The Calcium‐Binding Protein S100A10 (p11) Is Required for Normal Motor Performance by Regulating Vesicle Dynamics at Excitatory Synapses

ABSTRACT

Aim

Identifying interactors in sensorimotor processing and neurotransmission remains a current challenge for understanding neural information processing and brain function.

Methods

To evaluate the role of p11 in sensorimotor processing and excitatory synaptic neurotransmission, neuron‐specific lentivirus‐directed p11 silencing, small interfering RNA (siRNA_p11)‐induced p11 deletion, unitary extracellular recordings of hypoglossal motor neurons (HMNs), western blot, co‐immunoprecipitation, multiple immunolabeling, proximity ligation (PLA) assays, electron microscopy, and whole‐cell patch‐clamp recording of AMPA receptor–mediated excitatory postsynaptic currents in adult and/or neonatal rat HMNs were performed.

Results

p11 knockdown depressed baseline and chemoreceptor‐modulated inspiratory‐related activity in HMNs. Co‐immunoprecipitation and PLA assays indicated that p11 interacts with Munc13‐1, a presynaptic active zone (AZ) protein for vesicle priming, presumably at excitatory inputs in the hypoglossal nucleus. Interference with p11 resulted in Munc13‐1 downregulation, reduction in AZ length, and increased vesicle accumulation at excitatory boutons on HMNs, without affecting the number of docked vesicles at the AZ. p11 knockdown robustly reduced the synaptic strength of excitatory neurotransmission incoming to HMNs by affecting both the synchronous and asynchronous phases of neurotransmitter release. The decrease in synaptic strength was concurrent with a reduction in the size of the “functional” pool of readily releasable (RRP) vesicles and with the slowing down of the vesicle recruitment rate to replenish RRP.

Conclusion

p11 is proposed as a relevant mediator in the neurotransmitter release by regulating vesicle dynamics at central excitatory synapses. Here, p11 is highlighted as a multifaceted factor involved in neurotransmission and synaptic plasticity and, therefore, central for neural information processing.

1. Introduction

The protein p11 (also named S100A10, annexin II light chain, or calpactin I light chain) is a member of the S100 family of small, dimeric EF‐hand‐type Ca²⁺‐binding proteins that generally modulate cellular target proteins in response to intracellular Ca²⁺ signals. In contrast to all other S100 proteins, p11 is rendered Ca²⁺‐insensitive due to genetic mutations in its EF‐hand Ca²⁺‐binding loops, such as substitutions and deletions, which lock the protein into a permanently active state [1, 2]. This protein is widely distributed throughout the organism, suggesting that it may not be restricted to immune, cardiovascular, respiratory, digestive, excretory, and reproductive systems, but also impacts the functioning of the nervous system, both in health and in disease. Specifically, in the brain p11 is mainly expressed in neurons and neuroprotective reactive A2‐astrocytes [3, 4, 5]. In this context, it has been shown that p11 modulates the therapeutic actions of drugs used to treat several brain disorders, and that aberrant changes in its expression might be involved in the pathophysiology of several forms of cancer, psychiatric disorders, and neurodegenerative diseases. Of interest, p11, either as a tight dimer or forming heterotetramers with annexin A2 (AnxA2), regulates surface expression levels and function of various ion channels and G protein‐coupled receptors (GPCRs) [1]. p11 generally acts as an adaptor protein that regulates intracellular trafficking and plasma membrane expression of specific receptors (5‐HTs and mGluR5) and ion channels (CTFR, Nav1.8, ENaC, ASIC‐1, TASK1, and TRPV5/6 channels). Accordingly, p11/AnxA2 complex regulates the expression of TRPV5/6‐mediated Ca²⁺ currents in a variety of epithelial cells [6]. Besides, the altered interaction of p11 with 5‐HT, mGluR5, Nav1.8, or L‐type Ca²⁺ channels interferes with signaling and neural processing within nociceptive pathways and limbic circuits, underlying the development of depression and neuropathic pain behavior [7, 8, 9, 10]. The broad expression of p11 throughout the CNS [3], including structures such as cerebellum, thalamus, hypothalamus, hippocampus, amygdala, cerebral cortex, medulla, and spinal cord, leaves open the possibility that this protein plays a central role also in many other, still unexplored, pivotal processes underlying brain functioning. In this scenario, the expression pattern of this protein in motor nuclei, along with its causative role in motor neuron degeneration by regulating membrane insertion of TASK1 subunits, lies beneath a likely involvement of p11 in the management of motor output commands [11].

There are several lines of evidence, particularly those concerning the identity and function of the factors with which p11 has been shown to interact, suggesting that p11 may act as a relevant mediator in the dynamic process of synaptic communication, key for fine‐tuning of neuronal excitability, information processing in the brain, and hence also handing out motor commands. Firstly, Ahnak is a major binding partner and protein stabilizer of p11/AnxA2 in the brain, with which it forms a heterotetramer complex to recruit L‐type Ca²⁺ channels to the cell surface, thereby regulating L‐type‐specific Ca²⁺ influx into neurons [10]. It has been shown that Ahnak modulates excitatory neurotransmission in glutamatergic neurons in layer II/III of the prefrontal cortex, though a causal link between Ahnak‐dependent regulation of Ca²⁺ signaling and glutamate release has not been proven [12]. Consistent with the notion that the Ahnak/p11/AnxA2 complex exerts a regulatory role in neurotransmitter release, other studies have found that cell‐specific modulation of p11 levels rescues stress‐induced glutamatergic synaptic deficits [13] or elevates inhibitory neurotransmission [14]. Secondly, p11 also interacts with several types of GPCRs, modulating their trafficking, cellular localization, and function. For instance, p11 promotes cell surface expression of specific subtypes of metabotropic serotonin (5‐HT_1B, 5‐HT_1D, and 5‐HT₄) and glutamate receptors (mGluR5), all of which are mainly found at presynaptic sites, operating as autoreceptors or heteroreceptors that modulate neurotransmitter release [1]. Therefore, evidence supports that p11 participates in the control of neurotransmission by at least a subsidiary influence on neurotransmitter release. However, the mechanisms by which p11 interacts with neurotransmitter release are poorly disclosed.

Finally, evidence from studies on the involvement of p11 in the exocytosis of secretory vesicles/granules from neuroendocrine and other non‐neuronal cell types indirectly suggests that p11 is a reasonable member of the presynaptic machinery for vesicle dynamics, also at central synapses within the brain. Specifically, in adrenergic chromaffin cells, p11 mediates AnxA2 translocation to SNARE complexes to form part of the Ca²⁺ sensitive machinery for exocytosis [15, 16]. p11 can indeed physically interact with the SNARE protein VAMP2 at the plasma membrane of these cells, facilitating the formation of SNARE trans complexes between plasma and granule membranes during exocytosis. Therefore, p11 appears particularly well suited as a functional partner for SNAREs in the molecular machinery underlying exocytosis [16]. Interestingly, p11/AnxA2 also regulates von‐Willebrand factor secretion in endothelial cells, where p11 interacts with Munc13‐4 to tether Weibel‐Palade bodies to the fusion site and to promote their exocytosis [17, 18]. Given the homology between Munc13‐4 and Munc13‐1, being the latter a presynaptic active zone (AZ) protein required for vesicle priming as a general and essential step common to multiple fast neurotransmitter systems in the CNS [19, 20, 21], it is plausible to speculate that p11 may also operate as a molecular mediator in the control of vesicle dynamics at central synapses.

Here, we provide solid findings highlighting a leading role of p11 in managing motor output commands, at least in part, by regulating vesicle dynamics, neurotransmitter release, and thereby synaptic strength at excitatory synapses on motor neurons.

2. Results

2.1. Neuron‐Specific p11 Knockdown Reduces Inspiratory‐Related Synaptic Drive in Hypoglossal Motor Neurons (HMNs)

In this study, we used the hypoglossal nucleus (HN) as an experimental model to investigate the involvement of p11 in translating afferent synaptic drive into appropriate discharge patterns in motor neurons, which in turn determine the final stage for coding motor output commands. This model is helpful for such purpose because (i) the HN is a cranial motor nucleus easily accessible for the recording of unitary discharge activity of HMNs as a direct method for monitoring motor output commands; (ii) most HMNs exhibit a characteristic, and easily identifiable and assessable, discharge activity that consists of rhythmic bursts of action potentials (APs) synchronized with the inspiratory phase of breathing [22, 23]; (iii) this rhythmic, inspiratory‐related discharge activity is driven by excitatory glutamatergic afferences from the medulla and pons, mainly acting via postsynaptic AMPA receptors (AMPARs) [24, 25, 26, 27]; (iv) the respiratory‐related bursting discharge displayed by HMNs persists following decerebration; (v) in in vivo decerebrated preparations, output respiratory activity can be chemosensorially modulated by changes in arterial blood CO₂ [22, 23]; and (vi) profuse expression of p11 in the HN accounts for its pivotal contribution in setting intrinsic excitability of HMNs [11], as well as puts forward a probable, but yet unsolved, contribution of p11 in the presynaptic machinery for neurotransmission and hence for tuning motor output activity.

In an initial set of experiments performed in parallel with those previously reported [11], we first aimed to determine whether the interference with neuronal p11 expression exerts any effect on the synaptically driven rhythmic discharge of HMNs. For that purpose, neuron‐specific lentiviral constructions (LVVs; Figure 1A), which direct the expression of a small hairpin RNA against mRNA_p11 (LVV‐shRNA_p11) or against bacterial mRNA_lacz (LVV‐shRNA_lacz) as a control, were used. As reported before (see Figure 4H in our previous report [11]), a single intramedullary microinjection of LVV‐shRNA_p11, led to a drastic reduction of p11 expression (Figure 1B, bottom) in both HNs up to 24 days after viral administration.

FIGURE 1.

Cell‐type specific downregulation of p11 alters HMN response to baseline and chemoreceptor‐modulated inspiratory drive in adult rats. (A) Top, schematic diagram depicting microinjection of the indicated LVVs in the midline between both HNs. Bottom, images illustrating neuronal‐specificity of a LVV encoding the eGFP reporter gene driven under the same control system as that used by LVV‐shRNA_lacz and LVV‐shRNA_p11. Note that a large pool of neurons in the HN, mainly HMNs, have been transduced. GFAP‐immunolabeling in the HN reveals the absence of transduced astrocytes. Scale bars: Left panel, 200 μm; right panel, 50 μm. cc, central canal. (B) Experimental design for the recording (Rec) of the unitary discharge activity of HMNs in decerebrated rats. HMNs were identified by their antidromic activation from the stimulating electrode (St) implanted in the XIIth nerve. Right, traces represent the extracellularly recorded inspiration‐related spike discharge (top) and the instantaneous firing rate (middle, FR, in spikes/s; bin width = 25 μs) for a control HMN at baseline (ET_CO2 = 4.8%–5.2%). Mean FR (mFR) per burst and BR (in burst/min) were analyzed in relation to simultaneous ET_CO2 measurements. Bottom, effect of microinjection of LVVs on P11 protein expression in HNs. Neocortex was taken as a control for systemic transduction. α‐tub was the internal control for Western blotting. These results were obtained from the same tissue samples as those used in a previously published study (see Figure 4H in our previous report [11]) and are presented here to illustrate the effectiveness of LVV‐shRNA_p11 to induce downregulation of p11 in the HN. (C) Box‐plots and mean ± SEM of mFR and BR for HMNs recorded from animals treated with vehicle, LVV‐shRNA_lacz, and LVV‐shRNA_p11. (D) Instantaneous FR of a control HMN showing modulation of inspiration‐related discharge activity at different ET_CO2 levels. (E) Plots showing the relationships between mFR per burst (left) or BR (right) and ET_CO2 for the same HMN after grouping and averaging data at 0.2% intervals of ET_CO2. The slopes of the regression lines represent the neuronal sensitivity or gain to ET_CO2 changes (S_mFR, in spikes/s/%ET_CO2; S_BR, in bursts/min/%ET_CO2). (F) Box‐plots and mean ± SEM of S_mFR (left) and S_BR (right) for HMNs recorded under the indicated conditions. The number of HMNs analyzed per condition is indicated in parentheses in C and F. *p < 0.05, **p < 0.01; by one‐way ANOVA followed by post hoc Holm–Sidak method (C) or one‐way ANOVA on Ranks with post hoc Dunn's method (F) relative to vehicle and LVV‐shRNA_lacz‐injected groups.

FIGURE 4.

siRNA_p11 reduces vGLUT2 and Munc13‐1 expression in the HN of adult rats. (A) siRNA_p11 effectively reduced vGLUT2 and Munc13‐1 in the brainstem 3 days after administration relative to the control oligonucleotide. β‐Actin was taken as the internal control for WB. (B, C) Confocal images showing immunostaining for p11 (red), vGLUT2 (green), or Munc13‐1 (blue) in the HN of animals receiving the indicated oligonucleotides 3 days before. Higher magnification images of the boxed areas in merged panels of the right column are displayed in C. (D) From left to right, box‐plots and means ± SEM of the area covered by vGLUT2 or Munc13‐1 immunolabeled profiles relative to the total field (in percent), linear (per 100 μm) and surface densities (per 100 μm²) of these synaptic structures around HMN perikarya and in neuropil, respectively, and fraction of p11‐ir area which overlaps with vGLUT2‐ or Munc13‐1‐immunoreactivity under the indicated treatments. (E, F) Double confocal immunofluorescence for the presynaptic markers vGLUT2 and Munc13‐1 after siRNA_p11 or cRNA administration. Boxed areas are displayed at higher magnification in the right panels. Box‐plots and means ± SEM for the fractions of Munc13‐1‐ir patches overlapping with vGLUT2‐ir puncta (top) and vice versa (bottom) under the indicated conditions. Scale bars: 20 μm. The number of animals, fields, or neurons analyzed per condition is indicated in parentheses. *p < 0.05, ***p < 0.001; by unpaired Student's t‐test or non‐parametric Mann–Whitney U test.

