Final Report Summary - SPINAL CORD REPAIR (Spinal locomotor circuits: organization and repair after injury)
The spinal cord is the part of the central nervous system that executes practically all movements, receives sensory information about the body and interacts with the environment, and controls many autonomic functions. The spinal cord is by no mean a passive transmitting cable but contains a large number of neural networks that are directly involved in coordinating and processing many fundamental body functions. Dysfunctions and diseases, such as spinal cord injury (SCI), that directly or indirectly affect these spinal networks therefore lead to a loss or impairments of movements and sensation, chronic pain syndrome, bladder, bowel and sexual dysfunctions. The personal consequences for and economical costs of these disabling injuries and diseases are very large. An improved treatment and rehabilitation of patients with spinal cord dysfunctions will consequently be of socio-economical benefit.
In the present project, we have been studying the basic function and development of spinal networks generating locomotion and provide new knowledge about the repair of the spinal cord following injury with specific reference to motor functions. Our combined effort aimed at understanding the normal function and mechanisms underlying dysfunction of the spinal cord and upon such a base develop effective treatment strategies for spinal cord injury.
Neural circuits in the spinal cord, called central pattern generators (CPGs), can produce locomotor movements. The locomotor activity pattern is the result of the circuit design and the interplay between the firing properties of the CPG constituent neurons and their synaptic interactions. Spinal cord injury impairs the function of the locomotor circuitry and results in paralysis. An understanding of the molecular mechanisms of the assembly of the spinal locomotor circuitry, the function of its key excitatory components and the intrinsic plasticity in the healthy and injured spinal cord is a prerequisite for designing novel therapeutic methods to restore locomotor function after spinal cord lesion.
The work-packages (WPs) of the project have been designed to elucidate the key molecular pathways responsible for the development and assembly of the spinal circuitry for locomotion. The intrinsic function and modulation the spinal circuitry has been examined in the healthy spinal cord by combining sophisticated molecular, anatomical, pharmacological and electrophysiological tools. The mechanisms of plasticity and reorganisation of the circuitry have been examined in the injured spinal cord as have the mechanism to promote regeneration of the lesioned axons. This project has integrated knowledge on the development and normal spinal cord function together with biological interventions aiming at protecting and repairing the injured spinal cord.
To understand how the healthy spinal cord produces locomotor movements two main mechanisms have been studied; the first is the development and organisation of the spinal locomotor networks, the second is the modulatory mechanisms that regulate the locomotor activity. An important research task was therefore to elucidate how the spinal sensori-motor circuitry develops and define the key neuronal elements in the locomotor network that generate the motor rhythm.
In the injured spinal cord two main forms of biological intervention can promote recovery of function; the first is axon regeneration, the second is promotion and control of plasticity. Spinal cord injury invariably damages ascending and descending axons, and this is the main cause of disability. If axons can be induced to grow across the site of injury to make connections below and above it, useful function is returned. Describing the basic mechanisms of regeneration and how they relate to the functional circuits that are present in the spinal cord is therefore of great importance to advance functional recovery.
Project context and objectives:
The spinal cord contains neuronal networks responsible for the precise execution of different motor tasks ranging from locomotor movements, equilibrium control, hand - finger movements, and many autonomic functions. These networks are called into action by command signals from the brain. Spinal cord injury (SCI) disrupts the interaction between the brain and the networks in the spinal cord resulting in loss or impairments of movements, sensation and autonomic functions. To promote recovery of motor function after SCI, we have tried to understand:
(1) how the spinal sensori-motor network is assembled during development and identify the types of neurons responsible for the generation of locomotor activity in the healthy spinal cord,
(2) characterise at the cellular and synaptic level the mechanisms of plasticity that takes place within these networks after injury, and
(3) determine the mechanisms of regeneration to facilitate the reconnection between the brain and spinal cord through axonal regrowth.
Organisation of spinal locomotor networks
Neural circuits in the spinal cord, called central pattern generators (CPGs), can produce locomotor movements, such as walking. These networks interact continuously with centers in the brain and with sensory feedback to adapt the different phases of the step cycle to internal and external demands. The organisation of the spinal locomotor CPGs has been characterised in two lower vertebrate model systems, the lamprey and Xenopus tadpole. It consists of a network of excitatory glutamatergic and inhibitory glycinergic interneurons. The glutamatergic interneurons project ipsilaterally and provide the excitatory drive necessary to produce sustained rhythmic locomotor activity, while the glycinergic interneurons projecting to the contralateral side and mediate the reciprocal inhibition responsible for the alternation of activity between the two sides of the spinal cord.
In mammals, the motor pattern with flexion and extension coordination at different joints requires a more complex circuit organisation. Several types of interneurons have been identified based on the reflex pathways that activate them and their pattern of activation during locomotor rhythm. One key component of the locomotor network that was not yet defined was the neuronal assembly that is directly involved in generating the rhythm. Network pharmacology and lesion studies have shown that these rhythm-generating neurons are glutamatergic excitatory interneurons with ispilateral projections. Activation of these neurons from the brain provides the direct switch to turn on locomotion. Characterizing and defining the network organisation of these neurons is therefore of paramount importance for understanding the normal function of locomotor network and how these network are developmentally controlled. Moreover, these neurons are prime targets for intervention therapies aimed at improving locomotor function following spinal cord injury.
Molecular signalling defining the ipsilateral projection of excitatory interneurons
Axon guidance molecules EphA4 and ephrinB3 identify a group of excitatory spinal interneurons that are rhythmically active during locomotion. Mice with targeted deletions in the genes for ephA4 and ephrinB3 display a characteristic hopping rabbit-like gait. EphA4-positive neurons, whose cell bodies are located in the spinal cord, aberrantly cross the midline in ephA4 and ephrinB3 knockouts. These observations suggest that at least some EphA4-positive spinal neurons are excitatory CPG neurons, whose projections normally remain ipsilateral, but in mutants cross the midline and result in synchronous activity on the left and right side of the spinal cord. Recordings from EphA4 positive cells have now demonstrated that most EphA4 positive neurons are rhythmically active during locomotion. A subset of these neurons provides excitation to motoneurons in the same segment. However, it was not known if these excitatory interneurons constitute a homogenous cell population and how they are related to the interneurons characterised by the expression of specific transcription factors in the developing spinal cord.
Developmental characterisation of spinal interneurons
Spinal cord cell differentiation is orchestrated by a series of transcription factors expressed in temporal and spatial patterns. Four early metaclasses of ventral horn interneurons (V0, V1, V2, and V3) have been described based on transcription factor expression, which may relate to functional phenotypes in the adult spinal cord. V0 interneurons express Dbx1/Evx1 and are laminae VII-VIII excitatory and inhibitory commissural interneurons. V1 interneurons express En1 are all inhibitory and include Renshaw cells and Ia inhibitory interneurons. V2 neurons express Lhx3 early in development and then differentiate into V2a and V2b classes, expressing the transcription factors Chx10 and Gata2/3, respectively. V3 express Sim1 and are a mixed group of ipsi- and contralaterally excitatory projecting neurons. Experiments in zebrafish and mouse have shown that neurons expressing Chx10 and its zebrafish homolog Alx are all excitatory interneurons and project ipsilaterally, while Gata2/3 cells are mainly inhibitory ipsilaterally projecting neurons.
This molecular characterisation of early neuronal cell population has provided a new basis for genetically dissecting the neuronal circuits that control locomotion in the vertebrate spinal cord. Such methods are under development and include genetic silencing or activation of molecularly defined populations of interneurons. They have been also used to study the role of interneurons controlling left-right alternation and possibly the speed of locomotion. However, the individual contribution of the different types of ipsilaterally projecting excitatory interneurons in the generation of locomotion was still unclear before the project started.
Sensory contribution to spinal motor coordination
The intraspinal locomotor network receives a prominent feedback from the moving limb that helps regulate and adapt the locomotor movements to unexpected perturbations. These feedback systems include those that signal the progression of the step cycle (hip movements) and those that signal the load that the limb exerts on the ground. The efficacy of these feedback systems can be modulated during the locomotor cycle at both the presynaptic and the interneuronal levels, so that the feedback signals may be more efficacious in one part of the movement cycle than in another. These prominent feedback systems are critical for an optimal performance of the locomotor system and are present in all vertebrate systems investigated. In spinal animals they play a prominent role for a good coordination, but they need to be recalibrated through a continuous activation by locomotor movements, otherwise the motor pattern degenerates. Patients with partial spinal cord lesions can similarly profit from locomotor training on the treadmill, under conditions when the load and movement trajectory are carefully controlled.
Transcriptional programs for sensory connectivity
Transcriptional programs control the differentiation of sensory afferent projections at multiple steps. Recent evidence supports an important role for the Runt domain transcription factor Runx3 in early specification of proprioceptive afferents - that regulate muscle length (Ia-afferents) and force (Ib-afferents) and the regulation of the neurotrophic factor receptor TrkC. Moreover, graded expression of Runx3 within proprioceptive afferents controls laminar termination within the spinal cord, and thus indirectly determines which neuronal populations will be contacted by proprioceptive afferent subclasses. In addition to early transcriptional programs, proprioceptive afferents are also influenced by target-derived factors, which retrogradely induce transcriptional programs important for functionality of spinal networks. In particular, the neurotrophic factor NT-3 induces the expression of the transcription factor Er81 in proprioceptive afferents. Er81 expression within group Ia proprioceptive afferents in turn controls the elaboration of a ventrally projecting trajectory and thus the establishment of monosynaptic connections with motor neurons. Er81 mutation in mice has dramatic consequences for motor behaviour, resulting in highly uncoordinated movement due to the absence of group Ia proprioceptive afferent connectivity in the ventral spinal cord. However, it was still unclear before the start of the project how the activity of interneurons within the locomotor circuitry is affected by the lack of Ia afferent during development.
Modulation spinal networks in the healthy spinal cord
To account for how the spinal circuitry generates the appropriate locomotor pattern, it is not sufficient to characterise how the network develops and how its connectivity is assembled. The detailed properties of the synaptic transmission between the nerve cells and the membrane properties of the different types of nerve cells that form the network need also to be understood. The core part of the segmental network is composed by excitatory interneurons acting via ionotropic AMPA and NMDA receptors and glycinergic neurons via glycine receptors. These components are responsible for the cycle to cycle operation. There is, however, a number of modulatory systems that can fine tune the properties of the network, by modifying the properties by individual types of nerve cells or the synaptic properties. This modulation primarily involves activation of G-protein-coupled metabotropic receptors. The modulation can be a result of transmitters released from locomotor network neurons themselves (intrinsic modulation) such as glutamate activating via the family of metabotropic glutamate receptors (mGluRs) acting at both the pre- and postsynaptic levels. There are also transmitters released from neurons not being part of the network that activate for instance 5-HT1a, dopamine (D2), GABAB or tachykinin receptors (extrinsic modulation). These intrinsic and extrinsic modulatory systems play a fundamental role for the basic network function and for the induction of both short-term (seconds or minutes) or long-term (hours or days) plasticity of the activity of the locomotor circuitry. It has be important to further explore the modulator actions not only for an understanding of the neural mechanisms that turn on and optimise network operation, but also to find pharmacological tools for improving the function of the locomotor network in patients with an incomplete spinal cord lesion.
