Developmental expression of the SRF co-activator MAL in brain: role in regulating dendritic morphology

Authors

  • Jun Shiota,

    1. Department of Biological Chemistry, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama City, Japan
    Search for more papers by this author
    • 1

      Jun Shiota, Mitsuru Ishikawa and Akiko Tabuchi contributed equally to this work.

  • Mitsuru Ishikawa,

    1. Department of Biological Chemistry, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama City, Japan
    Search for more papers by this author
    • 1

      Jun Shiota, Mitsuru Ishikawa and Akiko Tabuchi contributed equally to this work.

  • Hiroyuki Sakagami,

    1. Division of Histology, Department of Cell Biology, Graduate School of Medicine, Tohoku University, Sendai, Japan
    Search for more papers by this author
  • Masaaki Tsuda,

    1. Department of Biological Chemistry, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama City, Japan
    2. CREST, JST, Tokyo, Japan
    Search for more papers by this author
  • Jay M. Baraban,

    1. Department of Neuroscience, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA
    Search for more papers by this author
  • Akiko Tabuchi

    1. Department of Biological Chemistry, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama City, Japan
    2. CREST, JST, Tokyo, Japan
    Search for more papers by this author
    • 1

      Jun Shiota, Mitsuru Ishikawa and Akiko Tabuchi contributed equally to this work.


Address correspondence and reprint requests to Akiko Tabuchi, Department of Biological Chemistry, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, 2630 Sugitani, Toyama City, Toyama 930–0194, Japan. E-mail: atabuchi@pha.u-toyama.ac.jp

Abstract

The dynamic changes in dendritic morphology displayed by developing and mature neurons have stimulated interest in deciphering the signaling pathways involved. Recent studies have identified megakaryocytic acute leukemia (MAL), a serum response factor (SRF) co-activator, as a key component of a signaling pathway linking changes in the actin cytoskeleton to SRF-mediated transcription. To help define the role of this pathway in regulating dendritic morphology, we have characterized the pattern of MAL expression in the developing and adult brain, and have examined its role in regulating dendritic morphology in cultured cortical neurons. In histological studies of mouse brain, we found prominent expression of MAL in neurons in adult hippocampus and cerebral cortex. MAL immunostaining revealed localization of this protein in neuronal cell bodies and apical dendrites. During development, an increase in MAL expression occurs during the second post-natal week. Expression of dominant negative MAL constructs or MAL siRNA in cortical neurons grown in primary culture reduces the number of dendritic processes and decreases the basal level of SRF-mediated transcription. Taken together, these findings indicate that the MAL-SRF signaling pathway plays a key role in regulating dendritic morphology.

Abbreviations used
BSA

bovine serum albumin

DIV

days in vitro

DMEM

Dulbecco's modified Eagle's medium

E

embryonic day

FBS

fetal bovine serum

GFP

green fluorescent protein

MADS

MCMI, Agamous, Deficiens, SRF

MAL

megakaryocytic acute leukemia

MAPK

mitogen-activated protein kinase

P

post-natal day

PBS

phosphate-buffered saline

PW7

7th week

PW11

11th week

SSC

saline sodium citrate buffer

SRF

serum response factor

TCF

ternary complex factor

Serum response factor (SRF) is a member of the MCMI, Agamous, Deficiens, SRF (MADS)-box transcription factor family, which binds to a consensus sequence CC(A/T)6GG called a CArG box, present in the promoters of several immediate early genes, including c-fos and egr-1, and cytoskeletal genes such as β-actin, vinculin and gelsolin (Miralles et al. 2003; Philippar et al. 2004; Salvaraj and Prywes 2004; Alberti et al. 2005; Ramanan et al. 2005). SRF-dependent transcription is controlled by two different co-activator families, ternary complex factors (TCFs), such as Elk-1, and myocardin-related transcription factors, such as megakaryocytic acute leukemia, referred to as MAL or MKL (Salvaraj and Prywes 2003;Cen et al. 2003, 2004). Elk-1 mediates activation of SRF by mitogen-activated protein kinase (MAPK) cascades; MAL mediates SRF activation by Rho signaling pathways (Miralles et al. 2003). Rho GTPases play important roles in regulating actin-based neuronal morphology (Van Aelst and Cline 2004). Although the mechanisms linking Rho signaling pathways to MAL activation are still being clarified, the available evidence indicates that MAL is normally sequestered in the cytoplasm by binding to G-actin. Stimulation of Rho proteins, which exert prominent effects on the actin cytoskeleton, leads to dissociation of G-actin from MAL. Release of MAL from G-actin enables it to translocate to the nucleus where it binds and activates SRF (Miralles et al. 2003). Even though TCF family members and MAL/MKL family members both bind to, and activate, the same transcription factor, SRF, they regulate transcription of different sets of target genes. For example, the TCF/SRF complex drives transcription of egr-1 and c-fos, whereas MAL/SRF has been implicated in stimulating transcription of cytoskeleton-related genes, such as vinculin.