Motor output activity was first evaluated by performing single‐unit extracellular recordings of HMNs under baseline conditions (end‐tidal CO₂ (ET_CO2) = 4.8%–5.2%), from 5 to 7 days after LVVs or vehicle microinjection. Burst parameters were measured and averaged using a minimum of 30 bursts per HMN. In LVV‐shRNA_lacz‐treated animals, mean firing rate (mFR: 59.7 ± 3.7 sp/s) and inspiratory burst rate (BR: 48.6 ± 1.6 bursts/min) were like those observed in HMNs from vehicle‐receiving rats (Figure 1B,C; Table S1). In contrast, mFR in basal conditions was significantly lower after injection of LVV‐shRNA_p11 (46.7 ± 2.7 sp/s). Interestingly, BR, a parameter indicative of the rhythmogenic activity produced in, and transmitted from, premotor respiratory centers, was not altered after virus application (45.1 ± 1.8 bursts/min) (Figure 1C; Table S1). Therefore, neuronal p11 knockdown is accompanied by specific changes in HMN basal activity, without apparently affecting the medullary‐pontine neuronal network involved in the rhythmogenesis of breathing. Although the attenuated discharge activity displayed by HMNs under p11 knockdown could be, at least to some extent, due to the increased surface expression of TASK1 “leak” potassium channels, and concomitantly to the expected reduction in intrinsic membrane excitability of HMNs [11], it is also plausible that alterations in synaptic strength may contribute to this effect.

To further discern between both effects, we analyzed the relationship that links the modulation of HMN firing activity to varying levels of afferent synaptic drive produced by changes in ET_CO2. Modifying ET_CO2 levels causes variations in chemoreceptor drive of afferent signals to the respiratory neuronal network of the brainstem which in turn, via its efferences to the HN, produces changes in both the mFR of the burst and the BR of HMNs (Figure 1D–F). Thus, discharge activity of HMNs increased when ET_CO2 rose and vice versa (Figure 1D). Relationships obtained between mFR per burst or BR and ET_CO2 were linear and positively correlated across the range of ET_CO2 tested (Figure 1E). The slope of the regression line represented the sensitivity (S) or gain of mFR or BR to ET_CO2 (S_mFR and S_BR, respectively) (Figure 1E). Specifically, the first of these parameters (S_mFR) can therefore be considered an indicator of the motor neuron sensitivity to changes in the afferent drive, which in turn, as we have shown previously [23, 28] is primarily determined by the strength of its synaptic inputs.

After injection of LVV‐shRNA_lacz, S_mFR and S_BR were similar to the vehicle condition (Figure 1E,F; Table S1). However, LVV‐shRNA_p11‐induced p11 downexpression was accompanied by a dramatic decrease in HMN sensitivity to the chemoreceptor‐modulated inspiratory drive, affecting S_mFR but not S_BR (Figure 1E,F; Table S1). The lack of effects on S_BR is crucial for the validation of these results because this parameter is indicative of the integrity and functional state of chemosensors and rhythm‐related premotor structures, indicating that the responsiveness of the respiratory centers was similarly preserved under all conditions. These results indicate that virally‐mediated neuronal p11 knockdown in the HN effectively induces impairment of the inspiratory‐related motor output command, and also suggests that this effect could be attributable, at least in part, to a decreased efficacy of glutamatergic excitatory transmission onto HMNs.

2.2. Anatomical Basis for a Presynaptic Role of p11 at Excitatory Synapses

We next performed confocal analysis of double immunolabeled HN from adult rats to investigate the possibility that p11 could have a direct relationship with synaptic vesicle proteins with special focus on excitatory presynaptic terminals (Synaptophysin, Syn; vesicular glutamate transporter 2, vGLUT2; Figure 2A,B). Analysis in neuropil, that is, excluding somata, showed p11‐immunoreactive (ir) puncta, patches, and fiber‐like structures throughout the HN. Double immunolabeling with Syn, a generic marker of vesicle pools, or vGLUT2, specifically expressed by vesicles at glutamatergic excitatory inputs, confirmed an evident proximity between p11 and vesicle pools at excitatory synapses in the HN. Thus, quantitative analysis revealed that a high proportion of Syn‐ir (88.6% ± 0.6%, n = 29 slices from four rats) and vGLUT2‐ir (81.09% ± 1.10%; n = 17 slices from four rats) puncta were adjacent or partially co‐localized with p11‐ir clusters (Figure 2A,B). These results strategically position p11 as a possible partner in managing vesicle dynamics at excitatory synapses.

FIGURE 2.

Evidence for a close proximity relationship between p11 and Munc13‐1 at excitatory synaptic terminals in the HN of the adult rat. (A, B) Confocal images obtained from the HN co‐immunolabeled for p11 (red) and the synaptic vesicle markers synaptophysin (Syn, green in A) or vesicular glutamate transporter 2 (vGLUT2, green in B). Boxed areas are displayed at higher magnification in the bottom panels. (C) Co‐immunoprecipitation assays showing that p11 form part of a complex together with Munc13‐1. Brain lysates from adult rats were immunoprecipitated (IP) with a non‐specific isotype antibody (IgG), as a negative control, or with a specific antibody directed against p11 linked to protein G‐magnetic beads. Total lysates before (LB) and after (LA) immunoprecipitation and elution of the bound fraction (BE) samples were resolved by SDS‐PAGE and analyzed by western blotting (WB) with the indicated antibodies. WBs for two rats are shown. (D) Confocal images of the HN immunolabeled for p11 (red) and the AZ presynaptic protein Munc13‐1 (green). Boxed areas are displayed at higher magnification in the right panels. (E) Duolink in situ proximity ligation assay (PLA) was performed on brainstem tissue sections at the level of the HN, by combining rabbit anti‐Munc13‐1 and goat anti‐p11 as primary antibodies to detect two protein complexes (top). Omission of the primary antibodies in the PLA assay served as a negative control (right). HN was visualized using confocal fluorescence microscopy. Positive PLA reactions are illustrated in red (arrows). Nuclei were counterstained with DAPI (blue). The arrow on the left panel points to a cell presumably surrounded by complexes containing p11‐Munc13‐1. (F) Left, multiple immunolabeling confocal images from the HN showing the proximity and partial co‐localization (arrowheads) of positive PLA reactions (p11‐Munc13‐1, red) with vGLUT2‐positive patches (green). Nuclei were counterstained with DAPI (blue). Right, 3‐D reconstructions of a positive PLA reaction with vGLUT2‐positive patches. The merged XY‐, XZ‐, and YZ‐planes are shown at the bottom. Scale bars: A, B, D, E, 20 μm; F, left, 5 μm; right, 2 μm.

To contribute further support to this hypothesis, we looked for a link of p11 with the presynaptic molecular machinery involved in neurotransmitter release. The members of the Munc13 protein family have been proposed to function as tethering/priming factors in Ca²⁺‐dependent exocytotic events [20]. Interestingly, p11 interacts with Munc13‐4 to tether Weibel–Palade bodies to the fusion site and to promote their exocytosis in endothelial cells [18]. Munc13‐1, the most abundant isoform in the brain, is selectively localized within the presynaptic AZs (but not on synaptic vesicles), where it plays an essential role in vesicle priming at glutamatergic synapses [29, 30, 31]. Given the structural and functional similarity between Munc13‐4 and Munc13‐1 [21], we investigated a reasonable interaction between p11 and Munc13‐1 as a feasible mechanism by which p11 could exert its influence on the release of neurotransmitters from excitatory synaptic terminals. To evaluate this hypothesis, we performed a combined analysis of co‐immunoprecipitation, confocal microscopy, and proximity ligation assays (PLA) in sections or homogenates obtained from adult rat brainstem tissue. First, as shown in Figure 2C, co‐immunoprecipitation assays consistently supported that p11 resides in a complex together with Munc13‐1 (Figure 2C). Confocal analysis further indicated that most Munc13‐1‐ir patches (85.88% ± 1.21%; n = 10 slices from 3 rats) were adjacent or partially co‐localized with p11‐ir structures in the neuropil of the HN (Figure 2D). Analysis of in situ interaction between p11 and Munc13‐1 by using specific antibodies for each of these proteins in PLA revealed likewise positive interaction signals in the HN (Figure 2E; PLA no primary antibodies: 0.07 ± 0.02 patches/100 μm², n = 11 slices from 3 rats; PLA p11‐Munc13‐1: 1.34 ± 0.27 patches/100 μm², n = 17 slices from 3 rats; p < 0.001, Mann–Whitney U test). Finally, co‐immunolabeling for vGLUT2 revealed that most of PLA reactions (57.0% ± 6.8%; n = 406 analyzed PLA reactions from 7 slices) were adjacent or partially co‐localized with pools of excitatory synaptic vesicles (Figure 2F, left). 3D confocal images illustrate some examples of straight relationship between p11‐Munc13‐1 PLA reactions and glutamatergic vesicle pools (Figure 2F, right).

Taken together, these outcomes strongly suggest an interaction between p11 and Munc13‐1, a presynaptic AZ protein essential for vesicle priming, thereby strengthening a pivotal role of p11 in the presynaptic machinery for neurotransmitter release at least at excitatory synapses.

2.3. Interfering With p11 Reduces the Expression of Other Presynaptic Proteins

The proximity relationship of p11 with presynaptic proteins like Syn, vGLUT2, or Munc13‐1 raises the possibility that this multifunctional factor may play a role in regulating their expression, trafficking, targeting, cellular localization, or function. Hence, a series of experiments was carried out subsequently to investigate whether p11 knockdown caused alterations in the expression of the aforementioned presynaptic proteins in the neuropil of the HN. For that purpose, single injections of a small interfering RNA directed against mRNA_p11 (siRNA_p11; 5 μg/5 μL) were administered into the fourth ventricle. This treatment efficiently reduced p11 expression in the brainstem and in the HN relative to administration of a non‐targeting siRNA (cRNA; 5 μg/5 μL) (Figure 3A–F; Table S2). This modified siRNA has been reported to induce a rapid and targeted neuron‐selective knockdown of the selected genes in the CNS, with absence of off‐target effects, toxicity, and recruitment of immune responses after a single intracerebroventricular injection in the adult rat [32]. This interfering tool has been broadly and successfully used in our lab to reduce specifically the expression of proteins of interest for extended periods (up to 2 weeks) after a single injection in the fourth ventricle [11, 33, 34, 35, 36].

FIGURE 3.

siRNA_p11 alters p11 expression and synaptic arrangement in the HN of adult rats. (A) Schematic diagram illustrating the experimental procedure for oligonucleotide microinjections. A solution (5 μL) containing a control non‐interfering siRNA (cRNA, 5 μg) or a small interfering RNA directed against mRNA_p11 (siRNA_p11, 5 μg) was administered into the fourth ventricle using a Hamilton syringe. (B, C) Microinjection of siRNA_p11 effectively reduced mRNA_p11 (B) and p11 protein (C) at the brainstem 3 days after its administration. Gapdh and β‐Actin were internal controls for qRT‐PCR and WB, respectively. (D–F) Confocal images showing p11 immunoreactivity (p11‐ir, red) in the HN from animals receiving the indicated oligonucleotides 3 days before (D). Box‐plots and means ± SEM of the area covered by p11‐ir relative to either the total field (in percent, E) or the neuropil area, that is, excluding perikarya, (in percent, F) are shown for the indicated conditions. (G) High‐magnification confocal images showing p11‐ir (red) and synaptophysin‐ir structures (Syn, green) in the HN under the indicated treatments. Boxed areas are displayed at higher magnification in the right panels. (H–K) Box‐plots and means ± SEM of the area covered by Syn‐ir patches relative to the total field (H), linear (per 100 μm) and surface densities (per 100 μm²) of these same synaptic structures around HMN perikarya (I) and in neuropil, respectively (J), and fraction of p11‐ir structures overlapping with Syn‐ir puncta (K) under each experimental condition. Scale bars: (D) 25 μm; (G) 20 μm. The number of animals (B, C), sections (E, F, H, J, K), or neurons (I) analyzed per condition is indicated in parentheses. **p < 0.01, ***p < 0.001; by unpaired Student's t‐test.

siRNA_p11 administration led to significant reductions in mRNA_p11 (−23.0% ± 3.0%) and p11 protein expression (−46.0% ± 11.0%) in the brainstem, and also in p11 immunolabeling both in the whole HN (−21.4% ± 3.3%) and in the neuropil (−19.0% ± 4.1%), i.e., excluding somata (Figure 3A–F; Table S2). Strikingly, p11 downregulation was accompanied by a drastic reduction of Syn‐ir in the HN as compared with the cRNA condition (Figure 3G–J; Table S3). In particular, siRNA_p11 reduced total area covered by Syn‐ir puncta (−35.7% ± 2.2%), as well as linear and surface densities of Syn‐positive patches close to neuron perikarya (−9.4% ± 3.3%) and in the neuropil (−24.2% ± 5.5%), respectively (Figure 3G–J; Table S3). Since the HN consists of 90% motor neurons and 10% interneurons [37], we can assume that the data obtained for neuron perikarya are mostly attributable to HMNs. Furthermore, the vesicle‐immunolabeled pools remaining after siRNA_p11 treatment were smaller (−11.7% ± 0.3%) and displayed lower mean intensity of fluorescence (−16.4% ± 0.1%) compared with the control condition (Table S3). Finally, while the Syn‐ir area overlapping with p11‐ir structures remained unchanged after treatment with the interfering oligonucleotide (39.0% ± 2.2%), the p11‐ir area overlapping with Syn‐ir was significantly reduced (−26.5% ± 2.9%) (Figure 3K; Table S3).

