Noradrenergic Suppression of Persistent Firing in Hippocampal CA1 Pyramidal Cells Through Camp-PKA Pathway

Home , Himbacine

Research Article: New Research | Neuronal Excitability Noradrenergic suppression of persistent firing in hippocampal CA1 pyramidal cells through cAMP-PKA pathway https://doi.org/10.1523/ENEURO.0440-20.2020

Cite as: eNeuro 2021; 10.1523/ENEURO.0440-20.2020 Received: 12 October 2020 Revised: 3 December 2020 Accepted: 20 December 2020

This Early Release article has been peer-reviewed and accepted, but has not been through the composition and copyediting processes. The final version may differ slightly in style or formatting and will contain links to any extended data.

Alerts: Sign up at www.eneuro.org/alerts to receive customized email alerts when the fully formatted version of this article is published.

Copyright © 2021 Valero-Aracama et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

1 Noradrenergic suppression of persistent firing in hippocampal CA1 2 pyramidal cells through cAMP-PKA pathway 3 4 Abbreviated Title: Noradrenaline suppress persistent firing 5 6 1Ɨ 2,3Ɨ 2 3 7 Maria Jesus Valero-Aracama , Antonio Reboreda , Alberto Arboit , Magdalena Sauvage , Motoharu 2,3 * 8 Yoshida 9 1 10 Faculty of Psychology, Mercator Research Group - Structure of Memory, Ruhr-University Bochum, 11 44780 Bochum, Germany 12 Current affiliation: Institute for physiology and pathophysiology, University of Erlangen, Germany 13 2 14 German Center for Neurodegenerative Diseases (DZNE), 39120 Magdeburg, Germany 15 3 16 Leibniz Institute for Neurobiology (LIN) 39118 Magdeburg, Germany 17 Ɨ Equal contributions ∗ 18 To whom correspondence should be addressed 19

20 Author contributions:

21 MY, AR, MJ and MS designed research; MJ, AR and AA performed research and analyzed data; 22 MJ, AR and MY wrote the paper.

24 Correspondence should be addressed to:

25 Dr Motoharu Yoshida 26 German Center for Neurodegenerative Diseases (DZNE) Magdeburg 27 Leipziger Str. 44 / Haus 64, 39120, Magdeburg, Germany 28 E-mail: [email protected] 29

30 Number of figures: 5 31 Number of tables: 2 32 Number of Multimedia: 0 33 Number of words for Abstract: 164 34 Number of words for Significance Statement: 35 Number of words for Introduction: 418 36 Number of words for Discussion: 1740 37

38 Acknowledgments: We thank Dr. Menno Witter for kindly sharing experimental procedures with us.

39 Conflict of Interest: Authors report no conflict of interest

40 Funding sources: This work was supported by the Mercator Stiftung and the German Research 41 Foundation (DFG) project YO177/4-1 and YO177/7-1. 42

43 Abstract

44 Persistent firing is believed to be a cellular correlate of working memory. While the effects of

45 noradrenaline (NA) on working memory have widely been described, its effect on the cellular

46 mechanisms of persistent firing remains largely unknown. Using in vitro intracellular recordings, we

47 demonstrate that persistent ﬁring is supported by individual neurons in hippocampal CA1 pyramidal

48 cells through cholinergic receptor activation, but is dramatically attenuated by NA. In contrast to the

49 classical theory that recurrent synaptic excitation supports persistent firing, suppression of persistent

50 ﬁring by NA was independent of synaptic transmission, indicating that the mechanism is intrinsic to

51 individual cells. In agreement with detrimental effects of cyclic adenosine monophosphate (cAMP) on

52 working memory, we demonstrate that the suppressive effect of NA was through cAMP-PKA pathway.

53 In addition, activation of β1 and/or β3 adrenergic receptors, which increases cAMP levels,

54 suppressed persistent firing. These results are in line with working memory decline observed during

55 high levels of NA and cAMP, which are implicated in high stress, aging and schizophrenia.

57 Significance Statement

58 While cholinergic modulation supports working memory, high concentrations of noradrenaline (NA), 59 which occurs under high stress for example, are detrimental for working memory. However, cellular 60 and molecular mechanisms underlying such working memory deficit remain largely unclear. In this 61 paper, we studied the effect of these two neuromodulators on persistent firing, the cellular correlate 62 of working memory. We demonstrate that a cholinergic receptor activation supports, while a 63 noradrenergic activation strongly inhibits persistent firing due to the PKA activation through specific 64 NA receptor types which upregulate cAMP. These data are in line with working memory deficits in 65 aging and schizophrenia in which cAMP levels are altered, and indicate potential intrinsic cellular 66 mechanism of working memory impairment.

68 Introduction

69 Persistent firing is a cellular response characterized by repetitive spiking that outlasts triggering 70 stimulus, and is believed to be a neural base of working memory (Goldman-Rakic, 1995; Major & 71 Tank, 2004). Persistent firing observed in humans and animals in vivo during working memory tasks 72 correlate with the task performance (Bukalo et al., 2004; Colombo & Gross, 1994; Funahashi, Bruce, 73 & Goldman-Rakic, 1989; Hampson & Deadwyler, 2003; Kamiński et al., 2017). In contrast to the 74 classical hypothesis that persistent firing is supported by recurrent excitatory synaptic connections, 75 recent studies have shown that persistent firing can also be supported in individual cells (Egorov, 76 Hamam, Fransén, Hasselmo, & Alonso, 2002; Jochems & Yoshida, 2013; Knauer, Jochems, Valero- 77 Aracama, & Yoshida, 2013). These studies in general used cholinergic receptor activations to induce 78 persistent firing, which is in agreement with the supportive role of acetylcholine on working memory 79 (Reboreda et al., 2018). However, roles of other neuromodulators such as the noradrenaline (NA) on 80 this type of persistent firing largely remain to be studied.

81 While cholinergic modulation generally supports working memory (Kaneko & Thompson, 82 1997; Weiss et al., 2000), NA could be detrimental for working memory (Arnsten, 2009; Roozendaal 83 & McGaugh, 2011). High levels of NA are usually present during states of stress (Bremner, Krystal, 84 Southwick, & Charney, 1996), which are known to impair working memory (Arnsten, Wang, & 85 Paspalas, 2012; Diamond, Ingersoll, Fleshner, & Rose, 1996; Qin, Hermans, van Marle, Luo, & 86 Fernández, 2009; Schoofs, Preuß, & Wolf, 2008). At the receptor subtype levels, β1 and α2 87 adrenoreceptors, which are positively and negatively coupled to cAMP signaling, impair and facilitate 88 working memory, respectively (Arnsten & Li, 2005; Jäkälä et al., 1999; Li, Mao, Wang, & Mei, 1999;

89 Ma, Qi, Peng, & Li, 2003; Ramos et al., 2005). In addition, elevated and decreased cAMP levels impair 90 and enhance, respectively, working memory performance (Arnsten, Mathew, Ubriani, Taylor, & Li, 91 1999; Runyan, Moore, & Dash, 2005; Taylor, Birnbaum, Ubriani, & Arnsten, 1999; M. Wang et al., 92 2007). Moreover, upregulated cAMP levels are associated with working memory impairment in aging 93 and schizophrenia (Barch & Ceaser, n.d.; Millar, 2005; M. Wang et al., 2011). Raising cAMP levels 94 would increase PKA activation that also is negatively correlated to working memory performance 95 (Kobori, Moore, & Dash, 2015), together supporting the importance of NA and cAMP-PKA in working 96 memory. 97 At the cellular level, a supportive role of NA on persistent firing through a reduction of cAMP 98 levels has been observed in the prefrontal cortex (PFC, Zizhen Zhang, Cordeiro Matos, Jego, 99 Adamantidis, & Séguéla, 2013). In addition, elevated cAMP levels caused a suppression of the 100 calcium-activated nonspecific cation (CAN) current and persistent firing in the PFC and hippocampus 101 (El-Hassar, Hagenston, D’Angelo, & Yeckel, 2011; Sidiropoulou et al., 2009; Zizhen Zhang et al., 2013). 102 However, it remains unclear whether NA has a supportive or detrimental effect on persistent firing in 103 the hippocampus. 104 We therefore tested the effects of cholinergic and noradrenergic receptor activation in both 105 induction and modulation of persistent firing in individual hippocampal CA1 pyramidal cells in mice. 106 We find that a noradrenergic stimulation strongly inhibits persistent firing while a cholinergic 107 activation supports it. In addition, the noradrenergic suppression of persistent firing is independent 108 of ionotropic synaptic transmission, but is through the cAMP-PKA pathway. We further demonstrate 109 specific NA receptor subtypes involved in this suppression. These observations indicate that high NA 110 and cAMP conditions, which often suppress working memory through PKA activation, also suppress 111 cellular mechanisms for persistent firing in the hippocampus. 112

