Acute intrathecal BDNF enhances functional recovery after cervical spinal cord injury in rats

Abstract

Unilateral C₂ hemisection (C₂SH) disrupts descending inspiratory-related drive to phrenic motor neurons and thus, silences rhythmic diaphragm muscle (DIAm) activity. There is gradual recovery of rhythmic DIAm EMG activity over time post-C₂SH, consistent with neuroplasticity, which is enhanced by chronic (2 wk) intrathecal BDNF treatment. In the present study, we hypothesized that acute (30 min) intrathecal BDNF treatment also enhances recovery of DIAm EMG activity after C₂SH. Rats were implanted with bilateral DIAm EMG electrodes to verify the absence of ipsilateral eupneic DIAm EMG activity at the time of C₂SH and at 3 days post-C₂SH. In those animals displaying no recovery of DIAm EMG activity after 28 days (n = 7), BDNF was administered intrathecally (450 mcg) at C₄. DIAm EMG activity was measured continuously both before and for 30 min after BDNF treatment, during eupnea, hypoxia-hypercapnia, and spontaneous sighs. Acute BDNF treatment restored eupneic DIAm EMG activity in all treated animals to an amplitude that was 78% ± 9% of pre-C₂SH root mean square (RMS) (P < 0.001). In addition, acute BDNF treatment increased DIAm RMS EMG amplitude during hypoxia-hypercapnia (P = 0.023) but had no effect on RMS EMG amplitude during sighs. These results support an acute modulatory role of BDNF signaling on excitatory synaptic transmission at phrenic motor neurons after cervical spinal cord injury.

NEW & NOTEWORTHY Brain-derived neurotrophic factor (BDNF) plays an important role in promoting neuroplasticity following unilateral C₂ spinal hemisection (C₂SH). BDNF was administered intrathecally in rats displaying lack of ipsilateral inspiratory-related diaphragm (DIAm) EMG activity after C₂SH. Acute BDNF treatment (30 min) restored eupneic DIAm EMG activity in all treated animals to 78% ± 9% of pre-C₂SH level. In addition, acute BDNF treatment increased DIAm EMG amplitude during hypoxia-hypercapnia but had no effect on EMG amplitude during sighs.

INTRODUCTION

Unilateral spinal cord hemisection at C₂ (C₂SH) is a well-established model of incomplete upper cervical spinal cord injury that causes cessation of inspiratory-related diaphragm muscle (DIAm) EMG activity on the ipsilateral side due to disruption of descending inspiratory-related excitatory synaptic drive to phrenic motor neurons (1–9). Over the first few weeks after C₂SH, a proportion of animals display recovery of rhythmic DIAm EMG activity on the ipsilateral side consistent with neuroplasticity and a strengthening of the spared synaptic input to phrenic motor neurons (1, 9–15). Paradoxically, C₂SH has a greater impact on DIAm EMG activity during quiet breathing (eupnea) as compared with more forceful ventilatory (e.g., sighs, airway occlusion) and airway clearance (coughing, sneezing) behaviors (16, 17). Recovery after C₂SH follows a similar pattern, with earlier and more pronounced recovery of more forceful behaviors compared with eupnea.

In previous studies, we showed that brain-derived neurotrophic factor (BDNF) acting through its high-affinity tropomyosin-related kinase type B (TrkB.FL) receptor plays an important role in promoting neuroplasticity following C₂SH (9, 16–23). For example, after C₂SH recovery of rhythmic ipsilateral DIAm activity during eupnea is enhanced by chronic intrathecal delivery of BDNF at the level of the phrenic motor neuron pool (9, 16). This effect of BDNF/TrkB.FL signaling may be mediated via either a postsynaptic effect at phrenic motor neurons or a presynaptic effect at excitatory glutamatergic (Glu) synaptic input to phrenic motor neurons. Postsynaptically, at phrenic motor neurons, we previously showed that prolonged inhibition of TrkB.FL kinase activity markedly blunts recovery of rhythmic DIAm EMG activity after C₂SH (9, 24). Thus, the chronic effects of BDNF/TrkB.FL signaling support postsynaptic neuroplasticity. However, we also showed that acute BDNF treatment enhanced synaptic transmission at DIAm neuromuscular junctions, whereas acute inhibition of TrkB.FL signaling impaired neuromuscular transmission (25–29). A similar enhancement of excitatory synaptic input to lumbar motor neurons occurs in the neonatal rat after acute BDNF treatment (30). Accordingly, we hypothesized that acute intrathecal delivery of BDNF to phrenic motor neurons will enhance recovery of rhythmic DIAm EMG activity after C₂SH.

