|
||||||||
|
NEUROSCIENCE Ion Channel Traced to Gene ExpressionL-type Calcium Channel in Neural Membrane Signals to Nucleus Through MAP Kinase Pathway Feats of neural plasticity like development, learning and memory hinge on the brain's capacity to continually reinvent itself. To do so requires exquisite tools and timing. When a nerve cell depolarizes, for example, in response to an outside signal, the event may trigger gene transcription in the faraway nucleus. Ion channels, peptide hormones, growth factors, and a host of regulatory proteins and transcription factors may come into play. All must coordinate precisely to convey the signal from the cell membrane to particular nuclear targets.
In the neural membrane, L-type calcium channels (LTC) sense local calcium concentration to activate gene transcription. Calcium-free calmodulin (CaM) is bound to the closed LTC channel in the resting neuron. The LTC opens in response to membrane depolarization and calcium enters the cell. After binding calcium, calmodulin shifts to the IQ domain of the channel and possibly triggers a change in LTC conformation. Michael Greenberg, Ricardo Dolmetsch (l to r, below), and colleagues propose that the calmodulin/LTC complex recruits scaffolding proteins that can activate kinases in the MAPK/Erk pathway, eventually leading to phosphorylation of CREB in the nucleus. Photo by Steve Gilbert
A study led by Michael Greenberg, HMS professor of neurology at Children's Hospital, and published in the Oct. 12 Science, offers a model for how one neural membrane component, the L-type calcium channel, might activate gene expression using a special calcium ion sensor. Just how calcium controls transcription is controversial. To clarify the issue, postdoc Ricardo Dolmetsch developed a novel pharmacological knock-in technique to modify the channel. He found that an L-type channel/calmodulin complex can detect local calcium at the mouth of the protein and that calcium binding to the complex may stimulate conformational changes that specifically activate the Ras/mitogen-activated protein kinase (MAPK/Erk) pathway. Greenberg thinks that they have just started to uncover the complexity of interactions. "If you want to make the analogy to growth factor receptors," he said, "there are going to be a lot of proteins bound to the L-type channel and also to the transcription complexes." The new knock-in technique, he says, offers a good, systematic way of testing among the various functions. The Calcium ConnectionAlthough the L-type calcium channel carries only a small fraction of the calcium influx into a depolarized primary neuron, it is responsible for the calcium-dependent activation of transcription triggered by several nuclear transcription factors, including the cAMP response element binding protein, or CREB. After nuclear kinases phosphorylate CREB, it recruits the molecular machinery necessary for transcription.Greenberg refers to CREB as a "sensitive barometer." It is the target of kinases from many interacting signaling pathways, and as such, it is finely tuned to changes in the nuclear environment. Yet this raises a question: in the company of the promiscuous and abundant kinases, what makes the link between mercurial CREB and the L-type channel specific? This is precisely the problem Dolmetsch wished to tackle. "In a sense," explained Dolmetsch, "this work represents the confluence of two different fields." On the one hand, there were studies that looked at signaling pathways, and on the other, there were structural studies on the channel. "We knew that L-type channels were particularly good at inducing genes," he said. "But to understand how, we needed to find out what regions of the molecule were involved." And that's where work from a third field came in handy. In its search for effective cardiovascular agents for hypertension, the pharmaceutical industry had developed a number of highly specific drugs, the dihydropyridine L-type channel blockers. Dolmetsch wanted to introduce the L-type channel into a cell and to distinguish it from endogenous L-type channels. To do this, he constructed a mutant that was insensitive to dihydropyridines and transfected it into primary cultured neurons. By exposing such cells to dihydropyridines and other channel blockers, endogenous L-type channels were blocked. Under these conditions any effects observed on CREB-activated transcription could be attributed to the introduced channel. Using the dihydropyridine-insensitive mutant, the researchers could begin to investigate L-type channel domains critical for signal transduction by introducing mutations in a second, cytoplasmic region of the protein. Transcription Ins and OutsEarlier work had shown the dynamics of CREB activation to be particularly intriguing. Normally, the transcription factor is phosphorylated within a minute of membrane depolarization and remains so for at least 40 minutes. But in the presence of dihydropyridines, CREB phosphorylation returned to resting levels after 20 minutes. This pattern suggested that phosphorylation occurred in two phases, of which only the latter was L-type channel-dependent. Since transcription from a CREB-dependent reporter gene was completely eliminated in the presence of dihydropyridines, this suggested that transcription activation occurred in the late, sustained phase.In order to understand what structural aspects of the channel were involved in this regulation, Dolmetsch focused on the channel's IQ region, a cytoplasmic domain known to bind calmodulin. By creating a series of systematic mutations in both the IQ region of the L-type channel and calmodulin, he showed that upon membrane depolarization, calmodulin binds calcium and shifts to the IQ site from a second site on the calcium channel. If this transition was blocked, there was a corresponding decrease in CREB-dependent transcription, suggesting that calcium binding allows calmodulin to bind to the IQ domain and that this binding activates transcription. Furthermore, the MAPK/Erk pathway shut down, as did CREB-activated transcription, in neurons expressing an L-type channel mutated in the IQ region. This pathway is therefore linked to the calcium-dependent binding of calmodulin and to gene activation. A Unifying ModelGreenberg hopes that the results of the current study will clarify a longstanding controversy. The field has been split into two camps: one believes activation of CREB is mediated through the family of calcium/calmodulin-dependent protein (CaM) kinases. "On the other hand," said Dolmetsch, "the Erk kinases can also activate CREB, but they tend to be activated later and for prolonged periods of time." Dolmetsch and Greenberg have proposed a model based on results from the study to account for this discrepancy. They say that a large increase in calcium concentrations in the nucleus may lead to CaM kinase activation and transient CREB phosphorylation. But for reasons the researchers do not understand, phosphorylation is only transient. "To get really robust transcription, what you need is sustained CREB phosphorylation and that is best accomplished with this other pathway, the Erk pathway," said Dolmetsch."A key question that remains unanswered," said Greenberg, "is how calcium, which is sensed locally by calmodulin, gets to Ras." Dolmetsch has preliminary evidence that suggests scaffolding proteins associate with other regions of the channel. "One challenge for the future will be to figure out which molecules mediate particular responses," Greenberg said. The researchers propose that the dynamics of CREB activation are a result of the mechanics of the channel operation. Upon membrane depolarization, the channel opens, then inactivates partially. During the initial phase, calcium floods the nucleus. As the second phase begins, and ion flow is curtailed, the calcium concentrated around the mouth of the channel becomes the dominant factor. Calmodulin binds calcium, then slides onto the IQ domain, perhaps inducing a conformational change that activates the MAPK/Erk pathway. "This is one of the reasons why the L-type channel is effective," said Dolmetsch. "It has this local calcium-sensing mechanism, whereas other channels do not." Greenberg adds, "It's starting to look like everyone was right. This study reconciles the data into a unified hypothesis for CREB regulation." --Anne Mahon |
