Homeostatic regulation of ionic currents is of paramount importance during periods of synaptic growth or remodeling. Our previous work has identified the translational repressor Pumilio (Pum) as a regulator of sodium current (I Na) and excitability in Drosophila motoneurons. In this current study, we show that Pum is able to bind directly the mRNA encoding the Drosophila voltage-gated sodium channel paralytic (para). We identify a putative binding site for Pum in the 3′ end of the para open reading frame (ORF). Characterization of the mechanism of action of Pum, using whole-cell patch clamp and real-time reverse transcription-PCR, reveals that the full-length protein is required for translational repression of para mRNA. Additionally, the cofactor Nanos is essential for Pum-dependent para repression, whereas the requirement for Brain Tumor (Brat) is cell type specific. Thus, Pum-dependent regulation of I Na in motoneurons requires both Nanos and Brat, whereas regulation in other neuronal types seemingly requires only Nanos but not Brat. We also show that Pum is able to reduce the level of nanos mRNA and as such identify a potential negative-feedback mechanism to protect neurons from overactivity of Pum. Finally, we show coupling between I Na (para) and I K (Shal) such that Pum-mediated change in para results in a compensatory change in Shal. The identification of para as a direct target of Pum represents the first ion channel to be translationally regulated by this repressor and the location of the binding motif is the first example in an ORF rather than in the canonical 3′-untranslated region of target transcripts.
Neuronal activity is regulated by homeostatic mechanisms that serve to maintain membrane excitability within predefined limits. This is achieved, at least in part, by continual adjustment of both ligand- and voltage-gated ionic conductances to maintain stable action potential firing rates in response to changing synaptic excitation (Turrigiano and Nelson, 2000; Marder and Prinz, 2002; Davis, 2006). Such regulation is predicted to be particularly predominant when neural circuit synaptic activity is changing rapidly, for example during both neuronal circuit development and in the formation of memory (Turrigiano, 1999). However, although now well established, the molecular pathways that underlie homeostatic regulation remain mostly unknown.
Previous studies indicate that activity-dependent regulation of voltage-gated sodium channels is central to the control of membrane excitability in both mammalian and invertebrate neurons (Desai et al., 1999; Baines et al., 2001; Baines, 2003; Mee et al., 2004). Studies in Drosophila have shown that increased synaptic excitation of motoneurons is countered by a decrease in sodium current (I Na) and membrane excitability in these cells (Baines, 2003). Similar, but opposite, changes in I Na and excitability are observed in mutants that display decreased synaptic excitability (Baines et al., 2001). These changes require the known translational repressor Pumilio (Pum), which we have shown previously is both necessary and sufficient for activity-dependent changes of I Na in Drosophila motoneurons (Mee et al., 2004). Our model predicts that prolonged change in exposure to synaptic excitation is countered by a reciprocal Pum-dependent regulation in translation of paralytic (para) mRNA and membrane excitability.
The role of Pum is well described from studies of early Drosophila embryogenesis (Barker et al., 1992; Murata and Wharton, 1995; Zamore et al., 1997; Wharton et al., 1998). Specification of the abdomen requires Pum-dependent repression of translation of hunchback (hb) mRNA. The first step begins with the recognition and binding of Pum to the Nanos response element (NRE)-motif located in the 3′-untranslated region (UTR) of hb mRNA (Zamore et al., 1997; Wharton et al., 1998). Once bound, Pum then recruits the cofactors Nanos (Sonoda and Wharton, 1999) and Brain Tumor (Brat) (Sonoda and Wharton, 2001) to form a repressor complex that results in the translational repression of hb mRNA. The mechanism of repression involves both deadenylation and poly(A)-independent silencing (Chagnovich and Lehmann, 2001). In addition to its characterized roles in repression of hb, Pum has also been shown to bind, and repress translation of, mRNAs encoding the eukaryotic initiation factor 4E (eIF4E) (Menon et al., 2004) and Cyclin B (CycB) (Asaoka-Taguchi et al., 1999; Kadyrova et al., 2007). Indeed, these few mRNAs may represent just the tip of the iceberg because the actual list of targets is likely to be extensive based on a recent demonstration that Pum associates with >1000 different mRNAs in the ovaries of adult flies (Gerber et al., 2006). Pum proteins are evolutionarily conserved from yeast to mammals (Spassov and Jurecic, 2002; Wickens et al., 2002), and, moreover, Pum expression is activity dependent in mammalian neurons in culture (Vessey et al., 2006).
In this study, we report that Pum is able to directly bind para mRNA (encoding the Drosophila voltage-gated Na+ channel). The mechanism of para translational repression shows similarities and differences to Pum-dependent repression of hb mRNA. We show that, unlike repression of hb, full-length Pum is necessary for para repression. As for most other Pum-dependent repressed transcripts described to date, para repression requires the presence of the cofactor Nanos. However, the requirement for the cofactor Brat is neuronal type specific. We also show that Pum is sufficient to downregulate nanos mRNA levels in the CNS, a property that may serve to protect neurons from the effects of overactivity of this translational repressor.
