Although most of us will never achieve the grace and dexterity of professional ballerina Misty Copeland, we each make sophisticated, complex movements every day. Even simple movements often involve coordinating many muscles throughout the body. Moreover, because we have so many muscles, there are often multiple ways that we could use them to make the same movement. So which ones do we use, and why?

Many studies into muscle control focus on how the muscles activate to perform a task like kicking a soccer ball. But muscles do more than just move the limbs; they also act on joints. Contracting a muscle exerts strain on bones and the ligaments that hold joints together. If these strains become excessive, they may cause pain and injury, and over a longer time may lead to arthritis. It would therefore make sense if the nervous system factored in the need to protect joints when turning on muscles.

The quadriceps are a group of muscles that stretch along the front of the thigh bone and help to straighten the knee. To investigate whether the nervous system selects muscle activations to avoid joint injuries, Alessando et al. studied rats that had one particular quadriceps muscle paralyzed. The easiest way for the rats to adapt to this paralysis would be to increase the activation of a muscle that performs the same role as the paralyzed one, but places more stress on the knee joint. Instead, Alessando et al. found that the rats increase the activation of a muscle that minimizes the stress placed on the knee, even though this made it more difficult for the rats to recover their ability to use the leg in certain tasks.

The results presented by Alessando et al. may have important implications for physical therapy. Clinicians usually work to restore limb movements so that a task is performed in a way that is similar to how it was done before the injury. But sometimes repairing the damage can change the mechanical properties of the joint – for example, reconstructive surgery may replace a damaged ligament with a graft that has a different strength or stiffness. In those cases, performing movements in the same way as before the surgery could place abnormal stress on the joint. However, much more research is needed before recommendations can be made for how to rehabilitate rats after injury, let alone humans.

The question of how the central nervous system (CNS) determines motor commands is usually answered in terms of task performance (Todorov and Jordan, 2002; Guigon et al., 2007; Scott, 2004): the CNS activates muscles according to how they contribute to task goals such as grasping an object or escaping a predator. Additional criteria based on energetics (Todorov and Jordan, 2002; Izawa et al., 2008; Kurtzer et al., 2006), dynamics (Uno et al., 1989), or kinematics (Flash and Hogan, 1985) can further constrain muscle activations, and the specific criteria used by the CNS to specify muscle activations underlying behavior has been the topic of considerable research (Scott, 2004; Todorov, 2004; Alessandro, 2016; Kistemaker et al., 2014; Prilutsky, 2000).

This previous work, however, has generally ignored another critical aspect of muscle actions: how muscles act on internal joint structures such as ligaments or articular cartilage between bones (Herzog et al., 2003). Although muscle actions on internal joint structures have been considered in clinical biomechanics or sports medicine in the context of injuries (Farrokhi et al., 2013; Pal et al., 2012; Felson, 2013; Konrath et al., 2017), their more general role in the context of the neural control of movement and how they are balanced with task goals remains poorly understood. We examine these issues in this study, evaluating the hypothesis that the nervous system chooses muscle activations to achieve task performance while minimizing internal joint stresses and strains.

Evaluating whether the CNS regulates stresses and strains within internal joint structures is challenging, primarily because of the complexity of joint mechanics. Since muscles can only affect task performance by acting through joint structures (Herzog et al., 2003), any manipulation affecting joint structures will usually also affect task performance. For example, damage to a ligament (Gutierrez et al., 2009; Needle et al., 2014; O'Connor et al., 1993) will alter both the distribution of stresses and strains within the joint (Chen et al., 2017; Gardinier et al., 2013) as well as how muscle forces are transmitted across the joint (Boeth et al., 2013). Similarly, although removal of sensory feedback from joint structures might compromise neural control of joint stresses and strains (O'Connor et al., 1993; O'Connor et al., 1992; Salo et al., 2002; Solomonow, 2006; Solomonow and Krogsgaard, 2001; Johansson et al., 1991), this sensory feedback can also provide information about task performance (Ferrell et al., 1985; Sjölander et al., 2002). These challenges have made it difficult to reach clear conclusions as to whether the CNS controls internal joint stresses and strains.

In the experiments reported here, we overcome these challenges by taking advantage of the unique properties of the rat knee joint (Figure 1a). Activation of the quadriceps muscles vastus lateralis (VL) and vastus medialis (VM) in the rat produces very similar forces at the distal tibia and so will produce similar joint torques (Sandercock et al., 2018). These muscles therefore have redundant contributions to task performance. On the other hand, VL and VM produce opposing mediolateral forces on the patella (Sandercock et al., 2018; Lin et al., 2004; Wilson and Sheehan, 2010). In the rat, these mediolateral patellar forces are not transmitted to the tibia but are balanced by contact forces between the patella and the femur within the trochlear groove (Sandercock et al., 2018). To avoid overloading patellofemoral contact forces and prevent joint pain or injury, the CNS should therefore balance activation of VM and VL to minimize net mediolateral forces on the patella while producing the knee torque necessary to achieve task goals. Since VM and VL in the rat have redundant contributions to task performance but opposing contributions to joint stresses, we can use the rat knee joint to design tractable experiments to evaluate whether the CNS regulates joint stresses and strains.

