Motor and Sensory Deficits in the teetering Mice Result from Mutation of the ESCRT Component HGS

Abstract

Neurons are particularly vulnerable to perturbations in endo-lysosomal transport, as several neurological disorders are caused by a primary deficit in this pathway. In this report, we used positional cloning to show that the spontaneously occurring neurological mutation teetering (tn) is a single nucleotide substitution in hepatocyte growth factor-regulated tyrosine kinase substrate (Hgs/Hrs), a component of the endosomal sorting complex required for transport (ESCRT). The tn mice exhibit hypokenesis, muscle weakness, reduced muscle size and early perinatal lethality by 5-weeks of age. Although HGS has been suggested to be essential for the sorting of ubiquitinated membrane proteins to the lysosome, there were no alterations in receptor tyrosine kinase levels in the central nervous system, and only a modest decrease in tropomyosin receptor kinase B (TrkB) in the sciatic nerves of the tn mice. Instead, loss of HGS resulted in structural alterations at the neuromuscular junction (NMJ), including swellings and ultra-terminal sprouting at motor axon terminals and an increase in the number of endosomes and multivesicular bodies. These structural changes were accompanied by a reduction in spontaneous and evoked release of acetylcholine, indicating a deficit in neurotransmitter release at the NMJ. These deficits in synaptic transmission were associated with elevated levels of ubiquitinated proteins in the synaptosome fraction. In addition to the deficits in neuronal function, mutation of Hgs resulted in both hypermyelinated and dysmyelinated axons in the tn mice, which supports a growing body of evidence that ESCRTs are required for proper myelination of peripheral nerves. Our results indicate that HGS has multiple roles in the nervous system and demonstrate a previously unanticipated requirement for ESCRTs in the maintenance of synaptic transmission.

Fig 1. Positional cloning and phenotypic analysis of the tn mice.

(A) Image showing reduced size of 4-week-old Hgs ^tn/tn mice relative to wild type HGS ^+/+ mice. (B) Body mass of 3- to 5-week-old HGS ^+/+, heterozygous Hgs ^tn/+ and Hgs ^tn/tn mice. n > 6 mice per genotype. A two-way anova was used to find a significant effect of genotype on body mass. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method (C) Kaplan-Meier survival curve of wild type (Hgs ^+/+) and Hgs-mutant mice. The Hgs ^KO allele does not complement the Hgs ^tn allele. A Mantel-Cox test with p<0.001 demonstrated a significant difference between the survival curves of the Hgs ^tn/tn and Hrs ^KO/tn mice as compared to the Hg ^KO/+ Hgs ^tn/+, and Hgs ^+/+ mice. (D) Meiotic linkage map depicting SNP markers that define the tn critical region. (E) Genomic sequencing of HGS revealed an adenine to guanine change in the Hgs ^tn/tn mice, resulting in a methionine to valine substitution at amino acid 89 of HGS. (F) Schematic of HGS protein structure in eukaryotes, demonstrating the conservation of the methionine residue at position 89 in the VHS domain (orange box). Data are shown as mean ± SE. *p < 0.05 and ***p < 0.001.

Fig 2. HGS expression in Hgs ^+/+ and Hgs ^tn/tn tissues.

(A) qPCR analysis of Hgs mRNA expression in 4-week-old Hgs ^+/+ tissues. Transcript level is expressed relative to Hgs level found in the brain. (B) Representative immunoblot of HGS expression in 4-week-old Hgs ^+/+ (wt) and Hgs ^tn/tn (tn) mice. β-actin was used as a loading control. (C) qPCR analysis of Hgs levels from the brains of Hgs ^+/+ mice during postnatal development. (D) Representative immunoblot analysis of HGS expression from embryonic day 15 (E15) to postnatal day 35 (P35) in Hgs ^+/+ (wt) and Hgs ^tn/tn (tn) brain lysates. β-tubulin is used as a loading control. (E) Quantitation of developmental time course of HGS expression in Hgs ^+/+ (wt) and Hgs ^tn/tn (tn) mice expressed as percent of E15 Hgs ^+/+ levels. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. A one way anova with a Geisser-Greenhouse adjustment demonstrated a significant difference between time points. (F) Quantitation of HGS expression in Hgs ^tn/tn mice expressed as a percent of Hgs ^+/+ controls at each developmental time point. Data are shown as ± SE. Symbols represent unpaired t-tests. *p<0.05 and ***p<0.001.

Fig 3. Comparison of hippocampal structure and protein expression between Hgs ^tn/tn and Hgs ^+/+ mice.

