The paper explained
PROBLEM:
Wolfram Syndrome (DIDMOAD) is an incurable disease characterized by a range of endocrine and neurological symptoms, including diabetes insipidus, diabetes mellitus, blindness due to optic atrophy and sensorineural deafness. Mutations in the genes encoding Wolframin (an endoplasmic reticulum protein involved in ion channel activity and the unfolded protein response) or Miner1 (a 2Fe‐2S cluster‐containing protein) lead to the development of DIDMOAD. Full understanding of DIDMOAD is hindered by limited insight into the role of Miner1 in the cell.
RESULTS:
We used fibroblasts immortalized from WT and Miner1 KO mouse embryos to elucidate the cellular processes impacted by Miner1. Deletion of Miner1 results in a more oxidized intracellular milieu, and this is associated with ER stress and the unfolded protein response, profound ER and mitochondrial Ca2+ dysregulation, and altered mitochondrial structure and function. Importantly, treatment with the sulphydryl anti‐oxidant N‐acetyl cysteine ameliorates many of the mitochondrial and ER‐stress related anomalies observed in the Miner1 KO cells.
IMPACT:
Our studies form the basis of a paradigm for the function of Miner1 and provide novel mechanistic insights into the underlying pathology of DIDMOAD. These experiments strongly suggest that oxidative stress lies at the root of the ER and mitochondrial changes caused by Miner1 deficiency, and importantly that sulphydryl anti‐oxidant treatment could be a rational therapeutic approach for treating this syndrome.
INTRODUCTION
Wolfram Syndrome (DIDMOAD) is an incurable disease characterized by a range of endocrine and neurological symptoms, including diabetes insipidus, diabetes mellitus, blindness due to optic atrophy and sensorineural deafness. Wolfram Syndrome is also associated with an increased incidence of psychiatric disorders and a significantly shortened life span, averaging only 30 years (Barrett et al,
1995; Strom et al,
1998). Two unrelated genes have been shown to be mutated in Wolfram Syndrome (Amr et al,
2007; Sam et al,
2001). These genes, originally designated as
WFS1 and
WFS2, have since been demonstrated to encode the Wolframin and Miner1 proteins.
Mutations in Wolframin (
WFS1), the first gene linked to Wolfram Syndrome, are responsible for the majority of Wolfram Syndrome cases. Recently, single nucleotide polymorphisms (SNPs) of the
WFS1 gene have also been implicated in the pathogenesis of type 2 diabetes (Wasson & Permutt,
2008).
WFS1 encodes a 100 kDa integral membrane protein of the ER that lacks known catalytic domains (Hofmann,
2003). Although there is some evidence that WFS1 is involved in Ca
2+ homeostasis and influences the stability of the ER stress sensor ATF6 (Fonseca et al,
2010; Osman,
2003; Takei et al,
2006), the exact function of WFS1, its regulation, and the molecular mechanisms linking its function to Wolfram Syndrome and type 2 diabetes are far from resolved.
WFS2 encodes the Miner1 protein (aka: ERIS, CISD2). We previously identified a small family of proteins with high sequence similarity that includes mitoNEET, Miner1 and Miner2 (Wiley et al,
2007a). The current names for the genes encoding these proteins are
CISD (CDGSH Iron Sulphur Domain)
1,
2 and
3, respectively. CDGSH domains are characterized by the presence of a Cys‐Asp‐Gly‐Ser‐His motif. MitoNEET and Miner1 each contain a single carboxy‐terminal CDGSH domain, while Miner2 possesses tandem CDGSH domains. In addition to the CDGSH domain, there is a predicted hydrophobic sequence at the amino‐terminus of Miner1. The CDGSH domain binds a redox active 2Fe2S cluster that is coordinated by an uncommon arrangement of 3 Cys and 1 His (Wiley et al,
2007b). The structures of the CDGSH domains of mitoNEET and Miner1 are virtually identical (Conlan et al,
2009; Paddock et al,
2007). The molecules exist as dimers with a unique fold. The His that ligates the 2Fe2S cluster is surface exposed and exquisitely sensitive to protonation if the pH is less than 7.0, resulting in release of the 2Fe2S cluster, a unique property among FeS cluster‐binding proteins. This feature has led to the speculation that this family of proteins may be involved in FeS cluster assembly or their mobilization within the cell; alternatively, it may function in a redox capacity, as has been found for other 2Fe2S cluster proteins. Our initial data suggested that Miner1 with a C‐terminal EGFP tag localized to the ER (Wiley et al,
2007a). However, a recent study concluded that Miner1 is predominantly a mitochondrial protein (Chen et al,
2009).