These findings suggest that Rho GTPases may affect neuronal morphology via direct and indirect mechanisms: directly, via their well established effects on the actin cytoskeleton; and indirectly, via transcriptional effects mediated by the MAL-SRF signaling pathway. To assess the possibility that the MAL-SRF pathway regulates neuronal morphology, we have investigated the pattern of MAL expression in developing and adult brain, and assessed the impact of suppressing MAL expression or function on the morphology of cortical neurons.

Materials and methods

Plasmids and antibodies

Dominant negative MAL constructs (ΔB1B2 and C471Δ) and SRF reporter vector (3D.ALuc) were provided by Dr R. Treisman (Cancer Research UK, London Research Institute). An internal control vector, phRL-TK, was from Promega (Madison, WI, USA). FLAG-tagged full-length MAL construct (FLAG-MAL) was generated as described below. RSV-βgal vector has been described previously (Tabuchi et al. 2002). Anti-BSAC/MAL antibody was provided by Dr H. Nakano (Juntendo University, Japan). The following antibodies were used at the indicated dilutions: fluorescein isothiocyanate (FITC)-conjugated anti-rabbit IgG (1 : 500; Vector Laboratories, Burlingame, Canada), rhodamine-conjugated anti-mouse IgG (1 : 200; Santa Cruz Biotechnology, Santa Cruz, CA, USA) made in goat, horshradish peroxidase (HRP)-conjugated anti-rabbit IgG and HRP-conjugated anti-mouse IgG made in donkey (1 : 5000; Amersham GE Healthcare, Little Chalfant, UK). The following rabbit polyclonal antibodies were used: anti-green fluorescent protein (GFP) (1 : 500; Invitrogen, Carlsbad, CA, USA), anti-BSAC/MAL (1 : 200) (Sasazuki et al. 2002) and biotinylated anti-rabbit IgG (1 : 500, Vector Laboratories). The following mouse monoclonal antibodies were used: anti-GFP (1 : 500 for western blotting; Santa Cruz Biotechnology), anti-FLAG (1 : 1000; Sigma, St Louis, MO, USA), anti-α-tubulin (1 : 1000; Sigma), anti-microtubule-associated protein 2 (MAP2) (1 : 1000; Sigma) and anti-β-actin (1 : 1000; Sigma).

Cloning of mouse MAL cDNA

To clone mouse MAL cDNA total RNA, derived from NIH 3T3 cells, was reverse-transcribed with SuperScript II (Invitrogen). MAL cDNA was amplified by PCR in two segments using the following primer pairs: (5′-CCGGAATTCACCCCCTTCCGTCATTGCT-3′, 5′-TGTGGAATTCTCATCACCCGTGCTGAG-3′) and (5′-TGATGAGAATTCCACACCTGGGGATGC-3′, 5′-CTAGTCTAGACTACAAGCAGGAATCCCAGTGG-3′), respectively. The downstream segment was ligated into the EcoRI/XbaI sites of the N-terminal FLAG expression vector, pFLAG-CMV2 (Sigma). The upstream portion was then ligated into the EcoRI site of the vector containing the downstream portion (Sigma). The nucleotide sequences are identical to those displayed in NCBI's website (http://www.ncbi.nlm.nih.gov/genome/seq/MmBlast.html).

In situ hybridization

In situ hybridization was performed as described previously (Sakagami et al. 1998). The antisense oligonucleotide probe was complementary to the mouse MAL gene (accession number: BC054801; nucleotide positions: 3363–3407). The sequence used was: 5′-CTAGGGACTGTGATTGTCGAGGCTCACAGTCACAGGAGCGTACAC-3′. Cryostat sections (25 µm) of mouse (C57BL/6 J) brains harvested at (E) day 18, post-natal day 0 (P0), P5, P10, P15, P21, 7th week (PW7) and 11th week (PW11) were hybridized overnight at 42°C with [α-35S]dATP-labeled oligonucleotide probes in a solution consisting of 50% deionized formamide, 4 × saline sodium citrate buffer (SSC), 1 × Denhardt's solution, 1% sodium N-lauroyl sarcosinate (sarcosyl), 0.1 m phosphate buffer (pH 7.2), 200 µg/mL yeast tRNA, 10% dextran sulfate, 100 mm dithiothreitol (DTT) and 35S-labeled oligonucleotide probe. After hybridization, the sections were washed four times with 0.1 × SSC/0.1% sarcosyl at 50°C for 30 min each. After exposing the slides to film to obtain an initial image, the slides were dipped in NTB2 nuclear emulsion (Eastman Kodak, Rochester, NY, USA) and exposed for one month at 4°C. To control for the specificity of the hybridization signal, parallel experiments were also performed in the presence of a 50-fold excess of unlabeled oligonucleotide probe.

