Targeting mRNAs to different functional domains within neurons is crucial to memory storage. In Aplysia sensory neurons, syntaxin mRNA accumulates at the axon hillock during long-term facilitation of sensory-motor neuron synapses produced by serotonin (5-HT). We find that the 3′ untranslated region of Aplysia syntaxin mRNA has two targeting elements, the cytosolic polyadenylation element (CPE) and stem-loop double-stranded structures that appear to interact with mRNA-binding proteins CPEB and Staufen. Blocking the interaction between these targeting elements and their RNA-binding proteins abolished both accumulation at the axon hillock and long-term facilitation. CPEB, which we previously have shown to be upregulated after stimulation with 5-HT, is required for the relocalization of syntaxin mRNA to the axon hillock from the opposite pole in the cell body of the sensory neuron during long-term facilitation, whereas Staufen is required for maintaining the accumulation of the mRNA both at the axon hillock after the treatment with 5-HT and at the opposite pole in stable, unstimulated sensory neurons. Thus, the cooperative actions of the two mRNA-binding proteins serve to direct the distribution of an mRNA encoding a key synaptic protein.
- long-term synaptic plasticity
- mRNA-binding proteins
- RNA trafficking
- double-stranded stem loop structures
- long-term memory
- cell culture
The distribution of mRNAs changes during the production of long-term synaptic plasticity (St Johnston, 1995; Knowles et al., 1996; Kiebler and DesGroseillers, 2000; Job and Eberwine, 2001; Steward and Schuman, 2001). Interfering with these changes in hippocampal pyramidal cells disrupts long-term potentiation and blocks memory storage (Steward et al., 1998; Miller et al., 2002). In Aplysia sensory neurons, the distribution of syntaxin mRNA depends on whether new synapses are forming (Hu et al., 2003). When synapses first form with target neurons, syntaxin mRNA and protein accumulate at the axon hillock. As the synapses become stable, the mRNA and protein accumulate in the cell body opposite the axon hillock. The mRNA and protein again accumulate at the axon hillock as new synapses develop during long-term facilitation of sensory-motor neuron synapses (Bailey and Chen, 1988; Glanzman et al., 1990; Hu et al., 2003; Kim et al., 2003). More syntaxin mRNA and protein are then transported into neurites presumably to supply the new synapses. Conditions that block the long-term facilitation that underlies sensitization of defensive reflexes (Kandel, 2001) prevent the accumulation of the mRNA and interfere with the transport (Hu et al., 2003).
The distribution of an mRNA depends on defined nucleotide sequences or structures (cis-acting elements) that are recognized by specific binding proteins [trans-acting (Kiebler and DesGroseillers, 2000)]. We now examine how two trans-acting proteins, the cytosolic polyadenylation element (CPE)-binding protein (CPEB) and Staufen, a double-stranded RNA-binding protein, regulate the distribution of the mRNA during long-term facilitation. The CPE was identified for its role in cytosolic polyadenylation, which activates the translation of dormant maternal mRNAs during oocyte maturation (Hake and Richter, 1994; Stebbins-Boaz et al., 1996; Walker et al., 1999). CPEB has been cloned in mollusks (Walker et al., 1999; Liu and Schwartz, 2003; Si et al., 2003) and in several vertebrates (Hake and Richter, 1994; Gebauer and Richter, 1996; Bally-Cuif et al., 1998; Welk et al., 2001). The binding protein, which is present at distal sites of neurons (Wu et al., 1998; Liu and Schwartz, 2003; Si et al., 2003), facilitates the transport of CPE-containing mRNAs in hippocampal neurons (Huang et al., 2003). Staufen, another RNA-binding protein critical for distributing mRNAs in neurons, recognizes double-stranded RNA stem-loop configurations (Ferrandon et al., 1994; Saunders and Barber, 2003). This protein is essential for Drosophila development, mediating the distribution of mRNAs in the egg and dividing neuroblasts (Schupbach and Wieschaus, 1986). In hippocampal neurons, Staufen is present in ribonucleoprotein particles (RNPs) (Kiebler et al., 1999; Tang et al., 2001; Macchi et al., 2003) that are moved along microtubules by kinesin motors (Kohrmann et al., 1999; Wickham et al., 1999; Mallardo et al., 2003; Kanai et al., 2004).
We find that disrupting the interaction between CPEB and Staufen with the 3′ untranslated region (UTR) of syntaxin mRNA abolishes the serotonin (5-HT)-induced changes in the distribution of the mRNA and also blocks long-term facilitation. Interaction between CPEB and CPE is decisive for the accumulation at the axon hillock after the treatment with 5-HT, whereas Staufen seems to be needed to maintain the accumulation at the axon hillock in sensory neurons treated with 5-HT and at the opposite pole of the cell body in stable, unstimulated sensory neurons.