Figure 1

Download asset Open asset

We examined neural adaptation strategies following paralysis of VL (Figure 1b). If the CNS prioritizes task performance, the easiest adaptation strategy would be to increase activation of VM. Because VM and VL have redundant contributions to task performance, this strategy would restore joint kinematics without requiring changes in other muscle activations. This straightforward restoration of task performance, however, would come at the cost of unbalanced mediolateral patellar forces. On the other hand, if the CNS prioritizes joint integrity, it should decrease activation of VM and increase activation of rectus femoris (RF) since RF produces minimal mediolateral force on the patella. This strategy to minimize joint stresses, however, would come at the cost of a more complex, potentially incomplete restoration of task performance: increased activation of RF will restore the knee torque lost by VL paralysis but will also introduce an additional hip flexion torque. This strategy will therefore either require compensatory activity in hip extensor muscles or introduce residual deviations in joint kinematics. Finally, if the CNS considers both task performance and internal joint stresses, it might preferentially increase RF activation while maintaining activation of VM. This strategy would minimize mediolateral forces on the patella while also limiting deviations of joint kinematics, reflecting a compromise between control of task performance and internal joint state. Observing either of the last two strategies would demonstrate that the CNS considers internal joint stresses since the preferential increase of RF activation reflects minimization of mediolateral patellar forces. Note that vastus intermedius (VI) is very small in the rat and so is unlikely to contribute substantially to the adaptation following VL paralysis. Our results show that rats compensated for the loss of VL by preferentially increasing activation of RF, consistent with the hypothesis that the CNS chooses muscle activations to minimize stresses and strains to internal joint structures.

We hypothesized that the CNS activates muscles to minimize internal joint stresses while achieving task requirements. We found that animals compensated for VL paralysis by increasing activation of RF while maintaining the activation of VM at levels similar to those observed before the perturbation. This strategy restores the knee extension torque lost after VL paralysis while causing minimal additional mediolateral patellar forces. The increased RF activation resulted in a persistent perturbation in joint kinematics due to the extra hip flexor torque from RF, but global aspects of limb kinematics were restored. These results suggest that the CNS considers internal joint stresses and strains when choosing muscle activations.

The main result of this study is the observation that animals compensated for VL paralysis by preferentially increasing activation of RF. If one only considers muscle contributions to task performance, this adaptation strategy is counterintuitive; because VM and VL have redundant contributions to knee extension movements (Sandercock et al., 2018), restoring task performance would be most easily accomplished by increasing activation of VM (Figure 1b). This strategy would restore the same joint kinematics observed before VL paralysis without requiring any additional compensation. In contrast, increased activation of RF, while restoring the knee extension torque lost by VL paralysis, also introduces an additional flexion torque at the hip (Johnson et al., 2008; Greene, 1963) (Figure 1). Thus, this strategy would either require additional changes in other muscles across the limb, or introduce residual deviations in joint kinematics. The increased activation of RF we observed is therefore difficult to reconcile with strategies for muscle coordination focused solely on considerations of task performance.

This strategy is consistent, however, with the hypothesis that the CNS takes into account the effect of muscle activity on the stresses and strains within a joint. Increasing activation of RF restores knee extension function while producing negligible additional mediolateral forces on the patella (Figure 1). On the other hand, higher activation of VM would increase the medial force on the patella and cause excess patellofemoral loading. Hence, our results suggest that the CNS considers these mediolateral forces when selecting muscle activations to compensate for VL paralysis.

We also observed a selective increase in the mass of RF that likely reflects a use-dependent strengthening of that muscle, although we did not confirm this by measuring peak forces or myosin content (Blaauw et al., 2013). This result indirectly confirms the preferential increase of RF activity after VL paralysis. Furthermore, it suggests that at late time points of the adaptation, the loss of VL was compensated both by higher neural activity and by additional force generating capability of RF. Finally, this result suggests that the increased RF activation was not specific to treadmill locomotion, but was also used in behaviors we did not measure here; if there was increased VM or VI activation in other behaviors, we would have observed an increase in the mass of those muscles. The observation of no change in the mass of VI also suggests that there was no consistent change in the activation of this muscle during the adaptation period, though we did not record from VI in these experiments due its small size and inaccessibility for implantation.

We did not observe a complete silencing of VM, even after 7 weeks of adaptation: VM activation levels were on average similar to those observed before VL paralysis. This implies that a residual medial force on the patella persisted throughout the adaptation period. Such residual VM activity might reflect a compromise between unbalanced mediolateral patellar forces and perturbations in joint kinematics caused by the increased activation of RF. In the context of optimization approaches to motor control (Todorov and Jordan, 2002; Scott, 2004), this result suggests a cost function that includes terms for task performance as well as internal joint stresses. Alternatively, the residual activation of VM might reflect habitual persistence of the strategy used by the CNS prior to VL paralysis (de Rugy et al., 2012; Loeb, 2012). Although our study cannot distinguish between these two explanations, in either case the fact that RF and not VM activity was increased demonstrates the influence of criteria related to the state of internal joint structures.

The particular aspects of task performance that drive adaptation are potentially complex. Although muscle paralysis caused persistent changes at individual joint kinematics, we observed restoration of global limb kinematics. This result is consistent with ideas such as those used in uncontrolled manifold analyses (Latash et al., 2007) and with previous results examining adaptation following paralysis of triceps surae in rats and cats (Chang et al., 2009; Bauman and Chang, 2013). In this context, our results suggest a hierarchical control of movement, in which higher level variables related to task goals (i.e. global limb variables) are preserved after VL paralysis while lower level variables related to task execution (i.e. individual joint angles) are allowed to vary.