Quantitative real-time PCR

Quantitative real-time PCR was performed in an ABI PRISM 7700 sequence detection system (Applied Biosystems, Foster City, CA, USA) using the SYBR Green PCR master mix (Applied Biosystems) according to the method described previously (Fukuchi et al. 2004). Total RNA was extracted from mouse (C57BL/6 J) brain regions using ISOGEN (Nippongene, Toyama, Japan) and cDNA was synthesized with SuperScriptII (Invitrogen). In brief, for detection of MAL mRNA levels, PCR was performed in 25 µL of 1 × SYBR system using the SYBR Green PCR master mix containing 1 µL cDNA solution and 0.5 µm primers (5′-TGGAGAGACGCTTTTTCTGG-3′ and 5′-TGAGCTTCTTCACCTTTGGC-3′). After pre-heating at 95°C for 10 min, the samples were denatured at 95°C for 45 s, annealed at 55°C for 45 s and extended at 72°C for 1 min for 45 cycles.

Immunostaining

Mice (C57BL/6 J) were anesthetized with chloral hydrate (400 mg/kg, i.p.) and perfused with 4% freshly depolymerized paraformaldehyde in phosphate-buffered saline (PBS). Brains were removed and post-fixed for 6–12 h in 4% paraformaldehyde in PBS, before being transferred to 25% sucrose in PBS for cryoprotection. Coronal sections (40 µm thickness) were cut on a sliding microtome and then processed for immunostaining using an antigen retrieval protocol as described in Reti et al. (2002). Briefly, sections were washed for 30–60 min in PBS before mounting on glass slides (Precleaned Superfrost Plus MicroSlides, VWR, West Chester, PA, USA). After drying for approximately 1 h, the mounted sections were re-hydrated in PBS and then boiled for 5 min in an antigen retrieval buffer (Tissue Retrieval Solution, Cell Marque, Hot Springs, AR, USA). After cooling, the slides were washed in PBS, then blocked with PBS containing 3% bovine serum albumin (BSA; Sigma) and 0.2% Triton X-100 at room temperature (25°C) for 1 h. Then, they were treated with anti-BSAC/MAL antibody in the presence of PBS containing 1% BSA overnight at 4°C. After the sections were washed with PBS, they were incubated for 1 h at room temperature (25°C) with a biotinylated anti-rabbit secondary antibody (1 : 500). After another washing, the sections were treated with avidin-biotin complex at room temperature (25°C) for 1 h. The sections were incubated in a solution containing tyramide (TSA Biotin System, Perkin Elmer, Boston, MA, USA). Finally, coverslips were mounted on the slides with Permafluor-DABCO (Beckman-Coulter, Fullerton, CA, USA). To visualize the localization of MAL protein, immunofluorescence images were acquired with a confocal microscope (Radiance 2100MP, Bio-Rad Laboratories, Hercules, CA, USA). In control experiments, we verified that omission of the primary antibody abolished staining. Furthermore, we confirmed in immunoblotting studies conducted with brain extracts that the anti-MAL antibody only detects a single band that corresponds to the expected size of MAL protein. The procedures used to process NIH 3T3 cells and cortical cultures for immunostaining have been described previously (Tabuchi et al. 2005).

Cell culture

NIH 3T3 cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen) containing 10% fetal bovine serum (FBS) (Invitrogen), 10% Nu-serum (BD Biosciences, San Jose, CA, USA), 2 mm glutamine (Invitrogen) and 1% penicillin/streptomycin (Invitrogen). For immunostaining, NIH 3T3 cells were seeded at 2.5 × 105 cells/well and grown on Permanox two-chamber slides (Nalge Nunc, Naperville, IL, USA). For reporter assays, cells were seeded at 5 × 105 cells/well and grown on six-well plates (Nalge Nunc). Dissociated cortical cell cultures were prepared from rat embryos at embryonic day (E) 17 or 18, as described previously (Tabuchi et al. 2002). In brief, cortical tissue was dissected quickly in DMEM containing 10% FBS and 100 µg/mL kanamycin sulfate (WAKO Pure Chemicals Industries, Osaka, Japan). After the dissected tissue was treated with trypsin solution and DNase I solution, cells were re-suspended in neurobasal medium containing 1 × B27 supplement (Invitrogen), 2 µg/mL gentamicin and 2 mm glutamine. Half of the conditioned medium was exchanged for fresh medium every 3 days. For immunostaining, cells were plated at a density of approximately 7 × 105 cells/well onto 18 mm circle coverslips coated with poly d-lysine that had been placed in 12-well plates. For reporter assays, cells were plated at 2.5 × 106 cells per well in six-well plates.

