1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

NP5579_proof ■ 2 September 2014 ■ 1/17

Neuropharmacology xxx (2014) 1e17

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential Q8

A. Ahnaou*, H. Huysmans, T. Jacobs, W.H.I.M. Drinkenburg Dept. of Neuroscience, Janssen Research & Development, A Division of Janssen Pharmaceutica NV, Turnhoutseweg 30, B-2340 Beerse, Belgium

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 April 2014 Received in revised form 18 August 2014 Accepted 20 August 2014 Available online xxx

Synchronization of electroencephalographic (EEG) oscillations represents a core mechanism for cortical and subcortical networks, and disturbance in neural synchrony underlies cognitive processing deficits in neurological and neuropsychiatric disorders. Here, we investigated the effects of cognition enhancers (donepezil, rivastigmine, tacrine, galantamine and memantine), which are approved for symptomatic treatment of dementia, on EEG oscillations and network connectivity in conscious rats chronically instrumented with epidural electrodes in different cortical areas. Next, EEG network indices of cognitive impairments with the muscarinic receptor antagonist scopolamine were modeled. Lastly, we examined the efficacy of cognition enhancers to normalize those aberrant oscillations. Cognition enhancers elicited systematic (“fingerprint”) enhancement of cortical slow theta (4.5 e6 Hz) and gamma (30.5e50 Hz) oscillations correlated with lower activity levels. Principal component analysis (PCA) revealed a compact cluster that corresponds to shared underlying mechanisms as compared to different drug classes. Functional network connectivity revealed consistent elevated coherent slow theta activity in parieto-occipital and between interhemispheric cortical areas. In rats instrumented with depth hippocampal CA1-CA3 electrodes, donepezil elicited similar oscillatory and coherent activities in cortico-hippocampal networks. When combined with scopolamine, the cognition enhancers attenuated the leftward shift in coherent slow delta activity. Such a consistent shift in EEG coherence into slow oscillations associated with altered slow theta and gamma oscillations may underlie cognitive deficits in scopolamine-treated animals, whereas enhanced coherent slow theta and gamma activity may be a relevant mechanism by which cognition enhancers exert their beneficial effect on plasticity and cognitive processes. The findings underscore that PCA and network connectivity are valuable tools to assess efficacy of novel therapeutic drugs with cognition enhancing potential. © 2014 Published by Elsevier Ltd.

Keywords: Cognition enhancers EEG oscillations Coherent functional network Neurodegenerative disorders Animal model Translational biomarker

Q1

1. Introduction A prominent property of neuronal networks is their tendency to engage in oscillatory rhythms. Neuronal oscillations represent a fundamental mechanism enabling coordinated activity during normal brain functioning, and are therefore an instrumental research target for neurological and neuropsychiatric disorders. Brain networks communicate with each other by means of neuronal oscillations at multiple temporal and spatial scales to integrate feature information during sensory processing and ensure higher order coordination of motor and cognitive functions ki and Watson, 2012; Hammond et al., (Babiloni et al., 2011; Buzsa 2007; Uhlhaas and Singer, 2010). Neuronal assemblies are able to

* Corresponding author. Tel.: þ32 14 60 72 26; fax: þ32 14 60 74 33. E-mail addresses: [email protected], [email protected] (A. Ahnaou).

modulate their level of synchrony without affecting their firing rate (Singer, 1999), which typically occur at different frequencies both ki, 2006; Steriade et al., 1993) and awake states during sleep (Buzsa (Fries et al., 2001; Pesaran et al., 2002). Oscillations in each frequency band may have a different underlying mechanism and subserve a different function. Synchronous oscillatory rhythms in the slow theta and gamma oscillations represent a core mechanism for sculpting temporal coordination of disparate brains networks during higher order cognitive tasks including attention and working memory (Cantero and Atienza, 2005; Kaiser and Lutzenberger, 2003; Lisman and ki, 2008; Singer, 1999). Several studies have focused on the Buzsa origin of theta rhythm, which generators have been found in speki and cific hippocampal and extra-hippocampal regions (Buzsa Watson, 2012; Montgomery et al., 2008; Vertes, 2005). Oscillations at theta frequency range are critical to establish precise temporal interactions to propagate and coordinate the flow of

http://dx.doi.org/10.1016/j.neuropharm.2014.08.015 0028-3908/© 2014 Published by Elsevier Ltd.

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 2/17

2

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

information across distributed neuronal networks in the sub regions of the hippocampus and the entorhinal cortex (Cappaert et al., 2009). Gamma oscillatory activity, which has repeatedly been shown to accompany neurocognitive functions in normal subjects, has its primary generators in the cortex where subsets of inhibitory GABAergic interneuron circuits modulate glutamatergic pyramidal cell activity (Debener et al., 2003; Fell et al., 2001; Fries, 2009; Sederberg et al., 2003; Tallon-Baudry et al., 2005). Oscillatory activity in the gamma frequency range establishes synchronization with great precision in short distance local cortical networks ki, 2006). (Buzsa There is growing recognition that aberrant oscillatory activity and network connectivity represent a core feature of a wide range of neurological and neuropsychiatric disorders such as Alzheimer, Parkinson, Schizophrenia and Attention Deficit Hyperactivity Disorder (ADHD) (for reviews see Herrmann and Demiralp, 2005; Uhlhaas and Singer, 2006, 2010). Impairment in synaptic plasticity and/or synaptic connectivity disrupts the synchrony of both local and long range EEG oscillations, which may lead to dysfunctional coordination of distributed neuronal activities between and within functionally specialized brain regions resulting in cognitive dysfunctions and alterations in regional cerebral blood flow and metabolism (Cook and Leuchter, 1996; Hammond et al., 2007; Sloan et al., 1995; Rodriguez et al., 1999; Uhlhaas and Singer, 2010; Williams and Gordon, 2007; Yener and Bas¸ar, 2013). In case of Alzheimer's disease (AD), the present mainstay pharmacological treatment is based on vascular prevention and symptomatic therapy, which does not decelerate or prevent the progression of the disease. Current first line pharmacotherapy approved for mild to moderate Alzheimer's disease includes acetylcholinesterase inhibitors (AChEIs) (donepezil, rivastigmine, galantamine and tacrine) and the N-methyl-d-aspartic acid (NMDA) receptor antagonist (memantine). While AChEIs reduce the enzymatic degradation of the neurotransmitter Ach, deficient in the AD brain, and thus enhance the cholinergic system (Singh et al., 2013), memantine is thought to reduce the over activation of NMDA receptors, while allowing normal activity to occur (Witt et al., 2004). Amphetamine was added in this study to differentiate dopaminerelated stimulant activity. Understanding humaneanimal correlations is instrumental in the development of novel drugs and efforts are devoted to search for an objective biomarker that would provide a quantifiable index of neuronal network activities. The information derived from the electroencephalogram (EEG) has contributed to understanding normal brain function and dysfunctions in neurological and neuropsychiatric disorders. The field potential EEG recorded from the human scalp provides a powerful noninvasive tool for studying network oscillations under different settings such as diagnosis, identification of altered biological pathways, assessment of diseases progression and efficiency of therapeutics. Typical hallmarks of EEG alterations were found in cortical dementias and described in AD such as general slowing and decrease of alpha activity which leads to increased delta and theta activities (Jeong, 2004). Additionally, reduction in higher EEG frequency components in occipital and temporal areas correlates with cognitive decline and the severity of the disease (Jeong, 2004; Jackson and Snyder, 2008). Therefore, the approach is suited in translational preclinical research to examine the rapid changing patterns in coordination and synchronization of EEG network oscillations that underlie cognitive as well as behavioral dysfunctions. Consequently, in the present study, the central effects of drugs were characterized by using epidural EEG measures of oscillations and network connectivity from different cortical areas involved in sensory and cognitive function. The study objectives were (1) to use

these neurophysiological oscillations and connectivity as valuable quantitative biological markers to reproduce cardinal EEG correlates of cognition deficits in conscious rats, and (2) to measure efficacy of cognition enhancers to normalize abnormalities in EEG oscillations. This pharmacological approach may therefore help to improve detection of more effective and targeted pharmacological interventions of neurological and psychiatric disorders involving cognitive deficits. An animal model of cognitive deficits was prepared with the muscarinic cholinergic antagonist scopolamine, which is a gold standard proven to be accurate and reproducible to induce cognitive impairments in preclinical and clinical settings (Klinkenberg and Blokland, 2010; Ebert and Kirch, 1998; Lenz et al., 2012; Riedel and Jolles, 1996; Sannita et al., 1987). The hypothesis tested whether cognition enhancers would produce opposite effects on EEG oscillations and network connectivity as compared to the aberrant EEG activities as modeled in rats and as described in patients with cognitive decline. 2. Materials and methods 2.1. Animals and surgical procedure The experiments were carried out in male adult Sprague Dawley rats, supplied by Harlan, (Netherlands) and weighing about 250 g at the time of surgery. Animals were housed in full-view Plexiglas cages (25  33 cm, 18 cm high) that are an integrated art of IVC-racks (individually ventilated cages) located in a soundattenuated chamber. Animals were maintained under controlled environmental conditions throughout the study: 22 ± 2  C ambient temperature, the relative humidity at 60%, 12:12 lightedark cycle (lights off 06:59e18:59; light intensity: ~100 lux,) and food and water available ad libitum. The experimental procedures were carried out in accordance with the U.K. Animals (Scientific Procedures) Act, 1986 and associated guidelines, the European Communities Council Directive of 24 November 1986 (86/609/EEC) or the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978), and received approval from the animal care and use committee of Janssen Research & Development. The surgery was performed under isoflurane inhalation anesthesia. A mixture of 30% O2, 70% N2O and 5% isoflurane was administered to animals as an initial induction for 2 min. Then, the animals were mounted in a stereotaxic apparatus and were given a continuous constant mixture of O2, N2O and 2% isoflurane. An analgesic Piritramide (dipidolor) 0.025 mg/kg (0.1 ml/100 g. B.W) was administered before the incision over the total length of the head. The oval area of the scalp was removed, and the uncovered skull was cleared of the periosteum. Animals were equipped with EEG and EMG electrodes as described in Fig. 1. Six stainless steel recording screws (diameter 1 mm) were inserted bilaterally in the left and right hemisphere along the antero-posterior axes at the locations (frontal left “FL”, parietal left “PL”, occipital left “OL” and frontal right “FR”, parietal right “PR”, occipital right “OR”, respectively). All electrodes were placed stereotaxically (FL/FR: AP þ2 mm, L ± 2 mm; PL/PR: AP 2 mm, L ± 2 mm and OL/OR: AP 6 mm, L ± 2 mm from Bregma; the incisor bar was 5 mm under the center of the ear bar, all according to the stereotactic atlas of Paxinos and Watson, 1998) and referenced to ground electrodes placed midline Q2 above of the cerebellum. The Biosemi ActiveTwo recording system (Biosemi, Amsterdam, The Netherlands) replaces the conventional ground electrodes by two separate electrodes: the common mode sense (CMS) active electrode and the driven right leg (DRL) passive electrode. This common mode reference for online data acquisition and impedance measures is a feedback loop driving the average potential across the montage close to the amplifier zero. In addition, stainless steel wires electrodes were placed in the muscles of the neck to record electromyographic (EMG) activity. The EMG signals were used to differentiate active versus passive behavior. Electrodes (stainless steel wire, 7N51465T5TLT, 51/46 Teflon Bilaney, Germany) were connected to a pin (Future Electronics: 0672-2-15-15-30-27-10-0) with a small insert (track pins; Dataflex: TRP-1558-0000) and were fitted into a 10hole connector, then the whole assembly was fixed with dental cement to the cranium. 2.2. Recording, spectral power, principal component analysis and functional network connectivity Two weeks after surgery, the animals were habituated to the recording EEG procedure, which occurred under vigilance controlled waking condition during the dark phase. In addition to EMG indices, general motor activity was measured by two passive infrared (PIR) detectors placed above each recording cage and the envelope activity was used in the analysis of motion activity following drug administration. Artefacts were rejected on a 4-s basis if the power exceeded a threshold based on a mean power value in the 0.5- to 4.5-Hz and 20- to 30-Hz bands and confirmed by

