Effect of endurance training on release of growth factors as well as their receptors
Neurotrophic factors are present in the nervous system throughout the whole of life. They play their most important role during early development, supporting the growth of nerve fibers and dendrites and the formation of synapses. In the adult body the concentrations of growth factors are decreased, but their role is crucial in the regeneration process after traumatic injuries. Previously we detected impacts of endurance training on some motor-related CNS mechanisms, e.g. on the serotoninergic system [21], activation of the NO/sGC/cGMP pathway in locomotion-relevant brain areas (striatum, midbrain and cerebellum), and lasting increase in BDNF and TrkB expression in the striatum, cerebellum and hippocampus [22].
Multiple experimental and clinical studies have confirmed that physical activity increases growth factor release, and it can improve spinal cord plasticity and spasticity and strengthen atrophied muscles [23–26]. However, the mechanism of such improvements is still not fully understood. For instance, Gomez-Pinilla et al. [27] studied the effect of 21-day pre-training on BDNF expression 48 hours after Th8-Th9 spinal cord transection. They found that complete spinal cord transection significantly reduced BDNF levels in the lumbar enlargement, but 21 days of pre- training compensated for the reduction in BDNF expression.
Intracellular signaling stimulated by neurotrophins is crucial for neuronal survival, morphogenesis and plasticity. Sasi et al. [28] reported that the BDNF-TrkB complex activated various intracellular signaling pathways, including MAPK/ERK, PLCγ and PI3K. Moreover, simultaneous triggering of PI3K and MAPK pathways could alter actin and microtubule dynamics and dendritic branching [7, 28]. In order to evaluate the influence of pre-training on the activation of genes which increase the regenerative capacity of the spinal cord, we examined the effects of endogenous stimulation of growth factors on the specific signaling pathways. The major finding of the present study is that six weeks of pre-training generated growth factors which in turn regulated the activity of the major signaling pathways (PLC-IP3-CAMK; PLC-PKC; PI3k/Akt, Ras/Erk1/2; Rac1/Cdc42) at the lesion site and in the segments immediately adjacent to it, and promoted motor recovery in paralyzed hindlimbs.
BDNF-dependent neuroplasticity is triggered by PLCγ-PKC signaling pathway
Our results indicate that six weeks of endurance training had a notable impact on activation of PLCγ pathways. The evidence that increased BDNF level could stimulate post-SCI neuroplasticity was found predominantly at the lesion site. Significant upregulation of gene expression (PLCγ, CAMK II and PKC mRNA) and PLCγ and PKC protein levels was observed at the lesion site in the pre-trained SCI group compared to SCI alone. These results indicate that endurance pre-training could contribute to neuroplasticity in the injured spinal cord via BDNF-dependent PLCγ - PKC signaling. Tashiro et al. [29] investigated molecular mechanisms by which two weeks of rehabilitation ameliorated allodynia and spasticity after spinal trauma. They found that short-term treadmill training after SCI upregulated the expression of BDNF and post-translational modification of potassium-chloride cotransporter-2 (KCC2) in the lumbar enlargement of the spinal cord. Application of TrkB-IgG revoked the activity-induced upregulation of KCC2 as well as the beneficial effects on allodynia and spasticity. The authors showed that PLCγ expression was downregulated in the subacute phase after SCI, which could have contributed to the change in function of BDNF. On the other hand, TrkB/PLCγ-mediated intracellular signaling could be crucial for sensory neuron plasticity [30]. Our results show that increased BDNF-TrkB levels, produced by prolonged physical training before SCI, could play a crucial role in PLCγ activation. Macias et al. [31] found that the moderate increase in BDNF shown to occur after long-term locomotor training (four weeks) did not significantly change overall TrkB mRNA expression, but was accompanied by an increase in the number of TrkB-expressing cells, including astro- and oligo-dendrocytes in the lumbar intumescence [30]. However, it is currently unclear which cell populations may be targeted through activation of the PLCγ signaling pathway in the spinal cord after pre-training and SCI.