Transfection into NIH 3T3 cells and cortical neurons

Transfection of NIH 3T3 cells was performed using Lipofectamine Plus reagents according to the manufacturer's instructions (Invitrogen). After cells were exposed to the transfection reagents for 4 h, the transfection medium was replaced with fresh medium. The cells were then used for immunostaining or reporter assays at the indicated times after transfection.

For cortical neurons, cells were transfected using the calcium phosphate precipitation method, as described previously (Tabuchi et al. 2005).

Reporter assay

Transcriptional activity was monitored using a dual luciferase assay, as described previously (Tabuchi et al. 2005). The Galacto-Star assay for β-galactosidase activity was used (Applied Biosystems) in experiments that employed a β-gal vector as an internal control. NIH 3T3 cells were transfected with plasmids (2 µg/well) in the following ratios: reporter vector : expression vector = 1 : 1; firefly luciferase : Renilla luciferase = 10 : 1. Cortical neurons were transfected with plasmids (4 µg/well) in the following ratios: reporter vector : expression vector = 1 : 3; firefly luciferase : Renilla luciferase = 5 : 1.

Morphological analysis

To monitor the effect of MAL constructs on the dendritic morphology of cortical neurons, these constructs (2 µg/well) were transfected with a GFP expression vector (2 µg/well) at 7 or 8 days in vitro (DIV) using the calcium phosphate precipitation method. Cells were fixed at the times indicated and then processed for microscopy (BX50-34-FLA-1, Olympus, Tokyo, Japan), as described previously (Tabuchi et al. 2005). A circle with a radius of 40 µm was centered on the cell body and the number of intersections with GFP-positive processes was recorded by an investigator (J.S.) who was unaware of the experimental condition being tested. For measurement of dendrite lengths, Scion Image software (Scion, Frederick, MD, USA) was used. Dendrite length was measured by tracing all the dendrites starting at the cell body. More than 20 neurons (for dendrite number) or more than 15 neurons (for dendrite length) were evaluated for each construct in each of at least three independent experiments.

Western blotting

Cortical neurons were harvested and protein was extracted with 160 µL buffer containing 25 mm HEPES, 0.3 m NaCl, 1.5 mm MgCl2, 0.2 mm EDTA, 0.1% Triton X-100, 20 mmβ-glycerophosphate, 10 µg/mL aprotinin, 10 µg/mL leupeptin, 1 mm sodium orthovanadate, 1 mm DTT and 1 mm phenylmethylsulfonyl fluoride (PMSF). After centrifugation, the cell lysates were mixed with an equal volume of 2 × Laemmli sample buffer (Bio-Rad Laboratories) and subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). To prepare protein extracts from mouse hippocampus, cortex and cerebellum, we homogenized 100 mg frozen tissue in 500 µL lysis buffer. After centrifugation, the samples were processed as described above for culture lysates. Protein detection was carried out with the enhanced chemiluminescence (ECL) protocol (Amersham).

RNA interference

To knock-down MAL expression, a vector expressing small interfering RNA (siRNA) (pSUPER-mrMAL) was constructed. The oligonucleotides (60-mers) 5′-GATCCCCGGTAGCAGACAGTTCCTCCTTCAAGAGAGGAGGAACTGTCTGCTACCTTTTTA-3′ and 5′-AGCTTAAAAAGGTAGCAGACAGTTCCTCCTCTCTTGAAGGAGGAACTGTCTGCTACCGGG-3′ were annealed and subcloned into HindIII/BglII sites of pSUPER vector (OligoEngine Platform, Seattle, WA, USA). The target sequences are identical to both mouse and rat MAL genes. The control vector, pSUPER-mrMALmut, which has two base substitutions, was also generated by annealing and subcloning of oligonucleotides 5′-GATCCCCGGTAGCAGtCAGTgCCTCCTTCAAGAGAGGAGGcACTGaCTGCTACCTTTTTA-3′ and 5′-AGCTTAAAAAGGTAGCAGtCAGTgCCTCCTCTCTTGAAGGAGGcACTGaCTGCTACCGGG-3′. Transfection of these pSUPER vectors was performed as described above and the effects were assessed by immunocytochemical and western blot analyses (Fig. 6).

Figure 6.

