(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Amyloid precursor protein modulates cerebellar Purkinje cell activity and motor function through regulation of Nav1.6 currents [1]

['Miao-Jin Ji', 'Jiangsu Province Key Laboratory Of Anesthesiology', 'Brain Science', 'Nmpa Key Laboratory For Research', 'Evaluation Of Narcotic', 'Psychotropic Drugs', 'School Of Anesthesiology', 'Xuzhou Medical University', 'Xuzhou', 'Tong-Xuan Wu']

Date: 2025-12 

Amyloid precursor protein (APP)-null mice exhibit significant deficits in motor performance, including reduced grip strength and impaired locomotion; however, the underlying neurophysiological mechanisms remain unclear. In this study, we show that conditional knockdown of APP selectively in Purkinje cells (PCs) recapitulates these motor deficits, while exogenous expression of APP in APP-null mice rescues motor function. Electrophysiological analysis revealed that APP deficiency leads to aberrant firing patterns in PCs and reduces inhibitory synaptic transmission onto neurons of the deep cerebellar nucleus (DCN). We identified a marked reduction in Nav1.6-mediated sodium currents as the key mechanism underlying abnormal action potential firing and propagation in APP-deficient PCs. Importantly, all electrophysiological and behavioral deficits were rescued by PC-specific APP reconstitution. These findings reveal a novel and essential role for APP in cerebellar motor control by regulating Nav1.6 channel activity and PC excitability.

Funding: This work was supported by the National Natural Science Foundation of China ( https://www.nsfc.gov.cn/ ) (32271010 to CL; 81974157 to CL), the Fund for Jiangsu Province Specially-Appointed Professor (to CL), the Starting Grant for Excellent Talents in Xuzhou Medical University ( https://www.xzhmu.edu.cn/ ) (to CL), and the Innovative Youth Team Project in Xuzhou Medical University ( https://www.xzhmu.edu.cn/ ) (TD202203 to MJJ). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

APP is abundantly expressed in cerebellar PCs [ 20 ]. Our prior study showed that APP increases Nav1.6 currents by enhancing its cell surface localization in vitro [ 21 ]. However, whether APP regulates the physiological function of cerebellar PCs and participates in motor control remains unclear. In this study, we investigated the role of APP in cerebellar PCs by analyzing the motor function deficits, PC firing patterns, and cerebellar circuit dysfunctions in APP-null mice. We found that APP in cerebellar PCs is critical for PC firing through the regulation of Nav1.6 and plays an essential role in PC-mediated motor control.

The cerebellum is a crucial subcortical region involved in motor control, comprising the cerebellar cortex, DCNs, and medulla. Purkinje cells (PCs) in the cerebellar cortex receive input from mossy fibers and climbing fibers, integrating these signals and projecting to DCN neurons. As the sole output neurons of the cerebellar cortex, PCs are critical for encoding motor function. In humans, cerebellar activity increases linearly with increasing grip force in the ipsilateral limb [ 14 ]. Patients with cerebellar lesions display muscle force production deficits [ 15 ]. The action potential (AP) firing of PCs is spontaneous and of high frequency, dependent on voltage-gated sodium channels. In the cerebellar cortex, Nav1.6 and Nav1.1 are predominant in PCs, while Nav1.2 is primarily expressed in cerebellar granule cells [ 16 ], as demonstrated by morphological [ 17 ] and electrophysiological studies [ 18 ]. Nav1.6 accounts for ~70% of persistent sodium currents and ~82%–92% of resurgent sodium currents in PCs [ 18 ]. Three types of sodium currents are typically studied in PCs: resurgent currents, responsible for the generation of spontaneous firing; transient sodium currents, contributing to the upstroke of APs, and persistent sodium currents, which regulate plateau potentials and amplify dendritic input [ 19 ].