In networks of this type, many factors co-vary at both the single cell and the network level, which makes it difficult to explore different possible network solutions intuitively. One important tool consists of detailed biophysically realistic modelling that in combination with experimental analysis has help exploring the function of the spinal locomotor network. This approach has successfully been applied to the lamprey and tadpole CPG. Within the proposed consortium, the Grillner laboratory has made this particular methodology available for analysis of zebrafish and mouse models.
Repair of the injured spinal cord
An injury to the spinal cord disrupts the interaction between the brain centres and the spinal networks and results in loss of motor and sensory functions. Although the circuitry necessary to generate movement located in the spinal cord is still intact, it lacks the appropriate drive to trigger its activity. In human patients, the spinal cord is usually partially transsected, with damage to both grey and white matter. Much of the disability results from the interruption of ascending and descending white matter tracts. However spontaneous axon regeneration does not occur. For a long time there has been an extensive focus on the promotion of axon regeneration in the injured cord. Three main factors prevent axon regeneration in the spinal cord:
(1) inhibition of regeneration by the glial scar;
(2) inhibition of regeneration by molecules associated with oligodendrocytes and CNS myelin; and
(3) poor spontaneous regenerative response of CNS neurons.
The first treatments designed to overcome these forms of inhibition are now available. However, their effect is limited because they only induce axon regeneration over short distances and in a small number of fibres. The main inhibitory molecules in the glial scar are chondroitin sulphate proteoglycans (CSPGs), and the axon guidance molecules semaphorin3 and ephrin B2. Digestion of CSPGs with chondroitinase ABC and blocking of ephrin signalling have been shown to promote axon regeneration with recovery of function. Oligodendrocytes express the growth inhibitory membrane proteins Nogo-A, Myelin associated glycoprotein (MAG) and Oligodendrocyte and myelin glycoprotein (Omgp). Inhibition of Nogo-A with antibodies and by interfering with its signalling pathway has promoted axon regeneration and recovery. The regenerative response of CNS axons may also be enhanced through neurotrophins, raised cAMP, protein kinase C inhibition and other means, again with the effect of increasing axon regeneration.
Thus, different strategies have been developed to promote regeneration of the damaged axons that consist of interfering with growth inhibitory molecules and transplantation techniques to favour neurons survival and axon regeneration. However, at the start of the project, there was a lack of information about the optimal dosage of treatment to be use, the timing of administration, their combination with one another and with rehabilitative training.
Plasticity in the injured spinal cord
After human spinal cord injury there is a period lasting up to one year during which there is spontaneous improvement in neurological function. A similar though more rapid improvement occurs after rodent spinal cord injury, which can be enhanced by appropriate rehabilitative therapy. For instance, even in completely severed spinal cords, stepping or weight-bearing behaviour can be induced by treadmill training or other means. Much of this recovery is probably due to plasticity, and recent studies have demonstrated plastic changes anatomically. It is probable that enhancing plasticity after injury will promote better recovery in human patients because young children show remarkable recoveries after partial spinal cord injury provided that they are injured before the end of the critical period for spinal cord plasticity. It is also apparent that the promotion of plasticity is an easier and more achievable aim than long-distance axon regeneration. An additional reason to focus on plasticity treatments is that they will probably help patients with chronic injuries.
Plasticity in the spinal cord encompasses axonal sprouting, terminal sprouting and changes of synaptic strength. In most cases of return of function after injury through plasticity it is not known which of these processes have occurred. The main focus of investigations has been the corticospinal tract. Three of the treatments that promote functional recovery after SCI, anti NogoA, chondroitinase ABC and inosine promote collateral sprouting of the corticospinal tract above the injury together with some axonal regrowth through the injury. While these treatments enhance functional recovery, the changes in spinal cord circuitry that underlie these improvements are not worked out. Neither is the mechanism by which the treatments affect synaptic behaviour understood.
In animal models of cortical plasticity, if appropriate changes in connectivity and cortical mapping are to occur, plasticity must be driven by voluntary behaviour. It is therefore often assumed that there will be co-operative effects between plasticity-inducing treatments and rehabilitation therapy. This has yet to be proven after spinal cord injury.
The overall objective of the SPINAL CORD REPAIR project is to restore motor function after spinal cord injury. The SPINAL CORD REPAIR partners have been working towards this goal by defining the key elements of the vertebrate locomotor network and how these elements are remodelled following SCI, by developing new strategies to protect and activate the remaining circuitry, as well as by promoting regeneration of the damaged axons.
The specific objectives of the project were to obtain:
Basic knowledge on:
- the molecular and cellular identification of the interneurons responsible for rhythm generation to unravel the intrinsic function of the network underlying locomotion;
- the influence of proprioceptive sensory feedback on spinal locomotor circuit development and plasticity;
- the cellular and synaptic mechanisms responsible for the modulation of locomotor circuitry activity;
- membrane proteins and extracellular matrix-related factors that restrict plasticity in the adult spinal cord;
- the changes in morphology, intrinsic properties and synaptic connectivity of locomotor network interneurons after SCI in animal models and spinal cord injured patients.
Translational knowledge on:
- restoration of locomotion and remodelling of the locomotor circuitry;
- plasticity of the morphology and connectivity of the component neurons of the locomotor circuitry following injury;
- functional effects on motor function of promoting spinal cord plasticity with extracellular matrix-related factors;
- the influence treatment that promotes regeneration and rehabilitative training on locomotion and network interneurons morphology, intrinsic properties and synaptic connectivity.
Throughput knowledge on:
- novel pharmacological tools to activate the locomotor circuitry and restore function;
- novel candidate molecules for protection and regeneration of damaged axons;
- identification of the optimal timing and method of administration of chondroitinase following partial SCI;
- identification of optimal training and treatment paradigms to optimise functional recovery in spinal cord injured animals and patients.
This collaborative project aims at gaining fundamental understanding of the function and plasticity of the spinal networks responsible for motor behaviour and their dysfunction after injury. The project has been designed to bridge the basic mechanism of spinal cord function with translational approach to protect the injured tissue and promote re-growth and recovery of function in animal models as well as patients. To address the scientific objectives, the research has been divided into seven interrelated strategic lines of work, which aim:
1. to identify the interneuron classes responsible for the generation of the locomotor rhythmic pattern using a combination of molecular, electrophysiological and anatomical tools;
2. to examine the molecular mechanisms responsible for the development and assembly of the sensori-motor circuitry and the role of sensory feedback in shaping the identity of locomotor network interneurons;
3. to determine cellular and synaptic mechanisms by which modulatory systems operating within the locomotor network regulate the frequency of the locomotor rhythm at the short- and long-time scale;
4. to examine the role of extracellular matrix in mediating the plastic changes in synaptic connectivity after spinal cord injury and assess novel therapeutic strategies to protect the injured tissue and promote regrowth;
5. to determine the changes in the spinal network interneurons' morphology, firing properties and synaptic connectivity after spinal cord injury and how to restore normal features by treatments suppressing the growth inhibitor Nogo-A and rehabilitative training;
6. to assess the plasticity in the activity of spinal interneurons in paraplegic patients and its optimisation by treatment and training to promote recovery of motor function in these patients;
7. to ensure cohesion, integration and collaboration within the the scientific programme.
To this end the network is supported by a management team, which deals with administration, management and exploitation issues.
Project results:
Work package (WP) 1
Central pattern generators (CPGs) in the spinal cord produce locomotor patterns that are expressed as a coordinated activity of motor neurons controlling muscle contractions. These intrinsic networks produce both the rhythm and the detail of the pattern.
A major goal of the WP1 was to identify excitatory CPG neurons and their role in the generation of locomotor activity. We have used two experimental models to obtain this goal: the neonatal mice (OK) and the young zebrafish (AEM). In mice, the team of OK has identified excitatory neural populations as essential for rhythm-generation and defined specific roles for these cells in coordinating the locomotor activity in the mammalian CPG. V2a neurons drive the locomotor central pattern generator (CPG) network that ensures left-right alternation and have some effect on rhythm-generation. Further studies have demonstrated that excitatory EphA4 neurons also contribute to excitation in the network. In zebrafish, the team of AEM has also examined the pattern of recruitment of V2a interneurons. These experiments have defined the principles of recruitment of neurons of the locomotor circuit and characterise V2a interneurons as the source of excitation necessary to generate the locomotor rhythm.
The natural variability in speed of locomotion requires an orderly recruitment of motor neurons from those innervating slow to those innervating fast muscle fibres. The team of AEM showed that motor neurons in zebrafish are organised in specific topographic locations and are incrementally recruited to produce swimming at different frequencies.
Another advance of WP1 was that the cellular properties of excitatory neurons have been characterised in both mice and zebrafish. In mice, although the cells display a number of rhythmogenic properties these properties are not strongly involved in generating the rhythm. In zebrafish, many neurons have pacemaker properties that may strongly boost the rhythm.
Together our studies have identified excitatory neural populations as essential for rhythm-generation and defined specific roles for these cells in coordinating the locomotor activity in the vertebrate CPG.
Technological achievements: Shox2-Cre and Sox14-EGFP mice have been exchanged from Jessell lab to Kiehn lab. BAC-Vglut2-Cre from Ole Kiehn to Tom Jessell lab. Furthermore, the BAC-Vglut2-Cre and BAC-Vglut2-ChR2 mice have been disseminated to several labs in US, Europe and Japan. Frozen embryos of the BAC-Vglut2-Cre mice have been deposited at the European Mouse Mutant Archive (EMMA).
WP2
During the first reporting period, they have been using genetically labelled synaptic terminals and specific antibodies to genetically-defined interneurons and been able to demonstrate that two principle categories of interneurons exist in the spinal cord. One category is located in domains of the spinal cord with sparse proprioceptive input. The second category of interneurons receives a much higher frequency of proprioceptive input mapped anatomically. We have thus been able to provide a map of neuronal populations with- and without direct proprioceptive input in the mouse spinal cord.
Following our work performed during the first reporting period, the teams have extended their analysis of spinal interneuron subtypes receiving proprioceptive afferent input. In particular, they have focused on V2a interneurons; an ipsilaterally projecting, excitatory interneuron class marked by expression of the transcription factors Sox14 and Shox2. To analyse synaptic input of proprioceptors to these interneurons, they have crossed Shox2Cre mice with floxed reporter alleles and performed high-resolution image analysis of vGlut1 input to these neurons. They found that V2a interneurons receive a high level of proprioceptive input, putting them in a perfect position to integrate and relay sensory information. For these reasons, analysis of the function of these neurons in motor control, as carried out by WP1 was a key task.