Many criteria have been proposed to influence the activation of muscles during behavior (Alessandro, 2016; Prilutsky, 2000), but they generally do not consider the internal state of joints. Although some of these criteria might indirectly minimize joint stresses and strains (e.g. minimum jerk [Flash and Hogan, 1985] or torque change [Uno et al., 1989]), they would not predict the selective increase in RF activation observed here. Including criteria that explicitly consider the consequences of muscle activation on joint stresses and strains might improve the predictions of such optimization approaches.

Alternatively, minimization of internal joint stresses and strains in intact subjects might be accomplished without explicit control. The co-evolution of neural control and musculoskeletal anatomy (Valero-Cuevas, 2015) might result in a neuromechanical system such that muscle activations optimizing task performance criteria also optimize internal joint state criteria. For example, when VM and VL are intact, a criterion that minimizes muscle metabolism (Kistemaker et al., 2014; Prilutsky, 2000), or that considers signal dependent variability (Harris and Wolpert, 1998; Haruno and Wolpert, 2005) would predict a distributed activation of VM and VL. This distributed activation of VM and VL might result from optimization over evolutionary time scales (Giszter, 2015) and be implemented in coordinative structures such as muscle synergies (Tresch and Jarc, 2009; Alessandro et al., 2013a; Alessandro et al., 2013b) within the spinal cord (Levine et al., 2014; Hart and Giszter, 2010; Takei et al., 2017) or other brain regions (Overduin et al., 2015). This distributed activation could in turn balance net mediolateral patellar forces, even though such a balance was not explicitly considered. After injuries that alter limb properties (Gutierrez et al., 2009; O'Connor et al., 1993), this implicit regulation of internal joint state might be compromised (Needle et al., 2014). In these situations, the CNS might be forced into a solution that requires a tradeoff between task performance and internal joint stresses and strains, such as that observed in our experiments.

These results have implications to clinical syndromes affecting joint health, such as knee osteoarthritis (Felson, 2013; Mahmoudian et al., 2017) and patellofemoral pain (Smith et al., 2018) (PFP) in humans. Although many factors have been shown to contribute to PFP (Smith et al., 2018; Fagan and Delahunt, 2008), one common suggestion is that PFP results from an imbalance in mediolateral patellar forces (Pal et al., 2012). Our results show that although the CNS regulates these mediolateral forces, this regulation is imperfect since VM activation persisted after VL paralysis. PFP might result from similarly imperfect regulation of patellar forces, potentially reflecting a compromise between criteria related to task performance and internal joint states. More broadly, our experiments highlight the importance of considering the consequences of rehabilitation on internal joint stresses and strains following injuries (Farrokhi et al., 2013; Shull et al., 2013; Simic et al., 2011), rather than solely focusing on restoration of task performance (Hollands et al., 2012).

Regulation of internal joint stresses and strains, either in intact subjects or following injury, might be accomplished by feedforward or feedback mechanisms. Sensory receptors in the joint can convey information about strains in ligaments or stresses between bones. Activity in these receptors can be conveyed to spinal interneuronal populations (Solomonow et al., 1987; Iles et al., 1990) and other areas (Sjölander et al., 2002) to drive rapid feedback control of muscles (Ferrell, 1980) or longer term adaptation of muscle activations in order to avoid potential injury. The CNS could also derive information about internal joint structures indirectly from muscle proprioceptors (Wilmink and Nichols, 2003), combining information about muscle forces and lengths with ‘internal’ models of muscle anatomy to estimate joint stresses and strains. Similarly, predictive models of muscle actions, combined with efference copy of motor commands, could be used to make feedforward predictions of internal joint stresses and strains. The strong coherence between VM and VL across frequencies (Laine et al., 2015) observed in humans is consistent with the importance of coordinated activation of these muscles (Brøchner Nielsen et al., 2017), whether by feedforward or feedback processes. This coordinated muscle activation is also similar to ideas of muscle synergies (Tresch and Jarc, 2009; Alessandro et al., 2013a; Alessandro et al., 2013b), so that the balanced activation of muscles within a synergy would reflect regulation of joint stresses and strains. Note that neural control of internal joint states might be relatively coarse, only responding to large deviations in joint states inducing discomfort or pain (Grigg and Greenspan, 1977). Such a coarse control of joint state might reflect the ‘good enough’ habitual strategy (de Rugy et al., 2012) described above, potentially resulting in the residual activation of VM observed in our experiments.

Further experiments will be necessary to better understand the neural control of joint stresses and strains. Although the muscle paralysis by denervation used in these experiments has been used in many studies to examine neural adaptation strategies (Bauman and Chang, 2013; Dambreville et al., 2016; Frigon and Rossignol, 2007; Bennett et al., 2012; Bouyer et al., 2001; Maas et al., 2007), the loss of sensory information from the paralyzed muscle might contribute to the observed changes in muscle activation (Akay et al., 2014). Perturbations directly manipulating patellar forces or paralyzing VM or RF might provide complementary experiments to further evaluate this hypothesis. Cutting the nerves to VM and RF selectively, however, is difficult because they branch extensively immediately after leaving the common quadriceps nerve, making them hard to isolate (Greene, 1963). The nerve to VL was simpler to cut because it persists as a single trunk for a long distance from the common quadriceps nerve.

