2.1 Electrophysiology

For electrophysiological experiments, 10 mice from the control group and 10 from the 3-NP treated group were used. Two days after the last injection of 3-NP or vehicle, the mice were anesthetized with halothane in an induction chamber; then, the mice were decapitated to obtain the brains. The brains were placed in low-calcium physiological Ringer's solution that was chilled (4°C) and constantly oxygenated (95% O₂, 5% CO₂) and contained, in mM, 130 NaCl, 3 KCl, 1 CaCl₂, 5 MgCl₂, 1.25 NaH₂PO₄, 26 NaHCO₃, and 10 glucose (298–300 mOsm, pH 7.4). Sagittal brain slices (400 µm) containing the striatum were obtained through a vibroslicer (Pelco 1000, Ted Pella Inc., Redding, CA, USA). The slices were maintained at room temperature (25°C) in oxygenated physiological saline for 1 hr prior to recording to recover from slicing.

2.2 Stimulation parameters and electrophysiological recordings

For these experiments, synaptic population activity was measured in a recording chamber perfused with physiological saline (in mM; 3 KCl, 125 NaCl, 2 CaCl₂, 1 MgCl₂, 26 NaHCO₃, 0.2 thiourea, 0.2 ascorbic acid, and 10 glucose, pH 7.4) that was constantly bubbled (95% O₂, 5% CO₂) at a constant flow rate (1 ml/min) and a controlled temperature (32 ± 1°C).

Population spikes were generated by stimulating cortical afferents to the striatum with stimuli of variable intensity and duration (5–30 V, 50–200 μs) at 0.1 Hz with an isolated stimulator (Digitimer DS2, Letchworth Garden City, UK) through a concentric bipolar tungsten electrode (50 µm, CBCEE75, FHC, Inc., Bowdoinham ME, USA) placed in the corpus callosum. To analyze neurotrophin-induced changes on corticostriatal transmission, a paired-pulse (PP) protocol was used with a 30- to 50-ms interval, and paired-pulse ratio (PPR, S2/S1 * 100) was evaluated. The PP protocol is based on the residual calcium hypothesis, which establishes that changes in PPR result from changes in intracellular levels of Ca²⁺ in response to a pair of stimulus of same intensity and duration given in a short period of time (Atluri & Regehr, 1996; Dittman, Kreitzer, & Regehr, 2000; Zucker & Regehr, 2002).

To record the signal, borosilicate glass electrodes (30-30-1, FHC, Inc.) were pulled (P-97, Sutter Instruments, Novato, CA, USA) and filled with NaCl, 0.9%. The electrophysiological signal was amplified (AC, GRASS P15, Quincy, MA, USA), filtered (1 kHz), and acquired with an interface coupled to a PC running LabVIEW 6.1 (National Instruments, Austin, TX, USA).

To isolate glutamatergic activity, (-)-bicuculline methiodide (10 µM) was used in all experiments to block GABAergic inhibitory responses. Population spikes were recorded for 10 min before neurotrophin treatment. After 10 min of stable recordings, BDNF or NT-4/5 (10–50 ng/ml, PeproTech, Inc.) dissolved in distilled water (following the vendor instructions) was applied, and after 20 min, the other neurotrophin was added (BDNF → NT-4/5 or NT-4/5 → BDNF) for 10–30 min.

The data were digitized, stored, and analyzed offline with the aid of Microcal Origin 9.1 (Microcal Origin, Lab Corp., Northampton, MA, USA, RRID:SCR_002815).

2.3 Western blot analysis of cerebral slices

To evaluate changes in receptor protein expression, cerebral slices from 10 control mice and 10 3-NP-treated mice containing the striatum were incubated at RT in physiological saline for electrophysiological experiments that contained bicuculline and was bubbled (95% O₂ and 5% CO₂). Then, the slices were exposed to (a) BDNF, (b) BDNF → NT-4/5, (c) NT-4/5, or (d) NT-4/5 → BDNF for 10 min. Then, striatal tissue from each group was homogenized manually in lysis buffer (Tris–HCl 26 mM, 1% Triton X-100, glycerol 1.3 M, NaCl 130 mM) with cOmplete Mini phosphatase inhibitor tablets (Roche, Indianapolis, IN, USA), and proteins were obtained following the methodology described previously (Torres-Cruz et al., 2019).