Materials and Methods
Animals and tissue preparation.
Aplysia californica weighing 70–100 g were obtained from the Mariculture Resource Facility of the University of Miami (Miami, FL). Ganglia were dissected from animals anesthetized with Mg2+ (Schwartz and Swanson, 1987). RNA was isolated from tissues using the TRIzol reagent (Invitrogen, Carlsbad, CA). Reverse transcription (RT) was done with random hexamers using Superscript II reverse transcriptase (Invitrogen). Protein was measured with the Micro BCA kit (Pierce, Rockford, IL) with bovine serum albumin as a standard. Immunoblotting was done as described by Liu and Schwartz (2003).
Cell culture and electrophysiology.
Sensory neuron–L7 cocultures (consisting of one or two sensory neurons with one L7 motor neuron) were prepared as described previously (Rayport and Schacher, 1986; Schacher and Montarolo, 1991) and maintained in culture for 4–6 d. Sensory neurons were isolated from pleural ganglia dissected from mature animals (80–100 g); L7 motor cells were isolated from abdominal ganglia of juvenile animals (1–3 g). In some cultures, sensory neurons were plated alone. Standard electrophysiological techniques were used to monitor changes in the amplitude of the EPSP evoked in L7 with stimulation of each sensory neuron, before and after treatments as described by Schacher and Montarolo (1991). To produce long-term facilitation in cultures, neurons were pulsed with 5 μm 5-HT five times for 5 min each at 20 min intervals after testing the EPSP amplitude (Montarolo et al., 1986). Cultures were incubated at 18°C and EPSP amplitudes tested again 24 h later.
The sequences of degenerate primers, 5′CCNACNGTNGARYTNAAYGC and 5′GTDAYRAARTTYTTCATRTG, were based on conserved regions of human and Drosophila Staufen (GenBank accession numbers CAC14085 and AAF57752). These primers were used in RT-PCRs with a temperature profile of 94°C (30 s), 55°C (30 s), and 72°C (60 s) for 35 cycles with random hexamer-primed Aplysia neuronal cDNA as a template. A 530 bp DNA fragment was isolated that encoded a partial sequence. The full sequence was obtained by 5′ and 3′ rapid amplification of cDNA ends (RACE) performed with Aplysia sensory neuron cDNA and gene-specific primers, 5′TTGCCACGATAACGAGGATCACGAC (for 3′ RACE) and 5′CCCGTCATAATGCCGCCCAAAA (for 5′ RACE), using the SMART RACE cDNA amplification kit (Clontech, Palo Alto, CA). The entire 5′ UTR of the Aplysia Staufen gene was obtained using the First Choice RLM-RACE kit (Ambion, Austin, TX). This sequence has been deposited in GenBank (accession number AY764181).
Isolation of 3′ UTR of syntaxin mRNA.
A gene-specific primer (5′GGACACCAAAAAGGCGGTCAAGTA) was designed based on a partial sequence (GenBank accession number U03123), with only a short incomplete 3′ UTR (75 nucleotides). 3′ RACE was done using this primer with a PCR profile: 94°C (5 s) and 72°C (3 min) for 3 cycles; 94°C (5 s), 70°C (10 s), and 72°C (3 min) for 5 cycles; 94°C (5 s), 65°C (10 s), and 72°C (4 min) for 25 cycles yielding the complete syntaxin mRNA 3′ UTR (GenBank accession number AY764183) that is 982 nucleotides and has a polyadenylation signal and cytoplasmic polyadenylation element near its poly(A) tail.
We used the QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA). Plasmids containing the 3′ UTR of syntaxin mRNA were the parental vectors. A pair of primers, each complementary to opposite strands of the vector and containing the CPE deletion, was used in PCR with PfuTurbo DNA polymerase and the parental plasmid as templates. After plasmid was removed by DpnI digestion, the vector incorporating the desired mutations was transformed into Escherichia coli and purified for in vitro transcription.
In vitro transcription.
