Journal of Molecular Medicine

, Volume 86, Issue 11, pp 1243–1254

Multiple therapeutic effects of valproic acid in spinal muscular atrophy model mice

  • Li-Kai Tsai
  • Ming-Shiun Tsai
  • Chen-Hung Ting
  • Hung Li
Original Article

DOI: 10.1007/s00109-008-0388-1

Cite this article as:
Tsai, LK., Tsai, MS., Ting, CH. et al. J Mol Med (2008) 86: 1243. doi:10.1007/s00109-008-0388-1


Spinal muscular atrophy (SMA) is a progressive disease involving the degeneration of motor neurons with no currently available treatment. While valproic acid (VPA) is a potential treatment for SMA, its therapeutic mechanisms are still controversial. In this study, we investigated the mechanisms of action of VPA in the treatment of type III-like SMA mice. SMA and wild-type mice were treated with VPA from 6 to 12 months and 10 to 12 months of age, respectively. Untreated SMA littermates and age-matched wild-type mice were used for comparison. VPA-treated SMA mice showed better motor function, larger motor-evoked potentials, less degeneration of spinal motor neurons, less muscle atrophy, and better neuromuscular junction innervation than non-treated SMA mice. VPA elevated SMN protein levels in the spinal cord through SMN2 promoter activation and probable restoration of correct splicing of SMN2 pre-messenger RNA. VPA also increased levels of anti-apoptotic factors, Bcl-2 and Bcl-xL, in spinal neurons. VPA probably induced neurogenesis and promoted astrocyte proliferation in the spinal cord of type III-like SMA mice, which might contribute to therapeutic effects by enhancing neuroprotection. Through these effects of elevation of SMN protein level, anti-apoptosis, and probable neuroprotection, VPA-treated SMA mice had less degeneration of spinal motor neurons and better motor function than untreated type III-like SMA mice.


Animal modelApoptosisMotor neuron diseaseNeurogenesisSpinal muscular atrophyValproic acid



amyotrophic lateral sclerosis


adenosine triphosphatase




choline acetyltransferase


compound muscle action potential


glial fibrillary acidic protein


histone deacetylase


neuronal nuclear


spinal muscular atrophy


survival motor neuron


valproic acid


von Willebrand factor

Supplementary material

109_2008_388_Fig1_ESM.gif (538 kb)
Supplementary Fig. 1

Immunohistological analysis of the spinal anterior horn in untreated SMA (S–, ac, n = 4), VPA-treated SMA (S+, df, n = 4), untreated wild-type (W−, gi, n = 4) and VPA-treated wild-type mice (W+, n = 4). Spinal cord sections were stained for choline acetyltransferase (ChAT, red) and SMN (green) and for nuclei using DAPI (blue). ChAT-immunopositive cells are divided into three categories, including cells without SMN immunoreactivity (SMN(−), empty arrow), with partial SMN immunoreactivity (SMN(+/−), white arrow), and with full SMN immunoreactivity (SMN(+)). j VPA-treated SMN mice showed higher percentage of SMN(+) cells and lower percentage of SMN(−) cells than untreated SMA mice. k Higher gem (arrowhead) numbers in the nuclei could be detected in spinal neurons of VPA-treated SMA mice in comparison with untreated SMA mice. Scale bars 50 μm; *P < 0.05; **P ≤ 0.01 (GIF 550 kb)

109_2008_388_Fig1_ESM.tif (742 kb)
High resolution image (TIF 759 kb)
109_2008_388_Fig2_ESM.gif (1.1 mb)
Supplementary Fig. 2

Immunohistological analysis of the muscles in a negative control (not adding SMN antibody), b untreated SMA, c valproic acid (VPA)-treated SMA, and d untreated wild-type mice. Muscle sections were stained for SMN (green) and for nuclei using DAPI (blue). There was more SMN signal in muscles of VPA-treated than non-treated SMA mice. Scale bars 50 μm (GIF 1189 kb)

109_2008_388_Fig2_ESM.tif (2.5 mb)
High resolution image (TIF 2639 kb)
109_2008_388_Fig3_ESM.gif (535 kb)
Supplementary Fig. 3

Histological analysis of the spinal anterior horn in untreated SMA (S−, a, d, n = 4), VPA-treated SMA (S+, b, e, n = 4), untreated wild-type (W−, c, f, n = 4) and VPA-treated wild-type mice (W+, n = 4). Spinal cord sections were stained for Bcl-2 (ac, red) or Bcl-xL (df, red) with NeuN (green) and for nuclei using DAPI (blue). NeuN/Bcl-2 or NeuN/Bcl-xL double-labeled cells appeared to be yellow, while NeuN-immunopositive but Bcl-2 or Bcl-xL-immunonegative cells displayed green color. g VPA-treated SMA mice showed a higher percentage of NeuN-immunopositive cells displaying Bcl-2 immunopositive signal in the spinal cord as compared with untreated SMA mice. h VPA-treated SMA mice had a higher percentage of NeuN immunopositive cells displaying Bcl-xL-immunopositive signal in the spinal cord than untreated SMA mice. Scale bars 50 μm; *P < 0.05; **P ≤ 0.01 (GIF 547 kb)

109_2008_388_Fig3_ESM.tif (859 kb)
High resolution image (TIF 880 kb)
109_2008_388_Fig4_ESM.gif (366 kb)
Supplementary Fig. 4

Blots of the indicated figures for reactivity of antibodies used in our manuscript. For experimental design, see legends of the indicated figures and the “Materials and methods” section (GIF 374 kb)

109_2008_388_Fig4_ESM.tif (510 kb)
High resolution image (TIF 522 kb)

Copyright information

© Springer-Verlag 2008

Authors and Affiliations

  • Li-Kai Tsai
    • 1
  • Ming-Shiun Tsai
    • 2
  • Chen-Hung Ting
    • 3
  • Hung Li
    • 3
    • 4
  1. 1.Department of NeurologyNational Taiwan University Hospital and National Taiwan University College of MedicineTaipeiTaiwan
  2. 2.Department of Bioindustry TechnologyDa-Yeh UniversityChanghuaTaiwan
  3. 3.Institute of Molecular BiologyAcademia SinicaTaipeiTaiwan
  4. 4.Department of Neurology, Center for NeuropsychiatryChina Medical University and HospitalTaichungTaiwan