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 3/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

3

Fig. 1. Time course of the effects of A/donepezil, B/rivastigmine and C/memantine on the EEG spectral patterns in fronto-parieto-occipital cortical areas (only right hemisphere is presented) during each 15 min block of the recording session after the administration. Mean EEG spectra were expressed as a percentage of the mean spectra in the baseline period. Peak changes were elicited in both slow theta (4e6.5 Hz) and gamma (32e48 Hz) frequency oscillations during 2 h after the administration. Changes from baseline are presented as a heat map with cold dark blue to warm red color indicates an order increase in the magnitude of oscillatory power density in particular frequency. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 4/17

4

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Fig. 1. (continued).

visual inspection. Only continuous waking artifact-free epochs with low-voltage fast EEG activity, high to moderate EMG and body activities were used in the analysis. Epochs with high-voltage slow cortical waves in the absence of EMG and locomotor activities were discarded. A notch FIR filter at 50 Hz was applied to avoid voltage related to power line interferences After a stable baseline recording session of 30 min, signals from 6 brain areas were recorded for 2 h after drug administration (n ¼ 8 for each condition). Continuous EEG and EMG field potentials were acquired at 2 kHz sample rate with an input range of ±500 mV. The signals were digitized with 24 bit resolution, amplified and analogue band-pass filtered between 1 and 100 Hz. 2.2.1. Spectral power Spectral density estimates were calculated using Fast Fourier Transform with Hanning window function; block size of 512 data points, giving 1.0 Hz resolution, and power was expressed as percentage of total power over 1e100 Hz. The average spectral density in each frequency oscillation was normalized across animals to obtain the full power spectrum (heat map with 1 Hz resolution) constructed from overlapping windows of 4 s data and plotted for baseline periods and post-drug periods windowing over 1e100 Hz. Frequency bands were averaged for each 15 min session into frequency windows: Delta band (1e4 Hz), Theta band (theta1: 4e6.5 Hz, theta2: 6.5e8 Hz), Alpha band (alpha1: 8e11 Hz; alpha2: 11e14 Hz), Beta band (beta1: 14e18 Hz; beta2: 18e32 Hz), Gamma band (gamma1: 32-48; gamma2: 52e100 Hz). 2.2.2. Principal component analysis The multivariate statistical principal component analysis (PCA) method was applied to reduce dimensionality of the datasets while retaining the variation that is originally present. The obtained vector component that represents the most important variables influencing the results diversity was used in projection technique to look for pattern recognitions (Dimpfel, 2003). The PCA used existing datasets of reference drug classes obtained in the same standardized EEG experiments described in the present work. By using the fit from the PCA analysis, we can project results of the new experimental compound into the multi-dimensional component space formed by the reference compounds and we fit a cluster dendrogram using each compounds geometric component space mean. The resulting dendrogram gives an indication as to which reference compounds' class the newly tested compound is most similar. The EEG signals were measured from six brain regions in animals that received vehicle and drugs (low, middle and higher doses; n ¼ 8 per condition) and spectral

dynamics were analyzed in 9 different EEG bands during 8 time-blocks of 15 min. For each drug-class dataset, the doses estimated to have a therapeutic response in vigilance states and behavioral tests in rodent were used in the analysis. As the PCA requires complete observations, animals with missing values in the dataset were excluded from the analysis. The procedure starts with a dataset containing all complete vehicle observations and all complete observations at the specified dosage level from the reference datasets. For each animal in the dataset, there were 432 variables, representing every possible combination between time-point, brain region and EEG bands. The basic effect of the PCA projection is to project all the variables in the dataset to a small set of components that account for the greatest proportion of the variance in the data. Using these 432 variables, a principal component analysis was fitted to all observations to the specified number of components. Neither the drug name nor the drug classes were passed to the PCA function. The component space is made purely as a function of the EEG information contained in each animal's 432 variables. The drug class is strictly used for coloring the PCA projections and dendrogram. Also note that the test compound dataset was not used to form the PCA projections and projections for the reference compounds were identical across runs of the PCA application. The variable loadings from the PCA projections were saved and applied to the newly test compound dataset. Then, the drug's geometric center was computed by averaging the locations of all observations (animals) for that drug. This averaging will have dimensionality equal to the number of specified principal components. Thus, for a PCA projection fit that is using 4 principal components, the geometric center will be a set of coordinates in 4-dimensional space. These components were fit into hierarchical clustering dendrogram using the Ward's linkage, which helps to see and understand the group structures in the data. Here, we have used as our unknown compound “Rivastigmine” dataset obtained in a standard EEG experiment, which serves as a good example of the application's abilities. 2.2.3. Functional network activity Functional network activity between cortical regions was computed by coherence, which is commonly used to probe the integrity of cortical neural pathways and index the functional coupling between different cortical structures at various frequencies (Thatcher, 2012). The amplitude of EEG coherence ranges between 0 and 1: a low value indicates no similarity between the two signals, whereas values close to 1 indicate a high similarity between two time series signals. Coherence between the time series in different electrodes was quantified at different time points before and after administration of drugs using a generalized additive mixed model where the coherence over the frequencies was fitted using a cubic spline for the different

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 5/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Q3

treatment groups and the heterogeneity between the animals (Priestley, 1981; Chatfield, 1975, 2000). Subsequently, coherence between each pair of electrodes were concatenated every 4 s and aggregated over 15 min in each frequency bands for each animal, time point, and treatment group. Particular frequency bands were selected for comparison between groups based on the most prominent oscillatory and coherent activities across different brain regions. To reduce the impact of volume conduction of signals propagated from common generators, computation of all edges in which the maximum absolute value of coherence occurred at zero time lag between time series were removed. 2.2.4. Epidural versus depth EEG recordings A separate group of 20 rats was stereotaxically instrumented with intracortical (M1: AP þ2 mm, L ± 2 mm, V þ2 mm from the Bregma) as well as intra-hippocampal electrodes (CA1: AP 3.14 mm, L ± 1.8 mm, V þ2.8 mm; CA3: AP 4.5 mm, L ± 3.8 mm, V þ4 mm from Bregma, respectively) in order to compare functional coherent activity between cortical (epidural) and hippocampal (depth) networks. Donepezil at the dose of 3 mg/kg shown in earlier studies to elicit peak coherent activity in epidural recordings was used in this study. 2.3. Drugs All drugs were synthesized in Janssen Research & Development laboratories. Donepezil (0.3, 1.0 and 3.0 mg/kg), rivastigmine (0.3, 1.0 and 3.0 mg/kg), tacrine (0.16, 2.5, 10 mg/kg) galantamine (0.16, 2.5, 10 mg/kg), memantine (1, 3 and 10 mg/ kg), scopolamine (0.16, 0.64 and 2.5 mg/kg) and amphetamine (0.16, 0.64, 2.5 mg/kg) were formulated in sterile saline solution, and were administered subcutaneously (s.c.) in a volume of 10 ml/kg body weight. Saline was administered in control groups. All drug classes used in the database were used at the doses that showed therapeutic efficacy in neurophysiological and behavioral studies in rodents. The database includes antidepressants (desipramine, nortryptiline, paroxetine, fluoxetine, trazodone buproprion); Anxiolytics (ipsapirone, oxazepam, buspirone, zolpidem); psychostimulants (amphetamine, PCP, ketamine); antipsychotics (olanzapine, risperidone, clozapine, paliperidone, haloperidol); mood stabilizers (lamotrigine, valproate). 2.4. Statistics In the dose response experiments, values of consecutive 4-s epochs were averaged for a 30 min period (baseline) followed by 15 min periods (treatment) averages for the remaining time after the administration of vehicle or test compound. In the pharmacological reversal challenge experiments, EEG oscillations were averaged for 30 min before the muscarinic cholinergic antagonist (baseline), and then for every 15 min after combined treatment with cognition enhancers. Drug-induced changes in EEG spectral power were calculated in blocks of 15 min for a total of 2 h as the ratio of mean spectral power obtained following the administration of test drug versus the mean spectral power obtained during 30 min baseline period. This procedure allows for assessment of drug-induced changes in different cortical networks and EEG power expressed in each frequency band as a percentage of the original power, compared to the vehicle condition. Time course EEG spectral changes were submitted to a two-way multivariate ANOVA for repeated measures with two main factors (treatment and period of recordings), followed by a one-way ANOVA with main factor treatment for each of the 15 min period. Data were presented as the means ± S.E.M. The strength of association between locomotor activity and EEG slow theta oscillation (4e6.5 Hz) was assessed with linear Spearman correlation analysis. The aggregated coherences as well as the changes from baseline were analyzed per frequency band using an ANOVA with time, group and its interaction as covariates while taking into account the heterogeneity between the different animals (Verbeke and Molenberghs, 2000). P-values from the associated Ftests were reported as well as Least squares means from the corresponding ANOVA models.

