaatca ttcaagagatgatttctcccacagtcg tttttggaaa; shX4-A reverse, agcttttccaaaaacgactggtgggagaaatca tctcttgaatgatttctcccaccagtcg gggg; shX4-B forward, gatccccgcagcgtggcttcacatta ttcaagagataatgtgaagccacgctgc tttttggaaa; shX4-B reverse, agcttttccaaaaagcagcgtggcttcactta tctcttgaataatgtgaagccacgctgc ggg. The oligonucleotides were annealed and then ligated in pSUPER.
Xenopus embryo manipulation, mRNA injection, and lipofection
Xenopus laevis embryos were obtained by in vitro fertilization, and staged according to Niewkoop and Faber. All embryo manipulations were in accordance with the UK Home Office regulations and guidelines. For mRNA injection, all plasmids (pCS2FLAG-NM23X1, X4, pCS2-p27Xic1, pCS2-nβgal) were linearized and transcribed in vitro using the mMessage mMachine kit (Ambion Austin, TX, USA)). The mRNAs were injected into one blastomere at the two-cell stage. Lipofection into eye primordia was performed as described . The indicated constructs were lipofected in the eye field of embryos at stage 15. At stage 41, the lipofected embryos were fixed and cryostat sectioned as previously described. For cell-tracing purposes, pGFP plasmid was co-lipofected with all plasmids. In some cases, co-lipofection was confirmed by immunohistochemistry against the myc or FLAG tag. Cell types were determined by morphology and location of cell body and nucleus along the retinal epithelium and confirmed by cell type specific antibody staining. About 400 cells from 6 retinas were counted in each condition. Each experiment was repeated at least three times. Statistical analysis was performed by one-way ANOVA.
BrdU experiments in Xenopus retina were performed as described in our previous paper . Briefly, the indicated construct was co-lipofected with a GFP plasmid at stage 15. BrdU was injected in the abdomen from stage 30 to stage 41 at intervals of 8 h. The embryos were then fixed in 4% paraformaldehyde and sectioned. The section was treated with 2 N HCl and subjected to BrdU antibody staining according to the manufacturer's protocol (Roche Lewes. East Sussex, UK)). We then determined the ratios of BrdU and GFP double positive cells of the inner nuclear layer (Müller glial cells, bipolar, amacrine and horizontal cells) and ganglion cell layer. For the proliferation assay in early embryos, mRNA was injected into a right blastomere at the two-cell stage and the injected embryos were fixed at stage 7. The cell diameter of approximately 20 cells (longest length) was measured after taking images using the Openlab software (Improvision, Coventry, West Midland, UK). Then, the ratio between the diameter of injected and uninjected sides was calculated (Figure 8C).
RNA extraction and RT-PCR
RNA extraction and RT-PCR were carried out as described . The primers used were: NM23-X1 (forward, 5'-GAGGGACTGAATGTGGTAAA-3'; reverse, 5'-TTTAAACCACAAGGCAATTT-3'); NM23-X4 (forward, 5'-AGCAGAACATTACCATGACC-3'; reverse, 5'-TAGCCTGGGAAGAGTCTGTA-3'). The primers for ornithine decarboxylase have been described previously .
In situ hybridization and immunohistochemistry
In situ hybridization of whole embryos and sections was carried out as described previously . Probes were prepared after linearization of pBluescript plasmids encoding the full-length cDNA of NM23 members or p27Xic1 and digoxigenin- or FITC-labelling in vitro with polymerase T7 or T3. In situ hybridization on cryostat sections was performed after untreated embryos were fixed, sucrose treated and embedded in OCT compound. Colour reactions were performed using BM Purple and Fast Red (Roche) according to standard procedures. Immunohistochemistry was preformed as described . Primary antibodies used were anti-R5 (1:1), anti-CRALBP (1:1000; from J Saari), anti-calbindin (1:100; Sigma St. Louis, MO, USA), anti-39.4 D5 (1:100), and anti-FLAG (1:100; Sigma). All immunostainings were visualized with fluorescent conjugated secondary antibodies under a fluorescent microscope.