Our experiments exploited the unique mechanical properties of the rat knee joint: the clear distinction between muscle contributions to task performance and to patellar forces in the rat allowed us to make simple predictions about adaptation strategies consistent or inconsistent with neural regulation of internal joint stresses and strains. Such distinctions are more difficult to make in humans because the patellar mobility at extended knee angles complicates how quadriceps muscles affect both patellar mechanics and limb kinematics (Farahmand et al., 1998). The rat knee joint might therefore provide a tractable experimental model to examine the neural control of joint stresses and strains in both healthy subjects and following injury.

We have uploaded files with all source data and code for data analysis for each of the figures showing primary data.

1. 1. Akay T
   2. Tourtellotte WG
   3. Arber S
   4. Jessell TM

   (2014) Degradation of mouse locomotor pattern in the absence of proprioceptive sensory feedback

   PNAS 111:16877–16882.

   https://doi.org/10.1073/pnas.1419045111

   • PubMed
   • Google Scholar
2. 1. Alessandro C
   2. Carbajal JP
   3. d'Avella A

   (2013a) A computational analysis of motor synergies by dynamic response decomposition

   Frontiers in Computational Neuroscience 7:191.

   https://doi.org/10.3389/fncom.2013.00191

   • PubMed
   • Google Scholar
3. 1. Alessandro C
   2. Delis I
   3. Nori F
   4. Panzeri S
   5. Berret B

   (2013b) Muscle synergies in neuroscience and robotics: from input-space to task-space perspectives

   Frontiers in Computational Neuroscience 7:43.

   https://doi.org/10.3389/fncom.2013.00043

   • PubMed
   • Google Scholar
4. Book

   1. Alessandro C

   (2016)

   Motor Control and Learning Theories

   In: Pons J, Raya R, Gonzalez J, editors. Emerging Therapies in Neurorehabilitation II. Switzerland: Springer International Publishing. pp. 225–250.

   • Google Scholar
5. 1. Barr DJ
   2. Levy R
   3. Scheepers C
   4. Tily HJ

   (2013) Random effects structure for confirmatory hypothesis testing: keep it maximal

   Journal of Memory and Language 68:255–278.

   https://doi.org/10.1016/j.jml.2012.11.001

   • Google Scholar
6. 1. Bauman JM
   2. Chang YH

   (2010) High-speed X-ray video demonstrates significant skin movement errors with standard optical kinematics during rat locomotion

   Journal of Neuroscience Methods 186:18–24.

   https://doi.org/10.1016/j.jneumeth.2009.10.017

   • PubMed
   • Google Scholar
7. 1. Bauman JM
   2. Chang YH

   (2013) Rules to limp by: joint compensation conserves limb function after peripheral nerve injury

   Biology Letters 9:20130484.

   https://doi.org/10.1098/rsbl.2013.0484

   • PubMed
   • Google Scholar
8. 1. Bennett SW
   2. Lanovaz JL
   3. Muir GD

   (2012) The biomechanics of locomotor compensation after peripheral nerve lesion in the rat

   Behavioural Brain Research 229:391–400.

   https://doi.org/10.1016/j.bbr.2012.01.040

   • PubMed
   • Google Scholar
9. 1. Blaauw B
   2. Schiaffino S
   3. Reggiani C

   (2013) Mechanisms modulating skeletal muscle phenotype

   Comprehensive Physiology 3:1645–1687.

   https://doi.org/10.1002/cphy.c130009

   • PubMed
   • Google Scholar
10. 1. Boeth H
    2. Duda GN
    3. Heller MO
    4. Ehrig RM
    5. Doyscher R
    6. Jung T
    7. Moewis P
    8. Scheffler S
    9. Taylor WR

    (2013) Anterior cruciate ligament-deficient patients with passive knee joint laxity have a decreased range of anterior-posterior motion during active movements

    The American Journal of Sports Medicine 41:1051–1057.

    https://doi.org/10.1177/0363546513480465

    • PubMed
    • Google Scholar
11. 1. Bouyer LJ
    2. Whelan PJ
    3. Pearson KG
    4. Rossignol S

    (2001) Adaptive locomotor plasticity in chronic spinal cats after ankle extensors neurectomy

    The Journal of Neuroscience 21:3531–3541.

    https://doi.org/10.1523/JNEUROSCI.21-10-03531.2001

    • PubMed
    • Google Scholar
12. 1. Brøchner Nielsen NP
    2. Tucker K
    3. Dorel S
    4. Guével A
    5. Hug F

    (2017) Motor adaptations to local muscle pain during a bilateral cyclic task

    Experimental Brain Research 235:607–614.

    https://doi.org/10.1007/s00221-016-4826-y

    • PubMed
    • Google Scholar
13. 1. Chang YH
    2. Auyang AG
    3. Scholz JP
    4. Nichols TR

    (2009) Whole limb kinematics are preferentially conserved over individual joint kinematics after peripheral nerve injury

    Journal of Experimental Biology 212:3511–3521.

    https://doi.org/10.1242/jeb.033886

    • PubMed
    • Google Scholar
14. 1. Chen T
    2. Wang H
    3. Warren R
    4. Maher S

    (2017) Loss of ACL function leads to alterations in tibial plateau common dynamic contact stress profiles

    Journal of Biomechanics 61:275–279.

