Degradation of high affinity HuD targets releases Kv1.1 mRNA from miR-129 repression by mTORC1

Abstract

Little is known about how a neuron undergoes site-specific changes in intrinsic excitability during neuronal activity. We provide evidence for a novel mechanism for mTORC1 kinase-dependent translational regulation of the voltage-gated potassium channel Kv1.1 messenger RNA (mRNA). We identified a microRNA, miR-129, that repressed Kv1.1 mRNA translation when mTORC1 was active. When mTORC1 was inactive, we found that the RNA-binding protein, HuD, bound to Kv1.1 mRNA and promoted its translation. Unexpectedly, inhibition of mTORC1 activity did not alter levels of miR-129 and HuD to favor binding to Kv1.1 mRNA. However, reduced mTORC1 signaling caused the degradation of high affinity HuD target mRNAs, freeing HuD to bind Kv1.1 mRNA. Hence, mTORC1 activity regulation of mRNA stability and high affinity HuD-target mRNA degradation mediates the bidirectional expression of dendritic Kv1.1 ion channels.

Figure 1.

mTORC1 kinase–dependent repression of Kv1.1 is not a result of mRNA stability. (A) SNs were isolated from DMSO- or rapamycin-treated DIV 21 cortical neurons. Representative Western blots and quantification indicate the relative level of p-mTOR/mTOR and Kv1.1/tubulin (loading control). ***, P < 0.001; *, P< 0.05; unpaired Student’s t test. p-mTOR, n = 7; Kv1.1, n = 5 over two independent cultures. (B) SN RNA was isolated from DMSO- or rapamycin-treated cortical neurons, and Kv1.1 mRNA was detected via RT-qPCR. Representative DNA gel of RT-qPCR samples showing amplification of a specific Kv1.1 band. DMSO, n = 5; rapamycin, n = 6 over three independent cultures. Error bars show SEMs.

Figure 2.

Overexpression of Kv1.1 3′UTR removes endogenous repression factors, leading to increased Kv1.1 protein. (left) Representative neurons infected with Sindbis virus coding for control RNA (Kaede-MAP2-DTS) or Kv1.1 3′UTR (DIV 21). Neurons were treated with DMSO or rapamycin. Arrowheads show Kv1.1 puncta (signal). Bar, 20 µm. (right) Quantification of surface expression of Kv1.1. ***, P < 0.001; *, P< 0.05; unpaired Student’s t test. Control + DMSO: n = 16 neurons, 21 dendrites; 3′UTR + DMSO: n = 17 neurons, 20 dendrites; control + rapamycin (Rapa): n = 7 neurons, 8 dendrites; 3′UTR + rapamycin: n = 10 neurons, 11 dendrites. Error bars show SEMs. A.U., arbitrary unit.

Figure 3.

miR-129 binds Kv1.1 RNA when mTORC1 kinase is active. (A) Sequence alignment of Kv1.1 3′UTR indicates that the miR-129 seed match sequence (binding to nt 2–8 of miR-129) is conserved among rat, mouse, guinea pig, and human, with nt 1 and 8 being less conserved. The miR-129 binding site is highlighted in red. This motif is considered to be a “weak” binding site, consistent with a role in translational repression over degradation (Vatolin et al., 2006; Grimson et al., 2007). Note that there are an additional 180 nt after the stop codon in the 3′UTR of the rat and mouse sequences that are not present in guinea pig and human sequences. Nucleotide number after the stop codon of each sequence shown: Rat, 181–212 nt; mouse, 177–207 nt; guinea pig, 1–34 nt; human, 1–35 nt. The NCBI accession numbers are rat M26161.1, mouse NM_010595, and human BC101733.1. For guinea pig, the University of California, Santa Cruz genome browser database was used. The sequence is located in scaffold_107:2955798–2955831. The asterisks represent the nucleotides that are conserved between rat, mouse, guinea pig, and human. (B, top) Schematic of RNA fragments used as bait to determine miR-129 binding to Kv1.1. mTRS is indicated by the gray box illustrating the miR-129 seed match sequence (FL) or the mutated sequence (ΔmiR-129). (bottom) qPCR of miR-129 pulled down from DMSO- or rapamycin-treated neurons. *, P < 0.05; ***, P < 0.001; one-way analysis of variance (ANOVA), Dunnett’s posttest compared with the FL DMSO. DMSO: FL n = 11, CR n = 8, and ΔmiR-129 n = 5; rapamycin (Rapa): FL n = 6, CR n = 8, and ΔmiR-129 n = 4 over at least four independent cultures. CR, coding region; FL, full length. (C, left) Representative neurons transfected with EGFP alone (control) or with LNA to specifically knock down miR-129 (KD). Arrowheads indicate Kv1.1 puncta in dendrites. (right) Quantification of dendritic Kv1.1 punctal signal intensity. Number of dendrites: control, n = 12; miR-129 KD, n = 14. Nine neurons for each condition. ***, P < 0.001; unpaired Student’s t test. (D, left) Representative neurons transfected with EGFP alone (control) or with LNA against miR-129 (KD). (right) Quantification of dendritic Kv4.2 puncta. Number of dendrites: control, n = 12; miR-129 KD, n = 17. n = 8 and 10 neurons for control and miR-129 KD, respectively. In C and D, yellow dotted lines were drawn to outline representative dendrite in Fig. 3 (C–E). (E, left) Hippocampal neurons transduced with either DsRed control or DsRed pre–miR129-2 lentivirus at DIV 14. DMSO- or rapamycin-treated DIV 21 neurons were stained and imaged for RFP and Kv1.1. (right) Quantification of signal intensity. ***, P < 0.001; one-way ANOVA, Tukey posttest. n = 9 neurons per treatment. Number of dendrites: DMSO-treated rapamycin control, n = 17; n = 14 and 15 for miR-129 overexpression (OE). A.U., arbitrary unit. Error bars show SEMs. Bars, 20 µm.