3. Results 3.1. Cognition enhancers enhanced EEG slow theta (4e6.5 Hz) and gamma (32e48 Hz) oscillations We present the effects of different cognition enhancers on EEG oscillations. AChEIs donepezil, rivastigmine, tacrine, galantamine and the NMDA antagonist memantine elicited common changes in EEG network oscillations across different brain areas. A concomitant dose-dependent increase in the slow theta (4e6.5 Hz) and gamma (32e48 Hz) oscillatory patterns were found in different brain regions (Fig. 1AeC heat maps of the right side hemisphere are displayed for donepezil and rivastigmine and memantine). Quantitative assessment of slow theta (4e6.5 Hz) and gamma (32e48 Hz) frequency power for all cognition enhancers is

5

indicated in Fig. 2A and B. A much pronounced effect on slow theta activity (4e6.5 Hz) was observed particularly in parietal and occipital cortical areas, whereas an enhanced effect on slow gamma (32e48 Hz) occurred in all cortical areas. However, no major effect was found in the higher gamma frequency range (52e100 Hz). Thus, oscillatory activities in the slow theta and gamma frequency range appear to be particularly responsive in the mechanism of action of cognition enhancers. 3.2. Correlation between slow theta oscillations (4e6.5 Hz) and locomotor activity It was assessed whether the time course of the increase in EEG slow theta activity (4e6.5 Hz) oscillations were associated with changes in motor behavior following the administration of cognition enhancers. Analysis of motion activity at different time points showed that administration of all cognition enhancers' was associated with a reduction in activity levels (Fig. 2C) as opposed to the effects of the classical psychostimulant amphetamine. The results of a Spearman rank correlations analysis at 30 min after the administration of cognition enhancers between the slow theta (4e6.5 Hz) and activity levels are shown in Fig. 2D, upper panels. The correlation analysis showed that slow theta (4e6.5 Hz) was negatively correlated with activity levels (donepezil: r ¼ 0.85, p < 0.001; rivastigmine: r ¼ 0.57, p < 0.001; tacrine: r ¼ 0.63, p < 0.001; galantamine: r ¼ 0.66; memantine: r ¼ 0.43, p < 0.05). The scatter plot in Fig. 2D, bottom panels shows a highly significant negative association between EEG slow theta activity (4e6.5 Hz) and activity for each time intervals across all treatments (Spearman rank correlation analysis for all time points: rS ranges between 0.35 and 0.97, all P-values < 0.05). However, the Spearman rank correlations analysis at 30 min after amphetamine administration showed a positive correlation between slow alpha activity (8e11 Hz) (r ¼ 0.83, p < 0.001) and motor activity, which was maintained across all time intervals of the recording session (rS ranges between 0.7 and 0.84, all Pvalues < 0.001). 3.3. Principal component analysis Principal component analysis was used to identify common compact clusters that correspond to shared underlying mechanisms (“fingerprinting’). An unknown compound “Rivastigmine” dataset (black filled circles; Fig. 3) was used, which was not considered in the original learning dataset, and was projected into the space defined by our historical drug-classes datasets. Here, the first two components account for 20.8% and 17.1% of the variance in the dataset, respectively (Fig. 3, left panel). Using the four components, we fit a hierarchical clustering and visualize the clustering with a dendrogram. The results show that rivastigmine was grouped in the common spatial pattern of cognitive enhancing drugs, indicating that its effect on EEG oscillations over time was most closely matched with the cognition enhancers donepezil, tacrine, galantamine, memantine (Fig. 3, right panel). 3.4. Parallel coherent patterns in epidural and depth cortical structures To examine whether frontal, parietal and occipital networks were functionally related, we computed coherence indices in fronto-parieto-occipital cortical areas following the administration of donepezil at different doses. This revealed consistent increases in slow theta (4e6.5 Hz) and gamma (32e48 Hz) activities accompanied by coherence increase in those frequency bands in

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 6/17

6

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Fig. 2. Changes in the power of A/slow theta (4e6.5 Hz), B/slow gamma (32e48 Hz) frequencies during the first 2 h after the administration of donepezil, rivastigmine, tacrine, galantamine and memantine (only right hemisphere is displayed). The red dots on the brain diagram mark the brain region in which the drug effects are displayed. Color coded bars underneath the curves indicate intervals in which oscillatory activity difference differed from vehicle Mixed model ANOVA. C/Activity levels during the first 2 h after the administration of donepezil, rivastigmine, tacrine and galantamine. Note that the classical psychostimulant amphetamine was included to display differences in motion levels. D/ Spearman correlations (upper panels) at 30 min after the administration of different cognition enhancers between the slow theta oscillations (4e6.5 Hz) and locomotor activity, and (lower panels) R-Values plotted for each time block of 15 min. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

NP5579_proof ■ 2 September 2014 ■ 7/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

7

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Fig. 2. (continued).

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 8/17

8

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Fig. 3. A/A scatter plot of the principal component analysis for all drug classes tested. The PCA analysis was based on 432 variables (six brain regions times 8 for 9 frequency bands). The space projection distance biplot displayed the mean adjusted observations of the main two obtained principal components. The colored symbols in the figure legend correspond to the six drug classes and vehicles. Different drugs classes are coded by different colors. B/Dendrogram drug clusters. The level at which branches merge (relative to the “root” of the tree) is related to their highest similarity: the height of the branch points indicates how similar or different are from each other: the greater the height, the greater the difference between clusters. Note that the drug Rivastigmine, which was not part of the initial learning datasets showed high similarity with other cognition enhancers in the left part of the plot. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

distributed fronto-parieto-occipital networks (Fig. 4A, coherent pairs left panels). The presence of slow theta (4e6.5 Hz) and gamma (32e48 Hz) frequency oscillations suggests that cortical networks might spontaneously generate those oscillations or be derived from volume conduction of the signals propagated from common source. Therefore depth EEG recordings within the motor cortex, CA1 and CA3 after the administration of vehicle (n ¼ 10) and of donepezil at the dose of 3 mg/kg (n ¼ 10) was performed. In accordance with epidural recordings, the presence of prominent slow theta (4e6.5 Hz) and gamma (32e48 Hz) power associated with peak phase coherent activity was found. Power spectra derived from intracortical and intrahippocampal sites in rats implanted with depth electrodes displayed peak coherent slow theta (4e6.5 Hz) and gamma (32e48 Hz) oscillatory activities. Cortical motor cortex M1/M2 areas displayed marked gamma oscillations in all leads whereas CA1 and CA3 exhibited slower theta (4e6.5 Hz) rhythm. Donepezil (3 mg/kg) elicited strong cortico-hippocampal slow theta (4e6.5 Hz) and gamma (32e48 Hz) coherent activities (Fig. 4, coherent pairs right panels). This pattern of synchronized activity of the hippocampus and cortical structures parallels the pattern seen in rats instrumented with epidural electrodes. These data suggest those epidural network oscillatory activities in the slow theta (4e6.5 Hz) and gamma (32e48 Hz) range are unlikely derived from the same source as removal of the edges with maximal absolute coherence values occurring at zero time lag excluded spurious edges attributable to volume conduction. Analysis of functional connectivity changes as evolving over time, showed maintenance of stable coherent activity in slow theta (4e6.5 Hz) frequency oscillations with consistent contralateral

coherent activities (Fig. 4B). In addition, a strong dose-dependent decrease in coherent slow alpha (8e11 Hz) was observed in fronto-parieto-occipital pairs. 3.5. Effects of scopolamine on EEG oscillations and network connectivity To model EEG correlates of cognition deficits observed in brain disorders involving cognitive functions, we dose-dependently influenced the cholinergic transmission by the muscarinic cholinergic receptor antagonist scopolamine (0.16, 0.64 and 2.5 mg/kg). Scopolamine evoked a dose-dependent leftward shift towards slow EEG network oscillations. A peak of functional effect expressed most clearly in the increased slow delta activity in different brain areas was found at about 15 min after administration of the middle and highest doses and lasted till at least 2 h posttreatment (Fig. 5). A striking and robust network coherent delta activity was found in fronto-parieto-occipital sites, whereas coherence in fast oscillatory rhythms was strikingly reduced (data not shown). 3.6. Efficacy of the cognition enhancers to attenuate scopolamineinduced aberrant EEG oscillations and network connectivity Quantitative evaluation of the efficacy of cognition enhancers to reverse abnormalities in EEG network connectivity associated with disorganized behavior was performed in scopolamine-treated rats. The cognition enhancers tested in this study were potent to significantly reverse the leftward shift and increased coherent slow EEG delta oscillation synchrony in scopolamine-treated rats, in

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 9/17

Fig. 4. A/Network coherence changes after subcutaneous administration of vehicle and donepezil in epidural recording (left panels; n ¼ 8 for each condition) and intracortical recording (right panels; n ¼ 10 for each condition). Donepezil at 3 mg/kg elicited similar synchronized slow theta (4e6.5 Hz) and gamma (32e48 Hz) pattern in epidural and intracortical recording in comparison with that of the control group. B/Functional network coherence changes from baseline in each frequency band of interest derived from consecutive 4s epochs concatenated and averaged in 15 min blocks. The width of the edges between pair electrodes is drawn proportional to the weight of changes in coherent activity from baseline (0e50%, 50e100% and 100e150%). Red color represents increases and blue color indicates decreases. Average networks derived from the first session recording session in each pharmacological condition and for each frequency band of interest d (0.5e4 Hz); q1 (4.5e6 Hz); q2 (6.5e8 Hz); a1 (8.5e11); a2 (11.5e13); b1 (13.5e18 Hz); b2 (18.5e30 Hz); g1 (30.5e50 Hz); g2 (50.1e70 Hz), g3 (70.1e100 Hz). Time-evolving dominant coherent activity in the frontal-parietal and occipital structures indicated most persistent edges scaled to q1 frequency during waking. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 10/17

10

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Fig. 5. A/Time course of the effects of subcutaneous administration of scopolamine (0.16, 0.64, and 2.5 mg/kg) on the EEG spectral patterns in fronto-parieto-occipital cortical areas (only right hemisphere is presented) during each 15 min block of the recording session. Mean EEG spectra were expressed as a percentage of the mean spectra in the baseline period. Peak changes were elicited in both slow delta (0.5e4 Hz) frequency oscillations during 2 h after the administration. Changes from baseline are presented as a heat map with cold dark blue to warm red color indicates an order increase in the magnitude of oscillatory power density in particular frequency. B/Changes in peak delta frequency during 2 h after the administration (only right hemisphere is displayed, n ¼ 8 for each condition). Color coded bars underneath the curves indicate intervals in which oscillatory activity difference differed from vehicle. Mixed model ANOVA. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

location dependent manner (Fig. 6A, only Donepezil displayed). The findings suggest restoration of abnormalities in the network coordination to its large-scale synchronization form. Although no behavioral cognitive task was considered in these studies, the leftward shift in EEG neuronal oscillation likely indicated a depression in brain activity, which could correlate with impairment in cognitive processing.

