Fly stocks:

The bss¹ mutation is an EMS-induced allele of bang senseless once thought to be an allele of bas (bang sensitive) and designated bas^MW1 (Jan and Jan 1978). Allelic relationships were clarified by Ganetzky and Wu (1982) with bas^MW1 renamed bss^MW1; its current designation is bss¹. A second allele, bss², is an EMS-induced mutation also known as bss^PC75 (Ganetzky and Wu 1982). The eas gene is located at map position 1-53.5 and encodes an ethanolamine kinase (Pavlidis et al. 1994). The Canton-S strain was used as a wild-type control, unless indicated. Transgenic UAS-Na_V1-1 and UAS-Na_V1-1^L1699F lines were generated by BestGene (Chino Hills, CA). UAS-paraRNAi lines were obtained from the Vienna Drosophila RNAi Center. All other lines, including Gal4 driver, P[EP] insertions, and duplication and deletion lines were obtained from the Bloomington Drosophila Stock Center.

Behavior and electrophysiology:

Behavioral testing for bang-sensitive (BS) paralysis was performed as described previously (Kuebler and Tanouye 2000). Flies were collected <1 day post eclosion and tested for BS paralysis and recovery time 24 hr after collection to mitigate variability due to age. Pools of flies are combined (in total, n = ∼100 for each genotype). For genotypes that display only partial penetrance of BS paralysis, only those flies that displayed paralysis were used for recovery time analysis. In vivo recording of seizure-like activity and seizure threshold determination in adult flies was performed as described previously (Kuebler and Tanouye 2000). Flies 2–3 days post eclosion were mounted in wax on a glass slide, leaving the dorsal head, thorax, and abdomen exposed. Stimulating, recording, and ground metal electrodes were made of uninsulated tungsten. Seizure-like activity was evoked by high-frequency electrical brain stimulation (0.5-msec pulses at 200 Hz for 300 msec) and monitored by dorsal longitudinal muscle (DLM) recording. During the course of each experiment, the giant fiber (GF) circuit was monitored continuously as a proxy for holobrain function. For each genotype tested, n > 20.

Electrophysiology of heterologously expressed para Na⁺ channels was as described previously (Tan et al. 2002; Olson et al. 2008). A DmNa_v1-1^L1699F cDNA subcloned into the pGH19 expression vector was linearized with NotI and transcribed with T7 polymerase (mmESSAGE mmACHINE kit, Ambion). cRNA (1 ng) was co-injected with tipE cRNA (1 ng), which encodes a unique transmembrane protein known to be required for robust Na⁺ channel expression (Warmke et al. 1997). Oocytes obtained surgically from female Xenopus laevis were incubated with collagenase (type IA, 1 mg/ml; Sigma) in Ca⁺⁺-free ND96 medium (96 mm NaCl, 2 mm KCl, 1 mm MgCl₂, 5 mm HEPES at pH 7.5). Oocytes were then incubated in normal ND96 medium containing 1.8 mm CaCl₂, 5 mm pyruvate, 0.5 mm theophylline, and 50 mg/ml gentamicin, and Na⁺ currents were recorded by a two-microelectrode voltage clamp. Borosilicate glass electrodes were filled with filtered 3 m KCl in 0.5% agarose and had resistance <1.0 MΩ. Currents were measured using an oocyte clamp (OC725C, Warner Instruments), Digidata interface (1200A, Axon Instruments), and pCLAMP9 software (Axon Instruments). All experiments were performed at room temperature (20°–22°). Capacitative transients and linear leak current were corrected by P/N subtraction or subtraction in tetrodotoxin (20 nM).

The voltage dependence of sodium channel conductance (G) was calculated by measuring the peak current at test potentials ranging from −80 to + 65 mV in 5-mV increments and divided by (V − V_rev), where V is the test potential and V_rev is the reversal potential for sodium ion. Peak conductance values were normalized to the maximum peak conductance (G_max) and fitted with a two-state Boltzmann equation of the form G/G_max = [1 + exp(V − V_1/2)/k]⁻¹, in which V is the potential of the voltage pulse, V_1/2 is the voltage for half-maximal activation, and k is the slope factor.

The voltage dependence of sodium channel fast inactivation was determined by using 100-msec inactivating prepulses ranging from −120 to 25 mV in 5-mV increments from a holding potential of −120 mV, followed by test pulses to −10 mV for 20 msec. The peak current amplitude during the test depolarization was normalized to the maximum current amplitude and plotted as a function of the prepulse potential. Data were fitted with a two-state Boltzmann equation of the form I/I_max = [1 + exp(V − V_1/2)/k]⁻¹, in which I is the peak sodium current, I_max is the maximal current evoked, V is the potential of the voltage prepulse, V_1/2 is the half-maximal voltage for inactivation, and k is the slope factor.

To determine the steady-state slow inactivation, oocytes were held at prepulse potentials ranging from −100 to −10 mV in 10-mV increments for 60 sec. A 100-msec recovery pulse to −120 mV and a 20-msec test pulse to −10 mV were given before returning to the holding potential of −120 mV. The peak current amplitude during the test depolarization was normalized to the maximum current amplitude and plotted as a function of the prepulse potential. The data were fitted with a two-state Boltzmann equation of the form I/I_max = [1 + exp(V − V_1/2)/k]⁻¹, in which I_max is the maximal current evoked, V is the potential of the voltage pulse, V_1/2 is the half-maximal voltage for inactivation, and k is the slope factor.