Figure 4.

Mutating the miR-129 seed match sequence in Kaede-Kv1.1 mRNA results in mTORC1-independent new translation in neuronal dendrites. Live imaging of cultured hippocampal neurons expressing Kaede-Kv1.1 with (top, FL) or without (bottom, ΔmiR-129) seed match sequence of miR-129 in aCSF containing 200 nM DMSO (left, control) or rapamycin (right) before, immediately after (0 time point), and 120 min after UV exposure to photoconvert Kaede-Kv1.1. (left) Entire representative neuron. (right) Enlarged representative dendrite, indicated by arrows, >60 µm from the soma. Bars: (main images) 50 µm; (insets) 10 µm. DMSO: FL, n = 58 puncta and ΔmiR129, n = 67 puncta; rapamycin (Rapa): FL, n = 51 puncta and ΔmiR-129, n = 64. *, P < 0.05; **, P < 0.01; one-way ANOVA, Tukey posttest, relative to Kaede-Kv1.1-FL DMSO. Yellow boxes indicate the region of the dendrite used in the enlarged images. The yellow lines outline the representative dendrite. Error bars show SEMs.

Figure 5.

Mutating the miR-129 seed match sequence in EGFP-Kv1.1 RNA increases protein levels without changing RNA levels. Representative images of dendritic localization of EGFP-Kv1.1 RNA (red) and protein (green) with the intact (left, FL) or mutated (right, ΔmiR-129) miR-129 binding site revealed by in situ hybridization (ISH) using a digoxigenin-labeled antisense oligo against EGFP. Bar, 20 µM. n = 12 and 13 neurons for FL and ΔmiR-129, 40 dendrites each. Error bars show SEMs.

Figure 6.

HuD binds Kv1.1 mRNA when mTORC1 kinase is inhibited and increases Kv1.1 expression that is reversed with cycloheximide. (A, left) RT-qPCR amplification of miR-129 from neurons treated with DMSO or rapamycin (Rapa). n = 3 independent cultures, and each sample was performed in triplicate. (right) Northern blot analysis showed no significant difference in the expression level of miR-129 between neurons treated with DMSO or rapamycin. The signal intensity for miR-129 was normalized to the signal intensity for let-7a, which remained constant between DMSO- or rapamycin-treated neurons. As a loading control, ethidium bromide–stained low molecular weight RNA is shown in the bottom blot (labeled as EtBr). n.s., not significant. (B) RT-PCR amplification of Kv1.1 mRNA pulled down by HuD in HEK293T cells. HEK293T cells were cotransfected with myc-HuD and Kaede-Kv1.1 FL, CR, or ΔmiR-129. Antimyc-coated beads were used to pull down HuD bound to Kv1.1 mRNA outlined above. Kv1.1 FL without HuD or Kaede in place of Kv1.1 (bottom) were transfected as controls and show no binding to Kv1.1 mRNA or Kaede RNA alone. n = 2 independent cultures. (C, left) Representative neurons (DIV 14) cotransfected with EGFP and either pcDNA or HuD cDNA. Neurons were treated with 50 µM DMSO or cycloheximide, and Kv1.1 and Kv4.2 were detected with specific antibodies. Arrowheads show Kv1.1 puncta (signal). Bars, 20 µm. (right) Quantification of Kv1.1 and Kv4.2 punctal intensity in dendrites. Number of dendrites: DMSO-treated Kv1.1 pCDNA, n = 15 and HuD, n = 18; Kv4.2 pCDNA, n = 17 and HuD, n = 16. Cycloheximide (cyclo)-treated Kv1.1 HuD, n = 14 and Kv4.2 HuD, n = 17. ***, P < 0.001; one-way ANOVA, Tukey’s posttest. ^#, Kv4.2 punctal intensity significant from DMSO + HuD. (D) RT-PCR amplification of Kv1.1 (top), GAP-43, or CaMKIIα (middle) mRNA copurified with GST-HuD or GST from DMSO- or rapamycin-treated cortical neurons (DIV 21). (bottom) GAPDH mRNA was detected in input but not pull-down. The ratio (rapamycin/DMSO; right of the images) is determined by subtracting signal intensity of the background GST band from the specific GST-HuD band and normalizing each band by their respective GAPDH input mRNA levels. HuD-RNA pull-down was replicated with three independent cultures for Kv1.1 mRNA and two independent cultures for CaMKIIα and GAP-43 mRNA. Significance for each mRNA was determined by a single Student’s t test. *, P < 0.05, indicating the binding is significantly different from 1. A value of 1 suggests equal binding in both treatments. Error bars show SEMs. rRNA, ribosomal RNA.