The team of Silvia Arber has further study the connectivity rules of V2a interneurons in the spinal cord. Using a recently developed transsynaptic virus method which allows selective visualisation of neurons with direct connections to motor neurons, they found that premotor neurons connecting to flexor motor neurons outnumbered the population connected to extensor motor neurons. Second, and as described in more detail in WP1, intracellular recordings from V2a interneurons also revealed a biased distribution towards flexor activity in V2a interneurons. Together, these findings support a model in which V2a interneurons are preferentially connected to flexor motor neurons when compared to extensor counterparts.
Finally, they have performed electrophysiological studies to genetically remove proprioceptors by binary mouse genetics, and characterised the consequences of genetic perturbation of proprioceptors. Neonatal isolated spinal cord preparations in vitro from mice expressing DTA in Parvalbumin-expressing proprioceptors were compared to wild-type preparations. The team of Silvia Arber has found that proprioceptor-ablated mice have severe defects in the stability of the rhythm, which can be induced by application of DA, 5HT and NMDA. Motor bursting pattern in proprioceptor-ablated mutants is highly irregular, and this phenotype is more severe at caudal than rostral lumbar levels.
Technological achievements:
Development of a transsynaptic virus method, which allows selective visualisation of neurons with direct connections to motor neurons (Stepien et al., 2010). We have injected the combination of these viruses into selective extensor and flexor muscle groups of the hindlimb and determined the position and percentage of premotor interneurons labelled with direct connections to corresponding motor neuron pools.
WP3
Major efforts have been placed during this second reporting period in WP3 to assess the intrinsic and extrinsic modulation of the spinal locomotor circuitry, frequency and plasticity. The team of Abdel El Manira has identified the identity of the endocannabinoid and the synaptic mechanisms of plasticity responsible for the plasticity of the locomotor network circuit. The signalling pathway activated by mGluR1 leads to formation of diacylglycerol (DAG), which is the precursor of the endocannabinoid 2-arachydonylglycerol (2-AG). The data suggest that 2-AG is the endocannabinoid within the spinal locomotor network that is synthesised on demand via mGluR1 activation to mediate long-term plasticity in locomotor circuitry.
In lamprey, the tachykinin substance P, like mGluR1 agonists, accelerates the burst rate and reduces the crossed inhibition in an activity-dependent fashion. Tachykinins increase the rate of locomotor activity, and substance P is known to be endogenously released during bouts of locomotor activity, thereby fine-tuning the neuronal activity within the locomotor network. The team of Sten Grilllner has shown that this enhanced neuronal activity, elicited by substance P, is partially mediated by endocannabinoids. Substance P has also additional cellular effects; it is known to enhance the synaptic currents via NMDA receptors and we show here that calcium currents in both motorneurons and inhibitory interneurons are depressed.
Knowledge of the mechanisms by which the neuronal activity in network interneurons can be modulated is of importance for the understanding of how one can optimise network function in the isolated spinal cord. This in turn will most likely be of importance for an understanding of how one can pharmacologically modulate spinal cord function after spinal cord injury. Certain patients with partial spinal cord injury can be trained to regain some walking function, and in these cases pharmacological fine-tuning of the spinal cord locomotor network can become of considerable importance (drugs presumably administered through intrathecal pumps).
Technological achievements:
Rodents are frequently used to model CNS damage and diseases that lead to functional deficits. Currently, available methods for the evaluation of impaired locomotor function such as scoring systems or biomechanical measures often suffer from serious limitations, e.g. due to subjectivity, non-linearity, low sensitivity or an exclusive focus on very specific aspects of a movement. The use of extensive, time- and space-consuming test batteries is recommended but not standard. Therefore, the development of a single, standardised set-up for quantitative and objective evaluation of locomotor functions was envisioned. A new experimental set up has been developed by the team of Martin Schwab to assess several locomotor tasks at once and in a standardised manner. Those tasks include tests of skilled locomotion (horizontal ladder) and basic locomotion with differing weight support (walking, wading, and swimming). The setup and knowhow has been protected as a registered trademark (MotoRater) and licensed out to TSE Systems GmbH, Germany.
WP4
Many of the molecules and receptors that guide axons during embryonic development are expressed in the adult CNS. It is probable that they have actions on synapses and therefore on plasticity. During this second period of the project, we have mainly examined the expression of chondroitin sulphate and heparan sulphate proteoglycans (CSPGs, HSPGs) after CNS damage, and during normal development. During the first part of the project the main effort was to describe the constituents and distribution of the PNNs. Examination of the timing of expression of the various constituent molecules showed that most components are produced from before birth, but diffusely distributed in the CNS.
The PNNs appear at the end of critical periods for plasticity, and because of their content of inhibitory CSPGs, their localisation around inhibitory interrneurons and synapses it was hyothesised that they are involved in the control of plasticity. The molecules that are upregulated in PNNs are link proteins and aggrecan. To further study the role of PNNs in plasticity, the team of James Fawcett has used cartilage link protein-1 (CRTL-1) null double mutants where link protein is knocked back in to cartilage, but not in the brain. One of the major advances of WP4 have shown that animals lacking link protein in the CNS have attenuate perineuronal nets and show continuing plasticity into adulthood. This also suggests that PNNs are the target of chondroitinase treatment.
Chondroitinase has multiple effects on the CNS. It digests CSPGs in the glial scar, allowing axon regeneration through it, it digests PNNs promoting plasticity, and it encourages sprouting of damaged and undamaged CNS axons. The team of James Fawcett has demonstrated that expression of axon growth-promoting integrins, or activation of integrins that have been de-activated by CSPGs can enhance axon regeneration both in vitro and in vivo.
Another major highlight of WP4 is the findings coming out of three major trials of behavioural recovery following chondroitinase treatment that have been completed within the course of the project. These experiments have demonstrated that
(i) axon regeneration, sprouting and recovery were all the same if the chondroitinase treatment alone started immediately or seven days after injury;
(ii) chondroitinase-induced plasticity opens a time window during which rehabilitation is dramatically more effective than normal; and
(iii) chondroitinase with rehabilitation is still as effective when treatment is started one month after injury, a timescale that is relevant to clinical use of the enzyme.
The treatment also increases the formation of synapses by axonal sprouts from the corticospinal tract.
The collaborative effort of the teams of James Fawcett and Martin Schwab to assess the combination of chondroitinase and anti NogoA treatments has demonstrated that the treatments must be given at different times, because combining both gives unexpected and poor behaviours. A second experiment demonstrated that both anti NogoA and chondroitinase individually promote functional recovery in paw reaching when combined with rehabilitation.
In order to determine the therapeutical time window for the chondroitinase treatment, we have study the length of time during which we could extract active enzyme from the injected tissue. The conclusion is that chondrotinase treatment may be delivered up to one month after injury, and that its action persists for at least two weeks. The team of James Fawcett has developed a method to slowly release chondroitinase in the tissue. This method releases chondroitinase for a minimum of three weeks.
After partial spinal cord lesions surviving fibers mediate only limited recovery. In this study, the teams of Martin Schwab and James Fawcett evaluated the physiological and anatomical status of spared fibers after unilateral hemisection of thoracic spinal cord in adult rats. In conclusion, this study demonstrated delayed decline of transmission through surviving axons to individual lumbar motoneurons during chronic stage of incomplete spinal cord injury in adult rats. These findings suggest a chronic pathological state in intact fibers and necessity for prompt treatment to minimise it.
Finally, the team of James Fawcett has discovered an unexpected effect of CPSGs and chondroitinase on transmission of action potentials. We showed that adjacent to spinal cord injuries the descending reticulospinal axons lose the capacity to transmit action potentials, although they are undamaged. In animals treated with chondroitinase axonal conduction is restored.
The mechanism of CSPG-related transmission block is unknown, but the CSPG responsible is probably NG2, since direct injection of this molecule into the cord causes a transmission block.
WP5
The aim of this work package is to investigate the rearrangement of connections and possible anatomical plasticity of spinal interneuronal circuits as well as the effect of Nogo-A antibody treatment and rehabilitative training on a behavioral and morphological level.
Rodent studies
The team of Martin Schwab performed experiments with adult rats that received either a cervical or thoracic hemisection - a paradigm mimicking the Brown-Séquard Syndrome in humans. During the first reporting period, they have shown that after such an injury, animals recover a simple locomotor function within three to four weeks whereas fine motor control remains very limited.
During the second reporting period, the team of Martin Schwab has performed detailed anatomical experiments to understand which plastic changes explain the disparity in the recovery process between the fore- and hindlimb function following such spinal cord injury (SCI). Retrograde and anterograde neuroanatomical tracing techniques focusing on different descending supraspinal systems were performed. Anterograde tracing confirmed an increase in midline crossing re-innervation of the ipsilesional lumbar enlargement by contralateral medullary reticular formation (MMRF) cells. Moreover, the mesencephalic locomotor region revealed significant plastic changes following SCI which could indicate important locomotor adaptations acting on the MMRF. In conclusion, the potential for anatomical plasticity was assessed for all spinally projecting brain centres relevant for motor function. The most important anatomical substrate for functional recovery of the hindlimbs after unilateral cervical SCI was identified as being a part of the MMRF.
Human studies
The team of Volker Dietz has studied the human condition corresponding to the rodent hemisection experimental model. In a recent study, they have thus compared the Brown-Séquard syndrome (BS), which is relatively rare compared to the central cord syndrome (CC). A time-course of neurological deficit, functional recovery and impulse conductivity was studied in BS and CC human subjects. Motor score, walking function, daily life activities and electrophysiological recordings were evaluated one and six months after SCI and were compared between age-matched groups of tetraparetic BS and CC subjects. For all analysed measures, no difference in the time-course of improvement was found.
They concluded that in contrast to the assumption of a better outcome of subjects with BS, no difference was found between the two incomplete SCI groups. This is of interest with respect to the different potential mechanisms leading to a recovery of functions in these two SCI subgroups with anatomically dissimilar lesions.
Nogo-A / Rehabilitation combined treatments
Proof of concept studies in spinal cord injured macaque monkeys with anti-Nogo-A antibodies have replicated the rodent findings performed by the team of Martin Schwab and others; recently, clinical trials in spinal cord injured patients have been started. However, the optimal time window for successful Nogo-A function blocking treatments was largely unclear. The team of Martin Schwab has thus studied the effect of acute as well as one or two weeks delayed intrathecal anti-Nogo-A antibody infusions on the regeneration of corticospinal tract (CST) axons and the recovery of motor function after large but anatomically incomplete thoracic spinal cord injuries in adult rats. The results have shown that the time frame for treatment of spinal cord lesions with anti-Nogo-A antibodies is restricted to less than two weeks in adult rodents.
