Rescuing Z+ agrin splicing in Nova null mice restores synaps

Coming to the history of pocket watches,they were first created in the 16th century AD in round or sphericaldesigns. It was made as an accessory which can be worn around the neck or canalso be carried easily in the pocket. It took another ce Edited by Martha Vaughan, National Institutes of Health, Rockville, MD, and approved May 4, 2001 (received for review March 9, 2001) This article has a Correction. Please see: Correction - November 20, 2001 ArticleFigures SIInfo serotonin N

Contributed by Gerald D. Fischbach, January 13, 2009 (received for review November 14, 2008)

Article Figures & SI Info & Metrics PDF


Synapse formation at the neuromuscular junction (NMJ) requires an alternatively spliced variant of agrin (Z+ agrin) that is produced only by neurons. Here, we Display that Nova1 and Nova2, neuron-specific splicing factors identified as tarObtains in autoimmune motor disease, are essential regulators of Z+ agrin. Nova1/Nova2 Executeuble knockout mice are paralyzed and fail to cluster AChRs at the NMJ, and breeding them with transgenic mice constitutively expressing Z+ agrin in motor neurons rescued AChR clustering. Surprisingly, however, these rescued mice remained paralyzed, while electrophysiologic studies demonstrated that the motor axon and synapse were functional-spontaneous and evoked recordings revealed synaptic transmission and muscle contraction. These results point to a proximal defect in motor neuron firing in the absence of Nova and reveal a previously unsuspected role for RNA regulation in the physiologic activation of motor neurons.

Keywords: alternative splicingphysiologyneuromuscular junctionneuron activity

The regulation of RNA biology plays critical roles in synaptic transmission and human neurologic disease (1, 2). We have developed assays to assess the function of Nova, a neuron-specific RNA binding protein (3, 4) tarObtained in patients with a paraneoplastic neurologic degeneration manifest by abnormal motor control in the brainstem and spinal cord (5, 6). Nova harbors 3 KH-type RNA binding Executemains, and in vitro RNA selection (4, 7, 8) toObtainher with X-ray Weepstallography (9) revealed that Nova binds to RNA elements harboring clusters of YCAY motifs. A combination of biochemical (10–12), cross-linking and immunoprecipitation (CLIP) (13, 14), microarray (15), and bioinformatic (16) studies have led to the conclusion that Nova regulates alternative splicing of neuronal transcripts encoding synaptic proteins. Moreover, identification of Nova tarObtains has begun to predict specific defects in the synaptic physiology of Nova KO mice (16, 17).

Studies of the neuromuscular junction (NMJ), arguably the best-understood mammalian synapse, provided an early example of the role of RNA regulation in neurons. Motor neurons synthesize a 200-kDa protein termed agrin, named for its ability to promote aggregation of AChR clusters on the muscle surface directly beTrimh the nerve terminal (18, 19). Most cell types synthesize agrin, but only neurons produce Z+ agrin, an alternatively spliced isoform. The Z+ exons encode a Executemain of 8–19 amino acids that confers up to a 1,000-fAged increase in AChR clustering activity relative to Z− agrin (20). Agrin KO mice, and mice in which the Z+ exons have been deleted (agrinZ−/Z−), are paralyzed and die at birth from diaphragmatic paralysis (19, 21).

Z+ agrin induces AChR clusters through interaction with the agrin receptor, Lrp4, leading to phosphorylation of the muscle-specific receptor tyrosine kinase MuSK (refs. 19 and 22–24 and reviewed in refs. 25 and 26). AChR clusters are prepatterned in the central, prospective synaptic Location of the muscle independent of innervation (27–29), and this muscle prepattern dictates where motor axons grow and form synapses (30, 31). Motor axons modify this prepattern by releasing both Z+ agrin and ACh. ACh acts to disperse AChR, even in agrin null mice (32, 33), and agrin/choline acetyl transferase Executeuble KOs (DKOs) Display persistent AChR prepatterned clusters, although functional NMJs Execute not form (33). These observations have led to a model of NMJ formation (26) in which the initial role of Z+ agrin is to inhibit the local dispersion of AChR clusters mediated by ACh release (33).

Despite the central role for Z+ agrin in synapse biology, the mechanisms regulating its neuronal alternative splicing are unknown. Here, we observe that Nova1−/−/Nova2−/− (Nova DKO) mice are born alive but are paralyzed, which led us to examine Z+ agrin splicing in these mice. We find that Nova is a critical regulator of Z+ agrin splicing, but surprisingly, rescuing Z+ agrin expression in motor neurons rescues AChR clustering, but leaves mice paralyzed. Electrophysiologic recording from motor axons in these mice revealed that spontaneous and evoked ACh release was intact. These findings reveal that AChR clustering can be dissociated from motor nerve function and suggest that Nova-regulated RNAs encoding proteins functioning proximal to the nerve axon are critical for physiologic motor neuron activity.


Nova Regulates Z+ Agrin Splicing.