Figure 7.

HuD binding to Kv1.1 mRNA coincides with reduced levels of other HuD-target mRNAs. (A) Western blot analysis of HuD from SNs isolated from neurons treated with DMSO or rapamycin. DMSO, n = 3; rapamycin (Rapa), n = 4 over two independent cultures. (B) RT-qPCR analysis of CaMKIIα, GAP-43, Homer1a, and Kv1.1 mRNA isolated from control or 200 nM rapamycin-treated neurons normalized to the internal housekeeping gene, GAPDH, which remains constant between the two conditions. CaMKIIα and Kv1.1, n = 3; GAP-43, n = 5; Homer1a, n = 4; three to five independent cultures. Control was normalized to 100% and is indicated by dotted line on the graph. The mRNA target of HuD is shown as percent remaining after rapamycin treatment. One-sample Student’s t test was performed to determine statistical significance from control. (C) Representative blot and quantification of SN CaMKIIα protein isolated from DMSO- or rapamycin-treated cortical neurons. DMSO, n = 5; rapamycin, n = 6. (D) Neurons were treated with 12 µM Actinomycin D (Act D) for 4–5 h before treating with DMSO or rapamycin for 75 min. Degradation was measured by RT-qPCR for CaMKIIα mRNA and reported as the percent decrease with the addition of rapamycin relative to actinomycin alone. n = 5 per treatment. The solid line is connecting the mean for actinomycin alone to the mean for 75- min DMSO + actinomycin treatment. The dotted line is connecting the mean for actinomycin alone to the mean for 75-min rapamycin + actinomycin treatment. *, P < 0.05; ***, P < 0.001. Error bars show SEMs.

Figure 8.

Overexpression of the CaMKIIα UTR with multiple HuD sites occludes the increase in dendritic Kv1.1 expression. (A) Representative neurons (DIV 21) transfected with cDNA coding for EGFP or dGFP CaMKIIα UTRs were treated with 200 nM DMSO or rapamycin (Rapa), fixed, and immunostained for EGFP and Kv1.1. Quantification of dendritic Kv1.1 signal intensity for control (EGFP) or CaMKIIα UTR overexpression normalized by baseline signal for dendritic Kv1.1 under control conditions (EGFP/DMSO). Number of dendrites: DMSO-treated control, n = 22 and CaMKIIα UTR overexpression (OE), n = 29; rapamycin-treated control, n = 33 and CaMKIIα UTR overexpression, n = 31. *, P < 0.05; **, P < 0.01; ***, P < 0.001; one-way ANOVA, Tukey’s posttest. (B) Representative neurons expressing dGFP CaMKIIα UTR without the putative HuD binding sites treated with 200 nM DMSO or rapamycin were quantitated as outlined in A of this figure. Number of dendrites: CaMKIIα UTR overexpression with DMSO control, n = 20; CaMKIIα UTR overexpression with rapamycin, n = 23; ΔHuD overexpression with DMSO, n = 20; ΔHuD overexpression with rapamycin, n = 24. ***, P < 0.001 relative to all other conditions; one-way ANOVA, Tukey’s posttest. Arrows show Kv1.1 puncta (signal). Yellow dotted lines were drawn to outline the representative dendrite. Error bars show SEMs. Bars, 20 µm. A.U., arbitrary unit.