Introduction

In 1993 two review articles were published that addressed functional, mechanistic and biochemical features of myosin phosphorylation mediated force potentiation in vertebrate skeletal muscle (Grange et al. 1993; Sweeney et al. 1993). These comprehensive reviews provided important theoretical and practical information for future investigations. Work performed in the past 20 years has greatly expanded our knowledge concerning myosin phosphorylation and force potentiation. Thus, the purpose of this review is to highlight both functional and mechanistic aspects of force potentiation, focusing on mechanistic studies on animal muscle models. This article includes recent evidence from both wildtype and transgenic mouse as well as rat skeletal muscle models expanding the physiological utility of this mechanism for contraction. Existing cross-bridge models for the influence of myosin phosphorylation on mechanical function will be modified to account for these results. The genetic, enzymatic and regulatory aspects of RLC phosphorylation have been recently reviewed by Stull et al. (2011) and will not be detailed in this article.

First described in the literature over 100 years ago (see Lee 1907) force potentiation has long been recognized as a fundamental property of fast twitch skeletal muscle. Potentiation is generally expressed as an increase in isometric twitch force independent of change to peak tetanic force, thus increasing the twitch to tetanus ratio (e.g., Bagust et al. 1974; Ramsey and Street 1941). In contrast, slow twitch skeletal muscle displays a posttetanic depression of isometric twitch force (Buller et al. 1981; Close and Hoh 1969). A list of studies describing potentiation of fast twitch vertebrate striated muscle is compiled in Table 1. Potentiation is readily induced in most fast twitch skeletal muscles studied and as such may be a normal operating feature of these muscles (Brown and Loeb 1998). In the laboratory setting, the precise manifestation of potentiation is highly paradigm dependent, however (Fig. 1). For example, staircase potentiation describes the progressive increase in isometric twitch force observed during a period of low frequency stimulation (e.g., Colomo and Rocchi 1965; Isaacson 1969).

Table 1 Potentiation in isolated skeletal muscles in which RLC phosphorylation was not measured
Full size table
Fig. 1
figure 1

Examples of potentiation in fast twitch skeletal muscle. Typical experimental paradigms producing posttetanic potentiation (PTP) (a) or staircase potentiation (b). In a a single control or unpotentiated isometric twitch is elicited under resting conditions prior to the delivery of a brief, but high frequency conditioning stimulation. The precise time course of PTP depends on the frequency and duration of stimulation; in mouse EDL muscle at 25 °C the peak tends to occur shortly after the conditioning stimulus before decaying over several minutes to pre-stimulus response. Twitch duration may or may not be abbreviated compared to the unpotentiated twitch. In b a series of evenly spaced isometric twitches is elicited at a fixed low frequency (typically 5–10 Hz). In most species, stimulation rates below 5 Hz do not produce staircase while stimulation rates above 10 Hz produce fused contractions. During staircase, an initial modest decrease in twitch force is often observed before progressively increasing. As for PTP, the precise magnitude of staircase observed depends on stimulation rate, species, temperature and muscle length. Note that in both the top and bottom panels the time scale is variable

Full size image

During staircase potentiation, isometric twitch amplitude may initially decrease before increasing to a peak that is dependent upon both stimulus rate and number; upon the cessation of stimulation, twitch force dissipates slowly at first and then with a more rapid time course (Krarup 1981a). On the other hand, posttetanic potentiation (PTP) describes the acute increase in isometric twitch force observed following high (or sometimes low) frequency stimulation (e.g., Close and Hoh 1968b). Like staircase, the posttetanic response is also stimulus frequency and duration dependent, often displaying an initial increase to a peak shortly after the cessation of stimulation before dissipating, rapidly at first and more slowly thereafter, over the course of several minutes (e.g., Krarup 1981a). In addition to stimulus parameters, both staircase and PTP are highly sensitive to experimental factors such as temperature (high>low), muscle length (short>long) and contraction type (concentric>isometric). Even though a single mechanism for staircase and PTP is often assumed, the existence of divergent mechanisms now seems probable.

Although much has been learned regarding how potentiation benefits the mechanical performance of isolated muscles, a definitive teleological role in vivo has not yet been identified. Interesting in this regard is evidence suggesting that the potentiated state is the normal operating state of fast twitch skeletal muscle in vivo (Brown and Loeb 1998). In addition, work on isolated mouse fast twitch skeletal muscles shows that beta-adrenergic stimulation prolongs the potentiated state, an outcome that may indicate a role for potentiation during the fight or flight response of mammals (Decostre et al. 2000). Thus, much remains to be discovered regarding how potentiation integrates into overall organismal function when extra-muscular signaling is intact.

