Introduction

Skeletal muscle contraction is facilitated by the translocation of actin thin filaments by myosin. This process is coupled to ATP hydrolysis and product release (Sweeney and Houdusse 2010; Houdusse and Sweeney 2016). In mammals, myosin-actin interactions are regulated at two levels: release of Ca2+ from the sarcoplasmic reticulum (SR) during excitation–contraction coupling (ECC) results in Ca2+ binding to troponin C (Gordon et al. 2000; Lehman 2016). Subsequent structural changes allow for the movement of tropomyosin on actin, revealing the myosin-binding sites and permitting interaction (Gordon et al. 2000; Lehman 2016). At the same time, the number of myosin heads extending from the thick filament backbone to bind to actin is regulated at the level of the thick filament; this occurs through thick filament mechanosensing (Linari et al. 2015; Fusi et al. 2016), and also likely through interfilament communication mechanisms (Woodhead & Craig 2015; Irving 2017), which remain largely unknown. In addition to these main regulatory pathways, modulatory mechanisms can also affect contraction.

A main modulatory mechanism is phosphorylation of the regulatory light chain (RLC) of myosin (Sweeney et al. 1993; Vandenboom 2017). RLC phosphorylation is mediated by skeletal Myosin Light Chain Kinase (skMLCK), which in turn is activated by a Ca2+–calmodulin complex (Stull et al. 2011). This cascade is initiated during ECC, but occurs on a slower timescale than the contractile events (Stull et al. 2011). Due to the slow rates of skMLCK inactivation and RLC dephosphorylation, phosphorylation can be cumulative over multiple Ca2+ release events and thus act as a “molecular memory” mechanism (Stull et al. 2011). In permeabilized mammalian fibers, RLC phosphorylation results in increased Ca2+ sensitivity of steady-state force at submaximal but not maximal [Ca2+] (Persechini et al. 1985; Sweeney and Stull 1986, 1990; Metzger et al. 1989; Stephenson & Stephenson 1993; Szczesna et al. 2002), as well as rate of force redevelopment at intermediate [Ca2+] (Metzger et al. 1989; Sweeney and Stull 1990) (c.f. Szczesna et al. 2002). In intact skeletal muscle, RLC phosphorylation is the main mechanism of post-tetanic potentiation (PTP) (Zhi et al. 2005) i.e., the increased twitch force observed following a tetanic stimulus (Close and Hoh 1968). Extensor digitorum longus (EDL) muscles from skMLCK knockout mice (skMLCK−/−), which do not exhibit stimulation-induced increases in RLC phosphate content, display either completely ablated (Zhi et al. 2005) or attenuated (Gittings et al. 2011; Overgaard et al. 2022) isometric PTP. The discrepancy between the initial observation of ablated PTP by Zhi et al. (2005) and the decreased PTP observed later in our lab (Gittings et al. 2011; Overgaard et al. 2022 and findings here) is likely related to the different stimulation protocols utilized. Importantly, this remnant PTP observed in the skMLCK−/− EDL (Zhi et al. 2005; Gittings et al. 2011, 2017; Bowslaugh et al. 2016) and other RLC-phosphorylation void models, such as wildtype (WT) mouse lumbrical muscles (Smith et al. 2013), indicates that additional mechanisms exist. Structurally, RLC phosphorylation disrupts the folded-back conformation of myosin heads on the thick filament backbone, known as the “interacting heads motif” (reviewed in Alamo et al. 2017, 2018), thus radially displacing the heads towards the thin filament (Levine et al. 1996; Yang et al. 1998; Yamaguchi et al. 2016). This has been demonstrated in isolated mammalian thick filaments (Levine et al. 1996; Yang et al. 1998) and permeabilized fibers (Yamaguchi et al. 2016), and is thought to result in enhanced probability of myosin-actin interactions, explaining the observed effects in functional studies (Levine et al. 1996; Yang et al. 1998).