To determine recovery from fast inactivation, sodium channels were inactivated by a 100-msec depolarizing pulse to −10 mV and then repolarized to −120 mV for an interval of variable durations followed by a 20-msec test pulse to −10 mV. The peak current during the test pulse was divided by the peak current during the inactivating pulse and plotted as a function of duration time between the two pulses.

Recovery from slow inactivation was measured using a 60-sec depolarization to −10 mV followed by variable intervals at −120 mV and a subsequent 10-msec test pulse to −10 mV. The available current recorded with the test pulse was normalized to peak current recorded immediately prior to the recovery protocol.

Molecular biology:

The Na_V1-1 adult isoform of para encodes a functional Na⁺ channel when expressed heterologously in Xenopus channels (Warmke et al. 1997). The L1699F missense mutation was introduced into pGH19-Na_V1-1 (corresponding to Na_V1-1 L1676F, a C → T point change at 5026 bp) using primers L1699F-F (CCATCTTAGGTCTTGTATTTAGCGATATTATCGAGAAGTACTTCG) and L1699F-R (CGAAGTACTTCTCGATAATATCGCTAAATACAAGACCTAAGATGG) with the Quikchange XL II kit (Stratagene). The entire Na_V1-1^L1699F sequence was determined to confirm the point change and ensure that no additional PCR-induced mutations were introduced. Na_V1-1 and Na_V1-1^L1699F cDNAs were excised from the pGH19 vector using KpnI and NheI and then subcloned into pUAST cut with KpnI and XbaI.

Data analysis:

Chi-square tests were used to compare the penetrance of seizures. Student's t-test and ANOVA were used to compare recovery times and seizure thresholds across genotypes, as appropriate. For ANOVA analysis, where the null hypothesis was rejected by the overall F ratio, multiple comparisons of data sets were performed by Fisher's least significant difference with t-test significance set at P < 0.05. For Figures 2, 4 and 5, error bars represent standard error of the mean, and statistical significance is indicated by *P < 0.05, **P < 0.01, and ***P < 0.0001.

bss¹ behavior—seizure sensitivity and tonic-clonic-like activity:

The behavior of bss¹ mutants is similar to other mutants of the BS paralytic class such as eas^PC80, sda^iso7.8, and tko^25t mutants (Royden et al. 1987; Pavlidis et al. 1994; Zhang et al. 2002). Under normal conditions, bss¹ flies display unexceptional feeding, grooming, and mating behaviors and show wild-type phototaxis and geotaxis responses. Overall activity levels are unaltered, and flies are neither sluggish nor hyperactive. Abnormal behavior is induced in bss¹ mutants by mechanical shock (a “bang”). The resulting behavioral phenotype is complex (Figure 1A) with six distinguishable phases: seizure; initial paralysis; tonic-clonic-like activity; recovery seizure; refractory recovery during which behaviorally normal flies cannot be induced to have further seizures; and, finally, complete recovery.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.—

Drosophila bss¹ mutant behavior and electrophysiology seizure-like phenotypes. (A) Cartoon depicting stereotypic behavioral phenotype of bss¹ flies subjected to a mechanical shock (10-sec vortex: BANG): initial seizure-like behavior, followed by complete paralysis and then a tonic-clonic-like period that is unique to bss¹ and not evident in other BS mutant genotypes. One clonus-like event is depicted, but the number can vary, as can the duration of the period. The tonic-clonic-like period is followed by a recovery seizure, and the fly then recovers. Not depicted is a quiescent period of variable duration often observed between the recovery seizure and recovery, as well as the refractory period during which flies are resistant to further seizures that occurs immediately following recovery. (B) For bss¹/Y hemizygous males, recovery time from behavioral paralysis is substantially longer than for bss¹/+ heterozygous females or for another BS mutant, eas^PC80/Y. (C) Electrical recording from a DLM fiber showing seizure-like neuronal activity and synaptic failure in a bss¹/Y animal subjected to a 4-V high-frequency electrical stimulus (HFS), as well as single-pulse stimuli to trigger the GF circuit, allowing assessment of synaptic function (giant fiber response, GFR). Following an initial seizure (IS) is a period of synaptic failure within which GF stimulation fails to evoke a DLM response (SF^1–5), unlike that seen prior to the seizure. As depicted, during the period of synaptic failure, spontaneous secondary seizures are observed (SS^1–4). Although in this trace four secondary seizures are observed, the number is variable. A final recovery seizure (RS) is observed, and shortly thereafter, GF system transmission is restored (response recovery, RR). Vertical calibration bar: 40 mV in all panels. Horizontal calibration bar: 10 sec in (C) and (D), 1.5 sec in (E), and 10 msec in (F). (D) For eas^PC80/Y, HFS stimulation causes an initial seizure, synaptic failure, and a recovery seizure; however, no secondary seizures are observed. (E) At a higher sweep speed, depicted are initial seizure, secondary seizures, and recovery seizure from the recording in B. (F) At a higher sweep speed, depicted are GF response (GFR), synaptic failures (SF), and response recovery (RR) from the recording in B.