Many early studies attempted to identify an ionic origin for potentiation, with mostly equivocal results (e.g., Brown and von Euler 1938; Bernhard et al. 1941; Walker 1948). The demonstration by Ramsey and Street (1941) that potentiation is a property of isolated skeletal muscle fibers indicated a muscle origin for this phenomenon. Indeed, by the 1960s both staircase and PTP were considered to be intrinsic properties of skeletal muscle (e.g., Standaert 1964) although the precise mechanism was still not known. Interestingly, the observation that the magnitude of PTP was inversely related to the twitch:tetanus ratio was seen as evidence that alterations to the “active state” were responsible (Close and Hoh 1968a; Ritchie and Wilkie 1955). It was not until the demonstration that the myosin II molecule of vertebrate striated muscle contained a phosphorylatable light chain subunit (Perrie et al. 1973) that a viable intracellular mechanism reconcilable with contractile data was available, however. Two small protein subunits, the essential light chain and regulatory light chain (RLC) wrap around the α-helical neck or light chain binding domain of the myosin heavy chain, providing mechanical stability (Rayment and Holden 1994; Trybus 1994). Although skeletal muscle contraction results from the Ca2+ —regulated formation of force-generating actomyosin complexes and the linking of myosin ATPase activity to structural changes in this complex (reviewed by Geeves and Holmes 2005; Vale and Milligan 2000), phosphorylation of the RLC subunit may modify these unitary reactions to enhance muscle cell force and shortening (Sweeney and Stull 1990).

Potentiation by RLC phosphorylation in rodent muscle

Work performed on isolated rat and mouse skeletal muscle demonstrates that isometric twitch potentiation is temporally correlated with RLC phosphate content following various stimulus regimes. A comprehensive list of studies examining the phosphorylation-potentiation relationship in different rodent muscle models is presented in Tables 2 and 3. In studies where multiple measurements of isometric twitch potentiation and RLC phosphorylation have been made, both non- linear (Klug et al. 1982) and linear relationships (Manning and Stull 1979; Moore et al. 1990; Palmer and Moore 1989; Vandenboom et al. 1995, 1997; Xeni et al. 2011) have been documented. However, when data from these respective studies are pooled according to species and temperature in Fig. 2, the resultant scatter plot depicting twitch potentiation versus RLC phosphorylation are decidedly linear. The robustness of these relationships is particularly striking given the inevitable differences in methodology between the different studies used to construct these plots. This analysis thus indicates that RLC phosphorylation has a direct and predictable influence on the isometric twitch force of unfatigued fast twitch rodent skeletal muscle, an effect that increases as muscle temperature approaches the physiological range. Because these results were obtained from mostly unfatigued muscle it is possible that more extreme stimulus regimes causing more severe fatigue will disrupt the linearity of these relationships, especially at high phosphorylation levels. In addition, experimental interventions that alter thin filament activation level may also alter the slope of any of these relationships (Krarup 1981b, c; MacIntosh and Gardiner 1987; MacIntosh and Kupsh 1987; Palmer and Moore 1989; Vandenboom and Houston 1996).

Table 2 Potentiation in isolated rat and rabbit skeletal muscle in which RLC phosphorylation was measured
Full size table
Table 3 Potentiation in isolated mouse skeletal muscle in which RLC phosphorylation was measured
Full size table
Fig. 2
figure 2

Quantitative relationship between stimulation induced isometric twitch potentiation and RLC phosphorylation in rodent skeletal muscle. a Data from studies employing rat gastrocnemius and EDL muscle with separate linear fits to results obtained at 23 and 35–37 °C (r value = 0.79 and 0.82, respectively). b Data from studies employing mouse EDL muscle with separate linear fits to data obtained at 25 and 30 °C (r value = 0.89 and 0.90, respectively). Although the scatter of data is increased, there is a clear effect of increasing temperature on the slope of the RLC phosphorylation vs isometric twitch force potentiation relationship in both species. Note that in both panels we have pooled data from experiments in which the conditioning stimulus was fixed and the decay in RLC phosphorylation and twitch force over time was tracked and in which different conditioning stimuli were used and RLC phosphorylation and twitch force were assessed at a fixed time point. Data in panel A taken from Klug et al. 1982; Manning and Stull 1979, 1982; Moore and Stull 1984. Data in panel B taken from Moore et al. 1990; Palmer and Moore 1989; Vandenboom et al. 1995, 1997; Xeni et al. 2011. Although not included, data from Moore et al. (1990) for mouse EDL muscle at 35 °C also shows a linear response