(A) CA3 hippocampal sections from 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice were stained for Nissl, myelin basic protein (MBP), glial fibrillary protein (GFAP) and activated caspase 3. Inset is positive control (E13 embryo) for activated caspase-3. Nuclei are stained with DAPI (blue). (B) Representative immunoblot of hippocampal lysates from 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice. Blots were probed for the ESCRT components HGS, STAM1, and CHMP2B, the receptor tyrosine kinases TrkA and TrkB, and the autophagic markers LC3 and p62. (C) Quantitation of immunoblots from hippocampal lysates. n = 3 mice per genotype. Data are shown as mean ± SE. *p<0.05 and ***p<0.001.

Fig 4. Gene dosage effects of HGS expression on motor and sensory function in 3- to 4-week-old mice.

(A) Reduced HGS expression results in clawed paws in Hgs ^tn/tn mice. Behavioral assays of (B) open field, (C) rotarod, (D) elevated beam, (E) von Frey and (F) forelimb grip strength for Hgs ^+/+ (black), Hgs ^tn/+, Hgs ^tn/tn and Hgs ^KO/+ mice. n = 6 mice per genotype for all assays except rotarod, where n = 4. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. Data are shown as mean ± SE and n>6 animals per genotype for all assays except rotarod, where n = 4. A two-way anova demonstrated a significant effect of genotype on rotarod and elevated beam performance. Data are shown as mean ± SE. *p<0.05, **p<0.01 and ***p<0.001.

Fig 5. Examination of sciatic nerves from 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice.

(A) Electron micrograph of sciatic nerves from 4-week-old Hgs ^tn/tn and Hgs ^+/+ mice. Scale bar, 2 μm. Arrowheads indicate hypermyelinated fibers, curved arrows indicated disorganized myelin and arrows indicate demyelination. (B) Quantitation of axon density in myelinated and unmyelinated nerves. (C) Quantitation of average myelinated and unmyelinated axon diameters. (D) Histogram of frequency of axon diameters demonstrating an increase in large diameter myelinated axons in the sciatic nerves of Hgs ^tn/tn mice relative to Hgs ^+/+ controls. Shaded region represents axonal size distribution from Hgs ^+/+ mice. An unpaired t-test with a Welch’s correction demonstrated a significant difference in the distribution of axonal size frequency between Hgs ^+/+ and Hgs ^tn/tn mice. (E) Quantitation of the ratio of axon diameter to total fiber thickness (G-ratio). Symbols represent unpaired t-tests. (F) Relationship between myelin thickness and axon diameter in Hgs ^+/+ and Hgs ^tn/tn sciatic nerves. Circled region depicts 1.0–2.0 μm diameter axons that are affected in the Hgs ^tn/tn sciatic nerves. (G) Representative micrographs of myelin pathology in Hgs ^tn/tn nerves demonstrating (1–2) Tomaculous fibers, (3–5) myelin infoldings compared to (6) Hgs ^+/+controls. n = 3 mice per genotype. Scale bar, 5 μm. Data are shown as mean ± SE. *p<0.05 and ***p<0.001.

Fig 6. Distribution of HGS in sciatic nerves of 4-week-old Hgs ^+/+mice.

Top panel, cross sections of sciatic nerves stained with antibodies against neurofilament (NF, green) and HGS (red). Bottom panel, sciatic nerves were stained with antibodies to the Schwann cell marker S100β (green) and HGS (red). Scale bar, 10 μm.

Fig 7. Analysis of ESCRT and RTK expression in the sciatic nerves of 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice.

(A) Representative immunoblot of ESCRT expression in sciatic nerve extracts. β-tubulin was used as a loading control. (B) Quantitation of immunoblots of ESCRT expression in sciatic nerve extracts. (C) qPCR analysis of ESCRT-0 components in the sciatic nerve. (D) Representative immunoblot of TrkB.FL and TrkB.T1, EGFR and ERBB2 in sciatic nerves. β-tubulin was used as a loading control. (E) Quantitation of receptor tyrosine kinases in the sciatic nerve. (F) qPCR analysis of TrkB and BDNF in the sciatic nerve. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. Data are shown as mean ± SE. n > 3 mice per genotype. **p<0.01 and ***p<0.001.

Fig 8. Levels of HGS-interacting proteins and putative substrates in spinal cord extracts of 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice.