113 Methods

114 115 All the experimental designs were approved by the local ethic committee (Der Tierschutzbeauftragte, 116 Ruhr-Universität Bochum, and Deutsches Zentrum für Neurodegenerative Erkrankungen) and carried 117 out in accordance with the European Communities Council Directive of September 22nd, 2010 118 (2010/63/EU). 119 120 Slice preparation 121 Acute hippocampal slices were obtained from adult (2 - 4 month) C57BL6 female mice (Charles 122 River). Animals were deeply anesthetized (intraperitoneal injection of 120 mg/Kg of ketamine and 16 123 mg/Kg of xylazin) and intracardiac perfusion with ice-cold modified artificial cerebrospinal fluid

124 (mACSF) was conducted directly without or after cervical dislocation. The mACSF contained (in mM) 125 87 NaCl, 25 NaHCO3, 10 Glucose, 75 Sucrose, 2.5 KCl, 1.25 NaH2PO4, 0.375 CaCl2, 3.28 MgCl2, 3 126 Pyruvix acid, 1 Ascorbic acid, or 87 NaCl, 25 NaHCO3, 10 Glucose, 75 Sucrose, 1.25 KCl, 1.25 127 NaH2PO4, 0.188 CaCl2, 1.64 MgCl2, 3 Pyruvix acid, 1 Ascorbic acid (pH was adjusted to 7.4 by 128 saturation with 95% O2 - 5% CO2). Following the decapitation, horizontal hippocampal slices (350 129 μM) were obtained with a vibrating-blade microtome (Leica VT1000S) in ice-cold mACSF. Brain slices 130 were then incubated for 30 minutes at 37°C in normal ACSF (nACSF) containing (in mM) 124 NaCl, 131 1.25 NaH2PO4, 1.8 MgO4S, 3 KCL, 10 Glucose, 26 NaHCO3, 1.2 CaCl2, or 124 NaCl, 1.25 NaH2PO4, 132 1.8 MgO4S, 1.5 KCL, 10 Glucose, 26 NaHCO3, 0.6 CaCl2 (pH was adjusted to 7.4 by saturation with 133 95% O2 - 5% CO2). The second nACSF with 1.5 mM KCL was used in an attempt to reduce synaptic 134 effect but was used only in seven cells in Figure 5C and D. In all other 105 cells, the first nACSF with 3 135 mM KCL was used. Brain slices were maintained at room temperature for at least 30 minutes before 136 recording. 137 138 Whole-cell recordings 139 Whole-cell procedures were performed as described previously (Valero-Aracama et al., 2015). Briefly, 140 CA1 cells were visually identified using an upright microscope (Zeiss Axioskop 2FS), and whole-cell 141 patch was obtained using glass pipettes filled with an intra-cellular solution containing (in mM) 120 142 Kgluc, 10 HEPES, 0.2 EGTA, 20 KCl, 2 MgCl2, 7 PhCreat di(tris), 4 Na2ATP and 0.3 Tris GTP (pH 143 adjusted to 7.3 with KOH). Electrical signals were amplified with an Axoclamp 2A amplifier (Axon 144 Instruments), low-pass filtered at 10 kHz and sampled at 20 kHz, using WinWCP (John Dempster, 145 University of Strathclyde) software. Liquid junction potential (~10 mV) was not corrected. To prevent 146 pharmachological contamination, only one cell per slice was recorded. Recordings were performed 147 at 34 ± 1 ºC. 148 149 Drugs 150 All drugs except for Carbachol (Cch; Alfa Aesar, Ward Hill, USA) were purchased from Sigma-Aldrich 151 (St. Louis, USA). Water-based stock solutions were made for all drugs except picrotoxin and 152 kynurenic acid (freshly dissolved in nACSF). Except for the Cch stock solution, which was stored at 4 153 °C, all other solutions were aliquoted and stored at -20 °C. Stock solutions were diluted at least 1000 154 times in the final solution. Recording solutions with carbachol were freshly made each day of 155 experiment and solutions containing other drugs were freshly made from stock solutions minutes 156 before being used. Recordings started at least 5 minutes after the application of the experimental 157 drug. 158

159 Data analysis 160 Analysis was conducted with Matlab R2009b (Math Works) and Anaconda (Python 3.6). Only 161 pyramidal cells with a silent resting membrane potential lower than -55 mV, and spike amplitudes 162 that overshoot 0 mV were included into the analysis. Persistent firing was induced by a current 163 injection (2s, 100 pA) from a membrane potential below spiking threshold. The membrane potential 164 and frequency of persistent firing were measured during the first 10s after the termination of the 165 stimulation. General excitability was measured from the current-frequency curve (50-350 pA, 50 pA 166 increment, 1s) starting from the fix potential of -65 mV. Input resistance (IR) was computed from the 167 voltage drop peak in response to a 50 pA negative current injection from -65 mV. Sag ratio was 168 computed from voltage response to a negative current step (1s, -300 pA). Medium and slow after- 169 hyperopolarizing potential (AHP) were measured after the induction of 6 spikes (0.1 to 1 nA, 3 ms, 50 170 Hz) from a membrane potential of -60 mV. In cases the AHP was positive compared to the baseline 171 (as in conditions with Cch) this property is referred to as ADP. Comparisons were made using paired 172 t-tests. Significance level α < 0.05 (*P < 0.05, **P < 0.01, ***P < 0.001 was used. Data is expressed as 173 means ± SEM. 174

175 Results

176 Carbachol supports persistent firing through Gq/11 and Gi/o G-protein coupled receptors

177 First, we tested the effects of cholinergic stimulation alone on the generation of persistent firing in 178 CA1 pyramidal cells. Once a whole-cell recording was obtained, induction of persistent firing was 179 tested by using a brief (2 s, 100 pA) current injection at a membrane potential right below the spike 180 threshold. In the nACSF, none of the cells tested continued spiking after the termination of the 181 stimulation (n = 34; Fig. 1A left). However, in the presence of carbachol (Cch; 5 µM), persistent firing 182 was induced in 62 % (21 out of 34) of the same set of cells (Fig.1A right). This is in agreement with 183 previous reports of similar persistent firing in rat and mouse CA1 pyramidal cells in Cch (Arboit, 184 Reboreda, & Yoshida, 2020; Knauer et al., 2013). For the quantification of persistent firing, spike 185 frequency and membrane potential after the stimulation were measured: post-stimulus frequency 186 and post-stimulus potential. The gray line below each voltage trace indicates the period during which 187 these measurements were taken. Cch significantly increased the post-stimulus frequency (Figure 1B 188 left; n = 34; T(33) = -5.65, p < 0.001) and the post-stimulus membrane potential (Figure 1B right; n = 189 34; T(33) = -6.67, p < 0.001).

190 Cch activates both Gq/11 and Gi/o G-protein coupled receptors (GPCRs) through M1 and M2/4 191 receptor subtypes, respectively. To study the role of each receptor subtype on persistent firing, we

192 selectively blocked M1 or M2/4 receptors. Bath application of the M1 antagonist pirenzepine (10 193 µM), in addition to Cch, resulted in a complete blockade of persistent firing (Fig. 1 C and D). The post- 194 stimulus frequency (Fig. 1D left; n = 5; T(4)=5.94, p<0.01) and depolarization (Fig 1D right; n = 5; 195 T(4)=7.10, p<0.01) were significantly reduced. Similarly to this, M2/4 antagonist himbacine (1 µM) 196 suppressed persistent firing significantly (Fig, 1E) reducing both the frequency (Fig. 1F left; n = 6; 197 T(5)=6.59, p<0.01) and the depolarization (Fig. 1F right; n = 6; T(5)=7.83, p<0.001). These results

198 point out that both Gq/11 and Gi/o coupled receptors are necessary to trigger persistent firing. 199 Furthermore, a control experiment was conducted to exclude the possibility that the suppressive 200 effect of pirenzepine and himbacine was due to a run-down of cellular mechanisms supporting 201 persistent firing during the whole-cell recording. In this experiment, persistent firing recorded at two 202 different time points after the break-in was compared in the presence of Cch. The first and second 203 time points were comparable to the times at which persistent firing was tested in Cch alone and Cch 204 + pirenzepine (or himbacine), respectively, in the above experiment. Persistent firing was not 205 reduced, as shown in Figure 1G and H (Frequency: n = 8, T(7) = 0.388, p = 0.709; Potential: n = 8, T(7) 206 = 0.730, p = 0.489), indicating that the suppression of persistent firing was not an artifact of our 207 experimental procedure.

