Shortening deactivation: quantifying a critical component of cyclical muscle contraction

Abstract

A muscle undergoing cyclical contractions requires fast and efficient muscle activation and relaxation to generate high power with relatively low energetic cost. To enhance activation and increase force levels during shortening, some muscle types have evolved stretch activation (SA), a delayed increased in force following rapid muscle lengthening. SA’s complementary phenomenon is shortening deactivation (SD), a delayed decrease in force following muscle shortening. SD increases muscle relaxation, which decreases resistance to subsequent muscle lengthening. Although it might be just as important to cyclical power output, SD has received less investigation than SA. To enable mechanistic investigations into SD and quantitatively compare it to SA, we developed a protocol to elicit SA and SD from Drosophila and Lethocerus indirect flight muscles (IFM) and Drosophila jump muscle. When normalized to isometric tension, Drosophila IFM exhibited a 118% SD tension decrease, Lethocerus IFM dropped by 97%, and Drosophila jump muscle decreased by 37%. The same order was found for normalized SA tension: Drosophila IFM increased by 233%, Lethocerus IFM by 76%, and Drosophila jump muscle by only 11%. SD occurred slightly earlier than SA, relative to the respective length change, for both IFMs; but SD was exceedingly earlier than SA for jump muscle. Our results suggest SA and SD evolved to enable highly efficient IFM cyclical power generation and may be caused by the same mechanism. However, jump muscle SA and SD mechanisms are likely different, and may have evolved for a role other than to increase the power output of cyclical contractions.

INTRODUCTION

Power generation for many forms of animal locomotion relies on antagonistic muscle pairs cyclically lengthening and shortening to move limbs or other body parts. Well-known muscle properties, such as the force-length and force-velocity relationships, help determine the amount of muscle power generated (1–7). Though less studied, activation and relaxation rates are equally important for power generation during cyclical contractions. A muscle must be activated before shortening, and faster activation enables higher force generation during shortening (8, 9). Similarly, and perhaps even more imperative for cyclical contractions, is a fast relaxation rate. If the agonist muscle does not relax fast enough, the antagonist muscle must expend energy to work against the force generated by a partially activated agonist muscle (2, 10, 11). Thus, increasing the rate of activation to increase force and work during shortening, and increasing the rate of relaxation to decrease the resistance to lengthening, will boost net muscle power output from cyclical contractions, especially during fast locomotion (2, 10–12).

However, faster activation and relaxation rates are energetically costly. More calcium release channels and a larger volume of sarcoplasmic reticulum are required to increase the activation rate, which reduces fiber space for contractile proteins (10). Even more expensive is increasing the relaxation rate because more Ca²⁺-ATPases are needed to increase calcium uptake from the myoplasm and into the sarcoplasmic reticulum (11, 13). Depending on a muscle’s speed, the sarcoplasmic reticulum Ca²⁺-ATPases are responsible for 25%–40% of the total energy expended during contraction (14–17).

To help improve power output and efficiency, some muscle types have evolved two properties that assist in modulating force levels during cyclical contraction, stretch activation (SA) and shortening deactivation (SD). SA is a delayed increase in force following rapid lengthening of an active muscle, whereas SD is a delayed decrease in force following rapid shortening (Fig. 1) (10, 18, 19). SA’s delayed increase in force is timed to increase force levels during the shortening portion of a contraction cycle. This supplements or even takes the place of calcium activation in some muscle types (10, 12, 20–22). Similarly, SD is timed to decrease force during lengthening, and reduces or eliminates the need to pump calcium out of the myoplasm during each muscle contraction cycle. Thus, working together, SA and SD assist or replace calcium cycling, which greatly increases net power and efficiency.

Figure 1.

Stretch activation and shortening deactivation protocol. A: representative tension traces (top) from a skinned Drosophila IFM fiber following a 1% increase in muscle length (ML) over 0.5 ms (bottom). The fiber was held at the longer length for 500 ms and shortened by 1% ML back to its original length over 0.5 ms. This length change perturbation was performed at two calcium concentrations: pCa 5.0 (blue) and pCa 8.0 (gray). B: SA portions (0.04–0.10 s) of the tension traces shown in A. A_SA is the total SA tension, whereas P_SA is the SA response during passive conditions. The net active SA tension, F_SA, is the difference between A_SA and P_SA. Similarly, A₀ and P₀ indicate total and passive isometric tension, respectively, whereas A₀ − P₀ is the net calcium-activated isometric tension, F₀. Small arrows indicate the four phases of the SA trace. C: SD portion (0.54–0.60 s) of the tension traces show in A. Shortening-deactivated tension (A_SD) was defined as the difference between the tension prior to shortening the muscle, and the local minimum following initial tension recovery (phase 3 valley). A_SD is the total SD tension, whereas P_SD is the SD tension during passive conditions. F_SD, the net active SD tension, is the difference between A_SD and P_SD. Small arrows indicate the four phases of the SD trace. IFM, indirect flight muscle; SA, stretch activation; SD, shortening deactivation.

In vertebrate heart muscle, SA is important for enhancing contractility as part of the Frank-Starling mechanism (8, 23), and in skeletal slow-twitch muscle, it has been hypothesized to increase muscle endurance (24). However, although SD has been theorized to assist heart muscle contraction (25), it has not been studied directly. For vertebrate skeletal muscle, SD has been almost completely neglected as we are not aware of any quantitative studies. The muscle type that has been the most widely studied for its SA properties is the insect indirect flight muscle (IFM), yet there have only been a few investigations into IFM SD (2, 10, 22, 26, 27).

IFMs power the wing stroke of insects during flight. In some species, especially ones with faster IFMs, the rate at which the IFMs are stimulated by their motor neurons is much lower than the muscle’s cyclical contraction frequency, hence the name asynchronous IFM (10, 12, 28, 29). This results in no appreciable calcium cycling during an individual muscle contraction cycle. Instead, SA and SD have taken over modulating force levels per cycle. Asynchronous IFMs, such as those found in Drosophila and Lethocerus, are composed of antagonistic muscle pairs, the dorsoventral muscles (DVMs) and the dorsal longitudinal muscles (DLMs) (30). The DVMs lift the wings and lengthen the DLMs whereas contraction of the DLMs lowers the wings and lengthens the DVMs (31).

Although SD is important for IFM function, there have been few investigations into whether SD is present in other muscle types and if so, how much it benefits these muscles. Even less is known about the mechanism behind SD because there have been no direct investigations into the SD mechanism for any muscle type. Possible reasons SD has been overlooked are that it is not a well-known phenomenon in the muscle field and because investigators have likely presumed it is just the opposite phenomenon of SA and caused by the same mechanism.

Currently, the thin filament-based mechanism has the most support for being the cause of SA in Drosophila and Lethocerus IFM (20, 22, 32, 33), and evidence is growing for a myosin-based SA mechanism for skeletal slow-twitch muscle (24). The thin filament mechanism for SA posits that calcium does not fully activate the thin filament. Rather, stretching a calcium-activated muscle moves the troponin-tropomyosin complex further off myosin binding sites, which allows for more myosins to bind and contribute to force generation. If this mechanism is correct, and if SA and SD are reciprocal phenomena, then SD should cause the troponin-tropomyosin complex to move back over the myosin binding sites on actin, which would detach myosin heads and decrease force generation.

