SEARCH

SEARCH BY CITATION

Keywords:

  • Astrocytes;
  • GFAP ;
  • intermediate filaments;
  • Alexander disease;
  • neurodegenerative diseases

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References

Alexander disease (AxD) is a neurodegenerative disorder with prominent white matter degeneration and the presence of Rosenthal fibers containing aggregates of glial fibrillary acidic protein (GFAP), and small stress proteins HSP27 and αB-crystallin, and widespread reactive gliosis. AxD is caused by mutations in GFAP, the main astrocyte intermediate filament protein. We previously showed that intermediate filament protein synemin is upregulated in reactive astrocytes after neurotrauma. Here, we examined immunohistochemically the presence of synemin in reactive astrocytes and Rosenthal fibers in two patients with AxD. There was an abundance of GFAP-positive Rosenthal fibers and widespread reactive gliosis in the white matter and subpial regions. Many of the GFAP-positive reactive astrocytes were positive for synemin, and synemin was also present in Rosenthal fibers. We show that synemin is expressed in reactive astrocytes in AxD, and is also present in Rosenthal fibers. The potential role of synemin in AxD pathogenesis remains to be investigated.

Alexander disease is the first known primary disorder of astrocytes. It is an autosomal dominant disease caused by missense mutations in the glial fibrillary acidic protein (GFAP) gene [1], which encodes GFAP, a member of the type III intermediate filament (nanofilament) protein family and the main astrocyte intermediate filament protein [2].

Alexander disease is a neurodegenerative disorder with prominent degeneration of the white matter and with fatal outcome. Its prevalence is estimated to be 1/2 700 000 in Japanese population [3]. The typical onset is before 2 years of age with the death most commonly occurring in the first decade of life. The condition is characterized by the presence of Rosenthal fibers that are found throughout the brain and spinal cord, especially in white matter, subpial, perivascular, and periventricular astrocytes, and are often surrounded by activated astrocytes and prominent reactive gliosis (for review see [4]).

Ultrastructurally, Rosenthal fibers contain aggregates of polymerized GFAP and electron dense deposits with small stress proteins HSP27 and αB-crystallin [5]. The pathogenesis of Alexander disease remains unknown but the presence in the Rosenthal fibers of αB-crystallin, which was shown to induce disassembly of GFAP filaments [6], might reflect a protective stress response against the toxic effect of mutant GFAP.

We recently showed that intermediate filament protein synemin is upregulated in reactive astrocytes after neurotrauma [7]. In the brain, synemin needs vimentin to associate with GFAP and integrate into intermediate filament networks in response to injury [7]. Here, we have examined immunohistochemically the presence of synemin in reactive astrocytes and Rosenthal fibers in two patients with Alexander disease. We report that both reactive astrocytes and Rosenthal fibers contain synemin.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References

For hematoxylin and eosin (H&E) staining, tissue was embedded in paraffin, cut into 5–8-μm-thick serial sections with a microtome, baked for 1 h at 65 °C in an oven and stained with H&E. For immunohistochemical staining in Fig. 1, deparaffinized and hydrated sections were boiled for 10 min in 10 mM citric acid (pH 6.0), blocked for 30 min at room temperature (RT) in 0.3% H2O2 in methanol, rinsed in dH2O, and blocked for 30 min at RT in 20% horse serum (Life Technologies Ltd, Paisley, UK) in PBS. Then, sections were incubated for 30 min at RT in either goat anti-synemin (1:200; Reglia AB, Gothenburg, Sweden) or mouse anti-GFAP (1:500, G3893; Sigma-Aldrich, St. Louis, MO, USA) in PBS plus 0.05% Tween 20 (P1379; Sigma-Aldrich), washed in PBS, incubated for 10 min at RT in biotinylated pan-specific IgG (one drop per ml PBS, BA-1300; Vector Laboratories, Burlingame, CA, USA), washed in PBS, and incubated for 30 min at RT in Vectastain ABC kit (PK-6100; Vector Laboratories). For antigen detection, sections were incubated for 8 min at RT in 3,3′-diaminobenzidine (DAB) with nickel (SK-4100; Vector Laboratories), rinsed in dH2O, incubated for 3 min at RT in eosin (1:2) in dH2O, washed in dH2O, and finally dehydrated and mounted with Vecta Mount medium (H-5000; Vector Laboratories). For combined immunohistochemical analysis (Fig. 2), sections were incubated with rabbit anti-synemin antibodies (1:200; 7) in 1% BSA in PBS; overnight at 4 °C, washed and incubated with donkey anti-rabbit Alexa 488 (1:1000, A11034; Life Technologies; 1 h at RT), washed, incubated with rabbit anti-GFAP antibodies (1:200, Z0334; DAKO, Glostrup, Denmark; overnight at 4 °C) in 1% BSA in PBS, washed, incubated for 1 h at RT in donkey anti-rabbit Alexa 555 (1:1000, A21429; Life Technologies) and TOPRO-3 (1:1000, T3604; Life Technologies) in PBS and mounted. In negative controls, the primary antibody was omitted.

image

Figure 1. Rosenthal fibers and reactive astrocytes in two patients with Alexander disease were positive for both glial fibrillary acidic protein (GFAP) and synemin. (A and G) Rosenthal fibers in the cortical white matter (arrows) were visualized by hematoxylin and eosin (H&E) staining. (B and H) GFAP immunohistochemical analysis in the white matter tissue showed GFAP-positive reactive astrocytes (arrows) and GFAP-positive Rosenthal fibers (C and I). Reactive astrocytes were positive for synemin (D and J arrows). Immunohistochemical analysis with antibodies against synemin combined with eosin staining showed synemin positivity in Rosenthal fibers (E and K) and in reactive astrocytes (F and L). Scale bar, 50 μm in A, B, D, G, H, J and 20 μm in C, E, F, I, K, L.

