Persistent effects of cyclic adenosine monophosphate are directly responsible for maintaining a neural network state

Abstract

Network states are often determined by modulators that alter the synaptic and cellular properties of the constituent neurons. Frequently neuromodulators act via second messengers, consequently their effects can persist. This persistence at the cellular/molecular level determines the maintenance of the state at the network level. Here we study a feeding network in Aplysia. In this network, persistent modulation supports the maintenance of an ingestive state, biasing the network to generate ingestive motor programs. Neuropeptides that exert cyclic adenosine monophosphate (cAMP) dependent effects play an important role in inducing the ingestive state. Most commonly, modulatory effects exerted through cAMP signaling are persistent as a consequence of PKA activation. This is not the case in the neurons we study. Instead maintenance of the network state depends on the persistence of cAMP itself. Data strongly suggest that this is a consequence of the direct activation of a cyclic nucleotide gated current.

Figure 1

PKA is not required for the induction of ingestive priming (see also Fig. S1). (A,B) PKI loading does not impact priming of B48 activity observed with repeated CBI-2 stimulation. Six cycles of motor activity were triggered by CBI-2 in preparations in which pairs of B48 neurons were loaded intracellularly with vehicle (control, black) or PKI (blue). Increased B48 firing, i.e. priming, was observed in both cases. (C,D) CBI-2 induced increases in B48 excitability persist in the presence of PKI. B48 excitability was measured by injecting constant current pulses before priming (baseline) and for 80 min after priming in neurons injected with vehicle (control, black) and in neurons injected with PKI (blue). Gray bars indicate priming (Stim CBI-2). PKI loading had no effect. Traces are membrane voltage recorded from bilateral pairs of B48 neurons, during CBI-2 elicited motor programs (A) and during excitability tests (C). Sample sizes: Panel B (N = 5), Panel D (N = 5), where N = number of preparations.

Figure 2

Ingestive priming induces a PKI insensitive inward current that persists and dissipates in parallel with changes in excitability. (A,B) CBI-2 stimulation increases B48 excitability (top traces in (A), top plot in (B), bottom plot in (B)), and activates an inward current (bottom traces in (A), middle plot in (B), bottom plot in (B)). (C,D) FCAP + CP2 superfusion (1 µM each) increases B48 excitability (top traces in (C), top plot in (D), bottom plot in (D), and activates an inward current (bottom traces in (C), middle plot in (D), bottom plot in (D)). In (A–D) excitability was measured by injecting a 3 s constant current pulse. Plotted data show currents measured during a 2 s step to −30 mV. The dotted red line in (A) and (C) marks zero current. In (B) and (D) data plotted are B48 excitability (top plot), and inward current induced (middle plot). In the bottom plot data shown above were replotted together as a percentage of the maximal response. (E,F) Peptides induce an inward current that is not blocked by PKI. Currents were measured during 2 s steps delivered between −90 and −20 mV in 10 mV increments from a holding potential of −60 mV. Recordings were made before (baseline) and after peptide superfusion (FCAP + CP2 1 µM each) in vehicle loaded neurons (control, black) and in PKI loaded neurons (blue). In (F) total currents are plotted in the top graph and difference (peptide-induced) currents in the bottom graph. Sample sizes: Panel B (N = 5), Panel D (N = 5), and Panel F (N = 4).

Figure 3

The current induced by ingestive priming is similar to a CNG sodium current (see also Fig. S2). (A,B) Peptide-induced currents are not observed in low sodium saline. Currents were measured during 2 s steps delivered between −100 and −20 mV in 20 mV increments from a holding potential of −60 mV. Recordings were made before (baseline) and after peptide superfusion (FCAP + CP2, 1 µM each) in normal ASW (nASW, black) and in a saline with 95% of the sodium removed (low Na, green). In (B) the top graph plots total currents, the bottom graph difference (peptide-induced) currents. (C and D) cAMP induced currents are not observed in low sodium saline. Currents were induced by iontophoresis of cAMP (horizontal black bar) in nASW (baseline, black), in a saline with 95% of the sodium removed (low Na, green), and after returning to nASW (wash, gray). (E–H) cAMP induced currents and changes in membrane potential are not observed in Rp loaded neurons. Recordings were obtained from vehicle-loaded neurons (control, black) and in neurons preloaded with Rp-cAMPS (Rp, red). (E) and (F) show currents induced by cAMP at −60 mV and (G) and (H) show changes in membrane potential at −60 mV. In (H) changes in firing frequency are on the left, and changes in membrane potential after spiking are on the right. Sample sizes: Panels B, D, F and H (N = 5).

Figure 4

(A) Paradigm used in (B–E). Currents were measured during 2 s steps delivered between −80 and −20 mV in 10 mV increments from a holding potential of −60 mV before priming (baseline), immediately after priming (Post CBI-2), and 12 min after priming in neurons loaded with either vehicle (Control) or Rp-cAMPS (Rp). (B–E) Currents in Rp loaded neurons were smaller than currents in control neurons. (F) Paradigm used in (G) and (H). The B48 firing frequency was measured before priming (Rested), immediately after priming (Primed) and 12 min after priming in Control (vehicle loaded) neurons and in neurons loaded with Rp-cAMPS. (G,H) Priming is not observed at 12 min when neurons are loaded with Rp-cAMPS. (G) Current clamp recordings from control (vehicle loaded) and Rp-loaded (Rp) neurons. The bars under the traces indicate radula protraction (black) and radula retraction (white). Sample sizes: Panel C (N = 9), Panel E (N = 6), Panel H (N = 10, unpaired [Control = 5, Rp = 5]).