Processing of retinal signals in normal and HCN deficient mice

Abstract

This study investigates the role of two different HCN channel isoforms in the light response of the outer retina. Taking advantage of HCN-deficient mice models and of in vitro (patch-clamp) and in vivo (ERG) recordings of retinal activity we show that HCN1 and HCN2 channels are expressed at distinct retinal sites and serve different functions. Specifically, HCN1 operate mainly at the level of the photoreceptor inner segment from where, together with other voltage sensitive channels, they control the time course of the response to bright light. Conversely, HCN2 channels are mainly expressed on the dendrites of bipolar cells and affect the response to dim lights. Single cell recordings in HCN1⁻/⁻ mice or during a pharmacological blockade of I(h) show that, contrary to previous reports, I(kx) alone is able to generate the fast initial transient in the rod bright flash response. Here we demonstrate that the relative contribution of I(h) and I(kx) to the rods' temporal tuning depends on the membrane potential. This is the first instance in which the light response of normal and HCN1- or HCN2-deficient mice is analyzed in single cells in retinal slice preparations and in integrated full field ERG responses from intact animals. This comparison reveals a high degree of correlation between single cell current clamp data and ERG measurements. A novel picture emerges showing that the temporal profile of the visual response to dim and bright luminance changes is separately determined by the coordinated gating of distinct voltage dependent conductances in photoreceptors and bipolar cells.

Figure 1. Transcript expression and immunohistochemistry of HCN channels.

A: HCN channel 1–2 mRNA expression in murine retina. The amount of HCN amplicons is compared to cyclophilin expression. B: Confocal images of retinal sections immunolabeled with rabbit polyclonal antibodies (green fluorescence) specific for HCN1 (upper panel) and HCN2 (bottom panel) in HCN^+/+, HCN1^−/− and HCN2^−/− mice. In addition to immunolabeling with the antibody for HCN2 (bottom panel), the retinas were also stained with an antibody against PKC, a specific marker for rod bipolar cells (red fluorescence). Scale bars, 10 µm.

Figure 2. Voltage-gated currents and flash responses of rods in HCN^+/+, HCN1^−/− and HCN2^−/− mice.

A (records above): currents recorded in rods in response to hyperpolarizing voltage clamp steps from a holding potential of −53 mV, to −60/−67/−74/−81/−88/−95/−102/−109 mV, and depolarization to −65 mV. A slow-activating I_h current was present in HCN^+/+ and HCN2^−/−, but not in HCN1^−/− rods. In the latter, the absence of I_h left an instantaneous inward rectifying current (dots). A (records below): currents recorded in the same rods in response to depolarizing voltage steps from −64 mV, to −57/−50/−43/−36/−29/−22/−15 mV, and repolarization to −60 mV. A slow-activating I_kx current was present in all rods (stars). B: photovoltage responses of dark adapted rods to flashes of green light (514 nm) of increasing strength, covering over 3-log units (range 0.2–780 photons/µm²). The fast initial nose following bright flashes was present in both normal and HCN deficient rods. Flashes were delivered at the rods' apparent dark membrane potential (V_d). Baselines are aligned to each other (max shift 2 mV). Records are averages of several sweeps and are ‘box car’ filtered with a window of 20 ms. Data obtained at 24°C.

Figure 3. Rod responses to bright flashes in HCN^+/+, HCN1^−/− and HCN2^−/− mice at different membrane potentials.

A: bright flashes (109 photons/µm²) were delivered in HCN^+/+ rods at the dark membrane potential (V_d) and at a more hyperpolarized potential maintained by constant current injection (upper traces). The same flashes were also delivered in voltage clamp while holding the rod at −40 and −50 mV, respectively. In current clamp the nose was more prominent at the hyperpolarized potential (star), but it was always absent in voltage clamp. B: in contrast to HCN^+/+, in HCN1^−/− rods the nose (flash strength 195 photons/µm²) was present at V_d but disappeared at more negative potentials. C: in a rod from an HCN2^−/− animal, hyperpolarization speeded up the nose (54 photons/µm²) similarly to what observed in normal HCN^+/+ mice. D: pharmacological blockade of I_h with 3 µM ivabradine (right traces) abolished the nose (236 photons/µm²) at hyperpolarized (dot) but not at depolarized potentials. Compare this with the behavior of a rod recorded in the same preparation prior to perfusion with ivabradine and stimulated with the same flash (left traces). Records are averages of several sweeps and are ‘box car’ filtered with a window of 20 ms. Data obtained at 24°C.

Figure 4. Summary graph of the degree of nose in the bright flash response in HCN^+/+, HCN1^−/− and HCN2^−/− mice, as a function of dark membrane potential (V_d).

The nose was quantified by taking the maximum slope of photovoltage trajectory in the first second following the flash (inset). The dark membrane potential was imposed by constant current injection. Inspection of the data shows that I_h is entirely responsible for generating the nose at hyperpolarized potentials, while at depolarized potentials this role is played by another current, presumably I_kx. There may exist a range of V_d within which both mechanisms cooperate to quicken the bright flash response of rods.

Figure 5. ERG response to flashes of increasing intensity.

A: averaged ERG responses of increasing light intensity in the HCN^+/+ (gray, n = 18), HCN1^−/− (red n = 18) and HCN2^−/− (blue, n = 10). Dim, intermediate and bright flash intensities are shown in the left, middle and right panel, respectively. B–C: collected data of the b-wave peak amplitude as a function of the flash intensity in HCN^+/+, HCN1^−/− and HCN2^−/−, relative amplitudes were normalized at their maximum value. The intensity of the flash is expressed as a number of photoisomerizations per rod (Φ) per flash. The dotted ovals in B indicate the dim, intermediate and bright flash responses illustrated in A.

Figure 6. ERG response to sinusoidal time varying luminance stimuli.

FRCs obtained by sinusoidal modulation of a mean luminance equivalent to 38.79Φ in HCN^+/+ (n = 15), HCN1^−/− (A, n = 17) and HCN2^−/− (B, n = 12) before, and after blocker injection (12 mg/kg; n = 7/n = 6 respectively for HCN1^−/− and HCN2^−/−). Relative amplitude was normalized at their resonance peak. Stimulus contrast, 85%; vertical bars = SEM.

Figure 7. Sinusoidal current injections explore the FRC of HCN^+/+ and HCN1^−/− rods.

A/B: ZAP stimuli consisting of small amplitude sinusoidal modulated current stimuli (0.1 and 30 Hz, duration 50 s) were delivered in current clamp in dark adapted rods, at various potentials by constant current injection. Voltage responses are shown below, together with the corresponding normalized input impedance profiles. Resonance is expressed in both mouse lines, although in HCN1^−/− it is entirely abolished when the membrane is hyperpolarized below −55/−60 mV. C: The same protocol delivered in rod bipolars highlights their resonant membrane properties, which disappeared upon perfusion with the specific HCN inhibitor ivabradine 3 µM. Records are averages of several sweeps. Data obtained at 24°C.