Body weight-dependent troponin T alternative splicing is evolutionarily conserved from insects to mammals and is partially impaired in skeletal muscle of obese rats

Abstract

Do animals know at a physiological level how much they weigh, and, if so, do they make homeostatic adjustments in response to changes in body weight? Skeletal muscle is a likely tissue for such plasticity, as weight-bearing muscles receive mechanical feedback regarding body weight and consume ATP in order to generate forces sufficient to counteract gravity. Using rats, we examined how variation in body weight affected alternative splicing of fast skeletal muscle troponin T (Tnnt3), a component of the thin filament that regulates the actin-myosin interaction during contraction and modulates force output. In response to normal growth and experimental body weight increases, alternative splicing of Tnnt3 in rat gastrocnemius muscle was adjusted in a quantitative fashion. The response depended on weight per se, as externally attached loads had the same effect as an equal change in actual body weight. Examining the association between Tnnt3 alternative splicing and ATP consumption rate, we found that the Tnnt3 splice form profile had a significant association with nocturnal energy expenditure, independently of effects of weight. For a subset of the Tnnt3 splice forms, obese Zucker rats failed to make the same adjustments; that is, they did not show the same relationship between body weight and the relative abundance of five Tnnt3 β splice forms (i.e. Tnnt3 β2-β5 and β8), four of which showed significant effects on nocturnal energy expenditure in Sprague-Dawley rats. Heavier obese Zucker rats displayed certain splice form relative abundances (e.g. Tnnt3 β3) characteristic of much lighter, lean animals, resulting in a mismatch between body weight and muscle molecular composition. Consequently, we suggest that body weight-inappropriate skeletal muscle Tnnt3 expression in obesity is a candidate mechanism for muscle weakness and reduced mobility. Weight-dependent quantitative variation in Tnnt3 alternative splicing appears to be an evolutionarily conserved feature of skeletal muscle and provides a quantitative molecular marker to track how an animal perceives and responds to body weight.

Fig. 1.

Characterization and quantification of Tnnt3 alternative splicing in rat gastrocnemius muscle. (A) Rat Tnnt3 pre-mRNA comprises 18 exons, including a 5′ alternatively spliced cassette containing exons 4–9 (white boxes), and the mutually exclusive exons 16 and 17. Twelve different Tnnt3 splice forms were detected by RT-PCR and cDNA sequencing (‘X’ denotes inclusion of specific exons). Forward (FAM-fTnt_F1, blue) and reverse (fTnt_R2 and fTnt_R4, red) RT-PCR primers were designed so that each possible Tnnt3 splice form amplicon size was unique and detectable by DNA fragment analysis (see Materials and methods). Splice forms containing exon 16 or exon 17 were given an ‘α’ or ‘β’ designation, respectively. (B) Fluorescently labeled DNA fragment peaks showing Tnnt3 splice form diversity and abundance (i.e. peak height) of each splice form in gastrocnemius muscle of two rats differing in size (inset shows spectrum for the smaller of the two rats). Internal size standards are represented by red traces.

Fig. 2.

Tnnt3 splice form abundance is adjusted to total mass in rat gastrocnemius muscle. (A) Relative abundance of Tnnt3 splice form α1 as a function of native body mass (left) and total mass (body+added load; right) for weight-loaded and control male Sprague–Dawley rats. Solid lines depict linear regression fits to data for weight-loaded (R²=0.84, P<0.0001, N=19) and control animals (R²=0.67, P<0.0001, N=31), separately (left panel) or combined (right panel) (R²=0.78, P<0.0001). (B) Expression of Tnnt3 splice form β2 (plotted here against total load) was not sensitive to native or experimental changes in body weight. Tnnt3 splice form β3 expression initially increased with total mass, but leveled off in rats >250 g. Both Tnnt3 β5 and β9 expression correlated negatively with total mass and were sensitive (Tnnt3 β5 marginally so) to experimental body weight manipulation (see Table 1). All relative abundance data were arcsine transformed to achieve normality.

Fig. 3.

