Immunohistochemistry for GFP-Aequorin and quantification of muscle fibers

Get Complete Project Material File(s) Now! »

An open-loop access: sensorimotor circuits in the spinal cord across vertebrates

In the particular case of spinal sensorimotor circuits, a great wealth of anatomical and electrophysiological data has been accumulated over the years. However, being able to elaborate broader models in order to fit those data onto observed behaviors still remains a challenge, largely due to the fact that available techniques have prevented monitoring sensory inputs concomitantly with motor outputs until recently.

Ascending sensory feedback

While descending inputs schematically provide the motor command to spinal sensorimotor circuits, ascending afferents to the spinal cord mainly provide sensory information. In mammalian vertebrates, ascending sensory inputs include proprioceptive inputs (group Ia and II afferents from, respectively, primary and secondary endings of muscles spindles, and Ib afferents from Golgi tendon organs), cutaneous inputs (chemosensitive group III/Aδ and group IV/C fibers from nociceptive receptors). They have been extensively studied in the context of local spinal reflex pathways (Knikou, 2008; Rossignol et al., 2006) (Figure 2B).
The simplest, and fastest, somatic reflex is the monosynaptic pathway between primary sensory afferents from primary muscle spindles (Ia) and homonymous alpha motoneurons in the ventral horn of the corresponding segment grey matter. This is the basic myotatic reflex that is elicited by a muscle stretch due to a tendon tap, but is also involved in tonus and postural adjustments (Guertin, 2013). The experimental analog of the Ia reflex, the Hoffman reflex (H-reflex), where the mechanical stretch is replaced by a sub-threshold electrical stimulation of the afferent nerve, has been extensively used to investigate spinal sensorimotor circuits, and in particular presynaptic and reciprocal inhibition (Jankowska, 1992; Knikou, 2008), see (section 2.2.1).
Golgi tendon organs are force-sensitive receptors located at the muscle-tendinous junction, that are activated by passive and active muscle force. The Ib reflex arc, also known as the “inverse myotatic reflex”, is a disynaptic pathway by which group Ib sensory afferents from Golgi tendon organs inhibit alpha-motoneurons. This is the reflex arc responsible for the abrupt termination of the myotatic reflex, the well-known “clasp-knife” phenomenon (Hultborn, 2006). Although stimulating the Golgi tendon organs at rest cannot induce any movement, the Ib reflex has been suggested to be important for regulating muscle stiffness (Knikou, 2008).
While group Ib afferents from Golgi tendon organs provide information about the tension developed during muscle contraction, and group Ia afferents from primary muscle spindles inform spinal circuits about the dynamic of changes in muscle length, group II afferents from muscle spindle secondary endings provide information of muscle length itself (Jankowska and Edgley, 2010). Group Ia, Ib and II muscle afferents taken together constitute what is generally termed the “proprioception” input. Together with cutaneous afferents from nociceptors (Aδ and C fibers) and other muscle afferents (thinly myelinated group III and unmyelinated group IV fibers), group II muscle afferents constitute the flexion reflex afferents (FRA) involved in the withdrawal reflex, by which a painful stimulus lead to withdrawal of the limb through ipsilateral flexion and contralateral extension (Eccles and Lundberg, 1958). This sensorimotor reflex, more sophisticated than the “myotatic” and “inverse myotatic” reflexes, involves at least to two interneurons to either activate or inhibit the ipsilateral flexor or extensor alpha-motoneurons over several spinal segments (Guertin, 2013).
Sensory feedback pathways in non-mammalian vertebrates still remain unclear. Indeed, there is no clear equivalent to mammalian peripheral proprioceptive receptors in swimming vertebrates. However, in the lamprey, intra-spinal mechanosensitive receptors called the “edge cells” (Grillner et al., 1984) might provide movement-related sensory feedback (Di Prisco et al., 1990). Interestingly, it has recently been proposed that edge cells could be modulated by GABAergic cerebrospinal fluid contacting neurons (CSFns) (Jalalvand et al., 2014). Similar CSFns, called “Kolmer-Agduhr” cells, have been described in the zebrafish, and were able to trigger slow swim upon optical activation (Wyart et al., 2009). Another sensory feedback pathway in larvae and adult zebrafish is the lateral line system (Ghysen and Dambly-Chaudière, 2007). Mechanosensory hair cells in the lateral line neuromasts provide information about the water flow, contributing to orientating the fish against the water, a behavior called rheotaxis (Olszewski et al., 2012; Suli et al., 2012) (Figure 3C).