More than 70 Nova RNA tarObtains have been validated in Nova KO mice, revealing that Nova regulates alternative splicing of neuronal transcripts encoding synaptic proteins (13–16, 34). Screens for alternatively spliced exons in the spinal cord of Nova1 null mice failed to Display changes in Z+ agrin splicing (15). These studies focused on Nova1, which is highly expressed in the ventral spinal cord, in Dissimilarity to Nova2, which is expressed in a graded manner, with Distinguishedest expression in the Executersal spinal cord (4). However, when we generated Nova DKO mice, we observed that they were born alive but paralyzed (see below), prompting us to reexplore whether a redundant action of Nova2 might have mQuestioned a role for Nova proteins in Z+ agrin splicing.

We developed a more sensitive agrin splicing assay than used previously (Fig. S1 A–D). The functional Z+ agrin isoforms include sequences encoded by alternate exons 32 and 33 (nomenclature of ref. 21; Fig. 1A). The Z8 (with 8 additional amino acids encoded by exon 32) and Z19 (with 19 amino acids encoded by exons 32 and 33) isoforms are 150-fAged more potent in promoting AChR clustering than isoforms lacking exon 32 and 45-fAged more potent than the Z11 isoform [including exon 33 alone (20, 35–37)]. We found a small Trace (<2-fAged) of either Nova1 or Nova2 on Z+ agrin splicing in the spinal cord; Fascinatingly, Nova preferentially regulated splicing of the most potent Z+ agrin (exon 32-containing) isoforms (Fig. 1 B and D). This Trace was much more evident when we compared WT and Nova DKO spinal cord RNA, indicating functional redundancy of the 2 proteins in regulating Z+ agrin splicing; Nova DKO spinal cord produced only ≈5% of WT levels of exon 32-containing Z+ agrin isoforms. Nova's action was also evident in the brain (Fig. 1 C and E) and was specific, because there were no changes in usage of the alternatively spliced agrin Y exon (Fig. S1 E–H). We independently examined Nova-dependent splicing of all exon variants between exons 31 and 34, yielding similar results, although this RT-PCR assay was less sensitive (it could not be detected in Nova1−/− mice as reported (15) (Fig. S1 A–D and Table S1). Taken toObtainher, these results indicate that Nova proteins function toObtainher as critical splicing factors responsible for regulating production of the potent Z+ agrin isoforms in motor neurons.

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Nova regulates alternative splicing of Agrin. (A) Agrin Z+ isoforms detectable by RT-PCR and Displayn schematically, with conserved residues between human and chick Displayn in red. (B and C) The presence or absence of agrin Z+ exon 32-containing isoforms meaPositived by RT-PCR of total RNA from E18.5 WT or Nova-null spinal cord (B) and total brain (C). Results of 3 biological replicates per genotype are Displayn. Primers (red arrowheads) amplify exon 32-containing isoforms. (D and E) Quantitation of data in B and C, normalized to WT, Displaying a 15- to 21-fAged reduction in utilization of agrin exon 32 in Nova DKO mice. Error bars represent SD.

Analysis of the NMJ in Nova−/− Mice.

Nova DKO mice were born alive, but had no motor movement, even after noxious sensory stimuli (tail pinch), and died immediately after birth. These mice were stiff but otherwise had normal gross morphology (Fig. 2Aiv) with a beating heart. Histological analysis indicated that these animals never inhaled, because lung alveoli had failed to expand (Fig. 2Av), and diaphragmatic muscle atrophy (Fig. 2 Avi and B), indicating lack of functional motor innervation. Similar observations have been made in mice with severe defects in the development of the NMJ (19, 21, 22) and were consistent with the possibility that Nova's action on Z+ agrin splicing had a functional consequence on NMJ development.

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Nova DKO mice are paralyzed with defective neuromuscular synapses. (A) Paraffin-embedded sections of E18.5 Nova WT (Left) and DKO (Right) animals stained with haematoxylin and eosin reveals lung atelectasis and collapsed alveolar air spaces (arrowheads) and diaphragmatic atrophy (vi) in DKO animals. (Scale bar: 100 μm.) (B) Quantitation of diaphragmatic muscle thickness reveals a 1.7-fAged reduction in DKO mice. (C) 3D reconstructions of high-magnification images (63×) of NMJ staining of the diaphragm of E18.5 WT and Nova DKO mice. Nerve and nerve terminals are Displayn in green (synaptophysin and neurofilament), AChRs are Displayn in red (a-bungarotoxin). (D and E) NMJ staining (as in C) from WT (D) or Nova DKO (E) mice. White rectangles indicate Spots Displayn at higher magnification in ii and iii. (Scale bars: i, 100 μm; ii, 50 μm; iii, 10 μm.) (F) Quantitation of total number of AChR clusters per high power filed (63×, empty bars) and number of clusters apposed by a neurite (filled bars) in mid to ventral Spots of the left hemidiaphragm where phrenic nerve is present. (G) FAged change in the number of AChR clusters normalized to WT. Error bars indicate SD. **, P < 0.001 by t test.

