Fig 2: N-terminus of FMRP interacts with Kv1.2 to control inhibitory output at BC nerve terminals. a Images of Kv1.2 immunostaining in cerebellar cortex showing reduced expression of Kv1.2 (green) in the Fmr1-KO BC axonal terminals. Individual pinceau marked with a single arrow (top) are revealed at a higher magnification (bottom). Dotted lines indicate Purkinje cell layer. ML molecular layer, GCL granule cell layer. b Comparison of the density of pinceau (top) and fluorescence intensity (bottom) of Kv1.2 labeling between WT and Fmr1 KO. c Representative western blot of Kv1.2 taken from the whole cerebellar homogenates of WT and KO samples. d Analysis of western blot revealing a significant decrease in Kv1.2 expression in KO (n = 7) compared to WT (n = 8) mice. e Quantitative RT-PCR analysis showed no difference in Kv1.2 mRNA abundance between the WT (n = 6) and KO (n = 6) cerebellum. f Co-IP by anti-Kv1.2 and anti-FMRP antibodies, indicating the two molecules interact at the protein–protein level. g Schematic depiction of paired recording configurations: top, presynaptic cell-attached mode for baseline measurements; bottom, infusion of N-terminus of FMRP binding antibody (n-FMRP-AB) or N-terminus FMRP fragment (n-FMRP) into the BC terminal until equilibrium. Presynaptic and postsynaptic holding potential was set at −80 and −60 mV, respectively. h–j sIPSCs (left panels) recorded from PNs before and 20 min after diffusion of n-FMRP-AB (NBP2-01770, Novus Biologicals, 1:2000 dilution) into a WT (h) or KO (i) synapse, and n-FMRP (H00002332-P01, Novus Biologicals, 1:100 dilution) into a KO (j) synapse. The amplitude and frequency of sIPSCs are summarized in the right panels for the conditions in h (n = 5), i (n = 6), and j (n = 6). k–o Averaged intra-terminal Ca2+ rise from 5 trials evoked by AP trains (100 Hz, 100 ms) delivered to the soma of BCs in KO synapses with (black) or without n-FMRP (control, grey). The amplitude (l), area integral (m), 10–90 rise time (n), and 90–10 decay time (o) of Ca2+ transients are summarized for the control (grey bars, n = 7) and n-FMRP (black bars, n = 5) groups

Fig 3: N-terminus of FMRP interacts with Kv1.2 to control inhibitory output at BC nerve terminals. a Images of Kv1.2 immunostaining in cerebellar cortex showing reduced expression of Kv1.2 (green) in the Fmr1-KO BC axonal terminals. Individual pinceau marked with a single arrow (top) are revealed at a higher magnification (bottom). Dotted lines indicate Purkinje cell layer. ML molecular layer, GCL granule cell layer. b Comparison of the density of pinceau (top) and fluorescence intensity (bottom) of Kv1.2 labeling between WT and Fmr1 KO. c Representative western blot of Kv1.2 taken from the whole cerebellar homogenates of WT and KO samples. d Analysis of western blot revealing a significant decrease in Kv1.2 expression in KO (n = 7) compared to WT (n = 8) mice. e Quantitative RT-PCR analysis showed no difference in Kv1.2 mRNA abundance between the WT (n = 6) and KO (n = 6) cerebellum. f Co-IP by anti-Kv1.2 and anti-FMRP antibodies, indicating the two molecules interact at the protein–protein level. g Schematic depiction of paired recording configurations: top, presynaptic cell-attached mode for baseline measurements; bottom, infusion of N-terminus of FMRP binding antibody (n-FMRP-AB) or N-terminus FMRP fragment (n-FMRP) into the BC terminal until equilibrium. Presynaptic and postsynaptic holding potential was set at −80 and −60 mV, respectively. h–j sIPSCs (left panels) recorded from PNs before and 20 min after diffusion of n-FMRP-AB (NBP2-01770, Novus Biologicals, 1:2000 dilution) into a WT (h) or KO (i) synapse, and n-FMRP (H00002332-P01, Novus Biologicals, 1:100 dilution) into a KO (j) synapse. The amplitude and frequency of sIPSCs are summarized in the right panels for the conditions in h (n = 5), i (n = 6), and j (n = 6). k–o Averaged intra-terminal Ca2+ rise from 5 trials evoked by AP trains (100 Hz, 100 ms) delivered to the soma of BCs in KO synapses with (black) or without n-FMRP (control, grey). The amplitude (l), area integral (m), 10–90 rise time (n), and 90–10 decay time (o) of Ca2+ transients are summarized for the control (grey bars, n = 7) and n-FMRP (black bars, n = 5) groups

