Proteins that accumulate with age in human skeletal-muscle aggregates contribute to declines in muscle mass and function in Caenorhabditis elegans

Protein aggregation increases with age in human muscle

Protein aggregation was compared for muscle from young and old adults after isolation of aggregates that were insoluble in a strong ionic detergent (1% sarcosyl) from equal wet weights of initial muscle protein. Aggregates from two young (23 and 26 year-old males) and two old adults (a female age 65, and a male age 68) were pooled and resolved by 2D gel electrophoresis. Staining the gels for total protein showed a greater diversity of proteins, and higher abundance for most of these, in muscle aggregates from aged adults than in aggregates from young adults (Figure 1, A and B). Total muscle-aggregate protein was then quantitated for individuals, comparing samples from 5 young adults (ages 23–27 years; one female) and 5 older adults (ages 64–80 years, two females) on 1-D gels. The results confirmed that protein aggregation was significantly higher (2.1±0.07 fold, P<0.0001) for the older group than for young adults (Figure 1C).

Specific proteins are enriched with age in aggregates from human muscle

Aggregate proteins were identified, and their relative abundances estimated, by mass spectrometry of young and old human muscle samples pooled by age group (as described above). Muscle-aggregate proteins were dissolved in strong detergent and reducing agent, electrophoresed in one dimension, and gel slices were digested in situ with trypsin. Peptides were then identified by matching fragmentation patterns to a reference database within Mascot software (Matrix Science, Boston MA). Spectral counts (the number of peptide identifications per protein) allow an unbiased comparison of relative abundances for any given protein, between young and old human muscle. A total of 515 proteins were positively identified from muscle aggregates. When results were filtered to require ≥1.5-fold difference by age and ≥3 peptide identifications for at least one group, 168 proteins were more abundant in aggregates from old muscle than young (mean old/young ratio ± SEM = 7.0 ± 2.4), whereas 60 proteins were reduced in old muscle (old/young ratio = 0.092 ± 0.025).

For each protein, we assessed the significance of spectral-count enrichment at either age (by Chi-squared tests, relative to total aggregate protein); 56 of the 515 proteins (11%) were significantly enriched in muscle aggregates from old donors while 21 (4%) were depleted with age (each P<0.05). The complete list of identified proteins is provided in Supplementary Table S1. A partial list of aggregate proteins (Table 1) includes 34 proteins that differ significantly in abundance between young and old muscle. Half of these (17) are cytoskeletal and/or involved in muscle contraction, and another (LMCD1) regulates muscle differentiation. Eight are involved in protein homeostasis: 5 heat-shock protein isoforms, peroxi-redoxin, and 2 ubiquitin-conjugating enzymes.

Candidate-gene knockdowns in C. elegans relieve age-associated aggregation

We selected nine C. elegans proteins orthologous to proteins identified in human muscle aggregates (Table 1, candidates listed in bold), for which the corresponding nematode genes are available in the Ahringer RNA-interference (RNAi) library [58]. These comprise four aggregate proteins that were significantly more abundant in old muscle (FTT-2, ADSS-1, MLP-1, and UEV-1), four that showed a suggestive enrichment with age (UNC-44, CAH-1, CAP-2, and DHS-21), and one that was more abundant in young muscle (FLN-1). Gene expression was individually knocked down in nematodes by life-long RNAi, beginning with hatching of L1 larvae from eggs, in strain AM141. This strain expresses Q40::YFP (a tract of 40 glutamines fused in-frame to yellow fluorescent protein) in body wall muscle and is considered a model of Huntington-like protein aggregation as it accumulates fluorescent aggregates during days 1–5 of adult life.

Two knockdowns (targeting adss-1 and dhs-21) delayed development and were not pursued. Six of the remaining seven gene knockdowns significantly reduced protein aggregation, as illustrated in Figure 2. Representative images are shown for day-3 adult controls (panel A), in which aggregates are consistently a bit larger and more numerous than in worms knocked down for a candidate gene (uev-1) (panel B). Computer quantitation of images indicated significant reductions in the number of aggregates (by 6–17%; P-values of 0.03 to <0.0001; Figure 2C) for all RNAi targets except cah-1, accompanied by reduced intensity of Q40::YFP fluorescence per aggregate (by 11–45%; P-values 0.01 to <0.001; Figure 2D) for all targets except cah-1 and mlp-1.