Modulation of spinal circuitry from extrinsic inputs

Both descending motor inputs and ascending sensory feedback can modulate the activity of the spinal CPG. Indeed, if the CPG is able to generate the basic locomotor patterns, dynamic sensorimotor interactions with both supraspinal and peripheral inputs continuously modulate these patterns to achieve a flexible adaptation to the environment. Such interactions take place in a phase-dependent (swing/stance) and state-dependent (forward/backward) manner, that is extrinsic inputs will result in different modulations depending on the ongoing phase of the locomotor cycle (Rossignol et al., 2006).
As discussed in section 2.1.1, supraspinal pathways, such as the MLR and its projections through the reticulospinal tract, can induce locomotion in “fictive preparations”, i.e. isolated spinal cord or decerebrated adult cat preparations. However, descending pathways, whether carrying sensory or motor information, can also modulate ongoing locomotion. Such modulation can be achieved either though modulation of brainstem command circuitry, or through direct modulation of spinal circuitry (McCrea, 2001).
Vestibular inputs (relaying information about balance and posture) modulate the activity of reticulospinal neurons with a phasic pattern during fictive locomotion in lampreys, thereby avoiding a counteractive drive from reticulospinal neurons during ongoing locomotion (Bussières and Dubuc, 1992). A recent study in zebrafish larvae suggested that vestibular inputs are able to differentially recruit dorsal and ventral premotor spinal microcircuits during postural correction, possibly prefiguring the mammalian modular organization of spinal flexor/extensor microcircuits (Bagnall and McLean, 2014). The influence of visual feedback on the control of locomotion can be experienced on a daily basis when one needs to anticipate and adjust his gait to avoid an obstacle (Rossignol, 1996). New experimental paradigms, such as the optomotor response in zebrafish (Orger et al., 2008), have started to shed light on the neural circuitry responsible for visually induced locomotion.

Implications for plasticity after spinal cord injury

The emerging concept that intrinsic spinal circuits can produce adaptive locomotion with modulation by sensory feedback, independently, at least to some extent, from supra-spinal inputs, bears important consequences for new neurorehabilitative strategies after spinal cord injury.
Experimental paradigms with adult cats walking on a treadmill have demonstrated that neither bilateral lesion of the dorsolateral spinal cord (interrupting cortico- and rubrospinal tracts) (Jiang and Drew, 1996), nor bilateral lesion of the ventrolateral spinal cord (interrupting vestibulo- and reticulospinal tracts) (Brustein and Rossignol, 1998), could permanently suppress quadrupedal locomotion. However, after unilateral complete hemisection at the lower thoracic (T13) level, interrupting both dorsal and ventral descending pathways, cats showed a complete paralysis of the ipsilateral hindlimb during the first three days, followed by a progressive recovery over the following three weeks (Rossignol and Frigon, 2011). Interestingly, this recovery was accompanied by a modification of the step cycle, forelimb/hindlimb and left/right coordination (Martinez et al., 2012). These results suggest that the intrinsic spinal circuitry is able to produce locomotion even after removal of all supraspinal inputs, and that this recovery is underpinned by extensive reorganization of the spinal sensorimotor network (Martinez and Rossignol, 2013). They also suggest that treadmill-induced locomotor training, by providing sensory feedback, is crucial to drive the reorganization of spinal circuits (Rossignol and Frigon, 2011).
To test this hypothesis of a plastic spinal CPG, Rossignol et al. designed a dual-lesion paradigm in which a first hemisection performed at the T10/T11 spinal level is followed, after several weeks of locomotor training and complete recovery, by a complete spinal transection at the T13 level (Barrière et al., 2008; Martinez and Rossignol, 2013). The major finding was that cats regained full locomotor performance after only 24 hours, without any training of pharmacological intervention (Barrière et al., 2008), therefore indicating that intrinsic changes within the spinal CPG had indeed occurred during the rehabilitation period, and could be retained after the complete removal of supraspinal inputs.

READ  Intelligent behavioral based IDS: Neural networks basis

A genetic toolbox for targeting populations of neurons

Considered the large number of cells involved into spinal sensorimotor circuits, even in a simple vertebrate such as the zebrafish, one crucial requirement to investigate their functional role is to be able to specifically target the neural subpopulation of interest. Rather than relying on morphological cues, identification of specific promoters, and new tools to efficiently generate and screen transgenic lines, have recently allowed researchers to take full advantage of the optical and genetic accessibility of the zebrafish model.
The most straightforward approach to target a given neuronal population is to identify a specific gene with selective expression in the population of interest, isolate its promoter sequence and generate a bacterial artificial chromosome (BAC) incorporating the putative promoter, the gene and an attached reporter such as GFP. The plasmid is then microinjected into embryos at the single-cell stage for homologous recombination to occur, and injected zebrafish are subsequently screened for fluorescence in order to establish the transgenic line (Asakawa et al., 2013). Such approach have been successfully used to produce transgenic lines labeling cranial motoneurons or trigeminal/Rohon-Beard sensory neurons under control of the Islet-1 promoter (Higashijima et al., 2000). This transgenic line was then used to investigate the role of Rohon-Beard and trigeminal neurons in the sensorimotor escape circuitry (Douglass et al., 2008).
This BAC approach can be combined with the bipartite Gal4/UAS system, widely used in drosophila, which relies on the specific expression of the yeast Gal4 transcriptional activator to drive the expression of the reporter gene placed under the control of repetitive Gal4-responsive upstream activator sequences (UAS) (Asakawa and Kawakami, 2009; Davison et al., 2007). Enhanced reporter expression can be obtained using Gal4-VP16 (Koster and Fraser, 2001) or Gal4FF (Asakawa and Kawakami, 2009) fusion sequences and multiple (14X) repeats of the UAS Stable zebrafish transgenic lines using the Gal4/UAS system has been achieved using Tol2-mediated transposition: a plasmid carrying the Tol2 element is injected in zebrafish embryos with the Tol2 transposase mRNA, generating genome-wide insertions in the zebrafish genome (Asakawa et al., 2008; Kawakami et al., 2000). Tol2-mediated Gal4-UAS transgenesis has been used to successfully generate wide enhancer-trap screens, leading to identification of a large number of stable transgenic lines selectively labeling subsets of spinal neurons (Abe et al., 2011; Asakawa and Kawakami, 2009; Satou et al., 2013; Scott et al., 2007).
Another recent approach for genetic targeting of neurons in zebrafish is to combine viral gene delivery, using for instance rabies of sindbis viruses, together with the Tet system (Zhu et al., 2009). The Tet system works in a similar fashion to the Gal4/UAS system, with the transactivator (itTA) binding to the tTA-responder element (Ptet) to drive transcription of the downstream gene (Gossen and Bujard, 1992). However, the Tet system has the advantage of being able to be regulated with doxycycline, which binds to tTA and dramatically reduces its affinity to Ptet, turning off the expression of the gene of interest (Zhu et al., 2009). Interestingly, such silencing could also be used to generate sparse labeling in pan-neuronal HuC transgenic lines (Zhu et al., 2009). Combing the Tet and Gal4 systems provide exciting opportunities for combinatorial gene targeting of several neuronal populations of interest in zebrafish.