Cell culture, immunoprecipitation, and western blotting
COS-7 cells were cultured in DME media/Glutamax (Invitrogen, Paisley, Renfrewshire, UK) containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin. Immunoprecipitation was carried out as described . Briefly, about 100,000 COS-7 cells/well were transfected using Lipofectamin 2000 (Invitrogen See above) according to the manufacturer's protocol. After 20 h, MG132 (final concentration 50 μM) or AMP-PNP (200 μM) was added to the medium. Then, 4 h later the cells were washed several times with 1× phosphate-buffered saline and lysed in 500 μL of lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 100 μM PMSF). The supernatants were isolated by centrifugation. We performed immunoprecipitation using about 90% of the cell lysate of COS-7 cells after measurement of the protein level and used about 10% of the lysate (20 μg of protein) for confirmation of levels of expressed proteins. For immunoprecipitation, the lysates were incubated with 2 μg of anti-FLAG M2 (Sigma) for 2 h at 4°C. Then, 20 μL of Protein G-sephrose (Sigma) was added to the mixture. The mixtures were incubated for 1 h at 4°C. After several washes by 1× phosphate-buffered saline, proteins were eluted from the sepharose with 1× SDS buffer. The obtained proteins were separated in 8–15% SDS-PAGE, blotted on the PVDF membrane (Immobilon-P, GE Healthcare, Little Chalfont
Buckinghamshire UK), and detected with chemiluminescence (GE Healthcare)). For detection, we used anti-FLAG M2 (Sigma) for FLAG-tagged protein, and 3F10 (Roche) for HA-tagged proteins. (Note that in the immunoprecipitation experiments, although the expression levels of NM23-X3, NM23-X6, NM23-H4, and p17Xic3 are lower than other proteins, these proteins resulted in significant amounts of immunoprecipitated proteins. We used about 10% of the lysate for checking the expression level and about 90% for immunoprecipitation, suggesting that the amounts of proteins in the immunoprecipitation were sufficient for the binding.)
cyclin dependent kinase
cyclin dependent kinase inhibitor
ciliary marginal zone
green fluorescent protein
nucleotide diphosphate kinase
short hairpin RNA.
We are thankful to Drs A Mazabraud and JC Saari for reagents, and Drs H Suzuki, M Shibata, I Iordanova, P Steeg and A Metha for technical advice and helpful comments. Also, we acknowledge WA Harris, LK Ferrigno, G Lupo, S Morris and J Watson for critical reading of the manuscript. This work is supported by Cancer Research UK and Fight for Sight.
- Edlund T, Jessell TM: Progression from extrinsic to intrinsic signaling in cell fate specification: a view from the nervous system. Cell. 1999, 96: 211-224. 10.1016/S0092-8674(00)80561-9.View ArticlePubMedGoogle Scholar
- Ohnuma S, Harris WA: Neurogenesis and the cell cycle. Neuron. 2003, 40: 199-208. 10.1016/S0896-6273(03)00632-9.View ArticlePubMedGoogle Scholar
- Harris WA: Cellular diversification in the vertebrate retina. Curr Opin Genet Dev. 1997, 7: 651-658. 10.1016/S0959-437X(97)80013-5.View ArticlePubMedGoogle Scholar
- Cepko CL: The roles of intrinsic and extrinsic cues and bHLH genes in the determination of retinal cell fates. Curr Opin Neurobiol. 1999, 9: 37-46. 10.1016/S0959-4388(99)80005-1.View ArticlePubMedGoogle Scholar
- Holt CE, Bertsch TW, Ellis HM, Harris WA: Cellular determination in the Xenopus retina is independent of lineage and birth date. Neuron. 1988, 1: 15-26. 10.1016/0896-6273(88)90205-X.View ArticlePubMedGoogle Scholar