Amyloid precursor protein (APP) is a type I transmembrane protein processed by secretases to generate β-amyloid peptides and other fragments. While APP is best known for its central role in Alzheimer’s disease pathogenesis, its physiological functions remain incompletely understood. Recent studies have demonstrated that APP plays critical roles in neuronal development [ 1 , 2 ], synaptic function [ 3 ], and intracellular signaling [ 4 ]. These processes influence a wide range of central nervous system-regulated functions, including learning, memory [ 5 – 7 ], and motor functions [ 8 – 10 ]. The APP-null mice exhibited impaired motor performance, learning capacity, and poor memory [ 5 – 7 ]. While the role of APP in cognitive impairment has been extensively studied, its involvement in motor control remains less explored. Interestingly, in contrast to other movement disorder models, APP-null mice specifically display deficits in locomotor activity and grip strength [ 8 , 9 ]. Initial hypotheses suggested that APP loss disrupts neuromuscular junctions. However, APP single-knockout mice exhibited no significant alterations in neuromuscular transmission or synaptic morphology at the neuromuscular junctions [ 11 ], suggesting that APP may instead regulate higher-order motor control circuits rather than the function of neuromuscular junctions. In mammals, voluntary movement is orchestrated by the corticospinal tract, a motor pathway connecting the motor cortex to lower motor neurons and involving multiple regions, including the motor cortex, subcortical nuclei, cerebellum, brainstem, and spinal cord [ 12 ]. Conditional APP knockout under the Nex promoter—driving Cre expression in corticospinal tract-associated regions, including the motor cortex, superior and inferior colliculi, medulla oblongata, pons, cerebellar granule cell layer, and deep cerebellar nuclei (DCNs) [ 13 ], does not impair grip strength or locomotor activity [ 7 ], suggesting that APP in these brain regions is unlikely to account for its role in motor function regulation.

To further examine the role of Nav1.6 in AP transduction, we performed sIPSCs recordings during perfusion with 4,9 ah-TTX. This Nav1.6 blocker reduced the amplitude and the frequency of sIPSCs in all groups ( Fig 7D ). Importantly, application of 4,9 ah-TTX almost completely abolished the rescue effect of APP expression in GABAergic transmission of PC-DCN neurons ( Fig 7D ). These findings indicate that APP plays a crucial role in maintaining the inhibitory synaptic transmission from PCs to DCN neurons, and this function was Nav1.6-dependent. In addition, mIPSCs were also analyzed to identify whether APP deficiency influenced the AP firing-independent synaptic transmission. No significant difference in the amplitude or the frequency of miniature inhibitory post-synaptic currents (mIPSC) was observed among WT, APP-null + vec, and APP-null + APP groups ( Fig 7E ).

(A) Schematic of the viral injection strategy to achieve APP overexpression in PCs and trans-synaptic tracing of downstream neurons in the DCN. (B) Representative example of fluorescence-assisted selection of cells for patch-clamp recordings. Images from an acute cerebellar slice observed in differential interference contrast (DIC, left), or fluorescence mode (EGFP, right). (C) Left panel: Representative traces of sIPSCs in the PC-projected DCN neurons held at 0 mV. Brain slices of App +/+ + vec, App −/− + vec and App −/− + APP mice were recorded; GABA B -receptor antagonist picrotoxin (PTX, 100 μM) was perfused at the end of the recordings to validate IPSCs. Right panel: statistics of IPSCs frequency and amplitude, n = 10 cells for each group. (D) IPSCs recorded from the PC-projected DCN neurons with 4,9-ahTTX (200 nM), n = 8 cells for each group. (E) Miniature IPSCs (mIPSCs) recorded from PC-projected DCN neurons in the presence of TTX (1 μM), n = 8 for each group. Scale bars as presented in the images. A 5-min recording segment from a stable period, defined as commencing at least 10 min after obtaining whole-cell access, was used for statistical analysis. Data are represented as the mean with SEM (C, D IPSCs amplitude) or the median with interquartile range (D IPSCs frequency, E). One-way ANOVA and Kruskal–Wallis test. * P < 0.05; ** P < 0.01; *** P < 0.001; ns, non-significant. The data underlying this Figure can be found in S1 Data .