The initial seizure for bss¹ is similar to other BS mutants, lasting several seconds and characterized by leg shaking, abdominal muscle contractions, wing flapping and scissoring, and proboscis extensions. Initial paralysis is also typical of BS mutants: bss¹ flies are immobile and unresponsive to mechanical stimulus, as described previously (Ganetzky and Wu 1982). Unlike other BS mutants, initial paralysis in bss¹ homozygotes is followed by an extended period of tonic-clonic-like activity (Figure 1A). During this period, the fly is mainly quiescent, resembling a tonic phase. The quiescence is broken up by multiple bouts of clonus-like activity. As in other BS mutant flies, bss¹ shows a recovery seizure, then a refractory period, followed by complete recovery. bss¹ recovery time is longer than for other BS mutants; mean recovery time for bss¹ is ∼240 sec, compared with sda or eas at 38 and 81 sec, respectively (Figure 1B and Figure 2B). Interestingly, analysis of the larval neuromuscular junction in bss¹ animals indicates that motoneurons are hyperexcitable, displaying an abnormal long-term facilitation after repeated stimulation resulting in multiple action potentials and a large, prolonged excitatory junction potential (EJP), where, in contrast, wild-type animals show only a single-action potential and a single small EJP (Jan and Jan, 1978; Ganetzky and Wu 1982).

The bss¹ allele is a severely seizure-sensitive mutation and a strong semidominant that causes the most extreme phenotype of the BS mutants. In comparison, eas and tko mutants behave as recessives, while sda mutants are weakly semidominant. BS seizure-like behaviors and paralysis were observed in heterozygous bss¹/+ flies at high penetrance (95%), similar to previous reports (Ganetzky and Wu 1982). Behavioral phenotypes seen in bss¹ homozygotes and heterozygotes are generally similar, the main difference being that heterozygotes lack tonic-clonic-like activity, thereby more closely resembling eas or sda homozygous mutants in behavior. There is also a large reduction in recovery time from ∼240 sec in the homozygote to ∼50 sec in the heterozygote (Figure 1B and Figure 2B). Behavioral phenotypes of heterozygotes are sensitive to genetic background. For example, although ordinarily highly penetrant, there can be considerable variation in the percentage of paralyzed heterozygous flies (50–95%), depending on the genetic background. Genetic background can also influence recovery time, as discussed below.

Extreme seizure sensitivity in bss¹ and the electrophysiology of tonic-clonic-like spontaneous firing:

The electrophysiology phenotype of bss¹ is generally similar to that seen in other BS mutants, although more extreme in some aspects. As was the case for behavior, electrophysiology is normally unexceptional in bss¹ mutants as determined by using the adult giant fiber system neural circuit as proxy (Tanouye and Wyman 1980). Thus, for example, single-pulse stimulation of the giant fiber produced evoked muscle potential responses in the dorsal longitudinal muscles that were normal in appearance and had a threshold of 2.04 ± 0.48 V and a latency of 2.07 msec. Also normal was the ability to follow high-frequency stimulation and respond to short interpulse twin-pulse intervals.

Seizure-like electrical activity in bss¹ mutants can be evoked with high-frequency stimuli (HFS; 0.5-msec stimuli at 200 Hz for 300 msec) at exceptionally low stimulation voltages (Pavlidis and Tanouye 1995; Kuebler and Tanouye 2000; Lee and Wu 2002). Seizure-like activity following HFS consisted of aberrant high-frequency firing (>100 Hz) lasting for ∼1 sec, which was present in the dorsal longitudinal muscle and all other muscle fibers and motoneurons examined (Figure 1, C and E). This bss¹ seizure-like phenotype is qualitatively similar to that seen for other BS genotypes and normal flies, differing only in the HFS voltage required for eliciting it (i.e., the seizure threshold), but otherwise indistinguishable in subsequent electrical activity (Figure 1D). For bss¹ males, the seizure threshold is 4.4 V HFS. This is substantially lower than for wild-type males (30.1 V HFS) and slightly lower than for eas and sda BS mutant males (5.13 and 6.2 V HFS, respectively). Thus, using the criterion of HFS threshold, bss¹ mutants are about seven times more seizure-sensitive than wild-type flies.

After an initial seizure, the next aspect of the electrophysiology phenotype for bss¹ mutants is the sudden failure of the giant fiber system neural circuit to drive muscle potentials in the dorsal longitudinal muscle. This is termed a “synaptic failure” period as it has been shown to be due to transmission failure at many central synapses (Figure 1F; Pavlidis and Tanouye 1995); it is likely the cause of behavioral paralysis in bss¹ mutants and in other BS genotypes. However, whereas the period of synaptic failure is short in most BS mutants, for example, ∼38 sec in sda, for bss¹, it is longer and more complex, apparently reflecting the complexity observed behaviorally as tonic-clonic-like activity (Figure 1, A and C).

For the purpose of describing the electrophysiology of the bss¹ tonic-clonic-like period, it is convenient to consider three seizure types: (1) initial seizure, occurring immediately after HFS; (2) secondary seizures, occurring within the tonic-clonic-like interval; and (3) recovery seizure, occurring immediately prior to long-term recovery of the giant fiber system neural-circuit response (Figure 1, C and E). Initial and secondary seizures resemble each other in waveform and duration; recovery seizures are somewhat different in waveform and longer in duration, although the significance of this is unclear. In young bss¹ flies (1 day post eclosion), the tonic-clonic-like period is fairly short, ∼45 sec, and contains only one secondary seizure. In older bss¹ flies (7 days old), the tonic-clonic-like period is longer, ∼75 sec, and typically contains three to four secondary seizures, although up to eight have been observed. This is consistent with observations of longer behavioral recovery times in older bss¹ flies.