For electrophoresis, each well of 10% acrylamide/bisacrylamide gels was loaded with 30 µg of protein (quantified by the Bradford method). For protein separation, a current of 90 V and 15 A was applied for 1 hr 40 min in an electrophoresis chamber (Mini-PROTEAN® Tetra Handcast System, Bio-Rad Laboratories, Inc., Hercules, CA, USA) containing running buffer. After electrophoresis, the proteins were transferred to polyvinylidene fluoride (PVDF) membranes and blocked with 10% nonfat milk and bovine serum albumin (BSA, 5%) in PBS overnight (4°C). Once the proteins had been transferred to the PVDF membranes, the membranes were incubated with primary antibodies against total TrkB (1:20,000, Cell Signaling Technology Cat# 4603, RRID:AB_2155125), phosphor-Tyr816-TrkB (1:500–1:2,000, Millipore Cat# ABN1381, RRID:AB_2721199), and actin (1:1,000; Diaz-Barriga et al., 1989) for 12 hr at 4°C. After rinsing, the membranes were then incubated for 1–2 hr at room temperature with secondary antibodies (1:20,000) conjugated with horseradish peroxidase (HRP). The secondary antibodies, used were horseradish peroxidase-conjugated goat anti-mouse IgG (Thermo Fisher Scientific Cat# 62-6520, RRID:AB_2533947) and horseradish peroxidase-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific Cat# 65-6120, RRID:AB_2533967). The proteins were visualized with the chemiluminescence method (Gel Documentation System, Biosens SC 645, Shanghai Bio-Tech Co., Ltd) and imaged with Kodak photographic plates. The plates were scanned, and the signal obtained was analyzed by densitometry with the aid of ImageJ software (NIH). To obtain the relative density, we used the following equation: Relative density = phosphorylated protein/total protein.

2.4 COS-7 cell culture and transfection

Cell culture and transfection was performed with methodology previously described (Torres-Cruz et al., 2019). Briefly, COS-7 cells from the American Type Culture Collection (ATCC; Manassas, VA, USA) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA), 2 mM L-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin and maintained in a humidified atmosphere (5% CO_2, 37°C). Once the cells reached 50%–60% confluence, the medium was changed to serum-free Opti-MEM specialized medium (Gibco), and the cells were cotransfected with 1 μg of TrkB DNA constructs [plasmids pGFP-N1-TrkB (RRID:Addgene_32500) and pRc/CMV HA-TrkB.T1 (RRID:Addgene_39980)] using the Lipofectamine 2000 reagent (Invitrogen) following the manufacturer's instructions. Four hours after transfection, the cells were incubated in fresh 10% FBS-supplemented DMEM and kept at 37°C for 48 hr. The transfected cells were washed in ice-cold phosphate-buffered saline (PBS, pH 7.4), and the cell cultures were incubated with 3-NP (100 µM) for 12 hr and later treated with 10 ng/ml BDNF followed by 10 ng/ml NT-4/5 (BDNF → NT-4/5) or 10 ng/ml NT-4/5 followed by 10 ng/ml BDNF (NT-4/5 → BDNF) for 30 min.