PC projections are the sole output of the cerebellar cortex and represent the major input to DCN neurons [ 29 ], forming approximately 70%–80% of all synaptic connections onto DCN cells [ 30 , 31 ]. To explore whether inhibitory transmission from PCs to DCN neurons was affected by APP deficiency, we performed whole-cell patch clamp recordings in cerebellar slices. DCN neurons innervated by PCs were labeled using trans-synaptic labeling, achieved by injecting AAV2/1-hSyn-CRE into the cerebellar cortex and AAV2/9-hSyn-DIO-EGFP in the DCN ( Fig 7A ). After 3 weeks of viral expression, EGFP-positive DCN neurons were recorded in brain slice ( Fig 7B ). No significant difference was observed in the number of EGFP-labeled DCN neurons between the different virus-injected groups. The amplitude and the frequency of spontaneous inhibitory post-synaptic currents (sIPSCs) in APP-null DCN neurons were lower than those in WT DCN neurons, and the abnormal GABAergic synaptic transmission APP-null mice was restored by APP expression in PCs ( Fig 7C ).

In the analysis of firing properties, we found that exogenous APP expression in APP-null PCs restored spontaneous firing patterns nearly to the WT level. Complex spikes were observed in the APP-expressing group (2/21, 9.5%), but not in the control group ( Fig 6E ). The spontaneous firing frequency of the APP-expressing group was almost restored to that of WT mice ( Fig 6F ). Additionally, exogenous APP-expression in APP-null PCs restored the rheobase ( Fig 6G ). Moreover, the spontaneous AP spikes from exogenous APP-expressing PCs displayed higher peak amplitude, lower threshold, shorter half-width, and smaller AHP ( Fig 6H and 6I ). These data indicate that exogenous APP expression in PCs restored Nav1.6 currents and firing properties in APP-null mice.

(A, B) Representative traces for whole-cell recording of (A) transient Na + current and (B) statistics of whole-cell I–V curve, peak transient currents, and activation curves in PCs of App +/+ + vec, App −/− + vec, and App −/− + APP mice. N = 8 cells for each group. (C) Representative persistent Na + current recorded in PCs. The right panels show statistics of the peak persistent currents and the voltage eliciting the peak currents. App +/+ + vec, n = 6; App −/− + vec, n = 8; App −/− + APP, n = 6 cells. (D) Representative resurgent sodium currents and I–V curves recorded in PCs. App +/+ + vec, n = 8; App −/− + vec, n = 8; App −/− + APP, n = 8 cells. (E) Pie chart showing the percentage of PC firing patterns in App +/+ + vec, App −/− + vec, and App −/− + APP mice. (F) Representative spontaneous simple spikes recorded in PCs. Quantification of firing rates analyzed from 1-min recordings during stable whole-cell access (≥10 min after breakthrough). App +/+ + vec, n = 12; App −/− + vec, n = 9; App −/− + APP, n = 13 cells. (G) Representative membrane potential responses to depolarizing ramp current injections (−50 to +500 pA over 1 s) recorded from PCs under whole-cell current-clamp. Right: Rheobase quantification. N = 9 cells/group. (H) Representative images of spontaneous action potentials and the dV/dt curves. (I) Statistics of peak amplitude, threshold, half-width, and afterhyperpolarization potential (AHP) of spontaneous action potentials shown in (H), App +/+ + vec, n = 10; App −/− + vec, n = 8; App −/− - + APP, n = 8 cells. Scale bars are indicated in the images. One-way ANOVA and post-hoc Turkey’s test. * P < 0.05; ** P < 0.01; *** P < 0.001. The data underlying this Figure can be found in S1 Data .

We next investigated whether the restoration of the behavioral phenotypes in APP-null mice via exogenous APP expression was linked to the improvement in Nav1.6 function in PCs. Whole-cell patch clamp recordings revealed that transient, persistent, and resurgent I Na were all rescued by exogenous APP expression in APP-null PCs. The peak transient I Na was restored to levels comparable to WT mice ( Fig 6A and 6B ). APP expression in PCs did not affect the activation curve ( Fig 6B ). Decreased persistent and resurgent current were also restored by APP expression in PCs ( Fig 6C and 6D ). These results indicated that exogenous expression of APP in PCs restore the Nav1.6 currents.