    https://doi.org/10.1016/j.jbiomech.2017.07.024

    • PubMed
    • Google Scholar
15. 1. Dambreville C
    2. Charest J
    3. Thibaudier Y
    4. Hurteau MF
    5. Kuczynski V
    6. Grenier G
    7. Frigon A

    (2016) Adaptive muscle plasticity of a remaining agonist following denervation of its close synergists in a model of complete spinal cord injury

    Journal of Neurophysiology 116:1366–1374.

    https://doi.org/10.1152/jn.00328.2016

    • PubMed
    • Google Scholar
16. 1. de Rugy A
    2. Loeb GE
    3. Carroll TJ

    (2012) Muscle coordination is habitual rather than optimal

    Journal of Neuroscience 32:7384–7391.

    https://doi.org/10.1523/JNEUROSCI.5792-11.2012

    • PubMed
    • Google Scholar
17. 1. Fagan V
    2. Delahunt E

    (2008) Patellofemoral pain syndrome: a review on the associated neuromuscular deficits and current treatment options

    British Journal of Sports Medicine 42:489–495.

    https://doi.org/10.1136/bjsm.2008.046623

    • PubMed
    • Google Scholar
18. 1. Farahmand F
    2. Tahmasbi MN
    3. Amis AA

    (1998) Lateral force-displacement behaviour of the human Patella and its variation with knee flexion--a biomechanical study in vitro

    Journal of Biomechanics 31:1147–1152.

    https://doi.org/10.1016/S0021-9290(98)00125-0

    • PubMed
    • Google Scholar
19. 1. Farrokhi S
    2. Voycheck CA
    3. Tashman S
    4. Fitzgerald GK

    (2013) A biomechanical perspective on physical therapy management of knee osteoarthritis

    Journal of Orthopaedic & Sports Physical Therapy 43:600–619.

    https://doi.org/10.2519/jospt.2013.4121

    • PubMed
    • Google Scholar
20. 1. Felson DT

    (2013) Osteoarthritis as a disease of mechanics

    Osteoarthritis and Cartilage 21:10–15.

    https://doi.org/10.1016/j.joca.2012.09.012

    • PubMed
    • Google Scholar
21. 1. Ferrell WR

    (1980) The adequacy of stretch receptors in the cat knee joint for signalling joint angle throughout a full range of movement

    The Journal of Physiology 299:85–99.

    https://doi.org/10.1113/jphysiol.1980.sp013112

    • PubMed
    • Google Scholar
22. 1. Ferrell WR
    2. Baxendale RH
    3. Carnachan C
    4. Hart IK

    (1985) The influence of joint afferent discharge on locomotion, proprioception and activity in conscious cats

    Brain Research 347:41–48.

    https://doi.org/10.1016/0006-8993(85)90887-X

    • PubMed
    • Google Scholar
23. 1. Filipe VM
    2. Pereira JE
    3. Costa LM
    4. Maurício AC
    5. Couto PA
    6. Melo-Pinto P
    7. Varejão AS

    (2006) Effect of skin movement on the analysis of hindlimb kinematics during treadmill locomotion in rats

    Journal of Neuroscience Methods 153:55–61.

    https://doi.org/10.1016/j.jneumeth.2005.10.006

    • PubMed
    • Google Scholar
24. 1. Flash T
    2. Hogan N

    (1985) The coordination of arm movements: an experimentally confirmed mathematical model

    The Journal of Neuroscience 5:1688–1703.

    https://doi.org/10.1523/JNEUROSCI.05-07-01688.1985

    • PubMed
    • Google Scholar
25. 1. Frigon A
    2. Rossignol S

    (2007) Plasticity of reflexes from the foot during locomotion after denervating ankle extensors in intact cats

    Journal of Neurophysiology 98:2122–2132.

    https://doi.org/10.1152/jn.00490.2007

    • PubMed
    • Google Scholar
26. 1. Gardinier ES
    2. Manal K
    3. Buchanan TS
    4. Snyder-Mackler L

    (2013) Altered loading in the injured knee after ACL rupture

    Journal of Orthopaedic Research 31:458–464.

    https://doi.org/10.1002/jor.22249

    • PubMed
    • Google Scholar
27. 1. Giszter SF

    (2015) Motor primitives--new data and future questions

    Current Opinion in Neurobiology 33:156–165.

    https://doi.org/10.1016/j.conb.2015.04.004

    • PubMed
    • Google Scholar
28. 1. Greene E

    (1963)

    Anatomy of the rat

    Transactions of the American Philosophical Society (American Philosophical Society).

    • Google Scholar
29. 1. Grigg P
    2. Greenspan BJ

    (1977) Response of primate joint afferent neurons to mechanical stimulation of knee joint

    Journal of Neurophysiology 40:1–8.

    https://doi.org/10.1152/jn.1977.40.1.1

    • PubMed
    • Google Scholar
30. 1. Guigon E
    2. Baraduc P
    3. Desmurget M

    (2007) Computational motor control: redundancy and invariance

    Journal of Neurophysiology 97:331–347.

    https://doi.org/10.1152/jn.00290.2006

    • PubMed
    • Google Scholar
31. 1. Gutierrez GM
    2. Kaminski TW
    3. Douex AT

    (2009) Neuromuscular control and ankle instability

    PM&R 1:359–365.