In an effort to locate a previously mapped longevity gene linked to human chromosome 4q (Puca et al,
2001), Tsai and colleagues generated mice in which the Miner1 (
Cisd2) gene is deleted (Chen et al,
2009). These mice have a complex and dramatic phenotype and, indeed, appear to be a remarkable model of early aging, in addition to recapitulating many of the features of Wolfram Syndrome. Within 3 weeks after birth, pups lacking Miner1 show signs of optic and sciatic nerve atrophy. These mice develop normally for a short period and then proceed to display sarcopenia, thinning of the subcutaneous fat, hair loss, hair greying, osteopenia, lordokyphosis and significantly shortened lifespan. Although
Cisd2 KO mice are not overtly diabetic, their glucose tolerance is impaired.
Our understanding of the biological function of CDGSH domain proteins is still in its infancy. The phenotype of the Cisd2 KO mice suggests that Miner1 is crucial for the maintenance of multiple organ systems throughout the body, including the pancreas, skin, musculoskeletal and nervous systems. Miner1 appears to be at the nexus of metabolism and lifespan control. Insights into the functions of Miner1 will not only provide knowledge regarding the etiology of Wolfram Syndrome, but should also shed light on an important new regulatory protein linking metabolic disease and aging.
Given the importance of ER/mitochondrial interactions to metabolic regulation, we have used mouse embryonic fibroblasts (MEFs) derived from Miner1 WT and KO mice to investigate the role of Miner1 in maintaining proper ER function and ER‐mitochondrial communication. Miner1 KO cells displayed a dramatic reduction in ER Ca2+ and profound mitochondrial Ca2+ loading. Although mitochondrial respiratory capacity was increased in the KO cells, there was an increase in the ADP/ATP ratio and impaired cell proliferation. Miner1 deficient cells also displayed signs of oxidative stress and initiation of the unfolded protein response (UPR). Remarkably, treatment with the anti‐oxidant N‐acetylcysteine (NAC) reversed many of the molecular abnormalities caused by Miner1 deletion. Current treatments for Wolfram Syndrome focus on managing the diabetic symptoms; however, none of these reverse the degenerative course of the disease. Our data suggest that a potential therapeutic approach utilizing sulphydryl anti‐oxidants, such as NAC, could more effectively target the underlying cause of the disease.
DISCUSSION
Wolfram Syndrome was long thought to be a disease of reduced ATP supply, suggesting mitochondrial dysfunction. This view was based in part on bioenergetic deficits detected in patients with this disease. In addition, the tissues affected tend to be those with high energetic demands (Bundey et al,
1992). The data presented here suggest that the metabolic dysfunction in WFS2 arises not from a primary mitochondrial defect, but as an indirect consequence of altered Ca
2+ homeostasis at the ER caused by an alteration of the sulphydryl redox status. We and others have previously demonstrated that a recombinant CDGSH domain protein expressed in
Escherichia coli is redox active
in vitro (Wiley et al,
2007b). However, data supporting a redox role for Miner1 in cells has been lacking. In the current study, we have established the ability of Miner1 to influence the general redox status in the cell. Miner1 deficient cells displayed changes in NAD
+/NADH ratio, NO production, GSH/GSSG ratio, protein glutathionylation and oxidation of sulphydryl groups, including those in CX
5R phosphatases. We found that Miner1 KO MEFs have a more oxidized cytoplasmic environment and show signs of oxidative stress. The orientation of the Miner1 CDGSH domain on the cytoplasmic face of the ER is consistent with a role in regulating cytoplasmic redox status, and Miner1 KO cells exhibit a general increase in sulphydryl oxidation. The catalytic Cys of the PTP CX
5R family of phosphatases is exquisitely sensitive to sulphydryl oxidation, resulting in inactivation of the enzymes. Thus, these phosphatases can be viewed as protein sentinels reflecting the cellular redox status. The increase in oxidation of catalytic Cys in phosphatases detected in the mutant MEFs is likely to have wide‐ranging effects on signalling by reversible phosphorylation.
In addition to redox regulation of phosphatase activity, the oxidative environment of the Miner1 KO MEFs could lead to oxidation of regulatory cysteines on ER Ca
2+ transporters and channels. Oxidative modifications of ER Ca
2+ transporters have been shown to affect transport activity and lead to a decrease in Ca
2+ in the ER lumen, resulting from alteration of the activity of SERCA and increased leak via RyR or IP
3 receptors (Sammels et al,
2010). We observed an increase in SERCA2 glutathionylation in Miner1 deficient cells, suggesting that the decrease in ER Ca
2+ levels may be caused by changes in SERCA2 transport activity. Oxidation of IP
3 receptors may also contribute to the Ca
2+ dysregulation observed in the Miner1 KO cells. As a redox‐active protein anchored to the ER and enriched in MAMs, Miner1 is in a prime location to directly interact with these transporters and keep them in a reduced state.