Weight-associated changes in TNNT3 protein bands. (A) A western blot for TNNT3 detected four bands in smaller Sprague–Dawley rats, whereas three bands (i.e. bands 1, 2 and 4) were apparent in larger rats. β-Actin levels (obtained after stripping and reprobing the same western blot) are provided to demonstrate relatively uniform protein loading. β-Actin levels were not significantly affected by changes in body weight (one-way ANOVA: F=2.57, P=0.163). (B) Relative abundance of TNNT3 protein bands in relation to body mass. In a sample of gastrocnemius muscles from 10 Sprague–Dawley individuals, the most abundant TNNT3 protein (band 2) matched the behavior of the most abundant mRNA splice form (Tnnt3 β3; Fig. 2B) in regard to its relationship with body mass (i.e. increasing at low body mass and leveling off above 250 g). The second most abundant TNNT3 protein, band 4, and the second most abundant mRNA splice form, Tnnt3 α1 (Fig. 2A), in large rats showed the same linear increase with body mass. These trends associated with body mass increase were evident when variation associated with protein loading was taken into account as a three-way ANOVA (i.e. with TNNT3 band relative abundance as responses, and total TNNT3 band density, β-actin density and body mass as factors) revealed significant effects of body mass only (supplementary material Table S1).

Fig. 4.

Experimental manipulation of body weight did not affect metabolism, locomotor activity or energy expenditure. (A) Mean respiratory exchange ratio (RER, i.e. ) as a function of time for weight-loaded and control rats maintained in indirect calorimetry chambers for 5 days, during the last ∼50 h of the experiment (N=16 rats). Shaded areas indicate the dark period of the 12 h light/dark cycle. Statistical analyses of the effects of weight loading on energy expenditure and activity (see also Table 3) were performed on data obtained during the last full day (i.e. animals had been weight loaded for 4 days). (B) Comparison of mean RER (0.88±0.009 vs 0.86±0.01), mean (2589.3±312.2 vs 2514.2±310.2 ml h^–1) and mean ambulatory activity (139.2±13.8 vs 150.2±18.5 counts 15 min^–1) during day 4. No differences were detected between weight-loaded and control animals. Means are presented with s.e.m. (error bars).

Fig. 5.

Body composition and Tnnt3 splicing response to changes in body weight in lean and obese Zucker rats. Body composition and Tnnt3 relative abundance data for Zucker rats are superimposed on data for Sprague–Dawley rats. (A) Body fat and lean mass as a function of body mass in male lean and obese Zucker rats (N=18). (B) Lean mass contribution to the total weight experienced by skeletal muscles was similar in obese Zucker and weight-loaded Sprague–Dawley rats, indicating that muscles were loaded equally in these groups. (C) Comparison of relative abundance for 3 out of 12 Tnnt3 splice forms between obese and lean Zucker rats across a range of total mass (N=22; see Table 3 for statistical analyses of all Tnnt3 splice forms). The splicing response to changes in total mass for Tnnt3 α1 is unaffected by obesity. Impaired responses were observed for Tnnt3 β3 and β5 (and β1, β8 and the first principal component characterizing the overall mixture; see Table 3). The Sprague–Dawley data are shown here only for reference and were excluded from all linear regression fits.

Fig. 6.

Weight-associated changes in TNNT3 protein band abundance in Sprague–Dawley and Zucker rat gastrocnemius muscle. (A) A western blot for TNNT3 detected four bands in small rats (i.e. 149–164 g), whereas the large Sprague–Dawley and lean Zucker rats (i.e. 332–370 g) displayed three bands. In large obese Zucker rats (i.e. 351–388 g), an expression pattern similar to that of small rats (i.e. four bands, indicated by black arrowheads) was observed. Upper and lower main panels originate from the same western blot, but are presented separately. β-Actin levels (obtained after stripping and reprobing the same western blot) are provided to demonstrate relatively uniform protein loading. β-Actin levels were not significantly affected by changes in body weight (one-way ANOVA: F=1.54, P=0.243). The small panel on the lower right of part A originates from a separate western blot and differs in exposure from the two main panels. This panel provides a more readily perceivable visualization of the small rat-like band pattern in large obese Zucker rats, but these band intensities were not used for quantification. (B) Relative abundance of TNNT3 protein bands in relation to body mass for Sprague–Dawley rats and lean and obese Zucker rats. Data fits are provided to demonstrate trends only, as sample sizes are insufficient for statistical inference. Large obese Zucker rat TNNT3 band relative abundance differed from that of Sprague–Dawley and lean Zucker rats for bands 2 and 3 in particular.

Fig. 7.

Comparison of body weight-dependent troponin T alternative splicing in insects and mammals. Insects [upper panels; graphs are modified from data originally published in Marden et al. (Marden et al., 2008)] and mammals (lower panels; data from this study) show very similar body weight-dependent reaction norms for the troponin T mRNA splice form profile. In both taxa the response is identical for growth-related changes in body mass (open circles) and experimental weight loading (filled circles). There is not a direct homology of the exon structure of the alternative troponin T mRNA transcripts for which relative abundance is shown here; precisely how these molecular variations affect function in either taxa remains to be determined.