Materials and Methods
Flies were maintained on apple juice agar plates supplemented with live yeast paste at 25°C. Wild type was Canton-S (CS). Tissue-specific expression of transgenes in the nervous system was achieved using the GAL4/UAS system (Brand and Perrimon, 1993). RN2-O-GAL4 (homozygous viable on the second chromosome) or RN2-E-GAL4 (homozygous viable on the third chromosome) were used to express UAS transgenes in aCC/RP2 motoneurons (Fujioka et al., 1999; Baines, 2003). These are identical transgenes inserted on different chromosomes. For real-time PCR experiments, 1407-GAL4 (homozygous viable on the second chromosome) was used to express transgenes in all CNS neurons. UAS-pum full-length (homozygous viable on the second chromosome) (Schweers et al., 2002), UAS-pum RBD (homozygous viable on the second chromosome) (Menon et al., 2004), UAS-pum RBD-V5 (homozygous viable on the third chromosome), UAS-PumG1330D (homozygous viable on the second chromosome) (Ye et al., 2004), and UAS-nanos (homozygous viable on second chromosome) (Ye et al., 2004) have been described previously. UAS-brat (on the second chromosome, rebalanced over CyO GFP for this work) is described by Frank et al. (2002). nanos 17 was provided by the Bloomington Stock Centre and rebalanced over TM3SerGFP. para was removed using a small deficiency [Df(1) D34] (Baines and Bate, 1998) and rebalanced over FM7GFP.
TAP-PumRBD pull-down assays.
Five grams of elaV-GAL4; UAS-TAP-pum RBD or elaV-GAL4; + mock control flies were collected 1–3 d after eclosure. Flies were frozen in liquid nitrogen and TAP-PumRBD was affinity-purified from extracts as previously described (Gerber et al., 2006). For reverse transcription (RT)-PCR, 1 μg and 100 ng of total RNA isolated from extracts and from tobacco etch virus (TEV) protease eluates, respectively, were mixed with oligo-dT [dNV(T)22] and random nonamer primers (5 μg each) and made up to 15 μl with RNase-free water. The mix was incubated at 70°C for 10 min followed by incubation on ice. First-strand buffer (Invitrogen, Carlsbad, CA) (15 μl) was supplemented with 0.5 mm deoxyribonucleoside triphosphates (dNTPs) and 20 U of RNaseOUT (Invitrogen) was added to the mixture. A fraction, 10 μl, was transferred into a second tube (−RT control) and 1 μl (100 U) of Superscript RT II (Invitrogen) was added to the remaining 20 μl. Samples were incubated for 2 h at 42°C, 5 min at 95°C, and put on ice. PCR was conducted with 1.5 μl of the RT reaction with oligo pairs para-T7Fw1 (5′-TAATACGACTCACTATAGGGCACCCAGTACATACGCTATG-3′, which bears sequences for the T7 promotor at the 5′ end) and para-Rev1 (5′-CAGACATCCGCCGTGCGCGACGTG-3′) for amplification of para transcripts, and gfat2F (5′-CTCCTCGCAGATTAGGATCG-3′) and gfat2R (5′-AAGGCCTACACCTCCCAGTT-3′) to amplify glutamine-fructose-6-phosphate aminotransferase 2 (gfat2) transcripts. PCR was performed for 2 min at 94°C, 28 cycles at 94°C for 30 s, 53°C for 30 s, 72°C for 1 min, and 2 min at 72°C. To amplify gfat2, annealing was done at 58°C.
Pull-down experiments with biotinylated RNAs.
Synthesis of biotinylated transcripts and pull-down assay were performed as described with minor modifications (Gerber et al., 2006). para was amplified by PCR from 100 ng of Drosophila genomic DNA with primer pairs para-T7Fw1 and para-Rev1. A total of 4 pmol of biotinylated RNA was mixed with 200 μl of extract (OD280, 25) prepared from elaV-GAL4; UAS-TAP-pum RBD adult flies.
Embryo and larvae dissection.
Newly hatched larvae or late-stage 17 (19–21 h after egg laying at 25°C) embryos were dissected, and central neurons were accessed for electrophysiology as described by Baines and Bate (1998). Late-stage 17 embryos were first dechorionated using 50% bleach for 2 min, and the vitelline membrane was then manually removed using sharp tungsten wires. The larva/embryo was visualized using a water immersion lens (total magnification, 600×) combined with Normarski optics (BX51W1 microscope; Olympus Optical, Tokyo, Japan).
Recordings were performed in young first instar larvae, 1–4 h after hatching, or late-stage 17 embryos (in the case of nonviable genotype) at room temperature (22–24°C). Whole-cell voltage-clamp recordings were done using thick-walled borosilicate glass electrodes (GC100F-10; Harvard Apparatus, Edenbridge, UK), fire polished to resistances of between 15 and 20 MΩ. Cells were initially identified based on both size and dorsal position in the ventral nerve cord. Unequivocal identification was determined after recording by labeling with 0.1% Alexa Fluor 488 hydrazyde, sodium salt (Invitrogen), which was included in the patch saline. Recordings were made using a Multiclamp 700B amplifier controlled by pClamp 9.2 (Molecular Devices, Sunnyvale, CA). Only cells with input resistance >1 GΩ were accepted for analysis. To better resolve I Na, an on-line leak subtraction protocol was used (P/4). Currents shown are the average of three trials for I Na and five trials for I K. Currents shown were normalized for cell capacitance. To determine the effect of gene expression on electrical properties, we analyzed the peak current for each ion (I Na at −20 mV, I Kfast and I Kslow at +45 mV, and I Ba(Ca) at −10 mV).