Knockdown of p11 was also accompanied by a marked reduction in the protein level of vGLUT2 (−39.0% ± 10.0%) and Munc13‐1 (−51.0% ± 9.0%) in the brainstem, and in the immunofluorescence for each of these markers in the HN (Figure 4A–C; Table S3). Covered area (vGLUT2: −27.9% ± 8.9%; Munc13‐1: −75.2% ± 4.6%), linear density close to neuron somata (vGLUT2: −29.4% ± 3.3%; Munc13‐1: −58.7% ± 4.8%), and surface density in the neuropil (vGLUT2: −32.3% ± 6.3%; Munc13‐1: −68.5% ± 6.0%) of puncta ir for each of these presynaptic proteins was severely affected by p11 loss (Figure 4B–D; Table S3). Furthermore, our analysis reveals that while the area of Munc13‐1‐ir overlapping with vGLUT2‐ir structures in the HN neuropil was unaltered by siRNA_p11, the area corresponding to vGLUT2‐positive vesicle pools overlapping with Munc13‐1‐ir patches was significantly reduced (Figure 4E,F), indicating a more pronounced effect on Munc13‐1 than on the glutamate vesicle transporter. Finally, the finding that siRNA_p11 reduced immunoreactivity for Syn, vGLUT2, and Munc13‐1 also in primary cultures of spinal motor neurons from mouse embryos (Figures S1–S3) points to p11 as a regulator of the expression of these presynaptic proteins throughout a neuron‐specific mechanism of action.

These outcomes indicate that p11 participates to some extent in the expression of presynaptic proteins involved in vesicle priming and neurotransmitter transport into vesicles, at least at excitatory synapses. Consequently, a reduction in p11 levels would be expected to impair synaptic communication.

2.4. p11 Knockdown Impacts on AZ and Vesicle Pools at Excitatory Boutons

Changes in synaptic arrangement that occurred after p11 knockdown could involve disorganization of the presynaptic machinery engaged to manage vesicle dynamics and neurotransmitter uptake, and/or even the loss of excitatory boutons. To gain insight into the nature of these synaptic alterations, we performed an ultrastructural analysis of excitatory synapses on HMNs from oligonucleotide‐treated adult rats using conventional transmission electron microscopy. HMNs were easily distinguishable from interneurons because the latter exhibit a noticeable invaginated nucleus [38, 39]. In our preparations, synaptic contacts did not show pre‐ and/or postsynaptic densities as thick as those described originally in the cerebral cortex [40]. Therefore, classification of synaptic boutons was done using vesicle morphology as the main criterion, assisted by the symmetric or asymmetric appearance of synaptic densities when these were evident [39]. Excitatory boutons were characterized by containing spherical, S‐type vesicles [39, 40, 41]. We analyzed several ultrastructural parameters of S‐type boutons attached to HMNs as indicators of their functional competence. In the control/untreated condition, the number and frequency of S‐type boutons attached to HMNs were similar (5.1 ± 0.3 boutons/HMN; 6.0 ± 0.4 boutons/100 μm; Table S4) to those previously reported [41]. Unexpectedly, the number (8.7 ± 0.5 boutons/HMN) and linear density (9.9 ± 0.5 boutons/100 μm) of excitatory boutons on HMNs were higher in siRNA_p11‐treated than in control (untreated, cRNA) cohorts (Figure 5A,B; Table S4). In accordance with a dismantling of the presynaptic machinery for neurotransmitter release, the length of the AZ was significantly reduced under p11 knockdown (relative to untreated: −16.6% ± 0.9%; relative to cRNA: −19.6% ± 0.9%) (Figure 5C,D; Table S4). Although the number of vesicles docked at the AZ was unchanged after siRNA_p11, the concomitant reduction in AZ length resulted in a rise of their linear density (Figure 5C,E; Table S4). Strikingly, the number of vesicles near the docked pool (relative to untreated: +38.9% ± 4.4%; relative to cRNA: +44.9% ± 4.6%) and the total number of vesicles within the terminal (relative to untreated: +26.5% ± 3.9%; relative to cRNA: +15.5% ± 3.6%) were higher in S‐type boutons from siRNA_p11‐treated animals than in control conditions (Figure 5F,G; Table S4). These results suggest that siRNA‐induced deletion of p11 disrupts the presynaptic machinery for neurotransmitter release, impairing vesicle exocytosis and leading to an imbalance of exocytosis/reformation ratio of vesicles at excitatory terminals.

FIGURE 5.

siRNA_p11 impacts on the presynaptic machinery for neurotransmission in excitatory boutons attached to HMNs in the adult rat. (A) Electron microscopy photomicrographs showing two HMNs obtained from animals that received treatment with the indicated oligonucleotides 3 days before perfusion. S‐type boutons are shaded in green. (B) Box‐plots and means ± SEM of the number of spherical synaptic vesicles (SVs)‐containing, S‐type, boutons per HMN (left) and per 100 μm of somatic membrane (linear density, right) for the untreated, cRNA, and siRNA_p11 conditions. (C) High‐magnification electron microscopy images illustrating two S‐type boutons contacting HMNs from animals under each treatment. Arrows define active zones (AZs) length, and the boxed areas delimit the pool of SVs near docked ones. Scale bars: (A) 5 μm; (C) 0.5 μm. (D–G) Box‐plots and means ± SEM of the a.z. length (D), number of docked SVs per a.z. (E, left) and per μm of a.z. (E, right), number of SVs in the pool near docked SVs (F), and number of SVs in the total pool (G) for the indicated conditions. The number of HMNs (B) and S‐type boutons (D–G) analyzed per condition is indicated in parentheses. #, *, § indicate comparisons relative to untreated, untreated/cRNA, or cRNA conditions, respectively. Two or three symbols indicate p < 0.01 or p < 0.001, respectively, by one‐way ANOVA post hoc Holm–Sidak method.

2.5. p11 Deletion Depresses the Strength of Excitatory Synaptic Inputs in HMNs by Operating Presynaptically

Functional, confocal, and electron microscopy studies at adulthood seem compatible with a functional disruption of neurotransmission at excitatory synapses on HMNs caused by p11 depletion, which could be underpinned by changes specifically affecting the presynaptic machinery for vesicle exocytosis. To explore this hypothesis, we conducted whole‐cell patch‐clamp recordings of spontaneously occurring and electrically evoked excitatory postsynaptic currents (EPSCs) in neonatal (P7‐P9) HMNs. As a preliminary step, we tested whether intracerebroventricular administration of siRNA_p11 in rat pups had similar effects to those observed in adults on the expression of other presynaptic proteins. The results show that a single injection of siRNA_p11 (2 μg/2 μL) into the fourth ventricle of P5 pups strongly reduced mRNA_p11 (−46.0% ± 2.0%) and p11 protein (−88.0% ± 2.0%) levels in the brainstem and p11 immunolabeling (−32.0% ± 4.0%) in the HN relative to cRNA administration (Figure 6A–C; Table S5). As expected, siRNA_p11‐induced p11 downregulation also provoked a drastic decrease of vGLUT2 and Munc13‐1 immunostaining in the HN as compared with the cRNA condition (Figure 6D–I; Table S6). Specifically, siRNA_p11 reduced total area covered by ir puncta (vGLUT2: −57.6% ± 9.8%; Munc13‐1: −83.8% ± 2.7%), linear density of positive patches near neuronal perikarya (vGLUT2: −60.7% ± 5.2%; Munc13‐1: −61.0% ± 4.9%), and density of synaptic markers in the neuropil (vGLUT2: −57.4% ± 7.4%; Munc13‐1: −49.1% ± 7.7%) (Figure 6D–I; Table S6). Furthermore, the size and mean intensity of the remaining patches were lower in siRNA_p11 than in cRNA‐treated groups (Table S6). Thus, the effects of p11 knockdown on the expression of presynaptic markers in the HN were similar in neonates and adults.

FIGURE 6.

siRNA_p11‐mediated p11 knockdown reduces the expression of vGLUT2 and Munc13‐1 in the HN of rat pups. (A) Schematic of timing for cRNA/siRNA_p11 administration and performance of the different experimental approaches. (B, C) Microinjection of siRNA_p11 into the fourth ventricle of neonatal rats (P5) effectively reduced mRNA_p11, p11 protein expression levels (B) in the brainstem, and p11 immunolabeling in the HN (C) 2 days after oligonucleotide administration. Gapdh and β‐Tubulin were internal controls for qRT‐PCR and WB, respectively. (D, E) Confocal images showing immunostaining for p11 (red), vGLUT2 (green), or Munc13‐1 (blue) in the HN of animals receiving the indicated oligonucleotides 3 days before. Higher magnification images of the boxed areas in merged panels of the right column are displayed in E. (F–I) Plots of the area covered by immunolabeling (F) for vGLUT2 and Munc13‐1 ir patches relative to the total field (in percent), their linear (per 100 μm) and surface densities (per 100 μm²) around HMN perikarya (G) and in neuropil (H), and fraction of p11‐ir area which overlaps with vGLUT2‐ or Munc13‐1‐immunoreactivity (I) under the indicated treatments. Scale bars: C, 100 μm; D, 20 μm. The number of animals (B), slices (C, F, H, I), and HMNs (G) analyzed per condition is indicated in parentheses. **p < 0.01, ***p < 0.001; by non‐parametric Mann–Whitney U test (B) or unpaired Student's t‐test (C, F‐I).

Subsequently, we examined the effects of siRNA_p11–induced p11 deletion on excitatory glutamatergic neurotransmission, specifically focusing our functional analysis on its AMPAergic component. As previously reported [42, 43], electrical stimulation of the ventrolateral reticular formation (VLRF) evokes monosynaptic EPSCs in HMNs in in vitro brainstem slice preparations from neonatal rats (Figure 7A). Both the amplitude and charge transfer of the AMPAergic component of these responses (eEPSCs_AMPA) were markedly attenuated (amplitude: −57.9% ± 5.3%; charge transfer: −61.7% ± 6.9%) after siRNA_p11 microinjection (Figure 7A; Table S7). Thus, these results indicate that endogenous p11 is required to maintain the strength of AMPAR‐mediated synaptic transmission in the HN.

FIGURE 7.

siRNA_p11‐associated impairment of AMPAergic neurotransmission involves presynaptic mechanisms. (A) Top, left. Schematic diagram of the in vitro experimental model used to analyze the effects of oligonucleotide administration on AMPAergic synaptic transmission in HMNs. Whole‐cell patch‐clamp recordings (Rec) were obtained from somata of large HN neurons in neonatal brainstem slices (P7‐P9). EPSCs were evoked by electrical stimulation (Stim) of the ventrolateral reticular formation (VLRF) in the presence of 50 μM APV, 10 μM bicuculline, 1 μM strychnine, and 30 μM D‐tubocurarine to isolate their AMPAergic component (eEPSC_AMPA). Top, right. Representative examples of pharmacologically isolated eEPSCs_AMPA from HMNs recorded 3 days after cRNA (black) or siRNA_p11 (red) administration. Bottom, summary data showing the changes in peak amplitude (left) and charge transfer (right) of the eEPSC_AMPA under siRNA_p11 relative to the control condition. (B) Current traces of sEPSCs_AMPA recorded from two HMNs after cRNA (black) or siRNA_p11 (red) administration. (C) Data values of sEPSC_AMPA amplitude for the siRNA_p11‐treated group of HMNs compared with their respective control group. (D) Normalized cumulative probability distributions of intervals between consecutive sEPSCs_AMPA (inter‐event interval) measured for each indicated condition. Bin width, 2 ms. (E) Average frequency of sEPSCs_AMPA occurrence for each treatment group. (F) Top, examples of eEPSCs_AMPA recorded in response to paired‐pulse stimulation of VLRF axons in two HMNs under cRNA (black) or siRNA_p11 (red) treatments. Stimulus intervals were 50 ms (@20 Hz, left) and 16.7 ms (@60 Hz, right). The siRNA_p11‐related recordings are scaled to the peak of the first eEPSCs_AMPA obtained in the cRNA condition. Bottom, comparison of paired‐pulse ratio (PPR) obtained from the amplitude of the first and second eEPSCs_AMPA by the formula eEPSCs_AMPA2/eEPSCs_AMPA1 under the indicated stimulation frequencies and treatments. (G) Superimposed average records illustrating succession of eEPSCs_AMPA in HMNs subjected to the indicated treatments, evoked by trains of 20 stimuli at 20 Hz (top) or 40 stimuli at 60 Hz applied to the VLRF. Traces are scaled so that the first eEPSCs_AMPA of train recorded under siRNA_p11‐treatment is equal in size to the first eEPSCs_AMPA of train under the control condition. Facilitation index of eEPSCs_AMPA throughout the train was calculated relative to the amplitude of the first eEPSCs_AMPA in the same train. (H) Mean eEPSCs_AMPA amplitude, normalized to the first eEPSC_AMPA (eEPSC_AMPAn/eEPSC_AMPA1) and plotted against the position number of eEPSCs_AMPAn within the train (1–20 or 1‐40) for the indicated treatment groups of HMNs. The number of HMNs analyzed per condition is indicated in parentheses. **p < 0.01, ***p < 0.001; by unpaired Student's t‐test (A, E). Statistical outputs by applying Kolmogorov–Smirnov (D) or two‐way ANOVA (H) tests are displayed in the plots.