MATERIALS AND METHODS

Experimental Animals

Adult male Sprague-Dawley rats (Harlan, Indianapolis, IN; initial body wt ∼300 g) were used in the present study. The Institutional Animal Care and Use Committee at Mayo Clinic reviewed and approved all experimental procedures. All surgical procedures and experimental measurements were performed under anesthesia with an intramuscular injection of ketamine (90 mg/kg) and xylazine (10 mg/kg) and maintained by intermittent repeat dosing as needed.

Spinal Cord Hemisection (C₂SH)

The C₂SH procedure has been previously described in detail (1–4, 6, 7, 9, 16, 17, 22, 23, 31). Briefly, a dorsal C₂ laminectomy was performed on anesthetized rats, and the right anterolateral spinal cord was partially transected using a microknife, such that the dorsal funiculus was preserved, and the lateral and ventral funiculi were transected. DIAm EMG activity was monitored at the time of surgery to ensure that eupneic activity on the ipsilateral side disappeared following transection. Subsequently, animals were observed daily after surgery and administered oral acetaminophen (100–300 mg/kg) and intramuscular buprenorphine (0.1 mg/kg) for the first 3 days postsurgery.

Chronic DIAm EMG Recordings

Three days before C₂SH surgery, two pairs of insulated wire electrodes (0.28 mm diameter; model AS631, Cooner Wire Inc., Chatsworth, CA), stripped ∼3 mm at the tip, were implanted into both the left and right sides of the DIAm in the midcostal region. The wires were then tunneled subcutaneously, and externalized in the back of the animal, as detailed previously (1, 9, 16, 17, 22, 23, 31–37). DIAm EMG activity was assessed during C₂SH surgery and at 3 days after C₂SH (C₂SH 3 D) to verify the effective removal of descending inspiratory-related excitatory drive as indicated by the absence of ipsilateral inspiratory-related DIAm activity. No animals were excluded on this basis. Subsequently, eupneic inspiratory-related DIAm EMG activity was recorded at 7, 14, and 28 days (C₂SH 7 D, C₂SH 14 D, and C₂SH 28 D, respectively). Only animals in which eupneic DIAm EMG activity was absent at C₂SH 28 D were included in this study (7 out of 17 animals). In these animals, DIAm EMG activity was additionally recorded during exposure to a hypoxia-hypercapnia gas mixture (10% O₂ and 5% CO₂) for 5 min (16, 34, 38, 39). In addition, DIAm EMG activity was recorded during the spontaneous incidence of sighs, defined as an inspiratory effort >2 times the amplitude of average eupneic DIAm EMG activity on the intact contralateral side (16, 34, 38, 39).

The DIAm EMG signal was amplified (×2,000), band-pass filtered between 20 Hz and 1 kHz (Model 2124, DATA Inc.), and analog-to-digital converted at 2 kHz using LabView data acquisition board and software (National Instruments, Austin, TX). The root mean square (RMS) EMG calculates the square root of the average power of the EMG signal for a given window period of time (50 ms) that moves across the entire recording period (31, 34, 39–41). Compared with integrated EMG, which is the mathematical integral of the absolute value of the raw EMG signal, the mean RMS value is a more robust measure of EMG amplitude, which is less affected by movement artifacts and signal noise (42). Recovery of ipsilateral DIAm EMG activity during eupnea was indicated by the presence of 1) rhythmic activity in phase with EMG activity on the contralateral side; 2) consistent activity occurring for at least 90% of eupneic bursts; 3) activity that comprised multiple motor unit discharge; and 4) RMS EMG amplitude that was >10% of the pre-C₂SH amplitude. In all cases, the incidence of sighs was detected from EMG activity on the contralateral side and RMS EMG sigh amplitude was calculated from the sighs present in the hypoxia-hypercapnia recording. Respiratory rate and duty cycle were calculated from contralateral DIAm EMG recordings.

