Bace1 and Neuregulin-1 cooperate to control formation and maintenance of muscle spindles

Abstract

The protease β-secretase 1 (Bace1) was identified through its critical role in production of amyloid-β peptides (Aβ), the major component of amyloid plaques in Alzheimer's disease. Bace1 is considered a promising target for the treatment of this pathology, but processes additional substrates, among them Neuregulin-1 (Nrg1). Our biochemical analysis indicates that Bace1 processes the Ig-containing β1 Nrg1 (IgNrg1β1) isoform. We find that a graded reduction in IgNrg1 signal strength in vivo results in increasingly severe deficits in formation and maturation of muscle spindles, a proprioceptive organ critical for muscle coordination. Further, we show that Bace1 is required for formation and maturation of the muscle spindle. Finally, pharmacological inhibition and conditional mutagenesis in adult animals demonstrate that Bace1 and Nrg1 are essential to sustain muscle spindles and to maintain motor coordination. Our results assign to Bace1 a role in the control of coordinated movement through its regulation of muscle spindle physiology, and implicate IgNrg1-dependent processing as a molecular mechanism.

Figure 1

Bace1 is required for the correct formation of muscle spindles. (A) Performance of Bace1^+/+, Bace1^+/− and Bace1^−/− mice (P180) in the inverted grid test. Horizontal bars show the average time the mice remain clinging to the grid. (B) Normal walking pattern of a mouse; the movement of homolateral (top) and homologous (bottom) limbs are indicated. (C) Representative walking patterns of control and Bace1^−/− mice; displayed are series of four consecutive steps. (D, E) Homolateral (D) and homologue (E) coupling values for the movements of control (inner circle) and Bace1^−/− mice (outer circle). A value of 0.5 defines coordinated movement of the paws. The grey area indicates the non-pathological interval (0.5±0.1). (F) Schematic representation of a muscle spindle; indicated are the central intrafusal fibres of the spindle, which are surrounded by a capsule and (extrafusal) muscle fibres. (G) Quantification of muscle spindles in hindlimbs of control, heterozygous and homozygous Bace1 mutant mice at P0. (H) Immunohistological analysis of muscle spindles from control and Bace1 mutant mice at P0. Intrafusal fibres express Egr3, the nascent outer capsule displays a pronounced collagen IV staining, and contacting sensory fibres are NF200⁺. Scale bar, 10 μm (H).

Figure 2

Bace1 activity regulates motor coordination and maintains muscle spindles during adulthood. (A) Western blot analysis of Bace1 expression and processing of endogenous Nrg1 in the midbrain of adult (P180) wild type (wt) or Bace1^−/− (B1^−/−) mice treated with vehicle (Vh) or the Bace1 inhibitor Ly2811376 (Ly). Nrg1 FL denotes the 130 kDa Nrg1 species corresponding to full-length CRD-Nrg1, detected by an antibody directed against the Nrg1 C-terminal sequence. The amount of Nrg1 FL is markedly increased upon Bace1 inhibition or ablation, indicating reduced cleavage, but Bace1 expression is unaffected by Ly2811376 treatment. Calnexin is used as internal control. (B) Quantification of muscle spindles in the tibialis anterior muscle of control, heterozygous and homozygous Bace1 mutant mice (P180) treated with the Bace1 inhibitor Ly2811376. (C) Performance of Bace1^+/+, Bace1^−/− and vehicle- or Ly2811376-treated wild-type mice in the inverted grid test. Note that Bace1^−/− and Ly2811376-treated mice perform similarly. (D) Representative walking patterns of vehicle- and Ly2811376-treated wild-type mice. (E, F) Homolateral (E) and homologue (F) coupling values for the movements of vehicle- (inner circle) and Ly2811376-treated (outer circle) wild-type mice. Bace1^−/− mice (middle circle) are included for comparison.

Figure 3

Nrg1β1 is a substrate of Bace1. (A) Structure of Nrg1 isoforms containing cysteine-rich domain (CRD) or Ig-like (Ig) domains. Full-length (Nrg1 FL), N-terminal (Nrg1 NtF) and C-terminal (Nrg1 CtF) fragments analysed in (E, F, H, I) are indicated. Green arrows show the Bace1 cleavage site present in β1 isotypes, yellow star the position of the HA-tag used for biochemical analysis; C-terminal transmembrane domain TM. (B) In situ hybridization (Bace1, IgNrg1) combined with immunohistology (neurofilament 200; NF200) demonstrates co-expression of Bace1 and IgNrg1 in sensory neurons. Left: Bace1 is broadly expressed in sensory neurons. Right: IgNrg1 and Bace1 (insert) are co-expressed in NF200⁺ sensory neurons. (C) qPCR of DRG mRNA encoding CRD-Nrg1 and α/β IgNrg1 isoforms (P0). (D) Sequence alignment of α/β Nrg1 isotypes; predicted Bace1 cleavage sites are indicated. Asterisk indicates stop codon in β3. (E, F) Western blot analysis of Bace1-dependent processing of (E) Ig- and CRD-Nrg1β1, (F) IgNrg1β1 and IgNrg1β2 in HEK293 cells. Antibodies are indicated (αCNX: anti-calnexin). Full-length precursors (Nrg1 FL; IgNrg1: 110 kD, CRD-Nrg1: 130 kD) and processed C-terminal fragment (Nrg1 CtF: 65 kD) are detected in RIPA lysates using an antibody against the Nrg1 C-terminus; HA-tagged N-terminal fragments (Nrg1 NtF) are detected in the supernatant using anti-HA. Calnexin serves as loading control. NT, non-transfected. (G) Quantification of Bace1-dependent shedding of IgNrg1β1^SEAP and IgNrg1β2^SEAP. The N-terminal sequence of IgNrg1^SEAP variants contains alkaline phosphatase whose enzymatic activity is detected in supernatants in the absence/presence of Bace1 cDNA, C3 or GM6001 inhibitors. (H) Western blot analysis of IgNrg1β1 cleavage in primary neurons. Detected are full-length and C-terminal IgNrg1β1 in RIPA lysates using an antibody recognizing the Nrg1 C-terminus. (I) The N-terminal fragment of IgNrg1β1 was directly detected in supernatant by western blotting using anti-HA (upper panels). Alternately, supernatant was immunoprecipitated using anti-HA (middle and lower panels), and Nrg1 NtF fragments were detected using anti-HA and 4F10 antibodies; the latter identifies a Bace1-specific Nrg1 cleavage product. Asterisks indicate cross-reactive proteins. Scale bar, 50 μm (B).