208

209 NA does not support persistent firing

210 We next tested the effects of NA (5 – 10 µM; n = 14) on the induction of persistent firing. Persistent 211 firing was not observed in any of the cells recorded and the post-stimulus firing frequency was 212 unchanged (Fig 2A). While the post-stimulus membrane hyperpolarization (AHP) was reduced by NA 213 (Fig 2C, n = 14; T(13)=-3.068, p=0.009), the membrane potential did not reach levels above the 214 baseline (averaged voltage before the stimulation) unlike carbachol.

215 It has been reported that intrinsic excitability of CA1 pyramidal cells is increased by both 216 cholinergic and noradrenergic modulation (Benardo & Prince, 1982; Madison & Nicoll, 1982). In 217 agreement with these, the numbers of spikes elicited during the stimulation (2s, 100 pA) were 218 significantly larger in both Cch and NA compared to those in the nACSF (Cch: nACSF 23 ± 1, Cch 31 ± 219 2, n = 34, T(33) = -4.69, p < 0.001; NA: nACSF 21 ± 1, NA 28 ± 2, n = 14, T(13) = -4.77, p < 0.001). A 220 reduced AHP mentioned above is also in agreement with the previous report (Madison and Nicoll, 221 1982), while the mAHP measured after five spikes (Table. 1) did not show significant decrease 222 probably due to a bad signal-to-noise ratio.

223 These indicate that persistent firing was not simply the result of increased excitability 224 because persistent firing was observed only in Cch. This is in line with the specific role of the CAN

225 current in support of persistent firing, which is activated downstream of the cholinergic receptor 226 (Knauer et al., 2013; Yan, Villalobos, & Andrade, 2009; Zizhen Zhang, Reboreda, Alonso, Barker, & 227 Séguéla, 2011). Changes of other basic properties by Cch and NA are listed in the Table 1.

228

229 Effect of NA on cholinergically-induced persistent firing

230 Based on the observation that NA alone did not support the induction of persistent firing, we next 231 tested whether NA has modulatory effects on persistent firing under cholinergic activation. In this 232 experiment, persistent firing was tested first in nACSF, then in Cch (5 µM), and finally in a solution 233 containing Cch and NA (5 – 10 µM). To ensure a full effect of NA, persistent firing was tested ~15 234 minutes after the initiation of the NA application. In ten neurons that showed persistent firing in Cch 235 condition, persistent firing was completely blocked in six, suppressed in three, and slightly increased 236 in one cell by NA (Fig. 2B). Post-stimulus firing frequency and membrane potential were both 237 significantly decreased by NA (Fig. 2D and E; Frequency: n = 12, T(11) = 4.118, p = 0.002; Potential: n 238 = 12, T(11) = 5.647, p < 0.001). Cells that did not exhibit persistent ﬁring in Cch did not show it in NA 239 either (n = 3; data not shown).

240 Next, to assess whether the reduction of persistent firing was synaptic or intrinsic, the effect 241 of NA on cholinergically induced persistent firing was examined in the presence of synaptic blockers 242 (SBs) that blocked the ionotropic excitatory and inhibitory transmissions (0.1 mM picrotoxin and 2 243 mM kynurenic acid). In line with our previous study in rat CA1 pyramidal cells (Knauer et al., 2013), 244 induction of persistent firing was not affected by the presence of the SBs, indicating that persistent 245 firing is supported by intrinsic cellular mechanisms (data not shown). As for the modulation of 246 persistent firing, an application of NA suppressed persistent firing significantly similarly to the case 247 without SBs (Fig. 2E; Frequency: n = 6, T(5) = 3.748, p = 0.013; Potential: n = 6; T(5) = -6.016, p = 248 0.002). Together, these data indicate that suppression of persistent firing was probably caused by 249 other mechanisms than a potential change in synaptic transmission due to NA.

250

251 Intracellular mechanisms underlying the suppression of persistent firing

252 Cholinergic and noradrenergic systems act through different receptors coupled to mainly three types

253 of GPCRs. While Ach activates only Gq/11 (M1, M3, M5) and Gi/o (M2, M4) pathways, NA can activate

254 Gs (β1, β2, β3) in addition to Gq/11 (α1) and Gi/o (α2, partially β2) pathways. The Gs pathway, unlike 255 other two pathways, increases the cAMP level, which is detrimental to working memory (Dash, 256 Moore, Kobori, & Runyan, 2007). In addition, an elevated cAMP level has been shown to suppress 257 persistent firing in vivo in the PFC, and suppresses the CAN current in the hippocampus as mentioned 8

258 above (El-Hassar et al., 2011). Therefore, we tested whether NA suppressed persistent firing by a

259 cAMP activation through the Gs pathway.

260 First, we evaluated the effect of an elevated cAMP concentration on persistent firing using 261 forskolin, a compound that activates the adenylate cyclase and therefore increases intracellular 262 levels of cAMP. Persistent firing was first tested in Cch (5 µM) and then forskolin (10 µM) was bath 263 applied in addition to Cch. This and other experiments in this section were conducted in the presence 264 of SBs. Forskolin clearly suppressed persistent firing (Fig. 3A and B; Frequency: n = 8, T(7) = 8.165, p < 265 0.001; Potential: n = 8, T(7) = 8.165, p < 0.001). We found that forskolin also decreased the number 266 of spikes during the stimulation (Cch: 23.56 ± 1.5; Cch + Forskolin: 19 ± 1.8; n = 8; T(7) = 3.789, p = 267 0.007). To study whether the suppression of persistent firing was simply due to the smaller number 268 of spikes elicited during the stimulation, response of the cells to a fixed number (six) of induced 269 spikes were additionally tested (Fig. 3C). Spikes were induced by using six brief current injections 270 (0.1-1 nA, 3 ms, 50 Hz) at -60 mV, and the post-stimulus potential was measured as an average 271 membrane depolarization during the two-second period after the simulation offset (gray line). The 272 post-stimulus potential was significantly smaller in the Cch+forskolin condition compared to that in 273 Cch alone (Fig. 4C; n = 7; T(6) = 4.803, p = 0.003) suggesting that the suppression of persistent ﬁring 274 by forskolin was not simply because of the reduced number of spikes triggered during the 275 stimulation.

276 The fact that forskolin is able to suppress persistent firing might be indicating that NA 277 suppression could be generated by an increase of cAMP that leads to PKA activation. To test this 278 hypothesis, we included the competitive PKA antagonist Rp-cAMPS (100 µM) in the recording 279 pipette. In this condition, NA mediated inhibition of persistent firing was strongly reduced (Fig. 4A, n 280 = 10). Figure 4B shows the effect of NA on the frequency and depolarization of persistent firing with 281 and without Rp-cAMPS. Rp-cAMPS prevented the suppressive effect of NA on persistent firing, 282 maintaining persistent firing similar to that in the control condition without NA (Fig. 4B). A multiple 283 comparison with a 2–way repeated measures ANOVA followed by pairwise comparisons with the 284 Tukey post-hoc test indicated that neither the frequency or depolarization changed significantly 285 when Rp-cAMPS was present (the third and fourth columns), while both were reduced significantly 286 by NA when Rp-cAMPS was not present (the first and second columns). In contrast, when persistent 287 firing in NA was compared with or without Rp-cAMPS (the second and fourth columns), both 288 frequency and depolarization were significantly larger when Rp-cAMPS was present (Fig. 4B; 289 Frequency: t(20) = 4.104, p=0.040; Potential: t(20) = 8.932, , p<0.001). These indicate that

290 suppression of persistent firing by NA is mainly by Gs activation of PKA.

291 As a control experiment, we repeated the forskolin application in the presence of Rp-cAMPS 292 Fig. 4C. Repeated measures two-way ANOVA coupled with the Tukey pot-hoc test indicated that 293 forskolin inhibition of persistent firing was strongly reduced in the presence of Rp-cAMPS (Fig. 4D 3rd 294 and 4th columns; frequency: t(13) = 3.164, p = 0.164; potential: t(13) = 1.321, p = 0.787). In contrast, 295 Rp-cAMPS had a significant effect on the frequency and depolarization in the presence of forskolin 296 (Fig.4D 2nd and 4th column; Frequency: t(13) = 5.903, p = 0.005; Potential: t(13) = 5.660, p = 0.007). 297 These indicate that the suppressive effect of forskolin was through PKA activation.