Acute BDNF Treatment

In animals where eupneic inspiratory-related DIAm EMG activity was absent on the ipsilateral side at C₂SH 28 D, the cervical spinal cord segments containing the phrenic motor nucleus (C₃–C₅) were surgically exposed and BDNF was instilled intrathecally (15 ng/µL, 1 µL/min for 30 min; R&D Systems, Inc., Minneapolis, MN) at C₄. Immediately before and 30 min after BDNF treatment, DIAm EMG activity was recorded during eupnea, hypoxia-hypercapnia, and spontaneous sighs.

Statistical Analyses

All statistical evaluations were performed using JMP statistical software (JMP 10.0, SAS Institute Inc., Cary, NC). Diaphragm RMS EMG amplitude was normalized to the eupneic value before C₂SH for the same animal. Differences in DIAm RMS EMG amplitude were compared using matched pairs t test for each behavior and side (ipsilateral and contralateral). The correlation between pre- and post-BDNF DIAm RMS EMG amplitude was calculated by linear regression for both ipsilateral and contralateral sides of the DIAm including all behaviors. The degree of correlation was expressed as r² values, with the effectiveness of the model (difference from the mean response) determined by ANOVA. Differences in inspiratory-related DIAm EMG respiratory pattern were compared using matched paired t test. Statistical significance was established at the 0.05 level. All experimental data are presented as means ± SE, unless otherwise specified.

RESULTS

C₂SH surgery was successfully performed in all animals, as indicated by the absence of inspiratory-related DIAm EMG activity on the ipsilateral side at the time of surgery and at C₂SH 3 D. Inspiratory-related DIAm EMG activity during eupnea continued on the contralateral side. During breathing (i.e., during eupnea and hypoxia-hypercapnia), spontaneous sighs were evident in all but one animal after BDNF treatment. Sighs were defined as individual deep inspiratory efforts on the contralateral side of the DIAm with RMS EMG amplitude that was at least twofold greater than the amplitude of eupneic breaths (16, 34, 39). Representative DIAm EMG recordings and RMS EMG tracings are shown in Fig. 1 during eupnea, hypoxia-hypercapnia, and sighs.

Figure 1.

Ipsilateral diaphragm muscle (DIAm) EMG activity after acute treatment with BDNF. Representative raw DIAm EMG recordings and root mean square (RMS) EMG tracings during eupnea, hypoxia-hypercapnia, and sigh at 28 days post-C₂SH (C₂SH 28 D) and after BDNF treatment (C₂SH 28 D + BDNF). Animals were chronically monitored with EMG recordings obtained via implanted DIAm electrodes. Successful C₂SH surgery was indicated by the absence of inspiratory-related DIAm EMG activity on the ipsilateral side at the time of surgery and at C₂SH 3 D. Intrathecal BDNF was given only to animals not displaying eupneic ipsilateral DIAm EMG activity at C₂SH 28 D (n = 7). The traces shown in this figure represent animal “△” in Fig. 2. BDNF, brain-derived neurotrophic factor.

Effect of Acute BDNF on Recovery of DIAm EMG Activity after C₂SH

As in our previous studies (9, 22, 23, 31), we found that at C₂SH 28 D, recovery of eupneic inspiratory-related DIAm activity on the ipsilateral side was absent in only ∼40% of animals (n = 7 out of 17 rats). In these rats, intrathecal BDNF treatment restored inspiratory-related DIAm EMG activity during eupnea on the ipsilateral side in all animals. DIAm EMG RMS amplitude was 78% ± 9% of pre-C₂SH EMG eupneic amplitude within 30 min of administration (Fig. 2; P < 0.01 compared with before BDNF treatment). Exposure to hypoxia-hypercapnia restored inspiratory-related DIAm EMG activity on the ipsilateral side but at a much lower amplitude (28% ± 2% RMS EMG) compared with the contralateral side (Figs. 1 and 2; P < 0.01). Acute intrathecal BDNF treatment markedly increased ipsilateral DIAm RMS EMG amplitude during hypoxia-hypercapnia by ∼128% compared with EMG amplitude before BDNF (Figs. 1 and 2; P = 0.02). Spontaneous sighs were present on the ipsilateral side at C₂SH 28 D, but at a much lower amplitude (55% of RMS EMG on the contralateral side; Figs. 1 and 2; P = 0.02). Acute BDNF treatment had no effect on DIAm RMS EMG amplitude during sighs (Figs. 1 and 2; P = 0.22). Acute BDNF treatment also had no effect on the incidence of sighs, with 0.81 ± 0.21 sighs/min at SH 28 D before BDNF treatment and 0.51 ± 0.12 sighs/min after BDNF treatment (P = 0.22).