Because many of the folding chaperones in the ER are Ca2+‐dependent, decreased ER Ca2+ content has been linked to ER stress and induction of the UPR. We found that MEFs derived from Miner1 KO mice also display the hallmark signs of ER stress/UPR. Miner1‐deficient MEFs exhibit elevated levels of the folding chaperone Bip and the pro‐apoptotic protein CHOP, as well as an expanded ER network.
Enhanced Ca
2+ leak from the ER as a result of oxidative protein modification can explain the increased mitochondrial Ca
2+ load that we observed in Miner1 KO cells and the increase in dephosphorylated PDH. Oxidation of RyR Ca
2+ channels have been associated with increased Ca
2+ leak from the ER, increased mitochondrial Ca
2+ loading and increased muscle wasting in mice (Andersson et al,
2011). Sarcopenia was also observed in the Miner1 KO mice (Chen et al,
2009). Enhanced ER Ca
2+ leak would also predict the elevated ADP/ATP ratio and the enhanced rate of ATP utilization we observed in the rates of endogenous O
2 consumption (Norris et al,
2010). These findings, along with a decreased growth rate, are consistent with an underlying mechanism of enhanced futile Ca
2+ cycling.
There is precedent for altered mitochondrial Ca
2+ loading, morphology and function caused by changes in ER protein function. Expression of a truncated isoform of SERCA1 results in ER Ca
2+ leak, mitochondrial Ca
2+ loading, changes in mitochondrial morphology and increased ER‐mitochondrial contacts (Chami et al,
2008). Similarly, mutations in presenilin2, an intramembrane protease that is localized to the ER and that is linked to the onset of Alzheimer's Disease, influences ER‐mitochondrial Ca
2+ cross‐talk and physical contacts (Zampese et al,
2011). Consistent with these paradigms of ER‐mitochondrial Ca
2+ dysregulation, we found that Miner1 KO MEFs have altered mitochondrial structure. Surprisingly, mitochondria isolated from Miner1 KO MEFs demonstrated much higher respiratory capacity than WT mitochondria, in contrast to the mitochondria isolated from muscle of Miner1 KO mice, which showed virtually no ADP‐stimulated respiration (Chen et al,
2009). It is possible that immortalization of the MEFs in culture promotes tolerance and adaptation to the mitochondrial Ca
2+ loading and elevated ROS, perhaps via a compensatory increase in cristae area and expression of ETC components. We predict that an electrically excitable cell type, such as a neuron or myocyte, would not tolerate this form of Ca
2+ dysregulation and would be more likely to suffer from mitochondrial dysfunction. The phenotype of a recently reported second mouse model of Miner1 deficiency is significantly milder than the mouse established by Chen and colleagues (Chen et al,
2009) (the source of our fibroblasts) and lacks features present in the human disease (Chang et al,
2012). However, the increase in cristae density that we observe in the fibroblasts is recapitulated in skeletal muscle myocytes of the mouse established by Chang and coworkers (Chang et al,
2012).
Wolfram Syndrome patients present with diabetes mellitus and diabetes insipidus, which result from compromised production of insulin by the pancreas and of vasopressin by the pituitary, respectively. Both of these are secretory organs that place a tremendous demand on the ER for protein folding. The chronic ER stress created by Miner1 deletion is likely to overwhelm the protein folding and secretory capacity of these organs in the short term, and ultimately lead to organ failure. The blindness and deafness observed in Wolfram Syndrome is of neuronal origin, and the chronic oxidative stress, mitochondrial Ca
2+ loading and lower ATP levels that we observed could lead to chronic dysfunction or death of electrically excitable cells. The sarcopenia observed in Miner1 KO mice could reflect low ER Ca
2+ levels in muscle. Lastly, oxidative stress has been correlated with both aging and the diabetic state (Giacco & Brownlee,
2010; Haigis & Yankner,
2010). Thus, our cell culture model has enabled us to develop a mechanistic model for the manifestations of Wolfram Syndrome observed in human patients and knockout mice (Supporting Information Fig S7).
Our studies form the basis of a new paradigm for the function of Miner1 and the etiology of Wolfram Syndrome. We conclude that Miner1 is a redox protein that resides in the ER and that regulates the UPR and mitochondrial function. Consequently, our data suggests that the defects observed in Miner1 MEFs stem from oxidative stress, which can be mitigated by treatment with anti‐oxidants. Indeed, application of the sulphydryl anti‐oxidant NAC reversed many of the mitochondrial and ER stress‐related anomalies observed in the Miner1 KO cells, hence anti‐oxidant treatment represents a potentially promising therapeutic strategy for treating this currently incurable disease.