298 Experiments with forskolin and Rp-cAMPS supported the idea that NA interfered with 299 persistent firing through a cAMP mediated activation of PKA. To further demonstrate the

300 involvement of receptor subtypes coupled to the Gs pathway, effects of individual receptor agonists 301 were tested. First, the effect of the α2 receptor agonist clonidine (10 µM), which decreases cAMP

302 levels through Gi/o activation, was examined in the presence of Cch. As expected, clonidine did not 303 inhibit persistent firing (Fig. 5A and B; Frequency: n = 10, T(9) = -0.646, p = 0.534; Potential: n = 10; 304 T(9) = 1.234, p = 0.252). As for the adrenergic β receptors, although all β receptors (β1, β2, β3) are

305 coupled to Gs, recent studies pointed out that β2 activates the Gi in addition to the Gs cascade (Hall, 306 2004; Schutsky, Ouyang, Castelino, Zhang, & Thomas, 2011; Schutsky, Ouyang, & Thomas, 2011). 307 Therefore, we aimed to activate β1 and β3 receptors without activating β2 by using the general β 308 receptor agonist isoproterenol (ISO; 1 µM) along with the β2 blocker ICI 118,551 (ICI, 1 µM) based on 309 a previous report (Schutsky, Ouyang, Castelino, et al., 2011) . Combination of ISO and ICI significantly 310 inhibited persistent firing (Fig. 5C and D; Frequency: n = 7, T(6) = 4.634, p = 0.004; Potential: n = 7, 311 T(6) = 4.488, p = 0.004). In contrast, a sole application of the general β agonist ISO did not affect 312 persistent firing either at 1 µM (Frequency: n = 7, T(6) = 1.788, p = 0.124; Potential: n = 7, T(6)= 313 0.777, p = 0.467) or at 10 µM (Fig 5E and F; Frequency: n = 8, T(7) = 0.008, p = 0.994; Potential: n = 8, 314 T(7)= 0.729, p = 0.128), indicating that, in line with previous literature, β2 might induce actions that 315 counteract the increase of cAMP. In summary, these results suggest that NA suppresses 316 cholinergically-induced persistent firing via PKA activation downstream of a β1 and/or β3 receptors 317 mediated increase of cAMP.

318

319 Discussion

320 Based on the differential modulation of working memory by acetylcholine and NA, we investigated 321 effects of these neuromodulators on the induction and modulation of persistent firing in 322 hippocampal CA1 pyramidal cells in brain slices taken from adult female mice. Results from this study 323 indicate that only the cholinergic agonist Cch, but not NA, is able to support induction of persistent

324 firing. However, when NA was applied in addition to Cch, NA strongly suppressed cholinergically 325 triggered persistent firing. We have further shown that the suppression of persistent firing by NA was 326 not a result of decreased intrinsic excitability or ionotropic synaptic transmissions. In contrast, 327 forskolin, the specific activation of β1 and PKA blockade with Rp-cAMPS indicated that cAMP

328 elevations through the Gs protein cascade activating PKA caused the suppression of the cholinergic 329 persistent firing. 330 331 Molecular mechanisms underlying suppression of persistent firing 332 Persistent firing similar to our study is shown to be supported by intrinsic cellular mechanisms in 333 multiple brain areas involved in working memory (Egorov et al., 2002; Fraser & MacVicar, 1996; Haj- 334 Dahmane & Andrade, 1998; Jochems & Yoshida, 2013; Knauer et al., 2013; Navaroli, Zhao, 335 Boguszewski, & Brown, 2012; Tahvildari, Alonso, & Bourque, 2008). These studies have pointed out 336 that this type of persistent firing is supported by the CAN current, which is believed to be supported 337 by TRPC channels (Arboit et al., 2020; Reboreda, Jiménez-Díaz, & Navarro-López, 2011; Z. Zhang, 338 Reboreda, Alonso, Barker, & Séguéla, 2011). In line with our data, the CAN current and TRPC 339 channels are inhibited by the cAMP (El-Hassar et al., 2011; Partridge, Swandulla, & Müller, 1990; 340 Sung et al., 2011). In particular, El-Hassar and colleagues (El-Hassar et al., 2011) used forskolin in the 341 hippocampal CA1 pyramidal cells as in our study to indicate that TRPC mediated CAN current is 342 suppressed. Suppression of similar persistent firing by cAMP was also reported in the entorhinal 343 cortex and prefrontal cortex (Sidiropoulou et al., 2009; Zizhen Zhang et al., 2013). Therefore, CAN 344 current suppression is a feasible mechanism for the suppression of persistent firing in our study. 345 However, it is possible that other mechanisms were additionally involved. One possibility 346 could be that an increase in GABAergic input mediated by NA (Kawaguchi, 1998; Sessler, 1995) 347 suppressed persistent firing. Inhibitory synaptic input is known to be higher in the presence of NA 348 (Kawaguchi & Shindou, 1998; Sessler et al., 1995). Inhibitory synaptic input is effective in terminating 349 persistent firing in the dentate gyrus (Anderson & Strowbridge, 2014). This interpretation was in line 350 with the decreased input resistance we observed in NA, which could be caused by the opening of 351 GABAergic receptors. To gain insight into this, we used the same synaptic blockers as in Zhang and 352 colleagues (Zizhen Zhang et al., 2013), which demonstrated facilitated persistent firing by NA in PFC 353 neurons. However, in the presence of the synaptic blockers, input resistance was still reduced and 354 persistent firing was suppressed by NA. Based on these, we suggest that changes in synaptic 355 transmission were not the main cause of persistent firing suppression in our study. 356 Another possibility is that the elevation of cAMP has modulated the hyperpolarization 357 activated cyclic nucleotide-gated (HCN) channels. In the prefrontal cortex, adrenergic stimulation of 358 α2 receptors improved working memory performance by suppressing HCN channels through cAMP

359 downregulation (M. Wang et al., 2007). This is in line with facilitated persistent firing in the PFC 360 neurons under the blockade of HCN channels (Zizhen Zhang et al., 2013). This mechanism might, 361 however, not be involved in the suppression of persistent firing in our study because the sag 362 potential, which indicates the amount of HCN current, was rather reduced by NA, suggesting a 363 reduction, if any, of HCN current. In addition, α2 receptor agonist clonidine did not reduce the sag 364 potential in our study. This could be because 1) the cAMP levels were already low due to Cch, and α2 365 activation could not further reduce the cAMP level, 2) modulation of HCN channels by cAMP is 366 pronounced in young but is weak in adult animals in the hippocampus (Surges et al., 2006), and 3) 367 the α2 receptor does not modulate the HCN channels because co-localization of the α2 receptor with 368 the HCN channels, which is reported in the PFC (M. Wang et al., 2007), does not occur in the 369 hippocampal CA1 pyramidal cells. 370 In addition, we cannot completely exclude the possibility that mGluR is involved in the 371 suppression of persistent firing by NA. Similarly to Cch, mGluRs activate the Gq (Group 1) and Gi 372 (Group 2 and 3) pathways. When Cch is not present, NA-induced mGluR activation could play a role 373 in modulating persistent firing as was the case in Zhang et al. (2013). However, in the presence of Cch 374 as in our study, Gi and Gq are already highly activated, and an additional activation of the same Gi 375 and Gq through mGluRs, if exists, could be marginal. This point has been tested by Zhang et al (2013) 376 by applying the mGluR blocker MPEP on persistent firing induced by Cch, and they showed no change 377 of persistent firing, indicating that the role of mGluR in persistent firing is limited when Cch is present. 378 In contrast, when mGluRs are the main activator of the Gq and Gi pathways, NA may have a similar 379 suppressive effect on persistent firing in the hippocampus. This view is supported by the suppression 380 of mGluR-induced “plateau potential” and persistent firing firing by cAMP and PKA in the 381 hippocampus and PFC (Sidropoulou et al. 2009; El-Hassar et al., 2011). 382 383 Suppressed persistent firing as a possible cellular mechanism of impaired working memory in high 384 NA levels 385 Roles of cholinergic receptor activation on persistent firing supported by intrinsic cellular 386 mechanisms have been studied intensively in vitro (Reviewed in (Major & Tank, 2004; Yoshida, 387 Knauer, & Jochems, 2012)). Similarly to these prior studies, persistent firing was observed under 388 cholinergic receptor activation in our experiment in hippocampal pyramidal CA1 cells in mice. This is 389 in line with the supportive role of cholinergic activation on working memory (Kaneko & Thompson, 390 1997; Weiss et al., 2000). In addition, we have shown that cholinergically-induced persistent firing is 391 suppressed by NA. In line with these results are the detrimental effects of high concentrations of 392 noradrenaline NA on working memory (Arnsten, 2009; Roozendaal & McGaugh, 2011). Together, 393 these observations indicate that both supportive (cholinergic) and detrimental (NA) modulations on