However, to elucidate if SA and SD are reciprocal phenomena, and if they are caused by the same molecular mechanism, we first need to develop methods to define, characterize, and quantify the SD response. To this end, we have developed a mechanical protocol for elucidating both SA and SD and applied it to three different muscles: Lethocerus IFM, Drosophila IFM, and Drosophila jump muscle (Fig. 1A). The two asynchronous IFM muscles power flight, with Drosophila IFM contracting ∼10-fold faster than Lethocerus IFM (27). Lethocerus IFM generates higher force levels than Drosophila IFM. The jump muscle (tergal depressor of the trochanter) is a synchronous muscle and is analogous to a fast-twitch mammalian skeletal muscle. As its name implies, it powers jumping and take-offs to initiate flight (34).

We measured skinned fiber SD characteristics and compared the results to SA characteristics to determine if SA and SD are opposite phenomena in all three muscles. Our results show that IFM SA and SD are more similar in their properties and cause much higher relative tension changes than jump muscle SA and SD. This suggests the IFM SA and SD mechanisms are more likely to be the same than jump muscle SA and SD mechanisms. The low amount of SA tension, relative to the overall force producing capacity in the jump muscle, suggests its SA may not be physiologically relevant. However, a 30% tension decrease due to SD could be important for muscle function. This first investigation to quantify SD sets the groundwork for future studies to characterize SD in other muscle types and to use the power of Drosophila genetics to investigate the mechanism(s) behind SD.

METHODS

The IFM SD data presented here are derived from further analysis of tension transients collected during the Glasheen et al. study (27), and some of the IFM SA data were previously published in Glasheen et al. (27). We specify which data have been previously published in the figure and table legends. All jump muscle data in this paper are new. Thus, we briefly describe the Drosophila and Lethocerus IFM preparation and procedures but go into more detail about the jump muscle.

Muscle Fiber Preparation

IFM fibers.

IFM fibers (n = 10) from 3-day-old wild-type Oregon-R Drosophila were dissected as previously described (27, 35). Briefly, an isolated thorax was immersed in skinning solution [-log₁₀ of the calcium concentration (pCa) 8.0, 12 mM MgATP, 1 mM free Mg²⁺, 5 mM EGTA, 300 mM CP, 300 U/mL CPK, 20 mM N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid (BES, pH 7.0), 200 mM ionic strength, adjusted with Na methane sulfonate, 1 mM DTT, 50% glycerol, and 0.5% Triton X-100] at 4°C and split open to allow the removal of fibers. Individual fibers were split lengthwise for better diffusion of fiber bathing solution components and mounted to the mechanical apparatus with aluminum T-clips. Our single chamber muscle mechanics apparatus, where fiber bathing solutions are exchanged into and out of one bathing chamber, was used for fibers from all three muscles.

Live specimens of Lethocerus indicus were imported from Thailand and skinned IFM fibers (n = 10) prepared as described in Glasheen et al. (27). Briefly, cold Ringer’s solution was flowed into the Lethocerus thorax through holes punctured into the anterior and posterior ends. The thorax was cut into dorsal/ventral halves and the DLMs remained in the hemithoraces for protection. The fibers were skinned by alternating glycerol and aqueous solutions, which were exchanged several times over 2 days. The hemithoraces can be stored in the final high glycerol solution for many years at −100°C with no apparent degradation of structure or mechanical function. In preparation for mechanical analysis, DLM fascicles were thawed and pared down to individual fibers of ∼0.5 mm length and then fitted with aluminum foil T-clips (35, 36). The resulting fiber length between clips averaged ∼0.3 mm, which was similar to the clipped Drosophila IFM fibers. Both species of IFM were transferred to the mechanics apparatus, connected to a force gauge and servo motor using aluminum foil T-clips, and submerged in relaxing solution [pCa 8.0, 12 mM MgATP, 30 mM creatine phosphate, 600 U/mL creatine phosphokinase, 1 mM free Mg²⁺, 5 mM EGTA, 20 mM BES with pH 7.0, 200 mM ionic strength (adjusted with sodium methanesulfonate), 1 mM DTT]. We performed the IFM and jump muscle mechanics at a physiological relevant temperature for ectothermic Drosophila, 15°C (37). Larger insects including Lethocerus, warm their IFMs to ∼35°C before and during flight (38). To enable comparison with our Drosophila muscle mechanics, the Lethocerus fiber mechanics were also performed at 15°C. Thus, the Lethocerus IFM experiments were below physiologically relevant temperatures.

To activate the Drosophila (n = 10) and Lethocerus IFM fibers (n = 10), the relaxing solution (pCa 8.0) was serially exchanged with activating solution (relaxing solution adjusted to pCa 4.0) to reach pCa 5.0. Sinusoidal analysis (see methods, Muscle Mechanics Protocols) was employed to find the length at which the fibers generated the most power, which was considered the optimal starting length for the IFM. Note that measuring sarcomere length in IFM while on a mechanics rig is not possible as the myofibrils within the IFM are not laterally aligned in sarcomeric register (39, 40). The cross-sectional area of each fiber for Drosophila IFM (0.00711 ± 1.2E-04 mm²) and Lethocerus IFM (0.00307 ± 7.0E-05 mm²) was calculated using the formula for an ellipse and two fiber diameters: the width measured from the bottom and the height measured from the side using a mirror (33, 41).

Jump muscle fibers.

Jump muscle fibers (n = 13) were dissected from 3-day-old w¹¹¹⁸ Drosophila melanogaster as described previously in detail (35) and mounted onto the same fiber mechanics apparatus as the IFM fibers. To summarize, the thorax was split along the sagittal plane and the jump muscle was teased out of each half so it could be chemically demembranated at 4°C in skinning solution. After 1 h, the skinned jump muscle fiber bundle was transferred to storage solution (skinning solution without Triton X-100), where it was reduced in size to 8–10 fibers. Note that the jump muscle has very small diameter fibers necessitating using a bundle of fibers rather than a single fiber. The muscle was connected to a force gauge and servo motor using aluminum foil T-clips and submerged in relaxing solution (pCa 8.0).

Jump muscle fibers (n = 13) were stretched to their optimal sarcomere length of 3.6 µm for mechanics experiments and the cross-sectional area was measured (0.00344 ± 5.1E-05 mm²) (33, 41). The fiber was shortened to a slack length to establish baseline tension and lengthened back to the starting sarcomere length of 3.6 μm. To activate the fiber, the relaxing solution (pCa 8.0) in the mechanics apparatus was first serially exchanged with preactivating solution (relaxing solution with 0.5 mM EGTA), followed by the appropriate amount of activating solution (pCa 4.0) to achieve pCa 5.0. The use of preactivating solution helped maintain sarcomere length homogeneity (35).

Muscle Mechanics Protocols

Isometric tension and tension-pCa curves.