2.5 Electrophoresis of cell extracts and Western blotting

After treatment with neurotrophins, the cells were processed for Western blot analysis as previously described (Torres-Cruz et al., 2019). In brief, the cell cultures were washed twice with PBS, and the cells were removed and lysed in radioimmunoprecipitation assay (RIPA) buffer containing a cocktail of protease inhibitors. The lysate was then centrifuged (4,500 rpm for 10 min), and the supernatant was collected for protein determination by the mini-Bradford assay (Bio-Rad Laboratories, Inc.). The proteins were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto nitrocellulose membranes for immunoblot analysis. The membranes were blocked in 10% nonfat dried milk + 5% BSA in PBS with 0.1% Tween 20 (PBS-tw) overnight (4°C) and then incubated with the primary antibodies used in tissue slices (Cell Signaling Technology, Danvers, MA, USA) against p-TrkB (1:2,000), TrkB (1:20,000), and actin (1:1,000) diluted in PBS-tw + 10% nonfat dry milk + 2.5% BSA. The membranes were rinsed and incubated with HRP-conjugated secondary antibodies (1:20,000, horseradish peroxidase-conjugated goat anti-mouse IgG, and horseradish peroxidase-conjugated goat anti-rabbit IgG) for 1–2 hr (RT). The immunoblots from at least two independent experiments were analyzed and visualized as described above, and the data were normalized to control condition.

2.6 Statistical analysis

All data were analyzed with Sigma Plot 12.3 (Systat Software, Inc, San Jose, CA, USA, RRID:SCR_003210) using parametric or nonparametric tests as appropriate. The data from the electrophysiological experiments were normalized to control condition and analyzed with one-way ANOVA followed by post hoc analysis if the variance and distribution were normal or with nonparametric analysis of variance if the normality test and equal variances failed. Significance was set at p < 0.05 and the results are expressed as the mean ± SEM unless otherwise specified.

2.7 Reagents

All reagents were purchased from Sigma-Aldrich Co., LLC (St Louis, MO, USA), unless otherwise stated.

3.1 BDNF → NT-4/5 treatment

All electrophysiological recordings were normalized to control as the 100% to analyze neurotrophin modulation on the population synaptic spike.

Figure 1 depicts the effect of BDNF followed by NT-4/5 (BDNF → NT-4/5). BDNF increased the amplitude of the synaptic spike (Figure 1a–c); nevertheless, NT-4/5 administration significantly inhibited the effect of BDNF as expected from our previous study ( = 6.5, p = 0.039, one-way repeated measures ANOVA on ranks, Tukey test, post hoc, multiple comparisons, n = 4, Figure 1a–c). Analysis of the PPR (S2/S1) did not show differences among experimental conditions (Figure 1d).

The application of BDNF to cerebral slices from mice treated with 3-NP in vivo significantly increased the spike amplitude (Figure 1e). This increase was not in the same proportion as that obtained in control slices (Figure 1f vs. 1b). The application of NT-4/5 after BDNF (BDNF → NT-4/5) significantly reduced the population spike amplitude (Figure 1g), indicating that NT-4/5 exerts an inhibitory effect in early striatal degeneration produced by 3-NP treatment as well as it did in control tissue (F_2,26 = 29.928, p = <0.001; one-way repeated measures ANOVA, n = 9, Holm–Sidak multiple comparisons test, n = 9). Analysis of the PPR did not show differences among the treatments (Figure 1h).

Four experiments receiving BDNF → NT-4/5 did not show any modulatory effect on synaptic transmission. Moreover, in two experiments where BDNF did not show any modulatory effect on corticostriatal synapses in tissue from 3-NP-treated mice, administration of NT-4/5 increased the population spike amplitude (data not shown).

3.2 NT-4/5 → BDNF treatment

Figure 2 illustrates the effects of NT-4/5 → BDNF treatment in control and 3-NP-treated tissue. In slices from control mice, NT-4/5 significantly increased the amplitude of the population spike, an effect that was not inhibited by BDNF when BDNF followed NT-4/5 (F_2,17 = 204.29, p = 0.001, one-way repeated measures ANOVA, Holm–Sidak multiple comparisons test, n = 6, Figure 2a–c). The PPR analysis found significant differences between control and treatments (F_3,19 = 73.413, p < 0.001; one-way ANOVA, Holm–Sidak multiple comparisons test), indicating that the effect of NT-4/5 was presynaptically mediated in control tissue (Figure 2d).

