Dystrophin-Associated Protein Scaffolding in Brain Requires α-Dystrobrevin

Abstract

Dystrophin and the α-dystrobrevins bind directly to the adaptor protein syntrophin to form membrane-associated scaffolds. At the blood-brain barrier, α-syntrophin co-localizes with dystrophin and the α-dystrobrevins in perivascular glial endfeet and is required for localization of the water channel aquaporin-4. We have previously shown that localization of the scaffolding proteins γ2-syntrophin, α-dystrobrevin-2, and dystrophin to glial endfeet is also dependent upon the presence of α-syntrophin. In the present study, we show that the expression levels of α-syntrophin, γ2-syntrophin, and dystrophin at the blood-brain barrier are reduced in α-dystrobrevin-null mice. This is the first demonstration that assembly of an astroglial protein scaffold containing syntrophin and dystrophin in perivascular astrocytes is dependent upon the presence of α-dystrobrevin.

INTRODUCTION

Since the discovery of dystrophin as the product of the Duchenne muscular dystrophy gene, numerous proteins have been identified which associate with dystrophin to form membrane-based scaffolds [reviewed in refs. ¹ and ²]. These dystrophin-associated scaffolds provide structural integrity to the cell by connecting the extracellular matrix with the cytoskeleton and create intracellular docking points for regulatory proteins. The dystrobrevins are a family of proteins that interact directly with dystrophin and syntrophin adaptor proteins providing a link to a network of signaling proteins that regulate key cellular processes such as water transport and nitric-oxide-mediated events [3–5] Two separate dystrobrevin genes named α- and β-dystrobrevin have been identified, both of which yield multiple transcripts [6–10]. α- and β-dystrobrevin are both expressed in brain but have distinct locations [8].

The α-dystrobrevin isoforms (currently six have been identified) have been characterized most thoroughly in muscle where several are required for proper formation of the neuromuscular synapse [11]. However, the dystrobrevins are also present in the brain where they are present in both neurons and glia[10–16]. Immunofluorescence studies show αDb at glial endfeet and endothelial cells at the blood-brain barrier [13,15], whereas αDb-1 and -2 localize specifically to glial endfeet [12,16]. We have previously shown that the localization of αDb-2 to glial endfeet at the blood brain barrier is specifically dependent upon the presence of α-syntrophin [16]. Using the α-dystrobrevin-null mouse, we now show that the expression levels of α-syntrophin, γ2-syntrophin, and dystrophin in the vasculature are dependent on α-dystrobrevin. These data indicate that α-syntrophin and α-dystrobrevin are mutually dependent on each other and that α-dystrobrevin plays a critical role in assembly of the DAPC at the blood-brain barrier.

METHODS

The syntrophin antibodies used in this study were mAb1351 (Affinity Bioreagents; MA1-745), which recognizes α-, β1-, and β2-syntrophin , and syntrophin-isoform specific affinity-purified rabbit polyclonal antibodies to α-syntrophin (Syn17; [17]) and γ2-syntrophin (Ab 212; [16,18,19]). The dystrophin (MANDRA-mouse monoclonal) was purchased from Sigma (St. Louis, MO) and the α-dystrobrevin antibodies (Ab 638, Ab αDB2-rabbit polyclonals) have been characterized previously [20,21].

For immunofluorescence, animals were euthanized in accordance with the NIH guide for the care of laboratory animals. After light sedation with CO₂ gas, 6-week and 3-month-old male C57Bl6 and α-dystrobrevin-null mice [22] and corresponding wild type littermates were quickly euthanized and brains were dissected. Cerebella were removed, bisected and placed in PBS containing 4% paraformaldehyde/10% sucrose for 12 min on ice. Cerebellar tissue was then transferred to PBS containing 25% sucrose and placed at 4°C overnight. Tissue was embedded in OCT and frozen in isopentane cooled with liquid nitrogen. Twenty µm thick cryosections were adhered to superfrost plus slides (VWR) and stored at −80°C. Immunoflurescence and confocal images were collected under identical conditions, as described previously [15].