Fragments of the syntaxin mRNA coding region and UTR containing the T7 promoter were amplified with the following primer pairs and Aplysia neuronal cDNA as templates: (1) GAATTGTAATACGACTCACTATAGGGACTACCATTGAAACAATGATTTGA and TGTCCATCTTGGCCGTCTCGATGTA (for the coding region; corresponding to nucleotides 37–60 and 886–910, respectively); (2) GAATTGTAATACGACTCACTATAGGGCCCTCCTCCCTCCCTCTAGTCTACT and GATAATGATTTTATTATGACATCTTC (for the UTR; corresponding to nucleotides 1025–1049 and 1974–1999, respectively). The underlined sequence is the T7 promoter. Amplified DNA was the template for the in vitro transcription reactions using mMESSAGE mMACHINE capped transcription kit (Ambion). The resulting capped RNA was polyadenylated using a Poly(A) Tailing kit (Ambion) and purified with the MEGAclear kit (Ambion). For microinjection, the final capped and polyadenylated RNA was dissolved in buffered 0.1 m KCl at a concentration of 200 μg/ml.
A rabbit anti-Aplysia Staufen antibody was raised and affinity purified by BioSource International (Hopkinton, MA). The inferred amino acid sequence of Staufen was analyzed using the Protean program from DNAStar (Madison, WI) to predict the surface probability and antigenic index. A peptide (NH2-TSPKDSPQAPKTSGRDSV-COOH corresponding to Thr301–Val318) was conjugated to keyhole limpet hemocyanin as an immunogen. This peptide has both a high surface probability and antigenic index.
After they were fixed in 0.1 m PBS, pH 7.4, containing 4% paraformaldehyde and 30% sucrose for 1 h and rinsed with 0.01 m PBS, pH 7.4, cultured neurons were blocked at room temperature with 0.01 m PBS, pH 7.4, containing 0.2% Triton X-100 and 3% normal goat serum for 1 h and incubated with the anti-Staufen antibody (1:1000 dilution) overnight at 4°C in 0.01 m PBS, pH 7.4, 1% normal goat serum, and 0.2% Triton X-100. After they were washed three times with 0.01 m PBS for 5 min, the neurons were treated at room temperature with Cy3-labeled goat anti-rabbit antibody (1:200 dilution) for 4 h and washed three times with 0.01 m PBS before photography. Preabsorption of primary antibody with the antigen peptide, omitting the primary or secondary antibodies, was done to test the specificity of immunostaining.
In situ hybridization.
As described by Hu et al. (2002), the probes used were 5′ biotin-CTTTGCTCTGGTACTTGACCGCCTTT (for syntaxin mRNA, complementary to nucleotides 925–940) and 5′ biotin-TGGGGCCCTTGTTCACAATGCCATCAGTGCTGGCAC [for Aplysia cell adhesion protein (ApCAM); complementary to nucleotides 1078–1113]. After the cells were fixed and treated with protease K, a biotin-labeled probe (1.5 μg/ml) was applied overnight at 42°C in 50% deionized formamide, 5× SSC, 0.02% SDS, and 2% blocking reagent. The unbound probe was washed out with 2× SSC (twice) at 42°C and with 0.1× SSC (twice) at 50°C. Cells were then incubated in streptavidin–FITC (1:200 dilution; Invitrogen) for 4 h at 4°C. The unbound antibody was washed out with 0.1 m PBS (3×). Hybridization or immunocytochemical signals were visualized directly with fluorescence microscopy and imaged with a Nikon (Tokyo, Japan) Diaphot microscope attached to an SIT video camera (Dage-MTI, Michigan City, IN). The specificity of staining with the biotin-labeled antisense oligonucleotides probe was tested either by substituting a labeled sense probe for the antisense probe or by incubating with excess unlabeled antisense probe.
Staufen antisense (CA-TGGGGTTGTACGTTGAATTATGC) and control (CGTATTAAGTTGCATGTTGGGGTAC, reversed sequence of the antisense) oligonucleotides and hnRNP L antisense oligonucleotides (CTCCCAGTTGAATCATAGCAGAC), dissolved in 0.5 m KCl, 10 mm HEPES-KOH, pH 7.4, at a concentration of 0.3 mm, were injected into sensory neurons with intracellular electrodes (resistance, 10–15 MΩ) using 50 ms pulses at 5–7 psi. The sensory neurons were hyperpolarized to −70 mV. Delivery of material into the neuron was confirmed by a small depolarization of 3–5 mV with each pressure pulse. Each cell was subjected to 50 pressure pulses given at 2 s intervals. Oligonucleotides were injected on day 1. The EPSP was recorded on day 2 before the treatment with 5-HT and on day 3 24 h after the treatment.