All three muscle types were subjected to an identical mechanical protocol to measure isometric tension, SA, and SD characteristics at increasing concentrations of calcium [Ca²⁺]. An initial mechanics perturbation was performed at pCa 5.0 to determine the maximum force generating capacity of each muscle. These measurements were used at the end of the protocol to help quantify fiber degradation and possible fiber exclusion (see below). A full solution exchange was then performed to reset the solution pCa to 8.0. The following concentrations (expressed as pCa) were tested in order: 8.00, 7.50, 7.00, 6.75, 6.50, 6.25, 6.00, 5.75, 5.50, 5.25, 5.00, 4.75, and 4.50. To adjust the [Ca²⁺], a predetermined volume of solution was removed from the bubble that surrounded the fiber and the same amount of activating solution (pCa 4.0) was added back to the bubble. The resultant tension following the solution exchange was recorded once the force reached a steady, maximum value (∼30–60 s) to ensure that diffusion of the new solution into the skinned fiber was complete. This calcium-activated tension produced by the fiber at a constant length is defined as isometric tension. Passive isometric tension (P₀) was measured at pCa 8.00. Total active isometric tension (A₀) was measured at successively increasing [Ca²⁺] between pCa 8.00 and 4.50. The net active isometric tension (F₀) was calculated as the difference between A₀ and P₀ at each calcium concentration (F₀ = A₀ – P₀). Average F₀ values for each pCa were first plotted to show changes in the magnitude of F₀. Each curve was then normalized to the maximum F₀ per muscle type (F_{0 MAX}) and fitted with the Hill equation (Eq. 1), where F was the observed force and the variable x was pCa.

$$\text{F}={\text{F}}_{\text{MAX}}/(1+{10}^{\text{nH}}{}^{\ast (x-\text{pCa}50)})$$ (1)

The parameter pCa₅₀ is the concentration of calcium that produces half-maximal tension, and the Hill coefficient (n_H) reflects cooperativity (42, 43).

Stretch activation and shortening deactivation.

SA was induced by elongating the fiber by 1% of its optimal muscle length over a period of 0.5 ms (Fig. 1A). The fiber was held stretched at this longer length for 500 ms and then SD was induced by returning the fiber to its original length over 0.5 ms. Immediately (∼1 s) following the trapezoid perturbation, A₀ was measured again at the original fiber length to determine if damage occurred to the fiber as result of the trapezoid shaped lengthening and shortening protocol (hereon referred to as Trapezoid Perturbation). A small drop in tension occurred relative to the tension before the perturbation: typically, 6%–8% for Drosophila IFM, 2%–3% for Lethocerus IFM, and 2%–3% for Drosophila jump muscle. If A₀ decreased by more than 10% after a single stretch, the data from this fiber were excluded. Activating solution (pCa 4.0) was exchanged into the bathing solution to reach the next pCa, and the measurements of A₀, F_SA, and F_SD were repeated at increasing [Ca²⁺]s (Supplemental Fig. S1; see https://doi.org/10.6084/m9.figshare.17026964.v1). To ensure full force recovery following the trapezoid perturbation and calcium exchange, we waited between 1 and 2 min before running the next perturbation.

The magnitude of the total stretch-activated force (A_SA) was measured at the peak of phase 3 (Fig. 1B), as described previously (41). The phase 3 peaks were selected manually as attempts to use automated routines did not improve accuracy and sometimes selected incorrect peaks. Passive stretch-activated tension, P_SA (pCa 8.0), was quantified at the same time that corresponded to the peak of phase 3 in active samples (Fig. 1B). The net active stretch-activated tension (F_SA) was defined as the difference between A_SA and P_SA at each pCa to help exclude contributions from passive sarcomere elements not found in cross bridges. However, we note that calcium may bind to titin making it become stiffer with increased [Ca²⁺]s (44). Thus, we also provide total SA tension values (A_SA) in our results (Table 3). To normalize the amount of F_SA to the overall tension generating ability of the fiber, F_SA was divided by F_{0 MAX} of that fiber.

Table 3.

Stretch-activated tension

Muscle	ASA, mN/mm2	PSA, mN/mm2	FSA, mN/mm2	FSA/F0 MAX (%)
	pCa 5.0	pCa 8.0	pCa 5.0–8.0
Drosophila IFM	8.5 ± 0.8Δ	5.6 ± 0.4#	3.8 ± 0.4Δ	233.1 + 32.8#
Drosophila jump muscle	4.5 ± 0.6Δ	0.7 ± 0.1†Δ	3.8 ± 0.5Δ	10.6 + 0.5†Δ
Lethocerus IFM	38.8 ± 2.6†#	23.6 ± 1.8#	15.2 ± 0.9†#	76.0 + 5.6#

Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. All values are means ± SE. †P < 0.05 compared with Drosophila IFM. #P < 0.05 compared with Drosophila jump muscle. ΔP < 0.05 compared with Lethocerus IFM (one-way ANOVA with Dunn’s method pairwise multiple comparisons). Results from Lethocerus and Drosophila IFM were previously reported in Glasheen et al. (27). A_SA, total stretch-activated tension (pCa 5.0); F_SA, net stretch-activated tension (F_SA = A_SA − P_SA); F_SA/F_{0 MAX}, net stretch-activated tension (F_SA) at pCa 5.0, divided by maximum net active isometric tension (F₀) for each fiber; IFM, indirect flight muscle; P_SA, passive stretch tension (pCa 8.0).

Similarly, total shortening-deactivated force (A_SD) was determined by subtracting the average phase 3 valley tension, again selected manually, from the tension level immediately before the muscle being shortened at pCa 5.0 (Fig. 1C). Passive shortening-deactivated tension, P_SD (pCa 8.0), was measured at the same time point as the phase 3 valley when the fiber was active (Fig. 1C). The net active shortening-deactivated tension (F_SD) was defined as the difference between A_SD and P_SD at each pCa, to exclude any contribution from passive elements to F_SD. To normalize the amount of F_SD to tension-generating ability of the fiber, F_SD was divided by F_{0 MAX}. Calcium response parameters, pCa₅₀ and n_H, were calculated by fitting the relationship between F_SA or F_SD and pCa with the Hill equation (Eq. 1).

Rates (r) of phases 2–4 of the SA tension transients were determined by fitting tension responses to the sum of three exponential curves (Eq. 2), with r₂ being the rate of rapid tension decrease in phase 2, r₃ the rate of tension increase during phase 3, and r₄ the rate of tension decay during phase 4:

$$T=a_2{e}^{(-r_{2}t)}+a_3(1-e^{(-r_{3}t)})+a_4{e}^{(-r_{4}t)}+\text{offset}$$ (2)

We have previously used this equation to fit SA traces (24, 27, 35, 45, 46). SD tension transients were fit using the same terms. But because the phases of the SD tension transients are inverted compared with SA, there are two increasing terms, r₂ and r₄, and one decreasing term, r₃, which is the SD term:

$$T=a_{2}e(1-e^{(-r_{2}t)})+a_3{e}^{(-r_{3}t)}+a_4(1-e^{(-r_{4}t)})+\text{offset}$$ (3)

The SD traces were fit using the nonlinear least squares method, a least absolute residuals robust setting, and a Trust-Region algorithm in MATLAB (v2019a, The MathWorks, Inc., Natick, MA). Because the SA traces were previously fit using SigmaPlot (SYSTAT Software, Erkrath, Germany), to ensure that using different software to fit the traces did not alter the rates generated, we refit the SA traces using the same MATLAB curve-fit program we used for the SD traces. We found that there was no statistical difference between the SA rates for phase 2, 3 and 4 generated by either program for any of the muscle types (data not shown).