To dissect the synaptic site of action for p11 deletion, we performed a combined electrophysiological analysis as described previously [36, 43]. Miniature quantal events (mEPSC) represent the postsynaptic responses to transmitter release from single vesicles. Their average amplitude is usually taken as a measure of the sensitivity and/or number of postsynaptic receptors, whereas their frequency is primarily, although not exclusively, dependent on the number of functional active sites and presynaptic release probability [44, 45, 46, 47]. In the first set of experiments, we examined the effect of siRNA_p11 administration on these parameters in pharmacologically isolated mEPSCs_AMPA. Our data show that treatment with the interfering oligonucleotide had no significant effect on either the amplitude (cRNA: 31.64 ± 2.56 pA; siRNA_p11: 26.18 ± 1.57 pA) or frequency (cRNA: 13.90 ± 2.74 Hz; siRNA_p11: 9.78 ± 1.54 Hz) of mEPSCs_AMPA (Figure S4; Table S8). Likewise, siRNA_p11‐mediated knockdown of p11 also did not affect the amplitude (−8.4% ± 11.4%) of spontaneously occurring EPSCs_AMPA (sEPSCs_AMPA) recorded under conditions of facilitated synaptic release. Interestingly, however, it significantly reduced the frequency of these types of events (−50.0% ± 4.9%). Consequently, the cumulative distribution of inter‐event intervals was significantly shifted to the right under p11 interference (Figure 7B–E; Table S8). Since in this condition, synaptic activity consists of a mixture of release events arising from both AP‐evoked Ca²⁺ influx and spontaneous fusion stochastically occurring at resting Ca²⁺, our outcomes are consistent with the idea that reduction in p11 expression impairs AP‐dependent glutamatergic transmission through a mechanism involving presynaptic alterations, without apparently affecting the probability of spontaneous vesicle fusion or postsynaptic responsiveness.

As a supplementary analysis, we also measured eEPSCs_AMPA facilitation using paired‐pulse and repetitive afferent stimulation protocols as previously described [36, 43]. Under repetitive stimulation, a change in the amount of facilitation is considered to be attributable to a presynaptic change in the release probability of neurotransmitter quanta [48]. In the control condition, paired‐pulse stimulation evoked a marked facilitation of second eEPSCs_AMPA at the two frequencies tested (Figure 7F; @20 Hz and @60 Hz; Table S9). As shown by the results of this analysis, siRNA_p11 treatment had no effect on the paired‐pulse facilitation ratio (PPR) at the two stimulus intervals tested in comparison with cRNA administration (Figure 7F; Table S9). Repetitive VLRF stimulation with trains of 20 shocks at 20 Hz or 40 pulses at 60 Hz also resulted in a powerful facilitation of eEPSCs_AMPA in cRNA condition (Figure 7G,H). In this case, the standard facilitation index was calculated by measuring the amplitude and charge transfer of the individual eEPSCs_AMPA throughout the stimulus train, normalized to the first eEPSC_AMPA. Noticeably, our data show that the facilitation index throughout the train was potentiated by siRNA_p11 in comparison to the control treatment at the two frequencies tested (Figure 7G,H). Therefore, findings from sEPSCs_AMPA and facilitation index during repetitive stimulation point to a presynaptic site of action for p11 in regulating excitatory neurotransmission in the HN.

2.6. p11 Deletion Alters Synchronous and Asynchronous Release of Neurotransmitter

Evoked neurotransmitter release often consists of two kinetically distinct Ca²⁺‐dependent components: one Ca²⁺ influx‐triggered rapid synchronous release and another slow asynchronous phase due to Ca²⁺ accumulation in the presynaptic terminal attended by different mechanisms of control [49, 50, 51, 52, 53]. In this work, we analyzed synchronous and asynchronous components of neurotransmitter discharge induced by single, double APs, and during and after HFS to look for a feasible differential contribution of p11 in both releasing processes.

After a single shock, cumulative charge transfer data were approximated by double‐exponential fits, reflecting two kinetics of neurotransmitter release (Figure 8A–C). Only responses that were well‐fitted to a double‐exponential were included in the analysis. Functions obtained by the fitting procedure allowed us to quantitatively estimate both the time constants of the fast and slow phases of vesicle release and also the relative charge transfer associated with each component [51]. As expected, under siRNA_p11 treatment, the charge transfer associated with the fast component of the eEPSC_AMPA was strongly decreased (−55.1% ± 9.9%), without affecting that transferred by the slow component (−2.2% ± 17.8%) or the time constants of each of the two phases of release (Figure 8A–G; Table S10). The lack of effect of p11 knockdown on the slow asynchronous component of the eEPSC_AMPA should, however, be interpreted with caution since bath application of 50 μM EGTA‐AM for 30 min, a membrane‐permeable form of the Ca²⁺‐chelating agent EGTA that virtually abolishes asynchronous release [51, 53], also had no significant effects on the charge transfer of both the fast and slow components of the response. These last results suggest that after a single stimulus, the asynchronous component of the evoked response is too subtle to unmask any potential effect of siRNA_p11 treatment on this release phase in our experimental conditions.

FIGURE 8.

siRNA_p11 alters synchronous and asynchronous release of neurotransmitter evoked by single and paired APs. (A, B) Single (A) and averaged (B) traces of eEPSCs_AMPA evoked by single shock VLRF stimulation in cRNA and siRNA_p11‐treated HMNs, and in HMNs pre‐incubated with 25 μM EGTA‐AM. (C) Examples of cumulative probability distributions of the eEPSCs_AMPA charge transfer over a time window of 100 ms for three HMNs under the indicated treatments. (D‐G) Box‐plots and means ± SEM of the charge transfer (D, E) and time constants (F, G) for the fast (D, F) and slow (E, G) components of the eEPSCs_AMPA evoked in HMNs by single stimulation under each of the indicated treatments. (H) Representative examples of paired‐pulse eEPSCs_AMPA recorded from HMNs under the indicated experimental conditions. The shaded portion of the trace represents the 2nd eEPSC_AMPA. (I‐N) Same as in B–G, but for the second eEPSCs_AMPA of the pair elicited by the paired‐pulse stimulation protocol. The number of HMNs analyzed per condition is indicated in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001 relative to cRNA condition; ^### p < 0.001 relative to siRNA_p11 condition; by unpaired Student's t‐test.

Given that asynchronous release is dependent on the intraterminal elevation of residual Ca²⁺ concentration during repetitive stimulation [51, 53], we analyzed the effect of the interfering oligonucleotide against p11 on the fast and slow components of the second eEPSC_AMPA of the pair elicited by application of a paired‐pulse stimulation protocol with an inter‐stimulus interval of 25 ms. Under these conditions, an elevation of residual Ca²⁺ in presynaptic terminals during the second stimulus would be expected to facilitate both release components. Consistent with this, EGTA‐AM completely abolished facilitation of the fast synchronous component of the second response (cRNA: 3.5 ± 0.3 pC; EGTA‐AM: 2.4 ± 0.2 pC), resulting in a charge transfer similar to that obtained by a single stimulus (2.2 ± 0.1 pC) and almost fully abolished slow release during the second EPSC_AMPA (cRNA: 3.7 ± 0.8 pC; EGTA‐AM: 0.8 ± 0.2 pC). Strikingly, the decay kinetics of the second EPSCs_AMPA was comparatively similar in siRNA_p11‐treated HMNs and control HMNs exposed to EGTA‐AM and markedly faster than that obtained for the cRNA‐treated group (Figure 8H,I). The acceleration of the second EPSC_AMPA decay kinetics under p11 knockdown or following incubation with EGTA‐AM produced a leftward shift in the cumulative distribution of the charge transfer associated with this second current response (Figure 8J). Our analysis further reveals that p11 knockdown dramatically attenuated both the synchronous and the asynchronous components of neurotransmitter release. Thus, under p11 deletion, the charge transfer of the fast and slow components of the eEPSC_AMPA was reduced by ~54% and ~64%, respectively, an effect comparable to the decline in charge transfer induced by EGTA‐AM pre‐incubation (−31% and −78%, respectively) (Figure 8K–N; Table S10). Likewise, downregulation of p11 produced a shortening in the time constants of the synchronous (from 8.7 ± 0.6 ms in cRNA‐ to 6.3 ± 0.5 ms in siRNA_p11‐treated HMNs) and asynchronous (from 67.2 ± 11.8 ms in cRNA‐ to 35.1 ± 4.8 ms in siRNA_p11‐treated HMNs) components, to an extent similar to that observed for the Ca²⁺ chelator (fast: 6.0 ± 0.2 ms; slow: 22.4 ± 3.3 ms) (Figure 8K–N; Table S10).

Finally, we addressed the role of p11 in synchronous and asynchronous release during HFS (@60 Hz, 10 s). As a positive control for the efficacy of the interfering oligonucleotide, the amplitude of the first AP‐evoked response in the train was substantially reduced following siRNA_p11 treatment, consistent with previous observations using single‐ or paired‐pulse stimulation protocols (Figure 9A–C). After an initial phase of facilitation (see also Figure 7G,H), the amplitude and total charge transfer of eEPSCs_AMPA during the train exponentially dropped to a final steady‐state of depression (Figure 9C–E). Noticeably, p11 deletion prolonged the facilitation interval during the train and slowed the progression toward steady‐state synaptic depression compared with the control condition (Figures 7H, bottom, and 9D). Although total charge transfer measured over the last 100 pulses of the train (once steady‐state depression has been reached) was lower in siRNA_p11‐ (0.41 ± 0.11 pC) than in cRNA‐ (2.08 ± 0.74 pC) treated HMNs (Figure 9E; Table S11), the degree of depression induced by HFS relative to the first response in the train was not significantly altered by treatment with the interfering oligonucleotide (cRNA: −95.3% ± 2.0%; siRNA_p11: −90.2% ± 4.5%; p = 0.303, Mann–Whitney U test).

FIGURE 9.

siRNA_p11 reduces the size of the readily releasable pool (RRP) and replenishment rate (RR) of synaptic vesicles. (A) Averaged current traces of eEPSCs_AMPA elicited in cRNA‐ (black) and siRNA_p11‐treated (red) groups of HMNs by high‐frequency stimulation (HFS, 60 Hz, 10 s). Insets show examples of 50 ms‐interval averaged traces where tonic charge transfer (shaded areas) was measured. (B) Detailed views of the traces boxed in blue in panel A. Gray and pink shaded areas delimitate the tonic charge component of the eEPSCs_AMPA primarily generated by asynchronous release. (C, D) Average peak amplitude of eEPSCs_AMPA during the train underwent an initial phase of facilitation, followed by a depression of synaptic strength reaching a steady‐state phase at the end in both conditions. In D, data were normalized to the first eEPSC_AMPA amplitude. (E, F) Like in C but for total (E) and synchronous (F) charge transfer. The inset in F shows the first 40 pulses under each treatment. (G) Comparison of the charge transfer of the synchronous component of release for the eEPSCs_AMPA elicited during the last 100 pulses in the train under each of two treatments. (H) Left, averaged tonic charge transfer for the entire pulse train for the two treatment groups. Inset shows a more detailed view of the differences in tonic charge transfer between both groups during the first 40 pulses. Middle and right, comparison of the tonic charge transfer for the synaptic current responses elicited by the 40th pulse in the train (middle), and at the steady‐state phase of depression (right), under each treatment condition. (I) Averaged tonic charge transfer measured after the end of the train was also reduced by interfering with p11. (J) Left, curve of cumulative charge transfer of the synchronous component of eEPSCs_AMPA plotted against pulse number during application of the 60 Hz train in the cRNA‐ and siRNA_P11‐treated groups of HMNs. Dotted lines represent linear regression fits to the last 100 eEPSCs_AMPA back‐extrapolated to pulse 0 on the y‐axis to estimate the initial RRP size (arrow). Estimation of the RRP size by this procedure was confined to HMNs that reached a steady‐state response in the late phase of the stimulus train. Right, the RRP calculated in HMNs from pups treated with siRNA_P11 was reduced in comparison with the control condition. (K) The release probability (calculated as the ratio: First eEPSC_charge/RRP) was similar in both groups. (L) Left, cumulative total charge transfer of eEPSCs_AMPA plotted against pulse number during the 60 Hz train. The slope of the linear regression fits to the last 100 responses rendered an estimate of the RR. Right, plot of the RR values calculated for each treatment group showing the siRNA_P11‐mediated reduction of this parameter. The number of HMNs analyzed per condition is indicated in parentheses. *p < 0.05; **p < 0.01; by unpaired Student's t‐test.