Saline for dissection consisted of the following (in mm): 135 NaCl, 5 KCl, 4 MgCl2·6H2O, 2 CaCl2·2H2O, 5 N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (TES), and 36 sucrose, pH 7.15. For isolation of I Na, the following solution was used (in mm): 100 NaCl, 6 KCl, 2 MgCl2·6H2O, 2 sucrose, 50 tetraethylammonium chloride (TEA), 10 4-aminopyridine (4-AP), and 10 HEPES, pH 7.1. For isolation of I K, the following solution was used (in mm): 135 NaCl, 5 KCl, 4 MgCl2·6H2O, 2 CaCl2·2H2O, 5 TES, 36 sucrose, and 10−6 tetrodotoxin (TTX) (Alomone Labs, Jerusalem, Israel), pH 7.1. For isolation of I Ba(Ca), the following solution was used (in mm): 50 NaCl, 6 KCl, 50 BaCl, 10 MgCl2·6H2O, 10 glucose, 50 TEA, 10 HEPES, and 10−6 TTX, pH 7.1. Internal patch solution was as follows (in mm): 140 K+ methylsulfonate (KCH3SO3), 2 MgCl2·6H2O, 2 EGTA, 2 KCl, and 20 HEPES, pH 7.4. When recording I Na or I Ba(Ca), CsCl2 was substituted for KCH3SO3.
RNA was extracted from whole late-stage 17 embryos or first instar larvae (para) or from isolated CNSs from these stages (Shal, slo, nanos, DmCa1A, and pum mRNA detection) using a Qiagen RNeasy Mini kit (Qiagen, Crawley, UK). Briefly, 35 late-stage 17 embryos or first instar larvae (or their isolated CNSs) were homogenized with a plastic mortar followed by repeated passage through a 20-gauge needle in 350 μl of lysis buffer containing 0.1 m β-mercaptoethanol. The lysate was then centrifuged, 1 vol of 70% ethanol added and passed through an RNeasy column. After washing in buffer, immobilized nucleic acids were then treated with ∼190 U of DNase I for 15 min, washed again in stages according to manufacturer's protocol, and then eluted in ∼35 μl of RNase-free water. Quantification of RNA concentration in eluates was made using a ND-1000 Nanodrop spectrophotometer (Nanodrop, Wilmington, DE).
Synthesis of cDNA was performed following the protocol in RevertAid First Strand cDNA Synthesis kit (Fermentas, York, UK). RNA (concentration ≥1.25 ng/μl) was mixed with 0.2 μg (1 μl) random hexamer primers (Fermentas) and made up to 11 μl with RNase-free water. The mix was incubated at 65°C for 5 min to denature RNA followed by incubation on ice for 1 min. A total of 4 μl of reaction buffer (in mm: 250 Tris-HCl, 250 KCl, 20 MgCl2, 50 DTT), 2 μl of 10 mm dNTPs, and 1 μl of Ribolock ribonuclease inhibitor (Fermentas) were added, and the mix was incubated at 25°C for 5 min. Then, after addition of 1 μl of RevertAid M-MuLV (monkey murine leukemia virus) reverse transcriptase (Fermentas), the reaction was subsequently incubated for 10 min at 25°C, 60 min at 42°C, and 15 min at 65°C. From the total reaction volume of 20 μl, 1 μl of cDNA was used for each PCR.
Clone Manager software (Sci-Ed, Cary, NC) was used to design primers for para, pum, nanos, slo, Shal, DmCa1A, and ribosomal protein 49 (rp49), a housekeeping gene. All primers are shown in 5′ to 3′ orientation: rp49 forward and reverse primers, CACCGGAAACTCAATGGATACTG and TTCTTCACGATCTTGGGCC; para forward and reverse primers, GATCTATATGGGCGTGCTCACGCAGAAGTG and TGCAGGCACACGTAATCGTCGTCGCATTG; pum forward and reverse primers, CGGCCCAACAGAATCTCTACTC and GCGGCGACCCGTCAA; nanos forward and reverse primers, CAATGGCGGCAACTTAATG and CCACACGTTGTTCAGATG; slo forward and reverse primers, CTTAACACACAAGGAAAAATTTCGTGG and GTGTTCGTTCTTTTGAATTTGAATTGG; Shal forward and reverse primers, ATGGCCAACGTGGTGGAGACGGTGCCGTGTGG and TTCGCTGGCGCAGGACTTGAGCGTGTAGCC; DmCa1A forward and reverse primers, TGTACTGCCATCTCCAGTTC and GTGCGTATCTTGGTGTTGTC; respectively.
A Roche Lightcycler 1.5 was used to undertake relative quantification of target mRNAs. Reactions contained 5 μl of Mastermix (3 mm MgCl2, Taq polymerase, dNTPs; Biogene, Kimbolton, UK), 0.5 μl of each forward and reverse primer (both 10 mm), 2 μl of water, and 1 μl of 1:1000 dilution SYBR Green (Invitrogen) and 1 μl of cDNA. Cycling was as follows: initial denaturation of 10 s at 94°C, and then 35 cycles of 5 s annealing at 54°C for para, 65°C for Shal, 60°C for slo, and 57°C for all the other primer pairs used (determination of rp49 was performed at either temperature), extension at 72°C for 10 or 20 s (for amplicons of <250 or >250 nt, respectively), and denaturation at 94°C. Reactions were performed in triplicate. Fluorescence was acquired at the end of each elongation step using the F1 detection channel with a gain of 1. Authenticity of PCR products was verified by melting-curve analysis and comparison with melting curves for a nontemplate control for each primer pair used. mRNA levels are expressed as relative fold change normalized against rp49 mRNA. The comparative cycle threshold (Ct) method (User Bulletin 2, 1997; Applied Biosystems, Foster City, CA) was used to analyze the data by generating relative values of the amount of target cDNA (Mee et al., 2004).