Gene knockdowns in C. elegans which decrease protein aggregation prevent age-associated loss of muscle mass

Targeting the same seven genes with RNAi also preserved muscle mass and function in aged nematodes (Figure 3), which otherwise lose 25–30% of muscle mass by day 3, and >70% by day 9 of adult life. Wild-type nematodes (strain Bristol N2) were exposed from hatch to empty vector (panel A) or gene-specific RNAi constructs (e.g. uev-1, panel B) and stained with rhodamine-tagged phalloidin to quantify muscle at 3 days of adult age. Knockdown of six genes (again, all except cah-1) consistently and significantly increased muscle mass relative to controls by 50–85% (P values ranged from 0.01 to <0.0001; panel C). Essentially the same preservation in muscle quantity was observed, without staining, in transgenic reporter strain RW1596 that expresses green fluorescent protein (GFP) driven by the muscle-specific myo-3 promoter (data not shown).

Gene knockdowns in C. elegans which decrease protein aggregation also counteract age-associated loss of muscle function

Finally, muscle function was assessed in nematode strain CL4176, which becomes paralyzed due to the formation of amyloid aggregates ∼48 hrs after induced expression of human Aβ_1–42 peptide in muscle. Knockdown from hatch, of the same six genes that significantly decreased protein aggregate counts (Figure 2) and increased muscle mass (Figure 3) also consistently rescued worms from amyloid paralysis, resulting in 39–67% (average of 47%) more motile worms per group at 48 hr post-induction (each P≤0.0005) than were seen among controls (Figure 4). By the time control worms reached zero motility (100% paralysis, 48 hr post-induction), 5–20% of RNAi-exposed worms remained motile (data not shown).

Human subjects

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki and national and international guidelines. After provision of informed consent, human muscle samples were collected under a protocol approved by the Institutional Review Board (IRB) of the University of Arkansas for Medical Sciences. Samples were de-identified for analysis at the Central Arkansas Veterans Healthcare System, and the CAVHS IRB determined the ensuing proteomic studies to be non-human research.

All subjects were medically cleared for participation by a study physician. Eligible individuals were between 19 – 30 years of age (“young” group) or 60 – 80 years of age (“old” group), had body mass indices of <30 kg/m², and were non-smokers. Individuals were excluded for history of lidocaine allergy or bleeding disorder, or for taking anti-inflammatory or anti-coagulant medications within seven days of their appointment. Subjects were required to abstain from strenuous physical activity for 3 days and alcohol consumption for 2 days prior to undergoing muscle biopsy. Muscle was sampled from the vastus lateralis under local anesthesia (1% lidocaine-HCl, buffered to pH 9:1 with sodium bicarbonate) using a 6-mm Bergstrom trocar (Millennium Surgical, Narberth, PA) with mild suction. The tissue was immediately stabilized in RNAlater (Life Technologies, Grand Island, NY), held overnight at 4°C, and then stored at −20°C until processing.

Human aggregate purification, protein separation and quantitation

Muscle (15 mg per specimen) was pulverized in a dry-ice cooled mortar and suspended in cold stabilizing buffer (20-mM Hepes pH 7.4, 0.3-M NaCl, 2-mM MgCl₂, 1% NP40 (w/v)) containing phosphatase/prote-ase inhibitors (CalBiochem). Crushed tissue was lysed on ice using a Teflon homogenizer (2 × 10 sec) and sonication (3 × 10 sec). After centrifugation (5 min, 2000g) to remove debris and organelles, protein concentrations of supernatants were determined (Bradford Protein Assay, Bio-Rad). Aggregate proteins were collected from equal amounts of total protein by centrifugation (15 min, 14000g), and pellets were suspended in ionic detergent buffer (1% sarcosyl (v/v); 0.1-M Hepes, pH 7.4; 5-mM EDTA) with protease inhibitors. Detergent-insoluble aggregate proteins were then pelleted by ultracentrifugation (30 min, 100,000 x g) and resuspended in 125-μl ampholyte buffer (8M urea; 2% CHAPS; 40-mM DTT; 0.2% w/v Biolyte, pH range 3–10). Proteins were separated by isoelectric focusing (pH 3–10) and then by orthogonal gradient gel electrophoresis (1% sodium dodecyl sulfate (SDS), 4–12% polyacrylamide). The separated proteins were visualized after staining with SYPRO Ruby (Invitrogen) and fluorescence intensities were quantified using ImageJ software (http://imagej.nih.gov).