 Effect of MAL siRNA on MAL expression and function. (a) Western blot analysis. FLAG-MAL (0.8 µg/well) was co-transfected with GFP (0.4 µg/well) and pSUPER (control, 0.8 µg/well) or pSUPER-mrMAL (MAL siRNA, 0.8 µg/well) or pSUPER-mrMALmut (MAL siRNAmut, 0.8 µg/well) into NIH 3T3 cells. The expression of FLAG-MAL protein was not markedly affected at 48 h, but displayed a clear reduction at 72 h after transfection with MAL siRNA vector. No reduction of FLAG-MAL expression was observed by transfection with another control vector, MAL siRNA mut. (b) Effect of MAL siRNA on SRF-reporter activity. SRF reporter vector, 3D.ALuc (1 µg/well) was co-transfected with control or MAL siRNA vector (1 µg/well) into NIH 3T3 cells, and firefly luciferase activity was measured at 72 h after ttransfection. Renilla luciferase vector was used as an internal control. Bar graphs present the mean ± SD from three independent experiments. (c) Immunocytochemical analysis of MAL siRNA in cortical neurons. FLAG-MAL (0.2 µg/well) was co-transfected with GFP (0.15 µg/well) and control (0.4 µg/well) or MAL siRNA (0.4 µg/well) or MAL siRNAmut (0.4 µg/well), and the cells were immunostained with anti-GFP and anti-FLAG antibodies at 72 h after transfection. Asterisks indicate significance level, as determined by Student's t-test, in comparison with control (left-most bar): ***p < 0.001.

Results

MAL expression in mouse brain

To examine the expression pattern of MAL mRNA in the adult mouse brain, we performed in situ hybridization studies. We found prominent MAL mRNA expression in the hippocampal formation, with weaker expression in the olfactory bulb, caudate putamen and cerebral cortex (Fig. 1a). Within the hippocampus, hybridization was localized to the pyramidal cell layer of the cornu ammonis (CA) regions and the granule cell layer of the dentate gyrus. To check the specificity of the hybridization pattern observed, we confirmed that it was completely abolished by inclusion of a 50-fold excess of unlabeled probe (Fig. 1b).

Figure 1.

 MAL mRNA and protein expression in mouse brain. Autoradiograms of sagittal sections from adult mouse brain (P11W) that were hybridized with 35S-labeled oligonucleotide probe for MAL in the absence (a) or presence (b) of a 50-fold excess of unlabeled probe. Note prominent expression of MAL in the hippocampal formation (Hip), with weaker expression in the olfactory bulb (OB), caudate putamen (CP) and cerebral cortex (Cx). Cb, cerebellar cortex; MO, medulla oblongata; Th, thalamus. Bar = 1 mm. (c) MAL protein expression. Lysates, extracted from adult mouse cerebral cortex (Cx), hippocampus (Hip) and cerebellum (Cb), were processed for western blotting with MAL antibody, demonstrating similar levels of MAL protein in Cx and Hip with less in Cb. The β-actin bands are shown as a loading control.

Next, we checked MAL protein expression in the adult mouse brain. Western blotting revealed almost the same amount of MAL protein in lysates derived from cerebral cortex and hippocampus, with less in cerebellum (Fig. 1c). Although in situ hybridization studies indicated lower expression of MAL mRNA in cortex than hippocampus, the levels of MAL protein expression appear to be comparable in these regions.

To assess the developmental profile of MAL mRNA expression, we processed brain sections from E18 to adult mice for in situ hybridization (Fig. 2). During hippocampal development, the expression level of MAL mRNA increases after P5 in both hippocampal pyramidal and dentate granule cell layers, peaks between P15 and P21, and then decreases slightly (Fig. 2a). This developmental profile was corroborated in quantitative real-time PCR studies (Fig. 2b). A similar profile was observed in developing cerebral cortex. In contrast, the cerebellum, which has lower MAL mRNA levels in the adult, displayed a gradual decrease in expression during development.

Figure 2.

 Regional expression of MAL mRNA in developing mouse brain. (a) Developmental expression of MAL mRNA in the hippocampal formation. Note prominent up-regulation of MAL mRNA in the hippocampal neuronal layers after P5. CA1–3, fields of CA1–3 of Ammon's horn; DG, dentate gyrus. (b) Quantitative real-time PCR analyses of MAL mRNA levels in developing mouse hippocampus (i), cerebral cortex (ii) and cerebellum (iii). E18, embryonic day 18; P0, post-natal day 0 (day of birth); P5, post-natal day 5; P10, post-natal day 10; P15, post-natal day 15. Bar graphs represent the mean ± SE from three or four independent experiments.

Localization of MAL protein in mouse brain

Next, we examined the localization of MAL protein in mouse brain. Consistent with the in situ hybridization studies, we found prominent staining of neurons in the hippocampus and cerebral cortex with anti-MAL antibodies (Fig. 3). In the hippocampus, pyramidal neurons and dentate granule neurons display cytoplasmic staining that spares the nucleus and extends into apical dendrites (Fig. 3a–d). A similar pattern of staining was apparent in pyramidal neurons of the cerebral cortex (Figs 3e and f).