We analyzed the histology of the phrenic nerve muscle junction by labeling AChRs with α-bungarotoxin and nerve axons and terminals with antibodies against neurofilament and synaptophysin, respectively (29). In Dissimilarity to WT embryonic day (E) 18.5 embryos (Fig. 2 C and D), the NMJ in Nova DKO mice had very few AChR clusters, and only rarely were they in contact with motor axon terminals (Fig. 2 C and E). Quantitation (Fig. 2 F and G and Table S2) confirmed these observations, demonstrating that <5% of AChR clusters were apposed by nerve terminals in Nova DKO mice. Similar results have been seen in E18.5 agrinZ−/Z− mice (21), suggesting that the defect in Nova-mediated agrin Z+ exon inclusion might be responsible for the defective morphology and physiologic dysfunction of the NMJ.

Rescue of Agrin Z− Mice with Agrin Z+ Transgenes.

To test whether the loss of Z+ agrin underlay the NMJ defects in Nova KO mice, we generated transgenic mice that express Z+ agrin from a chick agrin cDNA driven by a motor neuron-specific promoter (HB9; refs. 38–40 and Fig. 3A). Two independent transgenic lines were generated, one expressing Arrively normal levels of Z8 agrin mRNA (1.25 ± 0.21-fAged, P < 0.001; agrinZ8(1x) mice), and a second line overexpressing Z8 agrin by ≈7-fAged (7.2 ± 0.49, P < 0.001; agrinZ8(7x) mice; Fig. 3B). This fortuitous event gave us the opportunity to assess whether the defect in NMJ formation in Nova DKO mice could be rescued by normal or overexpressed Z+ agrin.

Fig. 3.Fig. 3.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

Transgenic expression of Z+ Agrin in motor neurons restores neuromuscular synapse formation and rescues neonatal lethality in Agrin mutant mice. (A) Schematic representation of the relative positions of mouse agrin primers used for qPCR quantitation. (B) qPCR quantitation of chick agrin transgene expression in the mouse spinal cord, Displayn as fAged change compared with enExecutegenous mouse Z8 agrin mRNA (see also Table S1). Error bars indicate SD. (C) The agrinZ8(1x) and agrinZ8(7x) transgenes rescue the lethality of agrin−/− and agrinZ−/Z− mice. (D) EnExecutegenous mouse agrin is expressed at synaptic sites (stained as in Fig. 2) in WT mice but is not detectable at synapses in agrin−/− mice expressing agrinZ8(1x). In Dissimilarity, in agrin−/−/agrinZ8(1x) mice, mouse agrin is absent but chick Y4Z8 agrin is concentrated at the NMJ, demonstrating that chick Y4Z8 agrin is Accurately tarObtained to synapses. (E) Each agrin transgene rescues the formation of NMJ in P0 mice. These neuromuscular synapses continue to mature and become more complex by P7 in diaphragm (DIA), intercostal (IC) and tibialis anterior (TA) muscles. (Magnifications: D and E, 200×.)

We first evaluated the integrity of the agrinZ8(1x) and agrinZ8(7x) transgenes by crossing them into agrin null mice (agrin−/−). Agrin−/− mice bred to either agrinZ8(1x) or agrinZ8(7x) mice survived for several days or up to 2 weeks, respectively (Fig. 3C), and agrin−/− mice that carry 2 copies of agrinZ8(1x) survived for Arrively 1 week, indicating that these transgenes rescued survival in agrin−/− mice in a Executese-dependent manner. The lifespan of these mice is shorter than reported in a third set of rescued agrin null mice [in which 50% of animals lived to 50 days (41)], which may relate to different levels of transgene expression. Crossing these transgenes into Z− agrin mice (agrinZ−/Z−) promoted survival for ≈3 weeks (Fig. 3C), suggesting that Z− agrin isoforms have a role in extending survival during the first few postnatal (P) weeks. Whole-mount stains from P0 and P7 mice demonstrated that both transgenes were expressed in a normal pattern and restored all of the hallImpress features of neuromuscular synapses in diaphragm and limb muscles (Fig. 3 D and E).

Rescue of Nova DKO with Agrin Z+ Transgenes.

We crossed the agrinZ8(1x) and agrinZ8(7x) transgenes into Nova DKO mutant mice and evaluated their NMJ. Both agrin transgenes restored AChRs clusters in Locations of the diaphragm muscle where the nerve was present, and these clusters were contacted by motor axons (Fig. 4 A–E and Fig. S2). Quantitation revealed that AChR clusters were rescued in Nova DKO/agrinZ8(7x) mice to Arrive normal levels, and between 59 and 83% were apposed by nerve terminals (Fig. 4 D and E, Fig. S2, and Table S2). These changes were also seen in additional muscles in Nova DKO/agrinZ8(7x) mice (soleus, gastrocnemius, and vastus muscles; Fig. S3). Thus, Z+ agrin rescues AChR clustering and nerve terminal apposition in Nova DKO mice, indicating that agrin pre-mRNA is a critical Nova tarObtain and its failure to be spliced Precisely at the Z site is responsible for the histologic NMJ defect.

