Charcot-Marie-Tooth disease and spinal muscular atrophy are united by similar symptoms of progressive motor neuron degeneration—and now, by a gene that is linked to some subtypes of these diseases. Three research groups, working independently with different families, homed in on mutations in a calcium channel called transient receptor potential vanilloid 4 (TRPV4). The papers were published simultaneously online by Nature Genetics on December 27. All three groups found that the affected people in these families had mutations in the same region of TRPV4, an area involved in protein-protein interactions. The mutations appear to affect membrane localization and activity of the channel.
The researchers studied families with Charcot-Marie-Tooth disease type 2C (CMT2C) and two forms of spinal muscular atrophy (SMA)—scapuloperoneal SMA and congenital distal SMA. All involve progressive motor neuron degeneration that can cause symptoms affecting the limbs, diaphragm, and vocal cords. The symptoms of the three disorders overlap considerably, so the researchers suspected they might have the same cause. Affected people were heterozygous for TRPV4 mutations, which were not found in control subjects.
TRPV4 is a calcium channel that functions as a tetramer. It opens to allow calcium into the cell in response to a variety of signals, including increased temperature, a hypotonic environment, and mechanical force. The channel’s involvement in neuron degeneration suggests an obvious mechanism for pathology. “Calcium is a mechanism of excitotoxicity in virtually all neurons,” said Teepu Siddique of the Northwestern School of Medicine in Chicago, Illinois, senior author of one of the papers. Calcium toxicity has also been linked to the pathology of Alzheimer disease (e.g., see ARF related news story on Yang et al., 2009) and amyotrophic lateral sclerosis (e.g., see ARF related news story on Jiang et al., 2009). Mutations in a different region of TRPV4 have also been linked to skeletal dysplasia (Rock et al., 2008; Krakow et al., 2009).
The researchers used multiple families to narrow their focus to a region of chromosome 12, then sequenced several genes to identify TRPV4 as the cause of disease. One group, based primarily at the National Institute of Neurological Disorders and Stroke (NINDS) in Bethesda, Maryland, included joint first authors Guida Landouré of NINDS and Anselm Zdebik, who along with Landouré is affiliated with the University College London laboratory of Robert Kleta. The other principal investigators are Kenneth Fishbeck of NINDS and Charlotte Sumner, formerly of Fishbeck’s lab and now at the Johns Hopkins School of Medicine in Baltimore, Maryland. They studied two multigenerational families with CMT2C and found TRPV4 mutations leading to replacement of the arginine at position 269 with either cysteine or histidine (R269C and R269H).
Siddique and first author Han-Xiang Deng, also at Northwestern Medical School, spearheaded another group. They investigated one extended family with scapuloperoneal SMA as well the same CMT2C family studied by the NINDS group. They found the R269H mutation as well as an R316C substitution.
Michaela Auer-Grumbach of the Medical University of Graz, Austria, led the third team. They discovered the TRPV4 link in a family whose afflicted members showed CMT2C, scapuloperoneal SMA, or congenital distal SMA. They then sequenced that genetic locus in members of four other families to find R269H, R315W, and R316C substitutions.
All of these mutations affect TRPV4’s six ankyrin repeats, positively charged domains likely involved in protein-protein interactions. However, the researchers’ work diverged somewhat when they delved into the mechanism of the mutations, possibly because they studied TRPV4 in different cell types.
The NINDS and Northwestern groups came to similar conclusions studying TRPV4 mutations in cell types that also express endogenous TRPV4. The NINDS scientists used both Xenopus oocytes and HEK293 kidney cells for their experiments; the Northwestern researchers also used HEK293 cells.
TRPV4 normally localizes to the plasma membrane. The Northwestern researchers transfected HEK293 cells with both wild-type, R269H, and R316C TRPV4 constructs, and found similar plasma membrane localization for each. Both they and the NINDS scientists examined calcium levels in transfected cells—HEK293s or Xenopus oocytes—and determined that the presence of the R269C, R269H, and R316C TRPV4 mutations increased calcium levels. Therefore, these papers suggest that the TRPV4 mutations cause a gain of function, increasing calcium influx and potentially poisoning the cell.