Continuous stimulation of the giant fiber system neural circuit during the tonic-clonic-like period shows that, during the tonic intervals, there is no transmission through the circuit to the dorsal longitudinal muscle fiber (Figure 1F). From this, we infer that the lack of transmission is likely due to synaptic failure at many central synapses as proposed for BS paralysis (Pavlidis and Tanouye 1995). Interestingly, approximately coincident with each secondary seizure, transmission through the giant fiber system neural circuit is transiently restored. That is, if a single-pulse stimulation is delivered to the giant fiber system within the period of the secondary-seizure burst, the stimulation is capable of evoking a muscle potential in the dorsal longitudinal muscle. Surprisingly, if a single-pulse stimulation is delivered to the giant fiber system anywhere outside of a secondary seizure burst (i.e., within the apparent tonic interval), it is ineffective at evoking a muscle potential (compare Figure 1, C and F). From these observations, we propose that tonic-clonic-like periods following initial seizure arise from periods of synaptic failure responsible for tonic intervals and a restoration of transmission at central synapses that are requisite for the bouts of clonus-like activity. What remains unclear is why secondary seizures lead to synaptic failure, while the recovery seizure does not. Following the recovery seizure, synaptic transmission not only is retained, but also is refractory to failure. Moreover, secondary seizures occur only in bss¹ mutants; only initial and recovery seizures are observed in sda and eas mutants (Figure 1D).

Similar to behavior, the tonic-clonic-like period observed electrophysiologically for bss¹ is sensitive to genetic background. For example, secondary seizures are completely absent in bss¹/+ heterozygotes. In these flies, initial seizure and synaptic failure are followed by a recovery seizure, without any occurrence of secondary seizures. Electrophysiological recordings from these heterozygotes resemble those taken from sda and eas flies, although with increased seizure thresholds (19.1 HFS V for bss¹/+ flies; Figure 5E).

A missense mutation of the para Na_V gene is responsible for bss¹:

Genetic localization by duplication/deletion mapping is somewhat problematic for bss¹ because the semidominant nature of its phenotype makes it difficult to distinguish homozygotes from heterozygotes. Since bss¹/+ heterozygote recovery time is shorter than homozygote recovery time, this feature of the phenotype can be used for mapping. However, since recovery time can also vary substantially with age, genetic background, and other factors, all comparisons were between age-matched 2-day-old siblings arising from the same cross to minimize variations due to these sources. Initial determination of the bss¹ chromosomal location indicated the 14E-15A9 cytological interval of the X chromosome (Figure 2, A–C). In particular, localization for bss¹ is based on its inclusion in the Dp(1)r+75c duplication, which reduces recovery time in bss¹/Y hemizygotes (Figure 2B) (15A9: position of its proximal breakpoint), and its exclusion from the Df(1)19 f deletion, which has no effect on recovery time in bss¹/Df(1)19 f heterozygotes, in comparison to bss¹/Df(1)r-D1 heterozygotes, which show increased recovery time (Figure 2C) (14E: approximate position of its proximal breakpoint). This is consistent with the combined findings from other overlapping chromosomal aneuploids of the region (Figure 2, A–C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.—

bss¹ maps to within the para gene. (A) Cytological region affected by chromosomal duplications and deletions used in mapping. Hatched region in Df(1)19 indicates inconsistency between cytological and molecular breakpoints. (B and C) Mean recovery time for bss¹ flies in the presence of duplications or deletions. Error bars represent standard error of the mean; significance was determined by Student's t-test. (B) The Dp(1;2)r⁺75c (14B13-15A9) includes bss¹, providing one boundary at 15A9. This is indicated by the finding that bss¹/Y;Dp flies have a significantly shorter mean recovery time than sibling control bss¹/Y flies (78 ± 3 sec, n = 137, vs. 243 ± 6 sec, n = 109, P = 1.6 × 10⁻⁷³). Confirmatory findings are observed with the larger Dp(1;4)r⁺f⁺ = 13F–16A7 (103 ± 5 sec, n = 100 test, vs. 232 ± 6 sec, n = 105 control, P = 4.4 × 10⁻³⁷). (C) Df(1)19 (13F–14E) fails to include bss¹, providing a boundary at 14E. This is indicated by the finding that the mean recovery time of bss¹/Df flies (46 ± 0.8 sec, n = 113) is not significantly longer than sibling control bss¹/Df;Dp(1;4)r⁺f⁺ flies (47 ± 1.1 sec, n = 134, P = 0.52). Similar findings are observed with the smaller Df(1)4b18 (14B8–14C1) (54 ± 1.0 sec, n = 98 test; 54 ± 1.0 sec, n = 105 control; P = 0.65). These results are in contrast to those observed using the larger Df(1)r-D1 (14C5–15B1) (70 ± 1.3 sec, n = 109 test; vs. 51 ± 1.2 sec, n = 122 control; P = 1.3 × 10⁻²⁰) that must contain bss¹. We conclude that the bss gene lies within the 14E-15A9 region and that the presence of a wild-type copy in a hemizygote mutant fly can reduce the duration of the behavioral phenotype to resemble that of a heterozygote (B), while the absence of a wild-type copy in a bss¹ heterozygote can increase mean recovery time (C). Statistical significance key here, and in Figures 4 and 5, is *P < 0.05, **P < 0.01, or ***P < 0.0001. (D) Recombination mapping placed bss¹ between P[EP] element insertions EY00852 and EY08038. Sequencing of recombinant chromosomes to track 40 polymorphisms detected between parental chromosomes placed bss¹ to the right of a 3-bp deletion at residue 16353592 (marked with an X) and indicates that the bss¹ lesion is in para. See File S1 and Table S1 for more detail.

