J Czarkowska-Bauch

Long-Term Locomotor Training Up-Regulates TrkBFL Receptor-like Proteins, Brain-Derived Neurotrophic Factor, and Neurotrophin 4 with Different Topographies of Expression in Oligodendroglia and Neurons in the Spinal Cord

Neurotrophins are potent regulators of neuronal survival, maintenance, and synaptic strength. In particular, brain-derived neurotrophic factor (BDNF), acting through full-length TrkB receptor (TrkB(FL)), is implicated in the stimulation of neurotransmission. Physical activity has been reported to increase BDNF expression in the brain and spinal cord. In this study we have evaluated the hypothesis that activation of a spinal neuronal network, due to exercise, affects the entire spinal neurotrophin system acting via TrkB receptors by modulation of BDNF, neurotrophin 4 (NT-4), and their TrkB receptor proteins. We investigated the effect of treadmill walking (4 weeks, 1 km daily) on distribution patterns and response intensity of these proteins in the lumbar spinal cord of adult rats. Training enhanced immunoreactivity (IR) of both neurotrophins. BDNF IR increased in cell processes of spinal gray matter, mainly in dendrites. NT-4 IR was augmented in the white matter fibers, which were, in part, of astrocytic identity. Training strongly increased both staining intensity and number of TrkB(FL)-like IR small cells of the spinal gray matter. The majority of these small cells were oligodendrocytes, representing both their precursor and their mature forms. In contrast, training did not exert an effect on expression of the truncated form of TrkB receptor in the spinal cord. These results show that both neuronal and nonneuronal cells may be actively recruited to BDNF/NT-4/TrkB(FL) neurotrophin signaling which can be up-regulated by training. Oligodendrocytes of the spinal gray matter were particularly responsive to exercise, pointing to their involvement in activity-driven cross talk between neurons and glia.

Locomotor exercise alters expression of pro‐brain‐derived neurotrophic factor, brain‐derived neurotrophic factor and its receptor TrkB in the spinal cord of adult rats

Previous evidence indicates that locomotor exercise is a powerful means of increasing brain‐derived neurotrophic factor (BDNF) and its signal transduction receptor TrkB mRNA levels, immunolabeling intensity and number of BDNF‐ and TrkB‐immunopositive cells in the spinal cord of adult rats but the contribution of specific cell types to changes resulting from long‐term activity is unknown. As changes in BDNF protein distribution due to systemic stimuli may reflect either its in‐situ synthesis or its translocation from other sources, we investigated where BDNF and TrkB mRNA are expressed in the spinal lumbar segments. We report on the cell types defined by size, BDNF mRNA levels and number of cells with TrkB transcripts in sedentary and exercised animals following 28u2003days of treadmill walking. In the majority of cells, exercise increased perikaryonal levels of BDNF mRNA but did not affect TrkB transcript levels. Bidirectional changes in a number of TrkB mRNA‐expressing cells occurred in small groups of ventral horn neurons. An increase in BDNF transcripts was translated into changes in pro‐BDNF and BDNF levels. A 7‐day walking regimen increased BDNF protein levels similarly to 28‐day treadmill walking. Our observations indicate that long‐ and short‐term locomotor activity of moderate intensity produce stimuli sufficient to recruit a majority of spinal cells to increased BDNF synthesis, suggesting that continuous tuning of pro‐BDNF and BDNF levels permits spinal networks to undergo trophic modulation not requiring changes in TrkB mRNA supply.