NT-4/5 application in slices from 3-NP-treated mice produced a small but statistically significant increase in the amplitude of the population spike (Figure 2e–g). Moreover, administration of BDNF after NT-4/5 produced a small increase in the spike amplitude in comparison to NT-4/5 alone ( = 9.0, p = 0.011, Friedman repeated measures ANOVA on ranks, Tukey test post hoc multiple comparisons test, n = 6). Analysis of the PPR did not show significant differences among treatments (Figures 2h). There were two experiments where NT-4/5 → BDNF treatment did not induce changes in the synaptic spike amplitude.

To specifically investigate the mechanism by which the sequential application of neurotrophins produced synergistic or antagonistic effects on corticostriatal synapses in slices from mice with 3-NP-induced striatal degeneration, we evaluated whether 3-NP treatment modified the protein levels and phosphorylation of TrkB in the presence of BDNF → NT-4/5 or NT-4/5 → BDNF by means of Western blot analysis of lysates from striatal slices.

3.3 3-NP treatment in vivo modifies the expression of TrkB in the striatum

Analysis of the protein expression in lysates of striatal slices did not show differences in the expression of p-TrkB or TrkB-T1 isoform in non-treated tissue under neurotrophin treatment (Figure 3a). The analysis of the protein expression of TrkB receptors in striatal lysates of slices from mice treated with 3-NP showed a reduction in the expression of the TrkB-FL which was very weak under all treatments (Figure 3b, b1), in comparison to non-treated tissue (Figure 3a, a1). There was a statistical difference in p-TrkB/TrkB level under BDNF, BDNF → NT-4/5 and NT-4/5, and NT-4/5 → BDNF compared to control. Also there were statistical differences between BDNF and BDNF → NT-4/5, and BDNF and NT-4/5 treatments (H₄ = 12.523, p = 0.014, Kruskal–Wallis ANOVA, Student–Newman–Keuls, multiple comparisons, Figure 3b, b2). The decrease in the phosphorylation of TrkB was accompanied by an increase in the level of the truncated isoform of TrkB in BDNF → NT-4/5, NT-4/5 but not in NT-4/5 → BDNF (F_4,14 = 12.593; p = 0.001, one-way ANOVA, Holm–Sidak method, multiple comparisons, Figure 3b, b3). The reduction in phosphorylation under BDNF → NT-4/5 exposure and the increase in TrkB-T1 are consistent with the antagonistic effect observed in electrophysiological experiments. No increase was observed in TrkB-T1 expression level under NT-4/5 → BDNF, which agrees with the synergic effect on population spike amplitude observed in the physiology.

There is experimental evidence that suggests that in the absence of TrkB, BDNF binds to the p75 receptor and elicits effects contrary to those activated through TrkB stimulation (Woo et al., 2005). Thus, we evaluated the expression of the p75 receptor under our experimental conditions. We found no evident changes in p75 protein expression under neurotrophins application in non-3-NP-treated slices (Figure 4a). In slices from 3-NP-treated mice, there was an increase in the p75 receptor when tissue was exposed to BDNF → NT-4/5 or NT-4/5 → BDNF (Figure 4b). Figure 4c illustrates protein expression of p75 receptors in non-3-NP-treated and 3-NP-treated tissue under the administration of BDNF → NT-4/5 or NT-4/5 → BDNF. Note that p75 receptor expression only showed differences under NT-4/5 → BDNF when compared to non-3-NP-treated tissue slices (t₆ = −2.938; p = 0.026, two-tailed, t test).

3.4 TrkB-FL and p-TrkB are reduced and TrkB-T1 is upregulated in 3-NP-treated COS-7 cell exposure to BDNF → NT-4/5

To further investigate the participation of TrkB isoforms and their changes under 3-NP treatment with neurotrophin exposure, we utilized a cell line system to avoid contamination with a mixed population of neuronal and glial cells, to express TrkB-FL, TrkB-T1, or TrkB-FL + TrkB-T1 receptor isoforms in COS-7 cells. The cells were transfected and treated with neurotrophins and thereafter (see methods), cell lysates from two independent experiments were analyzed by PAGE and Western blotting.