4. Discussion The present studies investigated the effects of cognition enhancers, which are approved for symptomatic treatment of patients with mild-moderate AD, on EEG oscillations in freely behaving rats and their potency to attenuate aberrant oscillations and network connectivity in the scopolamine-induced cognitive deficits model. Three major results are reported: i) EEG field potentials following cognition enhancers manifested common changes, notably a concomitant enhancement of oscillations and connectivity in slow theta (4e6.5 Hz) and gamma (32e48 Hz) in different brain networks associated with reduced motion activity, ii) Scopolamine promoted a consistent leftward shift coherent activity confined to the delta-range, iii) pre-treatment with cognition enhancers as shown with donepezil attenuated the leftward shift in network

connectivity dominated by enhanced coherence in slow EEG oscillations. 4.1. Translational value of the EEG recording sites The rat's frontal cortex and specifically the medial frontal cortex, is usually considered to be homologous to the dorsolateral prefrontal cortex of primates, which consists of cortex homologous to human anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (Seamans et al., 2008; Wallis, 2012). The cortex is hierarchically organized and some cortical areas serve higher-order integrative functions that are neither purely sensory nor purely moto. These higher-order areas of the cortex, known as association areas, serve to associate sensory inputs to motor response and perform those mental processes that intervene between sensory inputs and motor outputs. The cortex of each hemisphere contains one primary motor area (M1) and primary somatosensory areas. The M1 is not a simple transmission relay between the nonprimary motor areas (that anticipate, prepare) and the spinal cord, but rather a real crossroad playing an important role in cognitive motor integration as it could be active without any motor tasks (Carlsson et al., 2000; Saper et al., 2000). The primary somatosensory cortex and parietal association cortex provide information about the ongoing motor response. Because the multimodal

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 11/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

11

Fig. 6. A/Time course of the effects of combined subcutaneous administration of vehicle/vehicle, vehicle/scopolamine (0.16 mg/kg), donepezil (3 mg/kg)/vehicle and of donepezil (3 mg/kg)/scopolamine (0.16 mg/kg) on the EEG spectral patterns in fronto-parieto-occipital cortical areas (only right hemisphere is presented) during each 15 min block of the recording session. Mean EEG spectra were expressed as a percentage of the mean spectra in the baseline period. Peak changes were elicited in both slow delta (0.5e4 Hz) frequency oscillations during 2 h after the administration. Changes from baseline are presented as a heat map with cold dark blue to warm red color indicates an order increase in the magnitude of oscillatory power density in particular frequency. B/Changes in peak delta frequency during 2 h after the combined treatments (only right hemisphere is displayed, n ¼ 8 for each condition). Color coded bars underneath the curves indicate intervals in which oscillatory activity difference differed from vehicle. Mixed model ANOVA. C/Timeevolving dominant coherent activity in the frontal-parietal and occipital structures indicated the potential of donepezil to attenuate scopolamine-induced most persistent d frequency edges. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 12/17

12

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

association areas integrate sensory modalities and link sensory information to the planning of movement, they are thought to be the anatomical substrates of the highest brain function in human's conscious thought, perception, and goal-directed action. Hence, the M1 and primary somatosensory cortical recording sites were explicitly chosen in the present work. 4.2. Relevance of theta and gamma oscillations and connectivity in the mechanism of action of cognition enhancers All cognition enhancers consistently enhanced cortical oscillations and coherent networks activity in the slow theta (4e6.5 Hz) associated with lower motor activity. Dynamic neural oscillations are fundamental mechanisms for flexible communication between brain circuits to integrate sensory information and to coordinate higher-order cognitive and motor functions, suggesting that these coordinated patterns of activity could be a crucial target in cognition research (Varela et al., 2001; Womelsdorf et al., 2007). Neuronal oscillations in the theta and gamma range can be measured at different anatomical scales, ranging from invasive recordings of small neuronal populations till epidural regional networks. Synchronization of networks at theta rhythm establishes precise temporal interactions between disparate brain areas, whereas the gamma rhythm establishes synchronization with great precision in local cortical networks. Synchrony in the theta frequency oscillations represents one of the best-studied neuronal activities in mammalian's hippocampus, which is implicated in the top-down control of cognitive functions (Vertes, 2005). Theta oscillations, which dominates the EEG recorded from the hippocampal formation and the medial septum, acts as a pacemaker for the generation of this neuronal rhythm, which drives prefrontal cortical networks into theta oscillatory rhythm (Thierry et al., 2000). The strong theta rhythm synchronizes medial prefrontal cortex neurons and strengthens synaptic links between spatially distributed cortical areas, and therefore facilitates the transfer of hippocampal information to neocortex during learning and memory (Buszaki, 2002; Paz et al., 2008). It has been repeatedly demonstrated that perceptive, cognitive, and motor processes are associated with enhanced EEG coherence in the theta frequency oscillations between prefrontal and posterior association € lle et al., 2004; Seemüller et al., 2012; cortex and hippocampus (Mo Schmidt et al., 2013; Womelsdorf and Fries, 2006). Coherence is commonly used to probe the integrity of cortical neural pathways and index the functional coupling between different cortical structures at various frequencies (Thatcher, 2012). In the present study, all cognition enhancers modulated the slow theta (4e6.5 Hz) oscillations and its coherent activity, which may be important for memory and cognitive process. The strong and stable coherent slow theta (4e6.5 Hz) frequency between cortical areas may be associated with enhanced functional connectivity between frontal cortex and hippocampus for gating information flow and cortical processing. Changes in theta activity are strongly associated with changes in motor behavior in rats (Oddie and Bland, 1998; Vanderwolf, 1969). To rule out the possibility that EEG changes reflect the central effects of cognition enhancers rather than changes in motor behavior, we investigated the changes in motion activity and whether the time course of these alterations were correlated. In contrast to classical psychostimulants' effect on locomotor activity levels, the cognition enhancers induced slow theta (4e6.5 Hz) when motor activity is decreased revealing dissociation between theta power and motor activity. In addition, multiple spearman correlations analysis confirmed a strong negative association of slow theta (4e6.5 Hz) with low activity levels. These findings suggest that

cognition enhancers appear to influence cortical activation in these regions at lower motor activity. Although the theta oscillatory rhythm was less pronounced in frontal sites as compared to central and posterior recording sites, there were increased stable functional slow theta (4e6.5 Hz) coherent edges in bilateral frontal, parietal and occipital regions. Spectra based coherence derived from depth recording signals in donepezil treated animals exhibited a coherent pattern, which closely resembles the patterns of synchronized activity prevalent at epidural recording sites. Thus, it is presumable that the increased theta coherence in fronto-parieto-occipital cortical regions was due to temporal coupling of local networks activities generated in the underlying cortical structures i.e. the hippocampus and cortex to facilitate attention and encoding process. On a related note, it is worth mentioning that the hippocampus in rodents is located under the cerebral cortex, whereas in primates it is located in the medial temporal lobe, underneath the cortical surface containing two main interlocking parts such as the Ammon's horn and the dentate gyrus. Notwithstanding this anatomical disparity, functional neuroimaging and large scale microelectrode recording cognitive or sleep studies support the organization of networks, which provides an optimal functional brain state for effective hippocampus-to-neocortex processing and transfer of information (Burianov a et al., 2012; Cohen, 2011; Ferrara et al., 2012; Le Van Quyen et al., 2010; Staresina et al., 2013; Wagner et al., 2010). All cognition enhancers elicited cortical oscillations and coherent activity in the slow gamma (32e48 Hz) frequency across different cortical areas. Cortical gamma oscillations have received a great interest because of their role in sensory and cognitive functions such as selective attention (Doesburg et al., 2008; Fries et al., 2001; Lakatos et al., 2008; Womelsdorf and Fries, 2006; Womelsdorf et al., 2007), short-and long-term memory (Herrmann et al., 2004; Pesaran et al., 2002; Wu et al., 2008) and multisensory integration (Lakatos et al., 2007). The primary generators of gamma oscillations include the cortex, where subsets of inhibitory GABAergic interneuron circuits that modulate glutamatergic pyramidal cell activity, are thought to play a primary role (Gray and McCormick, 1996; Traub et al., 1999). However, gamma activity is also prevalent in the hippocampus formation, where it is postulated to assist cognitive functions such as encoding and retrieval of memory traces (Bragin et al., 1995; Chrobak and ki, 1998). The hippocampal gamma activity is segregated Buzsa into two distinct frequency rhythms, which are believed to rise from 2 distinct sources: in the CA1 nuclei, the slow gamma ranging from 25 to 50 Hz and the fast one from 65 to 140 Hz. The fast gamma oscillations are entrained by the entorhinal cortex, which couple CA1 to inputs from the medial entorhinal cortex, an area that provides information about the animal's position. In contrast, the slow gamma oscillations are generated within the hippocampal network and couple CA1 to CA3, a hippocampal subfield essential for storage of information input from entorhinal cortex (Bragin et al., 1995; Colgin et al., 2009; Csicsvari et al., 2003). Modulation of the slow and fast gamma frequency oscillations plays differential roles during specific cognitive processes. For instance, during the encoding phase of the delayed-match-tosample cognitive task, Vidal et al. (2006) examined the influence of attention-related information processes (subset of item differed in color: attention) and visual grouping of the display (homogenous: grouping condition) on gamma oscillations. The visuoperceptual processes during attention focusing properties of the stimulus required slow gamma frequency oscillations, whereas perceptual grouping of the stimulus modulated the fast gamma oscillations, which argues for functional specialization of both oscillatory rhythms in cognitive processing.