Figure 2.

Ipsilateral and contralateral diaphragm muscle (DIAm) root mean square (RMS) EMG amplitude pre- and post-BDNF treatment during eupnea (Eup), hypoxia-hypercapnia (HH), and sigh in animals at 28 days post-C₂SH (C₂SH 28 D). DIAm root mean square (RMS) EMG amplitude was normalized to the eupneic value before C₂SH for the same animal (bars represent mean ± 95% confidence interval; each symbol shape represents an individual animal). In C₂SH animals displaying lack of ipsilateral DIAm EMG activity at C₂SH 28 D, intrathecal BDNF significantly increased ipsilateral RMS EMG amplitude during eupnea and hypoxia-hypercapnia (*P < 0.05 compared with before BDNF). Acute BDNF treatment had no effect on ipsilateral DIAm RMS EMG amplitude during sigh. In addition, BDNF increased contralateral DIAm RMS EMG amplitude during eupnea compared to DIAm RMS EMG amplitude before BDNF but had no effect on hypoxia-hypercapnia or sigh. BDNF, brain-derived neurotrophic factor.

Acute intrathecal BDNF treatment also affected the amplitude of DIAm EMG activity on the contralateral side. During eupnea, acute BDNF treatment increased contralateral DIAm RMS EMG by ∼25% compared with EMG amplitude before BDNF (Fig. 2; P = 0.04). However, acute BDNF treatment had no effect on DIAm EMG amplitude on the contralateral side during hypoxia-hypercapnia (Fig. 2; P = 0.09) or during spontaneous sighs (Fig. 2; P = 0.32).

For each animal, the difference in RMS EMG (as a percent of pre-C₂SH eupnea) between post-BDNF and pre-BDNF was calculated for each behavior (Fig. 3). There was no difference in this within-animal RMS EMG change across behaviors on the ipsilateral (P = 0.89) or contralateral side (P = 0.62).

Figure 3.

The difference in diaphragm muscle (DIAm) root mean square (RMS) EMG (as a percent of pre-C₂SH eupnea) between post-BDNF and pre-BDNF for each animal during eupnea (Eup), hypoxia-hypercapnia (HH), and sigh. There was no difference in this within-animal DIAm RMS EMG change across behaviors on the ipsilateral (P = 0.89) or contralateral side (P = 0.62). Bars represent means ± 95% confidence interval; each symbol shape represents an individual animal. BDNF, brain-derived neurotrophic factor.

Correlation between Pre- and Post-BDNF DIAm EMG Activity

For each side of the DIAm, there was a significant positive correlation between DIAm EMG amplitude across the three motor behaviors (Fig. 4). On the ipsilateral side, there was a positive correlation between pre- and post-BDNF treatment DIAm RMS EMG amplitude (P < 0.01; r² = 0.45; slope = 1.26). Similarly, on the side of the DIAm contralateral to C₂SH, there was also a significant positive correlation between pre- and post-BDNF DIAm RMS EMG amplitude with a similar slope as the ipsilateral side (P < 0.01; r² = 0.77; slope = 1.16).

Figure 4.

Correlation between pre- and post-BDNF diaphragm muscle (DIAm) EMG activity across all behaviors. On both the ipsilateral and contralateral side, there was a significant, positive correlation between post- and pre-BDNF RMS EMG amplitude of similar slope (solid line; ipsilateral: 1.26 and contralateral: 1.16), indicating that BDNF-mediated enhancement of DIAm EMG activity is common across behavior and side. Each symbol shape represents an individual animal and symbol color represents behavior. The dotted line represents a slope of 1, such that post-BDNF equals pre-BDNF. BDNF, brain-derived neurotrophic factor; Eup, eupnea; HH, hypoxia-hypercapnia.