BDNF → NT-4/5 treatment induced the expression of p-TrkB, TrkB-FL, and TrkB-T1, but TrkB-T1 expression is upregulated and TrkB-FL expression is downregulated when TrkB-FL and TrkB-T1 are coexpressed in COS-7 cells (Figure 5a, 4th column). NT-4/5 → BDNF treatment induced p-TrkB and TrkB-FL expression. When both receptor isoforms were coexpressed in COS-7 cells, NT-4/5 → BDNF treatment activated TrkB in less proportion, as observed by the weak p-TrkB band, and slight expression of TrkB-FL. Note that the phosphorylation level was different from that obtained with BDNF → NT-4/5 treatment (8th column). The phosphorylation level of TrkB was different between BDNF → NT-4/5 and NT-4/5 → BDNF but densitometry analysis did not show statistical significance (Figure 5a, a2). However, the truncated isoform expression was different between BDNF → NT-4/5 and NT-4/5 → BDNF treatments (Figure 5a, a3; T₄ = 26, p = 0.029, Mann–Whitney U test).

The results obtained after the treatment of cultured COS-7 cells with 3-NP are illustrated in Figure 5b. The cells were incubated for 12 hr with 3-NP and later treated with BDNF → NT-4/5 or NT-4/5 → BDNF.

3-NP-treated COS-7 cells showed a reduction in the expression of p-TrkB after BDNF → NT-4/5 or NT-4/5 → BDNF treatment compared with untreated-COS-7 cells (Figure 5a vs. 5b). The reduction in p-TrkB was a specific consequence of 3-NP treatment and was indicative of poor activation of TrkB in the presence of neurotrophins.

The expression of the TrkB-FL band was obtained in the presence of BDNF → NT-4/5 or NT-4/5 → BDNF when the TrkB-FL isoform was transfected alone (2nd and 6th columns, Figure 5b), but when both the FL and T1 isoforms were cotransfected, pale bands were observed (4th and 8th columns, Figure 5b). Interestingly, when BDNF → NT-4/5 treatment was administered, an upregulation of the truncated isoform was observed in cells cotransfected with both the FL and T1 isoforms (column 4, Figure 5b, b1) that coincided with reductions in p-TrkB and TrkB-FL, consistent with an antagonistic effect of NT-4/5. The upregulation of the truncated isoform was not present as much in cells subjected to the NT-4/5 → BDNF treatment when the TrkB-FL and TrkB-T1 isoforms were coexpressed, and did not present significant differences between treatments (Figure 5b, b3); instead, there was a slight increase in p-TrkB (8th column, 5B,b1). This result is consistent with the synergistic effect of BDNF in the NT-4/5 → BDNF treatment in the electrophysiology experiments.

4.1 Antagonistic versus synergistic effects of BDNF and NT-4/5 and NT-4/5 → BDNF

BDNF and NT-4/5 exposure produced an antagonistic effect in 3-NP-treated tissue, as the one observed in non-3-NP-treated tissue. Individually, each neurotrophin increased the spike amplitude; however, mitochondrial inhibition lessened the effect of BDNF and NT-4/5 on corticostriatal transmission. The decrease in the modulatory effect of these neurotrophins on corticostriatal synapses is due, in part, to reductions in TrkB expression, as demonstrated through Western blot experiments, and these results are in agreement with our previous studies (Espíndola et al., 2012; Mendoza et al., 2014). BDNF → NT-4/5 exposure antagonized the effects of BDNF in slices from 3-NP-treated mice, in the majority of the experiments, as reported in healthy tissue (Torres-Cruz et al., 2019).

With the COS-7 cell system, we demonstrated that BDNF → NT-4/5 treatment resulted in the downregulation of TrkB-FL and p-TrkB when the TrkB-FL and TrkB-T1 isoforms were coexpressed in COS-7 cells treated with 3-NP (Figure 5b). The antagonistic effect of NT-4/5 can be explained by the stimulation of TrkB-T1 which has been shown to antagonize TrkB-FL activity (Ardelt, Flaris, & Roth, 1994; Biffo, Offenhauser, Carter, & Barde, 1995, Eide et al., 1996; Gupta, You, Gupta, Klistorner, & Graham, 2013; Quarta et al., 2018).