Initial mapping of bss¹ suggested a location farther to the right on the X chromosome than had been previously indicated (Ganetzky and Wu 1982). This, coupled with uncertainties of deficiency/duplication mapping due to bss¹ semidominance, prompted us to generate a more definitive location by recombination mapping with subsequent molecular determination of crossover sites (fine-structure recombinational mapping) (Figure 2D, Table S1). Recombination events were defined between bss¹ and one of two molecularly defined transposon insertions (P[EP]s). We scored 8977 male progeny from females of the genotype w⁻ bss¹ f⁻/w⁻ P[EP]EY00852. Four animals were found in which a crossover event occurred between bss¹ and P[EP]. The position of P[EP]EY00852 at 16,342,911 bp results in a derived position for bss¹ of 0.04 cM (or ∼10 kb) to the right of this location (Figure 2D). Similarly, we examined 2903 male progeny from females of the genotype w⁻ bss¹ f⁻/w⁻ P[EP]EY08038 and found nine animals in which a crossover event occurred between bss¹ and P[EP]. The location of this insertion at 16,427,296 bp gives a derived position for bss¹ at 0.34 cM (or ∼75 kb) to the left of this location (Figure 2D). These data are consistent with our deficiency/duplication mapping that placed bss¹ in the 14E-15A9 cytological region that corresponds to ∼16,400,000–16,700,000 bp.

Fine-structure recombinational mapping was used to determine more precisely the location of bss¹. This mapping took advantage of the fact that in the chromosomal region containing P[EP]EY00852 there are 40 DNA sequence polymorphisms that can be used to molecularly distinguish the w⁻ bss¹ f⁻ parental chromosome from the P[EP]EY00852 parental chromosome (Table S1). Each of the four male progeny with a crossover between bss¹ and P[EP] is expected to display one of the parental patterns up until the site of the crossover event and, at that point, abruptly shift to the other parental pattern. Each of the crossover locations was determined (Table S1) for the four recombinant chromosomes; these locations were ∼4, 8, 8, and 11 kb to the right of the insertion, respectively. The fourth recombination event, which provides the most useful information because it lies the farthest to the right of P[EP], took place between a TTG 3-bp deletion at location 16,353,592 and a A → G point change at location 16,354,339. This 747-bp segment lies within the extensive para 3′ UTR (Song and Tanouye 2007), 3.5 kb downstream of the para stop codon. Since the bss¹ lesion must lie farther to the right of the crossover point, it is either within the remaining para 3′ UTR or, more likely, within the C-terminal coding sequence of para.

The fine-structure mapping definitively rules out the possibility that bss¹ is a lesion in Rbp2, since all recombination events occurred to the right of the gene. It also makes it unlikely that bss¹ is a lesion in CG9906 because all recombination events occurred much farther to the left of this gene; however, we cannot rule out this possibility. We therefore sought to determine the molecular lesion responsible for the bss¹ mutation by sequencing all 31 para exons, including intron/exon boundaries, as well as the single CG9906 exon. In comparison to the published genome sequence (Adams et al. 2000), in bss¹ we found a single nucleotide change in para exon 30, a C → T point change in the coding sequence that lies at position 16,360,824. This change causes an amino acid substitution from leucine to phenylalanine, corresponding to L1699 in the ParaPA sequence (GenBank accession NP_523371; Figure 3A). The position of this point change in exon 30, 17.9 kb to the right of P[EP]00852, is entirely consistent with the recombination mapping data for bss¹. Only two other nucleotide changes were found in para and both were silent polymorphisms, one in exon 30 and another in exon 31; no changes were found in CG9906. Thus, L1699F is the only amino acid sequence change in the bss¹ para gene, the only candidate for the bss¹ lesion. The bss¹ lesion was confirmed by sequencing exon 30 from Canton-S, Oregon R, and w¹¹¹⁸ strains that all showed the wild-type sequence (Figure 3A). The wild-type strains also carried the same silent polymorphisms found in bss¹, suggesting that they may be sequencing errors in the published sequence. A second mutation, bss², has been reported as being of independent origin (Ganetzky and Wu 1982). Consistent with this, examination of the bss² sequence shows multiple polymorphisms in the para 3′ UTR compared to the bss¹ sequence. Surprisingly, however, and identical to bss¹, L1699 in bss² is substituted by a phenylalanine residue also as the result of a C5364T point change. The L1699 location corresponds to a highly conserved residue in the third membrane-spanning segment (S3b) of homology domain IV of the Na_V channel (Figure 3, B and C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.—

The bss¹ lesion causes an L → F change in a highly conserved residue of Para. (A) Sequencing of para exons from bss¹ mutants identified a C → T point change at nucleotide 5364 in the paraRA cDNA, which is not present in Canton S, Oregon R, or w¹¹¹⁸ strains, resulting in a L1699F residue change. (B) Cartoon depicting the structure of Na_V channels. The L1699 residue lies within homology domain IV, transmembrane segment S3 (star) of the Para protein. (C) L1699 is highly conserved across mammalian neuronal Na_V channels.

Phenotypic severity of para^bss1 can be manipulated by altering expression levels—Reduced expression suppresses phenotypes:

Since bss¹ and bss² mutations are alleles of para, we rename them here para^bss1 and para^bss2, respectively. The finding that they are allelic to para is surprising, given that none of the many other para mutations previously identified demonstrate bang sensitivity. In fact, all other para mutations examined for seizure susceptibility are loss-of-function alleles that have been shown to cause seizure resistance with thresholds above that of wild-type flies. Also, they have all been shown to act as seizure-suppressor mutations, capable of reducing seizure severity in other BS mutants (Kuebler et al. 2001; Song and Tanouye 2007). This implies that para^bss1 is a gain-of-function mutation of para, consistent with its dominant phenotype.