Figure 3.

 Immunohistochemical localization of MAL protein in mouse brain. Low and high power images of MAL immunostaining in the dentate gyrus (a and b), CA1 region of hippocampus (c and d) and the cerebral cortex (e and f) are shown. Low power views of hippocampus shows immunostaining in the dentate gyrus granule cell layer (a) and in the hippocampal CA1 pyramidal cell layer (c). Images in (b) and (e) show that MAL staining in the cell body spares the nucleus. (d, e) Examples of MAL staining extending into dendrites of hippocampal and cortical pyramidal neurons, respectively (arrows).

Effect of dominant negative MAL constructs on dendritic morphology

Previous studies have identified dominant negative MAL constructs that block the ability of endogenous MAL to stimulate SRF-mediated transcription (Miralles et al. 2003; Tabuchi et al. 2005). These were generated by deleting domains critical for either nuclear import (ΔB1B2) or transactivation (C471Δ). Accordingly, to assess the role of MAL in regulating neuronal morphology, we examined the impact of both of these dominant negative MAL constructs on cortical neurons grown in primary culture. Prior to proceeding with these experiments, we first confirmed that MAL protein is expressed in these cultures at 8 DIV, the age of cultures used for our initial transfection experiments (Fig. 4a). Furthermore, we checked that the dominant negative MAL constructs selected for use (Fig. 4a) are effective in these cultures. As expected, both of these dominant negative MAL constructs, ΔB1B2 and C471Δ, inhibited SRF-mediated transcriptional responses of rat cortical neurons (Fig. 4b). At 24 h post-transfection, C471Δ had a stronger inhibitory effect on SRF reporter activity than the ΔB1B2 construct (Fig. 4b, i and iii). However, the inhibitory effects of ΔB1B2 were greater at 48 h following transfection (Figs 4b, ii).

Figure 4.

 Dominant negative MAL constructs reduce SRF-mediated transcriptional response. (a) Schematic diagrams of full-length MAL and two dominant negative MAL constructs. The top bar represents full-length MAL (MAL) as previously reported (Miralles et al. 2003). The ΔB1B2 dominant negative construct, which is able to dimerize with endogneous MAL, is missing domains critical for nuclear import of MAL. The C471Δ dominant negative construct is missing the transactivation domain. RPEL, RPXXXEL motif; B1 and B2, basic region 1 and 2; SAP, SAF-A/B, Acinus, PIAS domain; PolyQ, polyglutamine repeat; LZ, leucine zipper. MAL protein expression is readily detected by western blotting in extracts harvested from cortical neurons at 8 DIV. (b) The SRF reporter vector, 3D.ALuc (1.0 µg/well), was co-transfected with either of the MAL dominant negative constructs, ΔB1B2 or C471Δ (1.5 or 3.0 µg/well), into cortical neurons. Firefly luciferase activity was measured at 24 h (i) or 48 h (ii and iii) post-transfection. For ΔB1B2 (i and ii), Renilla luciferase reporter vector (0.2 µg/well) was used as an internal control. For C471Δ (iii), β-galactosidase reporter vector (0.2 µg/well) was used. Bar graphs represent the mean ± SD from three independent experiments. Asterisks indicate significance level, as determined by Student's t-test, in comparison with control (left-most bar): *p < 0.05, **p < 0.01, ***p < 0.001.

To check the effect of these dominant negative MAL constructs on neuronal morphology, we co-transfected each of them with a GFP plasmid into cortical neurons and processed the cultures for immunostaining. As shown in Fig. 5(a), the dominant negative MAL constructs, ΔB1B2 or C471Δ, reduce the number of dendritic processes. In contrast, transfection with the wild-type MAL construct does not affect this parameter of dendritic morphology. To quantify these effects on dendritic morphology, we performed simplified Sholl analysis, as described previously (Tabuchi et al. 2005). We found that ΔB1B2 and C471Δ decreased the number of crossings in cultures harvested 48 h, but not 24 h, after transfection (Fig. 5b). Furthermore, both constructs also caused a reduction in total dendrite length at 48 h post-transfection (Fig. 5c). In addition to effects on dendrite number and length, we noted that dendrites of neurons transfected with dominant negative MAL constructs appear to be thinner. To investigate the possibility that the observed decrease in dendritic number occurs as part of an apoptotic response triggered by dominant negative MAL constructs, we checked for changes in nuclear morphology characteristic of apoptosis using 4′,6-diamidino-2-phenylindole (DAPI) staining. However, using this procedure we did not detect any evidence that these constructs trigger apoptosis.

Figure 5.