In this same phase of the steady‐state depression (100 last pulses), charge transfer of phasic/synchronous component of eEPSCs_AMPA fell to 0.08 ± 0.02 pC (−92.0% ± 2.4%) and 0.08 ± 0.03 pC (−90.6% ± 1.8%) after control and interfering oligonucleotide administration, respectively (Figure 9F,G; Tables S11 and S12), which is consistent with the almost full depletion of the vesicle pool mobilized during AP‐triggered synchronous release of neurotransmitter. Interestingly, the kinetics of depression of the phasic component from the peak of the initial facilitation phase in cRNA‐treated HMNs was best fitted to a double‐exponential function with fast (τ_fast) and slow (τ_slow) time constants of 0.21 ± 0.06 s and 1.39 ± 0.27 s, respectively (Table S11). This temporal pattern of progression is in accordance with the initial depletion of the readily releasable pool (RRP) and subsequent vesicle mobilization from the reserve pool (RP) [54, 55, 56, 57]. Remarkably, p11 downregulation did not alter τ_fast (0.41 ± 0.10 s) but resulted in the prolongation of τ_slow (4.85 ± 1.64 s) (Table S11). These findings conclusively suggest that p11 knockdown impacted synchronous release during HFS, altering the kinetics of vesicle depletion primarily by affecting recruitment from the RP.

Measurements of tonic charge transfer (TCT) have been broadly used as a method to estimate the magnitude of the asynchronous release during a train of stimuli. Although accumulation of neurotransmitter in the synaptic cleft could contribute, at least to some extent, to the tonic component of the total charge transfer, its impact is expected to be minimal at the onset of the stimulus train [51, 53, 58, 59]. Our analysis shows that TCT increased abruptly during the initial phase of the stimulus train (first 10 stimuli; ~167 ms), before significant glutamate accumulation could occur in the synaptic cleft. In siRNA_p11‐treated neurons, cumulative TCT during the first 10 pulses was strongly reduced (cRNA: 60.98 ± 8.66 pC; siRNA_p11: 19.63 ± 3.71 pC; p < 0.001, Student's t‐test), thereby indicating a drastic impairment of asynchronous neurotransmitter release during the initial phase of the train. Following this initial enhancement, TCT decreased exponentially with a similar τ_TCT in cRNA‐ and siRNA_p11‐treated HMNs (1.85 ± 0.33 s vs 1.98 ± 0.56 s, respectively), up to a steady‐state depression that was more pronounced after p11 knockdown (1.81 ± 0.64 pC vs 0.31 ± 0.07 pC for cRNA‐ and siRNA_p11‐treated groups, respectively) (Figure 9H; Table S11). At this point, it is interesting to remark that total charge transfer during the final phase of steady‐state depression was almost fully supported by TCT, with the synchronous release component being virtually exhausted. TCT in the siRNA_p11‐treated group was also lower immediately after the end of the train of stimuli, in a context where any possible interference from synchronous release would be no longer expected (Figure 9I; Table S11).

Taken together, these outcomes indicate that p11 participates in common mechanisms engaged in both the Ca²⁺ influx‐triggered rapid synchronous release and the slower asynchronous phase driven by Ca²⁺ accumulation in the presynaptic terminal. Furthermore, p11 is likely to act as a regulator of the time course of short‐term synaptic depression, potentially influencing vesicle depletion and/or mobilization from the RP during HFS.

2.7. p11 Knockdown Impacts RRP Size and Vesicle Resupply

We further extended our analysis of the synaptic responses elicited by HFS trains to estimate the RRP size, release probability (P_r), and replenishment rate (RR) of vesicles. One of the approaches used for the estimation of these parameters is based on the notion that depression of the synchronous component of synaptic transmission during HFS represents the depletion of the RRP, and once reached the steady‐state, synaptic release would be mostly sustained by ongoing replenishment of the depleted RRP [51, 53, 60]. Under these premises, to estimate the RRP size, the cumulative charge transfer of the synchronous component of eEPSCs_AMPA was plotted against the pulse number and a linear fit regression to the last 100 eEPSCs_AMPA of the steady‐state phase was back‐extrapolated to pulse 0 on the y‐axis. The intercept point on this axis represents an estimate of RRP size, expressed in terms of cumulative charge transfer of the eEPSCs_AMPA elicited by its release, and excluding the contribution of vesicle replenishment during stimulation (Figure 9J). Given that P_r is defined as the fraction of vesicles in the RRP that can be released by a single AP, its value can be estimated by dividing the charge transfer of the first eEPSC_AMPA in the train (eEPSC_1AMPA) by the RRP size (Figure 9K). Finally, the RR was determined from the slope of the linear regression fit to the cumulative plot of total charge transfer of eEPSCs_AMPA for the last 100 stimuli of the train (Figure 9L). siRNA_p11‐induced p11 knockdown resulted in substantial reductions in the RRP size (−67.7% ± 7.9%) and RR (−80.0% ± 4.8%) compared with controls but left the P_r unaltered (Figure 9J–L; Table S11). Thus, p11 has a pivotal role in fine‐tuning RRP size and vesicle recruitment from the RP to replenish RRP during HFS.

Because the balance between vesicle pool depletion and rapid vesicle replenishment during HFS is crucial for short‐term plasticity, we additionally determined the vesicle resupply rate after HFS‐induced short‐term depression by examining the recovery of eEPSC_AMPA elicited by a test pulse applied at different intervals following the train (Figure 10A). Transmission recovery was measured as the ratio between peak amplitudes or charges of the eEPSC_AMPA evoked by the test pulse and the first eEPSC_AMPA in the train. In our conditions, transmission recovery was best fitted to a double‐exponential function (Figure 10B), which is consistent with two phases of Ca²⁺‐dependent recovery as described for the calyx of Held and hippocampal synapses [53, 61, 62]. Despite the fact that the recovery curves started at similar initial levels in both conditions (cRNA: 8.0% ± 2.4%; siRNA_p11: 9.4% ± 1.8%, relative to their respective first eEPSC_AMPA in the train), the recovery rate of eEPSC_AMPA amplitude under p11 knockdown was slower (τ_fast: 0.33 ± 0.10 s; τ_slow: 5.0 ± 1.1 s) than in the control condition (τ_fast: 0.09 ± 0.02 s; τ_slow: 2.7 ± 0.7 s) (Figure 10C–E; Table S12). Outcomes followed the same trend when charge transfer of evoked responses was analyzed although the subtle increase in τ_slow under siRNA_p11 did not reach statistical significance (Table S12). Therefore, reducing p11 affects the vesicle resupply to AZs following HFS.

FIGURE 10.

p11 knockdown slows down and results in incomplete recovery of eEPSCs_AMPA from the HFS‐induced short‐term depression. (A) Averaged current traces showing the sequence of eEPSCs_AMPA elicited during the last 100 responses of HFS (60‐Hz, 10 s), and subsequently in response to a test pulse applied at different intervals (Δt = 0.01, 0.09, 0.2, 0.4, 3.8, 2, 4 and 5 s) following the 60 Hz train in c‐RNA‐ (black) and siRNA_p11‐treated (red) HMNs. Traces have been obtained by averaging individual records of 12 and 9 HMNs, respectively, for each condition. (B) Graph showing single‐ (green) and double‐exponential (black) fits to the time course recovery of evoked synaptic transmission (measured as the ratio between peak amplitudes of the eEPSC_AMPA elicited by the test pulse and eEPSCs_AMPA1 in the train) in two representative HMNs under each condition. Time constants and R² for each fit are shown in the plot. (C–E) Time course recovery of eEPSC_AMPA amplitude from STD (C) and comparison of values for τ_fast (D) and τ_slow (E) obtained for each treatment group. (F) Same as in C, but for eEPSC_AMPA charge transfer. (G) Early phase of the curves illustrated in F (shaded areas) at expanded time scales to show the significantly slower initial recovery in the siRNA_p11–induced knockdown condition. The number of HMNs analyzed per condition is indicated in parentheses. *p < 0.05; **p < 0.01; by unpaired Student's t‐test.

3. Discussion

This study highlights a pivotal role of p11 protein for the normal performance of motor commands, at least at the final output stage, by impacting on excitatory synaptic drive incoming to motor neurons. The possible mechanism of action engages the regulation of expression and interaction with the AZ protein Munc13‐1, which is involved in vesicle priming, thus impacting vesicle exocytosis. Our in vitro studies demonstrate that p11 is required to maintain synaptic strength by managing synchronous and asynchronous neurotransmitter release and vesicle dynamics, which comprise fine‐tuning of the RRP size and vesicle recruitment rate to replenish RRP. Thus, p11 is proposed as a multifaceted interactor protein in the processing of neurotransmitter release and activity‐dependent synaptic plasticity in the CNS.

The broad distribution of p11 throughout the CNS [3, 5] anticipates the participation of this adaptor factor in neural information processing underlying key brain functions. In particular, p11 acts as a retention factor at the endoplasmic reticulum for the “leak” potassium channel TASK1 in motor neurons, thereby fine‐tuning intrinsic excitability of this cell type [11]. Disruption of this cell‐autonomous mechanism, which would result in increased surface expression of TASK1 channels and subsequently in a decline in intrinsic excitability, might lie behind the depression of baseline inspiratory‐related activity of HMNs accompanying neuron‐specific p11 knockdown. One might anticipate that, under this cell‐type‐specific strategy for the experimental knockdown of p11 expression, the contribution of astrocytic p11 would be negligible [4]. It could be argued that the attenuated discharge activity exhibited by HMNs under p11 knockdown might be due, at least in part, to the decrease in surface expression of presynaptic 5‐HT receptors since p11 promotes the insertion of these receptors into the membrane of presynaptic terminals, where they regulate neurotransmitter release [1, 63]. However, this suggestion seems untenable, as microinjection of 5‐HT receptor agonists in the HN dampens spontaneous output activity from HMNs, mainly by acting on presynaptic 5‐HT_1B receptors, either by suppressing excitatory neurotransmission or by strengthening inhibitory inputs to HMNs [64]. In addition, Ahnak functionally scaffolds L‐type Ca²⁺ channels as a downstream effector of the Ahnak/p11/AnxA2 complex, thus regulating the influx of Ca²⁺ into neurons [10]. The C‐terminal region of Ahnak interacts with various auxiliary β subunits of L‐type Ca²⁺ channels [10, 65] which also bind to all voltage‐gated Ca²⁺ channels including L‐, N‐, R‐ and P/Q‐type channels, most of which conduct Ca²⁺ currents that initiate rapid synaptic transmission [66, 67, 68, 69]. Interestingly, Ahnak has been identified as part of a protein complex associated with presynaptic voltage‐gated Ca²⁺ channels at Torpedo electric organ synapses [70]. Furthermore, several lines of evidence (Refs. [11, 71], https://mouse.brain‐map.org/) support that the molecular partners of the Ahnak/p11/AnxA2 complex might coexist in the HN and in the VLRF, the main source of afferent inputs examined in this study. Thus, whether p11 downregulation impacts neurotransmitter release by reducing Ca²⁺ currents, thereby, cannot be ruled out and merits future research.

In this study, we provide evidence of a close proximity relationship between p11 and several presynaptic proteins in the HN, pointing to this protein as a partner of the molecular mechanisms that manage neurotransmitter release. In this sense, co‐immunoprecipitation analysis of brain tissue showed that p11 resides in a molecular complex that also includes Munc13‐1, a presynaptic AZ protein essential for vesicle maturation and priming at glutamatergic synapses [29, 30, 31]. Likewise, PLA specifically revealed direct p11‐Munc13‐1 interactions partially overlapping with glutamatergic vesicle pools. Further, as evidenced in this study, siRNA_p11 administration resulted in a reduction in Munc13‐1 levels in both the adult and neonatal models. Remarkably, our ultrastructural analysis shows that under the siRNA_p11‐mediated co‐downregulation of p11‐Munc13‐1, excitatory boutons (S‐type) attached to HMNs displayed a normal number of docked vesicles per AZ, assumed to represent the RRP, but increased numbers of vesicles both near the docked pool and in the entire pool of vesicles within presynaptic boutons. These findings argue against any defect in endocytosis or vesicle biogenesis caused by p11 knockdown, as synaptic vesicles remain abundant and clustered at release sites. However, despite vesicle availability, excitatory neurotransmission was impaired in the adult and neonatal models, and the size of “functional” RRP was reduced after interference with p11. These results suggest that p11 knockdown, potentially through downregulation of Munc13‐1, disrupts vesicle exocytosis, which in turn could also affect vesicle mobilization to the RRP. Similar vesicle accumulation was also observed in invertebrate genetic models bearing null mutations for the homolog genes unc‐13 and dunc‐13 [72, 73], for syntaxin or synaptobrevin [74], which all function downstream of vesicle docking, strengthening the hypothesis that synaptic vesicle exocytosis was impaired after siRNA_p11. The munc13‐1 knockout mouse, however, does not display changes in vesicle distribution or AZ length, as reported here, although synaptic transmission is blocked [29]. Further studies are needed, however, to determine whether the absence of alterations at the ultrastructural level involves more efficient feedback mechanisms controlling vesicle dynamics and/or partial redundancy with other members of the Munc13 family in the mouse [72]. We hypothesize that in our experimental model of p11 depletion, a disruption of vesicle tethering mediated by the p11‐Munc13‐1 pair cannot be ruled out by analogy with the function of p11‐Munc13‐4 in endothelial cells [18]. In any case, and based on our results, it could indeed be feasible that p11 might exert other actions beyond those mediated by interaction with Munc13‐1 and/or by the control of its expression, such as the regulation of AZ composition/structure, which merit further investigation. In addition, the findings that AnxA2 stabilizes intracellular p11 by preventing its proteasomal degradation [75], together with the co‐expression of p11 and AnxA2 in HMNs and in VLRF (Refs. [11, 71], https://mouse.brain‐map.org/) and the requirement of AnxA2 for the role of p11 in various exocytic processes [15, 16, 17, 18], collectively support the unresolved hypothesis of AnxA2 requirement for the p11 function in regulating neurotransmitter release at excitatory synapses reaching HMNs.