Glial inhibitors influence the mRNA and protein levels of mGlu2/3, 5 and 7 receptors and potentiate the analgesic effects of their ligands in a mouse model of neuropathic pain

ABSTRACT Metabotropic glutamate (mGlu) receptors, which are present on neurons and glial cells, have been shown to play a role in neuropathic pain. The present study sought to investigate how the glial inhibitors minocycline and pentoxifylline alter the effect that chronic constriction injury (CCI) has on the expression of mGlu receptors and on their associated ligands. RT‐PCR analysis revealed that seven days after CCI, the mRNA levels of glial markers C1q and GFAP, as well as those of mGlu5 and mGlu3, but not mGlu7, were elevated in the lumbar spinal cord – ipsilateral to the injury. The protein levels of the microglial marker OX42, the astroglial marker GFAP, and mGlu5 receptor protein were increased, whereas the levels of mGlu2/3 and mGlu7 receptor proteins were reduced. Preemptive and repeated intraperitoneal (i.p.) administration (16 and 1 h before nerve injury and then twice daily for seven days) of minocycline (30 mg/kg) and pentoxifylline (20 mg/kg) prevented the injury‐induced changes in the levels of mGlu3 and mGlu5 receptor mRNAs and the injury‐induced changes in the protein levels of all the receptors. Repeated administration of minocycline and pentoxifylline significantly attenuated CCI‐induced allodynia (von Frey test) and hyperalgesia (cold plate test) measured on day seven after injury and potentiated the antiallodynic and antihyperalgesic effects of single i.p. and intrathecal (i.t.) injections of mGlu receptor ligands: MPEP, LY379268 or AMN082. We conclude that attenuation of injury‐induced glial activation can reduce glutamatergic activity, thereby contributing to regulation of pain sensation.

Overexpression of BDNF increases excitability of the lumbar spinal network and leads to robust early locomotor recovery in completely spinalized rats.

Strategies to induce recovery from lesions of the spinal cord have not fully resulted in clinical applications. This is a consequence of a number of impediments that axons encounter when trying to regrow beyond the lesion site, and that intraspinal rearrangements are subjected to. In the present study we evaluated (1) the possibility to improve locomotor recovery after complete transection of the spinal cord by means of an adeno-associated (AAV) viral vector expressing the neurotrophin brain-derived neurotrophic factor (BDNF) in lumbar spinal neurons caudal to the lesion site and (2) how the spinal cord transection and BDNF treatment affected neurotransmission in the segments caudal to the lesion site. BDNF overexpression resulted in clear increases in expression levels of molecules involved in glutamatergic (VGluT2) and GABAergic (GABA, GAD65, GAD67) neurotransmission in parallel with a reduction of the potassium-chloride co-transporter (KCC2) which contributes to an inhibitory neurotransmission. BDNF treated animals showed significant improvements in assisted locomotor performance, and performed locomotor movements with body weight support and plantar foot placement on a moving treadmill. These positive effects of BDNF local overexpression were detectable as early as two weeks after spinal cord transection and viral vector application and lasted for at least 7 weeks. Gradually increasing frequencies of clonic movements at the end of the experiment attenuated the quality of treadmill walking. These data indicate that BDNF has the potential to enhance the functionality of isolated lumbar circuits, but also that BDNF levels have to be tightly controlled to prevent hyperexcitability.

Bcl-2 and Bax proteins are increased in neocortical but not in thalamic apoptosis following devascularizing lesion of the cerebral cortex in the rat: an immunohistochemical study.

The hypothesis that devascularization of somatosensory and motor cortex causes apoptosis in infarcted regions and in the linked thalamic nuclei was evaluated. To unravel whether Bcl-related proteins, known to regulate apoptosis, participate in neuronal and glial responses to devascularization, we analyzed immunohistochemically the distribution and intensity of staining of Bcl-2 and Bax proteins at different time points after lesion. Both early (up to 6 h) and late (1-7 days) responses were studied. Devascularization led to rapid (within hours) apoptosis in the cortex and to a delayed (within 3-7 days) apoptosis in thalamic nuclei. In control groups, Bcl-2 and Bax immunoreactivity (IR) was detected in neurons and oligodendrocytes but not in astrocytes or microglia. Following devascularization, Bcl-2 IR and Bax IR increased in neurons before the onset of the apoptosis. In the ischemic focus, the increase reached maximal values 3 h after the lesion. The increase was of slower onset in the penumbra zone (24 h and after), a region in which both proteins were induced in astrocytes also. The change of Bax IR intensity exceeded four times that of Bcl-2 at all time points investigated, indicating a diminution of Bcl-2/Bax ratio that may direct neurons to apoptotic pathway. In numerous neurons, an increase of IR in the cytoplasm was accompanied by induction of nuclear staining. No changes of Bcl-2 and Bax IR were found in thalamic nuclei. Our results point to different mechanisms underlying apoptosis of cortical and thalamic neurons. Nuclear appearance of Bcl-2 and Bax suggests they possess regulatory role of gene expression changes triggered by cortical infarct.