Download figure to PowerPoint

image

Figure 2. Colocalization of glial fibrillary acidic protein (GFAP) and synemin in reactive astrocytes and Rosenthal fibers in Alexander disease. Combined immunohistochemical analysis with antibodies against GFAP (red) and synemin (green) on a confocal image showed that white matter reactive astrocytes contained both GFAP and synemin (an arrow), and Rosenthal fibers were positive for both GFAP and synemin (arrowheads). Scale bar, 20 μm.

Download figure to PowerPoint

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References

The first patient developed clinical signs of Alexander disease at 5 months and died at 12 years of age. She carried a point mutation in GFAP at position 373 with glutamate being replaced by lysine [8]. H&E-stained cortical sections exhibited abundance of Rosenthal fibers, a typical morphological feature of Alexander disease (Fig. 1A). GFAP immunohistochemical analysis showed widespread reactive gliosis in the white matter (Fig. 1B). At a higher resolution, GFAP-positive Rosenthal fibers were apparent in the white matter and subpial regions (Fig. 1C).

Many reactive astrocytes in the white matter were positive for synemin (Fig. 1D). Immunohistochemical analysis with antibodies against synemin combined with eosin histochemical staining revealed synemin-positive Rosenthal fibers (Fig. 1E) as well as synemin-positive reactive astrocytes (Fig. 1F). Synemin expression was seen either throughout Rosenthal fibers or only on the fiber periphery. The latter phenomenon might be explained by the presence of cellular processes of reactive astrocytes with intermediate filaments containing synemin in close apposition of the fibers.

The second patient developed clinical signs of Alexander disease at 7 months and died at 13 years of age. She carried a point mutation in GFAP at position 79 with arginine being replaced by serine. H&E-stained cortical sections exhibited abundance of Rosenthal fibers (Fig. 1G). As in the first patient, GFAP immunohistochemical analysis showed widespread reactive gliosis in the white matter (Fig. 1H) with highly abundant Rosenthal fibers (Fig. 1I). Synemin expression (Fig. 1J) was seen in both Rosenthal fibers (Fig. 1K) and in reactive astrocytes (Fig. 1L).

Combined immunohistochemical analysis of tissue from the first patient with antibodies against GFAP and synemin confirmed that reactive astrocytes as well as Rosenthal fibers contained both GFAP and synemin intermediate filament proteins (Fig. 2).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References

In two cases of Alexander disease, the intermediate filament protein synemin was present in Rosenthal fibers as well as in many surrounding reactive astrocytes. This expands the spectrum of neuropathologies with synemin expression in reactive astrocytes from neurotrauma [7] to Alexander disease. Astrocyte activation and reactive gliosis accompany many CNS pathologies, e.g., trauma, stroke, or neurodegenerative diseases. Reactive astrocytes were proposed to have a positive role in the acute stage after neurotrauma [9-11], or stroke [12] and they seem to inhibit later plasticity-promoting and regenerative responses [9, 13-15] (for review see [16-19]). The role of astrocyte activation and reactive gliosis in neurodegenerative diseases has recently been implicated [20, 21], but remains incompletely understood. The mutations in the GFAP gene that cause Alexander disease result in a protein product that forms aggregates instead of assembling into filaments. These aggregates might contribute to the disease pathogenesis by interfering with signaling pathways important for cell survival [22, 23] and impairing the ability of astrocytes to deal with stress [24, 25]. Whether synemin presence in Rosenthal fibers or reactive astrocytes in Alexander disease affects the disease pathogenesis remains to be addressed. The potential involvement of synemin in protein aggregates that occur in more common neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease, or amyotrophic lateral sclerosis warrants further investigation.

Conflict of Interest

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References

The authors have no conflict of interest.

We would like to thank the late Prof. Milan Elleder for his involvement in this project. Rita Grandér, Ann-Marie Alborn, and Prof. Klas Blomgren are acknowledged for technical help. This work was supported by grants from the Swedish Medical Research Council (11548), AFA Insurance, Swedish Stroke Foundation, Torsten and Ragnar Söderberg Foundations, the Swedish Society for Medicine, the Region of Västra Götaland (RUN), Frimurare Foundation, Amlöv's Foundation, Sten A. Olsson Foundation for Research and Culture, Foundation Edit Jacobson's Donation Fund, Trygg-Hansa, Hjärnfonden, ALF Göteborg (11392), the EU FP7 Programs EduGlia (237956) and TargetBraIn (279017), and NanoNet COST Action (BM1002).

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conflict of Interest
  7. References