Naturally produced and synthesized opioid receptor agonists are often used as anesthetics and analgesics in clinical settings. Some patients who received regional or general anesthesia for surgery experience a short-term decline after surgery in cognitive function such as learning and memory; this is known as postoperative cognitive disorder (POCD; Moller et al., 1998; Rasmussen, 2006; Newman et al., 2007). Increasing evidence suggests that aging, duration of anesthesia, and postoperative infections rather than cerebral hypoxia are associated with POCD (Moller et al., 1998; Rasmussen, 2006; Newfield, 2009; Stenvall et al., 2012). However, the mechanisms that underlie POCD development remain largely unknown.

The hippocampus is a brain structure critical for learning and memory. Long-term potentiation (LTP) and long-term depression (LTD) of synaptic strength at the glutamatergic synapses in the hippocampus are well-characterized forms of synaptic plasticity and have long been believed to be a cellular mechanism underlying learning and memory (Bliss and Collingridge, 1993; Malenka and Nicoll, 1999). In rats, electrophysiological, immunohistochemical, and in situ hybridization experiments have demonstrated that the hippocampus expresses a high density of opioid receptors (Zieglgänsberger et al., 1979; Lupica et al., 1992; Drake and Milner, 2002; Stumm et al., 2004; Drake et al., 2007). Along with these studies, μ-opioid receptors (MOR) are selectively expressed on GABAergic interneurons within the CA1 region of the hippocampus (Drake and Milner, 1999; Stumm et al., 2004). Because of this interneuron expression profile of MOR, acute activation of MOR in the hippocampal CA1 area increases glutamatergic excitatory neuronal transmission by hyperpolarizing interneurons and inhibiting GABA release (Capogna et al., 1993; Svoboda and Lupica, 1998). Therefore, the enhanced excitatory neuronal transmission in the hippocampal CA1 region after opioid exposure is considered a secondary effect due to its disinhibiting effect. While activation of opioid receptors in the hippocampus by acute and chronic opioid exposure have demonstrated modulation of hippocampal plasticity and memory (Yang et al., 2004; Jafari-Sabet and Jannat-Dastjerdi, 2009), the role of acute MOR activation and washout in hippocampal synaptic plasticity of the Schaffer-collateral CA1 pathway remains elusive.

Fentanyl, an agonist of MOR largely used in human patients for anesthesia and analgesia associated with surgery, has been demonstrated to decrease GABA-mediated synaptic transmission from interneurons of the CA1 region 24 h after a single dose in vivo treatment (Kouvaras et al., 2008), which led to an increased susceptibility of pyramidal neurons to presynaptic stimulation. In in vitro cultured hippocampal neurons, synaptic α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid glutamate receptors (AMPAR) and dendritic spins were reported to be changed within 24 h of fentanyl exposure (Lin et al., 2009). In the peripheral C-fiber synapses, acute fentanyl exposure decreased presynaptic neurotransmission, but persistently potentiated neurotransmitter release from presynaptic sites when withdrawn from the acute treatment (Heinl et al., 2011). However, whether fentanyl acute exposure and washout directly affects glutamatergic synaptic function in the Schaffer-collateral CA1 pathway remains unknown.

In the present study, we exposed hippocampal slices to fentanyl and recorded its effects on glutamatergic synaptic transmission and plasticity in the presence or blockade of GABAergic inhibition. We aimed to see if fentanyl, similar to morphine, can modulate glutamatergic synaptic plasticity in the Schaffer-collateral pathway, and particularly to examine whether inhibitory neurotransmission is involved.

Many opioids such as morphine modulate basic synaptic transmission and plasticity in the hippocampal CA1 region, via activating several types of opioid receptor (μ-, σ-, κ-opioid receptors; Velísek et al., 2000; Yang et al., 2003, 2004; Dong et al., 2007, 2010; Nugent and Kauer, 2008; Hosseinmardi et al., 2009; Dacher and Nugent, 2011). In the present study, we report here that selective activation of MOR by in vitro fentanyl acute exposure and washout to hippocampal slices increased basic synaptic transmission in the Schaffer-collateral pathway, through its effect on disinhibition of GABAergic interneurons. More interestingly, fentanyl treatment shifted long-term plasticity of hippocampal synapses in favor of LTD induction, and the effect of fentanyl on synaptic plasticity was different from its effect on basic synaptic transmission, although both required MOR activation.