Statistical significance between control and experimental groups was calculated using a nonpaired t test with a confidence interval of *p ≤ 0.05 or **p ≤ 0.01.
Pumilio binds para mRNA
Our previous work has shown that increased expression of Pum is able to downregulate the mRNA of para and reduce the peak amplitude voltage-gated I Na in identified motoneurons aCC/RP2 (Mee et al., 2004). However, no evidence for binding between Pum and para mRNA has been demonstrated. To directly test whether Pum associates with para mRNA in the Drosophila nervous system, we generated transgenic flies that express a tandem-affinity purification (TAP)-tagged RNA binding domain (RBD) of Pum (Gerber et al., 2006) specifically in neuronal cells using the GAL4/UAS system. Tagged PumRBD was then recovered from adult fly extracts by affinity selection on IgG beads and subsequent cleavage with TEV protease as previously described (Gerber et al., 2006). As a control, the same procedure was performed in parallel with flies not expressing the tagged PumRBD construct (mock control). Microarray analysis of the associated RNAs revealed significant association of para mRNA in elaV:PumRBD versus mock controls (A. Gerber and S. Luschnig, unpublished data). To further substantiate these results, we performed RT-PCR on RNA from the affinity isolates. The para transcript was detected in RNA isolated from affinity-purified material of TAP PumRBD expressing flies but was not detectable in RNA isolated from mock controls (Fig. 1 A). In contrast, no particular enrichment was seen for the messages coding for gfat2 (glutamine-fructose-6-phosphate aminotransferase 2) or actin (act5), both of which are abundant messages that are not predicted to be Pum targets (Fig. 1 A) (data not shown). Thus, together, the microarray and RT-PCR data show that para mRNA associates with Pum in neurons.
To determine the location of potential Pum-binding sites in the para transcript, a bioinformatic search was undertaken using a consensus core 8 nt motif (UGUAAAUA) previously identified from an analysis of Pum-bound transcripts from ovaries (Gerber et al., 2006). This search revealed one exact match in the open reading frame (ORF) of para (chromosome coordinates X: 16358030, 16358037). To test whether this sequence was sufficient to bind Pum, a region encompassing this motif (chromosome coordinates X: 16357822, 16358541) was used in RNA pull-down experiments using synthetic biotinylated transcripts added to Drosophila extracts expressing TAP-PumRBD (Fig. 1 B). Similar to a positive control RNA encoding a fragment of Vha16 3′-UTR, previously shown to bind to PumRBD (Gerber et al., 2006), this region of para mRNA was able to bind to PumRBD (Fig. 1 B, lanes 2 and 3). Moreover, this binding was specifically competed by the addition of excess of a 10 nt RNA fragment comprising the Pum-binding consensus sequence (Fig. 1 B, lane 4) but not with a control RNA where the conserved core UGU was mutated to ACA (Fig. 1 B, lane 5). Finally, mutation of the Pum-binding consensus sequence in the para transcript (cGUcAAUA) is sufficient to abolish binding of Pum (data not shown). These results not only corroborate our previous binding observations gained from microarray and PCR but, additionally, show that a region within the para ORF is sufficient to bind Pum.
Full-length Pumilio is necessary for repression of para
The PumRBD consists of eight imperfect repeats that mediate the binding of the target mRNA and the cofactor Nanos to produce a translation repressor complex (Sonoda and Wharton, 1999). A high degree of conservation of the RBD has been described in proteins of the Pum family from yeast to humans (Zamore et al., 1997). Expression of only the PumRBD has been reported to be sufficient for a partial rescue of the pum mutant embryonic abdominal segmentation phenotype resulting from the lack of Pum-mediated hb mRNA repression (Wharton et al., 1998). However, the portion of Pum protein relevant to CNS-related processes remains controversial. On the one hand, PumRBD is sufficient to mimic the dendrite branching phenotype resulting from full-length Pum expression in dendritic arborization neurons (Ye et al., 2004). In contrast, full-length Pum is required to rescue the neuromuscular junction defects seen in pum mutants (Menon et al., 2004). Given this controversy, we investigated whether overexpression of the PumRBD on its own was sufficient to repress para mRNA and I Na in central neurons. To do this, we overexpressed UAS-pum RBD, a construct containing only the RBD of Pum. We also tested UAS-pum RBD-V5 and UAS-TAP-pum RBD, independent constructs also bearing only the pum RBD. Overexpression of these constructs was tested for their ability to downregulate both I Na in identified motoneurons and para mRNA in whole CNS. Figure 2, A–C, shows that only full-length Pum is able to repress I Na in aCC/RP2. Real-time RT-PCR quantification of para mRNA similarly shows that only full-length Pum is able to downregulate para mRNA in whole CNS (Fig. 2 D). These results suggest that the translational repression mechanism of para mRNA requires the participation of parts of the Pum protein outside the RBD. This represents a clear difference between the mechanism of repression of the para transcript compared with the known mechanism of repression of hb mRNA.