NT-4/5 → BDNF treatment of slices from 3-NP-treated mice produced a synergistic effect on spike amplitude. We assume that when NT-4/5 is not competing with BDNF to stimulate TrkB, it triggers the cellular response responsible for increasing spike amplitude (we have shown that PI3K, MAPK, and PLC are responsible for neurotrophin modulation in corticostriatal transmission; Torres-Cruz et al., 2019). The additive effect of further exposure to BDNF results from BDNF activation of TrkB as observed in WB experiments. In addition, 3-NP treatment causes a morphological disconnection of the corticostriatal pathway and also 3-NP reduces BDNF levels (Mendoza et al., 2014); thus, it should permit NT-4/5 and BDNF to cooperate to stimulate TrkB-FL isoforms. The experiments with COS-7 cells indicate that NT-4/5 → BDNF did not upregulate TrkB-T1 but increased p-TrkB (column 8, 1st and 2nd row, Figure 5b), explaining the synergistic effect of BDNF.

The differential responses to BDNF → NT-4/5 and NT-4/5 → BDNF exposure show that these neurotrophins elicit specific responses by interacting with specific TrkB isoforms, which in turn activate multiple signaling pathways. In the context of striatal pathology more attention should be paid to TrkB isoforms, especially because the overexpression of TrkB-FL receptor improves neural survival in an HD model (Silva et al., 2015). Moreover, while the expression of the TrkB-FL receptor regulates gene expression related to plasticity (Koponen et al., 2004), the overexpression of TrkB.T1 inhibits synaptic plasticity (Li et al., 1988). Truncated TrkB.T1 receptor is a regulator of TrkB-FL signaling (Biffo et al., 1995). Further experiments are necessary to understand the physiological role of the truncated isoforms in the healthy and degenerated striatum, especially because this TrkB receptor isoform is highly expressed in the adult brain.

We don't know why NT-4/5 is more prone to stimulate TrkB-T1. The role of NT-4/5 in striatal physiology of HD has not been explored in detail. Our results suggest that NT-4/5 inhibits BDNF modulatory effects; this antagonistic effect has not been described previously in HD. Most of the studies indicate that BDNF and NT-4/5 exhibit similar effects; however, they can activate the same receptor but in different regions, which produces diverse physiological responses (Fan et al., 2000; Minichiello et al., 1998; Proenca, Song, & Lee, 2016).

We evaluated if p75 receptor had a role in the antagonistic or synergistic effect observed in our experiments. However, p75 did not change under neurotrophin treatments, although it was upregulated with the 3-NP treatment and significant differences were found under NT-4/5 → BDNF treatment between non-3-NP-treated tissue and 3-NP-treated tissue. Further experiments will be necessary to understand if the upregulation of p75 has other physiological roles. The increase in p75 has been associated with apoptosis in excitotoxicity (Hanbury et al., 2002; Nykjaer, Willnow, & Petersen, 2005; Underwood & Coulson, 2008), and we have documented its augmentation in our model of striatal degeneration (Espíndola et al., 2012).

4.2 Mitochondrial inhibition alters the expression of TrkB

In vivo inhibition of mitochondrial complex II with 3-NP produces oxidative stress, mitochondrial dysfunction (Alston, Mela, & Bright, 1977; Beal et al., 1993; Brouillet et al., 1999; Coles et al., 1979; Gu et al., 1996; Rodriguez et al., 2010), as well as structural modifications of corticostriatal synapses, which cause anatomical and physiological disturbances such as reductions in the number of dendritic spines on MSNs (Mendoza et al., 2014), in which glutamate receptors are densely located (Deshpande et al., 2006; Nishino et al., 1997). In addition, 3-NP treatment reduces TrkB receptors in the striatum (Espíndola et al., 2012; Mendoza et al., 2014). In this study, we demonstrated that mitochondrial inhibition alters the expression of TrkB isoforms, modifying the modulation of synaptic activity, which is relevant to understanding HD pathophysiology as well as other neurodegenerative disorders that present with mitochondria dysfunction.