After traumatic injury of the CNS both anti-Nogo-A antibody treatment as well as rehabilitative physical training have been shown to induce functional recovery and anatomical plasticity. The team of Martin Schwab have shown no additional beneficial effect of the combined anti-Nogo-A antibody treatment with rehabilitative treadmill training of the hindlimbs, but rather decreased recovery of function when compared to every treatment individually. Further, they have demonstrated that there is a specific time window and order for these two treatments to show beneficial effects: physical rehabilitation has to follow acute anti-Nogo-A antibody treatment and should not be concomitantly given.
WP6
In patients with spinal cord injury, neither spontaneous nor treatment induced plasticity of local interneuronal circuits have been studied thus far. Locomotor activity and spinal reflexes (SR) have been shown to have a close relationship in different mammals, including humans.
To study changes in the spinal circuitry, the team of Martin Schwab has used the rat sacral level S2 complete transection model, where animals develop severe spastic cramps upon peripheral stimulation between two and five weeks after the injury. The study reveals changes in the intraspinal network below the injury site beginning one week after injury. The cholinergic input to motoneurons is progressively decreased and almost completely abolished 13 weeks after injury. This loss is accompanied by the disappearance of cholinergic interneurons around the central canal.
Conclusion:
The results suggest the occurrence of major changes in the local wiring of the spinal cord below a complete lesion; loss of inhibitory control of Ia fibres is consistent with their hyperexcitability during spastic cramps, and loss of the important excitatory cholinergic drive of motoneurons may contribute to the known progressive weakness of motor output after complete lesions.
Early after a SCI, neither locomotor nor spinal reflex activity can be evoked in patients. Once the spinal shock has resolved, plasticity is apparent in the human spinal locomotor circuitry: locomotor activity and an early spinal reflex component reappear in response to appropriate peripheral input. The team of Volker Dietz has demonstrated that a neuronal dysfunction below the level of lesion, reflected in a shift from dominant early to dominant late spinal reflex components and an exhaustion of leg muscle activity during assisted walking, is fully established one year after injury. In chronic complete spinal cord injured (cSCI) subjects only a late SR component was consistently present during upright stance. However, during assisted locomotion, an early SR component appeared, while amplitude of the late SR component became small. In contrast, in healthy subjects the early SR component dominated in all conditions, but a small late component appeared during assisted locomotion.
Conclusion:
A more balanced activity of early and late SR components occurred in both subject groups if an appropriate proprioceptive input was provided. These results suggest that early and late SR components are assumed to reflect the activity of separate neuronal circuits, which are associated with the locomotor circuitry possibly by shaping the pattern.
The neuronal dysfunction also occurs in non-ambulatory patients with incomplete SCI. The behaviour of spinal reflexes (early versus late component) is related to the locomotor capacity of SCI subjects, as it has been shown in spinal rats. Polysynaptic spinal reflexes are proposed as markers for the functional state of spinal locomotor circuitries.
Severely affected SCI subjects, unable to walk, show dominant late SR components, whilst in ambulatory SCI subjects an early reflex component dominates.
Conclusion:
Neuronal plasticity exploited by a functional training is reflected in both an improvement of locomotor ability and a change in balance of SR components towards the early SR component.
Potential impact:
The proposed projects capitalise on combined expertise in different area of spinal cord research. The proposal has therefore involved collaborative interactions that allowed us to merge our unique and complementary expertise in spinal cord research. The expected final results have / will shed light on:
- the molecular and cellular identity of the rhythm generating core of spinal locomotor circuitry;
- the development and plasticity of spinal locomotor circuits in response to modulation of proprioceptive afferent input;
- the cellular and synaptic mechanisms responsible for short- and long-term modulation of locomotor circuitry;
- the functional interactions among extracellular matrix, plasticity and scar formation;
- the plasticity of the spinal circuitry after injury;
- the plasticity of the locomotor circuitry after human spinal cord injury;
- the most promising and effective rehabilitation procedures for SCI patients.
The gained understanding obtained within this project offers molecular and cellular foundations that will be instrumental to develop strategies for restoring motor function following spinal cord injury.
Impact on science
An understanding of these mechanisms is not only required for elucidating the normal operation of motor networks in the healthy spinal cord, but is a prerequisite for restoring motor function in the injured spinal cord. Our results provide a detailed characterisation at the molecular and cellular level of the key components of the spinal locomotor circuitry. We have made major progress towards the identification of the signalling molecules that specify the identity and axonal projection of these neurons and the proprioceptive sensory input they receive. Second, we have been able to start defining their cellular and synaptic properties that allow them to generate locomotor activity and the mechanisms of plasticity. The obtained results will constitute the foundation for analysing the aberrations induced by spinal cord injury with regard to the morphology, cellular and synaptic properties of locomotor network neurons in animal and human models. Altogether this collaborative research will strongly advance the field and pave the way to translate these findings into rehabilitation of spinal cord injured patients.
Impact on society - socio economic benefits
In Europe overall, neurological damage accounts for 40 % of people severely disabled and who require daily help (Wade and Hewer, 1997; Office of Population Censuses and Surveys, 1998). It is estimated that 90 million people around the world currently suffer from some form of spinal cord injury. In Europe, there are estimated to be at least 330 000 people living with spinal cord injury with over 15 000 new cases reported each year. In two-thirds of cases, road accidents are the cause of injury, with sporting accidents making up another 10 %. Most occur at a young age: average age of 19; about 80 % of males with spinal cord injuries are aged 18-25 years.
The European Assembly recently stated it 'believes that, as part of a comprehensive policy for people with disabilities, more intensive efforts must be made to achieve further progress in research designed to bring about a cure for spinal cord injury' and 'highlights the importance of promoting prevention and financial support for spinal cord research'. This is further exacerbated by the pathologies that manifest themselves after the injury has occurred which further impacts both the sufferer and their family. Secondary effects have been noted to include: blood clots, CVD, deep vein thrombosis, edema and hypertension, gastrointestinal complications, orthopaedic / neurological complications and respiratory problems.
The cost of treatment and aftercare for sufferers is phenomenal: the average lifetime costs directly attributable to spinal cord injury for an individual injured at age 25 range from EUR 0.45 million to EUR 2.1 million and have to prepare to spend an average of forty years or more in a wheelchair. Patients typically require continuous physical and medical care depending on the degree of disability.
The burden of care giving most frequently falls on the partner. While spinal cord injury represents a significant physical and psychological burden to the affected individual, it has also a substantial economic burden to society. In addition to the direct medical costs, including patient care associated with spinal cord injury, which national health services have to cover, there is also the loss in economic productivity of the sufferer, which adds on another 30 % of these costs. Therefore, any scientific effort that leads to improvements of the ability to stand and walk for shorter distances and/or obtain a better bladder and bowel function will mean large gains for the individual and for society in reduced health costs.
Added value of consortium
The participants of SPINAL CORD REPAIR have been selected to provide the necessary first-rate and complementary expertise required to address the key questions outlined in the different work packages. This consortium provides a unique association of competences able to provide detailed insight into spinal cord injury and generate potential strategies for cure based on a fundamental understanding of the development and function of the healthy spinal cord matched with innovations to generate combination therapies matched with rehabilitation for real therapeutic advancement. The work and collaborations that have been initiated as part of this project will continue to generate breakthroughs after the end of the grant.
An important strength of SPINAL CORD REPAIR, based on its international character, directly pertains to the fact that many preclinical therapies have not been shown to be safe and efficacious by more than a single laboratory. Through the presence in the consortium of diverse specialists in animal models, we have been able to perform the independent replication in the different laboratories which is now considered imperative to determine the general applicability of a therapy. It is now accepted that before moving to clinical trials, potential therapies should be tested in models that closely approximate the human injury subtype to be treated. To address this issue SPINAL CORD REPAIR has planned extensive work directly correlating work we have been doing in animal models with paraplegic patients. In this instance, we have been analysing locomotor function in patients and assessing if this correlates with rat models specifically addressing electrophysiological characteristics.
In light of our clinical translation plans and clinical reverse engineering, we are convinced that SPINAL CORD REPAIR will have a significant impact.
Main dissemination activities
The project has been using all possible routes of dissemination to spread out the major findings to the scientific community, the stakeholders and the general public.
The primary route of dissemination of SPINAL CORD REPAIR foreground has been via scientific meetings and congresses. The findings of the consortium have been widely presented by the partners and disseminated to prestigious congresses and meetings. The secondary route of dissemination was through international peer-reviewed publications.
Exploitation of the results
Trademark
The collaboration between the teams of Martin Schwab and Marc Bolliger (Volker Dietz) has permitted the technical development of a new apparatus, called 'MotoRater', for the assessment of motor rehabilitation in rodents following spinal cord injuries (in relation with WP3). The locomotor tasks in our system include tests of skilled locomotion (horizontal ladder) and basic locomotion with differing weight support (walking, wading, and swimming). This combination of locomotor tests allows a complete, standardised and sensitive quantification of rodents after different types of CNS damage (maybe even PNS damage). In collaboration with the Tech Transfer Office of the University of Zurich (Unitectra), the University of Zurich has submitted a request to register a trademark for the MotoRater. The registered trademark has been licensed our to TSE Systems GmbH, Bad-Homburg, Germany in 2010.
Reference paper: Zörner B, Filli L, Starkey ML, Gonzenbach R, Kasper H, Rothlisberger M, Bolliger M, and Schwab ME (2010). Profiling locomotor recovery: comprehensive quantification of impairments after CNS damage in rodents. Nat Methods 7(9): 701-708
Further information on this new behavioural apparatus is available under: http://www.tse-systems.com/products/behavior/motor-function/motorater.htm
IP protection
The IP of James Fawcett (UCAM) for his work on 'Chondroitinase and extracellular matrix modification promotes CNS plasticity' (in relation with WP4) has been protected in 2010 by the Tech Transfer Office of the University of Cambridge (Cambridge Entreprise).
The IP of James Fawcett (UCAM) for his work on 'Alpha9 integrin promotes regeneration of axons in the CNS' (in relation with WP4) has been submitted for IP protection in 2011 by the Tech Transfer Office of the University of Cambridge (Cambridge Entreprise).
Dissemination of know-how and IP protection
Furthermore, the BAC-Vglut2-Cre and BAC-Vglut2-ChR2 mice have been disseminated by the laboratory of Ole Kiehn to several labs in US, Europe and Japan. Frozen embryos of the BAC-Vglut2-Cre mice have been deposited at The European Mouse Mutant Archive - EMMA.
Project website address
The project website (see http://www.spinalcordrepair.eu online) was developed early 2008 and was online in February 2008. The goal of this public website was to communicate about the SPINAL CORD REPAIR research to the EC, patients associations and general public. The project management team has placed particular attention to creating an attractive content that would be understandable for a lay audience. In line with the general communication strategy developed by the management team for SPINAL CORD REPAIR, we have created a logo for the project (see below).