Pleural and pedal ganglia were homogenized in buffer A (10 mm HEPES, pH 7.5, 40 mm KCl, 3 mm MgCl2, 0.32 m sucrose, and 2 mm DTT) containing protease inhibitors (1:100 dilution; Sigma, St. Louis, MO) and 1 mg/ml tRNA. The homogenates were centrifuged at 14,000 × g for 10 min at 4°C. The recovered supernatant was assayed for protein. Antibodies against Aplysia Staufen (1:50 dilution) were added to 0.5 mg of each sample and incubated for 2 h at 4°C. After protein G beads (Sigma) were added, the reaction mixture was incubated for 1 h. The beads were recovered by centrifugation, washed six times with 1× PBS containing 0.1% Triton X-100 and 0.5% NP-40, and washed twice with DEPC-treated H2O. The RNA from the immunoprecipitated samples was purified using TRIzol reagents, treated with DNase I, and subjected to RT-PCR. The primer pairs were as follows: (1) 5′GACTACATCGAGACGGCCAAGA and 5′CAGCCTAGCGTGCCACCTATC for syntaxin; (2) 5′gattacgtccctgccctttg and 5′gatcgagttcgag-cgtcttct for 18S ribosomal RNA; and (3) 5′aggatgatgtgacgccgtttg and 5′acctggtcgga-ggttgcagag for ApCAM. The last two pairs of primers have been used before successfully (Liu et al., 2004).
We used a fine microelectrode to dissect and isolate the cell bodies and their neurites (Schacher et al., 1999). Total RNA for the cell bodies and neurites was isolated using TRIzol (Invitrogen). RT was then done using random primer and Superscript II reverse transcriptase, and the resulting cDNA served as templates for PCR. Gene-specific primer sets were used to amplify specific genes.
Quantification and data analysis.
MCID software (version 7.0; Imaging Research, St. Catharines, Ontario, Canada) was used to quantify average pixel intensities for hybridized and immunostained signals from sensory neuron cell bodies (Hu et al., 2004b). Values for the overall differences in expression of mRNA or protein were determined by averaging pixel intensities in six areas of sensory neuron cytoplasm (25 μm2 each) positioned at 12:00, 2:00, 4:00, 6:00, 8:00, and 10:00 with respect to the center of the cell (Hu et al., 2003). For measuring differences in distribution, the middle of the axon hillock region is at the 6:00 position and staining intensity for each region of the cytoplasm is normalized to the staining intensity measured at the 12:00 position (Hu et al., 2003). Data are expressed as means ± SEM. Changes in EPSP amplitude were normalized to the initial amplitude for each culture. The effects of treatments were determined by ANOVA, and specific differences between treatments were determined by the Scheffe multi-comparison tests.
The 3′ UTR of syntaxin mRNA contains putative binding sites for CPEB and Staufen
Unlike several other mRNAs examined (Hu et al., 2002, 2003), Aplysia syntaxin mRNA accumulates in the region opposite the axon hillock in sensory neurons with stable synapses; this mRNA moves to the axon hillock during the formation of new synapses that occurs in long-term facilitation (Hu et al., 2003). Because the distribution of an mRNA can be mediated by targeting elements in its 3′ UTR (Ainger et al., 1993; Blichenberg et al., 1999; Zhang et al., 2001), we cloned the full-length 982-nucleotide-long 3′ UTR of syntaxin mRNA. The sequence contains a consensus CPE (nucleotides 1690–1695) and a polyadenylation signal, AAUAAA (nucleotides 1985–1990), just before the start of the poly(A) tail (GenBank accession number AY764183) (Fig. 1).
As predicted by the GeneQuest program (Lasergene), the cloned 3′ UTR can be folded into extensive stem-loop double-stranded configurations needed for binding Staufen (St Johnston et al., 1992; Saunders and Barber, 2003) (see supplemental Fig. 1, available at www.jneurosci.org as supplemental material). Because Staufen has been shown to be an important mRNA-binding protein in neurons (Kiebler et al., 1999; Tang et al., 2001; Macchi et al., 2003, 2004; Mallardo et al., 2003), we next cloned a Staufen gene from Aplysia nervous tissue (GenBank accession number AY764181), which is 6.2 kb long with an open reading frame (ORF) encoding 950 amino acid residues. The predicted protein, with a calculated molecular weight of 106,000 (Fig. 2A), is similar in size to its Drosophila homolog and has four regions with double-stranded RNA binding domains and a long N terminus. A small fifth region is also present at the C terminus (amino acids 896–916). Two mammalian homologs of Staufen (Stau1 and Stau2) have been described previously (Marion et al., 1999; Wickham et al., 1999; Macchi et al., 2004). The Aplysia protein is closer to Stau2, the brain-specific form. To test whether Aplysia Staufen interacts with syntaxin mRNA, we raised an antipeptide antibody that specifically recognizes a single Mr 125,000 component on immunoblotting (Fig. 2B), which is consistent with the calculated molecular weight. Immunoblot assays showed that Staufen is strongly expressed in nervous tissue (Fig. 2C). We then used a pull-down analysis with RT-PCR to determine whether Staufen interacts with mRNA. Staufen was immunoprecipitated using the antibody, and RNA was then isolated from the immunocomplex and treated with DNase I. Subsequent RT-PCR revealed that syntaxin mRNA and 18S ribosomal RNA is pulled down with Staufen. The coprecipitation is selective, because the neuron-specific mRNA that encodes ApCAM (Mayford et al., 1992) was not detected in the precipitate (Fig. 2D). The mRNA of ApCAM, like that of syntaxin, is concentrated at the axon hillock during long-term facilitation (Hu et al., 2002). The presence of ribosomal RNA in the precipitate suggests that syntaxin mRNA might be a component of an RNP (Mallardo et al., 2003).