RESULTS

The assembly of the dystrophin-associated protein complex at the blood-brain barrier is dependent upon the presence of α-syntrophin [16]. Specifically, α-syntrophin is required for the localization of γ2-syntrophin, α-dystrobrevin-2, and dystrophin to the perivascular astrocyte endfoot membrane. Since localization of αDB-2, but not αDB-1, requires α-syntrophin, we questioned how the DAPC would be affected by the loss of α-dystrobrevin isoforms. We performed immunolocalization for α-syntrophin on brains from α-dystrobrevin null animals [Figure 1D–F; ²²] and wild-type littermates (Figure 1A–C). The merged image in Figure 1C indicates that, in wild type mice, α-syntrophin co-localizes with α-dystrobrevin 1 in perivascular astrocyte endfeet, as reported previously. Furthermore, we observed a striking decrease in the level of α-syntrophin immunofluorescence in the α-dystrobrevin-null mice (compare 1E to 1B).

Figure 1. α-Syntrophin is Reduced at the Blood-Brain Barrier in the Absence of α-Dystrobrevin.

We performed double-immunolabeling for α-dystrobrevin-1 (red, A, D) and syntrophin (green, B, E) in wild type (A–C) and α-dystrobrevin-null (D–F) adult mouse cerebellum. α-Syntrophin and α-dystrobrevin-1 co-localize in perivascular astrocyte endfeet (see merged image, 1C). We observed a reduction in fluorescence intensity for α-syntrophin in the brains of α-dystrobrevin-null mice (E), indicating that localization of α-syntrophin to perivascular astrocyte endfeet is dependent on α-dystrobrevin. The pan-syntrophin antibody (mAb 1351; B, E) or an α-syntrophin-specific rabbit polyclonal antibody (syn17, data not shown) was used to recognize α-syntrophin, and an αDb-1 specific rabbit polyclonal [26] was used to localize αDb-1. As expected, αDb-1 is absent in the α-dystrobrevin-null cerebellum (D). C and F are merged images of [A, B] and [D, E], respectively. Scale bar, 25 µm.

We have previously shown that the localization of γ2-syntrophin to perivascular endfeet is dependent on α-syntrophin [16]. Since we find that expression of α-syntrophin at glial endfeet is dependent on the presence of α-dystrobrevin, we determined whether the localization of γ2-syntrophin was disrupted by the loss of α-dystrobrevin. We compared γ2-syntrophin immunolocalization (green, Fig 2B, E) in wild-type (Figure 2A–C) and α-dystrobrevin-null (Figure 2D–F) cerebellum. γ2-Syntrophin immunofluorescence was reduced in the surrounding cerebellar vasculature in the absence of α-dystrobrevin (compare 2E to 2B). As in the α-syntrophin null animals [16], γ2-syntrophin expression is unaffected in neurons (see arrows, Figure 2). Taken together, these data indicate that expression of γ2-syntrophin in non-neuronal cells is dependent on both α-syntrophin and α-dystrobrevin.

Figure 2. γ2-Syntrophin is Reduced at the Blood-Brain Barrier in the Absence of α-Dystrobrevin.

We performed double-immunolabeling for α-dystrobrevin-1 (red, A, D) and γ2-syntrophin (green, B, E) in wild type (A–C) and α-dystrobrevin-null (D–F) adult mouse cerebellum. As shown previously [16], γ2-syntrophin localizes to Purkinje neurons and to astrocyte endfeet in the adult cerebellum (B). In the absence of α-dystrobrevin (E), γ2-syntrophin is reduced in astrocyte endfeet but unaffected in Purkinje neurons. As expected, αDb-1 labeling is lost in the α-dystrobrevin-null animals (D). C and F are merged images of [A, B] and [D, E], respectively. Scale bar, 25 µm.

Dystrophin colocalizes with α-syntrophin in both the granular layer and the molecular layer of the cerebellum. The α-Db-1 and -2 isoforms exhibit discrete distributions in the vasculature of these layers but both colocalize with α-sybtrophin. Thus, dystrophin and a form of α-dystrobrevin are colocalized in the vasculature throughout the cerebellum. Dystrophin localization at the blood-brain barrier is contingent on the presence of α-syntrophin [16]. Since α-syntrophin localization is disrupted in the α-dystrobrevin-null mice, the localization of dystrophin might also be affected. We assessed immunofluorescence for dystrophin (green, Fig 3B, E) and αDb-2 (red, Fig 3A, D) in cerebellum from wild type (Fig 3A–C) and α-dystrobrevin-null mice (Fig 3D–F). As shown previously, αDb-2 is present in the granular layer of cerebellar cortex (Fig 3A, see arrows). As expected, the α-dystrobrevin-null cerebellum is negative for α-Db2 (Fig 3D). Dystrophin is associated with the blood-brain barrier throughout the cerebellar cortex (Fig 3B), and its expression levels are substantially reduced in the α-dystrobrevin-null mouse (compare 3E to 3B).

Figure 3. Dystrophin is Reduced at the Blood-Brain Barrier in the Absence of α-Dystrobrevin.

We performed double-immunolabeling for α-dystrobrevin-2 (red, A, D) and dystrophin (green, B, E) in wild type (A–C) and α-dystrobrevin-null (D–F) adult mouse cerebellum. αDB-2 is restricted to the blood-brain barrier in the granular layer of cerebellum (see arrows, A), whereas dystrophin localizes to the blood-brain barrier throughout the cerebellar cortex (B). In α-dystrobrevin null cerebellum (D–F), labeling for αDB-2 is absent (D), and immunolocalization of dystrophin is reduced (E). Arrows in C denote co-localization of αDB-2 and dystrophin in the granular layer of cerebellum. C and F are merged images of [A, B] and [D, E], respectively. mol, molecular layer; gran, granular layer. Scale bar, 25 µm.

DISCUSSION

Taken together with our previous published results [16], the main finding of this work is that α-syntrophin and α-dystrobrevin are mutually dependent upon one another for their expression and/or localization in the cerebrovasculature. Our earlier conclusion [16] that α-syntrophin is the central regulator of the DAPC at endfeet must now be modified to include the α-dystrobrevin/α-syntrophin subcomplex. Whether other syntrophin isoforms are required remains to be determined. Interestingly, co-dependence of α-syntrophin and α-dystrobrevin occurs also in skeletal muscle fibers. In mice lacking α-syntrophin, sarcolemmal staining intensity for αDB-1 and αDB-2 is also decreased [23]. Likewise, in α-dystrobrevin-null mice, levels of α-, β1- and β2-syntrophin were also decreased [11,22].

How might these protein members of the dystrophin complex be mutually dependent on each other for proper localization? One possibility is that they are assembled into a vesicular complex [24,25] prior to transport to glial endfeet. In this capacity, loss of either protein may have effects downstream in the trafficking pathway. Alternatively, α-syntrophin and α-dystrobrevin may act in concert to retain components of the DAPC at the cellular plasma membrane. Future experiments using endothelial-astrocyte co-cultures from normal and α-syntrophin null mice may be informative in revealing the dynamics of DAPC complex trafficking, assembly, and stability at the blood-brain interface.

Loss of either α-syntrophin or α-dystrobrevin results in a reduction of dystrophin from the vasculature. In skeletal muscle, dystrophin is considered the central component of the dystrophin-associated protein complex whose presence is critical for sarcolemmal localization of α-dystrobrevins and certain syntrophin isoforms [11,22,23]. The abence of α-dystrobrevins and syntrophins do not affect sarcolemmal dystrophin levels. Thus, despite some similarities discussed above, assembly and/or maintenance of this dystrophin-associated subcomplex may be regulated differentially in brain and skeletal muscle.

CONCLUSION

We show for the first time that the loss of α-dystrobrevin results in the decreased expression of α-syntrophin, γ2-syntrophin, and dystrophin in astrocytes. Taken together with our previous study [16], these data indicate that α-syntrophin and α-dystrobrevin are both essential for the proper assembly and maintenance of the dystrophin-based scaffold in the cerebrovasculature. Therefore, altered expression of either of these proteins could compromise both the structural and signaling functions of this scaffold.