The Austrian group came to a different conclusion using HeLa cervical cancer cells, which express no endogenous TRPV4. When they transfected TRPV4 into the cells, they saw that the wild-type localized to the plasma membrane, but the R269H, R315W, and R316C mutants instead formed cytoplasmic inclusions. Analyzing calcium levels, they found similar intracellular calcium concentrations in cells transfected with wild-type, R269H, R315W, or R316C TRPV4, but when they treated the cells with a hypo-osmotic solution to activate the channels, the mutants allowed less calcium influx than the wild-type channels. This work, then, suggests a loss of function in mutant TRPV4.
Two papers suggest normal localization of hyperactive mutant channels, and one suggests abnormal localization and less-active channels. How to explain the discrepancy? “Maybe there is some difference in the compensatory mechanisms of the cell itself,” Sumner suggested. Bernhard Keller of the University Medical Center of Göttingen, Germany, who was not involved with the studies, had a similar idea. “It is well known that alterations in calcium conductances of ion channels can show quite different effects on the overall cell physiology, depending on the localization and clustering of channels, their interaction with calcium-dependent enzymes, colocalization with calcium-dependent, transcription-relevant signal pathways, etc.” he wrote in an e-mail to ARF.
How do the mutations affect TRPV function? Because arginines are positively charged amino acids, the Austrian researchers hypothesized that loss of their positive charge was responsible for the pathogenic phenotype. Accordingly, they swapped R269 for a lysine, maintaining the positive charge. Cells expressing this mutant had calcium levels identical to wild-type cells, supporting their suspicions. This group also used immunohistochemistry to examine TRPV4 expression and localization in skeletal muscle from a person with CMT2C caused by an R316C substitution. The CMT2C tissue had less staining in the nerve fibers and in the cytoplasm of muscle cells than did tissue from a healthy person.
Ilya Bezprozvanny of the University of Texas Southwestern Medical Center in Dallas, who was also not part of the studies, put forth a hypothesis that brings the divergent results together. “If I had to bet, I would say that these arginines are involved in the interaction with PIP2” (phosphatidylinositol 4,5-bisphosphate), a cell membrane phospholipid, he told ARF. TRP channels are known to interact with PIP2 (reviewed in Nilius et al., 2008), and the positively charged ankyrin repeats are very typical of PIP2 binding sites, Bezprozvanny said.
Bezprozvanny proposed the following theoretical scenario: PIP2 is both a recruiter of TRPV4 and an inhibitor of its activity, and ankyrin domain mutations diminish the channel’s ability to interact with the phospholipid. In HeLa and other cells that lack endogenous TRPV4, the transfected mutant protein, unable to find the membrane without binding PIP2, remains cytoplasmic. This explains the loss of function in the Austrian group’s results. But in cell types such as HEK293 cells, where endogenous, wild-type TRPV4 is also present, the mutant and wild-type proteins sometimes come together in mixed tetramers. The interaction of the wild-type TRPV4 and PIP2 is sufficient to recruit the heterogeneous complex to the plasma membrane. However, the presence of the mutant channel means that PIP2 is not as effective an inhibitor of calcium influx as it normally is, thus causing the gain of channel function seen in the Northwestern and NINDS papers.
As for how mutations in one gene could cause so many different phenotypes, it is not unheard of, the Northwestern authors noted, for the same mutation to cause a wide spectrum of symptoms in different people. The scientists suggested that other hereditary or environmental factors could be responsible for the disparate diagnoses among their subjects.
The work suggests that TRPV4 channels could be a target for drug therapies. This is promising, Sumner said, because there are already successful treatments based on ion channels—for epilepsy, for example. Keller noted that membrane proteins are easier to target than cytoplasmic peptides. And TRPV4 therapies could have an impact beyond SMA and CMT, Siddique suggested, because these receptors have a variety of roles. Ultimately, TRPV4 treatments might be useful for conditions such as pain and pressure neuropathies or amyotrophic lateral sclerosis, another disease where motor neurons suffer excitotoxicity, he said.—Amber Dance