The impact of p11 in excitatory neurotransmission could be further emphasized by the finding that p11 loss also induced the downregulation of two additional presynaptic proteins, vGLUT2 and Syn, the latter of which has been proposed to participate in the vesicle exo‐endocytosis cycle [76]. Paradoxically, siRNA_p11 administration resulted in a gain in the number of S‐type boutons in apposition on HMN cell bodies, accompanied by vesicle accumulation inside them. Although it is tempting to speculate that, in our experimental model, Syn reduced expression might be causally linked to disruption of synaptic function and ultrastructure following p11 knockdown, this notion seems unlikely. Syn knockout mice do not exhibit an overt phenotype distinguishable from wild‐type mice, showing no evidence of synaptic dysfunction and a normal ultrastructural appearance of synaptic boutons. This suggests a certain degree of functional redundancy in Syn [76, 77, 78]. One could also hypothesize that the siRNA_p11‐induced downregulation of vGLUT2 might similarly represent a causal factor in the disruption of synaptic transmission observed in our experimental model, by the effect that the deletion of this transporter would be expected to have on vesicle refilling with glutamate. In line with this idea, it has been documented that complete targeted deletion of the Vglut2 gene causes lethality at birth and severely impairs evoked (> 95% reduction in eEPSCs amplitude) and spontaneous glutamate release (~50% and 90% reductions in mEPSCs amplitude and frequency, respectively) in thalamic neurons, a brain region where vGLUT2, like in the brainstem, is the most highly expressed isoform. A 50% reduction in vGLUT2 expression levels due to the loss of only one of the two alleles of the gene in heterozygous had no effects on motor performance in animals, or on the amplitude of evoked EPSCs, RRP size, or mEPSC frequency in this same neuronal type, but did lead to a 24% reduction of quantal size (mEPSC amplitude) [79]. On the contrary, overexpression of vGLUT1 at hippocampal excitatory synapses, where this isoform is predominantly expressed, results in both increased amounts of glutamate released per vesicle and enhanced evoked and miniature responses [80]. These observations support the idea that, although the magnitude of the postsynaptic response to the release of a single synaptic vesicle is clearly determined by the number and sensitivity of postsynaptic receptors, presynaptic mechanisms that affect filling of vesicles with neurotransmitter, and even the size of the vesicle, can also influence quantal size [81, 82]. In our study, vGLUT2 downexpression under p11 knockdown at neonatal stages did not affect mEPSC amplitude or frequency (Figure S4D,G), although AP‐evoked EPSCs were strongly attenuated (Figure 7A). One possible explanation for the lack of effect on mEPSCs amplitude is that while downregulation of vGLUT2 could lead to reduced glutamate content in vesicles, the amount of neurotransmitter released per vesicle could still be large enough to saturate all ionotropic postsynaptic receptors, thereby masking any effect of changes in vesicle filling. This idea is also in agreement with findings from other studies showing that decreased expression of the isoforms for vGLUT predominantly expressed in hippocampal circuits of mice or the neuromuscular junction of Drosophila had little effect on vesicle filling and quantal amplitudes of miniature synaptic events [83, 84]. Therefore, the dramatic reduction of AP‐evoked synaptic responses in HMNs under p11 knockdown is likely more closely associated with the reduced size of “functional” RRP and impaired vesicle mobilization to the RRP, rather than with impaired vesicle refilling with glutamate resulting from reduced vGLUT2 expression.

Our combined expression and electrophysiological analysis in neonatal rats demonstrates that the downregulation of Munc13‐1 and vGLUT2 induced by p11 interference was accompanied by a marked reduction in the synaptic strength of glutamatergic synapses, through presynaptic alterations affecting both the synchronous and the asynchronous components of neurotransmitter release. Currently, there is substantial evidence to support the notion that synchronous and asynchronous release differ in several key aspects, including the source and spatial domain of Ca²⁺ elevation required to trigger release, the type and affinity for Ca²⁺ of the sensors that trigger release, or the specific protein isoforms that participate in vesicle mobilization, fusion, or recycling (reviewed in [85, 86]). For instance, whereas fast synchronous release is triggered by rapid and local increases in Ca²⁺ concentration via the high‐affinity Ca²⁺ sensor synaptotagmin‐1, Doc2α and synaptotagmin‐7 have been proposed as the slow Ca²⁺ sensors that play a critical role in asynchronous neurotransmitter release, which is triggered by slower and more sustained Ca²⁺ signals [49, 50, 51, 52]. Despite the specificity of the regulatory signals and differences in the components of the molecular machinery that mediate each mode of release, it is highly probable that both share common key steps and regulatory mechanisms. At the molecular level, both modes of release seem to rely, and converge, on the SNARE‐dependent fusion process mediated by canonical SNARE proteins synaptobrevin‐2/VAMP2, SNAP‐25, and syntaxin‐1 [86].

Here, we show that p11 deletion strongly affected both fast‐ and slow‐release components, which is in line with its possible involvement in common processes, likely downstream vesicle tethering, and impacting on vesicle docking, priming, or fusion. In this context, a multifaceted role of p11 in different steps/processes of synaptic vesicle exocytosis cannot be ruled out. A key link between p11 and the regulation of synchronous and asynchronous neurotransmitter release could involve the interaction of Munc13‐1 with the Munc18‐1/syntaxin‐1 complex during the priming process. This interaction drives the opening of syntaxin‐1, thereby promoting its binding to VAMP2 for SNARE complex assembly [87, 88, 89]. Besides its possible role as a molecular interactor along with Munc13‐1 in the priming process of vesicle exocytosis, p11 could also assume additional roles in the final fusion step of vesicles. In this context, it is worth noting that the interaction between Munc13‐1 and Doc2α is required for evoked neurotransmission at cholinergic synapses in rat superior cervical ganglion neurons [90]. Whether p11 throughout its interaction with Munc13‐1 is also a key interactor for Doc2α‐mediated regulation of vesicle exocytosis is a challenge that merits further attention. Finally, p11 knockdown contributed to short‐term depression by reducing the replenishment rate of RRP during and after HFS, despite the increased vesicle accumulation around the AZ. This suggests a role of p11 in the rapid recruitment of vesicles to replenish the RRP, with negligible influence in the endocytic process, as reported for intersectin‐1 in fast neurotransmission [91].

In conclusion, our data contribute firm evidence for a role of the adaptor protein p11 in neurotransmitter release by controlling synchronous and asynchronous release of neurotransmitter, vesicle priming and exocytosis, the RRP size, and vesicle recruitment rate to replenish RRP, which may be of primary importance for information processing in the brain in health and neurological conditions.

4. Methods

4.1. Animals

Experiments were performed in neonatal (P5–P9) and adult (> 2‐month‐old) Wistar rats of either sex, and embryos extracted from CD1 pregnant mice (12.5‐days gestation). All these animals were provided by the local Animal Supply Services (SEPA, University of Cadiz). Animal care and handling followed the guidelines of the European Union Council (2010/63/EU, 86/609/UE) on the use of laboratory animals. Experimental procedures were approved by the local Animal Care and Ethics Committee (University of Cadiz, Cadiz, Spain) and the Counseling of Agriculture, Livestock, Fisheries and Sustainable Development (Project number 11/05/2022/064, Junta de Andalucía, Spain). Animals were individually housed in cages with water and food pellets available ad libitum, at 21°C ± 1°C, with a 12‐h light/dark cycle. Neonatal animals were housed with their mother. Efforts were made to minimize the number of animals used and their suffering. All surgical processes were carried out under aseptic conditions.

4.2. Unitary Extracellular Recordings of HMNs

Adult rats were prepared for extracellular recordings as previously described [23, 92]. Briefly, animals were anesthetized with 1.5%–3% isoflurane and additionally injected intramuscularly with atropine (0.2 mg/kg) and dexamethasone sodium phosphate (0.8 mg/kg). Teflon‐isolated silver bipolar electrodes were fixed around the right XIIth nerves (Figure 1B). Trachea, bladder, and femoral artery and vein were cannulated. Subsequently, animals were vagotomized, decerebrated, paralyzed with gallamine triethiodide (20 mg/kg, i.v., initially; 4 mg/kg, i.v., as needed), and mechanically ventilated. Expired CO₂ and O₂ were monitored continuously (Eliza duo; Gambro Engström, Bromma, Sweden). During the experiment, the ET_CO2 was changed (~3.5% to ~6.2%) by adjusting tidal volume and/or respiratory rate. Expired O₂ (14%–19%) was always higher than values below which hypoxia‐induced alterations have been reported [22]. Femoral arterial blood pressure (95 ± 15 mmHg) and rectal temperature (37°C ± 1°C) were continuously monitored and kept stable. Glass micropipettes (1–3 MΩ), filled with 2 M NaCl, were placed under visual guidance and advanced through the brainstem into the HN. The correct position of the micropipette was confirmed by recording the characteristic and rhythmic inspiratory pattern of neuronal discharge, and by the presence of the antidromic field potential elicited by electrical stimulation of the ipsilateral XIIth nerve. HMNs were identified by their antidromic activation from the XIIth nerve and by the collision test [23, 28]. The electrical signals were amplified and filtered at a bandwidth of 10 Hz to 10 kHz for display and digitalization purposes. Responses of HMNs were recorded in response to changes in ET_CO2 from hypocapnic (~3.5%) to hypercapnic (~6.2%) conditions. Only inspiratory HMNs discharging at basal conditions (ET_CO2 = 4.8%–5.2%) were considered in this study (Figure 1B). Unitary discharge activity, percentages of expired CO₂ and O₂, and arterial pressure recordings were amplified, filtered, digitized, and stored in a computer using a PowerLab/8SP analog‐to‐digital interface (ADInstruments, Castle Hill, Australia) for offline analysis. The mFR (in spikes per second) in each burst and the cycle duration, used to calculate BR (in bursts/min), were measured over the range of ET_CO2 tested. Because mFR and BR changed proportionally to alterations in ET_CO2 (Figure 1D), both parameters were positively correlated using linear regression analysis (Figure 1E). It was thus possible to obtain two equations characterizing the behavior of an HMN and the integrity of premotor structures in response to changes in ET_CO2: mFR = S_mFR × ET_CO2 + I_mFR where S is the slope (i.e., the neuronal gain or sensitivity of mFR relative to ET_CO2 variations), and I is the ordinate intercept (i.e., the theoretical mFR value when ET_CO2 = 0%); and BR = S_BR × ET_CO2 + I_BR. Data were grouped and averaged in intervals of 0.2% of ET_CO2. Only regressions with r ≥ 0.8 and p < 0.001 were considered for statistical analysis.

4.3. In Vitro Whole‐Cell Recordings From HMNs in Brainstem Slices

Whole‐cell patch‐clamp experiments were performed on HMNs from transverse brainstem slices of 7‐ to 9‐day‐old Wistar rats as previously described [42, 43]. In summary, rat pups were anesthetized by hypothermia (placed on ice for 10–15 min) and decapitated, and their brainstems were quickly removed. Dissection was in ice‐cold (~4°C) sucrose artificial CSF (S‐aCSF) bubbled with 95% O₂ and 5% CO₂. S‐aCSF composition was as follows (in mM): 26 NaHCO₃, 10 glucose, 3 KCl, 1.25 NaH₂PO₄, 2 MgCl₂, and 218 sucrose. Transverse slices (300–400‐μm thick) were obtained using a vibroslicer (NVSL; WPI). Slices were transferred to normal oxygenated aCSF (in mM: 26 NaHCO₃, 10 glucose, 3 KCl, 1.25 NaH₂PO₄, 2 MgCl₂, 130 NaCl, and 2 CaCl₂) and allowed to stabilize at ~37°C for 30 min. Next, slices were transferred to a recording chamber for whole‐cell patch‐clamp recordings. Recordings were performed at 31°C on slices superfused (rate ~ 3–4 mL/min) with artificial CSF solution equilibrated with 95% O₂ and 5% CO₂. HMNs were identified based on their location, characteristic size, and shape, and then patched under visual guidance using IR‐DIC (infrared‐differential interference contrast) optics. Patch electrodes (1.5–3 MΩ resistance) contained a KGluconate‐based internal solution (composition in mM: 17.5 KCl; 122.5 KGluconate, 9 NaCl; 1 MgCl₂; 10 HEPES; 0.2 EGTA; 3 Mg‐ATP, 0.3 GTP‐Tris with pH buffered to 7.2). Voltage‐clamp recordings were obtained and low‐pass Bessel‐filtered at 4 kHz with a MultiClamp 700B amplifier. Data were digitized at 10 kHz with a Digidata 1332A analog‐to‐digital converter and acquired using pCLAMP 9.2 software (Molecular Devices). Only recordings with access resistance between 5 and 20 MΩ were considered acceptable for analysis. The access resistance was checked throughout the experiments, and recording was abandoned if it changed > 15%. Series resistance was routinely compensated by 65%–75%. The pipette offset potential was zeroed before the cells were patched. Leak currents and liquid junction potentials were not corrected.