Full size image

The first direct evidence of a causal relationship between RLC phosphorylation and isometric force potentiation was provided by studies utilizing permeabilized skeletal muscle fibers. A list of these studies is compiled in Table 4. Persechini et al. (1985) were the first to show that the addition of exogenous skeletal myosin light chain kinase (skMLCK) to the media bathing rabbit psoas fibers both phosphorylated the RLC and increased steady state tension at submaximal, but not maximal, Ca2+ activation. Subsequent studies by this and other goups have elegantly documented how the potentiation of steady state force by RLC phosphorylation appears to be inversely related to thin filament activation levels (e.g., Davis et al. 2002; Metzger et al. 1989; Patel et al. 1996, 1998; Sweeney and Kushmerick 1985; Stephenson and Stephenson 1993; Sweeney and Stull 1986, 1990; Szczesna et al. 2002). An important feature of many of these studies is the increase in the rate constant for steady state force development (i.e. the kTR) that occurs following RLC phosphorylation, an effect that may be the mechanical basis for twitch potentiation in intact muscle. These studies have thus provided important mechanistic links between skMLCK catalyzed phosphorylation of the RLC and isometric twitch force potentiation in intact skeletal muscle models. A model for the influence of RLC phosphorylation on cross-bridge kinetics, and interactions with Ca2+ activation of the thin filament, is detailed below.

Table 4 Studies using permeabilized skeletal muscle fibers or in vitro motility assay
Full size table

RLC phosphorylation and cross-bridge structure

Skeletal muscle force is regulated by Ca2+ ion binding to regulatory proteins on the thin filament (Ebashi and Endo 1968). Although the ligand binding role of Ca2+ and regulatory protein conformation on the thin filament have been studied extensively (e.g., Lehman et al. 2009) there may be, in addition, thick filament structural constraints that participate in force regulation. For example, reconstructions of the thin filament of rested skeletal fibers from arthropods and mammals show that myosin head distribution is highly ordered with close proximity to the thick filament surface (Woodhead et al. 2005; Zhao et al. 2009).

Head to head interactions via highly conserved motifs present in both regulated and unregulated myosins may be responsible for this constraint, a profile that may minimize myosin head interactions with inactivated thin filaments (Jung et al. 2008). Interestingly, structural studies employing a variety of fiber types show that the addition of a negative charge to the RLC via phosphorylation disrupts these interactions to displace myosin heads away from the thick filament (Alamo et al. 2008; Brito et al. 2011; Craig et al. 1987; Hidalgo et al. 2001; Levine et al. 1991, 1996, 1998; Padrón et al. 1991; Ritz-Gold et al. 1980; Sweeney et al. 1994; Yang et al. 1998). It is proposed that these phosphorylation-induced changes to myosin head position on the thick filament surface increase the Ca2+ sensitivity of force development by promoting the formation of the actomyosin complex achieved by Ca2+ signalling alone (see below). These structure–function relations are consistent with findings from intact rat skeletal muscle showing that potentiation is greatest at short and minimized at long sarcomere lengths, respectively (e.g., Rassier et al. 1997, 1998; Rassier and MacIntosh 2000) (c.f. Moore and Persechini 1990). Finally, findings from experiments using in vitro motility assays showing that phosphorylation of the RLC does not improve intrinsic myosin motor function in isolation from the thin filament supports these arguments (Greenberg et al. 2009).

RLC phosphorylation and cross-bridge cycling kinetics

Huxley (1957) was the first to provide an analytical framework for understanding cross-bridge cycling in skeletal muscle. In his original scheme Huxley proposed that muscle force was regulated by the cyclic attachment and detachment of “side pieces” between the myofilaments. These rate constants, known as F and G, respectively, were considered intrinsic properties of independently operating force generators within the sarcomere; these structures were, in time, recognized as the actomyosin cross-bridge. Brenner and Eisenberg (1986) and Brenner (1988) advanced this two state model for cross-bridge action by incorporating the regulatory influence of Ca2+ binding to troponin C (TnC) on force. In their modern scheme, Ca2+ binding to TnC controls the rate constant describing the transition of cycling cross-bridges from non-force to force generating states (i.e. f app ). In contrast, the reverse rate constant describing the transition of cycling cross-bridges from force generating back to non-force generating states (i.e. g app ) is mostly unregulated and not under the influence of Ca2+. Thus, although only a fraction of the total available cross-bridge population may be involved, force is proportional to the fraction of cycling cross-bridges able to attain the force generating state (αFS). In turn, this term is determined by the balance between f app and g app according to the equation αFS = f app /(f app  + g app ). Within this scheme the influence of RLC phosphorylation on force may be understood by noting that RLC phosphorylation increases f app at all Ca2+ activation levels with little or no effect on g app (Sweeney et al. 1993). Thus, at low levels of myofilament activation where f app is small relative to g app , the RLC phosphorylation-mediated increases in f app greatly enhance cross-bridge formation (i.e. increase αFS). Conversely, at high levels of myofilament activation where f app is large relative to g app , crossbridge formation is little affected as αFS is already near unitary (Sweeney and Stull 1990). Moreover, the absence of any effect on g app accounts for why unloaded shortening velocity of permeabilized skeletal fibers (Persechini et al. 1985; Sweeney and Stull 1990) or intact skeletal muscle (Butler et al. 1983; Decostre et al. 2000; Gittings et al. 2011; Palmer and Moore 1989) are not altered.