Nanos is necessary for Pumilio-dependent para repression
The mechanism for translational repression of a majority of mRNAs by Pum requires the participation of the cofactor Nanos (Sonoda and Wharton, 1999; Kadyrova et al., 2007). Therefore, we tested whether this cofactor is also necessary for Pum to downregulate para mRNA and I Na. We overexpressed full-length Pum in a zygotic nanos 17 homozygous mutant background. This mutation contains a single amino acid change in the C-terminal region of Nanos that is sufficient to influence embryonic segmentation presumably through disrupted translational repression of hb mRNA (Curtis et al., 1997). nanos 17 is a weak allele that produces embryos with variable numbers of abdominal segments and, although stronger nanos alleles have been described, nanos 17 was chosen for this study because it produces viable first instar larvae. Whole-cell recordings from aCC/RP2 motoneurons in larvae overexpressing full-length Pum in a nanos 17 mutant background failed to show a reduction in I Na (Fig. 3 A,B), indicative that Nanos is necessary for Pum-dependent repression of para mRNA. One copy of wild-type nanos in a heterozygous nanos 17 mutant (nanos 17/+) rescues the ability of Pum to downregulate I Na. The extent of downregulation in the heterozygote is equivalent to that observed when two normal copies of nanos are present (i.e., wild type), suggesting that, although necessary, the dose of nanos does not determine the level of para repression (see also Fig. 5 A). This represents another key difference between para and hb repression in which Nanos is the principal factor limiting the translational repression of hb mRNA by means of a posterior-to-anterior concentration gradient (Barker et al., 1992). The nanos 17 mutation alone does not show any differences in I Na presumably because enough functional Nanos protein is present to allow endogenous Pum (but not increased Pum expression) to function normally (Fig. 3 A,B). Quantification of para mRNA after pan-neuronal overexpression of Pum in the nanos 17 mutant background also shows a clear necessity of Nanos for the downregulation of para mRNA (Fig. 3 C). Thus, it would seem that, as for most Pum mRNA targets examined so far, translational repression of para requires the cofactor Nanos. However, it should be noted that this conclusion is based on the use of nanos 17, which is a point mutation and not a genetic null.
A requirement of Brain Tumor for para repression is cell type specific
Brat is a second cofactor that is required for the translational repression of hb mRNA (Sonoda and Wharton, 2001). To analyze the requirement of Brat in the translational regulation of para, we took advantage of UAS-pum G1330D, a single amino acid substitution that renders Pum unable to recruit Brat to the repression complex and, as such, unable to translationally repress hb mRNA (Wharton et al., 1998). Patch-clamp analysis, after overexpression of UAS-pum G1330D in aCC/RP2 motoneurons, shows normal I Na (Fig. 4 A), implicating that Brat binding is a necessary step for the downregulation of para in these neurons. However, para mRNA quantification from whole CNS after overexpression of UAS-pum G1330D pan-neuronally suggests otherwise. This is because para mRNA levels are still downregulated when UAS-pum G1330D is overexpressed in all neurons (Fig. 4 B). This apparent dichotomy in requirement for Brat is consistent with the majority of cells in the CNS not requiring Brat for the translational repression of para, whereas Pum-dependent repression of para mRNA in motoneurons is Brat dependent. Pum repression without Brat involvement has been already reported. CycB mRNA, a gene important for germ cell proliferation, is regulated by Pum and Nanos but does not require Brat (Kadyrova et al., 2007). In conclusion, our results are suggestive of Brat being necessary for Pum function in some cell types (i.e., aCC/RP2 motoneurons), but not in other types of neurons.
Pumilio is the limiting factor in para repression
A key factor for Pum-mediated repression of hb mRNA in the Drosophila embryo is the spatial gradient of the cofactor Nanos. The posterior-to-anterior Nanos gradient defines the precise spatial zone of hb mRNA translational repression. Pum, however, is distributed homogeneously throughout the embryo (Barker et al., 1992). To test whether Nanos was also limiting for repression of para mRNA, we overexpressed UAS-nanos in aCC/RP2 motoneurons and recorded I Na. No changes in I Na were observed (Fig. 5 A). Consistent with this observation, real-time RT-PCR quantification of para mRNA, after overexpression of nanos pan-neuronally, also showed no significant difference (Fig. 5 B). Because our previous experiments demonstrate that the Pum protein is constitutively active [a mutation in pum results in an increase in para mRNA (Mee et al., 2004)], our observation that overexpression of nanos is without effect is unlikely to be attributable to lack of functional Pum. These results, together with the fact that the nanos 17 mutation alone does not affect the level of I Na (Fig. 3), suggest that Nanos is not a limiting factor in para repression. Similarly, overexpression of brat does not produce a change in I Na in aCC/RP2 (Fig. 5 A). Therefore, given that I Na in aCC/RP2 motoneurons is sensitive to Pum dosage (Fig. 2) (Mee et al., 2004), our data are consistent with Pum being a limiting factor for para mRNA translational repression in these neurons.