Ventilatory Parameters

Ventilatory parameters were measured from the contralateral DIAm EMG recordings during eupnea and hypoxia-hypercapnia pre-C₂SH, pre-BDNF at C₂SH 28 D, and post-BDNF at C₂SH 28 D (Table 1). During eupnea, C₂SH caused a significant increase in DIAm EMG burst duration compared with pre-C₂SH and BDNF treatment caused a further increase in burst duration, without a change in respiratory rate or duty cycle. There were no significant changes in ventilatory parameters induced by C₂SH or BDNF treatment during hypoxia-hypercapnia.

Table 1.

Ventilatory parameters before and after acute intrathecal BDNF treatment at 28 days after unilateral cervical spinal cord hemisection at C₂

		Time
		Pre-C2SH	C2SH 28 D	C2SH 28 D + BDNF
Behavior	Ventilatory Parameter		vs.		P Value
Eupnea	Respiratory rate, min−1	69 ± 4	73 ± 4	65 ± 6	0.297
	Burst duration, ms	240 ± 5	296 ± 18†	414 ± 48*	0.034
	Duty cycle, %	27 ± 1	36 ± 3	42 ± 6	0.160
Hypoxia-Hypercapnia	Respiratory rate, min−1		93 ± 8	66 ± 12	0.100
	Burst duration, ms		304 ± 20	458 ± 79	0.076
	Duty cycle, %		46 ± 4	43 ± 3	0.514

Data presented as means ± SE. Duty cycle was calculated as the percent of total time corresponding to inspiration. All analyses were conducted using matched pairs t test. Shown are only the P values for the treatment effect (pre- vs. post-BDNF). *P < 0.05 compared to C2SH 28D; †P < 0.05 compared with pre-C2SH. BDNF, brain-derived neurotrophic factor;C₂SH, spinal cord hemisection at C₂.

DISCUSSION

The results of the present study show that acute intrathecal BDNF treatment at the level of the phrenic motor neuron pool enhances recovery of ventilatory-related (i.e., eupnea and hypoxia-hypercapnia) DIAm EMG activity on the ipsilateral side after C₂SH. In those animals showing no rhythmic DIAm EMG activity during eupnea at 28 days after C₂SH, acute intrathecal BDNF restores a rhythmic pattern at an amplitude that was 78% of pre-C₂SH level. In C₂SH 28 D animals in which there was no eupneic DIAm EMG activity, exposure to hypoxia-hypercapnia resulted in a partial recovery of rhythmic inspiratory-related DIAm EMG activity. This partial recovery during hypoxia-hypercapnia was further enhanced by intrathecal BDNF treatment. By comparison, the amplitude of DIAm EMG activity during spontaneous sighs was not impacted by C₂SH and there was no effect of BDNF treatment. These results suggest an acute effect of BDNF/TrkB.FL signaling on neuroplasticity at phrenic motor neurons.

In previous studies, we (1) and others (43) showed that by disrupting descending excitatory (glutamatergic) bulbospinal inspiratory-related synaptic input to ipsilateral phrenic motor neurons, C₂SH silences eupneic DIAm EMG activity. Based on electron microscopic analysis of putative phrenic motor neurons, Tai and Goshgarian (43) reported an ∼40% loss in excitatory glutamatergic terminals by 30 days after C₂SH. This was consistent with the ∼50% overall loss in glutamatergic synaptic terminals at ipsilateral phrenic motor neurons by 7 days after C₂SH (1). In another recent study, we reported that the density of glutamatergic synaptic terminals is greater at smaller compared to larger phrenic motor neurons (44). It should be noted that smaller motor neurons are more excitable due to their intrinsic electrophysiological properties, which underlies the size principle for motor unit recruitment (45). Thus, the greater glutamatergic synaptic density at smaller phrenic motor neurons ensures their earlier recruitment during inspiration (38, 39, 46–51). Importantly, we found that following C₂SH the reduction in glutamatergic synapses at smaller phrenic motor neurons was much more pronounced compared to that at larger phrenic motor neurons (∼60% vs. 35% reduction) (1).