Results from skMLCK knockout models

Perhaps the most compelling evidence from intact skeletal muscles for a causative relationship between RLC phosphorylation and isometric twitch force potentiation comes from experiments using muscles from mice devoid of the skMLCK enzyme and which, as a consequence, do not display stimulation-induced elevations in RLC phosphorylation. In this regard, Zhi et al. (2005) showed that brief tetanic stimulation of EDL muscles from wildtype mice increased both RLC phosphorylation and twitch force by ~4 and 1.8-fold, respectively, but did not elevate either of these parameters in EDL muscles from skMLCK deficient or skMLCK−/− mice. Examples of posttetanic responses from wildtype and skMLCK−/− muscles is shown in Fig. 3. Interestingly, however, low frequency repetitive stimulation of skMLCK−/− muscles still produces a significant staircase potentiation, albeit attenuated by ~50 % compared to wildtype muscles. Indeed, this remnant potentiation in skMLCK−/− muscles accords with the presence of potentiation in muscle disease or atrophy models (MacIntosh et al. 2008a; Rassier et al. 1999). Interestingly, Ryder et al. (2007) showed that overexpression of skMLCK in mouse EDL muscle enhanced the rate of RLC phosphorylation and staircase potentiation relative to wildtype muscles. The absence of PTP but continued presence of staircase potentiation in this genotype suggests that although skMLCK catalyzed phosphorylation of the RLC is the primary mechanism for PTP, additional mechanisms may contribute to staircase potentiation (Zhi et al. 2005; Gittings et al. 2011).

Fig. 3
figure 3

Genotype dependence of potentiation in mouse fast skeletal muscle Comparison of posttetanic potentiation (PTP) of EDL muscles from and wildtype (WT) (a) and skeletal myosin light chain kinase (skMLCK) knockout (KO) (b) muscles (in vitro 25 °C). The influence of a tetanic conditioning stimulus (four-volleys of 100 Hz stimulation, each lasting 400 ms, in a 20 s time window) on isometric twitch force is shown. Note that although still present, the potentiation of twitch force was much greater in WT than skMLCK KO muscles. For example, in the WT muscle the post-twitches were potentiated by up to 22 % relative to the pre-twitch and were potentiated for more than 180 s. In the skMLCK KO muscle, post-twitches were potentiated by only 5 % relative to the pre-twitch and this potentiation lasted for only 30 s. Twitch time course was similar for each genotype regardless of differences in peak force. The time interval between tetani has been compressed for clarity. In both genotypes peak tetanic force was decreased similarly from the first to last tetanic volley. Previously unpublished data

Full size image

The influence of RLC phosphorylation on skeletal muscle mechanics may apply to both force development and relaxation. For example, evidence from skMLCK−/− muscles suggests that, in addition to increases in the rate and/or extent of force development, RLC phosphorylation may influence force relaxation kinetics as well. Gittings et al. (2011) compared high frequency tetanic contractions of EDL muscles from skMLCK−/− and wildtype mice following a stimulus protocol that elevated RLC phosphorylation in wildtype muscles only. They found that, despite no changes in peak tetanic force in either genotype, the relaxation rate of wildtype muscles was significantly slowed relative to skMLCK−/− muscles (in vitro, 25° C). This outcome corroborates findings from both intact cat (Brown and Loeb 1999) and permeabilized rabbit psoas skeletal fibers (Patel et al. 1998) but the functional significance of this effect is unknown. Moreover, although an increased rate of relaxation of potentiated twitch contractions is often observed, this effect may not be attributable to RLC phosphorylation as skMLCK−/− muscles also show this effect (Gittings et al. 2011).

An interesting corollary to the results from skMLCK−/− mice are results from insect flight muscle in which mutations to the serine residue have rendered the RLC unphosphorylatable. In these experiments, the power and flight characteristics of Drosophila melanogaster were attenuated related to wildtype flight characteristics (Dickinson et al. 1997; Miller et al. 2011; Tohtong et al. 1995).