The 3′ UTR regulates the distribution of syntaxin mRNA
Because the 3′ UTR of syntaxin mRNA contains elements that may bind both CPEB and Staufen, we injected excess amounts of the in vitro-synthesized 3′ UTR into the cell body of sensory neurons with the aim of influencing the distribution of the mRNA. We reasoned that the injected RNA would prevent the interaction between the two binding proteins and syntaxin mRNA in the sensory neuron. A similar strategy has been used successfully to disrupt the interaction between binding proteins and their nucleic acid-binding elements (Dash et al., 1990; Ferrandon et al., 1994). We found that the injected 3′ UTR disrupted the specific accumulation of syntaxin mRNA in both untreated sensory neurons (n = 6; opposite the axon hillock) and in neurons treated with 5-HT (n = 6; at the axon hillock) (Fig. 3A). Under both conditions, syntaxin mRNA became uniformly distributed throughout the cytoplasm (Fig. 3A, histogram). Control injections of the coding region (ORF; n = 6) did not affect the distribution of the mRNA in unstimulated cells (n = 6; opposite the axon hillock) or in cells treated with 5-HT (n = 6; at the axon hillock). Both types of injections did not change the overall amounts of syntaxin mRNA in sensory neuron cell bodies. These results suggest that the 3′ UTR contains cis-acting elements that both maintain the accumulations of the mRNA in unstimulated sensory neurons and the localization of the mRNA to the axon hillock after treatment with 5-HT.
Injecting the 3′ UTR also blocked long-term facilitation (Fig. 3B). The change in EPSP amplitude (a measure of synaptic strength) after the treatment with 5-HT in cells injected with the 3′ UTR was 15.9 ± 6.4% (n = 9) compared with 78.9 ± 19.8% when 5-HT was applied to the cells injected with the control ORF sequence (n = 9). Injection of the 3′ UTR or ORF (UTR: 5.0 ± 7.9%, n = 6; ORF: 4.3 ± 4.9%, n = 6) did not affect baseline synaptic transmission. These observations suggest that the interactions between the binding proteins and cis-acting elements in the 3′ UTR of syntaxin mRNA is essential for long-term facilitation.
The 3′ UTR contains at least two cis-acting elements
To examine the roles of specific targeting elements in the 3′ UTR of syntaxin mRNA in the distribution of the mRNA in stable as well as in 5-HT-treated neurons, we injected excess amounts of the CPE RNA (a 12-mer, UUGUUUUAUCCA) into sensory neuron cell bodies with the expectation that this would block the interaction between CPEB and the CPE in the 3′ UTR of syntaxin mRNA. As with the injections of the entire 3′ UTR, injecting the CPE sequence blocked the accumulation of the mRNA at the axon hillock after the treatment with 5-HT (n = 9), whereas injecting control oligonucleotide (reverse sequence of the CPE RNA; n = 9) did not interfere with the accumulation (Fig. 4A). In contrast to the results obtained with the 3′ UTR, injecting the CPE (n = 6) did not interfere with the accumulation of syntaxin mRNA in stable untreated cells; the mRNA remains situated opposite the axon hillock (Fig. 4A). The effect of injecting the CPE on the accumulation of the mRNA at the axon hillock is specific, because the distribution of the mRNA encoding the neuropeptide sensorin, which is also concentrated at the axon hillock (Hu et al., 2002, 2003), was not affected (Fig. 4A). Injecting the CPE also blocked long-term facilitation (a change in EPSP amplitude of −1.3 ± 4.6%, n = 10, compared with 52.6 ± 6.1%, n = 12) but did not affect baseline synaptic transmission (a change in EPSP amplitude of 0.3 ± 5.4%, n = 9, compared with 8.3 ± 11.1%, n = 6 for neurons injected with control RNA) (Fig. 4B).
