A Drosophila Model of Essential Tremor

Essential Tremor (ET) is one of the most common neurological diseases, with an estimated 7 million affected individuals in the US; the pathophysiology of the disorder is poorly understood. Recently, we identified a mutation (KCNS2 (Kv9.2), c.1137 T > A, p.(D379E) in an electrically silent voltage-gated K+ channel α-subunit, Kv9.2, in a family with ET, that modulates the activity of Kv2 channels. We have produced transgenic Drosophila lines that express either the human wild type Kv9.2 (hKv9.2) or the ET causing mutant Kv9.2 (hKv9.2-D379E) subunit in all neurons. We show that the hKv9.2 subunit modulates activity of endogenous Drosophila K+ channel Shab. The mutant hKv9.2-D379E subunit showed significantly higher levels of Shab inactivation and a higher frequency of spontaneous firing rate consistent with neuronal hyperexcitibility. We also observed behavioral manifestations of nervous system dysfunction including effects on night time activity and sleep. This functional data further supports the pathogenicity of the KCNS2 (p.D379E) mutation, consistent with our prior observations including co-segregation with ET in a family, a likely pathogenic change in the channel pore domain and absence from population databases. The Drosophila hKv9.2 transgenic model recapitulates several features of ET and may be employed to advance our understanding of ET disease pathogenesis.

Scientific REPORTS | (2018) 8:7664 | DOI: 10.1038/s41598-018-25949-w Shab I k dramatically increases NMJ transmission (up to 10-fold gain) during repetitive nerve stimulation 4 . The kinetics of the Shab channel are reported to be comparable to the classical (as described by Hodgkin-Huxley) delayed rectifier K + channel 6 .
Abnormal motor behavior and leg shaking (with or without anesthesia) has been described in Shab mutant files 4 . Many neurological disorders are associated with altered activity, function or expression of K + channels 7 and mutations in these channels can cause cerebellar dysfunction and ataxia. Notably, a tremor phenotype has been described in patients with mutations in KCNA1 (Kv1.1; EA; OMIM 160120) 8 and missense dominant negative mutations in KCNC3 (Kv3.3) are associated with hyperexcitability, cerebellar neurodegeneration and subsequent movement defects including spinocerebellar ataxia (SCA13; OMIM 605259) 9 . Kv3.1 and Kv3.3 mutant mice also display severe motor deficits, including tremor, myoclonus, and ataxic gait and behavioral alterations that include constitutive hyperactivity and sleep loss [10][11][12] . In addition to inherited channelopathies, several neurodegenerative disorders including Alzheimer's disease, Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, and SCAs exhibit altered properties of diverse K + channels characterized by protein aggregate induced hyperexcitability (Kumar et al. 13 ).
Currently, it is unclear how the mutation that we identified in Kv9.2 (p.D379E) could lead to an ET phenotype. However, the dominant inheritance pattern suggests a gain-of-function or dominant negative mechanism. To study the disease mechanism we have produced transgenic Drosophila lines that express either the wild type human Kv9.2 (hKv9.2) subunit or the ET causing mutant human Kv9.2 (hKv9.2-D379E) subunit in all neurons. Here, we show that the hKv9.2 subunit can modulate the channel activity of endogenous Drosophila Shab (Kv2), describe the behavioral manifestations of nervous system dysfunction, and the effects of hKv9.2 and hKv9.2-D379E on adult neuron activity.

Results
To study the disease mechanism of the Kv9.2 (p.D379E) mutation that we identified in an ET family and advance our understanding of disease pathogenesis, we have produced transgenic Drosophila lines that express the wild type hKv9.2 or ET causing mutant hKv9.2-D379E subunit.
Pan-neural expression of the wild type hKv9.2 or mutant hKv9.2-D379E subunit resulted in no apparent gross morphological differences. Western blot analysis was used to verify transgenic expression of hKv9.2 and hKv9.2-D379E in Drosophila. Using a hKv9.2 immunogen corresponding to amino acids 175-470, we detected a 54.2 kDa band of the correct size corresponding to human hKv9.2 (Fig. S1).

Behavioral Manifestations of Nervous System Dysfunction.
