A mutation—a change in a single gene—is termed “pathogenic” when it alters the gene product in a way that causes disease. Within the human genome, there are more than 20,000 genes. Each one of the cells within our body contains all of these 20,000 genes. Yet many mutations selectively impact some tissues or cell types, leaving others unaffected. An example is provided by sickle cell anemia. In this hereditary disorder, a mutation of the HBB gene results in production of an abnormal form of β-globin which is a component of hemoglobin, the iron-containing protein that carries oxygen within the blood from the lungs to other tissues throughout the body. Hemoglobin is present only in red blood cells, and these cells are profoundly affected by the HBB mutation, taking on a sickle shape and failing in their oxygen-carrying duty, while other types of cells within the body are unaffected. There are other genes, however, that are expressed in multiple cell types, and when these genes are mutated, each of these cell types is at risk. The paper that follows (Rush et al. 2006) illustrates an interesting complexity of biology: A single mutation can have different effects in different cell types. This phenomenon is, in part, a result of differences in “cell background.” The mutant protein may interact with one set of protein partners in cell type A, while interacting with a different set of protein partners in cell type B, or may be subject to one set of biological processes in cell type A while being subject to a different set of biological processes in cell type B.
The NaV1.7 sodium channel is present at high levels in two types of peripheral neurons: DRG sensory neurons that innervate the body, including pain-signaling neurons; and sympathetic ganglion neurons, located on either side of the spinal column as the most distal outposts of the sympathetic component of the autonomic nervous system. Sympathetic ganglion neurons participate in the regulation of many homeostatic mechanisms within the body including the control of blood vessel diameter.
Without even entering the laboratory, a skilled observer might conclude that autonomic function is affected in erythromelalgia, since the pain attacks in this disorder are accompanied by dramatic reddening of the skin. Although there was no evidence at the molecular level, this aspect of erythromelalgia had been attributed by others (Mork, Kalgaard, and Kvernebo 2002; Davis et al. 2003) to an abnormality of blood vessels or disturbed vasomotor control, i.e., an abnormality of neurons such as sympathetic ganglion neurons that control the width of blood vessels that irrigate the skin. Indeed, some patients who do not carry mutations of NaV1.7 develop “secondary” erythromelalgia as a result of having disorders of the blood that interfere with blood flow, such as abnormalities of the platelets that interfere with circulation, and these patients respond well to treatment with aspirin, which inhibits the action of platelets (Drenth, van Genderen, and Michiels 1994).
We addressed this issue in a paper by Rush et al. (Rush et al. 2006) by assessing the effect of a NaV1.7 mutation that had been identified in a patient with painful inherited erythromelalgia within both DRG neurons and neurons from a sympathetic ganglion (the superior cervical sympathetic ganglion). We expected from our previous studies that the mutation would produce hyperexcitability within DRG neurons, but in the absence of any prior experimental results we did not know what to expect in sympathetic ganglion neurons. In the initial part of this study, we expressed the NaV1.7 mutant channel in DRG neurons, and then in sympathetic ganglion neurons. In DRG neurons, we found that the mutant channels depolarized the resting membrane potential, which we had expected because the mutation hyperpolarizes channel activation. The shift in activation increases the overlap between activation and inactivation so as to produce an enhanced persistent current called, by electrophysiologists, a “window” current. Consistent with the idea that inappropriate firing of pain-signaling neurons underlies the pain experienced by this patient, we found that the mutation produces hyperexcitability of DRG neurons, manifested by a decrease in threshold (the amount of current needed to produce a single action potential) and an increase in the firing rate in response to graded suprathreshold stimulation.
The results in sympathetic ganglion neurons were strikingly different. We found that, as expected, the mutation depolarizes the resting potential of sympathetic ganglion neurons. To our surprise, however, we observed that, rather than rendering these cells hyperexcitable, the mutant channels had the opposite effect, reducing the excitability of sympathetic ganglion neurons, by increasing their threshold and decreasing their firing rate in response to graded suprathreshold stimulation. Repeating our experiments several times, we again and again observed that a NaV1.7 mutant channel could have different functional effects in different types of cells. Indeed, the NaV1.7 mutant channels produced opposite effects in the two types of neurons that express NaV1.7, hyperexcitability in pain-signaling sensory neurons and hypoexcitability in sympathetic ganglion neurons.
