Syne proteins anchor muscle nuclei at the neuromuscular junction

Abstract

Vertebrate skeletal muscle fibers contain hundreds of nuclei, of which three to six are functionally specialized and stably anchored beneath the postsynaptic membrane at the neuromuscular junction (NMJ). The mechanisms that localize synaptic nuclei and the roles they play in neuromuscular development are unknown. Syne-1 is concentrated at the nuclear envelope of synaptic nuclei; its Caenorhabditis elegans orthologue ANC-1 functions to tether nuclei to the cytoskeleton. To test the involvement of Syne proteins in nuclear anchoring, we generated transgenic mice overexpressing the conserved C-terminal Klarsicht/ANC-1/Syne homology domain of Syne-1. The transgene acted in a dominant interfering fashion, displacing endogenous Syne-1 from the nuclear envelope. Muscle nuclei failed to aggregate at the NMJ in transgenic mice, demonstrating that localization and positioning of synaptic nuclei require Syne proteins. We then exploited this phenotype to show that synaptic nuclear aggregates are dispensable for maturation of the NMJ.

Fig. 1.

Transgenic C-terminal domain of Syne-1 displaces endogenous Syne-1 from nuclei. (A) Structure of Syne-1 and the DNS transgene. The DNS protein was detected with an antibody to its N-terminal myc tag. Endogenous Syne-1 protein was detected by using a polyclonal antibody raised against a domain not included in the DNS transgene (11). (B and C) Longitudinal sections from wild-type and transgenic muscles, triply stained for Syne-1, myc, and DAPI. The myc-tagged DNS protein localizes to the nuclear envelope of muscle nuclei, and endogenous Syne-1 levels are decreased in myc+ nuclei. The preparations in B and C were processed in parallel and photographed with identical exposure times. (D–K) Cross sections of wild-type (D, F, H, and J) or DNS transgenic (E, G, I, and K) muscle were stained with antibodies against Syne-1 (D and E), emerin (F and G), lamin A/C(H and I), Sun2 (J and K), and myc (D′–K′). (L and M) Electron micrographs of nuclear envelopes of myonuclei. No differences were seen between control (L) and DNS muscles (M).

Fig. 2.

Decreased numbers of synaptic nuclei and increased numbers of perisynaptic nuclei at NMJs of DNS transgenic mice. (A–C) Wild-type (A) and transgenic (B and C) endplates from P30 mice, stained with BTX to mark synapses (red) and DAPI to mark nuclei (green). (A) Nuclei are clustered at the wild-type endplate. (B) A transgenic NMJ with no synaptic nuclei and multiple perisynaptic nuclei. (C) Another transgenic NMJ with two synaptic nuclei and multiple perisynaptic nuclei. (D and E) Numbers of synaptic (D) and perisynaptic nuclei (E) in wild-type and DNS transgenic mice (mean ± SEM, n = 150 wild-type and 160 transgenic synapses; defined as in text). (F–H) Some myc-negative nuclei are associated with the NMJ (F, arrows), but confocal rotation (G and H) and lamin staining (see Fig. 6) show that they are external to the muscle fiber. To assess cellular location of nuclei, a stack of z sections (G) was rotated and viewed in the z plane (H). Nucleus 1 is myc-negative and external to the postsynaptic membrane, likely a Schwann cell. Nucleus 2 is myc-positive and within the muscle fiber. (I and J) The same data set as in D and E corrected to exclude nonmuscle nuclei.

Fig. 3.

Ultrastructure of NMJs in DNS mice. Electron micrographs of NMJs from wild-type (A) and DNS transgenic (B) mice. Ultrastructure of the nerve terminal (nt), infolded postsynaptic membrane, and overlying Schwann cells (sch) is normal. However, aggregates of mitochondria (arrows), that normally flank synaptic nuclei (myo), lie directly beneath the postsynaptic membrane in its absence. Densities of active zones in nerve terminals (C) and of junctional folds in the postsynaptic apparatus (D) were normal.

Fig. 4.

Syne-1 is required for both formation and maintenance of synaptic nuclear aggregates. (A and B) Wild-type (A) and transgenic (B) endplates at P8, stained for BTX (red) and DAPI (green). Nuclei are displaced from the synapse by this stage in transgenic muscle. (C and D) Numbers of synaptic (C) and peripsynaptic (D) nuclei in wild-type (filled squares) and DNS transgenic muscles (open squares) as a function of age. (E) DNS-expressing fiber from a muscle transfected with DNS at P20, after synaptic nuclear aggregates had formed, then analyzed at P30. (F and G) Numbers of synaptic (F) and perisynaptic (G) nuclei in DNS-expressing and nonexpressing (control) fibers of muscles DNS-transfected muscles. (Data in C, D, F, and G are means ± SEM; n ≥ 50 synapses for each point; fibers from four separate mice were used in F and G.)

Fig. 5.

Normal synaptic architecture in DNS muscles. High-power views of branches from wild-type (A) and transgenic (B) endplates, stained with BTX, to show striated appearance resulting from junctional folds. (C–F) Double staining with BTX plus antibodies to the fetal AChR subunit (γ) at P3 (C and D) and P14 (E and F) shows that it is replaced by the adult subunit (ε) on schedule in DNS mice. (G and H) Normal distribution of the synaptic cytoskeletal protein, utrophin, in DNS muscle. (I and J) AChR-rich area (I) and number of branch points (bifurcations) per synapse (J) in NMJs of wild-type and DNS muscles (mean ± SEM, n = 30 of each genotype).