Further evidence for the gain-of-function nature of the para^bss1 allele is revealed by altering expression levels: reducing para^bss1 expression substantially suppressed BS phenotypes. Reduction of para^bss1 expression by para RNA interference utilized the Gal4-UAS system. In hemizygous para^bss1 males, when Na_V expression in all neurons was reduced using Elav-Gal4^3A to drive UAS-paraRNAi, only 65% of flies showed BS paralysis, a reduction of 35% compared to sibling controls that showed 100% BS paralysis (test: para^bss1/Y; Elav-Gal4^3A/UAS-paraRNAi ×2; control: para^bss1/Y; Elav-Gal4^3A/+) (Figure 4A). In contrast, heterozygous para^bss1/+ females showed nearly complete suppression: only 1.6% of flies are BS compared to 87% of sibling controls (genotype of test: para^bss1/para⁺; Elav-Gal4^3A/UAS-paraRNAi ×2; genotype of sibling control: para^bss1/para⁺; Elav-Gal4^3A/+). This finding suggests that the total amount of para^bss1 gene product is important in determining the severity of the phenotype and fits with the observation that para^bss1 hemizygote and heterozygote phenotypes differ significantly in severity. However, unexpected expression asymmetries due to differential paraRNAi knockdown or wild-type/mutant competition for membrane sites are also possibilities.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.—

Genetic manipulation of endogenous para^bss1 expression or function alters seizure phenotypes. (A and B) Penetrance of seizures is decreased when para expression or function is suppressed. (A) Decreasing para expression by driving UAS-paraRNAi pan-neuronally using Elav-Gal4 suppresses seizure occurrence both in para^bss1/+ heterozygous females (1.6% of flies are BS compared to 87% of sibling controls; P = 3.4 × 10⁻²⁴) and in para^bss1/Y hemizygous males (65% of flies are BS compared to 100% of sibling controls; P = 2.7 × 10⁻⁴). (B) Reduced TipE levels in para^bss1/+ heterozygote females suppress bang sensitivity. para^bss1/+; tipE¹/+ animals are 60.9% BS while para^bss1/+; tipE¹/tipE¹ animals are 29.6% BS; control siblings are 99% BS, and P = 8.9 × 10⁻²⁵. (C–E) Ectopic expression of endogenous para^bss1 exacerbates seizure phenotypes. Pan-neuronal ectopic expression of para in a bss¹ background was achieved using (C) cis and (D) trans configurations of the bss¹ mutation relative to P[EP] insertion EY08038. (E) When the bss¹ lesion is in trans to the P[EP] insertion, mean recovery time is virtually unchanged compared to control siblings (test: Elav-Gal4^C155, bss¹/para+ EY08038 = 70 ± 1.9 sec, n = 148, vs. control: Elav-Gal4^C155, bss¹/para+ = 65 ± 1.2 sec, n = 84; P > 0.05) whereas when bss¹ is in cis to the P[EP] insertion, mean recovery time is significantly increased (test: bss¹ EY08038/Elav-Gal4^C155 para⁺ = 116 ± 3.5 sec, n = 119, vs. control: bss¹ EY08038/para⁺ 67 ± 2 sec, n = 118; P = 1.03 × 10⁻²⁵), indicating that increased levels of mutant Para^bss1 protein result in more severe seizure phenotypes. All error bars represent standard error of the mean. Statistical significance was calculated for A and B by chi-square test and for E by Student's t-test.

Some insight into interpreting the paraRNAi results may be found from examining effects in wild-type animals. Interestingly, utilizing a stronger neuronal Gal4 driver (Elav-Gal4^C155) resulted in complete lethality; that is, no flies were observed of the test genotype compared to 196 control siblings (test: Elav-Gal4^C155; UAS-paraRNAi x2; control: Elav-Gal4^C155/+; TM6b/+). A similar level of lethality is seen in a para^bss1/+ heterozygote background. We interpret the lethality as an extreme para loss of function.

Consistent with this interpretation, a weaker pan-neuronal Gal4 driver (1407-Gal4) produced a temperature-sensitive paralytic phenotype, similar to that displayed by weaker para loss-of-function mutants (Suzuki et al. 1971; Siddiqi and Benzer 1976; Wu and Ganetzky 1980). Flies (112/121) were reversibly paralyzed when tested at 38°, compared to 0/120 sibling controls (test: 1407-Gal4/+; UAS-paraRNAi x2/+; control: 1407-Gal4/+; TM6b/+). These observations indicate that paraRNAi was effective at reducing para expression, thereby phenocopying para hypomorphic and null alleles, although with the necessity of using different Gal4 drivers.

Expression of para can also be altered by manipulating levels of TipE, an enhancer of para loss-of-function mutations (Ganetzky 1986; Jackson et al. 1986; Feng et al. 1995; Warmke et al. 1997; Hodges et al. 2002). The mechanism for this interaction is not completely clear; however, it appears to differ from paraRNAi. In contrast to the enhancement of para loss-of-function phenotypes, in para^bss1/+; tipE¹/+ flies, penetrance of BS phenotypes was significantly reduced compared to controls, and in para^bss1/+; tipE¹/tipE¹ flies, penetrance was reduced even further (Figure 4B). However, we saw no suppression of BS phenotypes in para^bss1/Y hemizygotes (126/126 para^bss1/Y; tipE¹/tipE¹ flies tested were BS).

