Carry-over effect of deep cerebellar stimulation-mediated motor recovery in a rodent model of traumatic brain injury

Abstract

Background

We previously demonstrated that deep brain stimulation (DBS) of lateral cerebellar nucleus (LCN) can enhance motor recovery and functional reorganization of perilesional cortex in rodent models of stroke or TBI.

Objective

Considering the treatment-related neuroplasticity observed at the perilesional cortex, we hypothesize that chronic LCN DBS-enhanced motor recovery observed will carry-over even after DBS has been deactivated.

Methods

Here, we directly tested the enduring effects of LCN DBS in male Long Evans rats that underwent controlled cortical impact (CCI) injury targeting sensorimotor cortex opposite their dominant forepaw followed by unilateral implantation of a macroelectrode into the LCN opposite the lesion. Animals were randomized to DBS or sham treatment for four weeks during which the motor performance were characterize by behavioral metrics. After four weeks, stimulation was turned off, with assessments continuing for an additional two weeks. Afterward, all animals were euthanized, and tissue was harvested for further analyses.

Results

Treated animals showed significantly greater motor improvement across all behavioral metrics relative to untreated animals during the four-week treatment, with functional gains persisting across two-week post-treatment. This motor recovery was associated with the increase in CaMKIIα and BDNF positive cell density across perilesional cortex in treated animals.

Conclusions

LCN DBS enhanced post-TBI motor recovery, the effect of which was persisted up to two weeks beyond stimulation offset. Such evidence should be considered in relation to future translational efforts as, unlike typical DBS applications, treatment may only need to be provided until such time as a new function plateau is achieved.

Introduction

Traumatic brain injury (TBI) disrupts normal brain function through primary (mechanical damage) and secondary (cascades of metabolic, cellular, and molecular processes) injury mechanisms^1,2, with chronic sequelae across sensorimotor, cognitive, and behavioral domains contributing to long-term disability^3–5. Although cognitive and affective abnormalities are more commonly thought of in relation to such injuries^4,6, persistent sensorimotor dysfunction may also occur, including impaired motor control, loss of muscle tone, balance, and coordination^7–11. Standard-of-care treatment generally involves inpatient and outpatient rehabilitation; however, recovery is often limited and many patients end up in assisted-living or residential care facilities^12,13. Given the limitations of current rehabilitation efforts, there is substantial need to identify and develop effective therapeutic approaches to promote significant and long-lasting behavioral improvements beyond what typically can be achieved^{7 11 14}.

Neurostimulation-based approaches to enhance chronic rehabilitation typically targets surviving brain tissue, with either non-invasive or invasive techniques^15–18. Our group has previously demonstrated that deep brain stimulation (DBS) of the dentatothalamocortical (DTC) pathway enhances the excitability of spared, perilesional cerebral cortex¹⁹ and promotes functional recovery in preclinical models of acquired brain injury^20–22. To achieve chronic DTC pathway activation, we target the lateral cerebellar nucleus (LCN; the rodent homologue of the human dentate nucleus) given its dense, glutamatergic efferent projections to both medial and ventrolateral thalamic nuclei. These projections uniquely position the LCN to influence thalamocortical interactions across widespread prefrontal, frontal, and parietal cortical circuits^23–25, areas that typically represent the perilesional zone following stroke or sensorimotor TBI. Given the link between the incidence of functional reorganization and changes in excitability across these spared cortical regions, we hypothesized that chronic upregulation of the net-excitatory DTC pathway should promote functional recovery²⁶. To date, we have shown that LCN DBS modulates cerebral cortical excitability in both naïve¹⁹ and post-stroke animals²⁷, with the magnitude and persistence of the effect sensitive to stimulus frequency. Moreover, in preclinical models of acquired brain injury, we have shown that treatment-related behavioral improvements are accompanied by increased expression of long-term potentiation markers, synaptogenesis, and neurogenesis in perilesional cerebral cortex^20–22,28.