Post-tetanic potentiation is inversely related to muscle or sarcomere length (Rassier et al. 1997, 1998; Rassier and MacIntosh 2000, 2002a; Rassier and Herzog 2002) (c.f. Moore and Persechini 1990). This relationship has been attributed to the aforementioned structural effects of RLC phosphorylation. Interfilament lattice spacing (i.e., the distance between thin and thick filaments) is decreased at longer sarcomere lengths (Millman 1998), and it has been proposed that this makes the displacement of the heads redundant (Yang et al. 1998; MacIntosh 2010). The phenomenon of length-dependent activation (LDA) might also explain the length-dependence of potentiation; LDA refers to the increased Ca2+ sensitivity observed at longer sarcomere lengths in striated muscle (Rassier et al. 1999; de Tombe et al. 2010). The mechanisms of LDA remain unclear, but likely include both the reduction in interfilament lattice spacing and structural mechanisms downstream of titin mechanosensing (Rassier et al. 1999; de Tombe et al. 2010; Williams et al. 2010, 2013; Mateja et al. 2013; Ait-Mou et al. 2016; Li et al. 2016; Zhang et al. 2017); in skeletal muscle, the latter might be related to partial thick filament activation (Reconditi et al. 2014; Fusi et al. 2016), as well as effects of other sarcomeric components like Myosin-Binding Protein-C (MyBP-C) (Reconditi et al. 2014). Thus, RLC-phosphorylation mediated increases in Ca2+ sensitivity would be expected to be less impactful at longer sarcomere lengths.

The length-dependence of potentiation has not been previously investigated in the context of RLC-phosphorylation independent potentiation. To this end, we utilized skMLCK−/− and WT mouse EDL muscles, and assessed PTP at a range of muscle lengths. As mentioned above, despite the absence of increased RLC phosphorylation following stimulation, skMLCK−/− EDL muscles still display a smaller amount of potentiation. We hypothesized that potentiation length-dependence would be attenuated in skMLCK−/− compared to wildtype muscles. In turn, this would become apparent as a significantly different pattern of length-dependence of potentiation between the two genotypes.

Methods

All experimental procedures were approved by the Brock University Animal Care and Use Committee. Wildtype mice with a C57BL/6 background (male and female, aged 10–24 weeks) were obtained from Charles River Laboratories (St. Constant, QC), while age-matched skMLCK−/− mice were obtained from our on-site colony (for information regarding the generation of the skMLCK−/− mouse see Zhi et al. 2005). Body mass was similar between genotypes (22.7 ± 1.1 g for WT, 23.9 ± 0.6 g for skMLCK−/−, mean ± SEM) (p > 0.05). Prior to the initiation of an experiment, mice were anaesthetized via inhalation of isoflurane gas and euthanized by means of cervical dislocation. Subsequently, both EDL muscles were excised and attached either vertically to a jacketed organ bath of the experimental apparatus (Model 1200A, Aurora Scientific Inc., Aurora, ON), or to a resting bath, using surgical silk suture. In both cases, muscles were incubated in continuously gassed (95% O2, 5% CO2) Tyrode’s solution (in mM: 121 NaCl, 24 NaHCO3, 5 KCl, 0.34 NaH2PO4, 0.23MgCl, 1.8 CaCl, 5.5 D-glucose, 0.07 EDTA). Experiments were done at 25 °C, and stimulation voltage was set at 1.25 the threshold required for maximal twitch force production (25–80 V depending on the muscle). Pulse duration was 0.1 ms for all the stimuli used during the experiments. Following suspension, each muscle underwent an equilibration period (≥ 30 min), with one twitch elicited every 3 min. Data were collected (1000 Hz sampling frequency) and monitored using Aurora Scientific’s 600A software (Aurora Scientific Inc., Aurora, ON). Contractile forces (either twitch or tetanic, in mN), + df/dt and − df/dt (i.e., rates of force development and relaxation, respectively, in mN/ms) were determined directly from the 600A analysis function. + Df/dt and − df/dt refer to peak values observed during a given twitch. Time to peak tension (TPT, in ms) and half-relaxation time (½ RT, in ms) were calculated from the raw data using custom Microsoft Excel (Microsoft Corp., 2018) spreadsheets.

Determination of optimal length (L o)

Following equilibration, doublets (3 ms spacing) were utilized to approximate optimal length for tetanic force (see Rassier and MacIntosh 2002b). An initial stimulus was given at 5 mN passive tension and subsequently at 0.5 mN intervals above and below this value, with 20 s between successive doublets. When maximal active force was detected (total force—passive force prior to initiation of stimulation), muscle length was measured using digital vernier calipers and defined as optimal length (Lo). In cases where force values were similar between lengths, the shorter length was selected. Baseline twitch force at Lo was then measured (mean of two twitches).