(A–D) Electrophysiological characterization of PC spontaneous firing in acute cerebellar coronal slices (300-µm-thick). (A) Representative traces and quantification of spontaneous firing rate and spike amplitude before/during application of 200 nM 4,9-ahTTX in App +/+ mice. n = 7 cells. (B) Representative traces and quantification of spontaneous firing rate and spike amplitude before/during application of 30 nM PoTX (Nav1.6 positive allosteric modulator) in App −/− mice. n = 5 cells/group. (C, D) Representative traces and quantification of spontaneous firing rate and spike amplitude before/during application of 350 nM ICA (Nav1.1 inhibitor) in App +/+ (C) or App −/− (D) mice. Statistics: Paired t test; * P < 0.05; ** P < 0.01; *** P < 0.001 vs. baseline; ns, non-significant. n = 4 cells/group. Scale bars are indicated in the images. (E) Morphology verification of cerebellar cannula placement in a coronal section of mice cerebellum. (F) Open field locomotion: Total distance traveled (m) in App +/+ (4,9-ahTTX/vehicle) and App −/− (PoTX/vehicle) mice. Mice received bilateral microinjections of saline (vehicle), 200 nM 4,9-ahTTX, or 20 pM PoTX into the cerebellar cortex 20 min prior to testing. (G) Forelimb grip strength: maximal force/body weight in same cohorts as (F). Statistics (F, G): App +/+ + vehicle, n = 10; App +/+ + 4,9-ahTTX, n = 8; App −/− + vehicle, n = 8; App −/− + PoTX, n = 9; One-way ANOVA with Tukey post-hoc; * P < 0.05; ** P < 0.01, *** P < 0.001; ns, non-significant. ctrl, control; 4,9-ahTTX, 4,9-anhydrotetrodotoxin; ICA, ICA-121431; PoTX, poneratoxin. The data underlying this Figure can be found in S1 Data .

However, Nav1.6 mRNA (Scn8a) levels in PCs did not differ between WT and APP-null mice ( S6 Fig ), suggesting that APP may not influence Scn8a gene expression. Co-immunoprecipitation (co-IP) assays in transfected HEK293 cells ( S7 Fig ) confirmed a specific physical interaction between APP and Nav1.6 ( Fig 4G ). This binding was dependent on holo-APP, as neither the soluble ectodomain sAPPα nor the AICD co-precipitated with Nav1.6 ( Fig 4H ), consistent with the requirement of holo-APP for functional rescue observed in vivo ( Fig 2 ).

(A) Representative traces for whole-cell recording of transient Na + currents (I Na ) in PCs of App +/+ or App −/− mice. (B) Whole cell I–V curve, peak currents, and activation curves of transient I Na as shown in (A). App +/+ , n = 8; App −/− , n = 14 cells. (C, D) Representative traces and statistical bar-charts showing amplitude of transient I Na (I NaT ) before or during bath application of (C) Nav1.6-specific blocker 4,9-anhydrotetrodotoxin (4,9-ahTTX, 200 nM) or (D) Nav1.1-specific inhibitor ICA-121431 (ICA, 350 nM). N = 6 cells for each group. The I Na1.6 or I Na1.1 were determined by subtracting the sodium currents recorded during bath application of specific blockers/inhibitors from that before drug application. (E) Representative persistent Na + currents recorded in PCs. The right panels show statistics of the peak persistent currents and voltage potentials eliciting peak persistent Na + currents. App +/+ , n = 6; App −/− , n = 12 cells. (F) Representative traces (the left panel) and statistics for whole-cell recording of resurgent I Na in PCs. The middle panel: Whole cell I–V curve of resurgent I Na ; the right panel: peak resurgent currents. App +/+ , n = 9; App −/− , n = 11 cells. Student t test; * P < 0.05; ** P < 0.01; ns, nonsignificant. Scale bars are indicated in the images. (G) APP co-immunoprecipitated with Nav1.6 but not Nav1.1. HEK293 cells co-expressing Flag-tagged APP and either Nav1.6 or Nav1.1-EGFP were lysed and immunoprecipitated (IP) with anti-Flag affinity beads (Smart-Lifesciences, #SA042001). Immunoblots (IB) were probed for: Nav1.6 (Alomone Labs #ASC-009), EGFP (Roche Applied Science, #11814460001) or Flag (AlpVHHs, #016-303-005). (H) Nav1.6 interacts with full-length APP (holo-APP), but not soluble APPα (sAPPα) and APP intracellular domain (AICD). HEK293 cells co-expressing Nav1.6 and Flag-tagged holo-APP, sAPPα, or AICD were lysed and IPed with anti-Flag affinity beads. IB were probed for Nav1.6 or Flag. Mock transfected controls are shown in the first lane. Input controls (7.5% lysate) shown below/beside corresponding lanes. Data representative of 3 biological replicates. The data underlying this Figure can be found in S1 Data .