Specific to TBI, our group has shown previously that LCN DBS was associated with enhancement of both motor²⁹ and cognitive³⁰ rehabilitation using a sensorimotor fluid percussion and bilateral, prefrontal controlled cortical impact (CCI) model of TBI, respectively. In both instances, however, evaluation of benefit was limited to the duration of treatment delivery and no data are available to support the long-term, post-treatment carry-over of benefits achieved with LCN DBS. Such insight has significant translational implications as it would guide the development and design of human trials, particularly with regard to determining whether DBS treatment would need to be delivered indefinitely to maintain benefit or, alternatively, could be short-term in nature and applied only until a new behavioral plateau is achieved. In light of our previous findings, which support underlying enhancement of cellular and molecular mechanisms consistent with plastic reorganization²⁹, we hypothesize that treatment-related benefits will persist beyond cessation of stimulation. Here, we directly evaluate the magnitude and persistence of motor rehabilitation and cellular effects of LCN DBS using a unilateral, sensorimotor CCI TBI model in rats.

Methods and Materials

Animals

Twenty-two male Long Evans rats, at about three- to four-month-old, weighing 225–250g at study onset, were housed on a 12:12, reversed light/dark cycle in a climate-controlled facility. All experiments were conducted under a protocol approved by the Institutional Animal Care and Use Committee of the Cleveland Clinic.

Experimental design – Overview

As summarized in Fig. 1A, animals underwent initial acclimation, pre-training, and baseline assessments over a period of four weeks (weeks 1 – 4). Assessments included the pasta matrix task (primary) as well as the cylinder and horizontal ladder tasks. Once baseline testing was complete, each underwent a stereotaxic-guided controlled cortical impact (CCI) injury centered over the sensorimotor cortex contralateral to the preferred forelimb (week 5). After a two-week recovery period (weeks 5 – 6), a second surgical procedure was performed to implant a concentric, bipolar macroelectrode (MS306; Protech, TX USA) in the LCN contralateral to the CCI-lesioned hemisphere (ipsilateral to the dominant forepaw). After an additional, one-week recovery period, assessments were repeated to quantify the post-CCI motor deficits. Data from the pasta matrix task were used to assign, pseudo-randomly (i.e., severity-matching), animals to one of two experimental groups: STIM- and STIM+. Thereafter, all animals were attached to an in-cage tethering system. Electrical stimulation was activated only for those animals in the STIM+ group for the next four weeks (weeks 9 −12). Throughout that time, pasta matrix testing continued at three days per week, with cylinder and horizontal ladder testing repeated during the fourth week of treatment (week 12). Once the treatment phase was complete, stimulation was turned OFF for those animals assigned to the STIM+ group. Carry-over effects were examined by continuing pasta matrix testing for an additional two weeks (weeks 13–14), with cylinder and horizontal ladder testing repeated again during week 14. At the end of week 14, all animals were euthanized, and brain tissue was harvested and preserved for further analysis by histology.

Fig 1:

A) Experimental timeline. B) Electrode placement. The location of the electrode tips observed in the cerebellar histology was co-registered onto the Paxinos rat brain atlas³⁴. C) Elimination of experimental subjects.

Motor assessments

All assessments were performed during the active phase of the animals’ circadian cycle under red lighting (200 Lux). Animals were food restricted at 11–12.5 g/day during task assessments to encourage performance on the food reward-based, pasta matrix task (weeks 1–4 & 8–14). The cylinder and horizontal ladder tasks were performed at four different time points, including the pre-CCI (naïve) baseline (week 4), two weeks post-CCI (week 8), during the fourth week of DBS treatment or sham (week 12) and at the end of the second week of the carry-over phase (week 14; Fig. 1A).

The pasta matrix task, an index of skilled reaching and forepaw dexterity³¹, was performed once per day, three days per week. Briefly, a rectangular block (7.5 cm wide × 12 cm long × 2.5 cm tall) with a matrix of pre-drilled holes spaced 2 mm apart (2.5 cm deep) was pre-filled with 5 cm long pasta pieces (~1.5 mm thick) and secured to the front panel of the home cage. During each 8-minute session, the animal was able to reach through a narrow, 1-cm wide vertical slit in order to grasp, break, and retrieve pasta pieces. Initially, pasta was placed on both sides of the matrix relative to the vertical slit and forepaw use preference was established by comparing the left-versus right-side retrieval numbers for each animal. Thereafter, pasta was loaded only on the side accessible using the dominant forepaw. Animals that failed to break a minimum of 15 pieces of pasta at behavioral plateau (week 4) were removed from the study (training failure). The pre-CCI baseline (# of pieces broken) was obtained following behavioral plateau (week 4), prior to the CCI induction, while the post-CCI, pre-treatment baseline was derived from data collected one week after the electrode implantation. Testing continued throughout the four-week DBS treatment phase with repeat evaluations performed once per week during the two-week, carry-over phase. The average performance per week was documented.