- Cepko CL, Austin CP, Yang X, Alexiades M, Ezzeddine D: Cell fate determination in the vertebrate retina. Proc Natl Acad Sci USA. 1996, 93: 589-595. 10.1073/pnas.93.2.589.PubMed CentralView ArticlePubMedGoogle Scholar
- McConnell SK, Kaznowski CE: Cell cycle dependence of laminar determination in developing neocortex. Science. 1991, 254: 282-285. 10.1126/science.1925583.View ArticlePubMedGoogle Scholar
- Ohnuma S, Hopper S, Wang KC, Philpott A, Harris WA: Co-ordinating retinal histogenesis: early cell cycle exit enhances early cell fate determination in the Xenopus retina. Development. 2002, 129: 2435-2446.PubMedGoogle Scholar
- Matter-Sadzinski L, Puzianowska-Kuznicka M, Hernandez J, Ballivet M, Matter JM: A bHLH transcriptional network regulating the specification of retinal ganglion cells. Development. 2005, 132: 3907-3921. 10.1242/dev.01960.View ArticlePubMedGoogle Scholar
- Levine EM, Green ES: Cell-intrinsic regulators of proliferation in vertebrate retinal progenitors. Semin Cell Dev Biol. 2004, 15: 63-74. 10.1016/j.semcdb.2003.09.001.View ArticlePubMedGoogle Scholar
- Farah MH, Olson JM, Sucic HB, Hume RI, Tapscott SJ, Turner DL: Generation of neurons by transient expression of neural bHLH proteins in mammalian cells. Development. 2000, 127: 693-702.PubMedGoogle Scholar
- Cremisi F, Philpott A, Ohnuma S: Cell cycle and cell fate interactions in neural development. Curr Opin Neurobiol. 2003, 13: 26-33. 10.1016/S0959-4388(03)00005-9.View ArticlePubMedGoogle Scholar
- Ohnuma S, Philpott A, Wang K, Holt CE, Harris WA: p27Xic1, a Cdk inhibitor, promotes the determination of glial cells in Xenopus retina. Cell. 1999, 99: 499-510. 10.1016/S0092-8674(00)81538-X.View ArticlePubMedGoogle Scholar
- Vernon AE, Devine C, Philpott A: The cdk inhibitor p27Xic1 is required for differentiation of primary neurones in Xenopus. Development. 2003, 130: 85-92. 10.1242/dev.00193.View ArticlePubMedGoogle Scholar
- Carruthers S, Mason J, Papalopulu N: Depletion of the cell-cycle inhibitor p27(Xic1) impairs neuronal differentiation and increases the number of ElrC(+) progenitor cells in Xenopus tropicalis. Mech Dev. 2003, 120: 607-616. 10.1016/S0925-4773(03)00010-8.View ArticlePubMedGoogle Scholar
- Dyer MA, Cepko CL: The p57Kip2 cyclin kinase inhibitor is expressed by a restricted set of amacrine cells in the rodent retina. J Comp Neurol. 2001, 429: 601-614. 10.1002/1096-9861(20010122)429:4<601::AID-CNE7>3.0.CO;2-V.View ArticlePubMedGoogle Scholar
- Le TT, Wroblewski E, Patel S, Riesenberg AN, Brown NL: Math5 is required for both early retinal neuron differentiation and cell cycle progression. Dev Biol. 2006, 295: 764-778. 10.1016/j.ydbio.2006.03.055.View ArticlePubMedGoogle Scholar
- Besson A, Gurian-West M, Schmidt A, Hall A, Roberts JM: p27Kip1 modulates cell migration through the regulation of RhoA activation. Genes Dev. 2004, 18: 862-876. 10.1101/gad.1185504.PubMed CentralView ArticlePubMedGoogle Scholar
- Baldassarre G, Belletti B, Nicoloso MS, Schiappacassi M, Vecchione A, Spessotto P, Morrione A, Canzonieri V, Colombatti A: p27(Kip1)-stathmin interaction influences sarcoma cell migration and invasion. Cancer Cell. 2005, 7: 51-63. 10.1016/j.ccr.2004.11.025.View ArticlePubMedGoogle Scholar
- Lee S, Helfman DM: Cytoplasmic p21Cip1 is involved in Ras-induced inhibition of the ROCK/LIMK/cofilin pathway. J Biol Chem. 2004, 279: 1885-1891. 10.1074/jbc.M306968200.View ArticlePubMedGoogle Scholar