(A) Representative immunoblot and (B) quantitation of the ESCRT-0 proteins HGS and STAM1, the ESCRT-I protein TSG101, the ESCRT-0 interacting proteins EPS15, and the receptor tyrosine kinases TrkB, TrkA and EGFR in the spinal cords of Hgs ^+/+ and Hgs ^tn/tn mice. β-tubulin was used as a loading control. (C) qPCR analysis of Hgs and Stam1 in the spinal cords of Hgs ^+/+ and Hgs ^tn/tn mice. Levels are expressed relative to levels found in wild type Hgs ^+/+ mice. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. Data are shown as mean ± SE. (D) Motor neuron counts from lumbar segments 4/5 from Hgs ^+/+ and Hgs ^tn/tn mice. n = 3 mice per genotype. (E) Immunostaining of Hgs ^+/+ and Hgs ^tn/tn L4/5 segments with GFAP. Scale bar, 100 μm. Data are shown as mean ± SE. n > 3 mice per genotype. ***p<0.001.

Fig 9. Alterations in muscles and motor endplates in the Hgs ^tn/tn mice.

(A) Wet weights of gastrocnemius muscles from 4 week-old Hgs ^+/+ and Hgs ^tn/tn mice. (B) Ratio of gastrocnemius muscle weights to body mass. n > 6 mice per genotype for each time point. (C) Gastrocnemius muscle fiber size measurements for Hgs ^+/+ and Hgs ^tn/tn mice. n > 6 mice per genotype. Symbols represent unpaired t-tests. (D) qPCR analysis of AChR-α, AChR-β, AChR-δ, AChR-ε, and AChR-γ mRNAs from the gastrocnemius muscles of 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice. n > 3 mice per genotype. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. (E) Motor endplate pathology in the Hgs ^tn/tn mice. TA muscle fibers from Hgs ^+/+ and Hgs ^tn/tn mice containing the Thy1-Yfp transgene (green) were stained with TRITC-α-bungarotoxin (red) to label the postsynaptic receptors. The presynaptic axons and nerve terminals are shown in green. Arrowheads mark ultra-terminal sprouting, and curved arrows mark swollen presynaptic terminals. Scale bar, 20 μm. (F) Quantitation of terminal swellings and terminal sprouting from Hgs ^+/+ and Hgs ^tn/tn mice. n > 6 mice per genotype. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. (G) Histogram of endplate area defined by TRITC-α-bungarotoxin (red) labeling of the postsynaptic AChR in Hgs ^+/+ and Hgs ^tn/tn mice. An unpaired t-test with a Welch’s correction demonstrated a significant difference in the distribution of endplate size frequency between Hgs ^+/+ and Hgs ^tn/tn mice. n > 6 mice per genotype. Data are shown as mean ± SE. *p<0.05, **p<0.01 and ***p<0.001.

Fig 10. Loss of HGS increases the number of endosome-like structures and results in synaptic transmission deficits at the NMJ.

(A) Representative electron micrographs of NMJs in the TA muscle from 4-week-old Hgs ^+/+ and Hgs ^tn/tn mice. Arrowheads point to endosomes-like structures. Asterisk marks an MVB in the Hgs ^tn/tn presynaptic terminal. No MVBs were observed in Hgs ^+/+ terminals. Scale bar, 500 μm. (B) Quantitation of endosome-like structures at the motor axon terminals. Symbol represents unpaired t-tests. (C) A 50% reduction in EPC amplitudes was observed in the endplates from 3-week-old Hgs ^tn/tn mice (n = 12 endplates from 6 mice) as compared to controls (n = 12 endplates from 5 mice). (D) MEPC amplitudes were reduced in the Hgs ^tn/tn mice (n = 19 endplates from 6 mice) to 46% of Hgs ^+/+ controls (n = 15 endplates from 8 mice). (E) Quantal content was significantly lower in Hgs ^tn/tn mice (n = 8 endplates from 6 mice) than in Hgs ^+/+ controls (n = 12 endplates from 5 mice). (F) Reduced HGS expression results in a 65% reduction in MEPC frequency at the TA muscles of Hgs ^tn/tn mice (n = 17 endplates from 6 mice) compared to Hgs ^+/+controls (n = 14 endplates from 5 mice). Symbol represents unpaired t-tests **p<0.01 and ***p<0.001.

Fig 11. Effect of reduced HGS expression on ubiquitin conjugates in the nervous system of Hgs ^tn/tn mice.

(A) Representative immunoblot of ubiquitin conjugates from the nervous system of 4-week-old Hgs ^+/+ (wt) and Hgs ^tn/tn (tn) mice. (B) Quantitation of immunoblots. n = 3 per genotype. Symbols represent unpaired t-tests corrected for multiple comparisons using the Holm-Sidak method. Data are shown as mean ± SE. **p<0.001.