Stimulus frequency and contraction type dependence for potentiation

Studies performed prior to the 1990s tended to study potentiation using single pulse, twitch contractions. Although single pulse contractions have been shown to represent ~30 % of all motor unit discharge events in hindlimb muscles of freely moving rats (Hennig and Lømo 1987), an influence of potentiation on higher frequency, multiple pulse (i.e. tetanic) contractions would greatly extend its physiological utility. The first to systematically study the stimulus frequency dependence for isometric force potentiation was Vandenboom et al. (1993) using a mouse EDL in vitro (25 °C) muscle model. These investigators showed that although a low frequency conditioning stimulus that elevated RLC phosphorylation to near maximal levels increased the rate of isometric force development at all frequencies between 1 and 200 Hz, the potentiation of peak force was restricted to frequencies below 20 Hz. Indeed, peak tetanic forces measured at higher frequencies were decreased rather than increased. It is important to point out that any threshold that is identified for peak force potentiation must be highly model dependent. As an example of this, more recent work from rat skeletal muscle in situ (35 °C) showed that potentiation of peak isometric force may in fact be observed at high stimulus frequencies (MacIntosh and Bryan 2002; MacIntosh and Willis 2000). This apparent difference may be related to the tetanus duration used to examine high frequency force; the enhanced +dF/dt that is a general characteristic of potentiated rodent muscle (Grange et al. 1995) may be able to enhance the peak force that is attained during brief, but not necessarily long, duration tetani.

Another vital mechanical factor that modulates the ability of RLC phosphorylation to potentiate muscle function is that of contraction type, i.e. isometric versus concentric. A host of studies performed on rat and mouse skeletal muscle demonstrates that the same stimulus regimes that potentiates isometric function also potentiate dynamic function (Abbate et al. 2000; Caterini et al. 2011; Gittings et al. 2012; Grange et al. 1995, 1998; MacIntosh et al. 2008b; Xeni et al. 2011). Evidence from these studies indicates that muscle shortening during concentric force development may actually sensitize the contractile apparatus to potentiation; as a result, dynamic force levels observed during shortening are potentiated to a much greater extent than might be predicted based on isometric responses alone. Representative force records depicting the potentiation of concentric forces during tetanic stimulation of wildtype and skMLCK−/− muscles are shown in Fig. 4. The first study showing this effect was Grange et al. (1995) who used mouse EDL muscles (in vitro, 25 °C) to show that work and power during after—loaded twitch contractions were enhanced more than isometric twitch force levels (see also Grange et al. 1998). Many years later, the sensitizing influence of muscle shortening was confirmed by Xeni et al. (2011) who showed that concentric twitch force was potentiated more than isometric twitch force when RLC phosphorylation levels were similar. Subsequent work from our laboratory has shown that the influence of muscle shortening speed on potentiation is progressive, i.e. twitch forces during fast, moderate and slow shortening were increased more than isometric twitches for the same increase in RLC phosphorylation (Caterini et al. 2011) an effect that also applies to higher frequency forces observed during partially fused tetani (Gittings et al. 2012). A speed-dependent increase in potentiation has also been found in rat gastrocnemius muscle studied in situ, suggesting that this response is a general feature of muscle function rather than just a species specific response (Abbate et al. 2000).

Fig. 4
figure 4

Influence of muscle shortening on potentiation in mouse muscle. Comparison of concentric force potentiation of EDL muscles from WT (a) and skMLCK knockout (b) mice. Each panel shows concentric force during shortening at 25 % Vmax while stimulated at 45 Hz for 100 ms before (thick trace) and after (thin trace) a tetanic conditioning stimulus (4 volleys of 100 Hz stimulation each lasting 400 ms). This maneuver increased mean concentric force by ~40 and by 20 % in the WT and skMLCK knockout muscles, respectively. In each panel, the horizontal line depicts peak force level during the tetanus before the conditioning stimulus. Previously unpublished data

Full size image

Work on intact frog skeletal fibers by Piazzesi et al. (2007) examining the influence of filament sliding on cross-bridge cycling kinetics (i.e. f app and/or g app ) may be able to account for the speed dependence for potentiation at the sarcomeric level. For example, their experiments show that while filament sliding may increase both fapp and gapp relative to isometric, the increase in gapp much exceeds the increase in fapp at both moderate and high, but not slow, speeds of shortening. Thus, the effect of shortening to decrease the number of attached cross-bridges (i.e. αFS) is consistent with a speed dependence for concentric force potentiation. Indeed, the relationship between fapp and gapp shown in Fig. 4d of Piazzesi et al. (2007) may be able to account for why although very slow shortening does not greatly increase the potentiation of concentric compared to isometric force, moderate speeds of shortening do greatly increase concentric compared to isometric force at most activation levels (i.e. Gittings et al. 2012).

