(A) In normal ASCF, pairing at 0 mV (arrow) induces homosynaptic LTP (top graph), while pairing at -40 mV induces homosynaptic LTD (bottom graph). (B). Bath application of the β-adrenergic agonist Isoproterenol (Iso: 10 μM, gray box) barely affects LTP (top), but suppresses LTD (bottom). (C) Bath application of the α-adrenergic agonist Methoxamine (Mtx: 5 μM, gray box) suppresses LTP (top), but not LTD (bottom). (D) Co-application of Iso+Mtx restores bidirectional plasticity. Filled circles in A-D: paired input; open circles: unpaired input. Note that neuromodulators without paired stimulation induce only transient changes in the responses (open circles in B-D). Traces are averages of ten EPSPs recorded before (thin line) and 40 min after (thick line) induction of plasticity. Scale-bars: 2.5 mV, 5 msec. Note the absence of changes in the normalized paired-pulse ratio (PPR) and membrane time constant (τ=RC), which are displayed below each group plot. Here and in subsequent figures data is presented as average ± SEM. The number of experiments is indicated in parentheses.

(A-D) Priming the adrenergic suppression of LTP and LTD . (A) Experimental temporal scheme: the seal was made and ruptured after the 10 min agonist application (10μM Iso, or 5 μM Metx). The time course of the changes in NMDAR-mediated responses are depicted below (thin line). (B-D) Pairing at 0 and -40 mV induced respectively LTP (filled circles) and LTD (open circles) in control conditions (B), but no LTD when the cells were pretreated with isoproterenol (C), and no LTP after pretreatment with methoxamine (D). (E-F) The priming of the adrenergic suppression does not require co-activation of NMDA. Bath application of APV (50μM) at the time of the agonist application (E) does not affect the induction of LTP and LTD (F), nor it affects the suppression of LTD by isoproterenol (G) or LTP by nethoxamine (H). The number of experiments is indicated in parentheses. (I-K) the duration of the priming depends on agonist exposure. (I) LTD magnitude (measured 60 min after pairing) induced at different times after a 10 min isoproterenol exposure. Open circles: individual experiments; filled circles: averages for the time intervals (in min) 21<t<40, 41<t<60, 60<t100. (J-L) One-hour agonist exposure causes a long-lasting suppression of LTD and LTD. (J) Experimental diagram. (K, L) Individual experiments showing the magnitude of LTP (open triangles) and LTD (inverted triangles) induced after prolonged wash out of isoproterenol (K) or methoxamine (L). Filled triangles: experiments in which methoxamine or Isopropterenol were respectively applied (10 min) right before induction of LTD (K) or LTD (L).

In all of the experiments, plasticity was evaluated in two independent pathways converging onto the same postsynaptic cell. (A) In normal ACSF, inputs recently depressed by pairing at -40 mV (downward arrow at -20 min) can be de-depressed by subsequent pairing at 0 mV (upward arrow at 0 min. de-depressed inputs: filled circles; control inputs: open circles). (B) Bath application of methoxamine (MTX: 5 μM grey box) does not interfere with de-depression (de-depressed inputs: filled circles; control inputs: open circles). (C) In normal ACSF pairing at 0 mV (upward arrow at 0 min) induced both de-depression in previously depressed inputs (filled circles), and LTP in naïve inputs (open circles). (D) Bath application of methoxamine suppresses LTP (open circles) but not de-depression (filled circles). (E-F) Pairing at -40 mV (downward arrow at 0 min: filled circles) de-potentiates previously potentiated inputs (filled circles) in either normal ACSF (E) or after bath applied isoprorenenol (F, Iso: 10 mM, grey box). LTP was induced by pairing at 0 mV (upward arrow at -20 min). Control LTP: open circles, Depotentiation: closed circles. (G,H) Pairing at -40 mV (downward arrow at 0 min) induced both depotentiation in previously potentiated inputs (filled circles), and LTD in naïve inputs (open circles) in normal ACSF (G), but only induces depotentiation after bath application of isoproterenol. The number of experiments is indicated in parentheses.

(A) Experimental diagram. Prior to recording, rats were injected with isoprorenol or methoxamine (15 mg/kg. i.p.) and subjected to 1 hour of monocular stimulation with drifting gratings (see methods). (B) Example mEPSCs from pyramidal neurons located in contralateral (contra) and ipsilateral (ipsi) hemispheres of a Mtx-treated animal. (C-E) Results obtained in animal treated with methoxamine (C), isoproterenol (D) or control vehicle solution (E). In each case the left panels show the average mEPSCs recorded from contralateral (stimulated, thick colored line) and ipsilateral (non-stimulated, thin black line) hemispheres superimposed with a scaled mEPSC (dashed line); the center panels show the average mEPSC amplitudes from contralateral (black filled dots) and ipsilateral (colored filled dots) hemispheres obtained from each rat (data connected by a straight line); the right panels show the cumulative distribution of mEPSC amplitude for contralateral (thick colored line) and ipsilateral (thin line) hemispheres and scaled distribution (dashed line). The inset shows mEPSC distributions. (F) CPP injection blocks the changes in average mEPSC amplitude in the stimulated cortex induced by methoxamibne (left) or Isoproterenol (right). (G-H) summary plots for average mEPSC amplitude (G) and frequency (H).

