Sp4-dependent repression of neurotrophin-3 limits dendritic branching

Abstract

Regulation of neuronal gene expression is critical to establish functional connections in the mammalian nervous system. The transcription factor Sp4 regulates dendritic patterning during cerebellar granule neuron development by limiting branching and promoting activity-dependent pruning. Here, we investigate neurotrophin-3 (NT3) as a target gene important for Sp4-dependent dendritic morphogenesis. We found that Sp4 overexpression reduced NT3 promoter activity whereas knockdown of Sp4 increased NT3 promoter activity and mRNA. Moreover, Sp4 bound to the NT3 promoter in vivo, supporting a direct role for Sp4 as a repressor of NT3 expression. Addition of exogenous NT3 promoted dendritic branching in cerebellar granule neurons. Furthermore, sequestering NT3 blocked the continued addition of dendritic branches observed upon Sp4 knockdown, but had no effect on dendrite pruning. These findings demonstrate that, during cerebellar granule neuron development, Sp4-dependent repression of neurotrophin-3 is required to limit dendritic branching and thereby promote acquisition of the mature dendritic pattern.

Figure 1. Sp4 transcription factor represses Neurotrophin-3 expression

Cerebellar granule neurons or Neuro2A cells were co-transfected with NT3 (−3275/+91)-luciferase reporter and either (A) FlagSp4 (shaded bar) or empty vector (Vec) (open bar), or (B) Sp4RNAi (shaded bar) or control RNAi (open bar). Data represent the resulting luciferase activity normalized to each control condition. (Two-tailed t-test, p<0.05 (*)) (C and D) Neuro 2A cells were transfected with pMSCVpuro U6/GFP RNAi or U6/Sp4 RNAi plasmid. (C) After selection, cell lysates were immunoblotted for Sp4 and Gapdh. (D) Total RNA was subject to RT-qPCR with primers specific for NT3, BDNF and Gapdh. Values represents mean ± SEM generated from experiments performed in triplicate and normalized to Gapdh expression levels. Student's t test was used to determine the significance between groups. Asterisk denotes statistically significant value relative to Control RNAi (* p<0.05).

Figure 2. Sp4 binds to the Neurotrophin-3 promoter in cerebellar granule neurons

(A) Schematic of NT3 promoter in mouse. Positions of two transcription start sites (1A) and (1B), as well as a TATA box and two inhibitory regions (shaded grey) enriched in Sp4 binding sites are indicated. Arrows represent the locations of the primers used to amplify region RIII for chromatin immunoprecipitation (ChIP) assays. (B) Sp transcription factors bound to the inhibitory region of the endogenous NT3 promoter. ChIP assays were performed in cerebellar granular neurons, using antisera specific for Sp1 and Sp4 transcription factors, or Luciferase as control IgG. Immunoprecipitated DNA was analyzed by real-time PCR with oligonucleotides that amplified a fragment within the inhibitory region RIII of the NT3 promoter or a control fragment within the coding region of the beta-actin gene. Percent input normalized to control IgG is shown.

Figure 3. NT3 promotes increased dendritic branchpoints in cerebellar granule neurons

Cerebellar granule neurons were transfected with control (U6/Scr) or Sp4 RNAi together with GFP expression vector at DIV 2. At DIV4, cells were treated with exogenous NT3 or bovine serum albumin (BSA) control for 48 hours. Neurons were fixed and subjected to immunocytochemistry with anti-GFP. Dendritic trees were visualized by fluorescence microscopy. (A) Representative images of neurons transfected with control RNAi (U6/Scr) after treatment with BSA control (left) or exogenous NT3 (right). Arrowhead indicates axon and asterisk cell bodies of transfected neurons. (B) Total number of primary and secondary dendrites was quantitated. (C) Total dendritic length was quantitated. Values represent mean ± SEM of number of dendrites (B) and total dendritic length (C) per neuron. ANOVA analysis was performed by comparing each condition to control neurons p<0.001 (***) or p< 0.01 (**), n= 157 neurons were analyzed. Two-tailed t-test was also performed to compare differences plus or minus NT3 p< 0.05 (*). Quantitation was performed in 4 independent experiments with similar results.

Figure 4. Sequestering NT3 prevents excess dendritic branch addition upon Sp4 knockdown

Cerebellar granule neurons were transfected with control or Sp4 RNAi as described for Figure 3. At DIV4, control bovine serum albumin or the chimeric protein TrkC-Fc was added. Neurons were fixed and subjected to immunocytochemistry with anti-GFP. Dendritic trees were visualized by fluorescence microscopy. (A) Total number of primary and secondary dendrites and total dendritic length was quantitated at DIV6. Values represent mean ± SEM of number of dendrites or total dendritic length per neuron. ANOVA analysis was performed by comparing each condition to control RNAi transfected neurons, p<0.001 (***), n= 145 neurons were analyzed. Two-tailed t-test was also performed to compare differences plus or minus addition of TrkC-Fc, p< 0.05 (*); ns= not significant. Quantitation was performed in 3 independent experiments with similar results. (B) Representative high magnification images at DIV 4 and DIV 6 of neurons transfected with control or Sp4 RNAi and treated with control BSA or TrkC-Fc (lower right panel). Asterisk indicates cell bodies of transfected neurons. (C) Total number of secondary dendrites was quantitated at DIV4, 5 and 6 as indicated. Values represent mean ± SEM of number of dendrites per neuron. ANOVA analysis was performed by comparing either Control or Sp4 RNAi transfected neurons at DIV4 with corresponding transfected neurons at DIV5 and 6, with or without TrkC-Fc treatment, p<0.001 (***) or p< 0.01 (**), n= 320 neurons were analyzed. Quantitation was performed in 3 independent experiments with similar results.

Figure 5. Transcription factor Sp4 activates TrkC expression

(A and B) Cerebellar granule neurons were co-transfected with TrkC-luciferase reporter and FlagSp4 expression vector (A) or Sp4RNAi (B) (shaded bar) or control vectors (open bar). The luciferase activity normalized to control is presented. (Two-tailed t-test, p<0.05 (*)). (C) Neuro 2A cells were stably transfected with control GFP or Sp4 RNAi as indicated. Total RNA was extracted and subject to RT-qPCR with primers for p75 (NGF receptor), TrKA, TrKB, TrKC, and Gapdh. Values represents mean ± SEM generated from experiments performed in triplicate and normalized to Gapdh expression levels. Student's t test was used to determine the significance between groups. Asterisk denotes statistically significant value relative to Control RNAi (* p<0.05).