The study of rare families—families carrying uncommon mutations—can point to critically important genes and the protein molecules they produce; this, in turn, can pinpoint important molecular players and pivotal mechanisms in that inherited disease. But the history of modern medicine also shows us that, in some cases, these mutations can also teach us more general lessons. Studies on rare genetic diseases can point to molecules or mechanisms that are relevant to common disorders and can inform the development of new therapies that may be useful in all of us. An example is provided by the statin medications, which control the levels of lipids within our blood. Discovery of the statins has had an immense impact on health worldwide, reducing the incidence of heart attacks and strokes, most of which occur in the population at large. A critical advance was provided by the discovery and study of rare families in which heart disease occurred prematurely due to a genetic disorder—inherited hypercholesterolemia—in which high levels of cholesterol plug up blood vessels.6 These families provided a platform for the identification of mutations in specific genes, and identification of those genes, in turn, pointed the way to culprit molecules that could be targeted by a new class of medications.
Here, I briefly provide a perspective on how rare families—I call them “pointer-kindreds”—have provided multiple clues about chronic pain and how we might more effectively treat it.
Beginning with the discovery, in 2004 to 2005, that pain in inherited erythromelalgia (IEM) is the result of hyperexcitability of pain-signaling dorsal root ganglion (DRG) neurons; due to gain-of-function mutations of sodium channel NaV1.7,1,2 IEM has been studied as a human genetic model of neuropathic pain. Most patients with IEM do not respond to any of the existing medications.3 However, like many tenets in clinical medicine, there was an exception to this rule. In 2009, we reported a kindred containing multiple family members with IEM with a previously unseen therapeutic response as part of their phenotype; pain in multiple members of this family was relieved by the sodium channel blocker carbamazepine.4
Electrophysiological analysis revealed that the Nav1.7 mutation in this family (V400M) hyperpolarized channel activation as is usual in IEM, thus making it easier to turn on the channel and causing DRG neuron hyperexcitability. The mutation also depolarized the voltage dependence of inactivation, adding to the “window” of overlap between activation and inactivation. These changes, together with slowed deactivation of the mutant channel, would be expected to make DRG neurons hyperexcitable. The unusual finding was that this mutation also sensitized the channel to carbamazepine, via a previously undescribed mode of action whereby the drug had a depolarizing effect on activation, shifting it back toward its wild-type value.4 This was especially interesting because hyperpolarized activation plays a dominant role in conferring hyperexcitability on DRG neurons.10 At that time, we did not follow-up on this action of carbamazepine in these mutant channels.
Two years later, Payandeh et al.9 published the crystal structure of the bacterial voltage-gated sodium channel. A postdoc in my laboratory, Yang et al.,12 promptly built upon these data to construct a structural model of the human NaV1.7 sodium channel with near-Ångstrom resolution. Reasoning that the location of the V400M mutation within the 3-dimensional folded structure of the channel might be critical for the pharmacoresponsiveness to carbamazepine, we used V400M as a “seed” and interrogated the model, looking for other IEM mutations that were located structurally close to V400M. This analysis revealed a second IEM mutation, S241T, that is located 159 amino acids away from V400M in the linear (unfolded) amino acid sequence, but only 2.4Å from V400M, measured from the closest hydrogen within the 3-dimensional folded channel (Fig. 1). This observation from structural modeling suggested that the V400M and S241T mutations are functionally coupled during activation. To test this, we performed thermodynamic mutant cycle analysis (a method that permits analysis of whether 2 residues at different sites within a protein molecule are energetically coupled). These experiments provided strong data for energetic coupling of V400M and S241T. We next hypothesized that the atomic proximity and energetic coupling of V400M and S241T might be paralleled by pharmacological coupling. To test this prediction, we exposed cells expressing the S241T mutant channel to clinically relevant (30 µM) concentrations of carbamazepine and assessed the effect of the drug on activation. Voltage-clamp recordings revealed that, as predicted, carbamazepine depolarized the activation of the S241T mutant channel. In a final set of in vitro experiments, we asked whether carbamazepine could attenuate the hyperexcitability of DRG neurons expressing S241T mutant channels and found that, indeed, exposure to carbamazepine resulted in a nearly 2-fold increase in current threshold together with a reduction in firing frequency.12 Here, we had a hint that it might be possible to predict effective treatment of pain in patients carrying the S241T mutation, on the basis of structural modeling and in vitro voltage-clamp and current-clamp analysis. To test that hypothesis, however, we would require finding patients carrying this mutation and design of an appropriate clinical study. That took us 4 years to do.
Our first task was to design a protocol, which would yield information from a small number of patients. This was a challenge because in the absence of a standardized stimulus to evoke pain, patients with IEM exhibit day-to-day variability; therefore, we opted to include assessment of pain evoked by heat (the trigger of pain in most IEM patients, including those carrying the S241T mutation) in our study design. Next, we had to locate and recruit the 2 patients known to carry this mutation within North America. Imagine receiving a telephone call and being told something like this: “from our genetic studies, we know that you have a mutation that causes your IEM. We have reason to believe, from analysis of your DNA, that your pain may be reduced by an existing medication. We cannot tell you the name of that medication, but we invite you to participate in a research study, in which you will be blinded, and first receive either placebo or the medication for 6 weeks and will then “cross over.” If you agree to participate, you will need to fill out a daily pain diary for 14 weeks. You will need to make 7 full-day trips to Yale Medical School. During each visit, we will warm you up and elicit your IEM pain. And while we do that, your head will be surrounded by a large magnet, and you will not be able to move.”