We also analyzed the kinetics of the three phases by measuring the time from the start of the length perturbation to the occurrence of phase 2 and 3 peaks/valleys of SA and SD. For SA, we subtracted the time when the muscle stretch began (50.4 ms into the Trapezoid Perturbation run) from the time when the phase 2 valley and phase 3 peak occurred for each fiber. Likewise, we subtracted the time when muscle shortening was initiated (550.8 ms into the Trapezoid Perturbation run) from the time when the phase 2 peak and phase 3 valley of SD occurred for each fiber.

A final exchange back into pCa 5.0 occurred at the end of the protocol where F₀, F_SA, and F_SD measurements were repeated to quantify fiber degradation. If the tension decrease over the entire protocol was greater than 30% for isometric tension, 15% for F_SA, and 15% for F_SD, the fibers were excluded from the study (27).

Sinusoidal length change analysis.

Sinusoidal length change analysis was performed as described previously (35). Briefly, Drosophila IFM (n = 10) and Lethocerus IFM (n = 10) fibers were oscillated through a small-amplitude sine wave, which administered a 0.125% peak-to-peak length change at 50 frequencies increasing serially from 0.5 to 650 Hz for Drosophila and 0.5 to 500 Hz for Lethocerus. In this study, maximum power output determined from sinusoidal analysis was used for setting the starting muscle length of IFM fibers. If an IFM fiber did not produce a maximum power greater than 80 W/m³, it was discarded and excluded from the study (27).

Statistical Analysis

Figures were plotted and statistical analyses were performed using SigmaPlot (Systat Software, Erkrath, Germany). The two-tailed Student’s t-test with unequal variance (significance threshold at P < 0.05) was used to compare the SA and SD parameters within the same muscle type. One-way ANOVA with Dunn’s method pairwise multiple comparisons (significance threshold at P < 0.05) was used to evaluate the differences between all three muscle types. Values are reported as the mean ± standard error of the mean.

RESULTS

Isometric Tension

We measured isometric tension at the start of each trapezoidal length change (Fig. 1B) to ensure fibers were performing optimally and to enable normalization of SA and SD measurements to the force generating ability of each muscle type. Active isometric tension (F₀) (F₀ = A₀ − P₀) at pCa 5.0 for Drosophila jump muscle was 21-fold and ∼1.9-fold greater than Drosophila IFM and Lethocerus IFM, respectively (Table 1 and Fig. 3A).

Table 1.

Isometric tension values

Muscle	A0, mN/mm2	P0, mN/mm2	F0, mN/mm2	pCa50	nH
	pCa 5.0	pCa 8.0	pCa 5.0–8.0
Drosophila IFM	4.3 ± 0.4Δ#	2.7 ± 0.2Δ	1.6 ± 0.3Δ#	6.0 ± 0.1 Δ#	1.9 ± 0.2#
Drosophila jump muscle	36.8 ± 3.5†	2.5 ± 0.3Δ	34.3 ± 3.4†Δ	5.6 ± 0.1†	12.4 ± 1.4†Δ
Lethocerus IFM	38.1 ± 4.0†	19.7 ± 2.3 †#	18.4 ± 1.9†#	5.6 ± 0.1†	1.0 ± 0.1#

Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. All values are means ± SE. †P < 0.05 compared with Drosophila IFM. #P < 0.05 compared with Drosophila jump muscle. ΔP < 0.05 compared with Lethocerus IFM (one-way ANOVA with Dunn’s Method pairwise multiple comparisons). Results from Lethocerus and Drosophila IFM were previously reported in Glasheen et al. (27). A₀, total active isometric tension (pCa 5.0); F₀, net active isometric tension (F₀ = A₀ − P₀); IFM, indirect flight muscle; n_H, Hill coefficient; P₀, passive isometric tension (pCa 8.0); pCa₅₀, the calcium concentration where isometric tension was half-maximal (calcium affinity).

Figure 3.

Comparison of isometric tension (F₀), SD tension (F_SD), and SA tension (F_SA) from Drosophila IFM, Drosophila jump muscle, and Lethocerus IFM. A: isometric tension at pCa 8.0 (P₀) was subtracted from isometric tension at pCa 5.0 (A₀) to obtain F₀. Box and whisker plots depict F₀ at pCa 5.0. B: SA tension at pCa 8.0 (P_SA) was subtracted from SA tension at pCa 5.0 (A_SA) to obtain SA tension, F_SA. C: SD tension at pCa 8.0 (P_SD) was subtracted from SD measurements at pCa 5.0 (A_SD) to obtain SD tension, F_SD. Plots depict F_SD at pCa 5.0. Plots depict F_SA at pCa 5.0. The upper and lower boundaries of each box represent the 75th and 25th percentile, respectively. The line within each box marks the median value. The error bars indicate the 90th and 10th percentiles. Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. Horizontal bars indicate statistically significant differences between the values at each end of the bar, P < 0.05 (one-way ANOVA with Dunn’s method pairwise multiple comparisons). *Significant difference compared with F_SA for the indicated muscle type (Student’s t test, P < 0.05). Isometric tension and SA data from Lethocerus and Drosophila IFM were previously reported in Glasheen et al. (27). A0, total calcium-activated isometric tension; ASD, total shortening-deactivated tension; F₀, net calcium-activated isometric tension; F_SA, net stretch-activated tension; F_SD, net shortening-deactivated tension; IFM, indirect flight muscle; P₀, passive isometric tension; P_SA, passive stretch-activated tension; P_SD, passive shortening-deactivated tension; SA, stretch activation; SD, shortening deactivation.

Passive isometric tension (P₀) at pCa 8.0 differed greatly between the two species as P₀ for Lethocerus IFM was sevenfold and eightfold greater than Drosophila IFM and jump muscle, respectively. However, as a percentage of total isometric tension (A₀), IFM P₀ values were half of the total tension (A₀), whereas jump muscle P₀ was less than one tenth of A₀ (Table 1). In other words, the Drosophila jump muscle’s passive tension contributed less to its total tension compared with that of Drosophila IFM and Lethocerus IFM.