Different effects of spinalization and locomotor training of spinal animals on cholinergic innervation of the soleus and tibialis anterior motoneurons.

Cholinergic input modulates excitability of motoneurons and plays an important role in the control of locomotion in both intact and spinalized animals. However, spinal cord transection in adult rats affects cholinergic innervation of only some hindlimb motoneurons, suggesting that specificity of this response is related to functional differences between motoneurons. Our aim was therefore to compare cholinergic input to motoneurons innervating the soleus (Sol) and tibialis anterior (TA) motoneurons following spinal cord transection at a low‐thoracic level. The second aim was to investigate whether deficits in cholinergic input to these motoneurons could be modified by locomotor training. The Sol and TA motoneurons were identified by retrograde labelling with fluorescent dyes injected intramuscularly. Cholinergic terminals were detected using anti‐vesicular acetylcholine transporter (VAChT) antibody. Overall innervation of motoneurons was evaluated with anti‐synaptophysin antibody. After spinalization we found a decrease in the number of VAChT‐positive boutons apposing perikarya of the Sol (to 49%) but not TA motoneurons. Locomotor training, resulting in moderate functional improvement, partly reduced the deficit in cholinergic innervation of Sol motoneurons by increasing the number of VAChT‐positive boutons. However, the optical density of VAChT‐positive boutons terminating on various motoneurons, which decreased after spinalization, continued to decrease despite the training, suggesting an impairment of acetylcholine availability in the terminals. Different effects of spinal cord transection on cholinergic innervation of motoneurons controlling the ankle extensor and flexor muscles point to different functional states of these muscles in paraplegia as a possible source of activity‐dependent signaling regulating cholinergic input to the motoneurons.

Enhancing Proprioceptive Input to Motoneurons Differentially Affects Expression of Neurotrophin 3 and Brain-Derived Neurotrophic Factor in Rat Hoffmann-Reflex Circuitry

The importance of neurotrophin 3 (NT-3) for motor control prompted us to ask the question whether direct electrical stimulation of low-threshold muscle afferents, strengthening the proprioceptive signaling, could effectively increase the endogenous pool of this neurotrophin and its receptor TrkC in the Hoffmann-reflex (H-reflex) circuitry. The effects were compared with those of brain-derived neurotrophic factor (BDNF) and its TrkB receptor. Continuous bursts of stimuli were delivered unilaterally for seven days, 80 min daily, by means of a cuff-electrode implanted over the tibial nerve in awake rats. The H-reflex was recorded in the soleus muscle to control the strength of stimulation. Stimulation aimed at activation of Ia fibers produced a strong increase of NT-3 protein, measured with ELISA, in the lumbar L3-6 segments of the spinal cord and in the soleus muscle. This stimulation exerted much weaker effect on BDNF protein level which slightly increased only in L3-6 segments of the spinal cord. Increased protein level of NT-3 and BDNF corresponded to the changes of NT-3 mRNA and BDNF mRNA expression in L3-6 segments but not in the soleus muscle. We disclosed tissue-specificity of TrkC mRNA and TrkB mRNA responses. In the spinal cord TrkC and TrkB transcripts tended to decrease, whereas in the soleus muscle TrkB mRNA decreased and TrkC mRNA expression strongly increased, suggesting that stimulation of Ia fibers leads to sensitization of the soleus muscle to NT-3 signaling. The possibility of increasing NT-3/TrkC signaling in the neuromuscular system, with minor effects on BDNF/TrkB signaling, by means of low-threshold electrical stimulation of peripheral nerves, which in humans might be applied in non-invasive way, offers an attractive therapeutic tool.