They both said yes.
The study5 was double blinded and placebo controlled. The results were striking. As predicted by the molecular modeling, thermodynamic analysis, and functional profiling, carbamazepine ameliorated pain in both of these subjects. Interestingly, the change in the pain experience was most marked for temporal aspects of pain. The average duration of pain attacks was significantly reduced in both subjects. Subject 1 reported a reduction of about 50% in mean time in pain per day and in total time in pain over the 15-day maintenance period while taking carbamazepine. Subject 2 reported a reduction of more than 80% in mean time in pain per day and in total time in pain during the maintenance period while taking carbamazepine.5
Notably, the attenuation of pain in this study was paralleled by a shift in brain activity as assessed by functional magnetic resonance imaging, from valuation and pain areas (which are characteristically activated in patients with chronic pain) while untreated or treated with placebo, to primary and secondary somatosensory motor and parietal attention areas while treated with carbamazepine.5 The observation of a rapid (within weeks) change in brain activity was a surprising result, which indicated that at least some changes in the brain in these individuals, with a decades-long history of chronic pain due to hyperexcitability of DRG neurons, were not “locked in” or hard wired.
A major caveat to this study is that the number of patients is small. Nevertheless, these results provide proof-of-principle that personalized pharmacotherapy guided by genomic analysis, molecular modeling, and functional profiling can attenuate pain in at least some selected kindreds carrying specific gene variants.
Since performing this study, we have extended it in 3 ways: First, in a reverse engineering approach, we used structural modeling to study another NaV1.7 mutation (I234T) from a patient known to respond to carbamazepine. This analysis showed that I234T is also located in atomic proximity to the carbamazepine-sensitizing V400M mutation and demonstrated that carbamazepine has an inhibitory action on I234T mutant channels similar to its action on V400M and S241T.11 Here, we had an additional demonstration that molecular modeling and functional assessment in vitro might be useful in predicting clinical pharmacoresponsiveness.
Second, we recently had the opportunity to study a patient with neuropathic pain and diabetic peripheral neuropathy, carrying the NaV1.8-S242T mutation, a mutation in sodium channel NaV1.8, that corresponds to the S241T mutation in NaV1.7.7 Molecular modeling showed us that, despite the substantial difference in primary amino acid sequence of the NaV1.7 and NaV1.8 channels, the S242 residue in NaV1.8 and the S241 residue in NaV1.7 have similar position and orientation within the same part of the channel. Voltage-clamp analysis showed that, like the NaV1.7 mutation, the S242T mutation in NaV1.8 hyperpolarizes channel activation and demonstrated that, similar to its action on the NaV1.7-S241T mutation, carbamazepine depolarizes activation of this mutant NaV1.8 channel and reduces the excitability of DRG neurons expressing these channels. This result extends this pharmacogenomically informed prediction of drug action from NaV1.7 to a second sodium channel subtype, NaV1.8, opening up the possibility that this precision medicine approach may be applicable to sodium channel variants in many channel subtypes beyond NaV1.7.
Most recently, the S241T kindred has provided still another lesson. In the study on the effect of carbamazepine in that kindred, we assessed pain in detail over a 14-week period; the mother displayed a much less severe pain profile (significantly lower number and shorter duration of pain attacks when untreated) compared with the son who carried the same gain-of-function NaV1.7 mutation. The pain profile in subject 1 (son), similar to that for many individuals with IEM, was severe: 424 minutes of pain per day, with a mean pain episode duration of 615 minutes while taking placebo. The pain profile in subject 2 (mother) was mild: 61 minutes in pain per day, and a mean pain episode duration of 91.5 minutes.5 So, the question was “might subject 2 be a carrier of a pain resilience gene?” Because this mother–son pair shared 50% of their genome and carried the same IEM-causative S241T mutation, we studied them as a “pointer-kindred” using whole exome sequencing and an induced pluripotent stem cell (iPSC)-derived sensory neuron (SN) model to begin to search for gene variants that modulate pain. That study by Mis et al.8 revealed hyperexcitability in the iPSC-derived SNs from both individuals, but with much less hyperexcitability in the mother (less pain) than the son. Thus, this analysis showed that, in at least some cases, interindividual differences in pain can be reproduced in a pain-in-a-dish model that uses subject-specific iPSC SNs.
