ALS-linked mutations in two chaperones of the protein disulfide isomerases (PDIs) family disrupt neuromuscular function, according to a new study published April 15 in The EMBO Journal. This finding suggests that PDI dysfunction may play a role in early events in the pathogenesis disease, at least in a subset of ALS cases.
“There is a strong likelihood that these ALS-linked PDI variants serve as risk factors or disease modifiers in ALS,” according to Smita Saxena, of the University of Bern, Switzerland, who was not involved in the study.
The PDIs are a large family of proteins that modify sulfur-sulfur bonds in target proteins as part of protein homeostasis pathways in the endoplasmic reticulum (ER). Two members of the family, PDI1A and Erp57, were discovered to be the most upregulated proteins in SOD1 mouse spinal cord in a proteomic analysis (Atkin et al., 2006), and subsequently in human sporadic ALS spinal cord (Atkin et al., 2008), as well as in patient blood (Nardo et al., 2011). In 2015, Claudio Hetz of the University of Chile in Santiago, together with Robert Brown at University of Massachusetts Medical School in Worcester, MA, screened for mutations in these two genes in over 100 familial and sporadic cases of ALS and 1000 control subjects, and identified missense variants in both at much higher frequency than in controls (Gonzalez-Perez et al., 2015) .
That led co-senior authors Hetz and Brown to explore, in the current study, the effects of these variants on the motor neuron function. First author Ute Woehlbier, of the University of Chile in Santiago, with colleagues from Chile, Finland and the US, expressed two variants of each of the two genes in zebrafish. Three of the four produced embryonic morphologic abnormalities. Among fish with normal appearance, the authors found that expression of the PDI variants led to motor axons with abnormal shape and, for two of the variants, disrupted synapses. These same variants, one of PDIA1 and one of Erp57, also caused motor performance deficits, a phenotype that resembled previously reported results from ALS studies in zebrafish (Babin et al., 2012).
In motor neuron cultures, including induced pluripotent stem cell (iPSC)-derived human motor neurons, minimal overexpression of the wild-type proteins promoted neurite outgrowth, an effect that was lost when the mutant variants were expressed instead. That same loss of outgrowth was found when the wild-type proteins were knocked down with interfering RNAs, suggesting that an important function of the two PDIs is to enhance neuronal outgrowth.
The variants of PDA1A, the authors found, reduced substrate specificity and altered the protein’s catalytic activity, with one variant increasing and the other decreasing the protein’s ability to reduce its substrate’s sulfide bonds to the disulfide form. Variants in Erp57 altered the binding to the chaperones calnexin and calreticulin, which participate in the folding of disulfide-containing proteins in the ER.
Because these results suggested that Erp57 variants impaired the protein’s normal function, the team next examined the effects of Erp57 deletion in mice. Haploinsufficiency led to a decline in rotarod performance, and loss of strength and coordination on the hanging test; mice with complete loss of the protein fared worse. But complete protein loss did not cause motor neuron loss; instead, it disrupted neuromuscular junctions. “This was entirely unexpected, and suggests we may be modeling early stages in the initiation of the disease” Hetz said.
A potential explanation for this disruption was the discovery that loss of ERp57 led to a decrease in the expression of synaptic vesicle protein SV2 in brain (though not in spinal cord, possibly because of generally lower levels of the protein in the spinal cord, the authors suggested). Like many synaptic proteins, mature SV2 is folded at the ER and reaches the synapse though the secretory pathway. The observed loss of expression was accompanied by aggregation of high molecular weight forms of SV2, possibly indicating that SV2 is a target of Erp57, and that reduction in the foldase activity had led to accumulation of the immature form of SV2.
The effects of a decline in Erp57 activity may not be limited to SV2, Hetz suggested. “We predict that many other proteins synthesized through the same secretory pathway are not going to be folded properly when these PDIs are affected, altering motoneuron proteostasis” he said.
While the described variants are rare in ALS cases, a reduction in PDI function may not be. “Nitrosylation at the active site, which inactivates the enzymes, is well described in mouse models and in human tissue,” Hetz pointed out. In this scenario, PDI dysfunction may be an early, though not precipitating, event in ALS pathogenesis, at least in some cases. “I think this is likely to be a component of the disease process for a subset of patients,” he said, “maybe by mutation, or maybe by inactivation through inflammation or oxidative stress.”
If this is the case, forced upregulation of the PDIs may help maintain the neuromuscular junction in the face of such stresses, a hypothesis Hetz and colleagues are gearing up to test using gene therapy.
“This study again highlights the importance of ER proteostasis to ALS pathology,” said Saxena, “and that the effects of impaired ER homeostasis are not just restricted to the ER or motor neuron soma, but can have far reaching consequences,” including at the NMJ. Saxena contributed a News and Views article about the study to The EMBO Journal.
“Due to the complexity of mechanisms involved in ER functioning, elucidating the role of individual components in disease pathology is often difficult,” added Gundars Goldsteins of the University of Eastern Finland in Kuopio, who was not involved in the study. “The current study is a clear step forward in understanding how alterations in PDI function, responsible for protein oxidative folding in general, may result in a selective defects resulting in motoneuron degeneration.” Goldsteins suggested that a key area for future investigation is whether the critical pathogenic events happen in microglia or astrocytes, rather than, or in addition to, motor neurons.
ER protein homeostasis has been linked to ALS through multiple studies, but often through the effects of SOD1 or other disease-causing mutations on the ER system, noted Adam Walker of Macquarie University in Sydney, Australia, who was not involved in the study. “This is a really interesting study, because it is the first to look at it from the other end, so to speak,” at the effect of the ER protein homeostasis network on the susceptibility to ALS. Now, he said, it will be critical to express these variants in the standard ALS models, to see if they hasten or exacerbate the disease process. Those experiments are underway, Hetz said.
1. Woehlbier U, Colombo A, Saaranen MJ, Pérez V, Ojeda J, Bustos FJ, Andreu CI, Torres M, Valenzuela V, Medinas DB, Rozas P, Vidal RL, Lopez-Gonzalez R, Salameh J, Fernandez-Collemann S, Muñoz N, Matus S, Armisen R, Sagredo A, Palma K, Irrazabal T, Almeida S, Gonzalez-Perez P, Campero M, Gao FB, Henny P, van Zundert B, Ruddock LW, Concha ML, Henriquez JP, Brown RH, Hetz C. ALS-linked protein disulfide isomerase variants cause motor dysfunction. EMBO J. 2016 Apr 15;35(8):845-65.[Pubmed].
2. Majarajan, N and Saxena, S. ER strikes again: Proteostasis Dysfunction in ALS. EMBO J. 2016 Apr 15;35(8):798-800.[Pubmed].
Homepage image credit: Neuromuscular junction – model. John Wildgoose, Wellcome Images