Figure 4

IgNrg1 isoforms control motor coordination. (A) Quantification of the expression of CRD and Ig Nrg1 isoforms in DRG neurons from control, IgNrg1^Δ/+, co-IgNrg1 and IgNrg1β1^Ov newborn mice using qPCR. (B) Performance of control and co-IgNrg1 mice (P180) in the inverted grid test. (C) Representative walking pattern of control and co-IgNrg1 mice. (D, E) Homolateral (D) and homologue (E) coupling values of limbs in control (inner circle) and co-IgNrg1 (outer circle) mice (P180).

Figure 5

Bace1 and IgNrg1 control the ontogeny of muscle spindles. (A) Quantification of muscle spindles in hindlimbs of newborn mice of indicated genotypes. (B) Immunohistological analyses of DRG from control, Bace1^−/− and IgNrg1β1^Ov animals at E15 and P0 using antibodies directed against TrkC, NF200 and Islet1/2. (C, D) Quantification of TrkC⁺ proprioceptive neurons (C) and Islet⁺ sensory neurons (D) in DRG of E15 and P0 animals. TrkC⁺ proprioceptive neurons persisted in increased numbers in newborn IgNrg1β1^Ov, while the overall quantity of Islet1/2⁺ sensory neurons was comparable. Bace1, co-IgNrg1 and wild-type mice display similar numbers of sensory or proprioceptive neurons. Scale bar, 25 μm (B).

Figure 6

The maturation of muscle spindles depends on Bace1 and IgNrg1. (A) Immunohistological staining of adult (P30) muscle spindles from the hindlimbs of mice of indicated genotypes. The capsule (collagen IV⁺) and sensory innervation (NF200⁺) are observable in all genotypes. Note that the intensity of calbindin staining of intrafusal fibres depends on the dose of IgNrg1 and on Bace1 (quantified in Supplementary Figure S6A). (B–E) Distribution of the diameter (B, C) and number of intrafusal fibres (D, E) in muscle spindles of Bace1 and IgNrg1 mutants. Scale bar, 20 μm (A).

Figure 7

Continuous Nrg1 signalling is required to sustain muscle spindles in the adult. (A) Schematic outline of Nrg1 ablation and spindle analysis in adult coTxNrg1 mice (Tx, tamoxifen). (B) qPCR analysis of Nrg1 transcripts in DRG of control and coTxNrg1 mice at P100. (C, D) Performance of coTxNrg1 mice (P120) on the beam walk test. coTxNrg1 mice display frequent missteps and distinctive hopping movements; the latter is also a feature of co-IgNrg1 mice. (E) 3D reconstruction of muscle spindles (white lines) in the tibialis anterior muscle of control and coTxNrg1 mice (P180). (F) Quantification of numbers of muscle spindles in tibialis anterior muscle of control (black bars) and coTxNrg1 (red bars) animals. (G) Distribution of muscle spindle length in coTxNrg1 and control mice. (H) Immunohistological analysis of muscle spindles from control and coTxNrg1 animals. The presence of NF200⁺ endings indicates that sensory projections contacted spindles in both genotypes (arrowheads), but calbindin was not detectable in intrafusal fibres of coTxNrg1 mice. In addition, collagen IV staining in outer capsules was weak and interrupted, indicating that capsules degenerated in remaining spindles of coTxNrg1 mice. (I) Schematic summary of gait impairment in Ly2811376-treated, Bace1^−/−, co-IgNrg1 and coTxNrg1 animals. Each panel displays the directions of paw movements (arrows) during a series of two steps. Both Bace1^−/− and Bace1-inhibited mice display defective anterior/posterior coordination resulting in a swaying walk. In addition, co-IgNrg1 and coTxNrg1 mice show more pronounced deficits of posterior homologue coupling resulting in frequent hopping. Scale bars, 1.5 mm (E); 15 μm (H).