Potentiation and locomotion

Although potentiation of isometric contractions may be physiologically relevant, the stimulus frequency and shortening speed dependence for potentiation described above has important ramifications for skeletal muscle function in vivo. During locomotion, for example, skeletal muscle must generate concentric work and absorb eccentric work on a cyclic basis (reviewed by Josephson 1993). Because concentric forces are potentiated in a speed dependent manner it seems possible to suggest that faster speeds of locomotion benefit in two ways: i.e., motive forces produced by agonists may be increased at each stimulus frequency and/or the stimulus frequency range over which motive forces produced by agonists is greatly increased relative to that observed during wholly isometric contractions. This effect would both increase work and power output of potentiated muscles under a wide variety of activation envelopes and also decrease the neural input required to achieve a given submaximal level of muscle work and power (as demonstrated in human skeletal muscles during isometric contractions by Klein et al. 2001; Inglis et al. 2011). On the other hand, because eccentric forces are largely unaffected, the forces required by antagonists may be largely unchanged during cyclic muscle activity typical of locomotion. Interesting in this regard are the results of Childers and McDonald (2004) who used permeabilized rabbit psoas skeletal fibers to show that RLC phosphorylation exacerbated damage caused by active lengthening despite the fact that eccentric forces were not increased.

Fiber phenotype and potentiation

The fiber type dependence for potentiation was first demonstrated by elegant studies showing reduced PTP in muscles exposed to chronic stimulation (Close and Hoh 1969). This fundamental difference is accounted for, at least in part, on the basis of differences in skMLCK and myosin light chain phosphatase (MLCP), the enzyme responsible for dephosphorylating the RLC. In general, skMLCK content is highest in fast-glycolytic and lowest in slow-oxidative fiber types of rodents; on the other hand, MLCP content is highest in slow-oxidative and lowest in fast- glycolytic fibers (Stull et al. 2011). Moreover, chronic stimulation of rabbit tibialis anterior muscle reduces skMLCK expression in a time dependent manner that anticipates attendant changes in myosin heavy chain isoform expression, indicating that the skMLCK enzyme is part of the fast, but not slow, muscle genetic program (Klug et al. 1986, 1992). Phenotypical differences in contractile response to stimulation, highlighting differences of fast and slow muscles from mouse from our laboratory, are shown in Fig. 5.

Fig. 5
figure 5

Phenotype dependence of potentiation in mouse fast and slow skeletal muscle. a Comparison of posttetanic responses of EDL (left) and soleus (right) muscles. (Top panels) The effect of a brief tetanic conditioning stimulus (four volleys of 100 Hz stimulation, each lasting 400 ms, in a 20 s time window) to cause posttetanic potentiation of EDL and posttetanic depression of soleus muscles from wildtype mice. (Bottom panels) The effect of the same conditioning stimulus on EDL and soleus muscles from skMLCK KO mice. In this case PTP of EDL is not evident but PTD of soleus is still present. b Western blots showing phosphorylation of RLC in wildtype EDL and soleus muscles before and after a tetanic conditioning stimulus. Note that in these examples, the RLC phosphate content of wildtype (WT) EDL muscles was increased significantly from rest by stimulation and PTP was evident. In contrast, the RLC phosphate content of WT soleus muscles was similar before and after stimulation and PTP was absent. The RLC phosphate content of skMLCK KO muscles was not determined in these experiments. Horizontal arrows at blots depict unphosphorylated (skRLC) and phosphorylated blots (skRLCP), respectively. Previously unpublished data

Full size image

Myosin heavy chain isoform expression may not singularly predict the capacity for potentiation. For example, although prolonged, high frequency stimulation may produce moderate elevations in RLC phosphorylation, little or no twitch potentiation is observed in rat soleus muscle (Manning and Stull 1982; Moore and Stull 1984). Consistent with this, Ryder et al. (2007) showed that overexpression of skMLCK in soleus muscle did not lead to twitch force potentiation despite stimulation induced phosphorylation of both slow and fast RLCs. The reason for the lack of twitch potentiation with substantial RLC phosphorylation in soleus muscles from transgenic mice is not clear. C57BL/6 mouse EDL muscle contains 70 % type IIb fibers (Gorselink et al. 2002). In contrast, the mouse soleus muscle contains as little as 6 % type IIb fibers with 59 % type IIa and 35 % type I fibers (Totsuka et al. 2003). Thus, the potentiation of contraction is correlated to type IIb fibers. Potentially, the sarcomeric interfilament spacing of type I and type IIa fibers may be sufficiently small for optimal force development even if RLC is not phosphorylated. It may also be possible to hypothesize that myosin light chain (MLC) isoform expression provides additional regulation. Up to five distinct MLC isoforms are present in mammalian skeletal muscle with the expression of slow (MLC1s, MLC2s) or fast (MLCf, MLC1f, MLC2f, MLC3f) isoforms generally mirroring myosin heavy chain profile (Bicer and Reiser 2004; Gonsalez et al. 2002; Schiaffino and Reggiani 1996, 2011). Interestingly, although it is unclear if alterations to MLC expression can occur independent of myosin heavy chain expression, skeletal muscle plasticity includes the MLC isoform. For example, the slow to fast phenotype transformation of rat muscle shown with hindlimb suspension or clenbuterol administration includes a change in the relative expression of slow to fast MLC isoforms (Bozzo et al. 2003). Moreover, the complementary, reverse, pattern of change to MLC and myosin heavy chain isoform was observed when a fast to slow phenotype transformation was induced (Bozzo et al. 2005). Similar results have been presented by Stevens et al. (2000, 2004). These studies suggest that, although the slow MLC isoforms may still be phosphorylatable, alterations to MLC isoform may participate in the fiber type dependence for RLC phosphorylation—mediated force potentiation. The mechanistic details for this regulation are unknown but it may be possible that RLC isoform composition differentially influences head to head interactions and displacements in the absence and presence of serine phosphorylation, respectively.