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 13/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Our in vivo findings indicate that cognition enhancers elicited slow gamma (32e48 Hz) oscillations, which may facilitate the interaction between short distance brain networks for attentional and information storage process. It is important to note that all cognition enhancers elicited large regional differences in low theta oscillations being more pronounced in the central and posterior recording sites as compared to the frontal cortex (Fig. 2A). Simultaneous multisite recordings from the hippocampus and the frontal cortex in behaving rats and mice demonstrated that hippocampal theta oscillations control the firing of prefrontal neurons, thereby delivering the temporal coordination of oscillating networks accounting for information transfer and storage (Sirota et al., 2008; Wierzynski et al., 2009). The activating cholinergic and glutamatergic drive directly interact with this oscillatory activity by modulating the hippocampal theta rhythm and shifting the frontal cortical activity from slow to fast oscillations (Metherate et al., 1992; Manns et al., 2000, 2003). Here, all cognition enhancers increased low gamma oscillations in frontal cortex areas (Fig. 3A). This observation suggests that hippocampal theta entrainment of distant frontal cortex neurons into low gamma rhythm represent dynamic functional binding mechanisms of widely distributed cortical assemblies, essential for cognitive processing. In addition, donepezil decreased coherent slow alpha (8e11 activity) in fronto-parieto-occipital recording pairs. The thalamo-cortical circuits generate a large network oscillatory activity in the alpha frequency range during cortical operations. However, this rhythm may also emerge during loss of attention in immobile state or during failure to process sensory input, and may serve to functionally disengage and reduce the processing capabilities of a given brain region. The alpha activity is decreased in engaged brain region whereas it increased in disengaged regions (Klimesch et al., 2007). In general, theta and gamma activities are accompanied by reduced synchrony in network alpha rhythm. Although the controversial views on translatability of alpha oscillatory rhythm between human and rodent, the present results suggest that the temporal enhancement of oscillations and coherent activity in theta and gamma frequencies induced by donepezil may damp cortical alpha activity and engage those cortical regions. However, a common limitation to epidural EEG based connectivity analysis may be biased by volume conduction of signals propagated from a common generators. The model of computation and removal of edges with maximal absolute coherence values occurring at zero time lag described in human recordings (Chu et al., 2012) and used here, likely excluded spurious edges attributable to volume conduction suggesting that epidural network connectivity in the slow theta (4e6.5 Hz) and gamma (32e48 Hz) range are unlikely derived from the same source. These results are particularly interesting and could prove that functional network connectivity measures in epidural recordings may provide sensitive indices of functional interregional network dynamics. 4.3. PCA revealed compact cognition enhancers cluster The principal component analysis captured a common clustering for cognitive enhancing as compared to different drug classes. In the present work, it was noted that the test dataset (rivastigmine) served as a good example to assess the application's abilities and to project exactly to where the reference drug lies in the principal component space. When looking at the distances between our test dataset and the reference drugs, rivastigmine was the first closest drug to acetylcholinesterase inhibitors and the cluster dendrogram indicated that the test drug was correctly

13

placed adjacent to the reference cognition enhancers. It is worth noting that in each experiment which contains a vehicle and three doses (low, middle and high), there was no standardized dosing schemes between experiments and these dosages were not recorded in the data (e.g. 15 mg/kg), nor were standardized to an EC50 or IC50 values. Thus, similar drugs might not be classified together because of a difference in the dosing scheme used. Secondly, the data do not provide indication of the Tmax for each drug. Since each of the variables in the PCA projection are simply independent combinations of time-point, brain region and EEG frequency band, it is possible that two similar compounds would not be clustered together due to a difference in their difference in Tmax. Nevertheless, the PCA identified compact clusters on the two main variables with common spatial patterns for cognition enhancing drugs displayed. The results showed the utility of PCA to keep fundamental features of variables to summarize data into compact clusters that correspond to common underlying mechanisms, and suggest that PCA technique holds a great capacity to successfully reveal oscillatory pattern classification. 4.4. Scopolamine-induced slow EEG network oscillations Scopolamine elicited a consistent leftward shift in the network oscillatory pattern, which was often accompanied by persistent increased coherent activity in slow frequency oscillations. Brain cholinergic neurons are believed to be involved in memory and cognitive functions, and extensive research in both animal and human demonstrated a key causal role of disruption of the cholinergic system in memory and cognitive decline (Klinkenberg and Blokland, 2012). Memory and cognitive functions are known to decline with advancing age, in mild cognitive impairment and AD. A major brain area of neuronal cell loss is the cholinergic nucleus basalis of Meynert (Whitehouse et al., 1982), which sends major cholinergic input to the hippocampus and cortex, both areas of which are highly associated with attention and memory formation and storage. Imaging studies revealed a consistent decreased activity in medial temporal lobe structures such as parahippocampal and hippocampal regions (Celone et al., 2006; Machulda et al., 2003; Bradley et al., 2002; Matsuda, 2001). Diffused and slowed EEG with peak in slow delta/theta network activity, higher coherence in delta and reduction in fast frequency oscillations reflect the major changes in cortical activity seen in patients with senile dementia or AD (Babiloni et al., 2009, 2011; Gianotti et al., 2008; Huang et al., 2000; Jeong, 2004; Lehmann et al., 2007; Ponomareva et al., 2003; Rossini et al., 2006; Yener and Bas¸ar, 2013). This loss of cholinergic drive has formed the basis of the cholinergic hypothesis of AD, which provided the use of scopolamine as a model of cognitive deficit. Scopolamine is a nonselective muscarinic receptor antagonist, which blocks the stimulation of post-synaptic receptors that may consequently impair the activity of hippocampal and cortical fast network oscillations leading to memory deficits associated with enhanced slow EEG oscillations in delta frequency bands (Ebert and Kirch, 1998; re et al., 2009; Dimpfel, 2005; Kikuchi et al., 1999). HowChauvie ever, the scopolamine model presents some limitations as it fails to recreate the cardinal pathological aspects and the progressive degenerative nature of AD (Van Dam and De Deyn, 2006). Despite this caveat, the scopolamine model remains a valuable pharmacological screening tool, which offers opportunities in clinical and translational research for discovery and characterization of potential cognition-enhancing drugs. The present results in rats demonstrate that acute disruption of cholinergic transmission transiently produced abnormalities in cortical network activities similar to those described in neurodegenerative disorders i.e. decreased cortical oscillations and

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 14/17

14

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

networks connectivity in high frequency oscillations, and shifted connectivity towards slow frequency rhythm. The findings that scopolamine-induced disruption of cortical and subcortical oscillations and networks connectivity further support the pivotal role of acetylcholine in the coordination of distributed network activities. The reduced coherence in fast oscillations associated with the shift into peak coherent activity in delta frequency range in response to scopolamine reflects a cortical component of cognitive impairment. 4.5. Cognition enhancers attenuated scopolamine-induced leftward shift of EEG oscillations and peak slow coherent activity The cognition enhancers effectively attenuated the scopolamine-induced aberrant leftward oscillatory shift accompanied with strong coherence in slow delta activity in different brain sites. Acetylcholine system is known to be crucial for triggering theta and gamma oscillation in hippocampal slices (Fellous and Sejnowski, 2000; Traub et al., 2000) and interaction of both rhythms is weakened following muscarinic receptor blockade in mice (Hentschke et al., 2007). Preclinical and clinical experiments were designed to examine the utility of the scopolamine-induced cognitive impairment model in predicting pharmacodynamics signals of putatively pro-cognitive compounds, utilizing the acetylcholinesterase inhibitor (Lenz et al., 2012). The efficacy of AChEIs to reverse slow oscillatory activity in EEG has been demonstrated in AD patients. A consistent reduction of EEG delta and theta oscillations and shift towards normalization was found after short as well as long term treatment with rivastigmine, donepezil and tacrine (Adler and Brassen, 2001; Adler et al., 2004; Balkan et al., 2003; Brassen and Adler, 2003; Kogan et al., 2001). In addition, convergent imaging data demonstrated hyper metabolism activation of brain structures known for their implication in memory and cognitive functions such as fronto-parietal and hippocampal structures (Bohnen et al., 2005; Stefanova et al., 2006). In the present work, all cognition enhancers consistently attenuated scopolamine-induced peak coherent delta activity in fronto-parietal networks. In line with previous results in AD patients showing the efficacy of AChEIs to attenuate the aberrant EEG network shift towards slow oscillations, the shift of EEG oscillations towards high frequency oscillations is presumably suggestive of network neuroprotective activities for cognitive ameliorating effect in the scopolamine model. 4.6. Slow theta (4e6.5 Hz) and gamma (32e48 Hz) oscillations and connectivity: implications for mnemonic processing Mnemonic processing requires a cross-interaction between hippocampal and cortical areas, while increased theta and gamma neuronal synchrony is generally observed during cognitive tasks performed during waking. The temporal synchronization and cross modulation of gamma and theta network oscillations is believed to be a fundamental mechanism by which the brain dynamically binds distributed sets of neurons to form coherent functional assemblies required for the integration of sensory input with information stored in memory (Buzsaki and Draguhn, 2004). The increased EEG oscillations and coherent activity in slow theta (4e6.5 Hz) and gamma (32e48 Hz) frequencies following cognition enhancers adds to the growing evidence that cognition impairments is associated not just with deficits in theta and gamma activity, but pathological networks disconnectivity as well. The current results underscore the sensitivity of global slow theta (4e6.5 Hz) and local gamma (32e48 Hz) oscillations and synchrony