In the central and peripheral nervous systems, acute exposure to opioids activated G-protein coupled receptors, promoted potassium efflux, and inhibited calcium influx, which has been shown to generally inhibit neurotransmitter release and result in neuronal activity (Al-Hasani and Bruchas, 2011). However, opioid receptors in the hippocampus are mainly expressed in interneurons (Drake and Milner, 1999; Stumm et al., 2004), and activation of opioid receptors in interneurons reduces GABA release that normally carries inhibitory input to pyramidal neurons. Therefore, in the hippocampus, opioid exposure leads to an increased glutamatergic excitatory neuronal transmission by hyperpolarizing interneurons and GABA release inhibition (Capogna et al., 1993; Svoboda and Lupica, 1998). We should note that the effect of in vitro acute exposure of hippocampal tissue to opioids on CA1 synaptic transmission remains controversial. An earlier study demonstrated that acute in vitro application of opioid receptor agonists enhanced synaptic transmission in the CA1 area of the hippocampus (Dingledine, 1981); recently, MOR activation by opioids was also reported to increase hippocampal network activity recorded in the CA1 region (Giannopoulos and Papatheodoropoulos, 2013). However, it is also characterized by lack of effect on glutamatergic synaptic transmission by acute in vitro exposure of organotypic rat hippocampal slice cultures to MOR-preferring agonist D-Ala2, MePhe4, Met(O)5-ol-enkephalin (FK 33-824) in the CA3 region of the hippocampus (Capogna et al., 1993). In the present study, our results indicated that acute fentanyl exposure to hippocampal slices increased the amplitude of fEPSPs in Schaffer-collateral CA1 synapses, and this effect was prevented by inactivation of MOR or removal of GABA-mediated inhibition. Therefore, our findings support that in vitro acute opioid exposure can lead to an increase in CA1 basic synaptic transmission, which is indirect from the disinhibition mediated by GABA-containing interneurons. In addition, we observed that the fentanyl-increased fEPSPs in vitro were decreased to baseline levels within 1 h after drug washout, which was also different from the in vivo acute treatment in which the facilitation lasted for several hours (Yang et al., 2004). This difference might derive from the activation of MOR in other brain regions following in vivo injection, rather than limited to hippocampus in in vitro treatment.

While the in vitro acute effect of fentanyl on hippocampal synaptic plasticity has not been fully investigated, it is known that in vivo acute opioid administration facilitated hippocampal LTD in the Schaffer-collateral CA1 pathway (Yang et al., 2004). In addition, opioid exposure during the prenatal period was reported to impair hippocampal LTP and facilitate LTD induction in Schaffer-collateral synapses (Velísek et al., 2000; Yang et al., 2003). Similar to acute in vivo administration, chronic exposure to opioids has been shown to reduce LTP in the CA1 area of the hippocampus in both in vivo (Pu et al., 2002) and in vitro experiments (Salmanzadeh et al., 2003). Our results in the present study revealed that acute fentanyl exposure and washout suppressed hippocampal LTP but facilitated LTD expression in the Schaffer-collateral pathway.

The significant finding of the present study was that activation of MOR by fentanyl might target distinct mechanisms for modulating basic synaptic transmission and activity-dependent synaptic plasticity. For the increase of basic excitatory synaptic transmission, the fentanyl-induced MOR activation in interneurons caused disinhibition of excitatory synaptic connection and transmission, which has been well described in the literature (Al-Hasani and Bruchas, 2011). However, our findings of the effect of fentanyl on synaptic plasticity seemed independent from GABA-mediated inhibitory neurotransmission within the CA1 region, despite the fact that activation of MOR was still required. In addition to the MOR-activation coupled inhibitory neurotransmission, newer findings have suggested that MOR activation also modulated other MAPK signal transduction pathways, including P38 (Tan et al., 2009), JNK (Macey et al., 2006), and ERK (Tan et al., 2009). Therefore, the alterations of these signaling transduction pathways following the MOR activation by fentanyl application might play a critical role in activity-dependent synaptic plasticity. In addition, activation of MOR in astrocytes has been reported (Miyatake et al., 2009). In the bath application of fentanyl in our present study, MOR activation might have also induced MAPK signaling transduction change among astrocytes, which could affect local environment stress and inflammation that eventually shaped the synaptic plasticity on the excitatory neurotransmission connections.

In addition, the bath application of fentanyl would globally activate MOR, including the interneurons within the CA1 region. Given that MORs are expressed in interneurons of other areas of the hippocampus such as in the CA3 region and dentate gyrus, the in vitro acute fentanyl exposure and washout to hippocampal slices could also activated MOR in those regions, which may affect Schaffer-collateral excitatory synaptic plasticity. Although our present study did not dissect a detailed pathway, it might provide a starting point for further investigations into the underlying mechanism of synaptic plasticity modulation induced by acute exposure and washout of MOR agonists.

Together our results indicate that the basic synaptic transmission in Schaffer-collateral CA1 synapses was increased by in vitro acute fentanyl exposure and washout, which is secondary to MOR activation-mediated GABA disinhibition among interneurons. However, in vitro acute fentanyl exposure and washout also shifted synaptic plasticity in favor of synaptic depression using a different mechanism. Understanding the underlying mechanism of fentanyl-mediated synaptic plasticity alterations might provide some clues for cognitive impairment after clinical fentanyl use in patients.

HT designed and performed experiments. YX and FL performed experiments and statistical analysis. GW arranged and supervised the drug purchasing and usage. HT and SH wrote the paper.