Consistent with APP enhancing Nav1.6 currents in heterologous systems [ 21 ], whole-cell recordings in cerebellar slices revealed altered sodium currents in APP-null PCs. Transient (I NaT ), persistent (I NaP ), and resurgent (I NaR ) currents were evoked using established protocols ( S4 Fig ) [ 18 ]. Given that Nav1.6 and Nav1.1 are the predominant voltage-gated sodium channels in mouse PCs [ 16 ], we next sought to determine which of these isoforms was affected by APP deficiency. To this end, we applied subtype-specific blockers/inhibitors: 4,9-anhydrotetrodotoxin (4,9-ah-TTX; selective Nav1.6 blocker) and ICA-121431 (Nav1.1 inhibitor) [ 19 ].

(A) Representative spontaneous firing patterns recorded in PCs of wild-type (App +/+ ) mice. Pie charts in the right panel showing the percentage of PC firing pattern. App +/+ , n = 22; App −/− , n = 19 cells. (B) Representative spontaneous simple spikes recorded in PCs. Quantification of firing rates analyzed from 1-min recordings during stable whole-cell access (≥10 min after breakthrough). App +/+ , n = 9; App −/− , n = 8. (C) Left: Representative voltage responses to depolarizing ramp current injections (−50 to +500 pA over 1 s) recorded PCs under whole-cell current-clamp. Right: Rheobase quantification (minimal current eliciting first action potential). n = 8 cells/group. (D) Representative spontaneous action potentials and the corresponding dV/dt curve. (E) Statistics of peak amplitude, threshold, half-width, and afterhyperpolarization potential (AHP) of action potentials shown in (D), App +/+ , n = 10; App −/− , n = 9. Scale bars are indicated in the images. Data are represented as the mean with SEM. Student t test: * P < 0.05; *** P < 0.001. The data underlying this Figure can be found in S1 Data .

Next, we investigated whether loss of APP altered the firing patterns of PCs. As established, PCs display three characteristic firing states: silent, simple spikes (SSs), and complex spikes (CSs) [ 18 , 25 , 26 ]. Most of 22 PCs (19/22, 86.4%) recorded from WT mice exhibited spontaneous firing, while 13.6% (3/22) of PCs maintained silent within 1 h of recording. Complex spikes recorded in 27.3% (6/22) of PCs from WT mice ( Fig 3A ). The percentage of silent PCs increased to 42.1% (8/19, 42.11%), and no CSs was recorded in 19 PCs from APP-null mice ( Fig 3A ). We further analyzed the spontaneous firing rate and various AP characteristics, including threshold, peak amplitude, half-width, and afterhyperpolarization potential (AHP) of PCs with SSs in brain slices recorded from APP-null and WT mice. The spontaneous firing rate of PCs in APP-null mice was significantly lower than that in WT mice ( Fig 3B ). The rheobase of AP in APP-null PC was higher than that in WT mice ( Fig 3C ). In addition, lower peak amplitude, higher threshold, longer half-width, and bigger AHP ( Fig 3D and 3E ) were observed in the spontaneous AP spikes of APP-null than those of WT PCs. These results indicate that APP-deficiency significantly impairs the firing properties of cerebellar PCs.

APP plays a crucial role in neurodevelopment [ 1 , 2 ], so we examined whether APP deficiency impaired the development of PCs. Golgi staining and Sholl analysis revealed that the dendritic branches and spine density were similar between 2-month-old APP-null and WT littermates (S3A–S3C Fig). Further, we compared the morphology of PC in 12-month-old mice. No significant difference in the dendritic branches and spine density was observed between APP-null and WT mice (S3D–S3F Fig). These results indicated that APP deficiency affects neither PC development nor provokes their degeneration.