to modulation of regional and local circuit's muscarinic and glutamatergic receptors. In general, theta rhythmicity organizes gamma frequency dynamics of neuronal populations (Bragin et al., 1995; Colgin et al., 2009; Mizuseki et al., 2009; Sirota et al., 2008). The increased slow gamma oscillations during the slow theta-entrained activity may constitute an important aspect of the communication among brain networks to facilitate the coupling of cortical-hippocampal CA3-CA1 networks in order to convey information in the memory stores. These experimental results and theoretical considerations suggest the relevance of slow theta and gamma network oscillations and modulation of these two oscillatory rhythms may, in part, underlie the pro-cognitive actions of the cognition enhancers in the early stages on AD where sufficient numbers of cholinergic and glutamatergic projection neurons are still available. Overall, the present findings are encouraging as they suggest that changes in cortical oscillations and networks connectivity can be relevant in the mechanism of action of cognition enhancers and the preservation of cognitive processes. Functional network connectivity measures in slow EEG delta as well as lower theta and gamma oscillations in conscious animals may prove to be useful efficacy indices for translational research, and provide a major opportunity for drug discovery of drug candidates with potential pro-cognitive efficacy. Uncited references Atri et al., 2008; Babiloni et al., 2004; Babiloni et al., 2013; Birks, n et al., 2012; De Graaf and 2006; Buckner et al., 2005; Carle Bonnard, 2011; Doody et al., 2001; Gray et al., 1989; Huxter et al., 2003; Hyman et al., 1990; Iversen, 1997; Jacobs et al., 2007; Jones and Wilson, 2005; Korotkova et al., 2010; Lee et al., 2005; Li et al., 2009; Mangialasche et al., 2010; Maurer and McNaughton, 2007; O'Keefe and Recce, 1993; Prichep et al., 2006; Principles of Neural Science, 2000; Reeves et al., 2002; Roberson and Mucke, 2006; Tallon-Baudry, 2009; Tallon-Baudry and Bertrand, 1999; Uhlhaas and Singer, 2006; Weingartner, 1985. Acknowledgments The authors are very grateful to Dr. Jason Waddell and Dr. Tom Van De Casteele for their statistical support. References Adler, G., Brassen, S., 2001. Short-term rivastigmine treatment reduces EEG slowwave power in Alzheimer patients. Neuropsychobiology 43, 273e276. Adler, G., Brassen, S., Chwalek, K., Dieter, B., Teufel, M., 2004. Prediction of treatment response to rivastigmine in Alzheimer's dementia. J. Neurol. Neurosurg. Psychiatry 75, 292e294. Atri, A., Shaughnessy, L.W., Locascio, J.J., Growdon, J.H., 2008. Long-term course and effectiveness of combination therapy in Alzheimer' disease. Alzheimer Dis. Assoc. Disord. 22, 209e221. Babiloni, C., Binetti, G., Cassetta, E., Cerboneschi, D., Dal Forno, G., Del Percio, C., Ferreri, F., Ferri, R., Lanuzza, B., Miniussi, C., Moretti, D.V., Nobili, F., PascualMarqui, R.D., Rodriguez, G., Romani, G.L., Salinari, S., Tecchio, F., Vitali, P., Zanetti, O., Zappasodi, F., Rossini, P.M., 2004. Mapping distributed sources of cortical rhythms in mild Alzheimer's disease. A multicentric EEG study. Neuroimage 22, 57e67. Babiloni, C., Frisoni, G.B., Vecchio, F., Pievani, M., Geroldi, C., De Carli, C., Ferri, R., Vernieri, F., Lizio, R., Rossini, P.M., 2009. Global functional coupling of resting EEG rhythms is abnormal in mild cognitive impairment and Alzheimer's disease: a multicenter EEG study. J. Psychophysiol. 23, 224e234. Babiloni, C., Vecchio, F., Lizio, R., Ferri, R., Rodriguez, G., Marzano, N., Frisoni, G.B., Rossini, P.M., 2011. Resting state cortical rhythms in mild cognitive impairment and Alzheimer's disease: electroencephalographic evidence. J. Alzheimers Dis. 3 (26 Suppl. l), 201e214. Babiloni, C., Del Percio, C., Bordet, R., Bourriez, J.L., Bentivoglio, M., Payoux, P., Derambure, P., Dix, S., Infarinato, F., Lizio, R., Triggiani, A.I., Richardson, J.C., Rossini, P.M., 2013. Effects of acetylcholinesterase inhibitors and memantine on

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

Q7

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

NP5579_proof ■ 2 September 2014 ■ 15/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17 resting-state electroencephalographic rhythms in Alzheimer's disease patients. Clin. Neurophysiol. 124, 837e850. ar, A., Yargiçog lu, P., 2003. Effect of Balkan, S., Yaras¸, N., Mihçi, E., Dora, B., Ag donepezil on EEG spectral analysis in Alzheimer's disease. Acta Neurol. Belg 103, 164e169. Birks, J., 2006. Cholinesterase inhibitors for Alzheimer's disease. Cochrane Database Syst. Rev. 1. CD005593. Bohnen, N.I., Kaufer, D.I., Hendrickson, R., Ivanco, L.S., Lopresti, B., Davis, J.G., Constantine, G., Mathis, C.A., Moore, R.Y., DeKosky, S.T., 2005. Cognitive correlates of alterations in acetylcholinesterase in Alzheimer's disease. Neurosci. Lett. 380, 127e132. Bradley, K.M., O'Sullivan, V.T., Soper, N.D., Nagy, Z., King, E.M., Smith, A.D., Shepstone, B.J., 2002. Cerebral perfusion SPET correlated with Braak pathological stage in Alzheimer's disease. Brain 125, 1772e1781. , G., Na dasdy, Z., Hetke, J., Wise, K., Buzsa ki, G., 1995. Gamma Bragin, A., Jando (40e100 Hz) oscillation in the hippocampus of the behaving rat. J. Neurosci. 15, 47e60. Brassen, S., Adler, G., 2003. Short-term effects of acetylcholinesterase inhibitor treatment on EEG and memory performance in Alzheimer patients: an open, controlled trial. Pharmacopsychiatry 36, 304e308. Buckner, R.L., Snyder, A.Z., Shannon, B.J., LaRossa, G., Sachs, R., Fotenos, A.F., Sheline, Y.I., Klunk, W.E., Mathis, C.A., Morris, J.C., Mintun, M.A., 2005. Molecular, structural, and functional characterization of Alzheimer's disease: evidence for a relationship between default activity, amyloid, and memory. J. Neurosci. 25, 7709e7717. Burianov a, H., Ciaramelli, E., Grady, C.L., Moscovitch, M., 2012. Top-down and bottom-up attention-to-memory: mapping functional connectivity in two distinct networks that underlie cued and uncued recognition memory. Neuroimage 63, 1343e1352. ki, G., 2006. Rhythms of the Brain. Oxford Univer. Press, Oxford, UK. Buzsa Buzsaki, G., Draguhn, A., 2004. Neuronal oscillations in cortical networks. Science 304, 1926e1929. ki, G., Watson, B.O., 2012. Brain rhythms and neural syntax: implications for Buzsa efficient coding of cognitive content and neuropsychiatric disease. Dialogues Clin. Neurosci. 14, 345e367. Cantero, J.L., Atienza, M., 2005. The role of neural synchronization in the emergence of cognition across the wake-sleep cycle. Rev. Neurosci. 16, 69e83. Cappaert, N.L., Lopes da Silva, F.H., Wadman, W.J., 2009. Spatio-temporal dynamics of theta oscillations in hippocampal-entorhinal slices. Hippocampus 19, 1065e1077. n, M., Meletis, K., Siegle, J.H., Cardin, J.A., Futai, K., Vierling-Claassen, D., Carle Rühlmann, C., Jones, S.R., Deisseroth, K., Sheng, M., Moore, C.I., Tsai, L.-H., 2012. A critical role for NMDA receptors in parvalbumin interneurons for gamma rhythm induction and behavior. Mol. Psychiatry 17, 537e548. Carlsson, K., Petrovic, P., Skare, S., Petersson, K.M., Ingvar, M., 2000. Tickling expectations: neural processing in anticipation of a sensory stimulus. J. Cogn. Neurosci. 12, 691e703. Celone, K.A., Calhoun, V.D., Dickerson, B.C., Atri, A., Chua, E.F., Miller, S.L., DePeau, K., Rentz, D.M., Selkoe, D.J., Blacker, D., Albert, M.S., Sperling, R.A., 2006. Alterations in memory networks in mild cognitive impairment and Alzheimer's disease: an independent component analysis. J. Neurosci. 26, 10222e10231. Chatfield, C., 1975. The Analysis of Time Series: an Introduction. Chapman and Hall. re, L., Rafrafi, N., Thinus-Blanc, C., Bartolomei, F., Esclapez, M., Bernard, C., Chauvie 2009. Early deficits in spatial memory and theta rhythm in experimental temporal lobe epilepsy. J. Neurosci. 29, 5402e5410. Chrobak, J.J., Buzs aki, G., 1998. Gamma oscillations in the entorhinal cortex of the freely behaving rat. J. Neurosci. 18, 388e398. Chu, C.J., Kramer, M.A., Pathmanathan, J., Bianchi, M.T., Westover, M.B., Wizon, L., Cash, S.S., 2012. Emergence of stable functional networks in long-term human electroencephalography. J. Neurosci. 32, 2703e2713. Cohen, M.X., 2011. Hippocampal-prefrontal connectivity predicts midfrontal oscillations and long-term memory performance. Curr. Biol. 22 (21), 1900e1905. Colgin, L.L., Denninger, T., Fyhn, M., Hafting, T., Bonnevie, T., Jensen, O., Moser, M.B., Moser, E.I., 2009. Frequency of gamma oscillations routes flow of information in the hippocampus. Nature 462, 353e357. Cook, I.A., Leuchter, A.F., 1996. Synaptic dysfunction in Alzheimer's disease: clinical assessment using quantitative EEG. Behav. Brain Res. 78, 15e23. Csicsvari, J., Jamieson, B., Wise, K.D., Buzs aki, G., 2003. Mechanisms of gamma oscillations in the hippocampus of the behaving rat. Neuron 37, 311e322. De Graaf, J.B., Bonnard, M., 2011. In: Peres, Julio F.P. (Ed.), Cognitive Integration in the Human Primary Sensory and Motor Areas: an Overview, Neuroimaging for Clinicians e Combining Research and Practice. http://dx.doi.org/10.5772/23783. Debener, S., Herrmann, C.S., Kranczioch, C., Gembris, D., Engel, A.K., 2003. Top-down attentional processing enhances auditory evoked gamma band activity. Neuroreport 14, 683e686. Dimpfel, W., 2003. Preclinical data base of pharmaco-specific rat EEG fingerprints (tele-stereo-EEG). Eur. J. Med. Res. 8, 199e207. Dimpfel, W., 2005. Pharmacological modulation of cholinergic brain activity and its reflection in special EEG frequency ranges from various brain areas in the freely moving rat (tele-stereo-EEG). Eur. Neuropsychopharmacol. 15, 673e682. Doesburg, S.M., Roggeveen, A.B., Kitajo, K., Ward, L.M., 2008. Large-scale gammaband phase synchronization and selective attention. Cereb. Cortex 18, 386e396. Doody, R.S., Stevens, J.C., Beck, C., Dubinsky, R.M., Kaye, J.A., Gwyther, L., et al., 2001. Practice parameter: management of dementia (an evidence-based review):