394 working memory are in agreement with the cellular level modulation of PF, which is a possible 395 cellular correlate of working memory. 396 Interestingly, persistent firing in our study was suppressed by β1/3 receptors activation (Fig. 397 5D), while the use of general β receptor agonist did not alter the response (Fig. 5E), suggesting that 398 receptor subtypes have different effects on persistent firing. Traditionally, it has been believed that 399 the actions of different β receptors were identical, all activating the Gs cascade. However, recent 400 studies have suggested that β2 receptors might act through different mechanisms than β1, possibly 401 by activating the Gi cascade in addition to the Gs cascade (Hall, 2004; Schutsky, Ouyang, Castelino, et 402 al., 2011; Schutsky, Ouyang, & Thomas, 2011). Since Gi cascade counteract the action of Gs cascade 403 by decreasing cAMP levels, unclear effect of the general β receptor agonist (Fig. 5E) could have been 404 resulted from this action of β2 receptors. Interestingly, the role of β receptors in working memory 405 was tested using generic agonists or antagonists, resulting in no apparent effect on this cognitive 406 function (Arnsten et al., 1999). On the other hand, the specific activation of β1 receptors has been 407 shown to impair working memory, while the activation of β2 has improved it (Ramos, Colgan, Nou, & 408 Arnsten, 2008; Ramos et al., 2005). Therefore, modulations of persistent firing and working memory 409 agree with each other at the level of receptor subtypes in this case. 410 Our data using Rp-cAMPS and forskolin further suggested that the suppression of persistent

411 firing was through cAMP upregulation, which can be induced by the activation of the Gs cascade. 412 Interestingly, multiple studies have indicated that increased cAMP levels impair working memory, 413 while a reduction of cAMP levels improves it (Ramos et al., 2003; Runyan et al., 2005; Taylor et al., 414 1999). Moreover, increased cAMP reduced persistent firing in vivo during working memory tasks 415 (Vijayraghavan, Wang, Birnbaum, Williams, & Arnsten, 2007; M. Wang et al., 2007). It is relevant to 416 note that Rp-cAMPS has been tested in vivo in monkeys during an oculomotor delayed response 417 working memory task (Vijayraghavan et al., 2007). In this study, Rp-cAMPS rescued D1 receptor 418 mediated reduction of neural firing while an application of Rp-cAMPS alone increased the neuronal 419 firing (Vijayraghavan et al., 2007). The rescue effect of Rp-cAMPS is in line with the suppressive roles 420 of cAMP on persistent firing observed in our study. In addition, increased neural activity by the 421 application of Rp-cAMPS in vivo is in line with the increased depolarization during persistent firing we

422 observed in our study (Fig. 4B). In summary, Gs mediated cAMP upregulation seems to be

423 detrimental for both working memory and persistent firing, while Gi mediated cAMP downregulation 424 seem to support both working memory and persistent firing. 425 426 Supportive role of moderate levels of NA on working memory 427 In contrast to the detrimental effect of high levels of NA, it has been proposed that moderate 428 concentrations of NA aid working memory through the α2 receptors due to relatively high affinity of

429 NA to the α2 receptors (Arnsten & Li, 2005). In fact, working memory is facilitated by the α2 receptor 430 activation in several studies (Arnsten & Li, 2005; Birnbaum, Podell, & Arnsten, 2000; Jäkälä et al., 431 1999; Li et al., 1999; Ma et al., 2003). In the present study, we tested two doses of NA (5 and 10 μM), 432 however, we did not observe facilitative effect of NA on persistent firing either with or without 433 cholinergic agonist Cch. The lack of facilitative effect of NA on persistent firing could be because the 434 two concentrations tested were both high doses equivalent to situations of stress, where β1/3 435 receptors were active in addition to the α2 receptors, masking the positive effect of the α2 receptors. 436 However, our experiment with the α2 receptor agonist indicates that there is no room for facilitation 437 of persistent firing through α2 receptor activation at least in the presence of cholinergic agonist, 438 meaning that in this condition, even a lower concentration of NA would not have facilitated

439 persistent firing. The α2 receptor activates the Gi cascade, which decreases cAMP levels. One 440 possibility is that carbachol at the concentration we used already decreased cAMP levels to a

441 minimum through Gi activation and therefore additional activation of the α2 receptors could not 442 lower the cAMP levels further. In contrast to our study, in the rat prefrontal cortex, application of NA 443 (10 μM) enhanced persistent firing with or without the cholinergic agonist Cch (Zizhen Zhang et al., 444 2013). In their study, the persistent firing facilitation was mediated by a combination of α1 and α2 445 receptors, and no regulation was observed by the β receptor activation. While the direction of 446 modulation of persistent firing by NA is opposite from our study, the effect of cAMP level is 447 consistent in both studies. While increased cAMP (through β receptors) suppressed persistent firing 448 in our study, a decreased cAMP (through α2 receptors) facilitated persistent firing in Zhang et al 449 (Zizhen Zhang et al., 2013). Therefore, the opposite effect of NA could be caused by activation of 450 different receptor subtypes in these studies, possibly due to different affinities or expression density 451 between the hippocampus and PFC. Although data available on these issues is limited, it has been 452 reported that the β3 receptors are expressed more in the hippocampus than in the cortex, while α2 453 and β1 receptors are similarly expressed in these areas in rats (McCune, Voigt, & Hill, 1993; A. P. 454 Nicholas, Pieribone, & Hökfelt, 1993; Anthony P. Nicholas, Pieribone, & Hökfelt, 1993; Summers, 455 Papaioannou, Harris, & Evans, 1995). Inhibitory effect of NA in our study might be linked to a

456 stronger activation of Gs and cAMP upregulation through β3 receptors in the hippocampal neurons.

457 458 459 Concluding remarks 460 Working memory impairment and compromised persistent firing have often been attributed to a 461 reduced recurrent synaptic excitation (Arnsten, 2011; X. J. Wang, 2001). For example, it is proposed 462 that activation of HCN channels and SK channels at dendritic terminals attenuates synaptic currents 463 and reduce persistent spiking under cAMP elevation (M. Wang et al., 2011, 2007). On the other hand,

464 our study together with recent studies (El-Hassar et al., 2011; Sidiropoulou et al., 2009) support the 465 idea that cAMP may modulate working memory through direct inhibition of the neuronal ability to 466 support persistent firing, possibly through the modulation of CAN/TRPC current. 467 Impaired working memory due to high levels of cAMP may also be involved in normal aging, 468 schizophrenia and traumatic brain injury in addition to high stress conditions (Arnsten, 2011; Kobori 469 et al., 2015; M. Wang et al., 2011). In support of this view, knocking down the disrupted in 470 schizophrenia 1 (DISC1) protein (a protein that participates in cAMP catabolism) results in a decrease 471 of the TRPC current in rat PFC neurons (El-Hassar et al., 2014). In addition, an infusion of Rp-cAMPS 472 in traumatic brain injury or aged rats with working memory deficits, improves the performance on 473 those tasks (Kobori et al., 2015; Ramos et al., 2003). Therefore, the mechanism of suppression of 474 intrinsic persistent firing we present here might be relevant to working memory impairment in 475 different conditions such as aging, traumatic brain injury and schizophrenia as well. 476 477

478 Acknowledgements

479 We thank Dr. Menno Witter for kindly sharing experimental procedures with us. 480 481

482 Tables

483 484 Table 1

485 Part I

50 pA 100 pA 150 pA 200 pA 250 pA 300 pA 350 pA nACSF 1±0 11±2 21±2 31±2 37±2 42±2 46±2 n = 47 Cch 6±2** 23±2*** 35±2*** 41±2*** 46±2*** 49±2*** 53±3** nACSF 1±1 7±3 15±3 22±4 27±5 n = 13 NA 2±1 12±3* 23±4** 31±4** 36±4** 486 487 Part II

IR (megΩ) TH (mV) SAG mAHP (mV) ADP (mV) n n n n n

nACSF 105±4 -47.9±.6 .166±.01 -1.16±.11 -.27±.07 48 51 48 42 42 Cch 139±6*** -48.9±.7** .151±.01 -.26±.14*** 1.24±.15*** nACSF 15 90±6 9 -46.4±2.0 15 .173±.01 9 -1.47±.34 9 -0.40±.25

NA 115±8*** -47.3±1.8 .115±.01** -1.33±.27 .23±.13* 488 489 Table 1. Effect of Cch and NA on general cellular properties. (Part I) Number of spikes elicited by 1s 490 current steps. (Part II) Input resistance (IR), spike threshold (TH), sag ratio (SAG), medium after- 491 hyperpolarizing potential (mAHP), after depolarizing potential (ADP). Values correspond to mean ± 492 SEM. 493 494 Table 2

495 Part I

50 pA 100 pA 150 pA 200 pA 250 pA 300 pA 350 pA Cch 6±3 26±4 37±3 41±3 47±3 50±3 53±3 n = 13 Cch + NA 19±5 29±5 42±4 48±4 54±5 57±6 60±5 496 497 Part II

IR (meg.Ohm) TH (mV) SAG mAHP (mV) ADP (mV) Cch n = 142±11 -51.0±1.6 .147±.02 -.14±.29 1.28±.26 Cch + NA 13 129±10* -53.7±1.6 .145±.01 -.27±.22 .84±.12 498 499 Table 2. Effect of NA on top of Cch, on general cellular properties. (Part I) Number of spikes elicited 500 by 1s current steps. (Part II) Input resistance (IR), spike threshold (TH), sag ratio (SAG), medium after- 501 hyperpolarizing potential (mAHP), after depolarizing potential (ADP). Values correspond to mean ± 502 SEM. 503 504