SD and SA Transients

A major goal for our investigation of SD was to develop an effective protocol to measure SD and quantitatively compare it with SA. The first step was to determine if the opposite action of our standard SA protocol of a 1% length step over 0.5 ms (shortening the fiber instead of lengthening it) would produce a four phase SD response that qualitatively mirrors the four phase SA response. Some attempted protocols were not successful. For example, we tried immediately shortening a pCa 5.0 activated Drosophila IFM fiber from its optimal starting length (the length where maximum power is generated) but found that when using an ∼0.5%–1% length decrease, the tension level often drops to zero and a four-phase response does not occur. However, if we performed a SA inducing length increase, which increases the force level, and waited 500 ms before shortening we could consistently evoke four SD phases for Drosophila IFM, Lethocerus IFM, and Drosophila jump muscle. We named this protocol the Trapezoid Perturbation (Fig. 1A).

In general, the SA and SD responses were inverted compared with each other (Fig. 2). The SA response consists of four phases: a rapid increase in force that is a direct and immediate response to lengthening (phase 1), a subsequent drop in force (phase 2), a delayed increase in force that we measure to find SA amplitude, F_SA (phase 3), and a slow decrease to a new steady-state force level (phase 4) (Figs. 1B and 2B). This response was consistent with our previous studies (24, 27, 45). The SD response also consisted of four phases, but these were inverted relative to SA (Figs. 1C and 2A). Phase 1 was a rapid, immediate drop in force, coinciding with muscle shortening. Phase 2 was a very brief increase in force that rapidly transitioned to phase 3, which was a delayed decrease in force. Phase 3 was only briefly maintained before force increased again (phase 4) yet at a slower rate than phase 2. SD phases 2 and 3 were not as obvious as SA phases 2 and 3, likely due to the overall lower tension levels induced by shortening the muscle (i.e., a lower signal to noise ratio). The phase 3 valley of SD was sometimes a large decrease in slope (decrease in rate of tension recovery) rather than a distinct reversal of slope (decrease in tension magnitude).

Figure 2.

Representative stretch activation and shortening deactivation tension traces from Drosophila IFM, Drosophila jump muscle, and Lethocerus IFM. A: SD tension traces. B: SA tension traces. All tension traces are from the same three individual fibers at pCa 5.0, and their isometric tension levels (F₀) were all set to 0 mM to enable comparisons of SA or SD levels. Dashed lines indicate where phase 3 amplitudes of SA and SD were measured for each muscle type. IFM, indirect flight muscle; F₀, net calcium-activated isometric tension; SA, stretch activation; SD, shortening deactivation.

Quantifying F_SD and F_SA: The Three Muscle Types Compared with Each Other

We defined SD amplitude, F_SD, as the difference between the isometric tension level immediately before the muscle length decrease and the lowest point of phase 3 (Fig. 1C). Drosophila IFM F_SD was the lowest of the three muscles examined, whereas Lethocerus IFM was the highest (Fig. 3C and Table 2). In other words, Lethocerus IFM displayed the largest delayed drop in tension. Lethocerus IFM F_SD was 9.9-fold higher than Drosophila IFM F_SD, and Drosophila jump muscle F_SD was 6.5-fold higher than Drosophila IFM. The relationship between muscle types for passive SD tension values, P_SD, was not congruent with the trend for active SD tension values (Table 2). Lethocerus IFM P_SD was the greatest, followed by Drosophila IFM P_SD, which was greater than Drosophila jump muscle P_SD.

Table 2.

Shortening-deactivated tension

Muscle	ASD, mN/mm2	PSD, mN/mm2	FSD, mN/mm2	FSD/F0 Max, %
	pCa 5.0	pCa 8.0	pCa 5.0–8.0
Drosophila IFM	7.1 ± 0.6 Δ#	5.2 ± 0.4#	2.0 ± 0.2Δ#	118.1 ± 16.7#
Drosophila jump muscle	13.6 ± 1.5†Δ	0.8 ± 0.2†Δ	12.9 ± 1.4†	37.1 ± 1.8†Δ
Lethocerus IFM	50.4 ± 3.6†#	30.6 ± 2.3#	19.8 ± 1.5†	96.6 ± 4.7#

Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. All values are means ± SE. †P < 0.05 compared with Drosophila IFM. #P < 0.05 compared with Drosophila jump muscle. ΔP < 0.05 compared with Lethocerus IFM (one-way ANOVA with Dunn’s method pairwise multiple comparisons). A_SD, total shortening-deactivated tension (pCa 5.0); F_SD, net shortening-deactivated tension (F_SD = A_SD − P_SD); F_SD/F_{0 MAX}, net shortening-deactivated tension (F_SD) at pCa 5.0, divided by maximum net active isometric tension (F₀) for each fiber; IFM, indirect flight muscle; P_SD, passive shortening-deactivated tension (pCa 8.0).

Normalizing to maximum isometric tension (F_SD divided by F_{0 MAX}) changed the relative order of SD amplitudes (Table 2 and Fig. 4). Drosophila jump muscle F_SD/F_{0 MAX} was the lowest, with Lethocerus IFM as the intermediate at 2.6-fold higher, and Drosophila IFM was the greatest at 3.2-fold higher than jump muscle F_SD/F_{0 MAX}.

Figure 4.

F_SA and F_SD normalized to maximum isometric force (F_{0 MAX}). Average F_SA divided by F_{0 MAX} is shown in the light colored box and whisker plots. Average F_SD divided by F_{0 MAX} is shown in the dark colored box and whisker plots. The upper and lower boundaries of each box represent the 75th and 25th percentile, respectively. The line within each box marks the median value. The error bars indicate the 90th and 10th percentiles. Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. *Significant differences between F_SA/F_{0 MAX} and F_SD/F_{0 MAX} within each muscle type (Student’s t test, P < 0.05). F_SA/F_{0 MAX} values for Drosophila and Lethocerus IFM were previously reported in Glasheen et al. (27). F_{0 MAX}, maximum net isometric tension; F_SA, net stretch-activated tension; F_SD, net shortening-deactivated tension; IFM, indirect flight muscle; SA, stretch activation; SD, shortening deactivation.

To gain insight into whether SD and SA are inverse phenomena, we also assessed SA properties of the three muscle types. F_SA values did not follow the same order among the three muscles as F_SD (Table 3 and Figs. 3, B and C), which one might expect if they are caused by the same mechanism in all three muscle types. The lowest F_SA values were observed for Drosophila IFM and jump muscle, which were not statistically different from each other, whereas Lethocerus F_SA was 4.6-fold higher than Drosophila F_SA values. Thus, Lethocerus IFM displayed the largest delayed increase in tension relative to the tension before the muscle was lengthened. The passive response to the length increase, P_SA, was statistically different between all three muscles with an order of Lethocerus IFM > Drosophila IFM > Drosophila jump muscle (Table 3).

Normalizing to maximum isometric tension (F_{0 MAX}) changed the relative order of SA amplitude values to Drosophila IFM > Lethocerus IFM > Drosophila jump muscle compared with F_SA values (Table 3 and Fig. 4) The jump muscle exhibited the lowest F_SA/F_{0 MAX} average value with Lethocerus IFM 7.2-fold higher and Drosophila IFM 22.9-fold higher than jump muscle F_SA/F_{0 MAX}.