Alternate Ca2+ based mechanism for potentiation?

Results from skMLCK−/− muscles indicates the presence of a secondary mechanism for potentiation that is highly stimulus regime dependent. In principal there are several mechanisms that may be able to account for this RLC phosphorylation-independent increase in isometric twitch force; clearly, the contribution of these mechanisms to potentiation may be obscured by RLC phosphorylation influences in wildtype muscles. Recent work using mouse lumbrical has provided evidence that stimulation-induced alterations to resting Ca2+ homeostasis can account for isometric twitch potentiation in the absence of stimulation-induced elevations in RLC phosphorylation (Smith et al. 2013). Indeed, the mouse lumbrical thus appears to be a unique fast-muscle in that it does not appear to contain the enzymatic apparatus for RLC phosphorylation (Ryder et al. 2007; Smith et al. 2013). In these experiments, stimulation did not produce any apparent change to the amplitude or kinetics of the intracellular [Ca2+]i transient itself but a short lived (20−30 s) increase in resting myoplasmic [Ca2+]i was observed, an increase that correlated temporally with the relatively small and short lived isometric twitch potentiation that was observed. The short lived nature of the increase in [Ca2+]i observed in these experiments could explain why staircase potentiation is less affected than PTP by the absence of RLC phosphorylation (MacIntosh et al. 2008a; Rassier et al. 1999). Although the mechanism of action remains unknown, stimulation induced increases in resting [Ca2+]i could enhance force by increasing the Ca2+ occupancy of troponin C or other Ca2+ buffers (e.g., parvalbumin) prior to twitch stimulation (e.g., Barclay 1992). On the other hand, the effect of increased resting [Ca2+]i could be more complex and be mediated at the level of the thick filament, perhaps by increasing the population of weakly-bound cross-bridges, a necessary precursor to the attainment of the strongly-bound, force generating state in some cross-bridge models (Kraft et al. 1999). Clearly, more work is needed to establish how muscle force is potentiated in the absence of RLC as well as the relative contribution of secondary mechanisms to potentiation in muscles with RLC phosphorylation.

Fatigue and potentiation

Twitch force potentiation and fatigue have been demonstrated to exist in a variety of single fiber and whole skeletal muscle preparations (Gordon et al. 1990; MacIntosh and Gardiner 1987; MacIntosh and Kupsh 1987; MacIntosh et al. 1993; Rankin et al. 1988; Tubman et al. 1996a; Vandenboom and Houston 1996; Vergara et al. 1977). This may suggest that RLC phosphorylation offsets fatigue or that certain metabolites arising during repetitive stimulation are able to potentiate twitch force. Interestingly, the relative depression in twitch force observed during the early, but not late, stages of repetitive stimulation of mouse EDL muscle is greater in skMLCK−/− than in wildtype muscles (Gittings et al. 2011). Although this result seems to corroborate a role for RLC during moderate, but not severe fatigue, it does not eliminate the contribution of other intracellular mechanisms. As an example, Barclay (1992) used mouse EDL muscles (25 °C) to show that the enhanced rate of force development observed simultaneous with fatigue could be attributed to increases in intracellular levels for inorganic phosphate ([Pi]). Increased [Pi] may influence cross-bridge kinetics in such a way that the rate of rise of isometric force is increased coincident with fatigue (e.g., Hibberd et al. 1985). Although plausible, this mechanism is difficult to isolate experimentally from other mechanisms, including RLC phosphorylation, in fatigued wildtype muscle.