The cylinder test is used to quantify spontaneous forepaw use during rearing³². Briefly, the animal was removed from the home cage and placed in a high-walled cylinder for eight minutes. Spontaneous rearing activity was recorded using a camcorder mounted above the cylinder. An operator who was blinded to group assignment recorded the number of either left or right forepaw touches against the wall of the cylinder during rearing. Forepaw preference was indexed as:

$$\mathit{\text{Forepaw}}\mathit{\text{preference}}=\frac{[\mathit{\text{Usage}}\mathit{\text{of}}\mathit{\text{affect}}\mathit{\text{paw}}]}{[\mathit{\text{Usage}}\mathit{\text{of}}\mathit{\text{affected}}\mathit{\text{paw}}+\mathit{\text{Usage}}\mathit{\text{of}}\mathit{\text{unaffected}}\mathit{\text{paw}}]}*100\text{\%}$$

The horizontal ladder test, slightly modified from a previous report³³, was used to evaluate motor coordination of each of the four limbs during walking. During testing, the animal was placed at one end of a 60cm long horizontal ladder with random inter-rung spacing (from 1–2cm). Each animal was required to traverse the ladder in a single direction a total of four times per session (the first three as acclimation trials, with data derived only from the final pass). Crossings were videotaped for off-line analysis, during which a blinded reviewer recorded the number of fore- and hind-paw misplacements. Misplacement of paw was defined as a failure of the palmer surface to contact the rungs combined with being grasped on the rung with the digits and palmer surface of the paw, or the paws slipped through the area between the rungs.

CCI induction

Animals were anesthetized using ketamine (50mg/kg, ip) and dexmedetomidine (0.15mg/kg) followed by placement in a stereotaxic frame (Kopf Instruments, Tujunga, CA USA). The scalp was infiltrated with 0.25% bupivacaine followed by a one-inch midline incision. A 6.0mm in diameter circular craniotomy was created over the motor cortex opposite to the dominant forepaw (centered at AP: +1.0mm; ML: +/−3.0mm, relative to bregma). The 5 mm-diameter impactor of the CCI system (Impact One, Leica, IL) was placed on the surface of the sensorimotor without depressing brain tissue. The parameters for impact were pre-set (impact speed: 2.25m/s, depth advanced 2.5mm and dwell time 0.5 sec) and applied uniformly across animals. After CCI induction, the craniotomy was covered with sterile cellulose paper and surgical mesh, then secured using VetBond. The scalp wound was sutured, and anesthesia was reversed by injection of atipemezole (1mg/kg, i.p.). Post-operative care included routine administration of buprenorphine (0.05 mg/kg, s.c.; 2 days post-surgery) and cephazolin (15mg/kg, s.c.) accompanied by regular monitoring for weight loss and signs of pain and distress.

Electrode implantation

Two weeks after the CCI induction, all animals underwent a second surgery for insertion of the LCN DBS electrode (MS306, Protech, VA, USA). Briefly, using the same protocol described above, animals were anesthetized and secured in the stereotaxic frame. The scalp wound was reopened and a bur hole created for the insertion of the electrodes into the contralesional LCN (AP: −2.30 mm; ML: ±3.6 mm; and DV: +3.7 mm; relative to the interaural point). Once in position, the electrode was secured by dental acrylic and four stainless screws anchoring were mounted into the exposed calvaria to serve as anchor points. Post-operative care was identical to that described for the CCI procedure above.

Post-CCI baseline and randomization

Two weeks after electrode implantation, all animals underwent behavioral testing to determine eligibility for continuation, facilitate group assignment, and establish the post-CCI, motor function baseline. As per our prior studies, it was established that animals failing to demonstrate at least a 30% reduction on the pasta matrix testing were considered lesioning failures and were withdrawn from the study, however all animals enrolled in the current study exceeded that threshold and were retained. Thereafter, group assignment was determined by severity-matching of the post-CCI performance on the pasta matrix task (pseudo-randomization).