15

report of the quality standards subcommittee of the American Academy of Neurology. Neurology 56, 1154e1166. Ebert, U., Kirch, W., 1998. Scopolamine model of dementia: electroencephalogram findings and cognitive performance. Eur. J. Clin. Invest. 28, 944e949. Fell, J., Klaver, P., Lehnertz, K., Grunwald, T., Schaller, C., Elger, C.E., Fernandez, G., 2001. Human memory formation is accompanied by rhinal-hippocampal coupling and decoupling. Nat. Neurosci. 4, 1259e1264. Fellous, J.M., Sejnowski, T.J., 2000. Cholinergic induction of oscillations in the hippocampal slice in the slow (0.5e2 Hz), theta (5e12 Hz), and gamma (35e70 Hz) bands. Hippocampus 10, 187e197. Ferrara, M., Moroni, F., De Gennaro, L., Nobili, L., 2012. Hippocampal sleep features: relations to human memory function. Front. Neurol. 17 (3), 57. Fries, P., 2009. Neuronal gamma-band synchronization as a fundamental process in cortical computation. Annu Rev. Neurosci. 32, 209e224. Fries, P., Neuenschwander, S., Engel, A.K., Goebel, R., Singer, W., 2001. Modulation of oscillatory neuronal synchronization by selective visual attention. Science 291, 1560e1563. Gianotti, L.R., Künig, G., Faber, P.L., Lehmann, D., Pascual-Marqui, R.D., Kochi, K., Schreiter-Gasser, U., 2008. Rivastigmine effects on EEG spectra and threedimensional LORETA functional imaging in Alzheimer's disease. Psychopharmacology 198, 323e332. Gray, C.M., McCormick, D.A., 1996. Chattering cells: superficial pyramidal neurons contributing to the generation of synchronous oscillations in the visual cortex. Science 274, 109e113. Gray, C.M., Konig, P., Engel, A.K., Singer, W., 1989. Oscillatory responses in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties. Nature 338, 334e337. Hammond, C., Bergman, H., Brown, P., 2007. Pathological synchronization in Parkinson's disease: networks, models and treatments. Trends Neurosci. 30, 357e364. Hentschke, H., Perkins, M.G., Pearce, R.A., Banks, M.I., 2007. Muscarinic blockade weakens interaction of gamma with theta rhythms in mouse hippocampus. Eur. J. Neurosci. 26, 1642e1656. Herrmann, C.S., Demiralp, T., 2005. Human EEG gamma oscillations in neuropsychiatric disorders. Clin. Neurophysiol. 116, 2719e2733. Herrmann, C.S., Lenz, D., Junge, S., Busch, N.A., Maess, B., 2004. Memory-matches evoke human gamma-responses. BMC Neurosci. 13 (5), 13. Huang, C., Wahlund, L.O., Dierks, T., Julin, P., Winblad, B., Jelic, V., 2000. Discrimination of Alzheimer's disease and mild cognitive impairment by equivalent EEG sources: a cross-sectional and longitudinal study. Clin. Neurophysiol. 111, 1961e1967. Huxter, J., Burgess, N., O'Keefe, J., 2003. Independent rate and temporal coding in hippocampal pyramidal cells. Nature 425, 828e832. Hyman, B.T., Van Hoesen, G.W., Damasio, A.R., 1990. Memory-related neural systems in Alzheimer's disease: an anatomic study. Neurology 40, 1721e1730. Iversen, S.D., 1997. Behavioural evaluation of cholinergic drugs. Life Sci. 60, 1145e1152. Jackson, C.E., Snyder, P.J., 2008. Electroencephalography and event-related potentials as biomarkers of mild cognitive impairment and mild Alzheimer's disease. Alzheimers Dement. 4 (1 Suppl. 1), S137eS143. Jacobs, J., Kahana, M.J., Ekstrom, A.D., Fried, I., 2007. Brain oscillations control timing of single-neuron activity in humans. J. Neurosci. 27, 3839e3844. Jeong, J., 2004. EEG dynamics in patients with Alzheimer's disease. Clin. Neurophysiol. 115, 1490e1505. Jones, M.W., Wilson, M.A., 2005. Theta rhythms coordinate hippocampal-prefrontal interactions in a spatial memory task. PLoS Biol. 3, 2187e2199. Kaiser, J., Lutzenberger, W., 2003. Induced gamma-band activity and human brain function. Neuroscientist 9, 475e484. Kikuchi, M., Wada, Y., Nanbu, Y., Nakajima, A., Tachibana, H., Takeda, T., Hashimoto, T., 1999. EEG changes following scopolamine administration in healthy subjects. Quantitative analysis during rest and photic stimulation. Neuropsychobiology 39, 219e226. Klimesch, W., Sauseng, P., Hanslmayr, S., 2007. EEG alpha oscillations: the inhibition-timing hypothesis. Brain Res. Rev. 53, 63e88. Klinkenberg, I., Blokland, A., 2010. The validity of scopolamine as a pharmacological model for cognitive impairment: a review of animal behavioral studies. Neurosci. Biobehav. Rev. 34, 1307e1350. Kogan, E.A., Korczyn, A.D., Virchovsky, R.G., Klimovizky, SSh, Treves, T.A., Neufeld, M.Y., 2001. EEG changes during long-term treatment with donepezil in Alzheimer's disease patients. J. Neural Transm. 108, 1167e1173. Korotkova, T., Fuchs, E.C., Ponomarenko, A., von Engelhardt, J., Monyer, H., 2010. NMDA receptor ablation on parvalbumin-positive interneurons impairs hippocampal synchrony, spatial representations, and working memory. Neuron 68, 557e569. Lakatos, P., Chen, C.M., O'Connell, M.N., Mills, A., Schroeder, C.E., 2007. Neuronal oscillations and multisensory interaction in primary auditory cortex. Neuron 53, 279e292. Lakatos, P., Karmos, G., Mehta, A.D., Ulbert, I., Schroeder, C.E., 2008. Entrainment of neuronal oscillations as a mechanism of attentional selection. Science 320, 110e113. Le Van Quyen, M., Staba, R., Bragin, A., Dickson, C., Valderrama, M., Fried, I., Engel, J., 2010. Large-scale microelectrode recordings of high-frequency gamma oscillations in human cortex during sleep. J. Neurosci. 30, 7770e7782. Lee, H., Simpson, G.V., Logothetis, N.K., Rainer, G., 2005. Phase locking of single neuron activity to theta oscillations during working memory in monkey extrastriate visual cortex. Neuron 45, 147e156.

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Q4

NP5579_proof ■ 2 September 2014 ■ 16/17

16

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17

Lehmann, C., Koenig, T., Jelic, V., Prichep, L., John, R.E., Wahlund, L.O., Dodge, Y., Dierks, T., 2007. Application and comparison of classification algorithms for recognition of Alzheimer's disease in electrical brain activity (EEG). J. Neurosci. Methods 161, 342e350. Lenz, R.A., Baker, J.D., Locke, C., Rueter, L.E., Mohler, E.G., Wesnes, K., Abi-Saab, W., Saltarelli, M.D., 2012. The scopolamine model as a pharmacodynamic marker in early drug development. Psychopharmacology 220, 97e107. Li, Y., Umeno, K., Hori, E., Takakura, H., Urakawa, S., Ono, T., Nishijo, H., 2009. Global synchronization in the theta band during mental imagery of navigation in humans. Neurosci. Res. 65, 44e52. Lisman, J., Buzs  aki, G., 2008. A neural coding scheme formed by the combined function of gamma and theta oscillations. Schizophr. Bull. 34, 974e980. Machulda, M.M., Ward, H.A., Borowski, B., Gunter, J.L., Cha, R.H., O'Brien, P.C., Petersen, R.C., Boeve, B.F., Knopman, D., Tang-Wai, D.F., Ivnik, R.J., Smith, G.E., Tangalos, E.G., Jack Jr., C.R., 2003. Comparison of memory fMRI response among normal, MCI, and Alzheimer's patients. Neurology 61, 500e506. Mangialasche, F., Solomon, A., Winblad, B., Mecocci, P., Kivipelto, M., 2010. Alzheimer's disease: clinical trials and drug development. Lancet Neurol. 9, 702e716. Manns, I.D., Alonso, A., Jones, B.E., 2000. Discharge properties of juxtacellularly labeled and immunohistochemically identified cholinergic basal forebrain neurons recorded in association with the electroencephalogram in anesthetized rats. J. Neurosci. 20, 1505e1518. Manns, I.D., Alonso, A., Jones, B.E., 2003. Rhythmically discharging basal forebrain units comprise cholinergic, GABAergic, and putative glutamatergic cells. J. Neurophysiol. 89, 1057e1066. Matsuda, H., 2001. Cerebral blood flow and metabolic abnormalities in Alzheimer's disease. Ann. Nucl. Med. 15, 85e92. Maurer, A.P., McNaughton, B.L., 2007. Network and intrinsic cellular mechanisms underlying theta phase precession of hippocampal neurons. Trends Neurosci. 30, 325e333. Metherate, R., Cox, C.L., Ashe, J.H., 1992. Cellular bases of neocortical activation: modulation of neural oscillations by the nucleus basalis and endogenous acetylcholine. J. Neurosci. 12, 4701e4711. ki, G., 2009. Theta oscillations provide Mizuseki, K., Sirota, A., Pastalkova, E., Buzsa temporal windows for local circuit computation in the entorhinal-hippocampal loop. Neuron 64, 267e280. €lle, M., Marshall, L., Gais, S., Born, J., 2004. Learning increases human electroMo encephalographic coherence during subsequent slow sleep oscillations. Proc. Natl. Acad. Sci. U. S. A. 101, 13963e13968. Montgomery, S.M., Sirota, A., Buzs aki, G., 2008. Theta and gamma coordination of hippocampal networks during waking and rapid eye movement sleep. J. Neurosci. 28, 6731e6741. Oddie, S.D., Bland, B.H., 1998. Hippocampal formation theta activity and movement selection. Neurosci. Biobehav. Rev. 22, 221e231. O'Keefe, J., Recce, M.L., 1993. Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus 3, 317e330. , D., 2008. Theta synchronizes the activity of medial prePaz, R., Bauer, E.P., Pare frontal neurons during learning. Learn. Mem. 15, 524e531. Pesaran, B., Pezaris, J.S., Sahani, M., Mitra, P.P., Andersen, R.A., 2002. Temporal structure in neuronal activity during working memory in macaque parietal cortex. Nat. Neurosci. 5, 805e811. Ponomareva, N.V., Selesneva, N.D., Jarikov, G.A., 2003. EEG alterations in subjects at high familial risk for Alzheimer's disease. Neuropsychobiology 48, 152e159. Prichep, L.S., John, E.R., Ferris, S.H., Rausch, L., Fang, Z., Cancro, R., Torossian, C., Reisberg, B., 2006. Prediction of longitudinal cognitive decline in normal elderly with subjective complaints using electrophysiological imaging. Neurobiol. Aging 27, 471e481. Priestley, M.B., 1981. Spectral Analysis and Time Series, vol. 1&2. Elsevier Academic Press. Principles of Neural Science, fourth ed. In: Kandel, E.R., Schwartz, J.H., Jessell, T.M. (Eds.), 2000. McGraw-Hill, New York; pp. 349e380. Reeves, R.R., Struve, F.A., Patrick, G., 2002. The effects of donepezil on quantitative EEG in patients with Alzheimer's disease. Clin. Electroencephalogr. 33, 93e96. Riedel, W.J., Jolles, J., 1996. Cognition enhancers in age-related cognitive decline. Drugs Aging 8, 245e274. Roberson, E.D., Mucke, L., 2006. 100 years and counting: prospects for defeating Alzheimer's disease. Science 314, 781e784. Rodriguez, G., Nobili, F., Copello, F., Vitali, P., Gianelli, M.V., Taddei, G., Catsafados, E., Mariani, G., 1999. Tc-HMPAO regional cerebral blood flow and quantitative electroencephalography in Alzheimer's disease: a correlative study. J. Nucl. Med. 40, 522e529. Rossini, P.M., Del Percio, C., Pasqualetti, P., Cassetta, E., Binetti, G., Dal Forno, G., Ferreri, F., Frisoni, G., Chiovenda, P., Miniussi, C., Parisi, L., Tombini, M., Vecchio, F., Babiloni, C., 2006. Conversion from mild cognitive impairment to Alzheimer's disease is predicted by sources and coherence of brain electroencephalography rhythms. Neuroscience 143, 793e803. Sannita, W.G., Maggi, L., Rosadini, G., 1987. Effects of scopolamine (0.25e0.75 mg i.m.) on the quantitative EEG and the neuropsychological status of healthy volunteers. Neuropsychobiology 17, 199e205. Saper, C.B., Iversen, S., Frackowiack, R., 2000. Integration of Sensory and Motor Function: the Association Areas of the Cerebral Cortex and the Cognitive Capabilities of the Brain.