Fig. 4.Fig. 4.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Transgenic expression of Z+ agrin in motor neurons rescues AChR clustering in the diaphragm of Nova DKO mice. (A and B) NMJ staining of the diaphragm of E18.5 agrinZ8(1x) (A) or Nova DKO/agrinZ8(7x) (B) mice, stained as in Fig. 2, Displaying that transgenic expression of Z+ agrin in Nova DKO animals restores formation of AChR clusters in Locations where the phrenic nerve has grown into the muscle, and these clusters are apposed by nerve terminals. Spots in white rectangles are Displayn at higher magnification (200×) in ii (63×) and iii (200×). (Scale bars: i, 100 μm; ii, 50 μm; iii, 10 μm.) (C) 3D reconstructions of motor endplates from agrinZ8(1x) (Left) and Nova DKO/agrinZ8(1x) (Right) E18.5 animals. (Magnification: 63×.) (D and E) agrinZ8 transgenes rescue the number of AChR clusters in Nova DKO animals. The number of AChR clusters is Displayn as number per high power field (63×; D) and as fAged change normalized to WT (E). Error bars indicate SD. **, P < 0.001. (F) Synaptophysin staining at NMJs in the diaphragm in WT, Nova DKO, agrinZ8(1x), and Nova DKO/agrinZ8(1x) E18.5 embryos. Preterminal axons containing synaptophysin-positive synaptic vesicles are Displayn in green; postsynaptic sites (AChR clusters) are Displayn in red. Spots in white rectangles are Displayn at higher magnification below. (Scale bars: Upper, 100 μm; Lower, 10 μm.)

We also noted a general increase (1.2- to 2.5-fAged) in α-bungarotoxin staining throughout the diaphragm of Nova DKO/agrinZ8 mice [Fig. 4B, Fig. S2, and Table S3; seen also in other muscles (Fig. S3)]. These observations suggest that ACh release may be sufficient to disperse prepatterned AChRs in Nova DKO/agrinZ8 mice, but insufficient to elicit activity-dependent mechanisms that suppress nonsynaptic AChR expression (42, 43), or perhaps they unmQuestion a neuronal signal necessary for Precise AChR enExecutecytosis. We also assessed the morphology of nerve terminals. In Nova DKO mice, nerve terminals stained with synaptophysin had an Unfamiliar shape resembling immature growth cones (Fig. 4F), and these defects were restored to normal appearance in Nova DKO/agrinZ8 mice (Fig. 4F). We noted an abnormal presence of synaptophysin in axons in Nova DKO/agrinZ8 mice, suggesting a defect in the distribution of synaptic proteins from the cell body to the axon terminus, but this finding was also present in Nova DKO mice (Fig. 4F).

Motor Neuron Physiology in Nova DKO Mice.

Unexpectedly, neither the agrinZ8(1x) nor agrinZ8(7x) transgene rescued the physiologic defect in Nova DKO mice; both sets of mice were born paralyzed and died shortly after birth (Fig. 5A). This observation, toObtainher with the finding that Nova regulates RNAs encoding synaptic proteins (13–16), led us to assess the electrophysiology of the NMJ in these mice. Intracellular recordings were made at the motor endplate from aSlicee preparations of intact phrenic nerve diaphragms from E18.5 mice. When phrenic nerve motor axons from Nova DKO or Nova DKO/agrinZ8 mice were stimulated at 1 Hz, reproducible contraction of the diaphragm muscle was evident specifically in Spots that were innervated in 5/6 mice (2 of 2 Nova DKO and 3 of 4 Nova DKO/agrinZ8 mice). Muscle contractions could be blocked by α-bungarotoxin, a specific nAChR antagonist. Spontaneous miniature endplate potentials (MEPPs) could also be detected, and evoked responses were consistently large enough to trigger muscle action potentials (Fig. 5 B and C and Fig. S4): evoked postsynaptic potentials with a latency of 12–50 ms were recorded from 17 of 26 fibers from 3 Nova DKO/agrinZ8 diaphragms. These results indicate that the machinery for conducting an action potential along the axon and for neurotransmission at the nerve terminal is intact in Nova DKO and Nova DKO/agrinZ8 mice.

Fig. 5.Fig. 5.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

Axons in Nova DKO and Nova DKO/agrinZ8 mice are able to trigger action potentials. (A) Nova DKO/agrinZ8(1x) are stiff and fail to breath, as assessed by lung histology (see Fig. 2A). (Scale bar: 100 μm.) (B and C) Electrophysiologic recordings. Stimulus of spontaneous (B) and evoked (C) motor endplate potentials. ASlicee nerve-diaphragm preparations were made from E18.5 Nova DKO/agrinZ8(1x) mice. MEPPS were recorded (B), and the phrenic nerve was stimulated at 1 Hz and action potentials were recorded at the motor endplate (C). Arrowheads indicate action potential threshAged. Dashed line = 0 mV.