Experimental protocol

Post-tetanic potentiation (PTP) was assessed at five relative muscle lengths (0.90 Lo, 0.95 Lo, Lo, 1.05 Lo, 1.10 Lo). The experimental protocol included an initial 200-s isometric twitch pacing period (one twitch every 20 s). For each contractile property, the mean of the last 3 twitches of the pacing period was considered as the “Pre” (i.e., unpotentiated) value. Thirty s after the pacing period, a conditioning stimulus was administered, consisting of 4 × 100 Hz, 400 ms isometric tetani within a 20-s window. Importantly, the CS was done at the same muscle length as the prior and following twitches, and not always at Lo. Peak tetanic force (Po) was defined as the highest active force value (total force—passive force prior to initiation of stimulation) recorded during the CS, at each length. Finally, at 10 and 30 s following the 20-s CS window, twitches were elicited to assess potentiation. Values from these twitches were considered as the “Post 10” and “Post 30” values respectively, for all contractile properties. To assess changes following the CS, post values were divided by pre values and expressed as % change (i.e., Post/Pre × 100%). Each muscle underwent this protocol at every experimental length in a randomized order, with 20 min of rest between lengths to allow for the effects of potentiation and fatigue to dissipate (see Fig. 1 for a visual summary of the main experimental protocol). After the end of this process, muscles were taken to Lo and went through a 30-min rest period, with one twitch elicited every 3 min. Subsequently, twitch force was assessed again (mean of two twitches) and compared to baseline values. Muscles were excluded from analysis if active force had declined by > 5%.

Fig. 1
figure 1

Overview of the main experimental protocol. Initially, there was a 200 s isometric twitch pacing period (1/20 s) (i.e., Pre). Thirty seconds following the end of pacing, a conditioning stimulus (CS) was elicited, comprising 4 × 100 Hz, 400 ms isometric tetani within a 20 s window. Isometric twitches were elicited 10 (Post 10) and 30 (Post 30) s after the CS 20-s window to assess potentiation. Each muscle underwent this protocol at five relative muscle lengths (0.90 Lo, 0.95 Lo, Lo, 1.05 Lo, 1.10 Lo) in a randomized order, with 20 m of rest between successive lengths. Please note that the CS and resting period took place at the same relative muscle length as the Pre and Post twitches in each case, and not always at Lo. Pre values for each measured variable were defined as the mean of the last three twitches of the pacing period, in each case

Full size image

Statistical analysis

Body mass was compared between genotypes using an independent samples t-test. A two-way mixed ANOVA was utilized for Po, with muscle length and genotype as the factors. For twitch force (Pt) and all other contractile properties as well as their potentiation values, three-way mixed ANOVAs were used with muscle length, genotype and time as the factors. Data were assessed for existence of outliers through boxplot inspection. Outliers detected were included in the final analyses, as no differences were apparent regarding the significance of interactions and main effects when tests were repeated without them. Normality was evaluated using Shapiro–Wilk’s test in all cases. Homogeneity of variance for the between-subjects factor was assessed using Levene’s test. Both for normality and homogeneity of variance, violations were noted in some cells of the design for all measured dependent variables. These were always in a minority of the cells, and the decision was made to carry on with the analyses. The tables of the assumption tests are provided in Online Resource 1, so the interested reader can assess our decisions. Mauchly’s test of sphericity was used where appropriate and a Greenhouse–Geisser correction was applied in cases where the assumption was violated. Post-hoc pairwise comparisons with Bonferroni corrections, or polynomial contrasts for trend analysis were utilized to further evaluate significant simple main effects or main effects. For each significant trend (linear, quadratic, cubic or higher order), its sum of squares was divided by the sum of squares of the total observed trend in each case, to assess the percentage of variance it could explain. Multiple simple main effect testing within a given interaction was also controlled for with Bonferroni corrections. Significance level was α = 0.05 and data are reported as mean ± SEM (standard error of the mean). All analyses were done in IBM SPSS Statistics for Windows, versions 27 and 28 (IBM Corp., Armonk, NY, USA).