To test the hypothesis that mutations in Kv9.2 cause nervous system dysfunction, we tested the effect of pan-neural expression of hKv9.2-D379E compared to wild type hKv9.2 or controls on climbing response throughout the fly lifespan 14 (Fig. 1). Flies expressing either wild type or mutant human channels displayed significantly faster climbing (p < 0.0003) throughout lifespan (Fig. 1A), consistent with the hKv9.2 channel expressing animals being behaviorally hyperexcitable as previously described for mutants of other Kv channels in locomotor assays 15,16 . Moreover, flies which only expressed the wild type or mutant hKv9.2 channel in adult neurons after development also displayed significantly faster climbing throughout lifespan (p = 0.04-0.0001; weeks [3][4][5] suggesting that the phenotype is not due solely to changes during development (Fig. 1B).
Because ET is associated with increased mortality 17 and Kv channel dysfunction in both humans and flies is associated with disease and decreased lifespan (Kumar et al. 13 ) we performed survival assays. Significant differences in lifespan were not detected between flies expressing hKv9.2 or hKv9.2-D379E channels during development compared to wild type ( Fig. 2A). Flies expressing the wild type or mutant hkv9.2 channel post-developmentally in neurons however did display significant differences in lifespan compared to wild type (Elav-gal4/+ Gal80 ts and Elav-gal4 > Gal80 ts hKv9.2 (p = 0.0012) or Elav-gal4/+ Gal80 ts and Elav-gal4 > Gal80 ts hKv9.2-D379E. (p = 0.0001)). (Fig. 2B). The reduced lifespan observed for flies expressing the hKv9.2 or mutant hKv9.2 channel post-developmentally is consistent with a reduced lifespan observed for other hyperexcitable Kv channel mutants 13 .
Wing posture and motility deficits were also assessed. Normally, flies hold their wings flat and rarely display an elevated or downturned wing posture. However, we observed that approximately 40% of flies expressing hKv9.2 or hKv9.2-D379E pan-neuronally displayed an abnormal wing posture, with bilateral wing elevation, with onset 7-21 days post eclosion (Supplementary Movie 1). This abnormality may be due to muscle and/or neural based effects and a downturned wing phenotype has been described in other Drosophila models of neurodegeneration including Huntington's disease 18 and Parkinson's disease 19,20 and bilateral wing elevation has been described for flies expressing the ion channel gene, TRPM8, in adult neurons 21 . Alternatively this effect could also be due to a developmental effect or muscle based hyperexcitability as described in the Drosophila ether-à-go-go (eag) Shaker double mutant 22 . To test whether the abnormal wing posture observed in flies expressing hKv9.2 or hKv9.2-D379E pan-neuronally was due to a developmental effect we also assessed flies expressing the wild type or mutant hKv9.2 channel only in adult neurons. We observed that approximately 13% of flies expressing hKv9.2-D379E in adult flies displayed an abnormal wing posture, with bilateral wing elevation, with onset 6 weeks post eclosion (p < 0.0001) (Fig. S2). Abnormal wing posture and motility defects were also observed in flies expressing the wild type hKv9.2 channel in post-developmental neurons (~20% also with onset 6 weeks after eclosion (p < 0.0001) (Fig. S2)).
Hyperexcitable mutants such as Shaker, Shab, Shaw, eag and Hyperkinetic display abnormal leg shaking and wing scissoring in etherised adults [23][24][25] . Adult flies expressing hKv9.2 or hKv9.2-D379E pan-neuronally also exhibited leg shaking, abdominal pulsations and body shuddering under ether anesthetization (Supplementary Movie 2). The ether anesthetization induced tremor in these flies is consistent with reports in the literature of excitatory effects of commonly used anesthestics in humans that may manifest as spontaneous movements including tremor, dystonia and myoclonus 26 . Interestingly, a number of observations suggest a link between anesthesia and ET, including an ET kindred with malignant hyperthermia 27 29,30 . Using a Shab specific toxin we were able to isolate the Shab current (a voltage-sensitive non-inactivating K + current) in these neurons consistent with previous reports 6 . Expression of either hKv9.2 and hKv9.2-D379E in these neurons caused an alteration in the kinetics of Shab (Kv2) (Fig. 3A). The current evoked from Shab when depolarized from −133 mV to −3 mV showed that the non-inactivating Shab becomes inactivating in the presence of the wild type hKv9.2 subunit and to a greater extent with the mutant hKv9.2-D379E subunit (Fig. 3A). The I-V relationships for Shab in the three given genotypes was also determined (Fig. 3C). The peak current shows that expression of the hKv9.2 subunits cause a shift in activation of Shab to more negative voltages. The sustained current shows that, at similar voltages, flies expressing either hKv9.2 subunit show reduced Shab current after 200 ms of depolarization. When the sustained current after 200 ms of depolarization is expressed as a