LCN DBS titration, treatment phase and carry-over phase

Immediately prior to the start of the treatment phase, electrode impedance was tested to ensure electrical integrity. Animals with impedance values greater than 50kΩ were removed from the study. Thereafter, the LCN DBS-induced motor side-effect threshold was established through visual observation by an experienced investigator as described previously^{21 30}. Briefly, 30Hz DBS was delivered in the form of charge-balanced, square wave pulses (pulse width: 400 μsec) using a constant-current biphasic stimulus isolation unit (SIU; Model BSI-2A, Microprobes, MD) controlled by a stimulator (Master-9, AMPI, Israel). The pulse amplitude was increase by 10–20μA and the action of the paws ipsilateral to the implanted electrodes were observed. The increase of pulse amplitude was stopped when the paws were twitching or shaking, which was at the motor threshold. Pulse amplitude was individually set for each animal at 80% of the value of motor threshold. At the start of the treatment phase (week 9), animals in the STIM+ group received DBS for 12 h daily during the active phase of the dark/light cycle. STIM- animals were similarly tethered to the commutator but no stimulation was delivered. The integrity of the electrodes and cables was regularly monitored through periodic impedance checks. Following the four-week treatment phase, animals in the STIM+ group had their SIUs turned off, but still remained tethered, for the two-week carry-over phase.

Euthanasia and tissue processing

At the end of the carry-over phase, all animals were euthanized with 100mg/kg pentobarbital (i.p.) followed by transcardiac infusion of cold normal PBS (1X) and cold 4% paraformaldehyde. The brains were harvested and fixed for 24 hours before cryoprotection in 30% sucrose solution in PBS. Brain blocks were initially stored at −80°C followed by cryostat sectioning of cortical and cerebellar tissue in 30μm sections collected onto gelatin-coated slides.

Volume measuring and validation of electrode placement – Nissl staining.

Cortical sections containing the CCI lesion at the sensorimotor cortex and cerebellar sections containing the electrode tracks were stained with 0.4% cresyl violet (buffered with 0.4M acetate) for 5 min followed by dehydration in ascending percentages of ethanol and defat in xylene. CCI lesion volume was measured using an in-house, MatLab-based program, SLICE³⁴. The electrode tracks of each experimental animals on the cerebellar sections were examined under the bright-field microscope and the location of tips were co-registered onto a rat brain atlas³⁵(Fig 1B).

Immunohistochemistry (IHC)

To characterize potential LCN DBS-mediated changes in perilesional cortex, which was defined as the area 100μm away from the border of the CCI lesion core, locating at the residual sensory and secondary motor cortical regions post-CCI, we used IHC techniques to compare the expression of BDNF and CaMKIIα between treated and untreated cohorts. Briefly, the cerebral sections were treated with goat serum containing blocking buffer after the citrate-based antigen retrieval and peroxidase quenching. Cerebral sections were then incubated with the primary antibodies for BDNF (n = 5 per animal; rabbit anti-BDNF, 1:500; Abcam, MA) or CaMKIIα (n = 5 per animal; rabbit anti-CaMKIIα; 1:500; MilliporeSigma, MA) overnight at 4°C. The secondary antibody, biotinylated anti-rabbit IgG, was incubated for one hour followed by the incubation of the avidin-biotin complex for an additional one hour. The immunocomplex was visualized by developing into brown color in diaminobenzene (DAB) reaction. After mounting the coverslips, the IHC was analyzed under a bright-field microscope.

Image analysis

The 10X images of the perilesional area on the immunostained sections were taken with a bright-field microscope (DM6 B, Leica). The numbers of the BDNF and CaMKIIα positive cells at a pre-defined area (approximately 100μm × 100μm), within 100μm from the borders of the lesion core, were manually counted. A minimum of five sections was analyzed for each subject by an individual blinded to the group assignment. The data are presented as density of cells ±S.E.M. (/μm²).