Pumilio downregulates nanos
To investigate further the Pum-dependent repression mechanism of para in the CNS, we examined whether Pum and Nanos could regulate one another. Such an interaction, if it exists, might be ideally suited to act as a control mechanism to safeguard neurons from excessive repression of para mRNA. Overexpression of UAS-nanos pan-neuronally does not produce any changes in pum mRNA levels (Fig. 5 B). In contrast, overexpression of UAS-pum pan-neuronally resulted in a very large and significant decrease in nanos mRNA (Fig. 5 C). The relative efficiency of Pum to downregulate nanos mRNA is approximately four times that of its ability to repress para mRNA. This regulation is possibly direct because binding of Pum to nanos mRNA has been reported (Gerber et al., 2006).
Pumilio indirectly regulates IKfast and Shal mRNA
In the motoneurons under study, activity-dependent homeostasis likely requires the coregulation of several proteins, including Pum, Para, and one or more K+ channel proteins (Baines et al., 2001; Mee et al., 2004). Here, we showed that changes in I Na current density are mediated by direct binding of Pum to para transcripts and subsequent translational repression. We therefore asked whether Pum can directly regulate transcript levels and ionic current densities of outward K+ channels.
In the first instance, we looked at the effects of increased expression of Pum on potassium currents (I K). In aCC/RP2 motoneurons I K exhibits characteristic fast (I Kfast) and slow (I Kslow) inactivating phases (Fig. 6 A) (Baines and Bate, 1998). Overexpression of full-length Pum, selectively in aCC/RP2 motoneurons, is sufficient to produce a significant decrease in the fast component of I K in these neurons (Fig. 6 B,D). In contrast, overexpression of Pum does not affect significantly the slow component (Fig. 6 C,E). In Drosophila neurons, the fast component of I K has been associated with the voltage-dependent potassium channel gene Shal (Shaker cognate I) (Tsunoda and Salkoff, 1995) and the voltage-dependent and calcium-activated potassium channel gene slowpoke (slo) (Pym et al., 2006). Therefore, we analyzed the effect of overexpression of Pum on both Shal and slo mRNA abundance. Real-time RT-PCR quantification shows significant diminution of Shal mRNA when Pum is overexpressed pan-neuronally, whereas slo mRNA levels are not affected (Fig. 6 F).
Although we demonstrated that Pum can affect I Kfast in addition to I Na in aCC/RP2, we were unable to identify a consensus Pum binding sequence in the Shal transcript. Furthermore, our microarray analysis did not identify Shal mRNA as a direct target of Pum binding (A. Gerber and S. Luschnig, unpublished data). Because of this, it is conceivable that the effect of Pum on I K may be an indirect consequence of a Pum-related reduction in I Na. This kind of compensatory mechanism has been described previously (Baines et al., 2001). To test this, we overexpressed Pum in aCC/RP2 motoneurons in a genetic background bearing a deficiency chromosome for the para locus. The para deficiency causes late embryonic lethality, and, although embryos do not hatch, aCC/RP2 motoneuron recordings can still be performed (Baines and Bate, 1998). Overexpression of pum, in the complete absence of para, results in an I K that is not significantly different than wild-type controls (Fig. 7 A). Real-time RT-PCR quantification of Shal mRNA confirms our physiology. Thus, overexpression of pum pan-neuronally, in the absence of para, results in no downregulation of Shal mRNA. On the contrary, there is a small, but significant, increase in Shal mRNA under these conditions (Fig. 7 B). We conclude, therefore, that the reduction in I Kfast and Shal observed when pum is overexpressed in wild-type backgrounds is most likely to be an indirect effect: Pum directly represses translation of para mRNA and this in turn produces a compensatory reduction of Shal mRNA abundance and I Kfast.
Identification of the molecular components that underlie homeostasis of membrane excitability in neurons remains a key challenge. Here, we show that the known translational repressor Pum binds para mRNA, which encodes the Drosophila voltage-gated Na+ channel. This observation provides a mechanistic understanding for the previously documented ability of Pum to regulate I Na and membrane excitability in Drosophila motoneurons (Mee et al., 2004). Thus, alteration in activity of Pum, in response to changing exposure to synaptic excitation, enables neurons to continually reset membrane excitability through the translational control of a voltage-gated Na+ channel.
Previous studies report several mRNAs subject to direct Pum regulation including hb (Murata and Wharton, 1995), bicoid (bcd) (Gamberi et al., 2002), CycB (Asaoka-Taguchi et al., 1999), eIF4E (Menon et al., 2004), and possibly the transcript destabilization factor smaug (smg) (Tadros et al., 2007). The majority of these identified transcripts concentrate the roles of Pum to the establishment of the embryonic anterior-posterior axis (hb and bcd) and germ-line function/oogenesis (CycB). However, in the last few years, new findings have expanded the role of Pum to encompass predicted roles in memory formation, neuron dendrite morphology, and glutamate receptor expression in muscle (Dubnau et al., 2003; Menon et al., 2004; Ye et al., 2004). Indeed, the role of Pum is likely to be very much more widespread given that Pum pull-down assays followed by microarray analysis of bound mRNAs have now identified a plethora of possible additional targets of translational regulation (Gerber et al., 2006). The ∼1000 or so genes identified are implicated to be involved in various cellular functions, suggesting that Pum-dependent translational repression might be a mechanism used in different stages of development and in diverse tissue function. To date, para is the first confirmed Pum target encoding a voltage-gated ion channel.