Results

Po and unpotentiated (i.e., Pre) Pt were not significantly different between WT and skMLCK−/− muscles at any muscle length (both p > 0.05) (e.g., Pre Pt at Lo was 51.3 ± 3.4 mN for WT and 48.9 ± 3.7 mN for skMLCK−/−). In contrast, while both genotypes exhibited PTP, WT Post 10 and Post 30 Pt were significantly greater compared to skMLCK−/− values at every muscle length (both p < 0.001) (e.g., 16.9 ± 1.6% potentiation for WT compared to 7.3 ± 1.5% for skMLCK−/− muscles at Lo, at Post 10). In both genotypes, Post 10 Pt values were greater than Post 30 Pt values, again at every muscle length (all p < 0.001) (e.g., at Lo, Post 10 potentiation was 16.9 ± 1.6% for WT and 7.3 ± 1.5% for skMLCK−/−, while Post 30 potentiation was 13.9 ± 1.4% for WT and 2.77 ± 1.2% for skMLCK−/−). All the above data are summarized in Table 1.

Table 1 Summary force data for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles
Full size table

Length-dependence of potentiation

PTP was dependent on muscle length in the WT muscles (p < 0.001 at both Post 10 and Post 30). For example, PTP was 44.3 ± 4.6% at 0.9 Lo compared to 16.9 ± 1.3% at 1.10 Lo, at Post 10. This was also observed in the skMLCK−/− muscles (p < 0.001 at both Post 10 and Post 30); PTP was 33.5 ± 6.2% at 0.9 Lo compared to 9.1 ± 1.8% at 1.10 Lo, at Post 10 (see Table 1 and Fig. 2) Contrary to our initial hypothesis, the pattern of potentiation length-dependence was not significantly different between genotypes, (p > 0.05 for the muscle length x genotype x time and muscle length x genotype interactions) (see Fig. 3). Additionally, trend analysis indicated that this pattern was similar between Post 10 and Post 30, and not distinctly linear; for example, at Post 10, there were significant linear, quadratic and cubic trends (all p < 0.001), which could explain 55.6%, 34.4% and 9.8% of the observed variance in potentiation with muscle length, respectively. These data are presented in Table 2.

Fig. 2
figure 2

Representative twitch traces from WT (left) and skMLCK−/− (right) mouse EDL muscles, at 0.90 Lo (top), Lo (middle) and 1.10 Lo (bottom). Post 10 twitches (solid lines) are superimposed on Pre twitches (dotted lines) to demonstrate potentiation. Force is normalized to Pre maximum values in each case. Potentiation was greater in WT muscles at every muscle length. In both genotypes, potentiation was greater at 0.90 Lo compared to the other two lengths shown

Full size image
Fig. 3
figure 3

Relative force (i.e., potentiation)—relative muscle length relationship for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles. WT values were significantly greater than skMLCK−/− values at both timepoints and at every muscle length. For both genotypes, Post 10 values (triangles) > Post 30 values (squares) at every muscle length. The length-dependence of potentiation was not different between genotypes (time × muscle length × genotype and muscle length × genotype interactions were not statistically significant). Additionally, potentiation length-dependence was similar between Post 10 and Post 30, as indicated by trend analysis (see Table 2). Error bars represent SEM. *Significantly different than skMLCK−/− at the same time point and muscle length, p < 0.05, †Significantly different than corresponding Post 10 value within genotype, p < 0.001

Full size image
Table 2 Trend analysis table for relative force (i.e., potentiation)
Full size table