F_SA/F_{0 MAX} Compared with F_SD/F_{0 MAX} within Each Muscle Type

All three muscle types showed different relationships between F_SA/F_{0 MAX} and F_SD/F_{0 MAX} values at pCa 5.0 (Fig. 4). Drosophila IFM was the only muscle type in which its F_SD/F_{0 MAX} was less than its F_SA/F_{0 MAX}, 1.9-fold, whereas Lethocerus F_SD/F_{0 MAX} was 1.3-fold greater than its F_SA/F_{0 MAX}, and jump muscle F_SD/F_{0 MAX} was 3.4-fold greater than its F_SA/F_{0 MAX}. Note that comparing F_SD and F_SA values would give the same results because the same F_{0 MAX} value is used to normalize both F_SD and F_SA.

Calcium Response

We observed the muscles’ responses to [Ca²⁺] as another way to characterize and compare SA and SD. What level of calcium elicits maximum F_SA is controversial. Our laboratory has demonstrated that maximum F_SA occurs at high [Ca²⁺] (27, 45), but earlier studies suggested that maximum F_SA occurs at moderate levels of calcium (8, 22). Thus, we wanted to examine the relationship between F_SD and [Ca²⁺]. F_SA and F_SD values showed a general increase in response to increasing [Ca²⁺] in all three muscle types (Fig. 5, A–C), with the typical sigmoidal shape characteristic of F₀ versus [Ca²⁺] reported previously (27, 47) and summarized in Table 1. The highest F_SD values occurred at the highest calcium concentrations, which is the same relationship we have previously reported for F_SA. Similar to F_SA and F_SD values at pCa 5.0, Drosophila IFM F_SA values were higher than their respective F_SD values at calcium concentrations where tension levels started increasing (pCa 6.75–4.5), whereas Drosophila jump muscle and Lethocerus IFM F_SD values were higher than their respective F_SA values.

Figure 5.

Relationship between stretch-activated tension, shortening-deactivated tension, and calcium concentration. Left: effect of calcium concentration on net stretch-activated tension (F_SA) and net shortening-deactivated tension (F_SD) for Drosophila IFM (A), Drosophila jump muscle (B), and Lethocerus IFM (C). Right: effect of calcium concentration on tension, normalized to maximum net stretch-activated tension (F_{SA MAX}) and maximum net shortening-deactivated tension (F_{SD MAX}) for Drosophila IFM (D), Drosophila jump muscle (E), and Lethocerus IFM (F). The relationship between normalized tension and [Ca²⁺] from individual fibers was fit with the Hill equation (Eq. 1) to derive the average Hill coefficient (n_H) and Ca²⁺ affinity (pCa₅₀) values shown in the insets. Values are means ± SE. Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. *Significant differences between F_SA and F_SD within each muscle type at each pCa (Student’s t test, P < 0.05). †P < 0.05 compared with Drosophila IFM, #P < 0.05 compared with Drosophila Jump, ΔP < 0.05 compared with Lethocerus IFM. F_SA values for Drosophila and Lethocerus IFM were previously reported in Glasheen et al. (27). IFM, indirect flight muscle; n_H, Hill coefficient; pCa₅₀, the calcium concentration where isometric tension was half-maximal (calcium affinity); SA, stretch activation; SD, shortening deactivation.

For the two IFM muscles, there were no statistical differences in pCa₅₀ or Hill coefficient (n_H) values between F_SD and F_SA when normalized to maximum tension (Fig. 5, D–F). In contrast, jump muscle SD and SA had similar pCa₅₀ values but the Hill coefficient was statistically different and 2.7-fold greater for SD (Fig. 5, D–F). These results suggest the calcium responses of SA and SD are similar in each of the IFM fibers but differ between SA and SD in jump muscle.

Compared between muscle types, the pCa₅₀ values for normalized F_SA and F_SD (F_SA/F_{SA MAX} and F_SD/F_{SD MAX}, respectively) were higher in Drosophila IFM compared with Lethocerus IFM and Drosophila jump muscle values. This suggests Drosophila IFM has the lowest affinity for calcium. Lethocerus IFM and Drosophila jump muscle pCa₅₀ values were not statistically different from each other (Fig. 5, D–F). Hill coefficient values were greatest in Drosophila jump muscle, suggesting a high degree of cooperativity and fast calcium activation, followed by Drosophila IFM and Lethocerus IFM, which were not statistically different (Fig. 5, D–F). Thus, all three muscles differed in their responses to [Ca²⁺].

SA and SD Kinetics

To determine SA and SD kinetics, we employed two methods. First, we measured the amount of time from the start of the length change to the peak/valley of phases 2 and 3 (Table 4). Second, we fit the tension traces with an equation consisting of three exponential terms corresponding to phases 2–4 (Fig. 6A). Equation 2 is the same equation we used for fitting SA traces in our previous studies (27, 35, 45). However, because the phases are inverted for SD (Fig. 1), we needed a new equation with two rising exponential terms and one falling (Eq. 3), as compared with SA where one phase is rising and the other two are falling (Eq. 2).

Table 4.

Timing of phase 2 and phase 3 of stretch activation and shortening deactivation

	Drosophila IFM, ms	Drosophila Jump Muscle, ms	Lethocerus IFM, ms
SA Phase 2	1.23 ± 0.1#	4.16 ± 0.2†Δ*	14.41 ± 0.9#*
SD Phase 2	1.38 ± 0.05	2.21 ± 0.1	2.40 ± 0.1
SA Phase 3	7.90 ± 1.8#Δ	48.04 ± 1.5†Δ*	158.94 ± 11.0†#*
SD Phase 3	4.51 ± 0.2#	2.63 ± 0.1†Δ	19.15 ± 2.7#

Values shown are the amount of time between the start of stretching or the start of shortening and the onset of phases 2 or 3, respectively. Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. All values are means ± SE. †P < 0.05 compared with Drosophila IFM. #P < 0.05 compared with Drosophila jump muscle. ΔP < 0.05 compared with Lethocerus IFM (one-way ANOVA with Dunn’s Method pairwise multiple comparisons). *Significant differences between timing of equivalent phases of SA and SD values for the indicated muscle type (Student’s t test, P < 0.05). IFM, indirect flight muscle; SA, stretch activation; SD, shortening deactivation.

Figure 6.

Rates of SD and SA obtained by fitting tension traces with Eqs. 2 and 3. A: example fits of phases 2–4 of tension transients following fiber shortening. Note that the y-axis scales are much smaller than those in Figs. 1 and 2 to show greater detail. Representative traces from a Drosophila IFM (blue), Drosophila jump muscle (green) and Lethocerus IFM (red) are shown. Left y-axis is for Lethocerus IFM and Drosophila jump muscle whereas the right axis is for Drosophila IFM. B: rates of phases 2–4 of an SD tension transient. C: rates of phases 2–4 of an SA tension transient. The upper and lower boundaries of each box represent the 75th and 25th percentile, respectively. The line within each box marks the median value. The error bars indicate the 90th and 10th percentiles. SA rates of Drosophila and Lethocerus IFM were previously reported in Glasheen et al., (27). Drosophila IFM, n = 10; Drosophila jump muscle, n = 13; Lethocerus IFM, n = 10. Horizontal bars indicate statistically significant differences between the values at the end of each bar, P < 0.05 (one-way ANOVA with Dunn’s method pairwise multiple comparisons). *Statistically significant from the equivalent SA rate in B. P < 0.05 (Student’s t test). IFM, indirect flight muscle; r₂, rate of force decrease (SA) or increase (SD) during phase 2 of the tension transient; r₃, rate of force increase (SA) or decrease (SD) during phase 3 of the tension transient; r₄, rate of force decrease (SA) or increase (SD) during phase 4 of the tension transient; SA, stretch activation; SD, shortening deactivation.