    https://doi.org/10.1016/j.pmrj.2009.01.013

    • PubMed
    • Google Scholar
32. 1. Harris CM
    2. Wolpert DM

    (1998) Signal-dependent noise determines motor planning

    Nature 394:780–784.

    https://doi.org/10.1038/29528

    • PubMed
    • Google Scholar
33. 1. Hart CB
    2. Giszter SF

    (2010) A neural basis for motor primitives in the spinal cord

    Journal of Neuroscience 30:1322–1336.

    https://doi.org/10.1523/JNEUROSCI.5894-08.2010

    • PubMed
    • Google Scholar
34. 1. Haruno M
    2. Wolpert DM

    (2005) Optimal control of redundant muscles in step-tracking wrist movements

    Journal of Neurophysiology 94:4244–4255.

    https://doi.org/10.1152/jn.00404.2005

    • PubMed
    • Google Scholar
35. 1. Herzog W
    2. Longino D
    3. Clark A

    (2003) The role of muscles in joint adaptation and degeneration

    Langenbeck's Archives of Surgery 388:305–315.

    https://doi.org/10.1007/s00423-003-0402-6

    • Google Scholar
36. 1. Hollands KL
    2. Pelton TA
    3. Tyson SF
    4. Hollands MA
    5. van Vliet PM

    (2012) Interventions for coordination of walking following stroke: systematic review

    Gait & Posture 35:349–359.

    https://doi.org/10.1016/j.gaitpost.2011.10.355

    • PubMed
    • Google Scholar
37. 1. Iles JF
    2. Stokes M
    3. Young A

    (1990) Reflex actions of knee joint afferents during contraction of the human quadriceps

    Clinical Physiology 10:489–500.

    https://doi.org/10.1111/j.1475-097X.1990.tb00828.x

    • Google Scholar
38. 1. Izawa J
    2. Rane T
    3. Donchin O
    4. Shadmehr R

    (2008) Motor adaptation as a process of reoptimization

    Journal of Neuroscience 28:2883–2891.

    https://doi.org/10.1523/JNEUROSCI.5359-07.2008

    • PubMed
    • Google Scholar
39. 1. Johansson H
    2. Sjölander P
    3. Sojka P

    (1991)

    A sensory role for the cruciate ligaments

    Clinical Orthopaedics and Related Research 268:161–178.

    • Google Scholar
40. 1. Johnson WL
    2. Jindrich DL
    3. Roy RR
    4. Reggie Edgerton V

    (2008) A three-dimensional model of the rat hindlimb: musculoskeletal geometry and muscle moment arms

    Journal of Biomechanics 41:610–619.

    https://doi.org/10.1016/j.jbiomech.2007.10.004

    • PubMed
    • Google Scholar
41. 1. Kistemaker DA
    2. Wong JD
    3. Gribble PL

    (2014) The cost of moving optimally: kinematic path selection

    Journal of Neurophysiology 112:1815–1824.

    https://doi.org/10.1152/jn.00291.2014

    • PubMed
    • Google Scholar
42. 1. Konrath JM
    2. Saxby DJ
    3. Killen BA
    4. Pizzolato C
    5. Vertullo CJ
    6. Barrett RS
    7. Lloyd DG

    (2017) Muscle contributions to medial tibiofemoral compartment contact loading following ACL reconstruction using semitendinosus and gracilis tendon grafts

    PloS One 12:e0176016.

    https://doi.org/10.1371/journal.pone.0176016

    • PubMed
    • Google Scholar
43. 1. Kurtzer I
    2. Pruszynski JA
    3. Herter TM
    4. Scott SH

    (2006) Primate upper limb muscles exhibit activity patterns that differ from their anatomical action during a postural task

    Journal of Neurophysiology 95:493–504.

    https://doi.org/10.1152/jn.00706.2005

    • PubMed
    • Google Scholar
44. 1. Laine CM
    2. Martinez-Valdes E
    3. Falla D
    4. Mayer F
    5. Farina D

    (2015) Motor neuron pools of synergistic thigh muscles share most of their synaptic input

    Journal of Neuroscience 35:12207–12216.

    https://doi.org/10.1523/JNEUROSCI.0240-15.2015

    • PubMed
    • Google Scholar
45. 1. Latash ML
    2. Scholz JP
    3. Schöner G

    (2007) Toward a new theory of motor synergies

    Motor Control 11:276–308.

    https://doi.org/10.1123/mcj.11.3.276

    • PubMed
    • Google Scholar
46. 1. Levine AJ
    2. Hinckley CA
    3. Hilde KL
    4. Driscoll SP
    5. Poon TH
    6. Montgomery JM
    7. Pfaff SL

    (2014) Identification of a cellular node for motor control pathways

    Nature Neuroscience 17:586–593.

    https://doi.org/10.1038/nn.3675

    • PubMed
    • Google Scholar
47. 1. Lin F
    2. Wang G
    3. Koh JL
    4. Hendrix RW
    5. Zhang LQ

    (2004) In vivo and noninvasive three-dimensional patellar tracking induced by individual heads of quadriceps

    Medicine & Science in Sports & Exercise 36:93–101.

    https://doi.org/10.1249/01.MSS.0000106260.45656.CC

    • PubMed
    • Google Scholar
48. 1. Loeb GE

    (2012) Optimal isn’t good enough

    Biological Cybernetics 106:757–765.