Force development and relaxation kinetics

Unpotentiated (i.e., Pre) + dF/dt was not significantly different between genotypes at any length (p > 0.05) (e.g., 6.3 ± 0.4 mN/ms for WT and 6.3 ± 0.3 mN/ms for skMLCK−/− at Lo). Post 10 and Post 30 values were significantly increased compared to Pre values for both genotypes at every length (all p < 0.001), but were significantly greater for WT muscles at both timepoints and at all lengths (p = 0.014 for Post 10 and p = 0.007 for Post 30) (e.g., 17.5 ± 1.7% potentiation for WT and 10.6 ± 1.6% potentiation for skMLCK−/− at Lo, at Post 10) (Fig. 4). -DF/dt, while also similar between genotypes at Pre across lengths (p > 0.05) (e.g., -2.8 ± 0.2 mN/ms for WT and -2.6 ± 0.3 mN/ms for skMLCK−/− at Lo), was increased to the same extent (p > 0.05) in both WT and skMLCK−/− muscles at both Post 10 and Post 30, again at all lengths (all p < 0.001) (e.g., 28.9 ± 4.3% potentiation for WT and 40.7 ± 5.7% for skMLCK−/− at Lo, at Post 10) (Fig. 5). ½ RT was not different at Pre between WT and skMLCK−/− muscles at any length (p > 0.05) (e.g., 13.7 ± 0.5 ms for WT and 14 ± 0.4 ms for skMLCK−/− at Lo). Following the CS, it was decreased at both Post 10 and Post 30, with no significant differences between genotypes (both p > 0.05), again at every length (all p < 0.001) (e.g., 11 ± 0.3 ms for WT and 9.7 ± 0.4 ms for skMLCK−/− at Lo, at Post 10) (Fig. 6). Pre TPT was significantly greater at every length in WT muscles (main effect of genotype, p < 0.05) (e.g., 17.1 ± 0.3 ms for WT and 16.6 ± 0.3 ms for skMLCK−/− at Lo). At Post 10 and Post 30, it was decreased in both genotypes (all p < 0.001) and values remained significantly higher for WT muscles at all lengths (e.g., 16.1 ± 0.2 ms for WT and 15 ± 0.3 ms for skMLCK−/− at Lo, at Post 10) (Fig. 7).

Fig. 4
figure 4

+dF/dt–relative muscle length relationship for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles. Unpotentiated values were not significantly different between genotypes (p > 0.05). Post 10 (p = 0.014) and Post 30 (p = 0.007) values were significantly greater in WT compared to skMLCK−/− muscles, at every muscle length. In addition, within each genotype, Post 10 values (triangles) > Post 30 values (circles) > Pre values (squares) at every muscle length (all p < 0.001). Error bars represent SEM. *Significantly different than skMLCK−/− at the same time point and muscle length, p < 0.05, †Significantly different than corresponding Post 10 value within genotype, p < 0.001

Full size image
Fig. 5
figure 5

− dF/dt—relative muscle length relationship for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles. There were no significant differences between genotypes at any time point and muscle length (p > 0.05). At every muscle length, Post 10 values (triangles) < Post 30 values (circles) < Pre values (squares) (all p < 0.001). Error bars represent SEM. *Significantly different than corresponding Pre value within genotype, p < .001, †Significantly different than corresponding Post 10 value within genotype, p < 0.001

Full size image
Fig. 6
figure 6

Half-relaxation time (½ RT)–relative muscle length relationship for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles. There were no significant differences between genotypes at any time point or muscle length. Post 10 values (triangles) and Post 30 values (circles) were significantly different than Pre values (squares) at every muscle length (all p < 0.001), but they were not significantly different from each other (all p > 0.05). Error bars represent SEM. *Significantly different than Pre value at the same muscle length, p < 0.001

Full size image
Fig. 7
figure 7

Time to peak tension (TPT)–relative muscle length relationship for WT (top, n = 11) and skMLCK−/− (bottom, n = 12) mouse EDL muscles. WT values were significantly greater than skMLCK−/− values at all time points and muscle lengths (main effect of genotype: p < 0.05). Within each genotype, Post 10 values (triangles) and Post 30 values (circles) were significantly greater than Pre values (squares) at every muscle length (all p < .001) but they were not significantly different from each other (all p > 0.05). Error bars represent SEM. *significantly different than Pre within genotype, p < 0.001

Full size image

The length-dependence of force kinetics properties was assessed through trend analysis. + Df/dt displayed a curvilinear relationship with muscle length, with values decreasing above and below Lo at all timepoints (quadratic trend could explain 94.3%, 97.5% and 99.7% of the total variance at Pre, Post 10 and Post 30, respectively; all p < 0.001) (Table 3); − dF/dt was maximal at 0.95 Lo and decreased above and below it, both before and after the CS. The − dF/dt—muscle length relationship was dominated by linear trends at all timepoints (linear trend could explain 74.8%, 92.2% and 90.2% of the total variance at Pre, Post 10 and Post 30, respectively; all p < 0.001) (Table 4). ½ RT and TPT increased with increasing muscle length, and this relationship was similar at all time points. For ½ RT, linear trends could explain 74.8%, 97.3% and 97% of the total variance at Pre, Post 10 and Post 30, respectively (all p < 0.001) (Table 5), while for TPT, there was no significant time x muscle length interaction (p > 0.05) and at all timepoints the linear trend could explain 99.4% of the total variance (p < 0.001) (Table 6).