We compared the timing of the peak of SA phase 3 with the valley of SD phase 3 (Table 4 and Fig. 2, A and B). Relative to the start of lengthening or shortening, the SA phase 3 peak occurs later than the phase 3 valley of SD for all three muscles. How much later is vastly different between muscle types. The difference in Drosophila IFM is 3.4 ms, which when you divide SA time by SD time means the SA peak of phase 3 took 1.75-fold longer to occur from the start of lengthening than the SD phase 3 valley took to occur from start of shortening. For Lethocerus IFM, the difference between SA and SD for the onset of phase 3 was 140 ms, or it took 8.4-fold longer to reach the peak of phase 3 of SA compared with the valley of phase 3 of SD. For Drosophila jump muscle, the phase 3 peak of SA presented 45 ms after the phase 3 valley of SD, therefore taking 18.2-fold longer compared with SD.

For phase 2, the relationship between the timing of this local minimum (valley) of SA compared with this local maximum (peak) of SD is not consistent across muscle types. Only the Drosophila IFM phase 2 SA valley occurred later than its phase 2 SD peak. Drosophila IFM SA phase 2 occurred 0.12 ms later (1.09-fold longer) than Drosophila IFM SD phase 2. Lethocerus IFM phase 2 occurred 5.8-fold earlier for SA than SD, and phase 2 Drosophila jump muscle SA occurred 1.84-fold earlier than that of SD.

To delve into more details about SA and SD kinetics, we fit the tension transients following the length changes to obtain rates for phases 2–4 (Fig. 6). The results showed that all three phases of SD occur faster than the SA phases. The only possible exception might be Drosophila IFM SD phase 4 (r₄), but there was a large error for that measurement. These results agree with the timing of the peaks and valleys (Table 4). Drosophila IFM rates were the fastest or tied for fastest with the jump muscle. Lethocerus was slowest or tied for slowest except for SA phase 3 (r₃) and phase 4 (r₄) where it was faster than jump muscle (Fig. 6, B and C).

DISCUSSION

To our knowledge, this is the first study to measure SD amplitude in any muscle type as well as to characterize shortening deactivation in Drosophila and Lethocerus muscle types. Specifically, we are the first to define and develop a measurement for phase 3 amplitude, F_SD, along with the rates of phases 2–4 of a SD transient. There have been a few studies that mentioned the SD response in the IFM of other insect species (2, 10, 11, 22, 29) because IFM is the most obvious muscle type that benefits from SD. The lack of prior investigation is surprising given the impact SD can have on muscle performance in terms of increased work, power, and efficiency. These benefits suggest that SD should be present in other muscle types in addition to IFM. SD has been identified as a potentially important property in vertebrate heart muscles (2, 25, 48, 49). However, it has not been investigated in other muscle types, such as vertebrate skeletal muscle. Our recent finding that SA might increase muscle endurance in mouse soleus skeletal muscle heightened our interest in studying SD in more muscle types, including elucidating its roles, benefits, and mechanisms (24).

We chose to start our inquiry by evaluating some extreme cases, muscles that clearly would benefit from SD, Lethocerus and Drosophila IFM, and compare them with a muscle type that presumably would not benefit from SD, Drosophila jump muscle. IFMs generate power from cyclical contractions of antagonistic IFM pairs (DLMs and DVMs). During the shortening portion of the contractile cycle, a mechanism of delayed force reduction is required because decreasing [Ca²⁺] is not the mechanism by which the muscles relax during each contractile cycle (10). In contrast, Drosophila jump muscle would seem unlikely to benefit from SD as its main function is to power jumping, which uses a single shortening contraction. Investigating jump muscle should also provide insight into whether SD occurs in all muscle types even if it may not benefit the muscle. In other words, is some level of SD intrinsic to all muscle types?

Therefore, given our presumption that jump muscle would benefit the least, it was somewhat surprising that we found the Drosophila jump muscle produced 6.5-fold more F_SD than Drosophila IFM (Table 2). If we had just examined Drosophila jump muscle and IFM, we might have assumed that jump muscle has higher F_SD because jump muscle generates 18-fold greater isometric tension than IFM (Table 1) and that, in general, a muscle that generates higher tension produces more F_SD. However, by including Lethocerus IFM in our study, this assumption was shown to be incorrect. The Lethocerus IFM, which generated about half the amount of maximum isometric tension (F_{0 MAX}) compared with jump muscle, produced greater F_SD than Drosophila jump muscle.

When normalized to maximum isometric tension, F_{0 MAX}, our results make more sense with the known functions of each muscle. We observed that the two IFM muscles generated similar F_SD/F_{0 MAX} ratios and that both values are ∼3-fold greater than jump muscle F_SD/F_{0 MAX} (Table 2). This result suggests that one must look at relative amounts of SD (F_SD/F_{0 MAX}), rather than only the raw tension values (F_SD).

Normalization to F_{0 MAX} is also important for correctly interpreting SA. Drosophila IFM and jump muscle F_SA values were not different, which does not make physiological sense until they are normalized to F_{0 MAX} (Table 3). Normalized, Drosophila IFM F_SA/F_{0 MAX} is 22-fold greater than jump F_SA/F_{0 MAX}. Lethocerus IFM displayed the highest F_SA but when normalized to its F_{0 MAX}, the resultant value is 68% less than Drosophila IFM. Therefore, we end up with the same result as we did for F_SD/F_{0 MAX}: the two IFMs generate greater normalized SA amplitudes than jump muscle. This again makes sense for the IFM’s main function of having to generate oscillatory power using antagonistic muscle pairs. The large relative changes in F_SA and F_SD for both IFM muscle types is what enables asynchronous IFM to produce positive power during cyclical contractions at constant calcium levels (10, 25, 26). In contrast, our previous experiments with jump muscle show it cannot generate positive work and power at constant calcium levels (45, 46).

Although the amplitudes of SA and SD are clearly important, another factor is the timing (rates) of SA and SD. For SA and SD to be useful, their rate of force gain or loss must correctly correspond with the oscillation frequency of the muscles contractile cycle. If the phase 3 peak for SA or valley for SD occurs too early or too late in a contraction cycle then SA or SD will end up working against the muscle by decreasing work, power, and efficiency rather than increasing these parameters.