(A) Lentivirus-mediated overexpression of full-length APP (holo-APP), sAPPα, or AICD in the cerebellar cortex of App −/− mice. Representative images showing co-expression of APP fragments/mCherry (red) and the PC marker calbindin (green) from coronal sections of the cerebellum. Scale bars: 100 µm (main); 40 µm (insets). (B) QPCR analysis of LCMed PCs (mcherry + ) from fresh-frozen sections. Top: Primer specificity schematic for APP fragments. Bottom: APP fragments mRNA level normalized to Gapdh. Data presentation: median with interquartile range. App −/− + vec, n = 4; App −/− + holo-APP, n = 5; App −/− + sAPPα, n = 4, App −/− + AICD, n = 4 samples. Kruskal–Wallis test. (C, D) Motor phenotyping of App −/− mice with Purkinje APP fragments rescue (n = 9 mice/group): (C) Open field total distance (10-min test). (D) Forelimb grip strength. N = 9 for each group. Data presentation: mean with SEM. Statistics: one-way ANOVA. ** P < 0.01; ns, non-significant. The data underlying this Figure can be found in S1 Data .

Proteolytic processing of full-length APP generates bioactive fragments including sAPPα (sufficient for spatial learning and long-term potentiation rescue in APP-null mice) [ 9 ] and APP intracellular domain (AICD) (with transcriptional activity). To determine whether the intact holoprotein is required for motor functions, we delivered holo-APP, sAPPα, or AICD to APP-null mice via cerebellar lentivirus injection. Fluorescence imaging confirmed targeted expression in PCs ( Fig 2A ). LCM-coupled RT-qPCR verified successful APP reconstitution in transduced neurons ( Fig 2B ). Behavioral analyses demonstrated that only full-length APP (holo-APP) significantly rescued motor deficits, exhibiting significant improvement in open field ambulation ( Fig 2C ) and grip strength ( Fig 2D ) in APP-null mice. sAPPα and AICD provided no significant rescue ( Fig 2C and 2D ). These results establish that full-length APP is necessary for cerebellum-mediated motor function.

To further investigate the contribution of APP to motor control, we utilized mice with conditional APP knockdown restricted to PCs. We achieved PC-specific App knockdown by co-injecting WT mouse cerebellar cortex with AAV2/9-L7-Cre [ 23 ] and AAV2/9-DIO-(U6-EGFP)-shApp [ 24 ], whereby Cre recombinase activates the expression of shRNA targeting App specifically in L7-positive cells. Immunofluorescence confirmed efficient PC-specific transduction mediated by AAV under the L7 promoter, as shown by its colocalization with calbindin ( Figs 1G and S2 ). LCM coupled with RT-qPCR verified dramatic APP downregulation in transduced neurons ( Fig 1H ). Subsequent behavioral analyses revealed significant motor deficits in conditional knockdown mice, including reduced ambulation in open field tests ( Fig 1I ) and impaired forelimb grip strength ( Fig 1J ).

(A) APP (red) and calbindin (green) immunofluorescence in 20-µm-thick cerebellar sagittal sections. The APP immunofluorescence in wild-type (App +/+ ) or APP-null (App −/− ) mice were acquired under identical imaging settings. Insets: magnified views. Scale bars: 100 µm (main), 20 µm (insets). The PC contours are indicated by dashed lines. (B) Lentivirus-mediated overexpression of APP695 in the cerebellar cortex of App −/− mice. Representative images showing co-expression of APP/mCherry (red) and the PC marker calbindin (green) in a sagittal section of cerebellum. Scale bars: 100 µm (main); 20 µm (insets). (C) Left: representative fluorescence images showing before or after laser capture microdissection (LCM) of LV-transduced PCs (mcherry + ) from fresh-frozen coronal sections. Right: App mRNA levels in dissected cells detected by qPCR (Data presentation: median with interquartile range. App +/+ + vec, n = 8; App −/− + vec, n = 4; App −/− + APP, n = 7 samples; Kruskal–Wallis test). Scale bar: 40 µm. (D–F) Motor phenotyping of App −/− mice with Purkinje APP rescue: (D) Open field total distance (10-min test). (E) Forelimb grip strength. (F) Body weight of all groups of mice. App +/+ + vec, n = 12; App −/− + vec, n = 13; App −/− + APP, n = 12. (G) Conditional knockdown of App in App +/+ mice using a mixed AAV injection strategy. Representative images show the CRE (mCherry + ) and APP-shRNA (EGFP + ) expression in PCs in a coronal section of the cerebellum. Scale bar: 100 µm (main), 20 µm (insets). The PC contours are indicated by dashed lines. (H) Left: representative fluorescence images before or after LCM of AAV-transduced PCs (mcherry + ) from fresh-frozen coronal sections of App +/+ mice. Right: App mRNA levels in dissected cells detected by qPCR (control, n = 5, App kd, n = 10 samples). (I, J) Motor deficits following Purkinje APP knockdown: (I) Open field mobility. (J) Forelimb grip strength decline. (Control, n = 10; App kd, n = 8 mice). Data presentation: mean with SEM. Statistics: (D–F) one-way ANOVA; (H–J) Unpaired t test; * P < 0.05; ** P < 0.01; *** P < 0.001. The data underlying this Figure can be found in S1 Data .