percentage of the peak current the flies with only native Shab (i.e. controls) show high percentages, indicating very low levels of inactivation (Fig. 3C). Flies expressing either the wild type hKv9.2 subunit (p < 0.001) or the mutant hKv9.2-D379E subunit (p < 0.001) were significantly different to flies with only endogenous Shab alone (control). Comparing the relative sustained currents (Fig. 3C), the mutant hKv9.2-D379E subunit shows significantly higher levels of inactivation than the wild type hKv9.2 subunit (n = 3, p = 0.0438). The observed change in the inactivation kinetics of Shab in the presence of mutant hKv9.2-D379E subunit would be predicted to result in neuronal hyperexcitability. To test whether the hKv9.2-D379E subunit does indeed cause neuronal hyperexcitability, we performed current clamp recordings of clock neurons. We found that expression of hKv9.2-D379E subunit caused a significant increase in firing frequency (a measure of hyperexcitability) compared to expression of the hKv9.2 subunit (Fig. 4). The spontaneous action potential firing rate in flies expressing either hKv9.2 subunit was significantly different with the mutant subunit having a higher frequency (p = 0.0102) (Fig. 4A,B). In order to further investigate the complex effects of expression of hKv9.2 and hKv9.2-D379E on Shab currents in Drosophila neurons, we developed a biophysical model that incorporated the major Kv currents separately. Major Kv current kinetics were fitted to the voltage-clamp data so that an accurate depiction of the specific effects of expression of hKv9.2 and hKv9.2-D379E subunits on the excitability of the whole-cell model could be determined. In particular, fitting the voltage clamp data (see Fig. 5) to classical Hodgkin-Huxley ionic current equations generated parameters consistent with the behaviors of flies with only native Shab or Shab perturbed by expression of hKv9.2 or hKv9.2-D379E on Shab (Fig. 5A,B). As illustrated in Fig. 6 the parameters obtained from the computational modeling are in concordance with experimental data. The hKv9.2 subunit activates at more negative values (activation Vh) than the native Shab channel as observed in the experimental data. The extent of inactivation is also higher in hKv9.2-D379E in the model than in the wild type hKv9.2 flies as evidenced by the lower inactivation K. The speed of inactivation is also higher in the mutant hKv9.2-D379E as shown by the higher inactivation σ. Importantly, the spontaneous firing rate of action potentials obtained in whole-cell model simulations (i.e. after combining the individual ionic currents in a biophysical model of the electrical activity of the cell), in neurons expressing either the wild type or mutant hKv9.2 subunit agrees very well with the experimental (current clamp) data (Fig. 6).
Expression of either hKv9.2 and hKv9.2-D379E causes a significant increase in night-time activity and reduction in sleep. Sleep dysfunction, including short duration of sleep, has been reported in patients with ET [31][32][33] . To determine whether the neuronal hyperexcitability observed in hKv9.2 and hKv9.2-D379E in Drosophila effects circadian locomotor rhythms and sleep we characterized their circadian behavior.

Discussion
We have created a model of ET by expressing the human Kv9.2 channel subunit in Drosophila. We show that the hKv9.2 subunit can modulate the endogenous Shab (Kv2 subfamily) channel activity. Behavioral manifestations of nervous system dysfunction consistent with a hyperexcitable phenotype were observed in flies expressing the hKv9.2 or hKv9.2-D379E subunit during development and in adult neurons only. Studying the effects of hKv9.2 and hKv9.2-D379E on CNS neuronal activity, we showed that the mutant hKv9.2-D379E subunit showed significantly higher levels of inactivation than the wild-type hKv9.2 subunit and a significantly higher frequency of spontaneous firing rate of action potentials consistent with neuronal hyperexcitibility. A biophysical model of the electrical activity of the cell, combining the individual ionic currents in flies expressing hKv9.2 subunits, was in agreement with experimental data. Characterization of circadian behavior in flies expressing hKv9.2 and hKv9.2-D379E channels showed that while rhythmicity was unaffected significant differences were observed for night time activity and night time sleep. This is consistent with previous studies where hyperexcitation of lateral  While effects on period circadian protein have been seen 36 , no differences were seen in the current study. While we recognise that there are limitations to the Drosophila model that we have developed because of the lack of a clear Drosophila ortholog for Kv9.2, our data is consistent with previous studies that have demonstrated that Kv9.2 modulates the activity of Kv2 channels 2 (such as Drosophila Shab), as well as observing significant differences in the phenotype (inactivation and spontaneous firing rate) between flies expressing the wild type and mutant hKv9.2 channels.