Spinalization and locomotor training differentially affect muscarinic acetylcholine receptor type 2 abutting on α-motoneurons innervating the ankle extensor and flexor muscles

Complete thoracic spinal cord transection (SCT) impairs excitatory cholinergic inputs to ankle extensor (soleus; Sol) but not to flexor (tibialis anterior; TA) α‐motoneurons (MNs) modifiable by locomotor training applied post‐transection. The purpose of this study was to investigate whether Sol and TA MNs adapt to changes in cholinergic environment by differential regulation of their muscarinic receptors M2 (M2R). We examined Chrm2 (M2R gene) transcript level, high‐affinity 3‐quinuclidinyl benzilate‐3H ([3H]QNB) ligand binding, distribution and density of M2R immunolabeling in lumbar (L) segments in intact and SCT rats, with or without inclusion of 5‐week treadmill locomotor training. We show that at the second week after SCT the levels of Chrm2 transcript are reduced in the L3–6 segments, with [3H]QNB binding decreased selectively in the L5–6 segments, where ankle extensor MNs are predominantly located. At 5 weeks after SCT, [3H]QNB binding differences between the L3–4 and L5–6 segments are maintained, accompanied by higher density of M2R immunolabeling in the plasma membrane and cytoplasm of TA than Sol MNs and by enriched synaptic versus extrasynaptic M2R pools (52% TA vs. 25% Sol MNs). Training normalized M2R in TA MNs, improved locomotion, and reduced frequency of clonic episodes. Our findings indicate higher sensitivity of TA than Sol MNs to cholinergic signaling after SCT, which might shorten flexor twitches duration and contribute to generation of clonic movements. Synaptic enrichment in M2R density may reflect a compensatory mechanism activated in TA and Sol MNs to different extent in response to reduced strength of cholinergic signaling to each MN pool.

Electrical Stimulation of Low-Threshold Proprioceptive Fibers in the Adult Rat Increases Density of Glutamatergic and Cholinergic Terminals on Ankle Extensor α-Motoneurons.

The effects of stimulation of low-threshold proprioceptive afferents in the tibial nerve on two types of excitatory inputs to α-motoneurons were tested. The first input is formed by glutamatergic Ia sensory afferents contacting monosynaptically α-motoneurons. The second one is the cholinergic input originating from V0c—interneurons, located in lamina X of the spinal cord, modulating activity of α-motoneurons via C-terminals. Our aim was to clarify whether enhancement of signaling to ankle extensor α-motoneurons, via direct electrical stimulation addressed predominantly to low-threshold proprioceptive fibers in the tibial nerve of awake rats, will affect Ia glutamatergic and cholinergic innervation of α-motoneurons of lateral gastrocnemius (LG). LG motoneurons were identified with True Blue tracer injected intramuscularly. Tibial nerve was stimulated for 7 days with continuous bursts of three pulses applied in four 20 min sessions daily. The Hoffmann reflex and motor responses recorded from the soleus muscle, LG synergist, allowed controlling stimulation. Ia terminals and C-terminals abutting on LG-labeled α-motoneurons were detected by immunofluorescence (IF) using input-specific anti- VGLUT1 and anti-VAChT antibodies, respectively. Quantitative analysis of confocal images revealed that the number of VGLUT1 IF and VAChT IF terminals contacting the soma of LG α-motoneurons increased after stimulation by 35% and by 26%, respectively, comparing to the sham-stimulated side. The aggregate volume of VGLUT1 IF and VAChT IF terminals increased by 35% and by 30%, respectively. Labeling intensity of boutons was also increased, suggesting an increase of signaling to LG α-motoneurons after stimulation. To conclude, one week of continuous burst stimulation of proprioceptive input to LG α-motoneurons is effective in enrichment of their direct glutamatergic but also indirect cholinergic inputs. The effectiveness of such and longer stimulation in models of injury is a prerequisite to propose it as a therapeutic method to improve inputs to selected group of α-motoneurons after damage.