    https://doi.org/10.1007/s00422-012-0514-6

    • Google Scholar
49. 1. Maas H
    2. Prilutsky BI
    3. Nichols TR
    4. Gregor RJ

    (2007) The effects of self-reinnervation of cat medial and lateral gastrocnemius muscles on hindlimb kinematics in slope walking

    Experimental Brain Research 181:377–393.

    https://doi.org/10.1007/s00221-007-0938-8

    • PubMed
    • Google Scholar
50. 1. Mahmoudian A
    2. van Dieёn JH
    3. Bruijn SM
    4. Baert IAC
    5. Faber GS
    6. Luyten FP
    7. Verschueren SMP

    (2017) Dynamic and static knee alignment at baseline predict structural abnormalities on MRI associated with medial compartment knee osteoarthritis after 2 years

    Gait & Posture 57:46–51.

    https://doi.org/10.1016/j.gaitpost.2017.05.024

    • PubMed
    • Google Scholar
51. 1. Needle AR
    2. Baumeister J
    3. Kaminski TW
    4. Higginson JS
    5. Farquhar WB
    6. Swanik CB

    (2014) Neuromechanical coupling in the regulation of muscle tone and joint stiffness

    Scandinavian Journal of Medicine & Science in Sports 24:737–748.

    https://doi.org/10.1111/sms.12181

    • PubMed
    • Google Scholar
52. 1. O'Connor BL
    2. Visco DM
    3. Brandt KD
    4. Myers SL
    5. Kalasinski LA

    (1992) Neurogenic acceleration of osteoarthrosis. the effects of previous neurectomy of the articular nerves on the development of osteoarthrosis after transection of the anterior cruciate ligament in dogs

    The Journal of Bone & Joint Surgery 74:367–376.

    https://doi.org/10.2106/00004623-199274030-00007

    • PubMed
    • Google Scholar
53. 1. O'Connor BL
    2. Visco DM
    3. Brandt KD
    4. Albrecht M
    5. O'Connor AB

    (1993) Sensory nerves only temporarily protect the unstable canine knee joint from osteoarthritis. evidence that sensory nerves reprogram the central nervous system after cruciate ligament transection

    Arthritis & Rheumatism 36:1154–1163.

    https://doi.org/10.1002/art.1780360817

    • PubMed
    • Google Scholar
54. 1. Overduin SA
    2. d'Avella A
    3. Roh J
    4. Carmena JM
    5. Bizzi E

    (2015) Representation of muscle synergies in the primate brain

    Journal of Neuroscience 35:12615–12624.

    https://doi.org/10.1523/JNEUROSCI.4302-14.2015

    • PubMed
    • Google Scholar
55. 1. Pal S
    2. Besier TF
    3. Draper CE
    4. Fredericson M
    5. Gold GE
    6. Beaupre GS
    7. Delp SL

    (2012) Patellar tilt correlates with vastus lateralis: vastus medialis activation ratio in maltracking patellofemoral pain patients

    Journal of Orthopaedic Research 30:927–933.

    https://doi.org/10.1002/jor.22008

    • PubMed
    • Google Scholar
56. Book

    1. Pinheiro JC
    2. Bates DM

    (2000)

    Mixed Effects Models in S and S-Plus

    Springer.

    • Google Scholar
57. Software

    1. Pinheiro J
    2. Bates D
    3. DebRoy S
    4. Sarkar D

    (2017)

    R Core Team. nlme: Linear and nonlinear mixed effects models

    R Package Version.
58. 1. Prilutsky BI

    (2000) Coordination of two- and one-joint muscles: functional consequences and implications for motor control

    Motor Control 4:1–44.

    https://doi.org/10.1123/mcj.4.1.1

    • PubMed
    • Google Scholar
59. 1. Salo PT
    2. Hogervorst T
    3. Seerattan RA
    4. Rucker D
    5. Bray RC

    (2002) Selective joint denervation promotes knee osteoarthritis in the aging rat

    Journal of Orthopaedic Research 20:1256–1264.

    https://doi.org/10.1016/S0736-0266(02)00045-1

    • PubMed
    • Google Scholar
60. 1. Sandercock T
    2. Wei Q
    3. Dhaher Y
    4. Pai DK
    5. Tresch MC

    (2018) Vastus lateralis and vastus medialis produce distinct mediolateral forces on the Patella but similar forces on the tibia in the rat

    Journal of Biomechanics.

    https://doi.org/10.1016/j.jbiomech.2018.09.007

    • Google Scholar
61. 1. Scott SH

    (2004) Optimal feedback control and the neural basis of volitional motor control

    Nature Reviews Neuroscience 5:532–545.

    https://doi.org/10.1038/nrn1427

    • PubMed
    • Google Scholar
62. 1. Shull PB
    2. Silder A
    3. Shultz R
    4. Dragoo JL
    5. Besier TF
    6. Delp SL
    7. Cutkosky MR

    (2013) Six-week gait retraining program reduces knee adduction moment, reduces pain, and improves function for individuals with medial compartment knee osteoarthritis

    Journal of Orthopaedic Research 31:1020–1025.

    https://doi.org/10.1002/jor.22340

    • PubMed
    • Google Scholar
63. 1. Simic M
    2. Hinman RS
    3. Wrigley TV
    4. Bennell KL
    5. Hunt MA

    (2011) Gait modification strategies for altering medial knee joint load: a systematic review

    Arthritis Care & Research 63:405–426.