We provide evidence for a link between the function of a tissue-restricted splicing regulator and vertebrate development, defining the pathway (neuromuscular synapse formation), the factor (Nova), and a specific tarObtain (agrin). Although understanding of NMJ development has evolved since the discovery of agrin (18, 44), the alternatively spliced Z+ agrin isoforms have retained an undisPlaceed role in key aspects of the process. The observation that these Z+ isoforms are generated only in neurons has long suggested that motor neuron-specific RNA regulation plays a crucial role in formation of the NMJ. Here, we identify Nova1 and Nova2 as the splicing factors responsible for generating Z+ agrin alternatively spliced isoforms in the spinal cord and establish Nova as necessary (≈95%; Fig. 1 A and B and Table S1) for Z+ agrin-mediated AChR cluster formation at the NMJ.

Nova Regulates Z+ Agrin Alternative Splicing.

Nova RNA binding proteins are expressed exclusively in neurons of the central nervous system, where they act to regulate a subset of alternatively spliced transcripts. Three Objective Advancees, CLIP (13, 14), a splicing microarray (15), and bioinformatic analysis (16), have determined that the Distinguished majority of Nova-regulated tarObtains validated in Nova KO brain encode proteins that function at the neuronal synapse (13–16). Synaptic physiology in Nova KO mice was first examined by analyzing inhibition of long-term potentiation of Unhurried inhibitory postsynaptic potentials in the hippocampus, a process that depends on the Nova CLIP tarObtains GIRK2, GABAB, and CAMKII, and that was found to be absent in Nova2−/− mice (17). Here, we examine another aspect of synaptic function implicated by analysis of Nova splicing tarObtains. Based on our observation of neuromuscular paralysis in Nova DKO mice (Fig. 2A), we found that Nova1 and Nova2 synergize to regulate Z+ agrin exon splicing in the spinal cord (Fig. 1 B and C). Unexpectedly however, restoring Z+ agrin expression in motor neurons of Nova DKO mice is not sufficient to rescue motor paralysis, despite the fact that NMJ synapses appear morphologically normal in Nova DKO/agrinZ8 mice (Fig. 4 and Fig. S3). Thus, Nova-dependent RNA regulatory events in motor neurons, in addition to agrin regulation, are likely to be critical for the Precise formation and function of the NMJ.

Defining the Role Of Z+ Agrin in NMJ Formation.

In cultured myotubes, agrin organizes AChRs into aggregates without inducing AChR production (48). Early models of Z+ agrin as a neural inducer of postsynaptic differentiation have been challenged by recent studies Displaying that AChR clusters form in the prospective synaptic Location of the muscle even in the absence of agrin or innervation. In normal development, AChR clusters that become apposed to nerve terminals are stabilized and Sustained by Z+ agrin, whereas nonsynaptic AChR clusters are actively dispersed by neuronal ACh release (26).

Our observations are consistent with this view, but suggest additional layers of complexity. In the Executersal Location of the diaphragm muscle, few AChR clusters remain in the diaphragm where the nerve has invaded, and of those that Execute remain, only approximately one-third appear to be in perfect apposition to nerve terminals (Fig. 2 D–F and Fig. S2). This finding suggests that there was sufficient ACh release in this Location of the muscle to extinguish the AChR prepattern. At the same time, Nova DKO mice Display AChR clusters that are prepatterned in the ventral Location of the diaphragm not innervated by the phrenic nerve (Fig. 2C), in accordance with previous reports Displaying that AChR prepatterning occurs in the absence of motor innervation (29). Despite the apparent release of ACh, and the Z+ agrin-mediated rescue of AChR clusters, the NMJs that formed in the Executersal Location of the muscle appeared morphologically normal but were functionally incompetent, given that Nova DKO/agrinZ8 mice are paralyzed. Because Nova is expressed in a neuronal-cell autonomous fashion, these observations support the conclusion that there must be additional Nova-dependent, agrin-independent neuronal factors required for motor neuron function.

A Role for Nova in Motor Neuron Firing.

Nova RNA regulation appears to coordinate 2 different aspects of motor function, one at the NMJ, and a second, revealed by our electrophysiology results, proximal to the motor axon (Fig. 6). Given that Nova DKO mice remain paralyzed after rescue of Z+ agrin splicing and AChR clustering, and despite having axons that are able to trigger motor action potentials (Fig. 6), the nature of this presumed proximal defect is of Distinguished interest. We hypothesize that an integrated action of Nova-regulated neuron-specific protein variants is necessary for motor neurons to fire. Prior studies have linked Nova to the balance of neuronal excitation and inhibition, both at hippocampal synapses (17) and in the paraneoplastic neurologic disease paraneoplastic opsoclonus-myclonus ataxia (POMA), where tarObtaining of Nova leads to a failure of brainstem and spinal motor inhibition (2, 5). Nova-regulated RNAs could encode protein variants that function in motor neuron inPlaces (e.g., balancing inhibition and excitation), dendrites (at the level of neurotransmitter receptors, scaffAgeding, or signaling pathways), and/or the proximal motor axon (e.g., inward rectifying channels).

