Distinct roles for two histamine receptors (hclA and hclB) at the Drosophila photoreceptor synapse

Abstract

Histamine (HA) is the photoreceptor neurotransmitter in arthropods, directly gating chloride channels on large monopolar cells (LMCs), postsynaptic to photoreceptors in the lamina. Two histamine-gated channel genes that could contribute to this channel in Drosophila are hclA (also known as ort) and hclB (also known as hisCl1), both encoding novel members of the Cys-loop receptor superfamily. Drosophila S2 cells transfected with these genes expressed both homomeric and heteromeric histamine-gated chloride channels. The electrophysiological properties of these channels were compared with those from isolated Drosophila LMCs. HCLA homomers had nearly identical HA sensitivity to the native receptors (EC(50) = 25 microM). Single-channel analysis revealed further close similarity in terms of single-channel kinetics and subconductance states ( approximately 25, 40, and 60 pS, the latter strongly voltage dependent). In contrast, HCLB homomers and heteromeric receptors were more sensitive to HA (EC(50) = 14 and 1.2 microM, respectively), with much smaller single-channel conductances ( approximately 4 pS). Null mutations of hclA (ort(US6096)) abolished the synaptic transients in the electroretinograms (ERGs). Surprisingly, the ERG "on" transients in hclB mutants transients were approximately twofold enhanced, whereas intracellular recordings from their LMCs revealed altered responses with slower kinetics. However, HCLB expression within the lamina, assessed by both a GFP (green fluorescent protein) reporter gene strategy and mRNA tagging, was exclusively localized to the glia cells, whereas HCLA expression was confirmed in the LMCs. Our results suggest that the native receptor at the LMC synapse is an HCLA homomer, whereas HCLB signaling via the lamina glia plays a previously unrecognized role in shaping the LMC postsynaptic response.

Figure 1.

Dose–response functions of HCLA and HCLB channels. A, Current recorded from an S2 cell transfected with hclA, whole-cell patch clamped at −60 mV. HA was applied using a 10-channel parallel-flow drug delivery system (Skingsley et al., 1995) at the concentrations indicated. B, Simple Hill fittings for data from cells transfected with hclA (circles), hclB (triangles), or both (diamonds). C, Composite Hill curve [Hill equation with weighted components for the contributions of different receptor types (Eq. 2)]; fits for data from cotransfected cells (diamonds and solid curve), as well as the individual components of the equation (dashed curves), are shown. Error bars represent ±1 SD.

Figure 2.

A typical noise analysis experiment. A, Whole-cell patch-clamp recording from an S2 cell transfected with hclA, held at −80 mV. Current was recorded in the presence (10 μm) and absence of HA. The capacitative transients generated by a +10 mV pulse were used to determine the clamp time constant, τ_RC (180 μs). B, Power spectrum (P) for channel noise at 10 μm HA, fitted with a double Lorentzian function. The component Lorentzian curves, L₁ and L₂, are also plotted (time constants, τ₁ = 0.08 ± 0.01 ms and τ₂ = 0.98 ± 0.03 ms, with normalized weights w₁ = 0.07 ± 0.01 and w₂ = 0.93 ± 0.01). The correction factor to compensate the variance was calculated to be 1.71. C, Plot of macroscopic current variance (σ²) against control-subtracted average current (I) at three different HA concentrations and without HA. The slope of the linear regression, hence single-channel current, was −4.47 pA (R² = 0.99), indicating a single-channel conductance (γ) of 55.9 pS (at −80 mV).

Figure 3.

Single-channel properties of HCLA channels. Left, Sample single-channel recordings from HCLA homomers (from hclA-transfected S2 cells) and native HA receptors (from Drosophila LMCs), recorded at the holding potentials indicated, with 10 μm HA in the recording pipette in both cases. At least 5000 transitions from each recording were idealized with SCAN. Middle, Amplitude (Amp) distributions of time course-fitted opening transitions, generated and fitted with multiple Gaussians by maximum likelihood with EKDIST (fitted amplitudes and relative contributions of subconductance states are indicated). Right, Open time distributions of fitted transitions, generated and fitted by multiple exponentials with EKDIST. Fitted time constants and relative contributions are indicated in insets. A 100 μs time resolution was imposed for all distributions.