Educational information was inserted in 'The spinal cord' link with appropriate links for adults (wiki), children (a University of Washington education resource) and for professionals (e-medicine.com) explaining the scientific concepts of the project in a simplified way. 'Spinal cord injury' addressed the socio-economic impacts of this injury on patients, families and healthcare systems, while 'Other initiatives' addressed other European Commission funded projects working in the field. 'Global initiatives' provided links to European and International associations and funding bodies. All the research teams and contact details were also present on the web site together with the description of the different committees, under 'Scientific teams'. Finally, there were two links 'Contact information' and a 'intranet / webconference' for the members of the consortium. The website has been updated every six months. The website has been discontinued at the end of June 2011, once the project has ended.
In the present project, we have been studying the basic function and development of spinal networks generating locomotion and provide new knowledge about the repair of the spinal cord following injury with specific reference to motor functions. Our combined effort aimed at understanding the normal function and mechanisms underlying dysfunction of the spinal cord and upon such a base develop effective treatment strategies for spinal cord injury.
Neural circuits in the spinal cord, called central pattern generators (CPGs), can produce locomotor movements. The locomotor activity pattern is the result of the circuit design and the interplay between the firing properties of the CPG constituent neurons and their synaptic interactions. Spinal cord injury impairs the function of the locomotor circuitry and results in paralysis. An understanding of the molecular mechanisms of the assembly of the spinal locomotor circuitry, the function of its key excitatory components and the intrinsic plasticity in the healthy and injured spinal cord is a prerequisite for designing novel therapeutic methods to restore locomotor function after spinal cord lesion.
The work-packages (WPs) of the project have been designed to elucidate the key molecular pathways responsible for the development and assembly of the spinal circuitry for locomotion. The intrinsic function and modulation the spinal circuitry has been examined in the healthy spinal cord by combining sophisticated molecular, anatomical, pharmacological and electrophysiological tools. The mechanisms of plasticity and reorganisation of the circuitry have been examined in the injured spinal cord as have the mechanism to promote regeneration of the lesioned axons. This project has integrated knowledge on the development and normal spinal cord function together with biological interventions aiming at protecting and repairing the injured spinal cord.
To understand how the healthy spinal cord produces locomotor movements two main mechanisms have been studied; the first is the development and organisation of the spinal locomotor networks, the second is the modulatory mechanisms that regulate the locomotor activity. An important research task was therefore to elucidate how the spinal sensori-motor circuitry develops and define the key neuronal elements in the locomotor network that generate the motor rhythm.
In the injured spinal cord two main forms of biological intervention can promote recovery of function; the first is axon regeneration, the second is promotion and control of plasticity. Spinal cord injury invariably damages ascending and descending axons, and this is the main cause of disability. If axons can be induced to grow across the site of injury to make connections below and above it, useful function is returned. Describing the basic mechanisms of regeneration and how they relate to the functional circuits that are present in the spinal cord is therefore of great importance to advance functional recovery.
Project context and objectives:
The spinal cord contains neuronal networks responsible for the precise execution of different motor tasks ranging from locomotor movements, equilibrium control, hand - finger movements, and many autonomic functions. These networks are called into action by command signals from the brain. Spinal cord injury (SCI) disrupts the interaction between the brain and the networks in the spinal cord resulting in loss or impairments of movements, sensation and autonomic functions. To promote recovery of motor function after SCI, we have tried to understand:
(1) how the spinal sensori-motor network is assembled during development and identify the types of neurons responsible for the generation of locomotor activity in the healthy spinal cord,
(2) characterise at the cellular and synaptic level the mechanisms of plasticity that takes place within these networks after injury, and
(3) determine the mechanisms of regeneration to facilitate the reconnection between the brain and spinal cord through axonal regrowth.
Organisation of spinal locomotor networks
Neural circuits in the spinal cord, called central pattern generators (CPGs), can produce locomotor movements, such as walking. These networks interact continuously with centers in the brain and with sensory feedback to adapt the different phases of the step cycle to internal and external demands. The organisation of the spinal locomotor CPGs has been characterised in two lower vertebrate model systems, the lamprey and Xenopus tadpole. It consists of a network of excitatory glutamatergic and inhibitory glycinergic interneurons. The glutamatergic interneurons project ipsilaterally and provide the excitatory drive necessary to produce sustained rhythmic locomotor activity, while the glycinergic interneurons projecting to the contralateral side and mediate the reciprocal inhibition responsible for the alternation of activity between the two sides of the spinal cord.
In mammals, the motor pattern with flexion and extension coordination at different joints requires a more complex circuit organisation. Several types of interneurons have been identified based on the reflex pathways that activate them and their pattern of activation during locomotor rhythm. One key component of the locomotor network that was not yet defined was the neuronal assembly that is directly involved in generating the rhythm. Network pharmacology and lesion studies have shown that these rhythm-generating neurons are glutamatergic excitatory interneurons with ispilateral projections. Activation of these neurons from the brain provides the direct switch to turn on locomotion. Characterizing and defining the network organisation of these neurons is therefore of paramount importance for understanding the normal function of locomotor network and how these network are developmentally controlled. Moreover, these neurons are prime targets for intervention therapies aimed at improving locomotor function following spinal cord injury.
Molecular signalling defining the ipsilateral projection of excitatory interneurons
Axon guidance molecules EphA4 and ephrinB3 identify a group of excitatory spinal interneurons that are rhythmically active during locomotion. Mice with targeted deletions in the genes for ephA4 and ephrinB3 display a characteristic hopping rabbit-like gait. EphA4-positive neurons, whose cell bodies are located in the spinal cord, aberrantly cross the midline in ephA4 and ephrinB3 knockouts. These observations suggest that at least some EphA4-positive spinal neurons are excitatory CPG neurons, whose projections normally remain ipsilateral, but in mutants cross the midline and result in synchronous activity on the left and right side of the spinal cord. Recordings from EphA4 positive cells have now demonstrated that most EphA4 positive neurons are rhythmically active during locomotion. A subset of these neurons provides excitation to motoneurons in the same segment. However, it was not known if these excitatory interneurons constitute a homogenous cell population and how they are related to the interneurons characterised by the expression of specific transcription factors in the developing spinal cord.
Developmental characterisation of spinal interneurons
Spinal cord cell differentiation is orchestrated by a series of transcription factors expressed in temporal and spatial patterns. Four early metaclasses of ventral horn interneurons (V0, V1, V2, and V3) have been described based on transcription factor expression, which may relate to functional phenotypes in the adult spinal cord. V0 interneurons express Dbx1/Evx1 and are laminae VII-VIII excitatory and inhibitory commissural interneurons. V1 interneurons express En1 are all inhibitory and include Renshaw cells and Ia inhibitory interneurons. V2 neurons express Lhx3 early in development and then differentiate into V2a and V2b classes, expressing the transcription factors Chx10 and Gata2/3, respectively. V3 express Sim1 and are a mixed group of ipsi- and contralaterally excitatory projecting neurons. Experiments in zebrafish and mouse have shown that neurons expressing Chx10 and its zebrafish homolog Alx are all excitatory interneurons and project ipsilaterally, while Gata2/3 cells are mainly inhibitory ipsilaterally projecting neurons.
This molecular characterisation of early neuronal cell population has provided a new basis for genetically dissecting the neuronal circuits that control locomotion in the vertebrate spinal cord. Such methods are under development and include genetic silencing or activation of molecularly defined populations of interneurons. They have been also used to study the role of interneurons controlling left-right alternation and possibly the speed of locomotion. However, the individual contribution of the different types of ipsilaterally projecting excitatory interneurons in the generation of locomotion was still unclear before the project started.
Sensory contribution to spinal motor coordination
The intraspinal locomotor network receives a prominent feedback from the moving limb that helps regulate and adapt the locomotor movements to unexpected perturbations. These feedback systems include those that signal the progression of the step cycle (hip movements) and those that signal the load that the limb exerts on the ground. The efficacy of these feedback systems can be modulated during the locomotor cycle at both the presynaptic and the interneuronal levels, so that the feedback signals may be more efficacious in one part of the movement cycle than in another. These prominent feedback systems are critical for an optimal performance of the locomotor system and are present in all vertebrate systems investigated. In spinal animals they play a prominent role for a good coordination, but they need to be recalibrated through a continuous activation by locomotor movements, otherwise the motor pattern degenerates. Patients with partial spinal cord lesions can similarly profit from locomotor training on the treadmill, under conditions when the load and movement trajectory are carefully controlled.
Transcriptional programs for sensory connectivity
Transcriptional programs control the differentiation of sensory afferent projections at multiple steps. Recent evidence supports an important role for the Runt domain transcription factor Runx3 in early specification of proprioceptive afferents - that regulate muscle length (Ia-afferents) and force (Ib-afferents) and the regulation of the neurotrophic factor receptor TrkC. Moreover, graded expression of Runx3 within proprioceptive afferents controls laminar termination within the spinal cord, and thus indirectly determines which neuronal populations will be contacted by proprioceptive afferent subclasses. In addition to early transcriptional programs, proprioceptive afferents are also influenced by target-derived factors, which retrogradely induce transcriptional programs important for functionality of spinal networks. In particular, the neurotrophic factor NT-3 induces the expression of the transcription factor Er81 in proprioceptive afferents. Er81 expression within group Ia proprioceptive afferents in turn controls the elaboration of a ventrally projecting trajectory and thus the establishment of monosynaptic connections with motor neurons. Er81 mutation in mice has dramatic consequences for motor behaviour, resulting in highly uncoordinated movement due to the absence of group Ia proprioceptive afferent connectivity in the ventral spinal cord. However, it was still unclear before the start of the project how the activity of interneurons within the locomotor circuitry is affected by the lack of Ia afferent during development.
Modulation spinal networks in the healthy spinal cord
To account for how the spinal circuitry generates the appropriate locomotor pattern, it is not sufficient to characterise how the network develops and how its connectivity is assembled. The detailed properties of the synaptic transmission between the nerve cells and the membrane properties of the different types of nerve cells that form the network need also to be understood. The core part of the segmental network is composed by excitatory interneurons acting via ionotropic AMPA and NMDA receptors and glycinergic neurons via glycine receptors. These components are responsible for the cycle to cycle operation. There is, however, a number of modulatory systems that can fine tune the properties of the network, by modifying the properties by individual types of nerve cells or the synaptic properties. This modulation primarily involves activation of G-protein-coupled metabotropic receptors. The modulation can be a result of transmitters released from locomotor network neurons themselves (intrinsic modulation) such as glutamate activating via the family of metabotropic glutamate receptors (mGluRs) acting at both the pre- and postsynaptic levels. There are also transmitters released from neurons not being part of the network that activate for instance 5-HT1a, dopamine (D2), GABAB or tachykinin receptors (extrinsic modulation). These intrinsic and extrinsic modulatory systems play a fundamental role for the basic network function and for the induction of both short-term (seconds or minutes) or long-term (hours or days) plasticity of the activity of the locomotor circuitry. It has be important to further explore the modulator actions not only for an understanding of the neural mechanisms that turn on and optimise network operation, but also to find pharmacological tools for improving the function of the locomotor network in patients with an incomplete spinal cord lesion.