Statistics analyses

Independent samples t-tests were used to confirm that pre-treatment performance on pasta matrix, cylinder, and ladder rung walking tasks was not significantly different between the STIM+ and STIM− groups. Two random intercepts mixed effects models were then applied to test for group differences in performance for each behavioral task – one including post-CCI baseline through treatment week 4 data, and one including post-CCI baseline through treatment week 6 data. An unstructured covariance matrix was implemented and time was treated as linear for all models. Subject and time point were considered random effects, while treatment group was a fixed effect. A difference in group rate of behavioral task improvement over time was tested with a time-by-group interaction term in each of the models. Analyses were conducted using SAS Studio. P values of < 0.05 were considered significant.

The density of BDNF and CaMKIIα positive cells in both STIM- and STIM+ groups were compared using One-way ANOVA and Bonferroni’s post hoc test (GraphPad Prism). P values of < 0.05 or 0.0001 were considered as statistically significant.

Results:

A total of eight animals were withdrawn over the course of the study: one due to premature implant failure, two due to off-target LCN DBS electrode placement, and an additional five as a result of failure of their SIUs during the course of the study. This yielded a total of seven animals in each group for statistical analysis (Fig. 1C). Impedance values across all animals ranged from 22 to 46kΩ and final pulse amplitude in the STIM+ cohort ranged from 102–189 μA.

Recovery effect of LCN DBS on the motor functions following CCI

As shown in Fig. 2, both groups showed an initial, 73% reduction in the number of broken pasta pieces at four weeks post-CCI, supporting the effect of the sensorimotor CCI on motor function. Thereafter, the mean number of broken pasta pieces in the untreated, STIM- group remained relatively stable throughout the four-week treatment period (up to DBS Wk4). In contrast, animals in the STIM+ group showed an increase in the number of pasta pieces broken at a rate 2.2 times per week more than that observed for the STIM- group across the treatment period (p < 0.001, Fig 2). This culminated in a 37% increase in the number of pieces retrieved by week 4 relative to baseline, supporting an effect of LCN DBS on motor recovery.

Fig 2: Effect of LCN DBS on post-CCI pasta matrix task performance.

Following the CCI induction at the sensorimotor cortex, performance on the pasta matrix task significantly decreased for both STIM- (blue) and STIM+ (red) cohorts. After four weeks of LCN DBS treatment, mean performance was significantly higher in STIM+ versus STIM- animals. Following the offset of the LCN DBS, this difference in task performance persisted between cohorts (DBS Wk6). Group differences at the DBS treatment phase: p < 0.001 and at the carry-over phase: p < 0.001 (by random intercepts mixed effects models).

The above finding was supported further by similar improvements on the secondary motor metrics. As depicted in Fig. 3, both groups demonstrated a marked reduction in use of the predominant forelimb post-CCI. During treatment, use of affected forelimb by animals in the STIM+ group increased by 2.5 percentage points more per week as compared to those in the STIM- group, yielding a week 4 mean difference of 11.88% (DBS Wk4; STIM-: 33.86% vs STIM+: 45.74%, group x time interaction p < 0.001; Fig 3). Notably, there was no significant difference in the incidence of spontaneous rearing activity between cohorts throughout the experimental phase, as indexed by the total number of rears observed at each test session (Fig S2). Finally, CCI increased the misplacement of both the fore- and hindpaws on the horizontal ladder walking test (Fig. 4). Recovery was significant in the STIM+ group, which showed reduced misplacements of the predominant forepaw in comparison to the STIM- group (DBS Wk4; STIM-: 2.86 vs STIM+: 1.71, group x time interaction: 0.36 slips per week, p = 0.037; Fig 4A). A similar effect was observed for hindpaw misplacements (DBS Wk4; STIM-: 1.86 vs STIM+: 1.00, group x time interaction: 0.43 slips per week, p = 0.040; Fig 4B).

Fig 3. Effect of LCN DBS on spontaneous rearing behavior.

CCI resulted in a decrease in dominant forepaw use in both STIM- (blue) and STIM+ (red) groups. In contrast to untreated animals, STIM+ animals showed an increase in dominant forepaw use at week 4. This effect persisted through week 6, two weeks after stimulation was turned OFF for the originally STIM+ animals. Group difference at the DBS treatment phase: p < 0.001 and at the carry-over phase: p < 0.001 (by random intercepts mixed effects models).