Pum-binding motifs have been identified in the 3′-UTRs of many mRNAs known to bind to this protein. Analysis of 113 such genes expressed in adult Drosophila ovaries has identified a consensus 8 nt binding motif [UGUAHAUA (Gerber et al., 2006)]. This sequence contains the UGUA tetranucleotide that is a defining characteristic of the NRE-like motif described in the 3′-UTR of hb mRNA (Zamore et al., 1997). We identified such an 8 nt motif within the ORF of para at the 3′ end of the transcript. Our biochemical binding data support the notion that this motif is indeed sufficient to bind Pum and as such represents the first such site to be localized to an ORF of any transcript. However, to translationally repress para mRNA, our data also show a requirement for regions of Pum in addition to the RBD. Interestingly, this kind of requirement has also been shown for another Pum target, eIF4E (Menon et al., 2004). The translational silencing of mRNAs is a complex mechanism on which only little information is available. It could involve deadenylation and degradation of the mRNA and/or the circularization of the mRNA and the recruitment of factors that would preclude translation (Chagnovich and Lehmann, 2001). The fact that different Pum targets may require only the RBD (hb) or the full-length protein (eIF4E and para) suggests that Pum-mediated translational repression may follow complex target mRNA-specific mechanisms, most probably involving the interaction of other domains of Pum with additional, so far unknown, factors. In this regard, it is interesting to note that the N terminus of Pum has regions of low complexity including prion-like domains rich in Q/R. These domains may provide a platform for other proteins that influence the fate of Pum targets.
The putative Pum binding motif that we identify lies within an exon that is common to all para splice variants identified (at least in the embryo) but is possibly subject to editing by adenosine deamination. Thus, in an analysis of splicing of para, a number of individual cDNA clones were sequenced and one splice variant was recovered that shows A-to-I editing in this motif (D. E. Wright and R. A. Baines, unpublished data). Together with a differential requirement for specific cofactors (see below), editing of this motif might serve to influence how para is affected by Pum and, as such, further increase diversity in level of expression of I Na in differing neurons or disease states (Song et al., 2004).
The known mechanism of action of Pum-dependent translational repression is absolutely dependent on additional cofactors. The most studied example, that of hb mRNA during early embryogenesis, requires the presence of both Nanos (Sonoda and Wharton, 1999) and Brat (Sonoda and Wharton, 2001). However, the requirement for these two cofactors is seemingly transcript dependent. Thus, Pum-mediated repression of CycB mRNA requires Nanos but not Brat (Sonoda and Wharton, 2001). However, Pum-dependent repression of bcd is apparently Nanos independent, because levels of Nanos in the anterior of the early embryo are undetectable (Gamberi et al., 2002). Although we clearly show that Pum-dependent repression of para mRNA in the Drosophila CNS requires Nanos, the requirement for Brat is less clear and seems to be neuronal cell type specific. A requirement for a different combination of cofactors for Pum-dependent translational regulation of a single gene transcript has not been reported previously, but clearly might represent an additional level of regulation. Such differential regulation might be required to spatially restrict the effect of Pum to certain cell types within the CNS. Voltage-gated Na+ currents are responsible for the initiation and propagation of the action potential and determine, together with other voltage-gated ion conductances, the membrane excitability of a neuron. Despite para being the sole voltage-gated sodium channel gene in Drosophila [compared with at least nine different genes in mammals (Catterall et al., 2005)], neuronal subpopulations nevertheless exhibit distinctive I Na characteristics (O'Dowd et al., 1995) (N. Muraro and R. Baines, unpublished observations). To achieve this, para is known to undergo extensive alternative splicing (Thackeray and Ganetzky, 1994; Thackeray and Ganetzky, 1995) and, additionally, RNA editing (Hanrahan et al., 2000). It is highly likely that both alternative splicing and RNA editing generate mRNAs that encode channels with differing electrophysiological properties (Song et al., 2004). It is also conceivable that these mechanisms might yield para transcripts that contain differing arrangements of Pum/Nanos binding sites, which may, or may not, recruit Brat. Indeed, it has been proposed that variations of the NRE consensus sequence may result in Pum–NRE–Nanos complexes with different topographies, resulting in altered recruitment abilities for additional cofactors such as Brat (Kadyrova et al., 2007). Additional work is necessary to clarify where, in para mRNA, the binding sites for the Pum/Nanos complex are localized and how the recruitment of Brat is facilitated in only some neurons. In the hb repression complex, Brat has been shown to interact with the cap-binding protein d4EHP (Cho et al., 2006). Therefore, additional cofactors might be necessary for Pum-dependent para repression in the Brat-independent neuronal cell subtypes that we propose here.
In contrast to translational repression of hb, our data show that Nanos is unlikely to be a limiting factor of Pum-dependent repression of para translation. Consistent with this finding is our observation that overexpression of pum is sufficient to downregulate (and probably translationally repress) nanos mRNA. However, the opposite is not true; overexpression of nanos does not affect levels of pum mRNA. These data suggest that Pum is at least a principal orchestrating factor (if not the prime factor) in regulation of para translation. Moreover, our demonstration that overexpression of pum is sufficient to greatly downregulate nanos mRNA (relative to para mRNA), together with a requirement of Nanos for Pum-dependent para mRNA repression, implicates the existence of a protective negative-feedback mechanism that prevents overrepression of para mRNA. In the absence of such feedback, it is conceivable that excessive overrepression of para mRNA might lead to neurons falling silent as their membrane excitability drops below a critical threshold. Were this to happen, then signaling in the affected neuronal circuit would be severely compromised.