Microglia respond to SCI through becoming activated and then developing into M1 (pro-inflammatory) or M2 (anti-inflammatory) phenotypes [32]. Zhang et al. [33] demonstrated that inhibition of aldose reductase, which plays a key role in a number of inflammatory diseases, significantly attenuated LPS-induced activation of protein kinase C (PKC) and phospholipase C (PLC), and that such inhibition can work as a switch which regulates microglia by polarization either to M1 or M2 activity after spinal trauma. They suggested that inhibition of aldose reductase regulates activity of the PLC-PKC pathway, and may be promising for future SCI therapies. Mohanraj et al. [34] examined the effect of Trehalose-6,6′-dibehenate (TDB) on LPS-induced neuroinflammation. They found that TDB could inhibit LPS-induced inflammatory response through the PLC-γ1/PKC/ERK signaling pathway and promote microglial polarization towards the beneficial M2 phenotype via the PLC-γ1/calcium/CaMKKβ/AMPK pathway. The peak proliferation of microglia in rats occurs at 48h post-SCI [35–36], but this strongly depends on the type and extent of injury. Within the CNS parenchyma, microglia actively interact with two main cell types, astrocytes and neurons, producing toxic substances (cytokine, nitric oxide, super oxide, TNF-α) which significantly accelerate neuronal damage after spinal cord injury, and pro-survival factors which affect polarization of microglia into their phenotypes [32]. Recently-published data show that microglia which exerted their beneficial effects during the first week post-SCI in a mouse model were necessary for the survival of neurons and oligodendrocytes following an insult [37]. We previously demonstrated that acute atorvastatin treatment effectively prevented the excessive infiltration of destructive M1 macrophages cranially, at the lesion site and caudally (by 66%, 62% and 52% respectively) one day post-injury, whereas the infiltration of beneficial M2 macrophages was less affected (by 27%, 41% and 16%) near the injured site [15]. To explain the mechanisms involved in regeneration, time-dependent regulation of PLCγ-CAMKII and/or PLCγ-PKC signaling in the spinal cord after training and SCI should be further investigated.
Although pre-training is rarely used in experimental medicine, we and others suggest that it can play a significant role in the treatment of CNS injuries [38]. It seems that prolonged endurance training prior to SCI creates a spinal cord milieu which better withstands the consequences of traumatic spinal cord injury.
Neurotrophic factors influence cell survival and regeneration through PI3k/Akt and ERK1/2 signaling
The PI3k/Akt pathway, activated by both BDNF and GDNF, is an important regulator of neuronal survival in the central and peripheral nervous systems. It is a major pathway blocking caspase-3 activation in neuronal cells [39]. By binding to their cognate receptors, neurotrophins elicit the recruitment of PI3k to the plasma membrane, which leads to the activation of several serine/threonine kinases, including Akt. At the plasma membrane, Akt activation depends on phosphorylation, which is in part achieved by PDK1 [40]. In our study, six weeks of endurance training upregulated the gene expression of PI3k, PDK1 and Akt in the low thoracic spinal cord. Evidence that pre-training could promote cell survival after SCI was clearly found at the lesion site and/or in its vicinity in our pre-trained SCI group, where the gene expression of PI3k, PDK1 and Akt was significantly elevated. Western blot analysis showed a similar course in the protein levels of PI3k and Akt. These results suggest that the PI3K/Akt signaling pathway is affected by pre-training and by the binding of growth factors to their receptors. Such changes could help to promote the survival of neural and glial cells in the inhospitable environment of the lesion site [41] and could help to influence intersegmental connections. In their experiments Zhong et al. [42] used bone marrow mesenchymal stem cells (BMSC) to treat neuropathic pain because of their ability to modulate inflammatory response. They demonstrated that BMSC could suppress neuroinflammation by transforming M1 microglial phenotype (destructive) into M2 phenotype (regenerative), and based on that BMSC could reduce pain possibly by suppressing the NF-κB pathway, while promoting PI3K/Akt signaling activation through producing GDNF. These findings support the potential therapeutic application of GDNF for neuropatic pain. An other study by Li et al. [43] presented that treatment with insulin-like growth factor-1 (IGF-1) in mouse microvascular endothelial cells (MVECs), which mitigated the apoptosis and cell damage due to LPS insult in an in vivo mouse spinal cord study. IGF-1 increased the activity of the PI3K/AKT pathway and significantly corrected the microenvironment of neural tissue repair, reducing the injured core area and improving functional recovery.