In networks of this type, many factors co-vary at both the single cell and the network level, which makes it difficult to explore different possible network solutions intuitively. One important tool consists of detailed biophysically realistic modelling that in combination with experimental analysis has help exploring the function of the spinal locomotor network. This approach has successfully been applied to the lamprey and tadpole CPG. Within the proposed consortium, the Grillner laboratory has made this particular methodology available for analysis of zebrafish and mouse models.
Repair of the injured spinal cord
An injury to the spinal cord disrupts the interaction between the brain centres and the spinal networks and results in loss of motor and sensory functions. Although the circuitry necessary to generate movement located in the spinal cord is still intact, it lacks the appropriate drive to trigger its activity. In human patients, the spinal cord is usually partially transsected, with damage to both grey and white matter. Much of the disability results from the interruption of ascending and descending white matter tracts. However spontaneous axon regeneration does not occur. For a long time there has been an extensive focus on the promotion of axon regeneration in the injured cord. Three main factors prevent axon regeneration in the spinal cord:
(1) inhibition of regeneration by the glial scar;
(2) inhibition of regeneration by molecules associated with oligodendrocytes and CNS myelin; and
(3) poor spontaneous regenerative response of CNS neurons.
The first treatments designed to overcome these forms of inhibition are now available. However, their effect is limited because they only induce axon regeneration over short distances and in a small number of fibres. The main inhibitory molecules in the glial scar are chondroitin sulphate proteoglycans (CSPGs), and the axon guidance molecules semaphorin3 and ephrin B2. Digestion of CSPGs with chondroitinase ABC and blocking of ephrin signalling have been shown to promote axon regeneration with recovery of function. Oligodendrocytes express the growth inhibitory membrane proteins Nogo-A, Myelin associated glycoprotein (MAG) and Oligodendrocyte and myelin glycoprotein (Omgp). Inhibition of Nogo-A with antibodies and by interfering with its signalling pathway has promoted axon regeneration and recovery. The regenerative response of CNS axons may also be enhanced through neurotrophins, raised cAMP, protein kinase C inhibition and other means, again with the effect of increasing axon regeneration.
Thus, different strategies have been developed to promote regeneration of the damaged axons that consist of interfering with growth inhibitory molecules and transplantation techniques to favour neurons survival and axon regeneration. However, at the start of the project, there was a lack of information about the optimal dosage of treatment to be use, the timing of administration, their combination with one another and with rehabilitative training.
Plasticity in the injured spinal cord
After human spinal cord injury there is a period lasting up to one year during which there is spontaneous improvement in neurological function. A similar though more rapid improvement occurs after rodent spinal cord injury, which can be enhanced by appropriate rehabilitative therapy. For instance, even in completely severed spinal cords, stepping or weight-bearing behaviour can be induced by treadmill training or other means. Much of this recovery is probably due to plasticity, and recent studies have demonstrated plastic changes anatomically. It is probable that enhancing plasticity after injury will promote better recovery in human patients because young children show remarkable recoveries after partial spinal cord injury provided that they are injured before the end of the critical period for spinal cord plasticity. It is also apparent that the promotion of plasticity is an easier and more achievable aim than long-distance axon regeneration. An additional reason to focus on plasticity treatments is that they will probably help patients with chronic injuries.
Plasticity in the spinal cord encompasses axonal sprouting, terminal sprouting and changes of synaptic strength. In most cases of return of function after injury through plasticity it is not known which of these processes have occurred. The main focus of investigations has been the corticospinal tract. Three of the treatments that promote functional recovery after SCI, anti NogoA, chondroitinase ABC and inosine promote collateral sprouting of the corticospinal tract above the injury together with some axonal regrowth through the injury. While these treatments enhance functional recovery, the changes in spinal cord circuitry that underlie these improvements are not worked out. Neither is the mechanism by which the treatments affect synaptic behaviour understood.
In animal models of cortical plasticity, if appropriate changes in connectivity and cortical mapping are to occur, plasticity must be driven by voluntary behaviour. It is therefore often assumed that there will be co-operative effects between plasticity-inducing treatments and rehabilitation therapy. This has yet to be proven after spinal cord injury.
The overall objective of the SPINAL CORD REPAIR project is to restore motor function after spinal cord injury. The SPINAL CORD REPAIR partners have been working towards this goal by defining the key elements of the vertebrate locomotor network and how these elements are remodelled following SCI, by developing new strategies to protect and activate the remaining circuitry, as well as by promoting regeneration of the damaged axons.
The specific objectives of the project were to obtain:
Basic knowledge on:
- the molecular and cellular identification of the interneurons responsible for rhythm generation to unravel the intrinsic function of the network underlying locomotion;
- the influence of proprioceptive sensory feedback on spinal locomotor circuit development and plasticity;
- the cellular and synaptic mechanisms responsible for the modulation of locomotor circuitry activity;
- membrane proteins and extracellular matrix-related factors that restrict plasticity in the adult spinal cord;
- the changes in morphology, intrinsic properties and synaptic connectivity of locomotor network interneurons after SCI in animal models and spinal cord injured patients.
Translational knowledge on:
- restoration of locomotion and remodelling of the locomotor circuitry;
- plasticity of the morphology and connectivity of the component neurons of the locomotor circuitry following injury;
- functional effects on motor function of promoting spinal cord plasticity with extracellular matrix-related factors;
- the influence treatment that promotes regeneration and rehabilitative training on locomotion and network interneurons morphology, intrinsic properties and synaptic connectivity.
Throughput knowledge on:
- novel pharmacological tools to activate the locomotor circuitry and restore function;
- novel candidate molecules for protection and regeneration of damaged axons;
- identification of the optimal timing and method of administration of chondroitinase following partial SCI;
- identification of optimal training and treatment paradigms to optimise functional recovery in spinal cord injured animals and patients.
This collaborative project aims at gaining fundamental understanding of the function and plasticity of the spinal networks responsible for motor behaviour and their dysfunction after injury. The project has been designed to bridge the basic mechanism of spinal cord function with translational approach to protect the injured tissue and promote re-growth and recovery of function in animal models as well as patients. To address the scientific objectives, the research has been divided into seven interrelated strategic lines of work, which aim:
1. to identify the interneuron classes responsible for the generation of the locomotor rhythmic pattern using a combination of molecular, electrophysiological and anatomical tools;
2. to examine the molecular mechanisms responsible for the development and assembly of the sensori-motor circuitry and the role of sensory feedback in shaping the identity of locomotor network interneurons;
3. to determine cellular and synaptic mechanisms by which modulatory systems operating within the locomotor network regulate the frequency of the locomotor rhythm at the short- and long-time scale;
4. to examine the role of extracellular matrix in mediating the plastic changes in synaptic connectivity after spinal cord injury and assess novel therapeutic strategies to protect the injured tissue and promote regrowth;
5. to determine the changes in the spinal network interneurons' morphology, firing properties and synaptic connectivity after spinal cord injury and how to restore normal features by treatments suppressing the growth inhibitor Nogo-A and rehabilitative training;
6. to assess the plasticity in the activity of spinal interneurons in paraplegic patients and its optimisation by treatment and training to promote recovery of motor function in these patients;
7. to ensure cohesion, integration and collaboration within the the scientific programme.
To this end the network is supported by a management team, which deals with administration, management and exploitation issues.
Project results:
Work package (WP) 1
Central pattern generators (CPGs) in the spinal cord produce locomotor patterns that are expressed as a coordinated activity of motor neurons controlling muscle contractions. These intrinsic networks produce both the rhythm and the detail of the pattern.
A major goal of the WP1 was to identify excitatory CPG neurons and their role in the generation of locomotor activity. We have used two experimental models to obtain this goal: the neonatal mice (OK) and the young zebrafish (AEM). In mice, the team of OK has identified excitatory neural populations as essential for rhythm-generation and defined specific roles for these cells in coordinating the locomotor activity in the mammalian CPG. V2a neurons drive the locomotor central pattern generator (CPG) network that ensures left-right alternation and have some effect on rhythm-generation. Further studies have demonstrated that excitatory EphA4 neurons also contribute to excitation in the network. In zebrafish, the team of AEM has also examined the pattern of recruitment of V2a interneurons. These experiments have defined the principles of recruitment of neurons of the locomotor circuit and characterise V2a interneurons as the source of excitation necessary to generate the locomotor rhythm.
The natural variability in speed of locomotion requires an orderly recruitment of motor neurons from those innervating slow to those innervating fast muscle fibres. The team of AEM showed that motor neurons in zebrafish are organised in specific topographic locations and are incrementally recruited to produce swimming at different frequencies.
Another advance of WP1 was that the cellular properties of excitatory neurons have been characterised in both mice and zebrafish. In mice, although the cells display a number of rhythmogenic properties these properties are not strongly involved in generating the rhythm. In zebrafish, many neurons have pacemaker properties that may strongly boost the rhythm.
Together our studies have identified excitatory neural populations as essential for rhythm-generation and defined specific roles for these cells in coordinating the locomotor activity in the vertebrate CPG.
Technological achievements: Shox2-Cre and Sox14-EGFP mice have been exchanged from Jessell lab to Kiehn lab. BAC-Vglut2-Cre from Ole Kiehn to Tom Jessell lab. Furthermore, the BAC-Vglut2-Cre and BAC-Vglut2-ChR2 mice have been disseminated to several labs in US, Europe and Japan. Frozen embryos of the BAC-Vglut2-Cre mice have been deposited at the European Mouse Mutant Archive (EMMA).
WP2
During the first reporting period, they have been using genetically labelled synaptic terminals and specific antibodies to genetically-defined interneurons and been able to demonstrate that two principle categories of interneurons exist in the spinal cord. One category is located in domains of the spinal cord with sparse proprioceptive input. The second category of interneurons receives a much higher frequency of proprioceptive input mapped anatomically. We have thus been able to provide a map of neuronal populations with- and without direct proprioceptive input in the mouse spinal cord.
Following our work performed during the first reporting period, the teams have extended their analysis of spinal interneuron subtypes receiving proprioceptive afferent input. In particular, they have focused on V2a interneurons; an ipsilaterally projecting, excitatory interneuron class marked by expression of the transcription factors Sox14 and Shox2. To analyse synaptic input of proprioceptors to these interneurons, they have crossed Shox2Cre mice with floxed reporter alleles and performed high-resolution image analysis of vGlut1 input to these neurons. They found that V2a interneurons receive a high level of proprioceptive input, putting them in a perfect position to integrate and relay sensory information. For these reasons, analysis of the function of these neurons in motor control, as carried out by WP1 was a key task.
The team of Silvia Arber has further study the connectivity rules of V2a interneurons in the spinal cord. Using a recently developed transsynaptic virus method which allows selective visualisation of neurons with direct connections to motor neurons, they found that premotor neurons connecting to flexor motor neurons outnumbered the population connected to extensor motor neurons. Second, and as described in more detail in WP1, intracellular recordings from V2a interneurons also revealed a biased distribution towards flexor activity in V2a interneurons. Together, these findings support a model in which V2a interneurons are preferentially connected to flexor motor neurons when compared to extensor counterparts.