- Tanaka H, Yamashita T, Asada M, Mizutani S, Yoshikawa H, Tohyama M: Cytoplasmic p21(Cip1/WAF1) regulates neurite remodeling by inhibiting Rho-kinase activity. J Cell Biol. 2002, 158: 321-329. 10.1083/jcb.200202071.PubMed CentralView ArticlePubMedGoogle Scholar
- Yokoo T, Toyoshima H, Miura M, Wang Y, Iida KT, Suzuki H, Sone H, Shimano H, Gotoda T, Nishimori S, Tanaka K, Yamada N: p57Kip2 regulates actin dynamics by binding and translocating LIM-kinase 1 to the nucleus. J Biol Chem. 2003, 278: 52919-52923. 10.1074/jbc.M309334200.View ArticlePubMedGoogle Scholar
- Kimura N, Shimada N, Ishijima Y, Fukuda M, Takagi Y, Ishikawa N: Nucleoside diphosphate kinases in mammalian signal transduction systems: recent development and perspective. J Bioenerg Biomembr. 2003, 35: 41-47. 10.1023/A:1023489722460.View ArticlePubMedGoogle Scholar
- Ouatas T, Salerno M, Palmieri D, Steeg PS: Basic and translational advances in cancer metastasis: Nm23. J Bioenerg Biomembr. 2003, 35: 73-79. 10.1023/A:1023497924277.View ArticlePubMedGoogle Scholar
- Postel EH, Berberich SJ, Flint SJ, Ferrone CA: Human c-myc transcription factor PuF identified as nm23-H2 nucleoside diphosphate kinase, a candidate suppressor of tumor metastasis. Science. 1993, 261: 478-480. 10.1126/science.8392752.View ArticlePubMedGoogle Scholar
- Carotenuto P, Marino N, Bello AM, D'Angelo A, Di Porzio U, Lombardi D, Zollo M: PRUNE and NM23-M1 expression in embryonic and adult mouse brain. J Bioenerg Biomembr. 2006Google Scholar
- Ouatas T, Abdallah B, Gasmi L, Bourdais J, Postel E, Mazabraud A: Three different genes encode NM23/nucleoside diphosphate kinases in Xenopus laevis. Gene. 1997, 194: 215-225. 10.1016/S0378-1119(97)00160-1.View ArticlePubMedGoogle Scholar
- Chuang LC, Zhu XN, Herrera CR, Tseng HM, Pfleger CM, Block K, Yew PR: The C-terminal domain of the Xenopus cyclin-dependent kinase inhibitor, p27Xic1, is both necessary and sufficient for phosphorylation-independent proteolysis. J Biol Chem. 2005, 280: 35290-35298. 10.1074/jbc.M506430200.View ArticlePubMedGoogle Scholar
- Hartsough MT, Morrison DK, Salerno M, Palmieri D, Ouatas T, Mair M, Patrick J, Steeg PS: Nm23-H1 metastasis suppressor phosphorylation of kinase suppressor of Ras via a histidine protein kinase pathway. J Biol Chem. 2002, 277: 32389-32399. 10.1074/jbc.M203115200.View ArticlePubMedGoogle Scholar
- Daniels M, Dhokia V, Richard-Parpaillon L, Ohnuma S: Identification of Xenopus cyclin-dependent kinase inhibitors, p16Xic2 and p17Xic3. Gene. 2004, 342: 41-47. 10.1016/j.gene.2004.07.038.View ArticlePubMedGoogle Scholar
- Perron M, Kanekar S, Vetter ML, Harris WA: The genetic sequence of retinal development in the ciliary margin of the Xenopus eye. Dev Biol. 1998, 199: 185-200. 10.1006/dbio.1998.8939.View ArticlePubMedGoogle Scholar
- Li M, Rohrer B: Gene silencing in Xenopus laevis by DNA vector-based RNA interference and transgenesis. Cell Res. 2006, 16: 99-105. 10.1038/sj.cr.7310013.View ArticlePubMedGoogle Scholar
- Zhou Y, Ching YP, Kok KH, Kung HF, Jin DY: Post-transcriptional suppression of gene expression in Xenopus embryos by small interfering RNA. Nucleic Acids Res. 2002, 30: 1664-1669. 10.1093/nar/30.7.1664.PubMed CentralView ArticlePubMedGoogle Scholar
- Miskevich F, Doench JG, Townsend MT, Sharp PA, Constantine-Paton M: RNA interference of Xenopus NMDAR NR1 in vitro and in vivo. J Neurosci Methods. 2006, 152: 65-73. 10.1016/j.jneumeth.2005.08.010.View ArticlePubMedGoogle Scholar