We next examined whether blocking the synthesis of CPEB protein affects the distribution of syntaxin mRNA. Injecting antisense oligonucleotides complementary to Aplysia CPEB mRNA (Si et al., 2003) abolished the accumulation of syntaxin mRNA at the axon hillock (Fig. 5A) and also blocked the long-term facilitation induced by 5-HT (Fig. 5B). It had no effect on the distribution of syntaxin mRNA in stable untreated cells, however (Fig. 5A). Thus, interfering with the interaction of CPEB with its targeting element (CPE) blocked both the accumulation of syntaxin mRNA at the axon hillock and long-term facilitation but failed to change the accumulation opposite the axon hillock in untreated neurons with stable synapses.
To distinguish the role of the CPEB from that of Staufen in regulating the distribution of syntaxin mRNA, we injected an altered 3′ UTR with the CPE deleted. Our rationale was that injecting excess amounts of the altered 3′ UTR would continue to interfere with the interaction of Staufen with the 3′ UTR. Both complete (Fig. 3A) and altered 3′ UTR sequences disrupted the accumulation opposite the axon hillock in stable untreated cells; the mRNA was distributed uniformly throughout the cell body (Fig. 6A). After injection of the altered 3′ UTR, treatment with 5-HT produced effects that are qualitatively different from those obtained with injections of the complete 3′ UTR. Although the complete 3′ UTR blocked both long-term facilitation (change in EPSP amplitude of only 0.9 ± 5.4%; n = 9) (Fig. 3) and the accumulation at the axon hillock (Fig. 6A), the altered 3′ UTR only reduced long-term facilitation moderately (an increase in EPSP amplitude of 37.5 ± 4.5%; n = 8) compared with an increase of 78.9 ± 19.8% (n = 9) after control injections (Fig. 3), and moderately reduced the accumulation of syntaxin mRNA at the axon hillock after 5-HT (only 25% increase over controls in Fig. 6A compared with 50% increase over controls in Fig. 3A). These results suggest that targeting elements other than the CPE (presumably Staufen) maintain the distribution of the mRNA opposite the axon hillock in cells with stable synapses and anchor the 5-HT-induced accumulation at the axon hillock. In contrast, the CPE is crucial for the relocalization of syntaxin mRNA to the axon hillock.
Staufen maintains syntaxin mRNA accumulations
The distribution of Staufen protein is congruent with that of syntaxin mRNA: in untreated stable sensory neurons, both Staufen and syntaxin mRNA are located together in the region of the cell body opposite the axon hillock (Fig. 7A); after the treatment with 5-HT, both Staufen protein and syntaxin mRNA are expressed at their highest levels at the axon hillock (Fig. 7A). The change in distribution does not require a net increase in Staufen protein because the overall immunostaining intensity does not change significantly after the treatment with 5-HT. Immunocytochemistry also shows that Aplysia Staufen is present both in the cell body and neurites of sensory neurons (Fig. 7B). Single-cell RT-PCR analysis reveals that syntaxin mRNA is also present in neurites (Fig. 7C). This colocalization of protein with mRNA and the binding data shown in Figure 2D suggest that Staufen interacts with the mRNA and might regulate the localization of the mRNA.
Injecting antisense oligonucleotides complementary to the Staufen gene alters the distribution of syntaxin mRNA in both stable untreated (Fig. 8A) and 5-HT-treated sensory neurons (Fig. 8B); the mRNA was distributed uniformly throughout the cell body. Control injections had no effect (Fig. 8 A,B). Injecting antisense oligonucleotides reduced the amount of Staufen protein in sensory neurons by 25%: 48 h after the injection and 24 h after treatment with 5-HT, the average intensity with control (reverse sequence of the antisense) was 54.1 ± 3.4 U and with antisense was 40.5 ± 3.5 U. This reduction in staining was significant (df = 1, 16, F = 7.751, p = 0.015). The effect of injecting Staufen antisense oligonucleotides is specific; it did not affect the distribution of ApCAM mRNA (Fig. 8C), an mRNA that does not interact with Staufen (Fig. 2). Furthermore, injection of antisense oligonucleotides complementary to the gene of another Aplysia RNA-binding protein, hnRNP L (GenBank accession number AY764182), had no effect on the distribution of syntaxin mRNA (Fig. 8D). Thus, the constitutive expression of Staufen is essential for the localization of the syntaxin mRNA in both stable neurons (accumulation opposite the axon hillock) and in 5-HT-treated sensory neurons (accumulation at the axon hillock).
