Molecular Mechanism of Rectification at Identified Electrical Synapses in the Drosophila Giant Fiber System

Summary Electrical synapses are neuronal gap junctions that mediate fast transmission in many neural circuits [1–5]. The structural proteins of gap junctions are the products of two multigene families. Connexins are unique to chordates [3–5]; innexins/pannexins encode gap-junction proteins in prechordates and chordates [6–10]. A concentric array of six protein subunits constitutes a hemichannel; electrical synapses result from the docking of hemichannels in pre- and postsynaptic neurons. Some electrical synapses are bidirectional; others are rectifying junctions that preferentially transmit depolarizing current anterogradely [11, 12]. The phenomenon of rectification was first described five decades ago [1], but the molecular mechanism has not been elucidated. Here, we demonstrate that putative rectifying electrical synapses in the Drosophila Giant Fiber System [13] are assembled from two products of the innexin gene shaking-B. Shaking-B(Neural+16) [14] is required presynaptically in the Giant Fiber to couple this cell to its postsynaptic targets that express Shaking-B(Lethal) [15]. When expressed in vitro in neighboring cells, Shaking-B(Neural+16) and Shaking-B(Lethal) form heterotypic channels that are asymmetrically gated by voltage and exhibit classical rectification. These data provide the most definitive evidence to date that rectification is achieved by differential regulation of the pre- and postsynaptic elements of structurally asymmetric junctions.

Two methods, neuronal dye coupling and electrophysiological recordings of muscle activity, were used to assess the degree of rescue of synaptic transmission in the GFS of shakB 2 mutants expressing UAS-shakB transgenes under the control of A307-GAL4 or c17-GAL4 (Figure 3;   Tables S1 and S2). The electrophysiological approach provided a more sensitive measure of GF-TTMn conductivity than dye coupling in so far as it allowed the detection of synaptic transmission in c17 transgenic flies that could not reliably be visualized by dye transfer. It was, however, less sensitive than dye transfer in detecting transmission through PSI. DLM responses were infrequently detected in A307, shakB(n+16) flies and never in c17, shakB(n+16) flies (Table S2) despite the presence of GF-PSI dye coupling ( Figure 3H; Table S1). These data can be reconciled if one considers that the GF activates the TTM through a monosynaptic pathway and the DLMs through a disynaptic pathway. We have partially rescued first order synapses; this is sufficient in the monosynaptic pathway to drive muscle responses but is often inadequate in the longer disynaptic pathway that (given the lack of any DLM response in shakB 2 ; Table S2) seems to depend more heavily on electrical connections than the pathway to the TTM.  Table S1. Controls are shakB 2 /+; UAS-shakB(n+16); shakB 2 are shakB 2 /Y; UAS-shakB(n+16). Parameters for controls and shakB 2 mutants are similar to those previously reported [S13, S14].    Figures 4D and 4E (heterotypic) and Figures 4I and 4J (homotypic). These data were fitted to a Boltzmann equation as described in Experimental Procedures. For heterotypic pairs, data were fitted to a single sigmoid; for homotypic pairs, separate curves were fitted for positive (+) and negative (-) Vjs. Gjmax and Gjmin are, respectively, maximum and minimum normalized Gj. V 0 is the Vj at which Gj is halfway between Gjmin and Gjmax. n is the number of cell pairs. N+16: ShakB(Neural+16); L: ShakB(Lethal)
Preparations were then washed in PBT-X, 6 x 10 min, incubated in secondary fluorescent antibodies (in blocking solution) for 2 hr at room temperature and washed as before. Labeled preparations were dehydrated through a glycerol or ethanol series and slide-mounted in Vectashield (Vector Laboratories) or methylsalicylate, respectively.

Intracellular Injection of Lucifer Yellow in the GFS
Nervous systems were dissected in cold Drosophila saline and mounted on poly-l-lysine (0.1% w/v)-coated glass coverslips in a chamber containing fresh saline. The chamber was transferred to the stage of a microscope equipped with fluorescence and Nomarski water-immersion objectives. One GF axon in the cervical connective was impaled with a Lucifer Yellowcontaining glass microelectrode and dye filled as previously described [S3].

Microscopy and Image Analysis
DIG labeled preparations were viewed and photographed under Nomarski or fluorescence optics with Zeiss Axiophot or Leica DMR microscopes. All other images are projections of confocal z series taken at 1 μm steps with Zeiss LSM 510 or Leica TCS SP2 confocal microscopes and associated software. Images were processed in Adobe Photoshop 6.0.

GFS Electrophysiology
Intracellular recordings of activity in the TTM and DLM muscles were made in response to extracellular activation of the GFs as described [S5].

Functional Expression of In Vitro Transcribed RNAs in Paired Xenopus Oocytes
shakB RNAs were transcribed in vitro as described [S8]. Stage V-VI Xenopus laevis oocytes were isolated and defolliculated as described [S9]. Oocytes were cultured in Barth's medium was removed after brief incubation in hypertonic medium [S11] and cells were paired with the vegetal membranes apposed. After incubation for 24-48 hr the formation of intercellular channels was recorded by dual voltage clamp [S12]. Each cell of a pair was impaled with two borosilicate glass electrodes filled with recording solution [S8]. Voltage clamp was carried out with two GeneClamp500 amplifiers interfaced to a PC via Digidata 1320A and data were recorded and analyzed using pClamp 8.0 software (Axon Instruments). The relationship between voltage and junctional conductance was quantified by fitting the data (Origin 7 software, OriginLab) to a Boltzmann equation of the form y = A2 + A1-A2/(1+exp((x-x0)/dx)) where A1 and A2 are maximum and minimum conductance, respectively, x0 is the voltage at which conductance is halfway between its maximum and minimum values and dx represents the change in conductance over the voltage range, a measure of voltage sensitivity.