505 Figure captions

506 507 Figure 1. Cholinergic induction of persistent firing in CA1 hippocampal neurons. (A) Representative 508 cell's responses in nACSF (left) and in Cch (right). The gray line indicates the 10 s period in which 509 characteristics of persistent firing were analyzed. Trace at the bottom shows the current step applied 510 to test persistent firing. (B) Post-stimulus frequency (left) and post-stimulus potential (right) in 511 nACSF (black) and Cch (white). (C) Representative trace of the effect of muscarinic M1 antagonist 512 pirenzepine on the cholinergic persistent firing. (D) Post-stimulus frequency (left) and post-stimulus 513 potential (right) in Cch (black) and Cch and pirenzepine (white). (E) Representative trace of the effect 514 of muscarinic M2/M4 antagonist himbacine on the cholinergic persistent firing. (F) Post-stimulus 515 frequency (left) and post-stimulus potential (right) in Cch (black) and Cch and himbacine (white). (G)

516 Representative responses from timed control experiment. (H) Post-stimulus frequency (left) and 517 post-stimulus potential (right) in the first test in Cch (black) and the second test in Cch (white). 518 519 Figure 2. NA suppresses persistent firing. (A) Representative traces of the effect of NA in nACSF. (B) 520 Representative traces of the effect of NA on the cholinergic persistent firing. (C) Effect of NA on post- 521 stimulus frequency (left) and post-stimulus potential (right) without Cch. (D) Same as in (C) but in 522 Cch. (E) Same as in (D) but in the presence of the synaptic blockers (SB). 523 524 Figure 3. Forskolin suppresses persistent firing. (A) Representative traces of the inhibition by 525 forskolin on the cholinergic persistent firing. (B) Post-stimulus frequency (left) and post-stimulus 526 potential (right) in Cch (black) and Cch + forskolin (white). (C) Representative after depolarization 527 following six spikes in Cch (black) and in a cocktail containing Cch and Forskolin (gray). Gray straight 528 line indicates the section analyzed. Vertical lines under truncated spikes indicate the timing of 529 current injection. The bar graph shows the membrane potential in Cch (black) and Cch plus Forskolin 530 (white). 531 532 Figure 4. Inhibition of PKA by Rp-cAMPS blocks the NA and forskolin mediated reduction of 533 persistent firing. (A) Representative traces of the effect of Rp-cAMPS on the NA-mediated inhibition 534 of the cholinergic persistent firing. (B) Post-stimulus frequency (left) and post-stimulus potential 535 (right) comparison with 2-way repeated measures ANOVA followed by Tukey post-hoc test. The first 536 two columns show measurements without Rp-cAMPS from previous experiments. (C) Representative 537 traces of the effect of Rp-cAMPS on the forskolin-mediated inhibition of the cholinergic persistent 538 firing. (D) Post-stimulus frequency (left) and post-stimulus potential (right) comparison with 2-way 539 repeated measures ANOVA followed by Tukey post-hoc test. 540 541 Figure 5. Roles of NA receptor subtypes in suppression of persistent firing. (A) Representative traces 542 of the effect of clonidine on the cholinergic persistent firing. (B) Post-stimulus frequency (left) and 543 post-stimulus potential (right) in Cch (black) and Cch and clonidine (white). (C) Representative traces 544 of the effect of β2 (inhibited by ICI 118,551) vs β1 and β3 adrenergic receptors (activated by 545 isoprotenerol in the presence of ICI 118,551) on the cholinergic persistent firing. (D) Post-stimulus 546 frequency (left) and post-stimulus potential (right) in a cocktail of Cch and ICI 118,551 (ICI; black) and 547 in a cocktail of Cch, ICI and isoproterenol (white). (E) Representative traces of the effect of β-receptor 548 stimulation by isoprotenerol. (F) Post-stimulus frequency (left) and post-stimulus potential (right) in 549 (black) and Cch and isoproterenol (white). 550

551 Bibliography

552 553 Anderson, R. W., & Strowbridge, B. W. (2014). Regulation of persistent activity in hippocampal mossy 554 cells by inhibitory synaptic potentials. Learning & Memory, 21(5), 263–271. 555 https://doi.org/10.1101/lm.033829.113 556 Arboit, A., Reboreda, A., & Yoshida, M. (2020). Involvement of TRPC4 and 5 Channels in Persistent 557 Firing in Hippocampal CA1 Pyramidal Cells. Cells, 9(2), 365. 558 https://doi.org/10.3390/cells9020365 559 Arnsten, A. F. T. (2009, June). Stress signalling pathways that impair prefrontal cortex structure and 560 function. Nature Reviews Neuroscience. Nat Rev Neurosci. https://doi.org/10.1038/nrn2648 561 Arnsten, A. F. T. (2011, June 15). Catecholamine influences on dorsolateral prefrontal cortical 562 networks. Biological Psychiatry. Biol Psychiatry. https://doi.org/10.1016/j.biopsych.2011.01.027 563 Arnsten, A. F. T., & Li, B. M. (2005, June 1). Neurobiology of executive functions: Catecholamine 564 influences on prefrontal cortical functions. Biological Psychiatry. Biol Psychiatry. 565 https://doi.org/10.1016/j.biopsych.2004.08.019 566 Arnsten, A. F. T., Mathew, R., Ubriani, R., Taylor, J. R., & Li, B. M. (1999). ??-1 Noradrenergic Receptor 567 Stimulation Impairs Prefrontal Cortical Cognitive Function. Biological Psychiatry, 45(1), 26–31. 568 https://doi.org/10.1016/S0006-3223(98)00296-0 569 Arnsten, A. F. T., Wang, M. J., & Paspalas, C. D. (2012, October 4). Neuromodulation of Thought: 570 Flexibilities and Vulnerabilities in Prefrontal Cortical Network Synapses. Neuron. NIH Public 571 Access. https://doi.org/10.1016/j.neuron.2012.08.038 572 Barch, D. M., & Ceaser, A. (n.d.). Cognition in schizophrenia: core psychological and neural 573 mechanisms. https://doi.org/10.1016/j.tics.2011.11.015 574 Benardo, L. S., & Prince, D. A. (1982). Cholinergic excitation of mammalian hippocampal pyramidal 575 cells. Brain Research, 249(2), 315–331. https://doi.org/10.1016/0006-8993(82)90066-x 576 Birnbaum, S. G., Podell, D. M., & Arnsten, A. F. T. (2000). Noradrenergic alpha-2 receptor agonists 577 reverse working memory deficits induced by the anxiogenic drug, FG7142, in rats. 578 Pharmacology Biochemistry and Behavior, 67(3), 397–403. https://doi.org/10.1016/S0091- 579 3057(00)00306-3 580 Bremner, J. D., Krystal, J. H., Southwick, S. M., & Charney, D. S. (1996). Noradrenergic mechanisms in 581 stress and anxiety: I. Preclinical studies. Synapse, 23(1), 28–38. 582 https://doi.org/10.1002/(SICI)1098-2396(199605)23:1<28::AID-SYN4>3.0.CO;2-J 583 Bukalo, O., Fentrop, N., Lee, A. Y. W., Salmen, B., Law, J. W. S., Wotjak, C. T., … Schachner, M. (2004). 584 Conditional Ablation of the Neural Cell Adhesion Molecule Reduces Precision of Spatial 585 Learning, Long-Term Potentiation, and Depression in the CA1 Subfield of Mouse Hippocampus.