Cooke and colleagues have performed a series of studies examining the complex interactions that may occur between metabolic changes during fatigue and RLC phosphorylation (Cooke 2007). It could be noted that, in unfatigued skeletal muscle, RLC phosphorylation is expected to influence pre—but not post power—stroke steps in the cross-bridge cycle (i.e. Sweeney and Stull 1990). Work by Franks-Skiba et al. (2007) and Karatzaferi et al. (2008) both suggest, however, that when the metabolites [Pi] and/or [H+] are elevated to levels that mimic those observed in fatigued skeletal muscle, the rate of release of ADP following the power stroke is delayed in phosphorylated cross-bridges (c.f., Stewart et al. 2009). This mechanism is in fact supported by studies on individual myosin molecules using the in vitro motility assay technique under non-fatigue conditions (Greenberg et al. 2010). It is of interest that these results could account for both the augmented twitch force and/or the slowed shortening velocity sometimes observed during and attributed to fatigue. Interestingly, no differences in either peak tetanic force or unloaded shortening velocity that are expected from these interactions were noted between skMLCK knockout and wildtype mouse muscles during severe fatigue at 25 °C (Gittings et al. 2011). A possible explanation for this discrepancy is that the interaction between fatigue related changes in metabolites and RLC phosphorylation in permeabilized fibers were found to be highly temperature dependent, with little interaction noted below 30 °C. Further work on intact muscle at higher temperatures (e.g., 30 °C) is perhaps needed to clarify this issue.

RLC phosphorylation and metabolism

The physiological ability of the RLC phosphorylation mechanism to modulate skeletal muscle function in vivo may hinge upon its influences on the metabolic cost or efficiency of muscle contraction. For example, if RLC phosphorylation (or other mechanisms) increases the energetic cost of contraction, potentiated contractions may prove to be unsustainable metabolically. Unfortunately, studies in this area have been equivocal, however. For example, early studies were interpreted to suggest that phosphorylation of the RLC decreased energetic demand during sustained or repetitive contractions of mouse EDL muscle in vitro (Crow and Kushmerick 1982a, b). Subsequent work also using a mouse muscle model questioned this outcome (Barsotti and Butler 1984; Butler et al. 1983). More recently, Abbate et al. (2001) have re-examined the issue by studying the energetic cost of potentiated contractions in rat muscle. These authors reported that high energy phosphate turnover was increased to a greater extent than was force, work and power, thus decreasing muscle economy of potentiated versus unpotentiated contractions. Critically, the decreased economy was attributed to differences in force, not RLC phosphorylation, implying that potentiated contractions may not be metabolically sustainable (Abbate et al. 2001). Although this result would seem to have resounding implications for the physiological utility of RLC phosphorylation to modulate repetitive type contractions, the robustness of this decrease to different contractile conditions (i.e. temperature, species, fiber type, stimulation rate, shortening speed etc.) is as yet unknown.

Potentiation in diseased muscle

Most work examining potentiation has been performed using non-diseased, healthy animal skeletal muscles. Exceptions to this trend suggest that certain peripheral diseases may influence potentiation, however. For example, Krarup (1983) showed that both staircase and PTP were attenuated in a neuropathy rat muscle model of chronic myasthenia gravis. In addition, potentiation has been shown to be reduced in parallel with disuse and/or atrophy of rat fast twitch muscle (MacIntosh et al. 1988; Tubman et al. 1996b, 1997). Another example is work by Smith et al. (2010) who showed that a high frequency conditioning stimulus caused greater PTP in EDL muscles from mdx mice than in EDL muscles from age matched controls, despite similar RLC phosphorylation at both ages and muscle types (see also Hoekman 1977). Thus, more information regarding interactions between potentiation and peripheral diseases such as sarcopenia and/or cachexia is required before a full understanding of how this fundamental response is altered in disease is achieved.

Summary

The different fiber types comprising skeletal muscles display a range of metabolic and mechanical properties geared to optimize function. Prominent among these differences is that of force potentiation, a history dependent enhancement of dynamic force, work and power at physiological activation rates in mammalian fast twitch muscle. Evidence from a variety of sources points to stimulation-induced elevations in RLC phosphate incorporation by skMLCK as the primary mechanism for potentiation, although stimulation-induced elevations in resting [Ca2+]i may provide a secondary mechanism in some muscle fibers. Evidence from skMLCK knockout and disuse models indicates that staircase potentiation observed during prolonged low frequency stimulation may be due to both RLC phosphorylation and alterations in resting [Ca2+]i while PTP observed after brief or high frequency stimulation is predominantly due to RLC phosphorylation. Further investigations are required to resolve the relative contribution of these respective mechanisms to potentiation phenomena.