The pathogenicity of the KCNS2 mutation (p.D379E) that we identified in an early-onset ET family 2 , that is the focus of the current study, is further supported by functional data from the Drosophila model and is consistent with previous observations reported by us including co-segregation with ET in a family, prediction as a pathogenic amino acid substitution by several variant prediction tools, absence from population databases, and its location in the pore domain of the Kv9.2 channel.
Our data is consistent with recent reports and observations that ET may represent a family of disorders of neurological channelopathies, with mutations identified in a voltage-gated K + channel α subunit (the focus of this study) in a family with pure ET 2 , in voltage-gated sodium channel α subunits in a family with epilepsy and ET (SCN4A) 37 and a family with familial episodic pain and ET (SCN11A) 38 . Further, the T-type calcium channel, Ca v 3, has been implicated in neuronal autorhythmicity 39,40 and is thought to underlie tremors seen in Parkinson's disease 41 , enhanced physiological tremor, and in ET 42 and T-type calcium channel antagonists have been shown to reduce tremor in mouse models of ET 43,44 . Nonetheless, the complete genetic basis for ET remains incomplete. Given the clinical and genetic heterogeneity observed in ET, further evaluation of ion channels as candidate genes for ET is warranted.

Methods
Transgenic Drosophila. Human Kv9.2 sequenced-verified cDNA was obtained from the Mammalian Gene Collection (Clone ID: 5199736)(GE Dharmacon, Lafayette, Co). The RNA source for the cDNA was from an anonymous pool of 6 male brains, age range 23-27 years old. The library was oligo-dT primed and directionally cloned into pCMV-SPORT6 vector. Human Kv9.2 was transferred and cloned from the pCMV-SPORT6 vector by TOPO ® cloning (Thermofisher scientific, Waltham, MA) into the pBID-UAS Drosophila vector 45 .
Site directed mutagenesis was used to generate mutant human Kv9.2, c.1137 T > A, (p.D379E) using the Quick-change II site directed mutagenesis kit (Agilent Technologies, Santa Clara, CA) from the human Kv9.2 sequenced-verified cDNA. Germ-line transformants were generated with PhiC31 integrase with Chromosome II attP40 landing site. Drosophila were maintained with standard conditions and food. Uas-hKv9.2 and uas-hKv9.2-D379E were crossed to Elav(c155)-Gal4 stock (Bloomington stock number-458) for pan-neural expression and pdf-Gal4, pdf-rfp for electrophysiological characterization 29,30 . Post-developmental effects utilized GAL80 TS ro, restricting expression to adult neurons 46-48 . For electrophysiological and circadian analysis, expression was restricted to the LNV clock neurons by use of pdf-Gal4 or throughout the clock system through tim-Gal4. Negative Geotaxis Climbing Assay. The loss of climbing response was used to monitor ageing-related locomotor changes in Drosophila 14,49 . The climbing assay was performed as previously described 14,49 . We assessed 20 flies per vial for each transgenic and control line. Five trials were conducted for each vial. The average climbing rate was determined by measuring the time for the first fly to climb 17.5 cm. Climbing response was assessed at the following time points: Day 7,14,21,28,35,42,49,and 56.
Expression of hKv9.2 and hKv9.2-D379E channels in adult neurons after development and effects on lifespan. The climbing assay was performed as described for transgenic and control lines except 10 flies per vial were assessed for each line.
Lifespan Assay. Lifespan assays were performed as described previously 50  Electrophysiology. Experiments were conducted on wild type control: pdf-Gal4; pdf-rfp, as well as the two experimental lines: pdf-Gal4 > uas-hKv9.2; pdf-rfp and pdf-Gal4/uas-hKv9.2-D379E; pdf-rfp. Flies were housed in 12 h light: 12 h dark cycles to facilitate recordings throughout the day. All flies were raised at 25 °C (incubators) at humidity of 60%.