Phenotypic severity of para^bss1 can be manipulated by altering expression levels—Increased expression enhances phenotypes:

If para^bss1 is a gain-of-function mutation, we reasoned that increasing the amount of mutant protein should increase the severity of the BS phenotype. We were able to show this by manipulating endogenous para^bss1 expression levels utilizing the GAL4/UAS system. The pan-neuronal Elav-Gal4^C155 driver was used to activate the UAS sequence in the P[EP]EY08038 insertion that lies immediately upstream of the para gene (e.g., in cis orientation with respect to the mutant gene). Test animals were generated where the P[EP] is cis to para^bss1, thereby driving overexpression of the mutated gene (genotype: w para^bss1 P[EP]EY08038/Elav-Gal4^C155 para⁺) (Figure 4C). The test flies have a much longer recovery time compared to sibling controls (genotype: w para^bss1 P[EP]EY08038/para⁺) that lack the Elav-Gal4^C155 driver (116 vs. 67 sec; Figure 4E). These data indicate that, by increasing the levels of Para^bss1 mutant protein through overexpression, we can significantly increase the severity of the BS phenotype, thereby providing strong phenotypic support that a gain-of-function lesion in para is responsible for bang sensitivity.

We also generated animals where the P[EP] is trans to para^bss1, that is, in cis with respect to the wild-type para gene and thereby driving expression of the wild type, not the mutant, form of para (genotype: Elav-Gal4^C155 para^bss1/w para⁺ P[EP]EY08038) (Figure 4D). These flies have a similar recovery time compared to sibling controls (genotype: Elav-Gal4^C155 para^bss1/w para⁺) that lack P[EP] (70 vs. 65 sec; Figure 4E). These data indicate that increasing the level of endogenous Para⁺ protein through overexpression does not seem to improve the BS recovery time phenotype above that of the heterozygote. A similar finding was observed in the duplication/deficiency analysis where a duplication of the para region in a heterozygous para^bss1/+ animal has no effect on recovery time (Figure 2C) and is seemingly a consequence of the gain-of-function nature of para^bss1 and its complicated gene-dosage relationships.

Expression of wild-type and para^L1699F mutant transgenes:

A wild-type para⁺ transgene was constructed and found to be effective at suppressing para^bss1 phenotypes. Constructs utilized the cDNA of the Na_V1-1 para isoform, the most common splice variant in adults (Olson et al. 2008; Lin et al. 2009), expressed in all neurons with an Elav-Gal4^c155 driver. Bang-sensitive paralysis was completely suppressed (0% BS) in para^bss1/+ heterozygotes carrying two copies of the transgene; in comparison, sibling controls showed 78.3% BS (Figure 5A) (test genotype: Elav-Gal4^C155 para^bss1/+; UAS-Na_V1-1 x2; control genotype: Elav-Gal4^C155 para^bss1/+). The wild-type transgene was also effective in rescuing phenotypes in hemizygous para^bss1/Y flies; bang sensitivity was suppressed to 17.3% (control: 100% BS; Figure 5A) (test genotype: Elav-Gal4^C155 para^bss1/Y; UAS-Na_V1-1 x2; control genotype: Elav-Gal4^C155 para^bss1/Y). Electrophysiologically, for these flies the seizure threshold increases to 19.4 V in animals carrying the transgene, compared to the lower 4.8-V seizure threshold of sibling control flies (Figure 5C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.—

Ectopic expression of para isoform Na_V1-1 wild-type and mutant forms alters seizure severity. (A–C) Pan-neuronal expression of Na_V1-1, driven by Elav-Gal4 with one or two copies of UAS-Na_V1-1, in a para^bss1 background suppresses seizure phenotypes in comparison to sibling controls. (A) Penetrance of bang sensitivity in para^bss1/+ heterozygotes is drastically reduced from 78% (n = 166) to 0.8% (n = 252) with one copy of UAS-Na_V1-1, while two copies completely suppresses the phenotype (n = 125) P = 6.7 × 10⁻⁶⁴. Bang sensitivity of para^bss1/Y hemizygotes is reduced from 98.9% (n = 88) to 59.6% (n = 193) with one copy and to 17.3% (n = 104) with two copies of UAS-Na_V1-1, P = 6.4 × 10⁻¹⁶. (B) Mean recovery time of para^bss1/Y hemizygotes is reduced from 163 ± 4.6 sec (n = 78) to 47 ± 1.6 sec (n = 59) with one copy of UAS-Na_V1-1, P = 8.4 × 10⁻⁴² and shows a similar recovery time to para^bss1/+ heterozygotes (33 ± 0.8 sec, n = 77). (C) Average seizure threshold of para^bss1/Y hemizygotes is significantly increased from 4.5 ± 0.3 V (n = 37) to 12.3 ± 2.6 V (n = 12) with one copy and to 19.4 ± 3.2 V (n = 14) with two copies of UAS-Na_V1-1; P = 1.6 × 10⁻⁸. (D and E) Pan-neuronal expression of Na_V1-1^L1699F, driven by Elav-Gal4 with two copies of UAS-Na_V1-1^L1699F in a para^bss1 background, enhances seizure phenotypes. (D) Mean recovery time of para^bss1/+ heterozygotes increases from 38 ± 1 sec, n = 113, in sibling controls to 52 ± 1 sec, n = 146, with two copies of UAS-Na_V1-1^L1699F; P = 1.3 × 10⁻¹⁷. (E) Average seizure threshold decreases from 19.1 ± 1.8 V (n = 24) in sibling controls to 12.2 ± 1.3 V (n = 18) with two copies of UAS-Na_V1-1^L1699F; P = 3.8 × 10⁻³. In all cases, error bars represent standard error. Statistical significance was calculated for A by chi-square test; for B, D, and E by Student's t -test; and for C by ANOVA.