APP is abundantly expressed in PCs of adult WT mice ( Fig 1A ). Given that cerebellar PCs play a critical role in motor function, we hypothesized that the deletion of PC-specific APP could be the primary cause of the motor dysfunction observed in APP-null mice. To specifically evaluate the effect of PC APP on motor performance, we injected a lentivirus encoding human APP695 with mCherry (LV-CAMK2A-SP-Flag-APP695-IRES2-mCherry) into the cerebellar cortex ( Fig 1B ). The CAMK2A promoter, which has been shown to drive gene expression specifically in PCs [ 22 ], was used in this experiment. Expression of mCherry was observed exclusively in the Calbindin-positive PCs ( Fig 1B ). Laser capture microdissection (LCM) coupled with Reverse Transcription Quantitative PCR (RT-qPCR) analysis further confirmed a significant increase of the APP expression in PCs ( Fig 1C ). PC-specific expression of APP in APP-null mice significantly improved their performance in locomotor activity ( Fig 1D ) and forelimb grip strength test ( Fig 1E ), but didn’t affect their body size ( Fig 1F ).

Consistent with previous study [ 8 , 9 ], we observed that APP-null (App −/− ) mice exhibited reduced locomotor activity ( S1A Fig ) and weaker forelimb grip strength ( S1B Fig ) compared with wild-type (WT, App +/+ ) littermates. Further, animals were tested for their capacity for motor learning and coordination using a rotarod and footprint test. APP-null mice showed normal motor learning over 3 days of training and their latency to fall was similar to that of WT mice in the test day ( S1C Fig ). During the footprint test, there was no differences between APP-null and WT mice in sway length and the overlap, while the stride length and the stance length were all significantly lower in APP-null group compared with that of WT (App +/+ ) group ( S1D and S1E Fig ). The reduced stride length in footprint test ( S1E Fig ) likely reflects smaller body size in APP-null mice ( S1F Fig ). To control for this confounder, we performed the balance beam assay. No significant differences were observed in traversal time or number of errors on 12- or 6-mm beams ( S1G Fig ), indicating intact gross motor coordination in APP-null mice. Thus, APP deletion specifically impairs open field locomotion and forelimb grip strength, independent of anthropometric differences.

Discussion

Our results demonstrate that the APP deficiency leads to aberrant sodium currents, altered firing patterns in PCs, reduced inhibitory synaptic transmission to DCN, and impaired motor function mediated by PCs. Specifically, we show that the reduction in Nav1.6 sodium currents is a key factor contributing to the motor deficits observed in APP-null mice. These defects were recapitulated by PC-specific APP knockdown and rescued upon exogenous expression of APP in PCs. Taken together, our findings reveal a novel role for APP in motor control and provide new insights into the molecular mechanisms underlying this process.

Motor deficits in grip strength and locomotion in APP-deficient mice have been reported by several groups in the past [8,32]. However, the underlying mechanisms have remained poorly understood. The absence of prominent learning and memory phenotypes in single APP knockout mice has been attributed to functional compensation by APP family members, APLP1 and APLP2, with more pronounced deficits observed in APP and APLP double knockout mice [10]. Interestingly, APLPs do not compensate for certain phenotypes seen in APP knockout mice, particularly those related to motor function and body size, suggesting that APP plays a unique role in the regulation of motor functions within the APP family. A key question remains: in which brain regions or cell types does APP mediate its effects on motor control? Within the motor system, APP knockout mice exhibit normal neuromuscular junction morphology [11]. Furthermore, mice with conditional APP knockout targeted to excitatory neurons via the Nex promoter show intact motor functions [7]. These observations guided our focus on cerebellar PCs to investigate APP’s role in motor control. Our exogenous rescue and conditional knockdown experiments demonstrate that APP in PCs is critical for motor function, appropriate expression of APP in PCs is essential for normal motor performance, particularly specific behaviors such as locomotion and forelimb grasping ability. However, it should be noted that this study focuses solely on cerebellar PCs; we cannot exclude the possibility that APP also modulates motor function via other regions of the cerebrospinal circuit, such as spinal motor neurons. Nav1.6-mediated persistent sodium current (I NaP ) contributes to postural tone and amplifies locomotor outputs in spinal cord bistable motor neurons [33]. Given that APP is also expressed in spinal motor neurons [34], it is likely that spinal APP may similarly participate in the regulation of motor function.