We had expected that the mutant channels might have different effects on the behavior of sensory and sympathetic ganglion neurons, but we had not anticipated the dramatic result of the experiment, with a single mutation having totally opposite effects in two different types of neurons. When a scientist makes an unexpected observation, one next step is to ask “why?” Reasoning that the effect of the mutant channel within each specific cell type would be shaped not just by the presence of the NaV1.7 mutant channel, but also by the presence of the mutant channel within a specific cell background—which reflects the entire ensemble of ion channels present in each type of neuron—we measured mRNA and protein for various types of sodium channels in the two types of cells. This analysis showed us that DRG neurons express five types of sodium channels, NaV1.1, NaV1.6, NaV1.7, NaV1.8, and NaV1.9, while sympathetic ganglion neurons express a different ensemble of sodium channels, NaV1.3, NaV1.6, and NaV1.7.
This is a very important difference, especially since the mutant NaV1.7 depolarizes cells. Most sodium channels are “inactivated” or put to sleep so they are not available for operation, by prolonged depolarization. The NaV1.8 sodium channel, which was originally called SNS because it is sensory neurons specific, is different. NaV1.8 is present in DRG neurons but not in sympathetic ganglion neurons. And NaV1.8 is remarkable in that, unlike other sodium channels which are silenced by depolarization, NaV1.8 is relatively resistant to inactivation (Akopian, Sivilotti, and Wood 1996; Cummins and Waxman 1997). We had shown a few years earlier that a major functional role of NaV1.8 is to support high-frequency firing in response to sustained depolarization in cells to which it is present (Renganathan, Cummins, and Waxman 2001). These earlier results suggested that the presence of NaV1.8 in DRG neurons, and its absence in sympathetic ganglion neurons, could have accounted for the different effects of the NaV1.7 mutation in these two types of cells. We set out to test the hypothesis, illustrated in figure 5.1, whereby (1) the NaV1.7 mutant channel depolarizes both sympathetic ganglion neurons and pain-signaling DRG neurons; (2) as a result of this depolarization, the sodium channels within sympathetic ganglion neurons are inactivated and it is harder for these cells to generate action potentials so that they become hypoexcitable; but (3) because NaV1.8 channels are present within pain-signaling DRG neurons, and are not inactivated by depolarization, these cells do not become hypoexcitable but, on the contrary, become hyperexcitable, because the depolarization brings the membrane potential closer to threshold for activation of the NaV1.8 channels. According to this hypothesis, the presence or absence of a single type of molecule, the NaV1.8 sodium channel, was critically important in determining whether a neuron would become hyperexcitable or hypoexcitable as a result of expression of NaV1.7 mutant channels within it.
Figure 5.1 Scheme by which NaV1.7 mutations, which depolarize resting membrane potential, produce hyperexcitability in dorsal root ganglion (DRG) neurons and hypoexcitability in sympathetic ganglion neurons. Note the enhanced action potential activity below the DRG neuron expressing the NaV1.7 mutation, in contrast to the electrical silence of the sympathetic ganglion neuron.
We tested this hypothesis in the experiment shown in figure 5 in Rush et al. (2006). In this experiment we asked whether we could convert a sympathetic ganglion neuron into a cell with properties similar to those of a DRG neuron, by inserting NaV1.8 channels within it. Insertion of the gene for NaV1.8 alongside the gene for the mutant NaV1.7 channel within sympathetic ganglion neurons was not easy, but Sulayman Dib-Hajj and our molecular biology team were successful in this effort, and the newly inserted genes produced their channels within their new host cells. Remarkably, we found that, indeed, it is possible to flip a sympathetic ganglion neuron expressing a mutant NaV1.7 channel from hypoexcitability, to hyperexcitability, by inserting NaV1.8 channels within it. We had essentially isolated a single molecule—NaV1.8—that was responsible for the opposing effects of NaV1.7 mutant channels within the two cell-types where NaV1.7 is present. These experiments taught us two lessons: first, that a single mutation can have dramatically different effects on the firing properties of different types of neurons; and second, that NaV1.8 acts as a molecular switch, with its presence or absence determining whether NaV1.7 mutations produce hyper- or hypoexcitability.
There was even more to the story. In subsequent studies, carried out together with Frank Rice, an expert on skin, we showed that NaV1.7 is also present within vascular myocytes, the muscle cells that form the walls of blood vessels within the skin (Rice et al. 2015). In these muscle cells, NaV1.8 is not present. We hypothesized that NaV1.7 mutations that depolarize resting potential produce hypoexcitability within these vascular muscle cells, thereby perturbing blood flow within the skin. We are currently testing this hypothesis, which would introduce another factor that contributes to the abnormal reddening of the limbs in people with erythromelalgia.
It is not often that one gets to see, in precise detail, how molecules like ion channels interact with each other. The Rush et al. (2006) paper provides an example of how a team effort—in this case combining ion channel biophysics, cellular electrophysiology, and molecular biology—allowed us to dissect, molecule by molecule, how a mutant ion channel interacts with other ion channels in its cellular environment.