- Kuriyama S, Lupo G, Ohta K, Ohnuma S, Harris WA, Tanaka H: Tsukushi controls ectodermal patterning and neural crest specification in Xenopus by direct regulation of BMP4 and X-delta-1 activity. Development. 2006, 133: 75-88. 10.1242/dev.02178.View ArticlePubMedGoogle Scholar
- Wagner PD, Steeg PS, Vu ND: Two-component kinase-like activity of nm23 correlates with its motility-suppressing activity. Proc Natl Acad Sci USA. 1997, 94: 9000-9005. 10.1073/pnas.94.17.9000.PubMed CentralView ArticlePubMedGoogle Scholar
- Postel EH, Weiss VH, Beneken J, Kirtane A: Mutational analysis of NM23-H2/NDP kinase identifies the structural domains critical to recognition of a c-myc regulatory element. Proc Natl Acad Sci USA. 1996, 93: 6892-6897. 10.1073/pnas.93.14.6892.PubMed CentralView ArticlePubMedGoogle Scholar
- Palacios F, Schweitzer JK, Boshans RL, D'Souza-Schorey C: ARF6-GTP recruits Nm23-H1 to facilitate dynamin-mediated endocytosis during adherens junctions disassembly. Nature Cell Biology. 2002, 4: 929-936. 10.1038/ncb881.View ArticlePubMedGoogle Scholar
- Milon L, Meyer P, Chiadmi M, Munier A, Johansson M, Karlsson A, Lascu I, Capeau J, Janin J, Lacombe ML: The human nm23-H4 gene product is a mitochondrial nucleoside diphosphate kinase. J Biol Chem. 2000, 275: 14264-14272. 10.1074/jbc.275.19.14264.View ArticlePubMedGoogle Scholar
- Rossig L, Badorff C, Holzmann Y, Zeiher AM, Dimmeler S: Glycogen synthase kinase-3 couples AKT-dependent signaling to the regulation of p21Cip1 degradation. J Biol Chem. 2002, 277: 9684-9689. 10.1074/jbc.M106157200.View ArticlePubMedGoogle Scholar
- Grimmler M, Wang Y, Mund T, Cilensek Z, Keidel EM, Waddell MB, Jakel H, Kullmann M, Kriwacki RW, Hengst L: Cdk-inhibitory activity and stability of p27Kip1 are directly regulated by oncogenic tyrosine kinases. Cell. 2007, 128: 269-280. 10.1016/j.cell.2006.11.047.View ArticlePubMedGoogle Scholar
- Ishida N, Kitagawa M, Hatakeyama S, Nakayama K: Phosphorylation at serine 10, a major phosphorylation site of p27(Kip1), increases its protein stability. J Biol Chem. 2000, 275: 25146-25154. 10.1074/jbc.M001144200.View ArticlePubMedGoogle Scholar
- Wagner PD, Vu ND: Histidine to aspartate phosphotransferase activity of nm23 proteins: phosphorylation of aldolase C on Asp-319. Biochem J. 2000, 346 (Pt 3): 623-630. 10.1042/0264-6021:3460623.PubMed CentralView ArticlePubMedGoogle Scholar
- Santos JL, Shiozaki K: Fungal histidine kinases. Sci STKE. 2001, 2001: RE1-10.1126/stke.2001.98.re1.PubMedGoogle Scholar
- Stock AM, Robinson VL, Goudreau PN: Two-component signal transduction. Annu Rev Biochem. 2000, 69: 183-215. 10.1146/annurev.biochem.69.1.183.View ArticlePubMedGoogle Scholar
- Masse K, Bhamra S, Eason R, Dale N, Jones EA: Purine-mediated signalling triggers eye development. Nature. 2007, 449: 1058-1062. 10.1038/nature06189.View ArticlePubMedGoogle Scholar
- Ohnuma S, Mann F, Boy S, Perron M, Harris WA: Lipofection strategy for the study of Xenopus retinal development. Methods. 2002, 28: 411-419. 10.1016/S1046-2023(02)00260-8.View ArticlePubMedGoogle Scholar
- Mochizuki T, Miyazaki H, Hara T, Furuya T, Imamura T, Watabe T, Miyazono K: Roles for the MH2 domain of Smad7 in the specific inhibition of transforming growth factor-beta superfamily signaling. J Biol Chem. 2004, 279: 31568-31574. 10.1074/jbc.M313977200.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.