Table 3 Trend analysis table for + dF/dt
Full size table
Table 4 Trend analysis table for − dF/dt
Full size table
Table 5 Trend analysis table for ½ RT
Full size table
Table 6  Trend analysis table for TPT
Full size table

Discussion

The main finding of this study was that the length-dependence of post-tetanic potentiation was not significantly different between WT and skMLCK−/− mouse EDL muscles. Previous findings in rat gastrocnemius in situ (Rassier et al. 1997; 1998; Rassier and MacIntosh 2000) and mouse EDL fiber bundles in vitro (Rassier and Herzog 2002; Rassier and MacIntosh 2002a) have consistently demonstrated that both PTP and staircase potentiation are dependent on muscle or sarcomere length, with potentiation diminishing as length is increased (c.f. Moore and Persechini 1990). However, all previous works have utilized wildtype models, and the length-dependence of RLC-phosphorylation independent potentiation had not been explored. Here, using skMLCK−/− EDL muscles we demonstrated that even though the absence of RLC phosphorylation results in lower potentiation magnitudes across the examined range of muscle lengths, it does not appear to alter its length-dependence, at least under the experimental conditions utilized. This finding extends prior knowledge on the alternative mechanisms of PTP, and provides additional information regarding potentiation length-dependence in general. Although we did not directly assess RLC phosphorylation here, between-genotype differences at rest and following conditioning stimuli similar to the one used in the current study have been documented repeatedly by our lab (Gittings et al. 2011, 2016, 2018; Bowslaugh et al. 2016; Bunda et al. 2018; Fillion et al. 2019; Morris et al. 2018). SkMLCK−/− EDL muscles consistently display reduced resting RLC phosphate content compared to WT muscles (e.g., Gittings et al. 2011; Bunda et al. 2018), and an absence of post-stimulation increases. In contrast, WT muscles typically display 2–4fold increases post-CS compared to baseline (e.g., Gittings et al. 2011, 2016). A difference between these data and our current experiments is that here we included conditioning stimuli at varying muscle lengths instead of Lo. However, past findings in rat gastrocnemius indicate that RLC phosphorylation is not significantly different across lengths (0.90 Lo, Lo, 1.10 Lo), at least following a staircase protocol (Rassier et al. 1997) (c.f. Moore and Persechini 1990).

The similar length-dependence of potentiation in the presence and absence of RLC phosphorylation is not readily explainable. Two main alternative potentiation mechanisms have been proposed; increased resting Ca2+ following stimulation (Smith et al. 2013, 2014), and S-glutathionylation of troponin-I, which increases calcium sensitivity of steady-state force in permeabilized fast, but not slow rat fibers (Mollica et al. 2012; Dutka et al. 2017). These mechanisms may interact with LDA in a currently unknown manner, and thus, regardless of the underlying mechanism, potentiation length dependence might simply be due to a ceiling effect of Ca2+ sensitivity at longer sarcomere lengths, as has been proposed before for RLC-phosphorylation dominant potentiation (Rassier et al. 1998; Rassier and MacIntosh 2000, 2002a; Rassier and Herzog 2002). However, if this were the case, a different length-potentiation relationship would be expected for the skMLCK−/− muscles since the magnitude of RLC-phosphorylation independent potentiation is smaller. Potentiation mechanisms might interact with both thin- and thick-filament related mechanisms of enhanced activation at longer sarcomere lengths (de Tombe et al. 2010; Mateja et al. 2013; Reconditi et al. 2014; Ait-Mou et al. 2016; Li et al. 2016; Zhang et al. 2017), as well as with the effects of spatial changes of the myofilament lattice on unitary actin-myosin interactions (Williams et al. 2010, 2013). As an example, thick filaments become partially activated at long sarcomere lengths in skeletal muscle (Reconditi et al. 2014; Fusi et al. 2016), and interactions of MyBP-C with the thin filaments might be lost (Reconditi et al. 2014). Both these factors could interact with RLC phosphorylation and/or other potentiation mechanisms and result in less potentiation at longer sarcomere lengths. Nevertheless, details regarding the molecular underpinnings of the alternative potentiation mechanisms and LDA are lacking, and any proposed model can currently only be speculative. On the other hand, the fact that potentiation is enhanced at short sarcomere lengths may be related both to increased interfilament lattice spacing (Millman 1998), as has been proposed before (Levine et al. 1996; Yang et al. 1998), and to the inhibited Ca2+ release at short sarcomere lengths that has been observed in mammalian skeletal muscle (Rassier and Minozzo 2016). However, the latter has only been seen with high-frequency stimulation, and it is unclear whether this effect could somehow influence PTP. As with long sarcomere lengths, changes in lattice spacing at short sarcomere lengths and consequent effects on unitary myosin-actin interactions (Williams et al. 2010, 2013) could also be a factor in the length-dependence of potentiation.