    https://doi.org/10.1002/acr.20380

    • PubMed
    • Google Scholar
64. 1. Sjölander P
    2. Johansson H
    3. Djupsjöbacka M

    (2002) Spinal and supraspinal effects of activity in ligament afferents

    Journal of Electromyography and Kinesiology 12:167–176.

    https://doi.org/10.1016/S1050-6411(02)00017-2

    • PubMed
    • Google Scholar
65. 1. Smith BE
    2. Selfe J
    3. Thacker D
    4. Hendrick P
    5. Bateman M
    6. Moffatt F
    7. Rathleff MS
    8. Smith TO
    9. Logan P

    (2018) Incidence and prevalence of patellofemoral pain: a systematic review and meta-analysis

    PloS One 13:e0190892.

    https://doi.org/10.1371/journal.pone.0190892

    • PubMed
    • Google Scholar
66. 1. Solomonow M
    2. Baratta R
    3. Zhou BH
    4. Shoji H
    5. Bose W
    6. Beck C
    7. D'Ambrosia R

    (1987) The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability

    The American Journal of Sports Medicine 15:207–213.

    https://doi.org/10.1177/036354658701500302

    • PubMed
    • Google Scholar
67. 1. Solomonow M
    2. Krogsgaard M

    (2001) Sensorimotor control of knee stability. A review

    Scandinavian Journal of Medicine and Science in Sports 11:64–80.

    https://doi.org/10.1034/j.1600-0838.2001.011002064.x

    • PubMed
    • Google Scholar
68. 1. Solomonow M

    (2006) Sensory-motor control of ligaments and associated neuromuscular disorders

    Journal of Electromyography and Kinesiology 16:549–567.

    https://doi.org/10.1016/j.jelekin.2006.08.004

    • PubMed
    • Google Scholar
69. 1. Takei T
    2. Confais J
    3. Tomatsu S
    4. Oya T
    5. Seki K

    (2017) Neural basis for hand muscle synergies in the primate spinal cord

    PNAS 114:8643–8648.

    https://doi.org/10.1073/pnas.1704328114

    • PubMed
    • Google Scholar
70. 1. Todorov E
    2. Jordan MI

    (2002) Optimal feedback control as a theory of motor coordination

    Nature Neuroscience 5:1226–1235.

    https://doi.org/10.1038/nn963

    • PubMed
    • Google Scholar
71. 1. Todorov E

    (2004) Optimality principles in sensorimotor control

    Nature Neuroscience 7:907–915.

    https://doi.org/10.1038/nn1309

    • PubMed
    • Google Scholar
72. 1. Tresch MC
    2. Jarc A

    (2009) The case for and against muscle synergies

    Current Opinion in Neurobiology 19:601–607.

    https://doi.org/10.1016/j.conb.2009.09.002

    • PubMed
    • Google Scholar
73. 1. Tysseling VM
    2. Janes L
    3. Imhoff R
    4. Quinlan KA
    5. Lookabaugh B
    6. Ramalingam S
    7. Heckman CJ
    8. Tresch MC

    (2013) Design and evaluation of a chronic EMG multichannel detection system for long-term recordings of hindlimb muscles in behaving mice

    Journal of Electromyography and Kinesiology 23:531–539.

    https://doi.org/10.1016/j.jelekin.2012.11.014

    • PubMed
    • Google Scholar
74. 1. Uno Y
    2. Kawato M
    3. Suzuki R

    (1989) Formation and control of optimal trajectory in human multijoint arm movement

    Biological Cybernetics 61:89–101.

    https://doi.org/10.1007/BF00204593

    • Google Scholar
75. 1. Valero-Cuevas FJ

    (2015)

    Biosystems & Biorobotics

    Fundamentals of Neuromechanics, Biosystems & Biorobotics, 8, London, Springer-Verlag.

    • Google Scholar
76. 1. Wilmink RJ
    2. Nichols TR

    (2003) Distribution of heterogenic reflexes among the quadriceps and triceps surae muscles of the cat hind limb

    Journal of Neurophysiology 90:2310–2324.

    https://doi.org/10.1152/jn.00833.2002

    • PubMed
    • Google Scholar
77. 1. Wilson NA
    2. Sheehan FT

    (2010) Dynamic in vivo quadriceps lines-of-action

    Journal of Biomechanics 43:2106–2113.

    https://doi.org/10.1016/j.jbiomech.2010.04.002

    • PubMed
    • Google Scholar

Author details

1. Cristiano Alessandro

   Department of Physiology, Northwestern University, Chicago, United States

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   For correspondence

   cristiano.alessandro@northwestern.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0655-4189

2. Benjamin A Rellinger

   Department of Biomedical Engineering, Northwestern University, Evanston, United States

   Contribution

   Conceptualization, Investigation

   Competing interests

   No competing interests declared

3. Filipe Oliveira Barroso

   Department of Physiology, Northwestern University, Chicago, United States

   Contribution

   Formal analysis, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0228-6447

4. Matthew C Tresch

   1. Department of Physiology, Northwestern University, Chicago, United States
   2. Department of Biomedical Engineering, Northwestern University, Evanston, United States
   3. Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago, United States
   4. Shirley Ryan AbilityLab, Chicago, United States

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Validation, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   m-tresch@northwestern.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9994-3989

• 1,262

  views

• 136

  downloads

• 31

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.