 Dominant negative MAL constructs decrease the number of dendrites in cortical neurons. (a) Immunofluorescence images of cortical neurons expressing GFP (2.0 µg/well) and either empty vector (empty) or one of the FLAG-tagged MAL constructs listed (2.0 µg/well). Images shown in each row show the same field. MAL fluorescence is shown in red, GFP in green and nuclei stained with DAPI in blue. Both dominant negative MAL constructs reduce the number of dendrites. (b) Simplified Sholl analysis. A measure of the number of dendritic processes was obtained for each cell scored by placing a circle of 40 µm radius over the center of the cell and counting the number of times the circle intersects GFP processes emanating from the cell body. Bar graphs (i) and (ii) indicate results obtained at 24 h and 48 h after transfection, respectively. (c) Dendrite length was measured by tracing GFP-positive dendrites from the cell body. Bar graphs represent mean ± SD from three to seven independent experiments. The statistical significance of differences between each group and the control group was analyzed by Student's t-test: **p < 0.01, ***p < 0.001.

MAL siRNA reduces the number of dendritic processes in cortical neurons

To confirm that the results obtained with dominant negative MAL constructs are due to blocking MAL function, we also checked whether RNAi-mediated knock-down of MAL mimicked their effect. First, we checked that the pSUPER-mrMAL (MAL siRNA) vector decreased MAL expression in NIH 3T3 cells, as expected (Fig. 6). Western blot analysis showed that MAL siRNA attenuated expression of FLAG-tagged MAL at 72 h post-transfection; in contrast, the control vector, MAL siRNA mutant (pSUPER-mrMALmut; MAL siRNAmut), which contains two base pair mismatches, did not (Fig. 6a). Moreover, to verify that MAL siRNA effectively suppressed MAL expression, we also checked its effect on SRF-mediated transcriptional responses. As expected, MALsiRNA inhibited the SRF transcriptional responses but the control siRNA mutant did not (Fig. 6b). When these cells were co-transfected with MAL siRNA and GFP constructs into cortical neurons, the fluorescence intensity of FLAG-tagged MAL was also markedly decreased 72 h after transfection (Fig. 6c).

Using these siRNA constructs, we assessed how loss of MAL affects the morphology of cortical neurons. After transfection at 7 DIV, cells were immunostained with anti-GFP and anti-MAP2 antibodies 72 h later. As found for dominant negative MAL mutants, expression of MAL siRNA, but not the mutant siRNA sequence, reduced the number of dendritic processes, as monitored with MAP2 immunostaining (Fig. 7a). Quantification of this response by Sholl analysis is presented in Fig. 7(b). Parallel experiments confirmed that MAL siRNA also reduces the basal level of SRF reporter activity present in these cultures (Fig. 7c). To assess whether MAL plays a key role in regulating dendritic morphology in more mature neurons, we assessed the effect of transfecting MAL siRNA constructs in cultures at 14 DIV instead of 7 DIV. We found that treatment with MAL siRNA decreased the number of dendritic processes in cultures scored 3 days after transfection. In contrast, the number of dendritic processes remained constant in control neurons (data not shown). Taken together, these findings imply that MAL is involved in regulating dendritic morphology at both 8 and 14 DIV, and may be involved in regulating both growth and stability of dendritic processes.

Figure 7.

 MAL siRNA reduces dendrite number and SRF transcriptional response of cortical neurons. (a) Immunofluorescent images of rat cortical neurons. Control, MAL siRNA or MAL siRNAmut plasmids (2.0 µg/well) were co-transfected with a GFP construct (2.0 µg/well) at 7 DIV, and neurons were immunostained with anti-GFP and anti-MAP2 antibodies at 72 h after transfection. (b) Sholl analysis of the experimental conditions shown in (a). (c) Effect of MAL siRNA on SRF transcriptional response in cortical neurons. SRF reporter vector (1 µg/well) was co-transfected with control or MAL siRNA vector (3 µg/well). Firefly luciferase activity was measured at 72 h post-transfection. Renilla luciferase vector was used as an internal control. Bar graphs shown represent the mean ± SD from three independent experiments. The statistical significance of differences between each group and the control group was analyzed by Student's t-test. **p < 0.05, **p < 0.01.

Discussion

Our studies of MAL expression and function in neurons have yielded several important, informative findings. In situ hybridization studies revealed that MAL mRNA is selectively expressed in several forebrain areas, with prominent expression in the hippocampus. Characterization of the developmental profile of MAL mRNA showed that it undergoes a rise during the second post-natal week, with expression sustained into adulthood. Immunohistochemical studies detected MAL in neurons where it is localized in the cytoplasm of the cell body and extends into apical dendrites. In addition, transfection of cultured cortical neurons with dominant negative or siRNA MAL constructs indicated that MAL is required for the growth and/or stability of dendritic processes. As MAL is a co-activator of SRF, these findings imply that the MAL-SRF pathway plays a key role in regulating dendritic morphology.

