Functional Characterization of Cnidarian HCN Channels Points to an Early Evolution of Ih

Abstract

HCN channels play a unique role in bilaterian physiology as the only hyperpolarization-gated cation channels. Their voltage-gating is regulated by cyclic nucleotides and phosphatidylinositol 4,5-bisphosphate (PIP2). Activation of HCN channels provides the depolarizing current in response to hyperpolarization that is critical for intrinsic rhythmicity in neurons and the sinoatrial node. Additionally, HCN channels regulate dendritic excitability in a wide variety of neurons. Little is known about the early functional evolution of HCN channels, but the presence of HCN sequences in basal metazoan phyla and choanoflagellates, a protozoan sister group to the metazoans, indicate that the gene family predates metazoan emergence. We functionally characterized two HCN channel orthologs from Nematostella vectensis (Cnidaria, Anthozoa) to determine which properties of HCN channels were established prior to the emergence of bilaterians. We find Nematostella HCN channels share all the major functional features of bilaterian HCNs, including reversed voltage-dependence, activation by cAMP and PIP2, and block by extracellular Cs+. Thus bilaterian-like HCN channels were already present in the common parahoxozoan ancestor of bilaterians and cnidarians, at a time when the functional diversity of voltage-gated K+ channels was rapidly expanding. NvHCN1 and NvHCN2 are expressed broadly in planulae and in both the endoderm and ectoderm of juvenile polyps.

Fig 1. Multiple amino acid sequence alignment comparing sea anemone, human and Drosophila HCN channels.

The alignment compares two putative HCN channels from the sea anemone Nematostella vectensis (NvHCN1 and NvHCN2) to HCN channels from Drosophila (DmHCN) and human (HsHCN1-4); residue positions are given at the right margin. Positions identical in all sequences are shaded in blue, and additional residues identical in Nematostella or the bilaterian sequences are shaded in green or brown, respectively. Approximate positions of 6 transmembrane domains (S1-S6) and the selectivity filter (PORE) are underlined and labeled in black or cyan, respectively. The C-linker is underlined in light green and the CNBD is underlined in red. Residues that contact cAMP in an HCN2 structure are boxed with a red outline. Accession numbers and full sequences are provided in S1 Table.

Fig 2. Bayesian inference phylogeny of the HCN channel family.

Channels in the HCN family are color coded by phylogenetic group: bilaterians (blue), cnidarians (red), sponge (light green), ctenophore (dark green) and choanoflagellate (purple). The HCN family and the EAG and CNG family outgroups are indicated with shading and labels at the right margin. All outgroup sequences are cnidarian, but are not colored. Posterior probabilities for nodes are indicated and the scale bar is in substitutions/site. Channel names are given at branch tips with species prefixes as follows:; Amil, Acropora millepora, stony coral; Amel, Apis mellifera, honey bee; Ccan, Corticium candelabrum, sponge; Cint, Ciona intestinalis, tunicate; Dmel, Drosophila melanogaster, fruit fly; Dgla, Dryodora glandiformis, ctenophore; Hsap, Homo sapiens, human; Hvul, Hydra vulgaris, hydra; Lgig, Lottia gigantea, limpet; Nvec, Nematostella vectensis, sea anemone; Ofav, Orbicella faveolata, coral; Pint, Panulirus interruptus, lobster; Spur, Stronglyocentrotus purpuratus, sea urchin; and Sros, Salpingoeca rosetta, choanoflagellate. Full and aligned sequences for all branches are given in S1 Table.

Fig 3. Nematostella HCN orthologs NvHCN1 and NvHCN2 are activated by hyperpolarization.

(A) Families of inward currents recorded from Xenopus oocytes expressing NvHCN1 (left) and NvHCN2 (right) in response to 4 s hyperpolarizing voltage steps ranging from -40 to -120 mV in 10 mV increments. The holding potential was -30 mV and tail currents were recorded at -50 mV. Scale bars indicate time and current size and the -100 mV sweep is indicated with an arrow. (B) GV curves for NvHCN1 and NvHCN2. Fractional open probability (G/G_MAX) was measured from isochronal tail currents recorded at -50 mV after 4 s hyperpolarizing steps to the indicated voltages as shown in (A). Data are normalized and show mean ± S.E.M. of measurements from individual cells. The smooth curves show single Boltzmann fits. V₅₀, slope values and sample numbers are reported in Table 1.

Fig 4. Cs⁺ block of NvHCN1 and NvHCN2.

(A) Current traces recorded at -100 mV from oocytes expressing NvHCN1 or NvHCN2 before (black) and after (red) application of 5 mM Cs⁺ to the bath. The holding potential was -30 mV and outward tail currents were recorded at 0 mV. (B) Fraction of current remaining after application of 5 mM Cs⁺ for NvHCN1 and NvHCN2. Data show mean ± S.E.M. (n = 6 and 7 for NvHCN1 and NvHCN2, respectively).

Fig 5. 8-Br-cAMP enhances activation of NvHCN2 but not NvHCN1.

(A) Examples of currents recorded in response to 4 s -100 mV voltage steps before (black) and after (red) bath application of 2 mM 8-Br-cAMP (-30 mV holding potential, tails recorded at 0 mV). (B) Half activation times at -100 mV for NvHCN1 and NvHCN2 before (black) and after (red) the addition of 2 mM 8-Br-cAMP. (C) 75% deactivation time at 10 mV for NvHCN1 and NvHCN2 with (red) and without (black) 2 mM 8-Br-cAMP. (D,E) GV curves for NvHCN1 and NvHCN2 with (red) and without (black) 2 mM 8-Br-cAMP. All data show mean ± S.E.M. and smooth curves in D and E show single Boltzmann fits of the data (parameters and sample number reported in Table 1). Sample numbers for (C) and (D) are shown on data bars. **Significance at p < 0.01 (t-test); ns, no significant difference.

Fig 6. PIP₂ depletion inhibits activation of NvHCN1.

(A) Normalized current traces recorded in response to a 4 s -100 mV voltage step for control (black) and with PIP₂ depletion by CiVSP (red). The holding voltage was -30 mV and tails were recorded at -50 mV. CiVSP was activated with a 2 s step to + 60 mV 400 ms prior to the hyperpolarizing current step. (B) Half activation time at -100 mV for NvHCN1 control (black) and for CiVSP-dependent PIP₂ depletion (red). (C) 75% deactivation time at +10 mV for NvHCN1 controls (black) and for PIP₂ depletion (red). (D) GV curves for NvHCN1 controls and NvHCN1 + CiVSP-dependent PIP₂ depletion. All data show mean ± S.E.M.; ** in B and C indicates significant difference (p < 0.01, t-test, n = 7–9). Smooth curves in D show single Boltzmann fits of the data; fit parameters and sample numbers are reported in Table 1.

Fig 7. In situ hybridization of NvHCN1 and NvHCN2 expression.

Examples of typical expression patterns in gastrulae, planulae and juvenile polyps are shown for NvHCN1 (top row) and NvHCN2 (bottom row). Animals were hybridized with anti-sense probes and detected colorimetrically (purple).