Blocking synthesis of Staufen protein also interferes with long-term facilitation. Injection of antisense oligonucleotides abolished long-term facilitation produced by 5-HT (the EPSP amplitude was reduced to 9.1 ± 6.4%; n = 8) (Fig. 8E). Injection of control oligonucleotides had no effect on long-term facilitation (EPSP amplitudes increased by 71.3 ± 15.8%; n = 8). Both control and antisense oligonucleotides did not significantly effect baseline synaptic transmission (Fig. 8E). The injection of antisense oligonucleotides complementary to the mRNA-binding protein hnRNP L also had no effect on the facilitation (data not shown). Thus, Staufen is essential for syntaxin mRNA accumulations and for the long-term facilitation produced by 5-HT.
Syntaxin is important for synaptic function, participating in transmitter release (Chen and Scheller, 2001), vesicle fusion (Jahn et al., 2003), and the generation and maintenance of neuronal polarity (Foletti et al., 1999; Horton and Ehlers, 2003). We find that two RNA-binding proteins, CPEB and Staufen, influence the distribution of syntaxin mRNA in Aplysia sensory neurons. Accumulation of the mRNA at the axon hillock leads to increases in the amounts of syntaxin protein at the axon hillock for transport into the axon (Hu et al., 2003), where it can contribute to axon growth and synaptic function (Tang et al., 2001; Horton and Ehlers, 2003). Because axon growth and the formation of new synapses occur both early during the initial interaction between the sensory neuron with a target motor neuron and later during the formation of long-term facilitation of stable sensory-motor neuron connections (Bailey and Chen, 1988; Glanzman et al., 1989, 1990; Kim et al., 2003), increased syntaxin in the axon would be required for the formation, maintenance, and functioning of the new synapses (Hu et al., 2003; Kim et al., 2003; Grabham et al., 2005). Disrupting the accumulation of syntaxin mRNA at the axon hillock by interfering with specific interactions between cis-acting elements and trans-acting RNA-binding proteins blocked long-term facilitation (Figs. 3⇑⇑⇑⇑⇑–8), indicating that the accumulation is important for the development of long-term facilitation.
Synthesis and accumulation of syntaxin protein correspond with the accumulation of its mRNA under all conditions (Hu et al., 2003). The mRNA is present in neurites of cells in culture, suggesting that synthesis of the protein may also take place at distal neurites. The axon hillock may be a docking site for the incorporation of the mRNA into RNPs for transport to synapses. Staufen antibodies precipitate syntaxin mRNA and 18S ribosomal RNA together with Staufen protein, suggesting that the mRNA, Staufen, and ribosomes are all components of the same RNPs. CPEB is also present in neurites (Liu and Schwartz, 2003; Si et al., 2003) and may be needed to move the mRNA to distal sites where local translation of the mRNA would also contribute to synaptogenesis and synaptic function (Casadio et al., 1999; Schacher et al., 1999; Sun et al., 2001; Giustetto et al., 2003; Hu et al., 2003).
RNA interaction: do the trans-acting proteins have distinctive functions?
Staufen plays a critical role in accumulating syntaxin mRNA at specific sites: opposite the axon hillock in stable untreated sensory neurons or at the axon hillock in 5-HT-treated neurons. During oogenesis, Drosophila Staufen is required for the accumulation of oskar mRNA at the posterior pole of the egg (St Johnston et al., 1991) and later is required for the accumulation of bicoid mRNA at the anterior pole (St Johnston et al., 1991; Ferrandon et al., 1994; Cha et al., 2001). In Aplysia, Staufen is distributed together with syntaxin mRNA: opposite the axon hillock in stable untreated sensory neurons and at the axon hillock in 5-HT-treated sensory neurons. Manipulations that disrupt the interaction of Staufen with syntaxin mRNA in Aplysia sensory neurons, intracellular injections of either Staufen antisense oligonucleotides or the 3′ UTR of syntaxin mRNA, disrupt the accumulation of the mRNA under all conditions. We found that reducing Staufen levels by 25% is sufficient to interfere with the distribution of syntaxin mRNA and long-term facilitation. Although small, this decrease is similar in magnitude to the reduction in the regulatory subunit of protein kinase A (PKA) that is required for long-term facilitation (Hegde et al., 1993). In addition to syntaxin mRNA, Staufen may also interact with other mRNAs, the distribution of which is important for changing synaptic function. We suggest that Staufen contributes to the formation of long-term synaptic plasticity by maintaining mRNAs at specific sites in the neurons. In accordance with the idea that this RNA-binding protein plays a role in the development of synaptic plasticity, Dubnau et al. (2003) found that Drosophila mutants of Staufen are unable to learn.