586 Journal of Neuroscience, 24(7), 1565–1577. https://doi.org/10.1523/JNEUROSCI.3298-03.2004 587 Colombo, M., & Gross, C. G. (1994). Responses of inferior temporal cortex and hippocampal neurons 588 during delayed matching to sample in monkeys (Macaca fascicularis). Behavioral Neuroscience, 589 108(3), 443–455. https://doi.org/10.1037//0735-7044.108.3.443 590 Dash, P. K., Moore, A. N., Kobori, N., & Runyan, J. D. (2007). Molecular activity underlying working 591 memory. Learning & Memory, 14(8), 554–563. https://doi.org/10.1101/lm.558707 592 Diamond, D. M., Ingersoll, N., Fleshner, M., & Rose, G. M. (1996). Psychological stress impairs spatial 593 working memory: Relevance to electrophysiological studies of hippocampal function. 594 Behavioral Neuroscience, 110(4), 661–672. https://doi.org/10.1037/0735-7044.110.4.661 595 Egorov, A. V, Hamam, B. N., Fransén, E., Hasselmo, M. E., & Alonso, A. A. (2002). Graded persistent 596 activity in entorhinal cortex neurons. Nature, 420(6912), 173–178. 597 https://doi.org/10.1038/nature01171 598 El-Hassar, L., Hagenston, A. M., D’Angelo, L. B., & Yeckel, M. F. (2011). Metabotropic glutamate 599 receptors regulate hippocampal CA1 pyramidal neuron excitability via Ca 2+ wave-dependent 600 activation of SK and TRPC channels. The Journal of Physiology, 589(13), 3211–3229. 601 https://doi.org/10.1113/jphysiol.2011.209783 602 El-Hassar, L., Simen, A. A., Duque, A., Patel, K. D., Kaczmarek, L. K., Arnsten, A. F. T., & Yeckel, M. F. 603 (2014). Disrupted in Schizophrenia 1 Modulates Medial Prefrontal Cortex Pyramidal Neuron 604 Activity Through cAMP Regulation of Transient Receptor Potential C and Small-Conductance K+ 605 Channels. Biological Psychiatry, 76(6), 476–485. 606 https://doi.org/10.1016/j.biopsych.2013.12.019 607 Fraser, D. D., & MacVicar, B. A. (1996). Cholinergic-Dependent Plateau Potential in Hippocampal CA1 608 Pyramidal Neurons. Journal of Neuroscience, 16(13), 4113–4128. 609 Funahashi, S., Bruce, C. J., & Goldman-Rakic, P. S. (1989). Mnemonic coding of visual space in the 610 monkey’s dorsolateral prefrontal cortex. Journal of Neurophysiology, 61(2), 331–349. 611 Goldman-Rakic. (1995). Cellular Basis of Working Memory Review. Neuron, 14, 477–485. 612 https://doi.org/10.1016/0896-6273(95)90304-6 613 Haj-Dahmane, S., & Andrade, R. (1998). Ionic mechanism of the slow afterdepolarization induced by 614 muscarinic receptor activation in rat prefrontal cortex. Journal of Neurophysiology, 80(3), 1197– 615 1210. https://doi.org/10.1152/jn.1998.80.3.1197 616 Hall, R. A. (2004). β-adrenergic receptors and their interacting proteins. Seminars in Cell and 617 Developmental Biology, 15(3), 281–288. https://doi.org/10.1016/j.semcdb.2003.12.017 618 Hampson, R. E., & Deadwyler, S. A. (2003). Temporal firing characteristics and the strategic role of 619 subicular neurons in short-term memory. Hippocampus, 13(4), 529–541. 620 https://doi.org/10.1002/hipo.10119

621 Jäkälä, P., Sirviö, J., Riekkinen, M., Koivisto, E., Kejonen, K., Vanhanen, M., & Jr, P. R. (1999). 622 Guanfacine and Clonidine, Alpha2-Agonists, Improve Paired Associates Learning, but not 623 Delayed Matching to Sample, in Humans. Neuropsychopharmacology, 20(2), 119–130. 624 https://doi.org/10.1016/S0893-133X(98)00055-4 625 Jochems, A., & Yoshida, M. (2013). Persistent firing supported by an intrinsic cellular mechanism in 626 hippocampal CA3 pyramidal cells. European Journal of Neuroscience, 38(2), 2250–2259. 627 https://doi.org/10.1111/ejn.12236 628 Kamiński, J., Sullivan, S., Chung, J. M., Ross, I. B., Mamelak, A. N., & Rutishauser, U. (2017). 629 Persistently active neurons in human medial frontal and medial temporal lobe support working 630 memory. Nature Neuroscience, 20(4), 590–601. https://doi.org/10.1038/nn.4509 631 Kaneko, T., & Thompson, R. F. (1997). Disruption of trace conditioning of the nictitating membrane 632 response in rabbits by central cholinergic blockade. Psychopharmacology, 131(2), 161–166. 633 https://doi.org/10.1007/s002130050279 634 Kawaguchi, Y., & Shindou, T. (1998). Noradrenergic excitation and inhibition of GABAergic cell types 635 in rat frontal cortex. Journal of Neuroscience, 18(17), 6963–6976. 636 https://doi.org/10.1523/jneurosci.18-17-06963.1998 637 Knauer, B., Jochems, A., Valero-Aracama, M. J., & Yoshida, M. (2013). Long-lasting intrinsic persistent 638 firing in rat CA1 pyramidal cells: a possible mechanism for active maintenance of memory. 639 Hippocampus, 23(9), 820–831. https://doi.org/10.1002/hipo.22136 640 Kobori, N., Moore, A. N., & Dash, P. K. (2015). Altered regulation of protein kinase a activity in the 641 medial prefrontal cortex of normal and brain-injured animals actively engaged in a working 642 memory task. Journal of Neurotrauma, 32(2), 139–148. https://doi.org/10.1089/neu.2014.3487 643 Li, B. M., Mao, Z. M., Wang, M., & Mei, Z. T. (1999). Alpha-2 adrenergic modulation of prefrontal 644 cortical neuronal activity related to spatial working memory in monkeys. 645 Neuropsychopharmacology : Official Publication of the American College of 646 Neuropsychopharmacology, 21(5), 601–610. https://doi.org/10.1016/S0893-133X(99)00070-6 647 Ma, C.-L., Qi, X.-L., Peng, J.-Y., & Li, B.-M. (2003). Selective deficit in no-go performance induced by 648 blockade of prefrontal cortical α2-adrenoceptors in monkeys. NeuroReport, 14(7), 1013–1016. 649 https://doi.org/10.1097/01.wnr.0000070831.57864.7b 650 Madison, D. V., & Nicoll, R. A. (1982). Noradrenaline blocks accommodation of pyramidal cell 651 discharge in the hippocampus. Nature, 299(5884), 636–638. https://doi.org/10.1038/299636a0 652 Major, G., & Tank, D. (2004). Persistent neural activity: prevalence and mechanisms. Current Opinion 653 in Neurobiology, 14(6), 675–684. https://doi.org/10.1016/j.conb.2004.10.017 654 McCune, S. K., Voigt, M. M., & Hill, J. M. (1993). Expression of multiple alpha adrenergic receptor 655 subtype messenger RNAs in the adult rat brain. Neuroscience, 57(1), 143–151.

656 https://doi.org/10.1016/0306-4522(93)90116-W 657 Millar, J. K. (2005). DISC1 and PDE4B Are Interacting Genetic Factors in Schizophrenia That Regulate 658 cAMP Signaling. Science, 310(5751), 1187–1191. https://doi.org/10.1126/science.1112915 659 Navaroli, V. L., Zhao, Y., Boguszewski, P., & Brown, T. H. (2012). Muscarinic receptor activation 660 enables persistent firing in pyramidal neurons from superficial layers of dorsal perirhinal cortex. 661 Hippocampus, 22(6), 1392–1404. https://doi.org/10.1002/hipo.20975 662 Nicholas, A. P., Pieribone, V. A., & Hökfelt, T. (1993). Cellular localization of messenger RNA for beta- 663 1 and beta-2 adrenergic receptors in rat brain: An in situ hybridization study. Neuroscience, 664 56(4), 1023–1039. https://doi.org/10.1016/0306-4522(93)90148-9 665 Nicholas, Anthony P., Pieribone, V., & Hökfelt, T. (1993). Distributions of mRNAs for alpha‐2 666 adrenergic receptor subtypes in rat brain: An in situ hybridization study. Journal of Comparative 667 Neurology, 328(4), 575–594. https://doi.org/10.1002/cne.903280409 668 Partridge, L. D., Swandulla, D., & Müller, T. H. (1990). Modulation of calcium‐activated non‐specific 669 cation currents by cyclic AMP‐dependent phosphorylation in neurones of Helix. The Journal of 670 Physiology, 429(1), 131–145. https://doi.org/10.1113/jphysiol.1990.sp018248 671 Qin, S., Hermans, E. J., van Marle, H. J. F., Luo, J., & Fernández, G. (2009). Acute Psychological Stress 672 Reduces Working Memory-Related Activity in the Dorsolateral Prefrontal Cortex. Biological 673 Psychiatry, 66(1), 25–32. https://doi.org/10.1016/j.biopsych.2009.03.006 674 Ramos, B. P., Birnbaum, S. G., Lindenmayer, I., Newton, S. S., Duman, R. S., & Arnsten, A. F. T. (2003). 675 Dysregulation of protein kinase a signaling in the aged prefrontal cortex: new strategy for 676 treating age-related cognitive decline. Neuron, 40(4), 835–845. 677 https://doi.org/https://doi.org/10.1016/S0896-6273(03)00694-9 678 Ramos, B. P., Colgan, L. A., Nou, E., & Arnsten, A. F. T. (2008). ??2 Adrenergic Agonist, Clenbuterol, 679 Enhances Working Memory Performance in Aging Animals. Neurobiology of Aging, 29(7), 1060– 680 1069. https://doi.org/10.1016/j.neurobiolaging.2007.02.003 681 Ramos, B. P., Colgan, L., Nou, E., Ovadia, S., Wilson, S. R., & Arnsten, A. F. T. (2005). The beta-1 682 adrenergic antagonist, betaxolol, improves working memory performance in rats and monkeys. 683 Biological Psychiatry, 58(11), 894–900. https://doi.org/10.1016/j.biopsych.2005.05.022 684 Reboreda, A., Jiménez-Díaz, L., & Navarro-López, J. D. (2011). TRP Channels and Neural Persistent 685 Activity. Transient Receptor Potential Channels, Advances in Experimental Medicine and Biology, 686 704, 323–339. https://doi.org/10.1007/978-94-007-0265-3 687 Reboreda, A., Theissen, F. M., Valero-Aracama, M. J., Arboit, A., Corbu, M. A., & Yoshida, M. (2018). 688 Do TRPC channels support working memory? Comparing modulations of TRPC channels and 689 working memory through G-protein coupled receptors and neuromodulators. Behavioural Brain 690 Research, 354, 64–83. https://doi.org/10.1016/j.bbr.2018.02.042