Our findings for force kinetics are in agreement with prior literature, both regarding the influence of muscle length (Wallinga-de Jonge et al. 1980; Rassier et al. 1997; Rassier and MacIntosh 2002b) and in relation to genotype differences prior to and following the CS (Vandenboom 2017). The greater post-tetanic increase in + dF/dt in WT compared to skMLCK−/− muscles has been previously observed in our lab for concentric twitches (Gittings et al. 2016) and may be a direct effect of RLC phosphorylation, parallel to increased force (Vandenboom 2017). The current findings complement earlier observations of correlations between post-tetanic increases in isometric + dF/dt and RLC phosphate content in WT mouse EDL (Vandenboom et al. 1995, 1997), and of RLC phosphorylation-mediated increases in rate of force redevelopment (i.e., ktr) in permeabilized mammalian fibers (Metzger et al. 1989; Sweeney and Stull 1990). The enhanced + dF/dt was accompanied by a slight, but significant decrease in TPT in both genotypes. Notably, TPT values were significantly greater in WT muscles both prior to, and following the CS. The reason for this difference is unknown, but it has been previously observed in our lab (Bunda et al. 2018). In regard to − dF/dt, it is known that the post-tetanic increase of twitch relaxation rate does not appear to be related to RLC phosphorylation, as this effect has been observed in mouse lumbrical muscles, which display potentiation without phosphorylation (Smith et al. 2013, 2014). Here, this observation has been recapitulated, as − dF/dt was increased to the same extent in both genotypes following the CS. Similarly, ½ RT was decreased post-CS with no significant differences between genotypes, in accordance with past findings (Gittings et al. 2011). The decreased relaxation time is likely related to the increased − dF/dt in both genotypes, the mechanisms of which remain unknown (Vandenboom 2017).

Limitations

It is known that active force calculation using the traditional method (total force—passive force prior to stimulation initiation, used here) might be problematic in fixed-end contractions, due to internal sarcomere shortening (MacIntosh and MacNaughton 2005; see MacIntosh 2017 for a review). Specifically, in whole muscle preparations sarcomeres are able to pull on in-series elastic components of the experimental apparatus and tendon, and thus passive tension at the peak of force production would be lower than prior to its initiation (MacIntosh 2017). Sarcomere or fascicle length measurements can be used to control for this change (MacIntosh and MacNaughton 2005; de Tombe and ter Keurs 2016), but unfortunately were not available here. This difference in active force calculation can result in effects that are pronounced at long muscle lengths, where passive tension is high: underestimation of active force, overestimation of potentiation due to force relaxation over time (i.e., passive tension would become progressively lower giving the impression of greater force increase than actually occurred) and potential influence of shortening-induced force depression (MacIntosh 2017). While this makes mechanistic interpretation of our findings more difficult, it is not yet clear exactly what the implications of internal shortening are at the molecular level (e.g., MacDougall et al. 2020). Regardless, our findings can be compared to existing potentiation length-dependence literature, as previous works also utilized the traditional method to calculate active force.

Conclusion

The pattern of potentiation length-dependence does not differ significantly in the presence and absence of RLC phosphorylation, with potentiation being greater at short compared to long muscle lengths. While the current findings are not sufficient for a mechanistic interpretation of this similarity, they provide additional information regarding the alternative mechanisms of potentiation. Further work is necessary to understand how potentiation without RLC phosphorylation is facilitated at the molecular level, and how these mechanisms interact with sarcomere length, as well as with RLC phosphorylation when it is present.