Schmidt, B., Hinman, J.R., Jacobson, T.K., Szkudlarek, E., Argraves, M., Escabí, M.A., Markus, E.J., 2013. Dissociation between dorsal and ventral hippocampal theta oscillations during decision-making. J. Neurosci. 33, 6212e6224. Seamans, J.K., Lapish, C.C., Durstewitz, D., 2008. Comparing the prefrontal cortex of rats and primates: insights from electrophysiology. Neurotox. Res. 14, 249e262. Sederberg, P.B., Kahana, M.J., Howard, M.W., Donner, E.J., Madsen, J.R., 2003. Theta and gamma oscillations during encoding predict subsequent recall. J. Neurosci. 23, 10809e10814. €sler, F., 2012. EEG-power and -coherence changes in a Seemüller, A., Müller, E.M., Ro unimodal and a crossmodal working memory task with visual and kinesthetic stimuli. Int. J. Psychophysiol. 83, 87e95. Singer, W., 1999. Neuronal synchrony: a versatile code for the definition of relations? Neuron 24, 49e65. Singh, M., Kaur, M., Korea, H., Chugh, R., Silakari, O., Singh, D., 2013. Acetylcholinesterase inhibitors as Alzheimer therapy: from nerve toxins to neuroprotection. Eur. J. Med. Chem. 70, 165e188. Sirota, A., Montgomery, S., Fujisawa, S., Isomura, Y., Zugaro, M., Buzs aki, G., 2008. Entrainment of neocortical neurons and gamma oscillations by the hippocampal theta rhythm. Neuron 60, 683e697. Q5 Sloan, E.P., Fenton, G.W., Kennedy, N.S., MacLennan, J.M., 1995. Electroencephalography and single photon emission computed tomography in dementia: a comparative study. Psychol. Med. 25, 631e638. Staresina, B.P., Cooper, E., Henson, R.N., 2013. Reversible information flow across the medial temporal lobe: the hippocampus links cortical modules during memory retrieval. J. Neurosci. 33, 14184e14192. € m, B., Stefanova, E., Wall, A., Almkvist, O., Nilsson, A., Forsberg, A., Långstro Nordberg, A., 2006. Longitudinal PET evaluation of cerebral glucose metabolism in rivastigmine treated patients with mild Alzheimer's disease. J. Neural Transm. 113, 205e218. Steriade, M., McCormick, D.A., Sejnowski, T.J., 1993. Thalamocortical oscillations in the sleeping and aroused brain. Science 262, 679e685. Tallon-Baudry, C., 2009. The roles of gamma-band oscillatory synchrony in human visual cognition. Front. Biosci. 14, 321e332. Tallon-Baudry, C., Bertrand, O., 1999. Oscillatory gamma activity in humans and its role in object representation. Trends Cogn. Sci. 3, 151e162. naff, M.A., Isnard, J., Fischer, C., 2005. Attention Tallon-Baudry, C., Bertrand, O., He modulates gamma-band oscillations differently in the human lateral occipital cortex and fusiform gyrus. Cereb. Cortex 15, 654e662. Thatcher, R.W., 2012. Coherence, phase differences, phase shift, and phase lock in EEG/ERP analyses. Dev. Neuropsychol. 37, 476e496. ge  ne tais, E., Glowinski, J., 2000. Hippocampo-prefrontal Thierry, A.M., GioaI, Y., De cortex pathway: anatomical and electrophysiological characteristics. Hippocampus 10, 411e419. Traub, R.D., Whittington, M.A., Buhl, E.H., Jefferys, J.G., Faulkner, H.J., 1999. On the mechanism of the gamma–> beta frequency shift in neuronal oscillations induced in rat hippocampal slices by tetanic stimulation. J. Neurosci. 19, 1088e1105. Q6 Traub, R.D., Bibbig, A., Fisahn, A., LeBeau, F.E., Whittington, M.A., Buhl, E.H., 2000. A model of gamma-frequency network oscillations induced in the rat CA3 region by carbachol in vitro. Eur. J. Neurosci. 12, 4093e4106. Uhlhaas, P.J., Singer, W., 2006. Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophysiology. Neuron 52, 155e168. Uhlhaas, P.J., Singer, W., 2010. Abnormal neural oscillations and synchrony in schizophrenia. Nat. Rev. Neurosci. 11, 100e113. Van Dam, D., De Deyn, P.P., 2006. Drug discovery in dementia: the role of rodent models. Nat. Rev. Drug Discov. 5, 956e970. Vanderwolf, C.H., 1969. Hippocampal electrical activity and voluntary movement in the rat. Electroencephalogr. Clin. Neurophysiol. 26, 407e418. Varela, F., Lachaux, J.P., Rodriguez, E., Martinerie, J., 2001. The brainweb: phase synchronization and large-scale integration. Nat. Rev. Neurosci. 2, 229e239. Verbeke, G., Molenberghs, G., 2000. Linear Mixed Models for Longitudinal Data. In: Springer Series in Statistics. Springer-Verlag, New-York. Vertes, R.P., 2005. Hippocampal theta rhythm: a tag for short-term memory. Hippocampus 15, 923e935. Vidal, J.R., Chaumon, M., O'Regan, J.K., Tallon-Baudry, C., 2006. Visual grouping and the focusing of attention induce gamma-band oscillations at different frequencies in human magnetoencephalogram signals. J. Cogn. Neurosci. 18, 1850e1862. Wagner, T., Axmacher, N., Lehnertz, K., Elger, C.E., Fell, J., 2010. Sleep-dependent directional coupling between human neocortex and hippocampus. Cortex 46, 256e263. Wallis, J.D., 2012. Cross-species studies of orbitofrontal cortex and value-based decision-making. Nat. Neurosci. 15, 13e19. Weingartner, H., 1985. Models of memory dysfunctions. Ann. N. Y. Acad. Sci. 444, 359e369. Wierzynski, C.M., Lubenov, E.V., Gu, M., Siapas, A.G., 2009. State-dependent spiketiming relationships between hippocampal and prefrontal circuits during sleep. Neuron 61, 587e596. Williams, L.M., Gordon, E., 2007. Dynamic organization of the emotional brain: responsivity, stability, and instability. Neuroscientist 13, 349e370. Witt, A., Macdonald, N., Kirkpatrick, P., 2004. Memantine hydrochloride. Nat. Rev. Drug Discov. 3, 109e110.

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6

NP5579_proof ■ 2 September 2014 ■ 17/17

A. Ahnaou et al. / Neuropharmacology xxx (2014) 1e17 Womelsdorf, T., Fries, P., 2006. Neuronal coherence during selective attentional processing and sensory-motor integration. J. Physiol. Paris 100, 182e193. Womelsdorf, T., Schoffelen, J.M., Oostenveld, R., Singer, W., Desimone, R., Engel, A.K., Fries, P., 2007. Modulation of neuronal interactions through neuronal synchronization. Science 316, 1609e1612.

17

Wu, W., Wheeler, D.W., Staedtler, E.S., Munk, M.H., Pipa, G., 2008. Standardized assessment of strategy use and working memory in early mental arithmetic performance. Neuroreport 19, 235e238. Yener, G.G., Bas¸ar, E., 2013. Brain oscillations as biomarkers in neuropsychiatric disorders: following an interactive panel discussion and synopsis. Suppl. Clin. Neurophysiol. 62, 343e363.

Please cite this article in press as: Ahnaou, A., et al., Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology (2014), http://dx.doi.org/10.1016/j.neuropharm.2014.08.015

7 8 9 10 11 12

Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential.

Synchronization of electroencephalographic (EEG) oscillations represents a core mechanism for cortical and subcortical networks, and disturbance in ne...
5MB Sizes 0 Downloads 8 Views