Our simplest way of testing the timing of SA and SD was to measure the time from the end of the imposed muscle length change until the phase 3 peak or valley of SA or SD, respectively (Table 4). This revealed that the fastest contracting muscle, Drosophila IFM, which powers a wing beat frequency (WBF) of 200 Hz at room temperature, displayed the shortest time to phase 3. For Lethocerus, with a WBF of ∼10 fold slower than Drosophila (38), the time to phase 3 was the longest of all three muscles tested.

Interpreting the results of jump muscle kinetics is difficult because the timing of its SA peak was similar to Lethocerus IFM, but the timing of its F_SD valley was comparable to Drosophila IFM. In other words, its F_SD occurred much sooner after its respective length change than F_SA. This, along with a large difference in jump muscle’s F_SA and F_SD magnitudes, made us wonder if these measures of F_SA and F_SD are truly equivalent to the F_SA and F_SD we measured in the IFM muscles. What we measured in all three muscle types meets the definition of SA and SD, a delayed increase in force and a delayed decrease in force, respectively. However, it could be that the evolution of SA and SD in IFMs were driven by the selective pressure of having to power cyclical muscle contractions, whereas for the jump muscle, there was not the same evolutionary pressure. Rather, SA and SD may simply be emergent properties emanating from thick and/or thin filament proteins together responding to a length increase or decrease. These properties likely either evolved randomly or were shaped by different functional pressures than those exerted on IFM.

To further explore SA and SD in jump muscle, we made measurements of tension following the respective length changes to see how the tension levels compared if SA and SD kinetics were synchronized in time, so as to be useful in a lengthening and shortening contraction cycle. This was done two ways. 1) We measured jump muscle tension levels at the same time point as jump muscle F_SA occurred. At this time point, the drop in tension from isometric averaged 3.8 mN/mm² (data not shown), which would only be ∼10% of F_{0 MAX}. Thus, there would not be much benefit from SD if it was timed with SA. 2) The reverse case, measuring jump muscle tension at the time point jump muscle F_SD occurred, is no more beneficial because the force gain from SA would only be around 5% of F_{0 MAX}. Based on these observations, jump muscle SA and SD did not evolve to work together in a sinusoidal shaped muscle contraction pattern.

Could either the jump muscle SA or SD be of physiological significance by itself? F_SA/F_{0 MAX} was 10%, which seems unlikely to be of physiological importance. Perhaps SD could be important as its F_SD/F_{0 MAX} value is higher, 30% of isometric tension. However, we do not see how it would assist in powering jumping, takeoff, or landing. Perhaps the jump muscle is also used for another function of which we are not aware.

SA and SD Characteristics of a Specific Muscle

Can we learn anything about the mechanisms behind SA and SD, such as the probability of SA and SD using the same molecular mechanism, by comparing the SA and SD characteristics of a specific muscle? In the Drosophila IFM, F_SD was 2-fold lower but occurred 1.7-fold sooner than F_SA, and the response to calcium was nearly identical (Hill coefficient and pCa₅₀; Tables 2 and 4, Fig. 5). Lethocerus IFM also showed a calcium response nearly identical between SA and SD, but showed an opposite response for relative amplitudes compared with Drosophila, with F_SD being 1.3-fold higher than F_SA. The timing was similar to Drosophila IFM with F_SD occurring 8.2-fold sooner than F_SA in Lethocerus IFM. The difference in timing between the phase 3 peak of SA and phase 3 valley of SD is larger for Lethocerus IFM than Drosophila IFM. However, if one considers the difference in the frequencies that generates maximum IFM power output at 15°C, 163 Hz for Drosophila versus 19 Hz for Lethocerus (27), it makes sense that the Lethocerus difference would be much greater if the difference in timing of F_SD and F_SA is related to the frequency of the contractile cycle used by the muscles.

Three possible explanations come to mind for F_SD phase 3 occurring sooner after a length change than F_SA phase 3. 1) If the contractile cycle of IFM is not a perfect sinusoid, such as if shortening occurs faster than lengthening, then F_SD needs to occur earlier following the length change than F_SA. 2) Perhaps the effects of SD need to be more immediate after shortening to keep the tension low in preparation to be lengthened again, whereas the effects of SA need to occur later after lengthening. A role of SA may be to counter the loss of force once a muscle has started shortening (due to the force-velocity relationship). 3) Finally, if the same mechanism drives both SA and SD, SD may simply be faster due to how the mechanism works, such as cross-bridge detachment perhaps being faster than attachment.

One could argue that the SD mechanism is more likely to be the same mechanism as SA for the two IFM muscles because their SA and SD traces much more closely mirror each other than the jump muscle SA and SD traces. The latest proposed SA mechanism for IFM with perhaps the strongest support is the thin filament mechanism proposed by Reedy, Bullard, and colleagues (22, 32, 38, 50). The primary mechanism postulates that stretching a partially or fully calcium-activated muscle moves the troponin-tropomyosin complex further off the myosin binding site enabling increased binding of cross bridges. Secondary mechanisms that may also help increase myosin attachment rate include the stretch of the muscle causing the thick filaments to twist for better alignment of myosins with their binding sites, and the compression of the thick and thin filament lattice bringing myosin closer to actin (38). The authors of these studies do not discuss SD, but if one presumes the opposite mechanism is responsible for SD (i.e., untwisting, decompression, and moving tropomyosin back over the myosin binding sites), our results of SD phase 3 being faster than SA phase 3 suggest these mechanisms are faster at detaching cross bridges (SD) than facilitating attachment of cross bridges (SA).

Method of Measuring SD Characteristics

Because this was our first attempt to quantify F_SD, we decided to start with a simple protocol that would replicate the ∼1% muscle length change amplitude that we have found produces maximum power output of skinned Drosophila IFM fibers (27, 41, 45). However, the length changes must be very fast (faster than in vivo) to make all four phases visible and distinguishable, particularly, phases 2 and 3. There are many variations that could be introduced to our protocol to potentially tease out more information about SD and SA. We think that the most promising protocols to test in the future include 1) increasing the number of length changes performed in a row to test for cumulative SA or SD responses, 2) adjusting the time between length changes, and 3) applying different amplitudes and rates of length changes to more closely replicate in vivo conditions. Our current paper provides a framework to pursue these studies. In addition, and of potentially greater interest, now that we have a method for measuring and quantifying SD, we can use the Drosophila system’s genetic engineering advantages to elucidate the details of the SD mechanism, similarly to what we and others have done for SA and muscle diseases (45, 46, 51–53).

SUPPLEMENTAL DATA

Supplemental Fig. S1: https://doi.org/10.6084/m9.figshare.17026964.v1.

GRANTS

This work was supported by NIAMS R01 Grant AR064274 to D.M.S.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.K.L. and D.M.S. conceived and designed research; A.K.L. and B.M.G. performed experiments; A.K.L., S.K.V.H., B.M.G., and D.M.S. analyzed data; A.K.L., S.K.V.H., B.M.G., and D.M.S. interpreted results of experiments; A.K.L., S.K.V.H., and D.M.S. prepared figures; A.K.L. and D.M.S. drafted manuscript; A.K.L., B.M.G., and D.M.S. edited and revised manuscript; A.K.L., S.K.V.H., B.M.G., and D.M.S. approved final version of manuscript.