The RNA-binding protein CPEB has been shown not only to mediate the polyadenylation and translation of certain mRNAs critical for long-term synaptic plasticity but also to transport select mRNAs (Huang et al., 2003; Liu and Schwartz, 2003; Si et al., 2003). In cultured hippocampal neurons, the CPE directs the movement of mRNAs in a microtubule-dependent manner (Huang et al., 2003). In Aplysia, the CPE in the 3′ UTR of syntaxin mRNA is critical for accumulation of syntaxin mRNA at the axon hillock during the development of long-term facilitation. Injection of CPE sequences or blocking the synthesis of CPEB protein with antisense oligonucleotides (Si et al., 2003) abolished both long-term facilitation and the accumulation of the mRNA at the axon hillock after the treatment with 5-HT. The effect of CPE injection was specific because sensorin mRNA accumulated at the axon hillock both in untreated cells and after 5-HT. These manipulations, however, did not affect the accumulation of syntaxin mRNA opposite the axon hillock in untreated cells. When the interaction with Staufen is interfered with following injections of the 3′ UTR lacking the CPE or blocking new synthesis of Staufen, syntaxin mRNA does not accumulate at either pole of the neuron but is distributed uniformly. The importance of the CPE in the accumulation of the mRNA at the axon hillock only after treatment with 5-HT is also supported by the observation that both the accumulation and the plasticity are attenuated but not abolished when sensory neurons are injected with the 3′ UTR lacking the CPE. Under these conditions, the interaction with CPEB may be sufficient to move the uniformly distributed mRNA to the axon hillock. Alternatively, injection with the 3′ UTR lacking the CPE may not affect the accumulation of syntaxin mRNA at the axon hillock to the same degree as reducing Staufen levels by injections with antisense or the full-length 3′ UTR. This reduction may not be sufficient to interfere with some accumulation at the axon hillock after CPEB-mediated movement of the mRNA. We found that movement from the opposite pole to the axon hillock is dependent on intact microtubules, because nocodazole blocked the 5-HT-induced accumulation of syntaxin mRNA at the axon hillock (data not shown).
How do the actions of 5-HT lead to relocalization of syntaxin mRNA from the opposite pole to the axon hillock? First, interaction between Staufen and syntaxin mRNA at the opposite pole of the neuron may be reduced through PKA and p42/44 mitogen-activated protein kinase (MAPK) catalyzed phosphorylations activated by 5-HT (Greenberg et al., 1987; Martin et al., 1997; Hu et al., 2004a; Sharma and Carew, 2004). Hu et al. (2003) found that the change in the distribution of syntaxin mRNA after treatment with 5-HT is PKA dependent. Moreover, the activation of MAPK by 5-HT (Martin et al., 1997; Purcell et al., 2003) is mediated by the secretion of the sensory neuron-specific neuropeptide sensorin, and the secretion of this peptide also affects the distribution of syntaxin mRNA (Hu et al., 2004a,b). In Xenopus oocytes, the transient phosphorylation of Staufen by MAPK during meiotic maturation may be important for the release of RNAs from their accumulation at the vegetal cortex (Allison et al., 2004). Aplysia Staufen has several possible PKA and MAPK phosphorylation sites (Fig. 2). Phosphorylation by PKA or MAPK after the treatment with 5-HT may induce conformational changes in Staufen, resulting in the release of syntaxin mRNA for interactions with newly synthesized CPEB. The mRNA would be directed to the axon hillock by CPEB, induced by the treatment with 5-HT (Liu and Schwartz, 2003; Si et al., 2003) in a microtubule-dependent manner. At the axon hillock, syntaxin mRNA would interact with newly synthesized Staufen, anchoring the mRNA at this new location for potential transport to distal sites.
This work was supported by National Institute of Mental Health Grants MH15174 (J.L.) and MH60387 (S.S.) and National Institute of Neurological Disorders and Stroke Grants MH048850 and NS29255 (J.H.S.). Animals were provided by the National Center for Research Resources, National Resource for Aplysia at the University of Miami under National Institutes of Health Grant RR1029.
- Correspondence should be addressed to Dr. Schacher at the above address. Email: