24 June 2008. Scientists are beginning to get a handle on a rare inherited form of amyotrophic lateral sclerosis (ALS). ALS8 is caused by mutation of the gene encoding VAPB (vesicle-associated membrane protein (VAMP)-associated membrane protein B), but exactly why this mouthful of a protein can cause ALS is unclear. Working independently, two research groups have now come to the same conclusion, that is, that aggregation of VAPB as a direct result of the ALS8 mutation leads to the sequestration of wild-type protein and, hence, loss of its normal function. What remains less clear is what that function is—at least in humans. Writing in the June 13 Cell, researchers led by Hugo Bellen at Baylor College of Medicine, Houston, Texas, report that the N-terminal of VAPB normally gets cleaved, secreted from the cell, and serves as a ligand for Ephrin receptors in fruit flies. In contrast, George Jackson and colleagues at University of California, Los Angeles, report that VAPB is important for bone morphogenetic protein signaling, which supports formation of fly neuromuscular junctions. This report is published in the June PLoS ONE. It is not known if these functions are mutually exclusive. Even so, they may both lead to a better understanding of the role of VAPB in inherited, and perhaps also in sporadic forms of ALS. “We believe this protein brings together all of the key aspects known from the disease. This is useful because we think by doing work in flies and C. elegans, we can put a pathway together that will incorporate many of the known genes that are involved in ALS,” said Bellen in an interview with ARF.
ALS8 is extremely rare (see Landers et al., 2008). So far VAPB mutations have only been found in certain families of Brazilian descent. All the same, the protein may be a bellwether for most forms of the disease, suggested Bellen. Recently, researchers in the Netherlands reported that VAPB levels are reduced in the spinal cord of patients with sporadic ALS and also in the spinal cord of SOD mouse models of the disease (see Teuling et al., 2007). “Although the mutation is unique and the disease is rare, it seems that this protein is involved so far in all cases that have been looked at,” said Bellen.
Researchers led by Mayana Zatz at the University of Sao Paolo, Brazil, were the first to report the link between the VAPB mutation (P56S) and ALS (see Nishimura et al., 2004). Since then studies have reported that the mutated protein forms aggregates in neurons (see Teuling et al., 2007 and Chai et al., 2008). The current papers support this idea and delve more deeply into the function of the protein.
Bellen and colleagues focused on the N-terminal of the protein. All VAPs contain a conserved domain called the major sperm protein (MSP), which in C. elegans is secreted into the reproductive tract, whereupon it binds to oocyte Eph receptors and functions in fertilization. In humans the function of VAPs is not at all understood, but first author Hiroshi Tsuda and colleagues now show that in Drosophila, the MSP domain of the fly homolog to VAPB is also cleaved from the protein and secreted. Using antibodies to specific epitopes on the protein, they found that human VAP (hVAP) expressed in flies suffers the same fate, and in human leukocytes protein fragments form that also correspond with the size of the human MSP domain. In fact, in human blood the researchers detected only the MSP domain. “Taken together, our data indicate that VAP MSP domains are secreted and suggest that the hVAP MSP domain is found in human serum,” write the authors.
What of the ALS-causing VAP mutant? Tsuda and colleagues found that the MSP domain of the mutant protein fails to be secreted in fly wing discs—unlike the wild-type—and that the protein ends up ubiquitinated and in aggregates in the cytoplasm. In addition, when the researchers looked at wing disc cells under the transmission electron microscope, they found abnormalities in the endoplasmic reticulum (ER). That, the presence of the ER marker Boca in the mutant VAP aggregates, and the upregulation of BiP/Hsc3, which is involved in the ER unfolded protein response (UPR), led the authors to conclude that the VAP mutation induces the UPR in vivo. The unfolded protein response has also been linked to Alzheimer disease pathology (see ARF related news story).
Bellen said that the effects of the mutation are complex, with an overall dominant-negative effect. The failure of MSP secretion in the mutant protein might suggest a loss of function, while the aggregation might work like a gain of function. On top of that, the aggregates recruit wild-type protein as well, which contributes to even greater loss of MSP cleavage and secretion, he said.
What might be the effect of losing secretion of the MSP domain? Tsuda and colleagues found that the MSP domain of VAPB plays an important role in Eph signaling, and that this role is crucial for the proper formation of the neuromuscular junction, where overexpressing wild-type dVAP increases bouton number and decreases bouton size (see Pennetta et al., 2002).
For their part, Jackson and colleagues report an important role for VAPB in the neuromuscular junction through a different signaling pathway. First author Anuradha Ratnaparkhi and colleagues found that the VAP P58S mutant (the fly equivalent of the human P56S mutation) impairs the activity of wild-type VAP in vivo, acting as a dominant-negative. They also found that it does this by attracting wild-type VAP into protein aggregates.
Ratnaparkhi and colleagues found two major correlates of mutant VAP expression in flies. First, they found that mutant VAP led to disorganization of microtubules in the neuromuscular junction (NMJ). This is in keeping with previous observations that VAP associates with these structures (see Skehel et al., 1995). In addition, the researchers observed “floating T bars” at the NMJ, which are electron-dense structures coupled to neurotransmitter vesicles that are not attached to the presynaptic membrane. These structures, which are not present in normal neuromuscular junctions, have been observed before, most notably in mutants of the bone morphogenetic protein (BMP) signaling pathway.
Could VAPB and BMP signaling pathways overlap? The authors used a “bristle” phenotype to test this relationship. Overexpression of wild-type VAP results in loss of hairs (bristles) in the fly thorax, but overexpressing a dominant-negative BMP receptor (thickvein) suppressed the bristle loss, which is consistent with the two proteins working in the same signaling pathway. “We also found that the mutant VAP protects against the bristle phenotype, which again suggests a dominant-negative effect,” said Jackson in an interview with ARF. He added that the bristle phenotype could be useful in genetic screens for other genes that affect the disease.
How will these findings change the study of ALS? “It is going to be interesting to see whether BMP, or in the case of Hugo’s work, Ephrin receptors, are legitimately involved in the role of VAP in humans,” suggested Jackson. That may take some doing, considering there are 16 Ephrin receptors in humans that have been implicated in just about everything one could think of, from sorting cells, to growth cone migration and collapse, to blood vessel wall tightness, to clustering of glutamate receptors. In the case of BMP, some evidence of a link already exists. BMP signaling leads to phosphorylation of SMAD transcription factors, which are found in round hyaline inclusions in spinal cord neurons from sporadic ALS patients (see Nakamura et al., 2008). These inclusions also contain TDP-43, a protein that may be a defining marker for ALS (see ARF related news story). Both Bellen and Jackson said that there is no direct evidence yet linking VAPB with TDP-43. Interestingly, blocking SMAD signaling may protect against AD-like pathology in mice (see ARF related news story).
One suggestion that has emerged from studying TDP-43 in ALS is that the disease involves much more than motor neurons (see ARF related news story). From the perspective of VAPB, Bellen agrees. “My gut feeling is that motor neurons are just more sensitive [to the mutant VAPB], but other neurons have problems as well, so simply solving the motor neuron problem in this disease would not be sufficient. It would probably delay the disease, maybe by years, but it is not going to solve the fundamental problem,” he suggested.
Can this new VAPB knowledge help to solve that fundamental problem? One idea is that simply supplying the MSP domain, which fails to be secreted from the cells when the protein ends up in intracellular aggregates, might help rescue neurons, much like insulin helps patients with diabetes. “That would be the hope,” said Bellen. But he said first it has to be determined if patients are actually deficient in the protein, and second, it will have to be shown that the peptide can rescue ALS phenotypes in mice.
The MSP domain may help explain the paracrine nature of ALS. The disease is not solely cell-autonomous, since expressing mutant SOD in glia, for example, can lead to neurodegeneration (see ARF related news story). That makes sense if there is a secreted molecule that either contributes to or prevents pathology. VAPB meets this criterion and also the cell-autonomous one, since accumulation of the protein in the ER is probably toxic to the host cell itself.—Tom Fagan.
Tsuda H, Han SM, Yang Y, Tong C, Lin YQ, Mohan K, Haueter C, Zoghbi A, Harati Y, Kwan J, Miller MA, Bellen HJ. The amyotrophic lateral sclerosis 8 protein VAPB is cleaved, secreted, and acts as a ligand for Eph receptors. Cell 2008, June 13;133:963-977. Abstract
Ratnaparkhi A, Lawless GM, Schweizer FE, Golshani P, Jackson GR. A Drosophila model of ALS: Human ALS-associated mutation in VAP33A suggests a dominant negative mechanism. PLoS ONE. 2008 Jun 4;3(6):e2334. Abstract