Table of contents :

Part A. Sensorimotor integration in the spinal cord, from behaviors to circuits: new tools to close the loop? 
1 A closed-loop approach to sensorimotor behaviors
1.1. Defining sensorimotor behaviors
1.1.1. Eliciting sensory input
1.1.2. Measuring motor output
1.2. Modulating sensorimotor behaviors
1.2.1. Sensory feedback
1.2.2. Neuromodulation
1.3. Modeling sensorimotor behaviors
1.3.1. Behavioral computations
1.3.2. Circuits computations
2 An open-loop access: sensorimotor circuits in the spinal cord across vertebrates
2.1. Extrinsic inputs to spinal sensorimotor circuits
2.1.1. Descending motor control
2.1.2. Ascending sensory feedback
2.2. Intrinsic spinal sensorimotor circuitry
2.2.1. Sensorimotor interneuronal networks
2.2.2. Spinal central pattern generator
2.3. Dynamic spinal sensorimotor interactions
2.3.1. Modulation of spinal circuitry from extrinsic inputs
2.3.2. Implications for plasticity after spinal cord injury
3 Closing the loop? Optogenetic manipulation of spinal sensorimotor circuits in zebrafish
3.1. Genetic targeting of spinal sensorimotor circuits in zebrafish
3.1.1. Identified sensorimotor neurons in the zebrafish spinal cord
3.1.2. A genetic toolbox for targeting populations of neurons
3.2. Optogenetic tools for monitoring and breaking neural circuits
3.2.1. Reporters: monitoring neural circuits
3.2.2. Actuators: breaking neural circuits
3.3. The escape response as a model for sensorimotor integration
3.3.1. The escape response and its supraspinal control
3.3.2. Monitoring spinal neurons during active locomotion
Part B. Mechanosensory neurons enhance motor output in the zebrafish spinal cord during active locomotion
1 Introduction
2 Results
2.1. Bioluminescence signals reflect the level of recruitment of motor neurons during movement
2.2. Spinal motor neurons recruitment is enhanced in the presence of mechanosensory feedback
2.3. Mechanosensory neurons are recruited during active but not fictive locomotion
2.4. Silencing mechanosensory neurons impairs escape responses
3 Discussion
3.1. Investigating sensorimotor integration in the spinal cord during ongoing locomotion
3.2. Non-invasive bioluminescence monitoring of genetically targeted neurons in motion
3.3. A closed-loop circuit within the spinal cord for mechanosensory integration
4 Methods
4.1. Zebrafish care and strains
4.2. Generation of transgenic lines
4.3. Immunohistochemistry for GFP-Aequorin and quantification of muscle fibers
4.4. Monitoring of neuronal activity with GFP-Aequorin bioluminescence
4.5. High-speed behavior recording
4.6. Bioluminescence analysis
4.7. Kinematics analysis
4.8. Calcium imaging of spinal motor neurons
4.9. Ventral nerve root recording (VNR)
4.10. Calcium imaging of spinal sensory neurons
4.11. Behavioral analysis of freely moving BoTxLCB larvae
4.12. Statistical analysis
Part C. From spatial to genetic targeting: a paradigm shift for neurosurgery..
1 Introduction
2 How we moved to genetically targeted neuroscience
2.1. From morphological to genetic identification of neurons
2.2. Genetic targeting of neurons in tractable animal models
2.3. A toolbox for manipulating genetically identified neurons
3 Moving toward genetically targeted neurosurgery
3.1. Candidate diseases for genetically targeted neurosurgery
3.2. Genetic identification and cellular targeting in the human brain
3.3. Genetically targeted neuromodulation and neuroablation in patients
4 Two challenges for a paradigm shift


Related Posts