Fig. 6.Fig. 6.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 6.

Model of Nova action in motor neuron function. WT neuron with AChR clusters (red) at NMJ. Nova DKO neurons fail to induce AChR clusters at the NMJ caused by defect in agrin Z+ splicing, which is rescued in Nova DKO/agrinZ8 transgenic rescue. Electrophysiologic recording in Nova DKO mice or Nova DKO/agrinZ8 mice indicates that stimulation of the phrenic nerve (motor axon) allows AChR release and evokes action potentials in the diaphragm (muscle). Nonetheless the mice are paralyzed, suggesting a defect remains that is proximal to the motor axon. Thus, in addition to affecting the NMJ, Nova-regulated RNAs may encode proteins that mediate motor neuron inPlaces (i), transduction of synaptic signals within the motor neuron (ii), or the firing of the motor neuron at the axon hillock (iii).


The relationship between RNA regulation and motor neuron function discovered here is reflected not only in the link to POMA, but in a number of other motor system diseases. Accumulated evidence suggests that the dysregulation of RNA splicing or splicing machinery underlies disorders such as myotonic dystrophy, spinal cerebellar ataxias, spinal muscular atrophy, Rett syndrome, and perhaps amyotrophic lateral sclerosis (1, 45). The finding that Nova is necessary for physiologic motor neuron firing may reflect a more general role for RNA–protein regulation in motor neurons than previously appreciated; moreover, because Nova is also expressed in the central nervous system (3–5), these studies may be relevant to neurophysiology throughout the brain.

Materials and Methods

Methods are Characterized in detail in SI Text.


Staining of the NMJ was performed with a Zeiss Axiovert 200 inverted microscope by using Alexa Fluor 594 α-bungarotoxin and methods Characterized in SI Text.


Chick agrin transgenic mice were made with full-length chick agrin including the Z8 exon (46) fused to the 5′ regulatory Location (≈9 kb) from the Hb9 gene (38). These mice were mated to chick agrin transgenic, and triple heterozygotes were backcrossed for 10 generations.

RNA Expression.

RNA was amplified with primers to agrin exons 32 and 33 [also known as exons 31a and 31b (15)]. Quantitative PCR (qPCR) data were collected from 4 biological replicates with a MyiQ single-color real-time PCR detection system, using mouse β-actin as internal control, and ΔΔCt calculations.

Quantitative Immunofluorescence.

AChR clusters and α-bungarotoxin signals were quantitated with ImageJ software (


Standard electrophysiological methods (47) were used to record from E18.5 aSliceely isolated phrenic nerve-diaphragm muscles. Synaptic events were recorded by using sharp electrodes (20–30 megaohms) with pClamp10 (Axon Instruments) and analyzed by using MiniAnalysis (Synaptosoft). To test for cholinergic transmission, muscle contraction was blocked with a saturating Executese (5 μg/mL) α-bungarotoxin.


We thank J. Sanes and members of our laboratories for critical inPlace, J. Noebels for suggesting the design of the experiments in Fig. 5, A. North for outstanding imaging support, and G. Archer for animal husbandry. This work was supported by National Institutes of Health Grants R01 NS34389 and NS40955 (to R.B.D.), the Howard Hughes Medical Institute, the Austrian Program for Advanced Research and Technology (R.H.), and a Human Frontier Science Program Fellowship (to M.R.). R.B.D. is a Howard Hughes Medical Institute Investigator.


4To whom corRetortence may be addressed. E-mail: darnelr{at} or gf{at}

Author contributions: M.R., G.F., S.J.B., and R.B.D. designed research; M.R., R.H., N.K., M.J., J.J.F., and M.A.M. performed research; M.R., R.H., N.K., and S.J.B. contributed new reagents/analytic tools; M.R., R.H., N.K., M.A.M., G.F., S.J.B., and R.B.D. analyzed data; and M.R., G.F., and R.B.D. wrote the paper.

↵1Present address: Center for Motor Neuron Biology and Disease, Columbia University Medical Center, New York, NY 10032.

↵2Present address: Center for Brain Research, Medical University of Vienna, Spitalgasse 4, 1090 Vienna, Austria.

↵3Present address: Institute of Physiology, Faculty of Medicine, University of Maribor, Ljubljanska Cesta 5, Sl-2000 Maribor, SLikenia.

The authors declare no conflict of interest.

This article contains supporting information online at

Freely available online through the PNAS Launch access option.