We then went on to assess the genetic/molecular basis for the difference in excitability in SNs in this mother–son pair. That analysis showed that the difference in excitability of the 2 lines of iPSC-derived SNs was due, in large part, to differences in resting membrane potential. Whole exome sequencing, filtered to focus on genes expressed in, and mechanisms that tune the excitability of, peripheral SNs, demonstrated that a variant of one particular gene known to be expressed in DRG neurons—KCNQ2 that encodes the KV7.2 potassium channel—was present in the mother. Dynamic clamp (which allowed us to subtract the current produced by the mutant KV7.2 channels in iPSC SNs from the mother and replace them with precisely titrated injections of wild type KV7.2 current) then allowed us to show that the variant in KV7.2 significantly reduces the excitability of iPSC SNs derived from the pain-resilient mother. In the aggregate, this study showed that gene variants expressed within peripheral neurons can contribute to interindividual differences in the pain experience. And, this study provided proof-of-concept that it is possible to use subject-specific iPSCs and whole exome sequencing to identify specific gene variants that may endow an individual with a degree of pain resilience.
I am occasionally asked, usually by nonscientists, why we surveil an immense population in multiple continents so that we can find and study families with rare disorders. The answer is that genetic “experiments of nature” provided by rare mutations can teach us important lessons. The V400M family provided a spring board for studies on atomic-level modeling and provided a basis for pharmacogenomic studies that suggest that personalized pain pharmacotherapy is not an unrealistic goal. And, the S241T family provided a platform for a study that is tracking down gene variants responsible for pain resilience.
Genetics is beginning to teach us important lessons about pain. We will undoubtedly learn more from pointer-kindreds. We can learn big lessons from rare families.
Conflict of interest statement
The author has served as an advisor or consultant to Biogen, Amgen, Chromocell Research, SiteOne Therapeutics, and GSK. S.G. Waxman reports other from SiteOne Therapeutics, personal fees from Biogen, personal fees from Amgen, personal fees from Chromocell Research, and personal fees from GSK, outside the submitted work.
Research in the author's laboratory has been supported in part by grants from the Rehabilitation Research and Development Service and Biomedical Laboratory Research Service, Department of Veterans Affairs, The Erythromelalgia Association, and The Nancy Taylor Foundation.
. Cummins TR, Dib-Hajj SD, Waxman SG. Electrophysiological properties of mutant NaV
1.7 sodium channels in a painful inherited neuropathy. J Neurosci 2004;24:8232–6.
. Dib-Hajj SD, Rush AM, Cummins TR, Hisama FM, Novella S, Tyrrell L, Marshall L, Waxman SG. Gain-of-function mutation in NaV
1.7 in familial erythromelalgia induces bursting of sensory neurons. Brain 2005;128:1847–54.
. Drenth JP, Waxman SG. Mutations in sodium-channel gene SCN9A cause a spectrum of human genetic pain
disorders. J Clin Invest 2007;117:3603–9.
. Fischer TZ, Gilmore ES, Estacion M, Eastman E, Taylor S, Melanson M, Dib-Hajj SD, Waxman SG. A novel NaV
1.7 mutation producing carbamazepine-responsive erythromelalgia. Ann Neurol 2009;65:733–41.
. Geha P, Yang Y, Estacion M, Schulman BR, Tokuno H, Apkarian AV, Dib-Hajj SD, Waxman SG. Pharmacotherapy for pain
in a family with inherited erythromelalgia
guided by genomic analysis and functional profiling. JAMA Neurol 2016;73:659–67.
. Goldstein JL, Brown MS. The LDL receptor defect in familial hypercholesterolemia. Implications for pathogenesis and therapy. Med Clin North Am 1982;66:335–62.
. Han C, Themistocleous AC, Estacion M, Dib-Hajj FB, Blesneac I, Macala L, Fratter C, Bennett DL, Waxman SG, Dib-Hajj SD. The novel activity of carbamazepine as an activation modulator extends from NaV
1.7 mutations to the NaV
1.8-S242T mutant channel from a patient with painful diabetic neuropathy. Mol Pharmacol 2018;94:1256–69.
. Mis M, Yang Y, Tanaka BS, Gomis-Perez C, Liu S, Dib-Hajj F, Adi T, Garcia-Milian R, Schulman BR, Dib-Hajj SD, Waxman SG. Resilience to pain
: a peripheral component identified using induced pluripotent stem cells and dynamic clamp. J Neurosci 2019;39:382–92.
. Payandeh J, Scheuer T, Zheng N, Catterall WA. The crystal structure of a voltage-gated sodium channel. Nature 2011;475:353–8.
. Sheets PL, Jackson JO II, Waxman SG, Dib-Hajj SD, Cummins TR. A NaV
1.7 channel mutation associated with hereditary erythromelalgia contributes to neuronal hyperexcitability and displays reduced lidocaine sensitivity. J Physiol 2007;581:1019–31.
. Yang Y, Adi T, Effraim PR, Chen L, Dib-Hajj SD, Waxman SG. Reverse pharmacogenomics
: carbamazepine normalizes activation and attenuates thermal hyperexcitability of sensory neurons due to NaV
1.7 mutation I234T. Br J Pharmacol 2018;175:2261–71.
. Yang Y, Dib-Hajj SD, Zhang J, Zhang Y, Tyrrell L, Estacion M, Waxman SG. Structural modelling and mutant cycle analysis predict pharmacoresponsiveness of a Na(V)1.7 mutant channel. Nat Commun 2012;3:1186.