691 Roozendaal, B., & McGaugh, J. L. (2011). Memory modulation. Behavioral Neuroscience, 125(6), 797– 692 824. https://doi.org/10.1037/a0026187 693 Runyan, J. D., Moore, A. N., & Dash, P. K. (2005). A role for prefrontal calcium-sensitive protein 694 phosphatase and kinase activities in working memory. Learning & Memory (Cold Spring Harbor, 695 N.Y.), 12(2), 103–110. https://doi.org/10.1101/lm.89405 696 Schoofs, D., Preuß, D., & Wolf, O. T. (2008). Psychosocial stress induces working memory 697 impairments in an n-back paradigm. Psychoneuroendocrinology, 33(5), 643–653. 698 https://doi.org/10.1016/j.psyneuen.2008.02.004 699 Schutsky, K., Ouyang, M., Castelino, C. B., Zhang, L., & Thomas, S. A. (2011). Stress and 700 Glucocorticoids impair memory retrieval via β 2-adrenergic, G i/o-coupled suppression of cAMP 701 signaling. Journal of Neuroscience, 31(40), 14172–14181. 702 https://doi.org/10.1523/JNEUROSCI.2122-11.2011 703 Schutsky, K., Ouyang, M., & Thomas, S. A. (2011). Xamoterol impairs hippocampus-dependent 704 emotional memory retrieval via Gi/o-coupled β2-adrenergic signaling. Learning & Memory (Cold 705 Spring Harbor, N.Y.), 18(9), 598–604. https://doi.org/10.1101/lm.2302811 706 Sessler, F. M., Liu, W., Kirifides, M. L., Mouradian, R. D., Lin, R. C. S., & Waterhouse, B. D. (1995). 707 Noradrenergic enhancement of GABA-induced input resistance changes in layer V regular 708 spiking pyramidal neurons of rat somatosensory cortex. Brain Research, 675(1–2), 171–182. 709 https://doi.org/10.1016/0006-8993(95)00060-4 710 Sidiropoulou, K., Lu, F.-M., Fowler, M. A., Xiao, R., Phillips, C., Ozkan, E. D., … Cooper, D. C. (2009). 711 Dopamine modulates an mGluR5-mediated depolarization underlying prefrontal persistent 712 activity. Nature Neuroscience, 12(2), 190–199. https://doi.org/10.1038/nn.2245 713 Summers, R. J., Papaioannou, M., Harris, S., & Evans, B. A. (1995). Expression of β3‐adrenoceptor 714 mRNA in rat brain. British Journal of Pharmacology, 116(6), 2547–2548. 715 https://doi.org/10.1111/j.1476-5381.1995.tb17205.x 716 Sung, T. S., Jeon, J. P., Kim, B. J., Hong, C., Kim, S. Y., Kim, J., … So, I. (2011). Molecular determinants 717 of PKA-dependent inhibition of TRPC5 channel. AJP: Cell Physiology, 301(4), C823–C832. 718 https://doi.org/10.1152/ajpcell.00351.2010 719 Surges, R., Brewster, A. L., Bender, R. A., Beck, H., Feuerstein, T. J., & Baram, T. Z. (2006). Regulated 720 expression of HCN channels and cAMP levels shape the properties of the h current in 721 developing rat hippocampus. European Journal of Neuroscience, 24(1), 94–104. 722 https://doi.org/10.1111/j.1460-9568.2006.04880.x 723 Tahvildari, B., Alonso, A. A., & Bourque, C. W. (2008). Ionic basis of ON and OFF persistent activity in 724 layer III lateral entorhinal cortical principal neurons. Journal of Neurophysiology, 99(4), 2006– 725 2011. https://doi.org/10.1152/jn.00911.2007

726 Taylor, J. R., Birnbaum, S., Ubriani, R., & Arnsten, A. F. (1999). Activation of cAMP-dependent protein 727 kinase A in prefrontal cortex impairs working memory performance. The Journal of 728 Neuroscience, 19(18), RC23. 729 Valero-Aracama, M. J., Sauvage, M. M., & Yoshida, M. (2015). Environmental enrichment modulates 730 intrinsic cellular excitability of hippocampal CA1 pyramidal cells in a housing duration and 731 anatomical location-dependent manner. Behavioural Brain Research, 292, 209–218. 732 https://doi.org/10.1016/j.bbr.2015.05.032 733 Vijayraghavan, S., Wang, M., Birnbaum, S. G., Williams, G. V, & Arnsten, A. F. T. (2007). Inverted-U 734 dopamine D1 receptor actions on prefrontal neurons engaged in working memory. Nature 735 Neuroscience, 10(3), 376–384. https://doi.org/10.1038/nn1846 736 Wang, M., Gamo, N. J., Yang, Y., Jin, L. E., Wang, X. J., Laubach, M., … Arnsten, A. F. T. (2011, August 737 11). Neuronal basis of age-related working memory decline. Nature. Nature Publishing Group. 738 https://doi.org/10.1038/nature10243 739 Wang, M., Ramos, B. P., Paspalas, C. D., Shu, Y., Simen, A., Duque, A., … Arnsten, A. F. T. (2007). 740 Alpha2A-adrenoceptors strengthen working memory networks by inhibiting cAMP-HCN channel 741 signaling in prefrontal cortex. Cell, 129(2), 397–410. https://doi.org/10.1016/j.cell.2007.03.015 742 Wang, X. J. (2001). Synaptic reverberation underlying mnemonic persistent activity. Trends in 743 Neurosciences, 24(8), 455–463. https://doi.org/10.1016/S0166-2236(00)01868-3 744 Weiss, C., Preston, A. R., Oh, M. M., Schwarz, R. D., Welty, D., & Disterhoft, J. F. (2000). The M1 745 muscarinic agonist CI-1017 facilitates trace eyeblink conditioning in aging rabbits and increases 746 the excitability of CA1 pyramidal neurons. Journal of Neuroscience, 20(2), 783–790. 747 https://doi.org/10.1523/jneurosci.20-02-00783.2000 748 Yan, H.-D., Villalobos, C., & Andrade, R. (2009). TRPC Channels Mediate a Muscarinic Receptor- 749 Induced Afterdepolarization in Cerebral Cortex. The Journal of Neuroscience, 29(32), 10038– 750 10046. https://doi.org/10.1523/JNEUROSCI.1042-09.2009 751 Yoshida, M., Knauer, B., & Jochems, A. (2012). Cholinergic modulation of the CAN current may adjust 752 neural dynamics for active memory maintenance, spatial navigation and time-compressed 753 replay. Frontiers in Neural Circuits, 6, 10. https://doi.org/10.3389/fncir.2012.00010 754 Zhang, Z., Reboreda, A., Alonso, A., Barker, P. A., & Séguéla, P. (2011). TRPC channels underlie 755 cholinergic plateau potentials and persistent activity in entorhinal cortex. Hippocampus, 21(4). 756 https://doi.org/10.1002/hipo.20755 757 Zhang, Zizhen, Cordeiro Matos, S., Jego, S., Adamantidis, A., & Séguéla, P. (2013). Norepinephrine 758 Drives Persistent Activity in Prefrontal Cortex via Synergistic α1 and α2 Adrenoceptors. PLoS 759 ONE, 8(6), e66122. https://doi.org/10.1371/journal.pone.0066122 760 Zhang, Zizhen, Reboreda, A., Alonso, A., Barker, P. A., & Séguéla, P. (2011). TRPC channels underlie

761 cholinergic plateau potentials and persistent activity in entorhinal cortex. Hippocampus, 21(4), 762 386–397. https://doi.org/10.1002/hipo.20755 763