Although we have examined the effect of suppressing MAL expression or function in developing neurons in culture, the high level of MAL expression in adult neurons suggests that this pathway also plays an important role in regulating dendritic morphology in mature neurons. This inference is supported by our observation that MAL siRNA decreases the number of dendrites in neurons treated at 14 DIV, which have reached a steady-state level of dendrite number. Of note, while our studies have demonstrated that MAL is expressed in the forebrain, other brain areas, such as thalamus, cerebellum and brain stem, display much lower levels of MAL expression. It is conceivable that other MAL family members or, alternatively, other transcriptional pathways, play analogous roles in those regions. In either case, variation in the level of MAL expression may contribute to the marked heterogeneity in dendritic morphology found in distinct neuronal populations.

These studies indicate that the MAL-SRF complex should be included in the growing list of transcription factors involved in regulating dendritic morphology. cAMP-response element binding protein (CREB), a transcription factor that plays crucial roles in learning and memory, has also been implicated in regulating dendritic morphology (Redmond et al. 2002). Furthermore, Neuro D, a basic helix-loop-helix (bHLH) transcription factor, controls the generation of dendrites but not axons (Gaudilliere et al. 2004).

Classical studies have implicated SRF in mediating activity-dependent changes in gene expression (Johnson et al. 1997). These responses, such as induction of c-fos, appear to be mediated via the TCF/SRF pathway following activation of MAPK cascades. Recent studies of mice harboring conditional deletions of the SRF gene have corroborated these earlier findings by demonstrating that activity-dependent induction of multiple immediate early genes, including c-fos and egr-1, is markedly impaired in these mice (Ramanan et al. 2005). Of note, Ramanan et al. (2005) did not detect alterations in neuronal morphology when SRF deletion is triggered by Cre recombinase under the control of synapsin I or CamKIIα promoters. However, subsequent studies have demonstrated that deletion of SRF earlier in development does elicit prominent alterations in neuronal migration and morphology (Alberti et al. 2005; Knöll et al. 2006). Although it is unclear whether these morphological effects are due to loss of the MAL-SRF or TCF-SRF pathways, our results suggest that the former is involved. Knöll et al. (2006) have reported that dominant negative MAL constructs also reduce outgrowth of neuronal processes in hippocampal cultures. The study of SRF knock-out mice indicates that the neurite outgrowth is triggered by SRF-mediated transcription of a set of genes encoding guidance receptors and synapse formation. We have also checked the effect of dominant negative MAL constructs on neurite outgrowth in our culture system and have found that they cause a slight reduction in dendrite length of neurons at 8 DIV (Fig. 5c). These findings suggest that further studies aimed at assessing the role of the MAL-SRF pathway on neuronal morphology in vivo are warranted. In particular, while these studies establish that MAL affects dendritic morphology, further studies are needed to test directly whether its effects on neuronal morphology are mediated via activation of SRF.

In previous studies, we have studied activation of the MAL-SRF pathway in cultured cortical neurons by Tech, a RhoA GEF enriched in hippocampus and cortex (Tabuchi et al. 2005). We found that constitutively active Tech constructs trigger nuclear translocation of MAL and stimulate SRF-mediated transcription. Of note, Tech, by activating RhoA, also produces a marked reduction in the number of dendritic processes elaborated by cortical neurons (Marx et al. 2004). The results of the present study indicate that the MAL-SRF pathway acts to stabilize or increase the number of dendritic processes. Accordingly, we infer that activation of the MAL-SRF pathway by RhoA may represent a compensatory response to the retraction of dendrites caused by the direct effect of RhoA pathways on the actin cytoskeleton.

As we have found that cortical cultures have a basal level of SRF-mediated transcription that can be blocked by dominant negative or siRNA MAL constructs, these studies indicate that there is an endogenous signal driving this response. As Rho signaling pathways have been linked to activation of the MAL-SRF pathway, these represent likely candidates. Accordingly, it would be interesting in future studies to identify the upstream signaling pathways that regulate the activity of the MAL-SRF pathway in neurons. Furthermore, the ability of this transcriptional pathway to regulate dendritic morphology suggests that identification of its target genes in neurons would help elucidate the molecular events that regulate the dynamic changes in dendritic morphology observed in both developing and mature neurons.

Acknowledgements

We thank Dr R. Treisman (Cancer Research, London, UK) for providing us with dominant negative MAL constructs and the 3D.ALuc reporter construct. We thank Yen Wu (Johns Hopkins University, Baltimore, MD, USA) for assistance with immunohistochemical studies. This study was supported by a grant-in-aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (project number: 17790055, AT) and by research grants from Hayashi Memorial Foundation for Female Natural Scientists (AT) and from Foundation of the first Bank of Toyama (AT).

Ancillary