(A, B) Inclusion of 10μM propanolol in the bath blocked the induction of LTP with 0 mV pairing (A) and allowed the induction of LTD with -20 mV pairing (B). Filled circles: in propanolol; open circles: control in ACSF. (C) mEPSC recorded ex vivo from monocularly stimulated rats injected with propanalol (10 mg/Kg) and maprotiline (10mg/Kg) and stimulated monocularly as described in Figures 7 and 8. The left panel show the average mEPSCs recorded from contralateral (stimulated, thick orange line) and ipsilateral (non-stimulated, thin black line) hemispheres superimposed with a scaled mEPSC (dashed line); the center panel show the average mEPSC amplitudes from contralateral (black dots) and ipsilateral (orange dots) hemispheres obtained from each rat (data connected by a straight line); the right panels shows the cumulative distribution of mEPSC amplitude for contralateral (orange line) and ipsilateral (thin line) hemispheres and scaled distribution (dashed line). The inset shows mEPSC distributions.

(A) Pairing at -20 mV induces no net change in EPSP slope in normal ACSF (open circles), but induces LTP in the presence of isoproterenol (10 μM Iso: black circles) or LTD in the presence of methoxamine (5 μM Mtx: gray circles. (B) Pairing at different voltages induces LTD and LTP (measured 30-40 min after pairing) in control ASCF (open circles), but only LTP after bath applied isoproterenol (black circles), and only LTD after applied methoxamine (gray circles). (C) Pairing at -10 mV (open triangles) and -30 mV (open circles) induce little change in normal ACSF, but induce LTP and LTD after co-application of 10 μM isoproterenol and 5 μM methoxamine (black triangles and circles). (D) Co-application of isoproterenol and methoxamine enhances bi-directional changes (Iso+Mtx: black symbols; control: open symbols). (E-F) LTP-only plasticity also results after activation of the Gs-coupled PGE2 receptor with Butaprost (10 μM: black circles), whereas LTD-only plasticity results after bath application of the Gq11-coupled M1 receptor agonist McN (3 μM: grey circles; 10μM: open circles). The number of experiments is indicated in parentheses.

Bath applied Methoxamine (Mtx: 5 μM, gray box in B and D) suppresses LTP in S831A phosphomutants and their wild-type littermates (A) Wild-type mice express LTP (filled circles) and LTD (open circles) in control conditions, and suppression of LTP by methoxamine (B). (C) S831A mice express LTP (filled circles) but no LTD (open circles), yet methoxamine still suppresses LTP (D). Note that the transient depression of glutamatergic responses by methoxamine was virtually absent in S831A phosphomutant mice. The number of experiments is indicated in parentheses.

(A-C) White matter to layer IV pathway expresses LTP and LTD (A), as well as suppression of LTD by bath-applied isoproterenol (B), and suppression of LTP by methoxamine (C). LTP (filled circles) and LTD (open circles) were simultaneously induced in two independent pathways in the same cells using a coordinated pairing (Suppl. Figure 1). (D, E) In the Schaffer collateral input to CA1, the induction LTD of the fEPSP with LFS (1Hz, 15 min, down arrow) is reduced by pre-exposure to 10 μM isoproterenol (D), whereas the induction of LTP with TBS (arrow) is reduced by pre-exposure to 5 μM methoxamine (E). Filled circle: exposed slices; open circles interleaved controls. The grey bar in A-E depicts the agonist application duration. (F, G) The extent of the suppression depends on the agonist incubation. Experimental scheme is indicated on the top of the graphs. Left: Agonist was applied for 15 - 60 min and washed for 1 hour before testing. Right: Agonist was applied for 1 hour and washed out for 2 hours. (F) Average LTD (60 min.post-LFS) induced after isoproterenol incubation. (G) Average LTP (60 min. post-TBS) induced after methoxamine incubation. Filled circles: exposed slices; open circles interleaved controls. The number of experiments is indicated in parentheses.

(A) Experimental design. Prior to recording LTP and LTD, rats were injected with isoprorenol or methoxamine (15 mg/kg i.p.) and subjected to 1 hour of monocular stimulation with drifting gratings (see methods). (B) Normal pairing induced LTP (filled circles) and LTD (open circles) in both hemispheres (ipsi: left; contra: right) in control rats, stimulated but injected with vehicle solution. (C) In methoxamine-injected rats no LTP was induced in the ipsilateral cortex (non-stimulated), and no LTD was induced in the contralateral (stimulated) cortex. (D) In isoproterenol-injected rats, no LTD was induced in the ipsilateral cortex, and no LTP was induced in the contralateral cortex.