Whole‐cell voltage‐clamp recordings were used to analyze the effects of oligonucleotide administration on synaptic strength at excitatory synapses on HMNs. To clearly determine these effects and avoid interference with other synaptic mechanisms, we focused our analysis on the AMPAergic component of the excitatory postsynaptic responses (EPSC_AMPA). To distinguish between presynaptic and postsynaptic effects of oligonucleotide administration, we performed a combined electrophysiological analysis of (1) mEPSCs_AMPA, (2) sEPSCs_AMPA, (3) eEPSCs_AMPA evoked by electrical stimulation of afferent axons from the VLRF, and (4) synaptic facilitation of eEPSCs_AMPA following repetitive or paired‐pulse stimulation. Whole‐cell AMPAergic responses were recorded at a holding potential of −65 mV with the KGluconate‐based intracellular solution.

mEPSCs_AMPA were pharmacologically isolated in the presence of 1 μM tetrodotoxin (TTX), 1 μM strychnine hydrochloride, 10 μM bicuculline methochloride, 10 μM D‐tubocurarine, and 50 μM (DL)‐APV applied to the bath perfusion. All these drugs were purchased from Tocris Cookson. The recording of sEPSCs_AMPA was performed under conditions of facilitated synaptic release without TTX in a modified extracellular solution containing high‐Ca²⁺ (4 mM), high‐K⁺ (9 mM), and the receptor antagonists indicated above. In a previous report [43], the AMPAergic nature of both types of events (mEPSCs_AMPA and sEPSCs_AMPA) was confirmed in some cells by blocking with NBQX (20 μM; Tocris Cookson) at the end of the experiment. For each cell and treatment condition, we obtained segments of continuous recording (5–8 min) of spontaneous activity. During analysis, the threshold level for detection of events was set at three times the root mean square noise (~2–3 pA, detection threshold ~6–9 pA). Spurious events and noise artifacts were manually rejected after visual inspection of the automatically screened events. For each cell, peak amplitudes and inter‐event intervals of single mEPSCs_AMPA and sEPSCs_AMPA were measured across the different recording epochs and used to calculate mean values and generate cumulative probability plots.

eEPSCs_AMPA were elicited using a concentric bipolar tungsten electrode placed in the VLRF ~0.5–1 mm from the border of the HN. Evoked AMPAergic responses were recorded in the absence of TTX and pharmacologically isolated with blockers of glycine, nicotinic, NMDA, and GABA_A receptors. Stimulation parameters (< 400 μA, 50 μs pulse duration, delivered at 0.05 Hz) were chosen to generate a sizable and reproducible postsynaptic response. Usually, the stimulus intensity was adjusted so that the postsynaptic response was ~50% of maximal amplitude and then it was maintained constant throughout the recording period. A minimum of 10 responses was recorded under each pharmacological condition and averaged for analysis of eEPSCs_AMPA amplitude or charge transfer (area contained in the evoked postsynaptic response).

Facilitation of eEPSCs_AMPA was studied using conventional repetitive stimulation protocols. In these experiments, the AMPA receptor–mediated component of the responses was pharmacologically isolated as described above. For these studies, VLRF was stimulated with brief trains of electrical stimuli at frequencies of 20 Hz and 60 Hz. PPR was computed by dividing the peak amplitude and charge transfer of the second response in the train by the peak amplitude or charge transfer of the first one, respectively. The repetitive stimulation protocol consisted of a train of 20 pulses at 20 Hz or 40 pulses at 60 Hz applied to the VLRF. Estimates of facilitation are expressed relative to the amplitude or charge transfer of the first response in the train. The synchronous charge component of each response was calculated as the integral of the current relative to an individual baseline determined 1 ms prior to each stimulus. Asynchronous release during the train was estimated, for each response in the train, as the difference between total transferred charge and charge transfer during synchronous release, i.e., the TCT, as previously described [51, 53, 58]. Asynchronous release was also calculated by measuring charge transfer during the period from 50 to 100 ms after the last eEPSC_AMPA in response to the 10 s train at 60 Hz.

Finally, the rate of recovery from synaptic depression induced by HFS was assessed by applying single stimuli 0.01, 0.1, 0.3, 0.7, 4.5, 6.5, 10.5, 15.5, and 20.5 s after the end of a 10‐s, 60 Hz stimulus train. Recovery kinetics of eEPSCs_AMPA amplitude and charge transfer were best fitted to a double‐exponential function with two time constants (τs; τ_fast, τ_slow).

4.4. Primary Cultures of Spinal Cord MNs (SMNs)

Primary cultures of SMNs were prepared from the spinal cord of mouse embryos at 12.5 days of gestation (E12.5), following a well‐established protocol in our lab [35]. Briefly, desiccated ventral cords were first chemically (0.025% trypsin in glucose–HEPES buffer solution supplemented with 20 i.u./mL penicillin and 20 mg/mL streptomycin) and mechanically dissociated to be collected under a 4% bovine serum albumin cushion. Centrifugation (10 min, 520 × g) on an Iodixanol density gradient (OptiPrep, Axis‐Shield, Oslo, Norway) allowed the isolation of the largest cells, which were once more centrifuged through a bovine serum albumin cushion. Isolated cells were pooled in a tube containing culture medium and plated at various densities depending on the type of experiment. Cultured SMNs were unambiguously identified by SMI32 immunofluorescence or by morphological criteria [11, 35]. SMNs were cultured in Neurobasal medium (Gibco, Invitrogen, Paisley, UK) supplemented with B27 (Gibco; Invitrogen), horse serum (2% v/v), L‐glutamine (0.5 mM; Gibco; Invitrogen), and 2‐mercaptoethanol (25 μM; Sigma‐Aldrich), and a cocktail of neurotrophic factors (NTFs): 1 ng/mL brain‐derived neurotrophic factor, 10 ng/mL glial cell‐line‐derived neurotrophic factor, 10 ng/mL ciliary neurotrophic factor, and 10 ng/mL hepatocyte growth factor (PreProtech, London, UK).

4.5. Quantitative Real‐Time Reverse Transcriptase PCR ( qRT‐PCR)

As previously described [11], total RNA was extracted from the medulla using the TRIsure Isolation Reagent (BioLine). This procedure allowed the isolation of total RNA, DNA, and protein fractions from a single sample. Subsequent treatment with the RNase‐free DNase set (Invitrogen), following the manufacturer's recommendations, was performed to reduce DNA contamination. The concentration and purity of RNA samples were determined by spectrophotometry at 260 and 280 nm. Retrotranscription (RT) was done using random hexamers, 500 ng of total RNA as template and iScriptTM cDNA Synthesis Kit (Bio‐Rad). For real‐time RT‐PCR, each specific gene product was amplified with the MiniopticonTM System (BIO‐RAD), using iQ SYBR Green Supermix (Bio‐Rad). The PCR primer sequences were for p11, forward 5′‐ATGGAAAGGGAGTTCCCTGGGT‐3′, reverse 5′‐TCCAGTTGGCCTACTTCTTCTG‐3′ and for gapdh, forward 5′‐AGAACATCATCCCTGCATCCA‐3′, reverse 5′‐AGATCCACGACGGACACATTG‐3′. The cDNA levels for the different samples were determined using the 2^−∆∆CT method, being gapdh the housekeeping gene. All analyses were performed in triplicate, with each experiment repeated at least twice.

4.6. Western Blot Analysis

Tissue samples from the medulla or microdissected HNs were homogenized in TRIsure, as above, or in lysis buffer [50 mM Tris/HCl, pH 7.4, 1% (v/v) Triton X‐100, 0.5% (w/v) sodium deoxycholate] supplemented with protease (Sigma‐Aldrich) and phosphatase (Pierce) inhibitors using a 1‐mL syringe. Protein concentrations were determined by Bradford protein assay (Bio‐Rad) or by micro BCA protein assay kit (Thermo Scientific). Proteins were separated by SDS‐PAGE and blotted to a polyvinylidene difluoride membrane (Amersham). Membranes were blocked with BSA 5%, 0.1% Tween for 1 h at room temperature and subsequently blotted with specific antibodies against p11 (1:1000, R&D systems, Cat# AF2377, RRID:AB_2183469) developed in goat, β‐actin (1: 2500000, Sigma‐Aldrich, Cat# A5441, RRID:AB_476744) and α‐tubulin (1:1000000, Sigma‐Aldrich, Cat# T9026, RRID:AB_477593) developed in mouse, vGLUT2 (1:1000, Millipore Bioscience Research Reagents, Cat# AB2251, RRID:AB_1587626) developed in guinea pig and Munc13‐1 (1:1000; Synaptic systems, Goettingen, Germany, Cat# 126102, RRID:AB_887734) developed in rabbit. Proteins were visualized by using alkaline phosphatase‐conjugated secondary antibodies (Pierce) and the chemiluminiscent immunodetection system (Invitrogen). Analysis was performed using ImageJ 1.48 v software from the National Institutes of Health (NIH).

4.7. Immunoprecipitation

Immunoprecipitation assays were performed using Pierce Classic Magnetic co‐immunoprecipitation kit (Thermo Scientific) following the manufacturer's instructions with subtle modifications. Briefly, the brain of adult rats was mechanically homogenized in immunoprecipitation assay buffer (5 μL/mg of tissue; 50 mM Tris, 150 mM NaCl, 1 mM EDTA, 1% NP40, and 5% glycerol, pH 7.4, supplemented with a protease inhibitor cocktail from Sigma). Following centrifugation at 13 000 × g for 10 min at 4°C, the clarified supernatants were collected. To adjust the protein concentration in samples prior to immunoprecipitation, it was measured using a micro BCA protein assay kit (Thermo Scientific). Then, the samples were incubated overnight at 4°C with 5 μg of anti‐p11 antibody (R&D Systems). As a negative control, the samples were incubated with 5 μg of an irrelevant antibody (anti‐goat IgG) under the same conditions. The homogenates were added to a tube with pre‐washed magnetic beads and incubated at room temperature for 1 h with mixing. The beads were subsequently collected with a magnetic stand, washed three times with lysis buffer and eluted with SDS‐PAGE sample buffer. Finally, the referenced samples were boiled and subjected to western blotting analysis with the p11 and Munc13‐1 antibodies.

4.8. Proximity Ligation Assay (PLA)

PLAs were performed using Duolink In Situ Red Starter Kit Goat/Rabbit (Sigma‐Aldrich DUO92105) according to the manufacturer's instructions. Briefly, brainstem slices of adult rats previously fixed with paraformaldehyde were deposited on glass slides and washed with PBS. Then, the samples were blocked in a humidity chamber with Duolink Blocking Solution for 1 h at 37°C and incubated overnight at 4°C with anti‐p11 and anti‐Munc13‐1 antibodies diluted in Duolink Antibody Diluent. Subsequently, slices were washed with wash buffer and incubated first with PLUS and MINUS probes of the kit for 1 h at 37°C and then with a ligation solution in a humidity chamber for 30 min at 37°C. Following this, the amplification buffer with the polymerase was added over slices for 100 min at 37°C, allowing the amplification of the signal. Some sections were additionally processed for vGLUT2‐immunofluorescence as below. The control samples were carried out under the same conditions by using secondary antibodies only. Finally, slices were mounted with a coverslip using a Duolink In Situ Mounting Medium with DAPI to counterstain nuclei and analyzed using a 63 × oil‐immersion objective of a ZEISS LSM 900 AIRYSCAN 2 confocal microscope for fluorescence detection.

4.9. Immunohistochemistry

Animals, under deep anesthesia (1.5%–3% isoflurane), were intraventricularly injected with heparin and immediately perfused transcardially with phosphate‐buffered saline (PBS), followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB), pH 7.4, at 4°C. The medulla was rapidly dissected and postfixed for 2 h in 4% PFA. Cryoprotection was performed by immersion in 30% sucrose in 0.1 M PB (4°C, overnight). Serial coronal sections (30‐μm thick), obtained using a microtome, were kept at −20°C in a cryoprotectant solution (glycerol/PBS, 1:1 v/v) until processing. After PBS washing, sections were immersed in a blocking solution [2.5% (w/v) bovine serum albumin and 0.1% (v/v) Triton X‐100 in PBS] for 30 min, followed by incubation (4°C, overnight) with the polyclonal primary antibodies against GFAP (1:250; Santa Cruz Biotechnology, Cat# sc‐58766, RRID:AB_783554) developed in mouse, p11 (1:100, R&D systems, Cat# AF2377, RRID:AB_2183469) developed in goat, vGLUT2 (1:2000, Millipore Bioscience Research Reagents, Cat# AB2251, RRID:AB_1587626) developed in guinea pig, Munc13‐1 (1:200; Synaptic systems, Goettingen, Germany, Cat# 126‐102, RRID:AB_887734), and Synaptophysin (Syn) (1:200; Zymed Laboratories, San Francisco, CA, Cat# 18‐0130) developed in rabbit. Next, the tissue was rinsed with PBS and incubated for 1 h at room temperature with Cy2‐conjugated, Cy3‐conjugated, and/or Cy5‐conjugated anti‐goat, anti‐mouse, anti‐guinea pig, and/or anti‐rabbit IgGs (1:200; Jackson ImmunoResearch Laboratories) as secondary antibodies. Finally, sections were washed with PBS and mounted on slides with a solution of propyl gallate (0.1 mM in PBS/glycerol, 1:9 v/v). Slides were analyzed using a ZEISS LSM 900 with AIRYSCAN 2 confocal microscope for fluorescence imaging. Omission of the primary antibodies resulted in no detectable staining.