Fig 4: Effect of LCN DBS on post-CCI horizontal ladder walking.

Post-CCI animals showed a marked increase in the number of forelimb (A) and hindlimb (B) misplacements for the dominant hemibody. Although both cohorts showed a trend towards improved performance by week 4, the effect only achieved significance for the STIM+ cohort. Following DBS deactivation (STIM+ DBS Wk6), the improvement of hindlimbs was still significant while that of forelimbs was not when compared to the STIM- groups. Group differences of forepaw at the DBS treatment phase: p = 0.037 and at the carry-over phase p = 0.07. Group differences of hindpaw at the DBS treatment phase: p = 0.040 and at the carry-over phase p = 0.003 (by random intercepts mixed effects models).

Carry-over effect of LCN DBS on the motor recovery.

At the end of the treatment phase, stimulation was turned off for those animals assigned to the STIM+ group in order to determine if treatment-related gains would be maintained in the absence of ongoing stimulus delivery. A significant difference in pasta broken was still observed between the STIM+ and STIM- groups at DBS Wk6_carry-over, with STIM+ group breaking pasta at an average rate of 1.2 more pieces per week than the STIM- group (group x time interaction: p < 0.001, Fig 2). In the cylinder test, use of the predominant forelimb in the STIM+ group remained stable at the period of DBS Wk4 to Wk6_carry-over, which resulted in a rate of improvement significantly greater than that of the STIM- group from post-CCI baseline through the carry-over period. The STIM+ group increased predominant forepaw use by 1.5 percentage points more per week on average than the STIM- group (group x time interaction: p < 0.001; Fig. 3). Similarly, on the horizontal ladder test, the improvement in predominant hindpaw performance persisted between DBS Wk4 and Wk6_carry-over for animals in the STIM+ group, which resulted in a rate of improvement significantly higher than that in the STIM- group when measuring from post-CCI baseline through the carry-over period (group x time interaction: p = 0.0030, Fig. 4A). However, from post-CCI baseline through the carry-over period, there was no evidence of a difference in improvement rates of affected forepaw placement in the ladder rung walking task between the STIM+ and STIM – groups. Animals in the STIM+ group committed only 0.17 fewer predominant forelimb slips on average than the STIM- group (group x time interaction: p = 0.070; Fig. 4B).

LCN DBS-mediated increase of expression of BDNF and CaMKIIα.

Following the LCN DBS treatment, there was a significant increase in the density of BDNF positive cells in the perilesional cortex for animals in the STIM+ group when comparing to the STIM- group (One-way ANOVA: df=3, F=16.51, p < 0.0001, Bonferroni test: STIM+ ipsilateral: 0.55±0.05/μm² vs STIM- ipsilateral: 0.37±0.07/μm², p < 0.05; Fig 5 A–B). Similarly, the density of CaMKIIα positive cells in the ipsilesional cortex of STIM+ animals was significantly greater than that of STIM- animals (One-way ANOVA: df=3, F=69.45, p < 0.0001, Bonferroni test: STIM+ ipsilateral: 1.40±0.08/μm² vs STIM- ipsilateral: 0.52±0.04/μm², p < 0.0001; Fig 6A–B). These findings support the persistence of stimulation-related increases in BDNF and CaMKIIα for at least two weeks following the offset of stimulation.

Fig 5: Effect of LCN DBS on BDNF-positive cell density.

A) The density of BDNF-positive cells in the ipsilesional (Ipsil), perilesional cortex of STIM+ significantly greater than that of STIM-. Data was presented as density of cells/μm²±S.E.M., analyzed by one-way ANOVA followed by Bonferroni’ post hoc test. * p < 0.05 when compared to the STIM- Ipsil while ### p < 0.0001 when compared to the STIM+ Contra. B) The representative photomicrographs demonstrated the BDNF-positive cells in both ipsilesional (Ipsil) and contralesional (Contra) perilesional motor cortex of STIM- and STIM+ animals. Scale bar is 100μm.

Fig 6: Effect of LCN DBS on CaMKIIα-positive cell density.