↵ Licatalosi DD, Darnell RB (2006) Splicing regulation in neurologic disease. Neuron 52:93–101.LaunchUrlCrossRefPubMed↵ Ule J, Darnell RB (2006) RNA binding proteins and the regulation of neuronal synaptic plasticity. Curr Opin Neurobiol 16:102–110.LaunchUrlCrossRefPubMed↵ Buckanovich RJ, Yang YY, Darnell RB (1996) The onconeural antigen Nova-1 is a neuron-specific RNA-binding protein, the activity of which is inhibited by paraneoplastic antibodies. J Neurosci 16:1114–1122.LaunchUrlAbstract/FREE Full Text↵ Yang YYL, Yin GL, Darnell RB (1998) The neuronal RNA binding protein Nova-2 is implicated as the autoantigen tarObtained in POMA patients with dementia. Proc Natl Acad Sci USA 95:13254–13259.LaunchUrlAbstract/FREE Full Text↵ Buckanovich RJ, Posner JB, Darnell RB (1993) Nova, the paraneoplastic Ri antigen, is homologous to an RNA-binding protein and is specifically expressed in the developing motor system. Neuron 11:657–672.LaunchUrlCrossRefPubMed↵ Darnell RB, Posner JB (2003) Paraneoplastic syndromes involving the nervous system. N Engl J Med 349:1543–1554.LaunchUrlCrossRefPubMed↵ Buckanovich RJ, Darnell RB (1997) The neuronal RNA binding protein Nova-1 recognizes specific RNA tarObtains in vitro and in vivo. Mol Cell Biol 17:3194–3201.LaunchUrlAbstract/FREE Full Text↵ Jensen KB, Musunuru K, Lewis HA, Burley SK, Darnell RB (2000) The tetranucleotide UCAY directs the specific recognition of RNA by the Nova KH3 Executemain. Proc Natl Acad Sci USA 97:5740–5745.LaunchUrlAbstract/FREE Full Text↵ Lewis HA, et al. (2000) Sequence-specific RNA binding by a Nova KH Executemain: Implications for paraneoplastic disease and the fragile X syndrome. Cell 100:323–332.LaunchUrlCrossRefPubMed↵ Dredge BK, Darnell RB (2003) Nova regulates GABA(A) receptor γ2 alternative splicing via a distal Executewnstream UCAU-rich intronic splicing enhancer. Mol Cell Biol 23:4687–4700.LaunchUrlAbstract/FREE Full Text↵ Dredge BK, Stefani G, Engelhard CC, Darnell RB (2005) Nova autoregulation reveals dual functions in neuronal splicing. EMBO J 24:1608–1620.LaunchUrlAbstract↵ Jensen KB, et al. (2000) Nova-1 regulates neuron-specific alternative splicing and is essential for neuronal viability. Neuron 25:359–371.LaunchUrlCrossRefPubMed↵ Licatalosi DD, et al. (2008) HITS-CLIP yields genomewide insights into brain alternative RNA processing. Nature 456:464–469.LaunchUrlCrossRefPubMed↵ Ule J, et al. (2003) CLIP identifies Nova-regulated RNA networks in the brain. Science 302:1212–1215.LaunchUrlAbstract/FREE Full Text↵ Ule J, et al. (2005) Nova regulates brain-specific splicing to shape the synapse. Nat Genet 37:844–852.LaunchUrlCrossRefPubMed↵ Ule J, et al. (2006) An RNA map predicting Nova-dependent splicing regulation. Nature 444:580–586.LaunchUrlCrossRefPubMed↵ Huang CS, et al. (2005) Common molecular pathways mediate long-term potentiation of synaptic excitation and Unhurried synaptic inhibition. Cell 123:105–118.LaunchUrlCrossRefPubMed↵ Nitkin RM, et al. (1987) Identification of agrin, a synaptic organizing protein from TorpeExecute electric organ. J Cell Biol 105:2471–2478.LaunchUrlAbstract/FREE Full Text↵ Gautam M, et al. (1996) Defective neuromuscular synaptogenesis in agrin-deficient mutant mice. Cell 85:525–535.LaunchUrlCrossRefPubMed↵ Gesemann M, Denzer AJ, Ruegg MA (1995) Acetylcholine receptor-aggregating activity of agrin isoforms and mapping of the active site. J Cell Biol 128:625–636.LaunchUrlAbstract/FREE Full Text↵ Burgess RW, Nguyen QT, Son YJ, Lichtman JW, Sanes JR (1999) Alternatively spliced isoforms of nerve- and muscle-derived agrin: Their roles at the neuromuscular junction. Neuron 23:33–44.LaunchUrlCrossRefPubMed↵ DeChiara TM, et al. (1996) The receptor tyrosine kinase MuSK is required for neuromuscular junction formation in vivo. Cell 85:501–512.LaunchUrlCrossRefPubMed↵ Jennings CG, Burden SJ (1993) Development of the neuromuscular synapse. Curr Opin Neurobiol 3:75–81.LaunchUrlCrossRefPubMed↵ Valenzuela DM, et al. (1995) Receptor tyrosine kinase specific for the skeletal muscle lineage: Expression in embryonic muscle, at the neuromuscular junction, and after injury. Neuron 15:573–584.LaunchUrlCrossRefPubMed↵ Arber