The most straightforward explanation for rescue of BS phenotypes by wild-type Na_V1-1 is that the cDNA is biologically active in vivo and rescues in a manner similar to genetic duplications carrying the para⁺ genomic region. We expect, however, that the explanation likely is more complex since control crosses indicate that, although the cDNA transgene is effective at rescuing the gain-of-function phenotypes of para^bss1, it is ineffective at rescuing the loss-of-function phenotypes in para^ts1, para^ST76, and para^lk2 mutants. These differences may be related to the complexity of the para gene that produces a large number of different splice variants; 53 different variants have been characterized to date, and the cDNA Na_V1-1 used here corresponds to only one of these variants (O'Dowd et al. 1995; Olson et al. 2008; Lin et al. 2009). It is unlikely that a single isoform could effectively rescue the entire range of para functions.

A mutant para transgene was constructed by engineering the L1699F mutation into the Na_v1-1 cDNA. The expectation was that since the para^bss1 allele is a gain-of-function mutation, the Na_V1-1^L1699F transgene would confer BS phenotypes on wild-type flies. The results obtained showed the mutant transgene to be less effective than expected: only a few flies carrying it displayed bang-sensitive paralysis (1.6%, 4/250 BS flies), although the average recovery period for the few flies paralyzed (33 sec) resembled the recovery period of para^bss1/+ heterozygotes (genotype of test flies: Elav-Gal4^C155/+; UAS-Na_V1-1^L1699F/UAS-Na_V1-1^L1699F). Increasing copy numbers of either UAS or Gal4 transgenes in an attempt to increase penetrance resulted in larval lethality. Nrv2-Gal4, another pan-neuronal driver, showed a similarly low penetrance of BS.

We then questioned whether or not the causative para^bss1 lesion was indeed the L1699F residue change. If not, then the Na_V1-1^L1699F transgene should behave as a wild-type transgene and rescue para^bss1 phenotypes. The results obtained showed that the mutant transgene could not rescue para^bss1, but instead exacerbated BS phenotypes. para^bss1/+ heterozygous females carrying the transgene were tested for recovery time and electrophysiological seizure threshold compared to their control siblings (test: Elav-Gal4^C155, para^bss1/+; UAS-Na_V1-1^L1699F x2/+; control: Elav-Gal4, para^bss1/+). A significant increase in recovery time was observed from 38 to 49 sec (Figure 5D), as well as a decrease in seizure threshold from 19.1 to 12.1 V (Figure 5E) in test flies carrying the transgene compared to their control siblings. From this, we conclude that Na_V1-1^L1699F does not behave as a wild-type transgene; rather, the results are most consistent with L1699F being responsible for the para^bss1 lesion with low phenotypic penetrance of the mutant transgene.

Na_V1-1^L1699F channels show altered inactivation:

The functional consequences of the L1699F mutation in cDNA Na_v1-1 were examined by expressing wild-type and mutant channels in Xenopus oocytes followed by electrophysiological testing. Mutant channel expression in oocytes appears to be substantially reduced compared to wild-type channel controls (∼10-fold) as indicated by significantly decreased peak current amplitudes (Figure 6A). Mutant and wild-type channels are similar in their voltage dependence and kinetics of channel activation (Figure 6, B and G). In contrast, inactivation for the mutant channel was significantly different from that of wild type. Voltage dependence of fast inactivation is shifted to more positive potentials for mutant channels, relative to wild type (V_1/2 = −38.3 ± 1.63 mV compared to −49.43 ± 2.48 mV, P < 0.01; Figure 6, C and G). While voltage dependence of slow inactivation was not altered, the slope factor (k) was significantly changed in the mutant channel (k = 5.32 ± 0.5 compared to 9.26 ± 1.85 for wild type, P < 0.01; Figure 6, D and G). Recovery from fast or slow inactivation was comparable in mutant and wild-type channels (Figure 6, E and F). Although a heterologous expression assay cannot be said to definitively account for the effect of the L1699F lesion in vivo, it indicates that a likely consequence of this mutation is the alteration of fast inactivation such that neurons expressing Para^bss1 will be more excitable, giving rise to increased risk of seizure.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.—

Channel inactivation is altered by the L1699F mutation. (A) Histogram comparing amplitudes of peak current from Na_V1-1 wild-type (WT) and Na_V1-1^L1699F mutant channels. Peak current was elicited by a 20-msec depolarization to −10 mV from the holding potential of −120 mV. The same amount of cRNA (5 ng/oocyte) of wild-type or mutant channels was injected into oocytes, and a recording was made at day 4. (B) Voltage dependence of activation in wild-type and mutant channels is comparable. (C) Voltage dependence of fast inactivation in mutant channels shows a shift to more positive potentials relative to wild-type channels. (D) Voltage dependence of slow inactivation in mutant channels is comparable to wild type; however, the slope value is altered. Recovery from (E) fast inactivation and (F) slow inactivation is comparable in wild-type and mutant channels. (G) Gating properties of Na_V1-1 and Na_V1-1^L1699F channels. V_1/2 activation voltages are similar for wild type and mutant; however, the V_1/2 fast inactivation difference is significant. Although V_1/2 slow inactivation is comparable for wild-type and mutant channels, the slope value (k) is significantly decreased. For A–F, graphical representations of recording protocols are shown above each graph; for details, see materials and methods.