A) The density of CaMKIIα-positive cells in the ipsilesional (Ipsil), perilesional cortex of STIM+ significantly greater than that of STIM-. Data was presented as density of cells/μm²±S.E.M., analyzed by one-way ANOVA followed by Bonferroni’ post hoc test. *** p < 0.0001 when compared to the STIM- Ipsil while ### p < 0.0001 when compared to the STIM+ Contra. B) The representative photomicrographs demonstrated the CaMKIIα-positive cells in both ipsilesional (Ipsil) and contralesional (Contra) perilesional motor cortex of STIM- and STIM+ animals. Scale bar is 100μm.

LCN DBS does not affect the lesion volume.

The volumes of the CCI lesion induced at the sensorimotor cortex of experimental animals in both STIM- and STIM+ groups were not found to differ significantly (STIM-: 35.57±1.23 mm³ vs STIM+: 36.03±1.10 mm³, p = 0.7846; Fig S1), suggesting that LCN DBS does not affect lesion volume.

Discussion

Durability of beneficial effects is an essential element of any post-TBI rehabilitation strategy. Our current results provide not only further evidence for a role of stimulating the ascending cerebellar output circuits in enhancing post-TBI motor recovery but, importantly, support proof-of-principle demonstration of durable change. Specifically, four weeks of LCN DBS, initiated four weeks after CCI induction, improved motor performance across all three motor outcome metrics, consistent with our prior findings²⁹. The therapeutic benefit was maintained after cessation of treatment, with maintenance of performance across all three metrics. The findings are consistent with our hypothesis, based on treatment-enhanced cellular and molecular perilesional reorganization in both chronic post-stroke and post-TBI models^21,22,29,30. The findings are also consistent with carry-over effects observed in our recent first-in-man human translation of DN-DBS for post-stroke rehabilitation³⁶. In the present study, similar changes were found to persist at the end of the two-week carry-over phase, specifically including up-regulation of BDNF and CaMKIIα at the perilesional cortex in the treated group.

The carry-over effect of LCN DBS on motor recovery possibly relies on treatment-induced cellular and molecular modification s across the perilesional area. Chronic LCN DBS-associated motor recovery in the current CCI TBI model was linked to evidence of changes that are associated with calcium-dependent neuroplasticity, a finding that is further consistent with our prior work which revealed evidence of synaptogenesis, in perilesional cortex of LCN DBS-treated, post-stroke animals^21,29. Combined, these data support that treatment-related benefits, once achieved, should be maintained even after stimulation has been discontinued. Notably, although this effect contrasts markedly with most current DBS indications, where continuous stimulation is required to maintain benefit and cessation of stimulation results in motor sign recurrence within seconds to minutes³⁷, there is evidence of stimulation-induced acute and long-term cellular changes in certain instances^38,39.

The transcranial alternating current stimulation at alpha frequency induced a carry-over effect of EEG power at the occipital cortex of healthy human in the range of the stimulation frequency, which is potentially dependent on the spike-timing dependent reorganization of neural activities^40,41. On the other hand, the DBS-mediated cellular and molecular reorganization plays an important role of disease amelioration or symptom-relieving effect in certain diseases. Single-pulse DBS of the internal segment of the globus pallidus (GPi) in patients with dystonia produces long-term potentiation-like neuronal activities, assessed by repetitive pairs of GPi DBS and transcranial magnetic stimulation onto the primary motor cortex and has been argued to play a mechanistic role in GPi-DBS mediated improvements in that population⁴². Furthermore, DBS to the subthalamic nucleus increased the motor evoked potential of the Parkinson’s disease patients, promoting the long-term potentiation in the motor cortex⁴³. In preclinical studies, DBS delivered to the mesencephalic locomotor region in a rat spinal cord injury model promoted the expression of PSD-95 and BDNF in the lesioned spinal tissue associated with the improvement of locomotor functions⁴⁴. The DBS of the nucleus accumbens of rats remotely proliferated the expression of tyrosine hydroxylase and the dendritic length in the prefrontal cortex associated with the anxiolytic effect⁴⁵. Taken together, these findings support the potential for DBS to yield cellular and molecular changes that, similar to the current findings, may play a role in achieving beneficial effects that persist beyond the actual delivery of stimulation.