Finally, they have performed electrophysiological studies to genetically remove proprioceptors by binary mouse genetics, and characterised the consequences of genetic perturbation of proprioceptors. Neonatal isolated spinal cord preparations in vitro from mice expressing DTA in Parvalbumin-expressing proprioceptors were compared to wild-type preparations. The team of Silvia Arber has found that proprioceptor-ablated mice have severe defects in the stability of the rhythm, which can be induced by application of DA, 5HT and NMDA. Motor bursting pattern in proprioceptor-ablated mutants is highly irregular, and this phenotype is more severe at caudal than rostral lumbar levels.
Technological achievements:
Development of a transsynaptic virus method, which allows selective visualisation of neurons with direct connections to motor neurons (Stepien et al., 2010). We have injected the combination of these viruses into selective extensor and flexor muscle groups of the hindlimb and determined the position and percentage of premotor interneurons labelled with direct connections to corresponding motor neuron pools.
WP3
Major efforts have been placed during this second reporting period in WP3 to assess the intrinsic and extrinsic modulation of the spinal locomotor circuitry, frequency and plasticity. The team of Abdel El Manira has identified the identity of the endocannabinoid and the synaptic mechanisms of plasticity responsible for the plasticity of the locomotor network circuit. The signalling pathway activated by mGluR1 leads to formation of diacylglycerol (DAG), which is the precursor of the endocannabinoid 2-arachydonylglycerol (2-AG). The data suggest that 2-AG is the endocannabinoid within the spinal locomotor network that is synthesised on demand via mGluR1 activation to mediate long-term plasticity in locomotor circuitry.
In lamprey, the tachykinin substance P, like mGluR1 agonists, accelerates the burst rate and reduces the crossed inhibition in an activity-dependent fashion. Tachykinins increase the rate of locomotor activity, and substance P is known to be endogenously released during bouts of locomotor activity, thereby fine-tuning the neuronal activity within the locomotor network. The team of Sten Grilllner has shown that this enhanced neuronal activity, elicited by substance P, is partially mediated by endocannabinoids. Substance P has also additional cellular effects; it is known to enhance the synaptic currents via NMDA receptors and we show here that calcium currents in both motorneurons and inhibitory interneurons are depressed.
Knowledge of the mechanisms by which the neuronal activity in network interneurons can be modulated is of importance for the understanding of how one can optimise network function in the isolated spinal cord. This in turn will most likely be of importance for an understanding of how one can pharmacologically modulate spinal cord function after spinal cord injury. Certain patients with partial spinal cord injury can be trained to regain some walking function, and in these cases pharmacological fine-tuning of the spinal cord locomotor network can become of considerable importance (drugs presumably administered through intrathecal pumps).
Technological achievements:
Rodents are frequently used to model CNS damage and diseases that lead to functional deficits. Currently, available methods for the evaluation of impaired locomotor function such as scoring systems or biomechanical measures often suffer from serious limitations, e.g. due to subjectivity, non-linearity, low sensitivity or an exclusive focus on very specific aspects of a movement. The use of extensive, time- and space-consuming test batteries is recommended but not standard. Therefore, the development of a single, standardised set-up for quantitative and objective evaluation of locomotor functions was envisioned. A new experimental set up has been developed by the team of Martin Schwab to assess several locomotor tasks at once and in a standardised manner. Those tasks include tests of skilled locomotion (horizontal ladder) and basic locomotion with differing weight support (walking, wading, and swimming). The setup and knowhow has been protected as a registered trademark (MotoRater) and licensed out to TSE Systems GmbH, Germany.
WP4
Many of the molecules and receptors that guide axons during embryonic development are expressed in the adult CNS. It is probable that they have actions on synapses and therefore on plasticity. During this second period of the project, we have mainly examined the expression of chondroitin sulphate and heparan sulphate proteoglycans (CSPGs, HSPGs) after CNS damage, and during normal development. During the first part of the project the main effort was to describe the constituents and distribution of the PNNs. Examination of the timing of expression of the various constituent molecules showed that most components are produced from before birth, but diffusely distributed in the CNS.
The PNNs appear at the end of critical periods for plasticity, and because of their content of inhibitory CSPGs, their localisation around inhibitory interrneurons and synapses it was hyothesised that they are involved in the control of plasticity. The molecules that are upregulated in PNNs are link proteins and aggrecan. To further study the role of PNNs in plasticity, the team of James Fawcett has used cartilage link protein-1 (CRTL-1) null double mutants where link protein is knocked back in to cartilage, but not in the brain. One of the major advances of WP4 have shown that animals lacking link protein in the CNS have attenuate perineuronal nets and show continuing plasticity into adulthood. This also suggests that PNNs are the target of chondroitinase treatment.
Chondroitinase has multiple effects on the CNS. It digests CSPGs in the glial scar, allowing axon regeneration through it, it digests PNNs promoting plasticity, and it encourages sprouting of damaged and undamaged CNS axons. The team of James Fawcett has demonstrated that expression of axon growth-promoting integrins, or activation of integrins that have been de-activated by CSPGs can enhance axon regeneration both in vitro and in vivo.
Another major highlight of WP4 is the findings coming out of three major trials of behavioural recovery following chondroitinase treatment that have been completed within the course of the project. These experiments have demonstrated that
(i) axon regeneration, sprouting and recovery were all the same if the chondroitinase treatment alone started immediately or seven days after injury;
(ii) chondroitinase-induced plasticity opens a time window during which rehabilitation is dramatically more effective than normal; and
(iii) chondroitinase with rehabilitation is still as effective when treatment is started one month after injury, a timescale that is relevant to clinical use of the enzyme.
The treatment also increases the formation of synapses by axonal sprouts from the corticospinal tract.
The collaborative effort of the teams of James Fawcett and Martin Schwab to assess the combination of chondroitinase and anti NogoA treatments has demonstrated that the treatments must be given at different times, because combining both gives unexpected and poor behaviours. A second experiment demonstrated that both anti NogoA and chondroitinase individually promote functional recovery in paw reaching when combined with rehabilitation.
In order to determine the therapeutical time window for the chondroitinase treatment, we have study the length of time during which we could extract active enzyme from the injected tissue. The conclusion is that chondrotinase treatment may be delivered up to one month after injury, and that its action persists for at least two weeks. The team of James Fawcett has developed a method to slowly release chondroitinase in the tissue. This method releases chondroitinase for a minimum of three weeks.
After partial spinal cord lesions surviving fibers mediate only limited recovery. In this study, the teams of Martin Schwab and James Fawcett evaluated the physiological and anatomical status of spared fibers after unilateral hemisection of thoracic spinal cord in adult rats. In conclusion, this study demonstrated delayed decline of transmission through surviving axons to individual lumbar motoneurons during chronic stage of incomplete spinal cord injury in adult rats. These findings suggest a chronic pathological state in intact fibers and necessity for prompt treatment to minimise it.
Finally, the team of James Fawcett has discovered an unexpected effect of CPSGs and chondroitinase on transmission of action potentials. We showed that adjacent to spinal cord injuries the descending reticulospinal axons lose the capacity to transmit action potentials, although they are undamaged. In animals treated with chondroitinase axonal conduction is restored.
The mechanism of CSPG-related transmission block is unknown, but the CSPG responsible is probably NG2, since direct injection of this molecule into the cord causes a transmission block.
WP5
The aim of this work package is to investigate the rearrangement of connections and possible anatomical plasticity of spinal interneuronal circuits as well as the effect of Nogo-A antibody treatment and rehabilitative training on a behavioral and morphological level.
Rodent studies
The team of Martin Schwab performed experiments with adult rats that received either a cervical or thoracic hemisection - a paradigm mimicking the Brown-Séquard Syndrome in humans. During the first reporting period, they have shown that after such an injury, animals recover a simple locomotor function within three to four weeks whereas fine motor control remains very limited.
During the second reporting period, the team of Martin Schwab has performed detailed anatomical experiments to understand which plastic changes explain the disparity in the recovery process between the fore- and hindlimb function following such spinal cord injury (SCI). Retrograde and anterograde neuroanatomical tracing techniques focusing on different descending supraspinal systems were performed. Anterograde tracing confirmed an increase in midline crossing re-innervation of the ipsilesional lumbar enlargement by contralateral medullary reticular formation (MMRF) cells. Moreover, the mesencephalic locomotor region revealed significant plastic changes following SCI which could indicate important locomotor adaptations acting on the MMRF. In conclusion, the potential for anatomical plasticity was assessed for all spinally projecting brain centres relevant for motor function. The most important anatomical substrate for functional recovery of the hindlimbs after unilateral cervical SCI was identified as being a part of the MMRF.
Human studies
The team of Volker Dietz has studied the human condition corresponding to the rodent hemisection experimental model. In a recent study, they have thus compared the Brown-Séquard syndrome (BS), which is relatively rare compared to the central cord syndrome (CC). A time-course of neurological deficit, functional recovery and impulse conductivity was studied in BS and CC human subjects. Motor score, walking function, daily life activities and electrophysiological recordings were evaluated one and six months after SCI and were compared between age-matched groups of tetraparetic BS and CC subjects. For all analysed measures, no difference in the time-course of improvement was found.
They concluded that in contrast to the assumption of a better outcome of subjects with BS, no difference was found between the two incomplete SCI groups. This is of interest with respect to the different potential mechanisms leading to a recovery of functions in these two SCI subgroups with anatomically dissimilar lesions.
Nogo-A / Rehabilitation combined treatments
Proof of concept studies in spinal cord injured macaque monkeys with anti-Nogo-A antibodies have replicated the rodent findings performed by the team of Martin Schwab and others; recently, clinical trials in spinal cord injured patients have been started. However, the optimal time window for successful Nogo-A function blocking treatments was largely unclear. The team of Martin Schwab has thus studied the effect of acute as well as one or two weeks delayed intrathecal anti-Nogo-A antibody infusions on the regeneration of corticospinal tract (CST) axons and the recovery of motor function after large but anatomically incomplete thoracic spinal cord injuries in adult rats. The results have shown that the time frame for treatment of spinal cord lesions with anti-Nogo-A antibodies is restricted to less than two weeks in adult rodents.
After traumatic injury of the CNS both anti-Nogo-A antibody treatment as well as rehabilitative physical training have been shown to induce functional recovery and anatomical plasticity. The team of Martin Schwab have shown no additional beneficial effect of the combined anti-Nogo-A antibody treatment with rehabilitative treadmill training of the hindlimbs, but rather decreased recovery of function when compared to every treatment individually. Further, they have demonstrated that there is a specific time window and order for these two treatments to show beneficial effects: physical rehabilitation has to follow acute anti-Nogo-A antibody treatment and should not be concomitantly given.