The fact that descending excitatory inspiratory-related drive to phrenic motor neurons is predominantly ipsilateral is supported by the silencing of eupneic DIAm EMG activity after C₂SH (3, 8, 23, 24, 52). Although rhythmic DIAm EMG activity during eupnea is silenced after C₂SH and DIAm EMG activity during hypoxia-hypercapnia is markedly reduced, DIAm EMG activity during spontaneous sighs and inspiratory efforts against an occluded airway are not as affected following a C₂SH (16, 17). Again, these differences in the impact of C₂SH on DIAm EMG activity reflect the disproportionate loss of ipsilateral descending glutamatergic synapses on smaller phrenic motor neurons (1).

Considerable attention has focused on the contribution of spared contralateral descending excitatory input to phrenic motor neurons in restoring DIAm activity following C₂SH. In 1895, Porter (53) first described the “crossed-phrenic” phenomenon suggesting that enhancement of contralateral descending excitatory input to phrenic motor neurons could restore inspiratory-related DIAm activity on the ipsilateral side. In the study by Porter, the phrenic nerve on the contralateral side was transected after C₂SH and immediately inspiratory-related DIAm activity reappeared on the ipsilateral side. It was speculated that the crossed-phrenic phenomenon was mediated by strengthening of contralateral inputs to phrenic motor neurons that were previously sub-threshold for inspiratory-related recruitment. Consistent with a strengthening of extant excitatory synaptic input, it was subsequently reported that any augmentation of inspiratory drive unmasks subthreshold activation of phrenic motor neurons after C₂SH and restores inspiratory-related DIAm EMG activity. Indeed, this is a well-accepted model for neuroplasticity in the respiratory system.

The respiratory pattern is generated in the pre-Bötzinger complex (preBötC) (54), and inspiratory drive is transmitted primarily via premotor neurons located in the ventrolateral respiratory group (VRG) of the medulla (55, 56). Ventilatory behaviors of the DIAm include eupnea, stimulated breathing during hypoxia-hypercapnia, spontaneous deep inspiration or sighs, and forceful efforts against an occluded airway. Previously, we found that DIAm EMG activity during sighs and airway occlusion were far less impacted by C₂SH (16, 17). Our current study and previous studies from our laboratory show that DIAm hemi-paralysis, either due to phrenic nerve denervation or C₂SH, does not affect ventilation (9, 37, 57). Recent evidence indicates that sighs reflect enhanced inspiratory efforts that are mediated via projections from a small population of ∼200 peptidergic neurons in the retrotrapezoid nucleus/parafacial respiratory group (RTN/pFRG) to the preBötC, which either convert a normal breath to a sigh (58) or alternatively, reconfigure neuronal activity in the preBötC (59). Most importantly, the premotor output to phrenic motor neurons would increase in strength (i.e., inspiratory drive) during sighs. Our results suggest a more balanced bimodal distribution of premotor drive to phrenic motor neurons such that unilateral disruption associated with C₂SH does not diminish activation of the phrenic motor neuron pool during sighs. By contrast, the normal enhancement of VRG premotor drive during hypoxia-hypercapnia is disrupted by C₂SH and shows a similar pattern of recovery to that of eupnea. Thus, by eliminating the predominant ipsilateral premotor drive to phrenic motor neurons C₂SH silences DIAm EMG activity during eupnea and hypoxia-hypercapnia. Premotor input for higher force, ventilatory (sigh, airway occlusion) and airway protective (e.g., coughing, sneezing) behaviors appears to be more evenly distributed and not predominantly ipsilateral in contrast to ventilatory behaviors (1, 60). In the present study, it did not appear that acute BDNF treatment enhanced premotor drive to phrenic motor neurons mediating higher force, ventilatory behaviors and airway protective behaviors.

