Neuromodulation and neuroprotective effects of chlorogenic acids in excitatory synapses of mouse hippocampal slices

Abstract

The increased healthspan afforded by coffee intake provides novel opportunities to identify new therapeutic strategies. Caffeine has been proposed to afford benefits through adenosine A_2A receptors, which can control synaptic dysfunction underlying some brain disease. However, decaffeinated coffee and other main components of coffee such as chlorogenic acids, also attenuate brain dysfunction, although it is unknown if they control synaptic function. We now used electrophysiological recordings in mouse hippocampal slices to test if realistic concentrations of chlorogenic acids directly affect synaptic transmission and plasticity. 3-(3,4-dihydroxycinnamoyl)quinic acid (CA, 1-10 μM) and 5-O-(trans-3,4-dihydroxycinnamoyl)-D-quinic acid (NCA, 1-10 μM) were devoid of effect on synaptic transmission, paired-pulse facilitation or long-term potentiation (LTP) and long-term depression (LTD) in Schaffer collaterals-CA1 pyramidal synapses. However, CA and NCA increased the recovery of synaptic transmission upon re-oxygenation following 7 min of oxygen/glucose deprivation, an in vitro ischemia model. Also, CA and NCA attenuated the shift of LTD into LTP observed in hippocampal slices from animals with hippocampal-dependent memory deterioration after exposure to β-amyloid 1-42 (2 nmol, icv), in the context of Alzheimer's disease. These findings show that chlorogenic acids do not directly affect synaptic transmission and plasticity but can indirectly affect other cellular targets to correct synaptic dysfunction. Unraveling the molecular mechanisms of action of chlorogenic acids will allow the design of hitherto unrecognized novel neuroprotective strategies.

Figure 1

The chlorogenic acid 3-(3,4-dihydroxycinnamoyl)quinic acid (CA, 1 or 10 µM) did not affect either basal synaptic transmission or synaptic plasticity. CA (1 µM, light blue symbols or 10 µM, dark blue symbols) did not alter basal synaptic transmission (A), measured as the descending slope of field excitatory postsynaptic potentials (fEPSP) recorded extracellularly in the stratum radiatum of the CA1 area upon stimulation of the afferent Schaffer collaterals, illustrated as inserts in panel A. Likewise, CA (1 or 10 µM) did not modify the input–output responses (B) or the paired-pulse facilitation ratio, i.e. the ratio of the fEPSP slopes evoked by two consecutive pulses (ratio of P2/P1) with inter-pulse intervals of 25 or 50 ms (C). The average time course experiments of LTP induction (D), i.e. the variation of fEPSP slope over time upon induction of LTP with a high-frequency stimulation train (HFS: one train of 100 pulses of delivered at 100 Hz) in hippocampal slices untreated (control, black symbols) or treated with CA (1 or 10 µM), revealed that LTP magnitude was not affected by CA (F). (E) Representative fEPSP recorded in ACSF (control conditions, upper row) and in the presence of 10 µM CA (lower row) before (dashed traces) and 60 min after the HFS (filled traces). Data are mean ± S.E.M. of 6–12 experiments.

Figure 2

The neochlorogenic acid 5-O-(trans-3,4-dihydroxycinnamoyl)-D-quinic acid (NCA, 1 or 10 µM) did not affect either basal synaptic transmission or synaptic plasticity. NCA (1 µM, light green symbols or 10 µM, dark green symbols) did not alter basal synaptic transmission (A), measured as the descending slope of field excitatory postsynaptic potentials (fEPSP) recorded extracellularly in the stratum radiatum of the CA1 area upon stimulation of the afferent Schaffer collaterals, illustrated as inserts in panel A. Likewise, NCA (1 or 10 µM) did not modify the input–output responses (B) or the paired-pulse facilitation ratio, i.e. the ratio of the fEPSP slopes evoked by two consecutive pulses (ratio of P2/P1) with inter-pulse intervals of 25 or 50 ms (C). The average time course experiments of LTP induction (D), i.e. the variation of fEPSP slope over time upon induction of LTP with a high-frequency stimulation train (HFS: one train of 100 pulses delivered at 100 Hz) in hippocampal slices untreated (control, black symbols) or treated with NCA (1 or 10 µM), revealed that LTP magnitude was not affected by NCA (F). (E) Representative fEPSP recorded in ACSF (control conditions, upper row) and in the presence of 10 µM NCA (lower row) before (dashed traces) and 60 min after the HFS (filled traces). Data are mean ± S.E.M. of 4–8 experiments.