As a key component in the formation of internal models of the cerebellum, PCs have the capacity to process a vast array of signals [35]. The output information from PCs is encoded by two distinct types of APs: high-frequency SSs and low-frequency CSs. A single PC receives excitatory input from between 100,000 and 200,000 parallel fibers, which modulate the intrinsically driven high-frequency SS discharge [18,36]. In monkeys, increased SS discharge during grasping and lifting movements was observed in 56% of PCs [37], supporting the hypothesis that the cerebellum plans or controls movements through a forward internal model. Complex spikes are characterized by a large Na+ somatic spike followed by a burst of smaller spikelets, which are generated by a massive depolarization of the entire PC induced by climbing fiber activation [35]. These CSs serve as error signals crucial for online correction and motor learning [38]. Although the precise representations of PC firing patterns remain incompletely understood, there is a clear correlation between PC firing and motor output. Therefore, aberrant firing patterns in PCs are likely to influence both grip force and locomotion. Specifically, the decline in SS firing rates and the reduction in CS generation may contribute to the motor impairments, such as the deficits in grip strength and locomotion observed in APP-null mice.

We previously reported that APP interacts with Nav1.6 and regulates its function in HEK293 cells by enhancing its cell surface distribution [21]. Nav1.6 is critical to the electrophysiological activity of PCs, comprising 82%–92% of resurgent sodium currents, 69% of persistent sodium currents, and more than 37% of the transient sodium currents [18]. Previous studies have shown that Nav1.6 and Nav1.1 are two major sodium channel subtypes expressed in PCs. We employed selective blockers/inhibitors combined with eliciting protocols to isolate different sodium current components produced by Nav1.1 or Nav1.6. Our data indicate that APP regulates Nav1.6 rather than Nav1.1 currents in transient I Na . APP regulates the persistent and resurgent I Na , which is predominantly produced by Nav1.6. Thus, APP plays a crucial role in the AP firings of PCs by modulating Nav1.6 currents. We further demonstrated that PoTX, a positive allosteric modulator of Nav1.6 that reduces channel inactivation, consequently enhanced the persistent I Na —, key determinant of repetitive AP firing. This PoTX-mediated enhancement likely compensates for the functional deficits resulting from reduced surface distribution of Nav1.6 in APP-null PCs, as evidenced by the restoration of firing properties and amelioration of motor deficits. These results further support the conclusion that Nav1.6 serves as the primary effector mediating APP-dependent regulation of cerebellar PC function. However, as shown in Fig 2, APP-null mice exhibited a higher threshold for AP generation and an enhanced AHP. These electrophysiological alterations may collectively contribute to the observed reduction in spontaneous firing frequency of PCs. In addition to regulating Nav1.6, APP may also modulate the threshold for AP generation and AHP through other mechanisms. Additionally, Nav1.6 mediates the saltatory propagation of APs, reflected by the output of PCs. APP-null mice exhibited abnormal synaptic output to post-synaptic DCN neurons, contributing to impaired motor functions. Notably, the motor deficiencies observed in APP-null mice differ from those in Nav1.6-knockout mice (such as med mice). This discrepancy may be explained by the fact that Nav1.6 is only one of the downstream targets of APP, and the partial modulation of Nav1.6 by APP is not as dramatic as the complete knockout of the Scn8a gene.

In summary, this study builds on previous findings that APP-null mice exhibit motor function deficits and identifies a novel mechanism through which APP regulates motor control via Nav1.6 in PCs.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3003513

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/