Identification of proteins in human muscle aggregates

Individual proteins in the aggregate fractions were resolved and identified as follows. Sarcosyl-insoluble aggregate proteins (the 100,000 x g pellet described above) were dissolved by heating (5 min, 95°C) in Laemmli buffer containing 2% SDS (w/v) and 0.5% Δ-mercaptoethanol (v/v). Proteins were resolved on gradient gels (1% SDS, 4‒12% polyacrylamide) and stained with SYPRO Ruby or Coomassie Blue. Each lane was robotically partitioned into 1-mm slices and the proteins were digested in situ with trypsin [16]. The resulting peptides were analyzed by high-resolution LC tandem MS (ThermoFisher Orbitrap Fusion Tribrid) as reported previously [56;57]. Proteins were identified by MASCOT (www.matrixscience.com) matching of pep-tide fragments to a database of predicted patterns [57].

Nematode strains

The wild-type C. elegans and two disease-model strains used to assess the role of protein aggregation in muscle aging were obtained from the Caenorhabditis Genetics Center which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440). Strains were maintained on 2% agar plates containing nematode growth medium overlaid with E. coli (strain OP50) at 20°C. Cohort age was synchronized by alkaline-hypochlorite lysis of well-fed gravid hermaphrodites at day 3 post-hatch (day 1 of adulthood) to release unlaid eggs. Eggs were plated on 100-mm plates seeded with empty-vector control or RNAi bacteria (Ahringer RNAi library [58]) designed to knock down production of each targeted protein. Nematode images were captured using an Olympus BX51 fluorescence microscope equipped with an Olympus DP71 camera, and aggregates were quantified using ImageJ software (http://imagej.nih.gov).

Nematode strains were selected to facilitate measurement of the effects of aging on muscle protein aggregation, muscle mass, and paralysis. Wild-type strain Bristol-N2 (N2) is a relatively long-lived DRM stock of Bristol-N2 maintained by the Caenorhabditis Genetics Center. Strain AM141 (unc-54p/Q40::YFP) expresses a tract of 40 glutamines (Q40) fused in-frame to yellow fluorescent protein (YFP), constitutively in muscle. This strain emulates the threshold pathologic mutation of huntingtin protein that elicits Huntington's disease, and forms yellow fluorescent aggregates in muscle progressively with adult age [16]. Aggregate foci were counted, either manually or using DotCount (http://reuter.mit.edu/software/dotcount/) software. Strain CL4176 (smg-1^ts [myo-3p/Aβ_1–42/long 3′-untranslated region] ) expresses human amyloid β_1–42 (Alzheimer's disease peptide, Aβ_1–42) in muscle after induction by temperature upshift from 20° to 25°C at the L3/L4 transition [16]. Induction of Aβ_1–42 expression causes paralysis that was measured at 29 h post-induction, as movement of the head but not the body in response to touch. All experiments were performed at least in triplicate.

Statistical analysis

Differences between groups were assessed for significance by the Fisher-Behrens heteroscedastic t test (appropriate to samples of unequal or unknown variance). Differences in peptide abundance, based on mass spectrometry spectral counts, were assessed for significance by chi-squared or 2-tailed Fisher exact tests. Significance values reported are nominal P values without adjustment for multiple comparisons. To calculate mean fold differences of proteins that were uniquely present in one age group, 0.1 was substituted for zero to prevent division by zero.