Figure 3

The chlorogenic acids 3-(3,4-dihydroxycinnamoyl)quinic acid (CA, 10 µM) and 5-O-(trans-3,4-dihydroxycinnamoyl)-D-quinic acid (NCA, 10 µM) accelerated and increased the extent of recovery of both hippocampal synaptic transmission and synaptic plasticity after exposure to oxygen–glucose deprivation (OGD), modelling ischemia. (A) A period of 7 min of OGD (red bar) caused a profound depression of hippocampal synaptic transmission, measured as the slope of field excitatory postsynaptic potentials (fEPSP) recorded extracellularly in the stratum radiatum of the CA1 area upon stimulation of the afferent Schaffer collaterals, without appreciable modification (< 15% around average) of the presynaptic volley amplitude throughout the whole protocol (data not shown). (A) Upon glucose and oxygen readmission (re-oxygenation), there was a discrete recovery of synaptic transmission in control conditions (no added drugs), whereas this recovery upon re-oxygenation was more robust in the presence of either CA (10 µM; blue symbols) or NCA (10 µM; green symbols), added when indicated by the arrow. (B,C) Representative superimposed fEPSPs before OGD (filled lines), 8 min after OGD (red/orange dashed lines) and upon reoxygenation, 24 min after OGD (dotted lines) in the absence (B) and in the presence of 10 µM NCA control conditions (C). (D) Quantification of fEPSP slopes between 20 and 25 min of re-oxygenation after OGD (D). The average time course experiments of LTP induction (E), i.e. the variation of fEPSP slope over time upon induction of LTP with a high-frequency stimulation train (HFS: one train of 100 pulses delivered at 100 Hz) in hippocampal slices subjected to OGD followed by re-oxygenation recovery revealed that LTP magnitude was discrete in slices not exposed to any drug (control, black symbols) but was more evident in slices treated with CA (10 µM) or NCA (10 µM) during re-oxygenation (H). (F,G) Representative superimposed fEPSPs before (filled lines) and 58 min after HFS (dotted lines) in the absence (F) and in the presence of 10 µM NCA control conditions (G). Data are mean ± S.E.M. of 3–8 experiments. * p < 0.05 (or exact p value) versus control using a Kruskal Wallis test followed by a Dunn’s post hoc multiple comparison test. Please note that the recordings from panels B and F are from the same slice in the absence of added drugs and the recordings from panels B and F are from the same slice in the presence of 10 uM NCA applied from reoxygenation onwards.

Figure 4

Mice injected intracerebroventricularly (icv) with β-amyloid peptide 1–42 (Aβ_1-42) displayed a selective disruption of hippocampal-dependent memory performance, accompanied by a paradoxical alteration of hippocampal synaptic plasticity, which was attenuated by exposure to the chlorogenic acids 3-(3,4-dihydroxycinnamoyl)quinic acid (CA, 10 µM) and 5-O-(trans-3,4-dihydroxycinnamoyl)-D-quinic acid (NCA, 10 µM). Fifteen days after an icv challenge with Aβ_1-42 (2 nmol), Aβ-treated mice (red symbols) displayed no alteration of spontaneous locomotion (A) assessed by the number of crossings in an open field test or anxiety, assessed as the time in the more aversive central area of the open field (B), compared with control (vehicle-treated mice; black symbols). In contrast, Aβ-treated mice displayed hippocampal-dependent memory deficits, as assessed in the object displacement test (C) and two-visits Y-maze test (D). Slices from vehicle-treated mice displayed a pattern of long-term depression (LTD) after repeated low-frequency stimulation (LFS), which was not affected by CA (10 µM; blue symbols) (E–F) or NCA (10 µM; green symbols) (G,H). Slices from Aβ-treated mice displayed a shift from LTD to LTP after repeated low-frequency stimulation (LFS), which was attenuated by CA (10 µM) (E–F) or NCA (10 µM) (G,H). Data are mean ± S.E.M. of 4–6 experiments. * p < 0.05 versus control using a Mann Whitney test (C,D) or a Kruskal Wallis test followed by a Dunn’s post hoc multiple comparison test (F,H).