WP6
In patients with spinal cord injury, neither spontaneous nor treatment induced plasticity of local interneuronal circuits have been studied thus far. Locomotor activity and spinal reflexes (SR) have been shown to have a close relationship in different mammals, including humans.
To study changes in the spinal circuitry, the team of Martin Schwab has used the rat sacral level S2 complete transection model, where animals develop severe spastic cramps upon peripheral stimulation between two and five weeks after the injury. The study reveals changes in the intraspinal network below the injury site beginning one week after injury. The cholinergic input to motoneurons is progressively decreased and almost completely abolished 13 weeks after injury. This loss is accompanied by the disappearance of cholinergic interneurons around the central canal.
Conclusion:
The results suggest the occurrence of major changes in the local wiring of the spinal cord below a complete lesion; loss of inhibitory control of Ia fibres is consistent with their hyperexcitability during spastic cramps, and loss of the important excitatory cholinergic drive of motoneurons may contribute to the known progressive weakness of motor output after complete lesions.
Early after a SCI, neither locomotor nor spinal reflex activity can be evoked in patients. Once the spinal shock has resolved, plasticity is apparent in the human spinal locomotor circuitry: locomotor activity and an early spinal reflex component reappear in response to appropriate peripheral input. The team of Volker Dietz has demonstrated that a neuronal dysfunction below the level of lesion, reflected in a shift from dominant early to dominant late spinal reflex components and an exhaustion of leg muscle activity during assisted walking, is fully established one year after injury. In chronic complete spinal cord injured (cSCI) subjects only a late SR component was consistently present during upright stance. However, during assisted locomotion, an early SR component appeared, while amplitude of the late SR component became small. In contrast, in healthy subjects the early SR component dominated in all conditions, but a small late component appeared during assisted locomotion.
Conclusion:
A more balanced activity of early and late SR components occurred in both subject groups if an appropriate proprioceptive input was provided. These results suggest that early and late SR components are assumed to reflect the activity of separate neuronal circuits, which are associated with the locomotor circuitry possibly by shaping the pattern.
The neuronal dysfunction also occurs in non-ambulatory patients with incomplete SCI. The behaviour of spinal reflexes (early versus late component) is related to the locomotor capacity of SCI subjects, as it has been shown in spinal rats. Polysynaptic spinal reflexes are proposed as markers for the functional state of spinal locomotor circuitries.
Severely affected SCI subjects, unable to walk, show dominant late SR components, whilst in ambulatory SCI subjects an early reflex component dominates.
Conclusion:
Neuronal plasticity exploited by a functional training is reflected in both an improvement of locomotor ability and a change in balance of SR components towards the early SR component.
Potential impact:
The proposed projects capitalise on combined expertise in different area of spinal cord research. The proposal has therefore involved collaborative interactions that allowed us to merge our unique and complementary expertise in spinal cord research. The expected final results have / will shed light on:
- the molecular and cellular identity of the rhythm generating core of spinal locomotor circuitry;
- the development and plasticity of spinal locomotor circuits in response to modulation of proprioceptive afferent input;
- the cellular and synaptic mechanisms responsible for short- and long-term modulation of locomotor circuitry;
- the functional interactions among extracellular matrix, plasticity and scar formation;
- the plasticity of the spinal circuitry after injury;
- the plasticity of the locomotor circuitry after human spinal cord injury;
- the most promising and effective rehabilitation procedures for SCI patients.
The gained understanding obtained within this project offers molecular and cellular foundations that will be instrumental to develop strategies for restoring motor function following spinal cord injury.
Impact on science
An understanding of these mechanisms is not only required for elucidating the normal operation of motor networks in the healthy spinal cord, but is a prerequisite for restoring motor function in the injured spinal cord. Our results provide a detailed characterisation at the molecular and cellular level of the key components of the spinal locomotor circuitry. We have made major progress towards the identification of the signalling molecules that specify the identity and axonal projection of these neurons and the proprioceptive sensory input they receive. Second, we have been able to start defining their cellular and synaptic properties that allow them to generate locomotor activity and the mechanisms of plasticity. The obtained results will constitute the foundation for analysing the aberrations induced by spinal cord injury with regard to the morphology, cellular and synaptic properties of locomotor network neurons in animal and human models. Altogether this collaborative research will strongly advance the field and pave the way to translate these findings into rehabilitation of spinal cord injured patients.
Impact on society - socio economic benefits
In Europe overall, neurological damage accounts for 40 % of people severely disabled and who require daily help (Wade and Hewer, 1997; Office of Population Censuses and Surveys, 1998). It is estimated that 90 million people around the world currently suffer from some form of spinal cord injury. In Europe, there are estimated to be at least 330 000 people living with spinal cord injury with over 15 000 new cases reported each year. In two-thirds of cases, road accidents are the cause of injury, with sporting accidents making up another 10 %. Most occur at a young age: average age of 19; about 80 % of males with spinal cord injuries are aged 18-25 years.
The European Assembly recently stated it 'believes that, as part of a comprehensive policy for people with disabilities, more intensive efforts must be made to achieve further progress in research designed to bring about a cure for spinal cord injury' and 'highlights the importance of promoting prevention and financial support for spinal cord research'. This is further exacerbated by the pathologies that manifest themselves after the injury has occurred which further impacts both the sufferer and their family. Secondary effects have been noted to include: blood clots, CVD, deep vein thrombosis, edema and hypertension, gastrointestinal complications, orthopaedic / neurological complications and respiratory problems.
The cost of treatment and aftercare for sufferers is phenomenal: the average lifetime costs directly attributable to spinal cord injury for an individual injured at age 25 range from EUR 0.45 million to EUR 2.1 million and have to prepare to spend an average of forty years or more in a wheelchair. Patients typically require continuous physical and medical care depending on the degree of disability.
The burden of care giving most frequently falls on the partner. While spinal cord injury represents a significant physical and psychological burden to the affected individual, it has also a substantial economic burden to society. In addition to the direct medical costs, including patient care associated with spinal cord injury, which national health services have to cover, there is also the loss in economic productivity of the sufferer, which adds on another 30 % of these costs. Therefore, any scientific effort that leads to improvements of the ability to stand and walk for shorter distances and/or obtain a better bladder and bowel function will mean large gains for the individual and for society in reduced health costs.
Added value of consortium
The participants of SPINAL CORD REPAIR have been selected to provide the necessary first-rate and complementary expertise required to address the key questions outlined in the different work packages. This consortium provides a unique association of competences able to provide detailed insight into spinal cord injury and generate potential strategies for cure based on a fundamental understanding of the development and function of the healthy spinal cord matched with innovations to generate combination therapies matched with rehabilitation for real therapeutic advancement. The work and collaborations that have been initiated as part of this project will continue to generate breakthroughs after the end of the grant.
An important strength of SPINAL CORD REPAIR, based on its international character, directly pertains to the fact that many preclinical therapies have not been shown to be safe and efficacious by more than a single laboratory. Through the presence in the consortium of diverse specialists in animal models, we have been able to perform the independent replication in the different laboratories which is now considered imperative to determine the general applicability of a therapy. It is now accepted that before moving to clinical trials, potential therapies should be tested in models that closely approximate the human injury subtype to be treated. To address this issue SPINAL CORD REPAIR has planned extensive work directly correlating work we have been doing in animal models with paraplegic patients. In this instance, we have been analysing locomotor function in patients and assessing if this correlates with rat models specifically addressing electrophysiological characteristics.
In light of our clinical translation plans and clinical reverse engineering, we are convinced that SPINAL CORD REPAIR will have a significant impact.
Main dissemination activities
The project has been using all possible routes of dissemination to spread out the major findings to the scientific community, the stakeholders and the general public.
The primary route of dissemination of SPINAL CORD REPAIR foreground has been via scientific meetings and congresses. The findings of the consortium have been widely presented by the partners and disseminated to prestigious congresses and meetings. The secondary route of dissemination was through international peer-reviewed publications.
Exploitation of the results
Trademark
The collaboration between the teams of Martin Schwab and Marc Bolliger (Volker Dietz) has permitted the technical development of a new apparatus, called 'MotoRater', for the assessment of motor rehabilitation in rodents following spinal cord injuries (in relation with WP3). The locomotor tasks in our system include tests of skilled locomotion (horizontal ladder) and basic locomotion with differing weight support (walking, wading, and swimming). This combination of locomotor tests allows a complete, standardised and sensitive quantification of rodents after different types of CNS damage (maybe even PNS damage). In collaboration with the Tech Transfer Office of the University of Zurich (Unitectra), the University of Zurich has submitted a request to register a trademark for the MotoRater. The registered trademark has been licensed our to TSE Systems GmbH, Bad-Homburg, Germany in 2010.
Reference paper: Zörner B, Filli L, Starkey ML, Gonzenbach R, Kasper H, Rothlisberger M, Bolliger M, and Schwab ME (2010). Profiling locomotor recovery: comprehensive quantification of impairments after CNS damage in rodents. Nat Methods 7(9): 701-708
Further information on this new behavioural apparatus is available under: http://www.tse-systems.com/products/behavior/motor-function/motorater.htm
IP protection
The IP of James Fawcett (UCAM) for his work on 'Chondroitinase and extracellular matrix modification promotes CNS plasticity' (in relation with WP4) has been protected in 2010 by the Tech Transfer Office of the University of Cambridge (Cambridge Entreprise).
The IP of James Fawcett (UCAM) for his work on 'Alpha9 integrin promotes regeneration of axons in the CNS' (in relation with WP4) has been submitted for IP protection in 2011 by the Tech Transfer Office of the University of Cambridge (Cambridge Entreprise).
Dissemination of know-how and IP protection
Furthermore, the BAC-Vglut2-Cre and BAC-Vglut2-ChR2 mice have been disseminated by the laboratory of Ole Kiehn to several labs in US, Europe and Japan. Frozen embryos of the BAC-Vglut2-Cre mice have been deposited at The European Mouse Mutant Archive - EMMA.
Project website address
The project website (see http://www.spinalcordrepair.eu online) was developed early 2008 and was online in February 2008. The goal of this public website was to communicate about the SPINAL CORD REPAIR research to the EC, patients associations and general public. The project management team has placed particular attention to creating an attractive content that would be understandable for a lay audience. In line with the general communication strategy developed by the management team for SPINAL CORD REPAIR, we have created a logo for the project (see below).
Educational information was inserted in 'The spinal cord' link with appropriate links for adults (wiki), children (a University of Washington education resource) and for professionals (e-medicine.com) explaining the scientific concepts of the project in a simplified way. 'Spinal cord injury' addressed the socio-economic impacts of this injury on patients, families and healthcare systems, while 'Other initiatives' addressed other European Commission funded projects working in the field. 'Global initiatives' provided links to European and International associations and funding bodies. All the research teams and contact details were also present on the web site together with the description of the different committees, under 'Scientific teams'. Finally, there were two links 'Contact information' and a 'intranet / webconference' for the members of the consortium. The website has been updated every six months. The website has been discontinued at the end of June 2011, once the project has ended.