Figure 4.

ERGs from hclA (ort) and hclB (hisCl1) mutants. A, ERGs recorded in response to 2 s light steps of increasing intensity in wild-type (wt; i. e., w¹¹¹⁸) control flies. The conspicuous transients at light on and off represent the contribution of the LMCs in the lamina. B, C, ERGs recorded using identical stimulation in a null hclB mutant (w¹¹¹⁸;;hisCl1¹³⁴; B) and a null hclA mutant (w¹¹¹⁸;;ort^US6096; C). Examples of “on” (left) and “off” (right) transients in response to increasing intensities (7 steps covering 6 log units of intensity) are shown on the right on an expanded scale after aligning baselines immediately before light on or off. Transients were completely eliminated in ort null (note larger scale). Transients in hisCl1¹³⁴ were qualitatively similar in waveform to wild-type controls, but the “on” transients were approximately twofold larger. D, Response intensity functions for the maintained negative plateau (photoreceptor component) in wild type (wt; w¹¹¹⁸), ort^US6096, and hisCl1¹³⁴ (left); and both “on” and “off” transients in wt and hisCl1¹³⁴ (right). Data (mean ± SD) are based on ERGs from n = 5–6 flies for each genotype.

Figure 5.

Intracellular LMC recordings from wild-type (wt; w¹¹¹⁸) and hclB (hisCl1¹³⁴) mutants. A, B, Responses to brief (10 ms) flashes of increasing intensity (log₁₀ steps) in wild-type (w¹¹¹⁸; A) and w¹¹¹⁸;;hisCl1¹³⁴ mutants (B). C, Averaged, normalized response intensity (V/log I) functions in wild-type (w¹¹¹⁸) and w¹¹¹⁸;;hisCl1¹³⁴ mutants. D, Averaged time to peak at different intensities (mean ± SD; n = 6 cells).

Figure 6.

Enhancer analysis of the two histamine receptor genes. A, ort (hclA)-GAL4-driven expression of GFP. The reporter was detected in lamina (la) monopolar cells L1–L3 (see also inset) and medulla (me) cells (indicated by open arrowhead). cb, Cell bodies. The L3 terminal layer is indicated by a filled arrowhead. The inset shows a 40× magnification of the distal medulla. Layers M1, M2, M3, and M5 were immunopositive, corresponding to the specializations of L1 (M1 and M5), as well as the terminals of L2 (M2) and L3 (M3, indicated by filled arrowhead). Scale bars, 20 μm. B, hclB-GAL4-driven GFP expression. Intrinsic cells in the lamina, but not the monopolar cells, were labeled by this line (compare with A). Scale bar, 20 μm. Inset, Cartridge cross sections reveal that the fibers are epithelial glia that surround the neuroommatidia. Note that the axon terminals of R1–R6 are also not stained (one indicated by arrowhead). Scale bar, 5 μm. C, The same driver as in B combined with a nuclear-localized lacZ reporter. The cell bodies of the immunopositive epithelial glia are within the lamina neuropil (outline indicated by dots), not in the cell body layer, which is located more distally (arrow). Scale bar, 20 μm. D, Schematic of the genomic structure of the two histamine receptor genes hclA and hclB. Light gray lines indicate enhancer regions that were used for expression of GAL4. Scale bar, 1 kb.

Figure 7.

Expression profile of hclA and hclB determined by mRNA tagging. Shown is a RT-PCR gel using primers for hclA and hclB, as well as elav and repo controls (to identify neurons and glia, respectively) and rpl32 (“housekeeping” control). All transcripts were detected in whole-brain tissue using conventional RT-PCR (top row). mRNA immunoprecipitated from all neurons using elav-Gal4 × UAS-hPF included both hclA and hclB transcripts, as well as elav (positive control); the lack of repo signal (negative control) confirms that there was no contaminating signal. Glial mRNA (using repo-gal4) contained hclB (and repo) but no hclA (or elav). In the LMCs (L2-Gal4), only hclA and elav were detected. All samples expressed the general housekeeping control gene rpl32 (ribosomal protein). ntc, No template control; m, markers.