Whole fly brains were dissected from CO 2 anaesthetized adult flies aged between 0-5 days post-eclosion just before patch clamping using a previously established protocols 29,30 . Dissections were conducted in standard Drosophila external solution. The brain was then transferred to a recording chamber and held in place using a brain harp.
Whole-cell and patch-clamp recordings were made from the red fluorescent protein (RFP)-tagged pigment dispersing factor (PDF)-positive large Lateral Neuron ventral (l-LNv) clock neurons using a Multiclamp 700B amplifier and Axon Digidata 1440A digitizer in a recording chamber filled with Drosophila external solution as described previously 29,30 . Glass pipettes (8-15 MΩ) were pulled using a Sutter P-1000 puller and filled with internal solution. The resulting signal was then monitored and recorded using pClamp10 Clampex software and pipette offsets were zeroed prior to cell contact with the pipette capacitance being compensated upon contact. Subsequent signals were passed through a 10 kHz low-pass Bessel filter and sampled at 20 kHz. The cell-attached configuration was established by gentle suction applied through the pipette holder, and subsequent whole-cell configurations utilized a stronger pulse of negative pressure to break into the cell through the membrane.
The standard Drosophila external solution consisted of (in mM) 101 NaCl, 1 CaCl 2 , 4 MgCl 2 , 3 KCl, 5 glucose, 1.25 NaH 2 PO 4 , and 20.7 NaHCO 3 with pH 7.2 and osmolality 250 mOsm. The internal solution consisted of (in mM) 102 K-gluconate, 0.085 CaCl 2 , 1.7 MgCl 2 , 17 NaCl, 0.94 EGTA, and 8.5 HEPES with pH 7.2 and osmolality 235 mOsm. The pH was increased with NaOH for the external solution and KOH for the internal solution and decreased with HCl for both. Stock solutions of guangxitoxin-1E (GxTX, Alomone labs) were made up using external solution. The resulting junction potential for these solutions has been calculated as being −13 mV (data shown has been adjusted to account for this). During the recordings, the drug solutions are added by pipette to 1 ml of external solution in the recording chamber to achieve the final drug concentration.
Voltage-clamp recordings were initially held at a membrane potential of −93 mV. For I-V relationships, a standardized protocol was used consisting of a hyperpolarizing step of −40 mV to −133 mV for 500 ms and then steps in increments of 10 mV from −93 mV to −3 mV for a duration of 200 ms before a return to the holding potential of −80 mV. To measure Shab currents, voltage-clamp recordings were performed before and after application of 10 nM GxTX 51 , the difference between the two conditions gives the subtractive Shab current. Wash-out of the drug recovered the channel current (88.64% ± 3.06%). Computational modeling. Computational models of the Shab channel ionic current, with addition of hKv9.2 or hKv9.2-D379E channels were generated by a nonlinear optimization algorithm fitting the electrophysiological data to Hodgkin-Huxley equations 53 of the form:

Drosophila
where the current, I, depends on the maximal conductance (gmax), the reversal potential of the channel (E), the membrane voltage (Vt), and the activation and inactivation ion channel gating variables (m and h). The gating variables are given by the equations: The individual ionic current models fitted to our data were then combined in a model of the electrical behavior of the whole cell. The resulting whole-cell computational model builds upon a previous model of suprachiasmatic nucleus (SCN) clock neurons 54 which incorporates Na + , K + , Ca 2+ , and leak currents. However, in our model the original composite K + current is separated into the four currents mediated via the major voltage-gated K + channels (Shaker Kv1, Shab Kv2, Shaw Kv3 and Shal Kv4) found in the l-LNvs based on electrophysiological recordings; while Kv2 refers to either the native Shab current, native Shab with human hKv9.2, or native Shab with hKv9.2-D379E. The current balance equation then is: where E Na = 52 mV, E Ca = 132 mV, E K = −90 mV, and E leak = −7 mV are the reversal potentials. These were calculated using the Nernst and Goldman-Hodgkin-Katz voltage equations based on the internal and external solutions used.
Statistical Analysis. Statistical analysis was carried out using the Prism 7.0 (GraphPad Software, Inc., La Jolla, CA) software. Analysis of climbing response during lifespan was performed using unpaired student's t-test (one unpaired t-test per row) in Prism 7.0. Data were represented as the mean and standard error of the mean (S.E.M.) from at least ten independent experiments at each week.