Activation of ERK1/2 signaling is generally associated with both cell survival and proliferation. However, depending on the cell types and stimuli involved, its activation could also have a death-promoting effect. Earlier studies demonstrated that the balance between intensity and duration of pro-apoptotic vs anti-apoptotic signals transmitted by ERK1/2 determines whether a cell undergoes cell death or survives [44]. We determined the expression and protein quantity of signaling molecules involved in the RAS/ERK pathway at the lesion site and in the surrounding spinal cord tissue. Molecular analysis showed that Ras and Raf expression was elevated at the epicentre of injury and in the surrounding area six weeks after SCI, while the levels of MEK, ERK1 and ERK2 were comparable or lower than the control values. In pre-trained SCI animals however, the RAS, RAF, MEK, ERK1 and ERK2 gene expressions were significantly higher than after SCI alone. Moreover, ERK2 expression was lower at the epicentre of injury and in the nearest caudal segments in pre-trained SCI animals compared to intact controls. Low levels of ERK1/2 proteins were also detected across the whole studied area in the SCI and pre-trained SCI groups. These results correlate with a significant decrease in BDNF mRNA and the protein level at the epicentre of injury after SCI. We suggest that neurotrophin expression in pre-trained animals could regulate the Ras/Raf/ERK1/2 signaling pathway, but its role in neuroregeneration is not entirely clear. Are these changes helpful or harmful? Previous studies have shown that upregulation of Ras/Raf/ERK1/2 signaling in the spinal cord impairs neural cell migration, neurogenesis and synapse formation [45–47]. On the other hand, Kim et al. [25] confirmed that BMSC transplantation in combination with treadmill exercise potently reduced Bax expression, increased Bcl-2 expression and effectively enhanced BDNF-TrkB expression in injured spinal tissue. The combination of training with BMSC transplantation facilitated ERK1/2 and c-Jun expression, and these findings demonstrate the neuroregenerative effect through activation of the ERK1/2 signaling pathway. Several studies have shown that BDNF and ERK1/2 activation could correlate with the development of pain after SCI, although ERK expression at the lesion epicenter is not merely due to the SCI per se [6, 48]. We measured the pain sensation using hot-plate testing in pre-trained SCI and non-trained SCI animals, and no significant differences were found between the two experimental groups. It seems that pre-training leading to upregulation of neurotrophins has an impact on ERK1/2 signaling and could play a role in neuroregeneration. Although changes in this particular pathway may be helpful for neural tissue regeneration, an in-depth analysis should be performed to specify whether signals transmitted by ERK1/2 are anti- or pro-apoptotic.
Locomotor function
BDNF- and GDNF- related activation of signaling pathways at and around the lesion site probably mediates the favorable motor effects of endurance training. Six weeks of pre-training contributed to the health benefit associated with better post-SCI functional recovery. BBB locomotor scores indicated the neuroprotective effect of physical activity as early as 14 days after SCI and also at the end of the survival period (42 days). Matsuda et al. [49] pointed out that low-energy extracorporeal shock-wave therapy (ESWT) promoted BDNF expression and improved the locomotor functions evaluated by both BBB scale and ladder rung walking test, in addition to the sensory function measured with a von Frey test. Furthermore, exercise increased BDNF levels not only in the brain and plasma [22], but in skeletal muscle as well [50]. The effects of various stimuli on BDNF and GDNF expression have been reported in several other studies. Côté et al. [51] compared the task-dependent effects of bike- and step-training after complete spinal transection in relation to changes in the levels of neurotrophic factors. In addition to measuring the BDNF, NT-3, NT-4 and GDNF levels, they recorded the H-reflexes from interosseus foot muscles following tibial nerve stimulation (0.3, 5, or 10 Hz). Their results showed that bike- and step-training significantly increased the levels of BDNF, Neurotrophin 3 and 4 in the lumbar enlargement of the spinal cord, whereas step-training alone increased the GDNF levels. Elevated growth factor protein levels positively correlated with the recovery of H-reflex, which suggests that neurotrophic factors play a role in reflex normalization after SCI.