We show that overexpression of full-length Pum in aCC/RP2 motoneurons not only causes a decrease in I Na but also a significant decrease in I Kfast. Additionally, pan-neuronal overexpression of Pum causes a significant decrease in Shal mRNA, a gene encoding a potassium channel known to contribute to I Kfast (Tsunoda and Salkoff, 1995). This result was surprising given that we did not identify Shal as a Pum target from our microarray analysis (A. P. Gerber and S. Luschnig, unpublished results). That this mechanism might, therefore, be indirect is corroborated by our finding that I Kfast and Shal mRNA remain at wild-type levels when Pum is overexpressed in a para-null background. It is, perhaps, counterintuitive that a reduction in I Na, to achieve a reduction in membrane excitability, should be accompanied by a similar decrease in outward I Kfast. However, changes in ionic conductances should not be considered in isolation and such a relationship might serve to maintain action potential kinetics within physiological constraints (Baines, 2003). Covariation of I Na and I K as a mechanism for changing neuronal excitability has been described in these motoneurons previously (Baines et al., 2001). Moreover, there is precedent for coupling between transcripts: injection of Shal mRNA into lobster PD (pyloric dilator) neurons results in an expected increase in I A but also an unexpected linearly correlated increase in I h, an effect that acts to preserve membrane excitability. Injection of a mutated, nonfunctional, Shal mRNA is also sufficient to increase I h indicative that this coregulation is activity independent (MacLean et al., 2003). It remains to be shown whether genetic manipulation of para mRNA levels in Drosophila motoneurons will similarly evoke compensatory changes in Shal expression.
In a previous study, it was shown that blockade of synaptic release, through pan-neuronal expression of tetanus toxin light chain, was sufficient to evoke a compensatory increase in membrane excitability in aCC/RP2 that was accompanied by increases in I Na, I Kfast, and also I Kslow (Baines et al., 2001). In contrast, we show here that overexpression of pum is sufficient to decrease I Na and I Kfast but does not significantly affect I Kslow (although there is a small nonsignificant reduction in this current). Clearly, the complete absence of synaptic input is a more severe change that likely elicits a greater compensatory change in these neurons than when Pum is overexpressed. However, whether removal of synaptic excitation also invokes additional compensatory mechanisms that act preferentially on I Kslow remains to be determined. What is consistent, however, is that change in synaptic excitation of these motoneurons is countered by Pum-dependent regulation of both para mRNA translation and magnitude of I Na.
A key question remains as to what the mechanism is that transduces changes in synaptic excitation to altered Pum activity. Perhaps the most parsimonious mechanism will be one linked to influx of extracellular Ca2+. Indeed, experimental evidence supports a role for Ca2+, because blocking its entry can preclude changes in neuronal excitability observed as a result of activity manipulation (Offord and Catterall, 1989; Desarmenien and Spitzer, 1991; Golowasch et al., 1999). In addition, changes of gene expression resulting from activity-mediated Ca2+ entry have been described both in vitro (Xiang et al., 2007) and in vivo after plasticity changes such as long-term potentiation (Miyamoto, 2006). Whether Ca2+ influx influences translation and/or transcription of Pum remains to be shown. Stimulation of mammalian neurons in culture with glutamate, after a preconditioning period of forced quiescence, results in an increase of Pum2 protein levels after just 10 min (Vessey et al., 2006). The rapidity of this response suggests that it is mediated by a posttranscriptional mechanism. We examined the role of Pum on Ca2+ channel activity. We find that neither I Ba(Ca) nor levels of the voltage-gated calcium channel coded by Dmca1A (cacophony, Calcium channel α1 subunit, type A) (Peng and Wu, 2007) are affected in aCC/RP2 motoneurons in which pum [full length (FL)] is overexpressed (data not shown). The fact that Pum does not affect Ca2+channel activity directly could reinforce the idea of its serving as a primary sensor of activity changes.
In summary, we show that Pum is able to bind to para mRNA, an effect that we previously showed to be sufficient to regulate both I Na and membrane excitability in Drosophila motoneurons (Mee et al., 2004). This mechanism requires the cofactor Nanos but does not obligatorily require Brat. Given that mammals express two Pum genes, Pum1 and Pum2 (Spassov and Jurecic, 2002), it will be of importance to determine whether this protein is also able to regulate sodium channel translation in the mammalian CNS.
This work was supported by The Wellcome Trust United Kingdom (R.A.B.) and a Career Development Award from the International Human Frontier Science Program Organization (A.P.G.). We thank Dr. Wei-Hsiang Lin for construction of a mutated Pum-binding motif. We also thank Kaushiki Menon, Bing Ye, Yuh Nung Jan, and Debora Frank for sharing fly lines.
- Correspondence should be addressed to either Nara I. Muraro or Richard A. Baines, Faculty of Life Sciences, Stopford 1.124, University of Manchester, Oxford Road, Manchester M13 9PT, UK. or