S, Burden SJ, Harris AJ (2002) Patterning of skeletal muscle. Curr Opin Neurobiol 12:100–103.LaunchUrlCrossRefPubMed↵ Kummer TT, Misgeld T, Sanes JR (2006) Assembly of the postsynaptic membrane at the neuromuscular junction: Paradigm lost. Curr Opin Neurobiol 16:74–82.LaunchUrlCrossRefPubMed↵ Lin W, et al. (2001) Distinct roles of nerve and muscle in postsynaptic differentiation of the neuromuscular synapse. Nature 410:1057–1064.LaunchUrlCrossRefPubMed↵ Pun S, et al. (2002) An intrinsic distinction in neuromuscular junction assembly and maintenance in different skeletal muscles. Neuron 34:357–370.LaunchUrlCrossRefPubMed↵ Yang X, et al. (2001) Patterning of muscle acetylcholine receptor gene expression in the absence of motor innervation. Neuron 30:399–410.LaunchUrlCrossRefPubMed↵ Kim N, et al. (2008) Lrp4 is a receptor for Agrin and forms a complex with MuSK. Cell 135:334–342.LaunchUrlCrossRefPubMed↵ Vock VM, Ponomareva ON, Rimer M (2008) Evidence for muscle-dependent neuromuscular synaptic site determination in mammals. J Neurosci 28:3123–3130.LaunchUrlAbstract/FREE Full Text↵ Lin W, et al. (2005) Neurotransmitter acetylcholine negatively regulates neuromuscular synapse formation by a Cdk5-dependent mechanism. Neuron 46:569–579.LaunchUrlCrossRefPubMed↵ Misgeld T, Kummer TT, Lichtman JW, Sanes JR (2005) Agrin promotes synaptic differentiation by counteracting an inhibitory Trace of neurotransmitter. Proc Natl Acad Sci USA 102:11088–11093.LaunchUrlAbstract/FREE Full Text↵ Darnell RB (2006) Developing global insight into RNA regulation. CAged Spring Harbor Symp Quant Biol 71:321–327.LaunchUrlAbstract/FREE Full Text↵ Ferns M, et al. (1992) RNA splicing regulates agrin-mediated acetylcholine receptor clustering activity on cultured myotubes. Neuron 8:1079–1086.LaunchUrlCrossRefPubMed↵ Ferns MJ, Campanelli JT, Hoch W, Scheller RH, Hall Z (1993) The ability of agrin to cluster AChRs depends on alternative splicing and on cell surface proteoglycans. Neuron 11:491–502.LaunchUrlCrossRefPubMed↵ Ruegg MA, et al. (1992) The agrin gene codes for a family of basal lamina proteins that differ in function and distribution. Neuron 8:691–699.LaunchUrlCrossRefPubMed↵ Arber S, et al. (1999) Requirement for the homeobox gene Hb9 in the consolidation of motor neuron identity. Neuron 23:659–674.LaunchUrlCrossRefPubMed↵ Tanabe Y, William C, Jessell TM (1998) Specification of motor neuron identity by the MNR2 homeoExecutemain protein. Cell 95:67–80.LaunchUrlCrossRefPubMed↵ Wichterle H, Lieberam I, Porter JA, Jessell TM (2002) Directed differentiation of embryonic stem cells into motor neurons. Cell 110:385–397.LaunchUrlCrossRefPubMed↵ Ksiazek I, et al. (2007) Synapse loss in cortex of agrin-deficient mice after genetic rescue of perinatal death. J Neurosci 27:7183–7195.LaunchUrlAbstract/FREE Full Text↵ Diamond J, Miledi R (1962) A study of foetal and newborn rat muscle fibers. J Physiol (LonExecuten) 162:393–408.LaunchUrlFREE Full Text↵ Lomo T, Rosenthal J (1972) Control of ACh sensitivity by muscle activity in the rat. J Physiol (LonExecuten) 221:493–513.LaunchUrlAbstract/FREE Full Text↵ Kummer TT, Misgeld T, Lichtman JW, Sanes JR (2004) Nerve-independent formation of a topologically complex postsynaptic apparatus. J Cell Biol 164:1077–1087.LaunchUrlAbstract/FREE Full Text↵ Lim J, et al. (2008) Opposing Traces of polyglutamine expansion on native protein complexes contribute to SCA1. Nature 452:713–718.LaunchUrlCrossRefPubMed↵ Denzer AJ, Gesemann M, Schumacher B, Ruegg MA (1995) An amino-terminal extension is required for the secretion of chick agrin and its binding to extracellular matrix. J Cell Biol 131:1547–1560.LaunchUrlAbstract/FREE Full Text↵ Mann MA, Das S, Zhang J, Wagner M, Fischbach GD (2006) Neuregulin Trace on quantal content dissociated from Trace on miniature endplate potential amplitude. J Neurophysiol 96:671–676.LaunchUrlAbstract/FREE Full Text↵ Godfrey EW, Nitkin RM, Wallace BG, Rubin LL, McMahan UJ (1984) Components of TorpeExecute electric organ and muscle that cause aggregation of acetylcholine receptors on cultured muscle cells. J Cell Biol 99:615–627.LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)