There is considerable evidence that BDNF treatment acting through its high-affinity TrkB.FL receptor promotes neuroplasticity following spinal cord injury (9, 18–21), and enhances spontaneous functional recovery over time (9, 17, 22, 23, 61). For example, we previously reported that chronic intrathecal delivery of BDNF in the cervical spinal cord promotes recovery of rhythmic DIAm EMG activity following C₂SH (9, 16). The half-life of BDNF is ∼1 h (62), which could affect the duration of the short-term beneficial effect of BDNF on the recovery of eupneic DIAm EMG activity. We also showed that transplantation of bone marrow derived stem cells engineered to release BDNF in the cervical spinal cord adjacent to the phrenic motor neurons pool enhances functional recovery of rhythmic DIAm activity (22, 63). At the postsynaptic level, chronic BDNF treatment upregulates mRNA expression of NMDA glutamatergic receptors in phrenic motor neurons (7, 64, 65). Furthermore, BDNF/TrkB.FL signaling may mediate NMDA receptor phosphorylation (66, 67) and such an effect could mediate more acute neuroplasticity following C₂SH. As such, post-translational modifications may be one possibility for the acute effect of BDNF. There is also evidence that BDNF/TrkB signaling promotes recovery of rhythmic DIAm activity following C₂SH by upregulating serotonergic receptor (5-HT2A) expression in phrenic motor neurons (7, 64, 65). In previous studies, we reported that siRNA knockdown of TrkB.FL expression in phrenic motor neurons or chemically induced inhibition of TrkB kinase activity in a chemical-genetic mouse model (TrkB^F616A knockin mice) prevented recovery of eupneic DIAm EMG activity after C₂SH (9, 24). Together, these results indicate a postsynaptic effect of BDNF/TrkB.FL signaling at phrenic motor neurons.

A presynaptic target in neuroplasticity is suggested by the effect of acute BDNF treatment in enhancing neuromuscular transmission in the DIAm (29, 68). In addition, we showed that acute 1NMPP1-induced inhibition of TrkB.FL kinase activity in TrkB^F616A knockin mice impairs neuromuscular transmission in the DIAm (25–28). Acute BDNF delivery has been implicated in synaptic transmission and plasticity in motor neurons and hippocampal cultures (30, 69–71). Specifically, in a phrenic long-term facilitation (pLTF) model of increased phrenic motor output induced by intermittent hypoxia (an example of neuroplasticity), acute treatment with BDNF enhances pLTF, whereas RNAi-mediated interference of BDNF expression blocks pLTF (69). Similarly, BDNF plays a role in regulating long-term potentiation (LTP), the best studied form of synaptic plasticity in the hippocampus, by enhancing synaptic responses to tetanic stimulation; however, several hours of BDNF treatment is necessary for this response (72, 73). Acute BDNF treatment has also been shown to enhance excitatory synaptic input to lumbar motor neurons in the neonatal rat (30). Presynaptically, acute BDNF increases glutamate release, thus enhancing the frequency of miniature excitatory postsynaptic currents (mEPSCs) in hippocampal microcultures without affecting their amplitude or kinetics (66, 70). In hippocampal slices, two hours of BDNF treatment increases the probability of neurotransmitter release at hippocampal excitatory synapses by modulating the size of a readily releasable pool of synaptic vesicles (71). Postsynaptically, BDNF elicits a dose-dependent increase in phosphorylation of NMDA subunit 1 within 5 minutes of exposure in hippocampal neurons (74), and augments NMDA receptor single channel opening probability in hippocampal neurons within 25 minutes (75).

In summary, the results of the present study show that acute intrathecal BDNF treatment is effective in promoting short-term neuroplasticity in phrenic motor neurons and thereby partially restoring ventilatory-related DIAm activity following C₂SH. These results indicate that acute BDNF treatment may be used therapeutically in treating patients with upper cervical spinal cord injury to restore eupneic DIAm activity and improve ventilation. However, how long the effects of acute intrathecal BDNF persist beyond 30 min will need to be systematically explored to establish a potential therapeutic benefit after spinal cord injury.

GRANTS

This study was supported by National Institutes of Health (NIH) Grants R01 HL96750, HL146114 and by the Mayo Clinic.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

G.C.S., H.M.G., and C.B.M. conceived and designed research; W.-Z.Z. performed experiments; G.C.S., H.M.G., and C.B.M. analyzed data; G.C.S., H.M.G., and C.B.M. interpreted results of experiments; G.C.S., H.M.G., and C.B.M. prepared figures; G.C.S., H.M.G., and C.B.M. drafted manuscript; G.C.S., H.M.G., W.-Z.Z., and C.B.M. edited and revised manuscript; G.C.S., H.M.G., W.-Z.Z., and C.B.M. approved final version of manuscript.