Isolated SMNs were plated on 24‐mm Corning glass coverslips (Corning, NY) placed in 35‐mm culture dishes for immunocytochemistry (50 000 SMNs per well). The cultures of SMNs were fixed in 4% PFA (15 min at room temperature) and 100% methanol (10 min at 4°C). After PBS washing, SMNs were blocked [5% (w/v) bovine serum albumin, and 0.2% (v/v) Triton X‐100 in PBS] for 1 h, followed by incubation overnight at 4°C with polyclonal antibodies as above. Micrographs were obtained using a ZEISS LSM 900 AIRYSCAN 2 confocal microscope.

Images for quantification were acquired under a 40 × oil‐immersion objective through a z‐plane in which maximum antibody penetration was evidenced. Immunofluorescent synaptic puncta denoting clusters of presynaptic protein markers on motor neurons perikarya were analyzed in a z‐plane containing the nucleus as previously reported [93]. Pinholes were set to the same airy unit for each laser to obtain identical optical section thickness. For comparison between different experimental conditions, acquisition settings were kept identical. Animals and tissue were processed in parallel. In all cases, the area delimiting p11‐positive somata, dendrites, or neurites was manually traced. Treatments were blinded to the researcher who performed measures. For multichannel image acquisition, fluorophores were excited alternately. The perimeter, density, fluorescent intensity, and dimensions of immunolabeled structures were assessed with ImageJ 1.54 k (NIH). Images for quantification were flattened and background‐filtered to enhance cluster outlines and a user‐defined intensity threshold was applied to select puncta as previously described [93, 94]. Briefly, to delineate synaptic structures, Syn, vGLUT2, and/or Munc13‐1 channels were first processed for background subtraction to obtain the maximum dynamic range of intensities (from 0 to 255). Next, all images were subjected to the same smoothing filter and subsequently were binarized based on local differences in intensity. Only puncta reporting an area of > 0.04 μm² were taken as specific, as established before [93]. Synaptic puncta apposed to HMNs somata and in the neuropil were differentially analyzed. Immunoreactive (ir) puncta apposed to somata are presented as the number of synaptic structures per HMN and as linear density (puncta/100 μm of HMN perimeter). On the other hand, a grid composed of 16 squares (10 μm × 10 μm) was randomly distributed along the HN in each section to characterize puncta in the neuropil. Only those synaptic structures included in squares that did not contain somata were analyzed. Area density (puncta/100 μm²) in neuropil was compared between treatments. Treatments were blinded to the researcher who performed the measurements.

Furthermore, we analyzed the overlapping degree between p11, Syn, Munc13‐1, and/or vGLUT2 immunoreactivity in confocal images. To this end, we used the BIOP‐JACoP plugin of Image J 1.54 k, which enabled us to determine the overlapping area of two proteins relative to the total area for each protein. The immunofluorescent signals located in neuronal somata were excluded. The ratio between the two areas was compared among the different treatments.

4.10. Transmission Electron Microscopy

Adult rats were deeply anesthetized with 1.5%–3% isoflurane 3–5 days after oligonucleotide administration. The animals then received intraventricular heparin injection and transcardial perfusion with PBS, followed by 1%:1.25% PFA: glutaraldehyde in 0.1 M PB, pH 7.4, at 4°C. Brainstems were removed and postfixed for 2 h in the same fixative solution. Next, coronal sections (~60–80 μm thick) were cut in a vibratome, washed in PBS and postfixed with 1% osmium tetroxide in 0.1 M PB plus 7% glucose for 120 min at room temperature. Slices were then rinsed with distilled water, dehydrated in a graded series of ethanol, 30% (twice, 5 min), 50% (twice, 10 min), and 70% (10 min), and bloc‐stained for 2 h and 30 min in 2% uranyl acetate in 70% ethanol at 4°C. Sections were dehydrated again in a graded series of ethanol, 70% (5 min), 96% and 100%; then washed with propylene oxide, embedded in Durcupan resin, and left to polymerize for at least 24 h at 70°C. Ultrathin sections (60–80 nm thick) obtained with an ultramicrotome (Leica EM UC6) were analyzed under a TEM Zeiss LIBRA 120 electron microscope at high magnification (43 000 ×) attached to a digital camera (Morada, Soft Imaging System, Olympus, Japan). Captured images were analyzed offline using ImageJ. HMNs were analyzed at the level of the nucleolus.

4.11. Lentiviral Constructs, Oligonucleotides, and Administration

Neuron‐specific LVVs were designed, constructed, and produced under the supervision of Dr. S. Kasparov (Bristol University, Bristol, UK) during a research stay of G.R.B. in his lab. They are the same as those used in parallel studies previously reported [11]. Briefly, lentiviral constructions were made based on the vector pTYF‐mCMV/SYN. This system uses a transcriptional amplification strategy (TAS) to bidirectionally enhance the transcriptional activity of human synapsin‐1 (SYN) promoter by joining upstream a minimal core promoter (65 bp) derived from the human cytomegalovirus (mCMV) in the opposite orientation. This construction creates a compact TAS‐amplified neuron‐specific promoter useful for the development of powerful gene expression systems. Two cloning steps were necessary to generate the LVV‐shRNA_lacz plasmid. First, the shRNA_lacz sequence (5′‐AAATCGCTGACTTGTGTAGTCGTTTTGGCCACTGACTGACAACCACACATCAGCGATTT‐3′) was cloned into the vector pcDNA6.2‐WemGFP (Invitrogen) to obtain the shRNA_lacz‐GFP (green fluorescent protein) plasmid. Finally, a PCR product containing the shRNA_lacz‐GFP was cloned into the vector pTYF‐CMV/SYN to generate the final lentiviral plasmid. This same procedure was also followed to obtain the LVV‐shRNA_p11 plasmid using the shRNA_p11 sequence (5′‐TAGTCATTGCATGCAATGGTGGTTTTGGCCACTGACTGACCACCATTGTGCAATGACTA‐3′). Transient co‐transfection of HEK293FT cells with shuttle plasmids, the packaging vector pNHP, and the envelope plasmid pHEF‐VSVG was performed to produce the LVV plasmid system derived from HIV‐1 used for this study. The virus‐containing medium was harvested 48 or 72 h after transfection and filtered using a 0.45 μm filter (Millipore). Filtered medium was carefully added on top of a 20% sucrose solution and subsequently centrifuged at 75 000 × g for 2 h. After removal of the supernatant, the pellet was resuspended in PBS. Finally, the titer of the virus (> 10⁹ iu/μL) was determined using the ELISA kit Lenti‐X p24 rapid Titer (Clontech), which measures the amount of lentiviral capsid protein p24. Cells were transduced with the different LVVs at a five times multiplicity of infection.

Adult rats under deep anesthesia (1.5%–3% isoflurane in 100% O₂) were placed in a stereotaxic device and surgically prepared to receive a single injection of LVVs or vehicle (sterile PBS) [11, 95]. Briefly, a midline dorsal incision exposed the caudal dorsal medulla and, subsequently, glass micropipettes (∼25 μm‐tip diameter), connected to an oil‐filled system, were visually advanced to perform microinjections of the LVVs (Figure 1A). LVVs were microinjected at least 4 days before HMN recordings since a delay of 3–4 days is expected for effective action of LVV‐shRNA_p11 [96]. Furthermore, this schedule increases the success rate of the experiment, aiming to minimize the number of animals required. Four days before electrophysiological recordings or tissue extraction for WB, a microinjection (0.5 μL) of LVV‐shRNA_lacz, LVV‐shRNA_p11, or the vehicle was performed in the midline and 0.7 mm below the dorsal surface of the medulla at the obex level. The intramedullary injection rate was 0.1 μL/min and the micropipette remained in situ for 10 min before being slowly withdrawn at the end of injection. The incision was sutured and cleaned with an aseptic solution (povidone‐iodine).

Sequence‐specific gene silencing using siRNA is a promising technique with potential clinical applications, enabling the inhibition of target genes’ expression and knockdown of associated proteins both in vitro and in vivo [32, 97]. Here, we have used a modified siRNA that induces a rapid and targeted neuron‐selective knockdown of the selected genes in the CNS after a single intracerebroventricular injection in the adult rat, without off‐target effects, toxicity, and recruitment of immune responses [32]. The siRNA_p11 (Accell, Dharmacon Inc., Lafayette, CO) or the cRNA used in this work were the same as we previously characterized [11]. The target sequence for the siRNA_p11 was 5′‐CGGUUGCCCUCAAUAAAGA‐3′. Neonatal (P5) and adult rats were anesthetized (1.5%–3% isoflurane in 100% O₂) and placed in a Kopf stereotaxic instrument. The needle of a microsyringe (5 μL, Hamilton Company, Tokyo, Japan) crossed the skull through the middle point of the interparieto‐occipital suture and was advanced parallel to the occipital bone up to the fourth ventricle (Figure 3A). The final position of the needle ending was confirmed visually employing a surgical microscope. The animals then received a single injection of siRNA_p11 or cRNA (P5: 2 μg/2 μL per pup, adult: 5 μg/5 μL per rat) dissolved in RNase‐free PBS at a rate of 0.5 μL/min. After the injection, the needle was left in place for 5 min, then slowly removed. Brainstems of treated rats were quickly extracted at least 2 days after oligonucleotide administration for qRT‐PCR, WB, immunohistochemistry, and electrophysiological analyses. Rostral and caudal portions of the brainstem region containing the HN, obtained from P7 rats used for electrophysiological studies, were additionally processed for qRT‐PCR. Only recorded HMNs from animals that reported at least 40% knockdown of mRNA_p11 were included in this study.

Primary cultures of SMNs were established as above. After 72 h, cells were triple washed with PBS to remove cell debris. In separate tubes, 5 μL of either cRNA or siRNA_p11 (each 100 μM) were mixed with 250 μL Accell siRNA Delivery Media supplemented with B27 (2% v/v) and a cocktail of recombinant neurotrophic factors as described above. After the addition of siRNA oligos, SMNs were incubated for 72 h at 37°C according to the manufacturer's protocol. Subsequently, the transfection medium was removed, and the cells were cultured in Neurobasal complete medium as described above. The sequence for siRNA_p11 in this experimental series was 5′‐CUUCUGAGUUUUAUAUUGU‐3′.

4.12. Statistics

Summary data are all presented as mean ± SEM. The number of analyzed specimens per experimental paradigm, and statistical tests applied to each data set are indicated in figure legends, in the results section, or in tables. Statistical analysis was performed using SigmaPlot (Systat Software Inc.). The minimum significance level was set at p < 0.05. Statistical tests were employed for all data sets with similar variance. For comparison between two groups, normally distributed data were analyzed by unpaired Student's t‐test, unless otherwise stated, while non‐parametric data sets were assessed by Mann–Whitney U test. One‐way or two‐way ANOVA followed by post hoc Holm–Sidak method was employed for comparison of three or more groups that passed the normality test. Box‐plots show the 25%–75% interquartile range as the box, the median (horizontal line), and the whiskers indicate 5%–95% range. No data points were excluded from the statistical analysis unless otherwise noted.

Author Contributions

B.M.‐L. conceptualized, designed the study, and wrote the paper. D.G.‐F. supervised electrophysiological experiments and reviewed the first draft of the paper. E.V.‐H., G.R.‐B., Á.G.‐C., and R.G.H. performed experiments, data analysis, and figures. Á.M.P. supervised electron microscopy experiments. A.C.‐C. supervised co‐immunoprecipitation and PLA assays. B.M.‐L., D.G.‐F., and Á.M.P. obtained the funding. All authors participated in the critical review of the manuscript.

Funding

This publication is part of the grants PID2019‐110960GB‐I00 and PID2023‐152180NB‐I00, funded by MICIU/AEI/10.13039/501100011033 and by ERDF/EU to BML and DGF and PID2021‐124300NB‐I00 to AMP. RGH is a Ramón y Cajal fellow (RYC2021‐033058‐I) funded by the MICIU/AEI and the NGEU to RGH.

Ethics Statement

Animals were obtained from an authorized supplier (Animal Supply Services, University of Cádiz, Cádiz, Spain). Animal care and handling followed the guidelines of the European Union Council (2010/63/EU, 86/609/UE) on the use of laboratory animals. Experimental procedures were approved by the local Animal Care and Ethics Committee (University of Cadiz, Cadiz, Spain) and the Counseling of Agriculture, Livestock, Fisheries and Sustainable Development (Project N. 11/05/2022/064, Junta de Andalucía, Spain).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting Information.

Vilches‐Herrando E., Rodríguez‐Bey G., Hernández R. G., et al., “The Calcium‐Binding Protein S100A10 (p11) Is Required for Normal Motor Performance by Regulating Vesicle Dynamics at Excitatory Synapses,” Acta Physiologica 242, no. 2 (2026): e70158, 10.1111/apha.70158.

Data Availability Statement

The datasets generated and/or analyzed during the current study are available from the first author and the corresponding authors on reasonable request and in Supporting Information.