Fig 4: Direct interaction between FMRP and Kv1.2 in a phosphorylation-specific manner. a Illustration of an alpha subunit of Kv1.2 channel with extensive details of intracellular C-terminus. b K+ current generated by a voltage ramp (−90 to 100 mV, top) from CHO cells expressing WT (Kv1.2-WT, middle) or truncated Kv1.2 (Kv1.2-Δ429-500, bottom) before and after infusion of N-terminus FMRP protein (n-FMRP, H00002332-P01, Novus Biologicals, 1:100 dilution). c The changes in amplitude of K+ current by n-FMRP is summarized for Kv1.2-WT (n = 8) and Kv1.2-Δ429-500 (n = 8) constructs. d Co-IP of Kv1.2 and FMRP in the absence or presence of three peptides targeting the sequence (436–457) of C-terminus of Kv1.2 as highlighted in a phosphorylated (PhosKv436-457), non-phosphorylated (Kv436-457), and phosphorylated scrambled peptide. e Summary of effect of the three peptides. Note that only PhosKv436-457 significantly reduces the production of Kv1.2 pull down as compared to the control conditions. f–h sIPSCs (left panels) recorded from PNs before and 20 min after infusion of the three peptides (2.5 mM) into WT synapses. The changes in the amplitude and frequency of sIPSCs are summarized in the right panels for PhosKv436-457 (n = 6, f), non-phosphorylated Kv436-457 (n = 6, g), and scrambled (n = 5, h) peptides. Consistent with the co-IP experiment, only PhosKv436-457 increases the amplitude and frequency of sIPSCs, acutely imparting a WT synapse with the KO phenotype. i Same quantifications for injecting PhosKv436-457 (2.5 mM) into KO synapses (n = 5) showing no effect on sIPSCs

Fig 5: N-terminus of FMRP interacts with Kv1.2 to control inhibitory output at BC nerve terminals. a Images of Kv1.2 immunostaining in cerebellar cortex showing reduced expression of Kv1.2 (green) in the Fmr1-KO BC axonal terminals. Individual pinceau marked with a single arrow (top) are revealed at a higher magnification (bottom). Dotted lines indicate Purkinje cell layer. ML molecular layer, GCL granule cell layer. b Comparison of the density of pinceau (top) and fluorescence intensity (bottom) of Kv1.2 labeling between WT and Fmr1 KO. c Representative western blot of Kv1.2 taken from the whole cerebellar homogenates of WT and KO samples. d Analysis of western blot revealing a significant decrease in Kv1.2 expression in KO (n = 7) compared to WT (n = 8) mice. e Quantitative RT-PCR analysis showed no difference in Kv1.2 mRNA abundance between the WT (n = 6) and KO (n = 6) cerebellum. f Co-IP by anti-Kv1.2 and anti-FMRP antibodies, indicating the two molecules interact at the protein–protein level. g Schematic depiction of paired recording configurations: top, presynaptic cell-attached mode for baseline measurements; bottom, infusion of N-terminus of FMRP binding antibody (n-FMRP-AB) or N-terminus FMRP fragment (n-FMRP) into the BC terminal until equilibrium. Presynaptic and postsynaptic holding potential was set at −80 and −60 mV, respectively. h–j sIPSCs (left panels) recorded from PNs before and 20 min after diffusion of n-FMRP-AB (NBP2-01770, Novus Biologicals, 1:2000 dilution) into a WT (h) or KO (i) synapse, and n-FMRP (H00002332-P01, Novus Biologicals, 1:100 dilution) into a KO (j) synapse. The amplitude and frequency of sIPSCs are summarized in the right panels for the conditions in h (n = 5), i (n = 6), and j (n = 6). k–o Averaged intra-terminal Ca2+ rise from 5 trials evoked by AP trains (100 Hz, 100 ms) delivered to the soma of BCs in KO synapses with (black) or without n-FMRP (control, grey). The amplitude (l), area integral (m), 10–90 rise time (n), and 90–10 decay time (o) of Ca2+ transients are summarized for the control (grey bars, n = 7) and n-FMRP (black bars, n = 5) groups