Cellular and molecular mechanisms:

Loss of motor control was compensated by the treatment-related synaptogenesis and the sprouting of new fibers⁴⁶. These calcium-dependent cellular modifications are associated with reorganization of preserved perilesional tissue, compensating for loss of neural transmission in the injured area^46–48. We propose that chronic LCN DBS upregulates neural activity across the ascending, glutamatergic DTC pathway, thereby creating conditions that promote calcium/CaMKII-dependent synaptogenesis²¹ and BDNF expression³⁰. Although BDNF promotes plastic reorganization and rehabilitation in both stroke and TBI^49,50, the causal relationship between CaMKIIα and BDNF under the condition of LCN DBS remains unclear. One possibility is that activation of the CaMKIIα-dependent long-term potentiation signaling pathways increases the expression and release of BDNF which signals other adjacent neurons to undergo synaptogenesis and dendritogenesis^51,52 and facilitates also the angiogenesis⁵³.

LCN DBS effect on different TBI models:

The current line of work, including its use of a sensorimotor TBI model and emphasis on motor rehabilitation, arose primarily from our prior experience in applying LCN DBS to treat chronic, post-stroke motor deficits^20–22. As noted previously, however, chronic motor deficits can arise post-TBI and negatively impact overall outcomes and quality of life^3,4. To that end, our initial studies in the sensorimotor, fluid percussion injury TBI model showed promise for improving motor outcomes²⁹, but did not include an examination of the post-treatment carry-over effects reported here. The current work is further distinguished from that prior effort in its use of the CCI model, which was adopted for its higher consistency as regards the size and location of the injury⁵⁴. Moreover, we had since adopted the CCI approach as it provides similarly high consistency in establishing a midline, prefrontal injury model in order to examine cognition-related behaviors³⁰, a model that is more suitable for representing post-TBI cognitive deficits in humans^55,56. Additional work is needed, however, to further characterize the effect of therapy on cognitive behaviors post-TBI as well as to examine the overall durability of any such effects.

Limitation:

Interpretation of the current data set is limited by the nature of the study and its design trade-offs. Most importantly, the study does not include several cohorts with multiple treatment durations. Such an experiment would require large groups of animals and could have its own downsides, such as the need for non-concurrent cohorts due to space constraints. Nevertheless, additional experiments can, in the future, identify the limits of stimulation carry-over as well as explore the possibility that longer periods of stimulation may drive even longer carry-over effects. Likewise, shorter treatments can be examined, in order to learn the shortest possible treatment durations that will produce non-inferior results. The implication of experiments examining temporal effects of stimulation are significant for human translation, as they can guide cost-effectiveness of treatment and device design.

It cannot be ruled out that LCN DBS could, as its primary effect, increase home cage, spontaneous motor activity, which, by itself, could promote neuroplasticity and motor recovery. Unfortunately, home cage activity was not monitored throughout the study, however the available evidence from the cylinder test suggests that spontaneous behavior (i.e., total number of rears observed) did not differ significantly between the treated and untreated animal across the trial.

Finally, interpretation of our findings is limited to male animals, which were selected due to the higher rate of TBI in the male population combined with preliminary data in our laboratory that showed evidence of stronger post-CCI spontaneous recovery in female animals⁵⁷. Further work is underway to characterize that difference in order to power and design comparable work using female animals as part of a future endeavor.

Conclusion:

LCN DBS promotes recovery of motor function following sensorimotor CCI injury and is associated with greater expression of CaMKIIα and BDNF in the surviving cortex. These data have significant implications for the design of future translational efforts, as they suggest that DBS targeting DTC pathways may only need to be delivered temporarily until such time as recovery effects have plateaued.

Abbreviations:

ANOVA

Analysis of variance

AP

anterioposterial

BDNF

Brain Derived Neurotrophic Factor

CaMKIIα

Calmodulin dependent kinase II alpha

CCI

controlled cortical impact

DTC

dentatothalamocortical

DAB

3,3’-diaminobenzidine

DBS

deep brain stimulation

DV

dorsoventral

GPi

internal segment of the globus pallidus

IHC

immunohistochemistry

LCN

lateral cerebellar nucleus

PBS

phosphate-buffered saline

PSD-95

Postsynaptic density protein-95

SIU

stimulus isolation unit

TBI

traumatic brain injury