Gamma power abnormalities in a Fmr1-targeted transgenic rat model of fragile X syndrome

Fragile X syndrome (FXS) is characteristically displayed intellectual disability, hyperactivity, anxiety, and abnormal sensory processing. Electroencephalography (EEG) abnormalities are also observed in subjects with FXS, with many researchers paying attention to these as biomarkers. Despite intensive preclinical research using Fmr1 knock out (KO) mice, an effective treatment for FXS has yet to be developed. Here, we examined Fmr1-targeted transgenic rats (Fmr1-KO rats) as an alternative preclinical model of FXS. We characterized the EEG phenotypes of Fmr1-KO rats by measuring basal EEG power and auditory steady state response (ASSR) to click trains of stimuli at a frequency of 10–80 Hz. Fmr1-KO rats exhibited reduced basal alpha power and enhanced gamma power, and these rats showed enhanced locomotor activity in novel environment. While ASSR clearly peaked at around 40 Hz, both inter-trial coherence (ITC) and event-related spectral perturbation (ERSP) were significantly reduced at the gamma frequency band in Fmr1-KO rats. Fmr1-KO rats showed gamma power abnormalities and behavioral hyperactivity that were consistent with observations reported in mouse models and subjects with FXS. These results suggest that gamma power abnormalities are a translatable biomarker among species and demonstrate the utility of Fmr1-KO rats for investigating drugs for the treatment of FXS.

Fragile X syndrome (FXS) is a debilitating neurodevelopmental disorder caused by a CGG repeat expansion mutation in the fragile X mental retardation 1 (FMR1) gene on the X chromosome 1 that results in loss of the fragile X mental retardation protein (FMRP). FXS is estimated to affect 1 in 4000 men and 1 in 8000 women 2 , and patients are characterized by intellectual disability, hyperactivity, anxiety, seizures, autism-like symptoms, and abnormal sensory processing 3,4 . Auditory processing deficits are also a common feature in subjects with FXS [5][6][7][8] . Further, electroencephalography (EEG) abnormalities are observed in 74% of males with FXS, and hyperactivity, a major behavioral symptom of FXS, is observed in 50-66% 9 . EEG abnormalities in subjects with FXS are usually characterized by an enhancement in the amplitude of the N1 of the event-related potential (ERP) in response to auditory stimuli 5,10-12 and basal gamma power 6,13,14 . Furthermore, recent evidence suggests that cortical oscillatory activity contributes to sensory hypersensitivity and social communication deficits in FXS, and that auditory steady state response (ASSR) at gamma frequencies is reduced in FXS 6,13 , the abnormalities that is widely used as a translational biomarker in neuropsychiatric disorders such as schizophrenia 15 and developmental disorders 16 .
Fmr1 knock out (KO) mice have been used as a preclinical model of FXS for more than 20 years 17 . Complete loss of FMRP causes neuronal morphological alterations such as changes to spine shape and density 18,19 , and behavioral abnormalities such as hyperactivity and hypersensitivity to sensory stimuli 20,21 24 . Fmr1-KO rats have been reported to display disrupted cortical processing of auditory stimuli 25 and memory impairment based on hippocampal cellular and synaptic deficits 18 . For drug development in general, rat model can provide beneficial information to evaluate the safety risk of drug candidates compared with mouse model since toxicity studies are usually conducted with rats. However, reports on other characteristics of Fmr1-KO rats remain limited.
In present study, we examined EEG phenotypes, such as basal EEG power and ASSR, in Fmr1-KO rats and conducted a brief behavioral assessment to examine its utility as an alternative preclinical model for drug development in FXS. ASSR is an electrophysiological response entrained to both frequency and phase of rapid, periodic acoustic stimuli 26,27 . The entrainment is generally defined as two indices with time-frequency decomposition; inter-trial coherence (ITC) indicating phase consistency across trials and event-related spectral perturbation (ERSP) indicating event-related alteration of EEG frequency spectrum as a function of time. Accumulated evidence demonstrates the excellent test-retest reliability of ASSR in clinical setting, indicating the use of this method for drug development trial [28][29][30][31] . We employed the methodology of ASSR with ERSP and ITC that we recently established in rodent model 32 .

Results
Fmr1-KO rats display hyperactivity in a novel environment. WT and Fmr1-KO rats displayed a time-dependent decrease in activity counts during the 90-min test period in novel cages (Fig. 1A). In the first 30 min, Fmr1-KO rats showed a significant increase in total activity counts compared with WT rats (Fig. 1B). In contrast, in the 30-60 min and 60-90 min time periods, total activity counts were not significantly different between Fmr1-KO rats and WT rats.

Fmr1-KO rats display enhanced gamma power.
Power spectrum analysis was calculated from baseline power with no sound stimuli. Spectra of absolute power at around 40-80 Hz was higher in Fmr1-KO rats than in WT rats ( Fig. 2A). Absolute power at the gamma frequency band (30-80 Hz) was significantly increased in Fmr1-KO rats compared with WT rats, while that at the alpha frequency band (8-12 Hz) tended to be decreased (p = 0.05) (Fig. 2B). Relative power at the gamma frequency band was significantly increased in Fmr1-KO rats compared with WT rats (Fig. 2C). In contrast, relative power at alpha and beta frequency bands (12-30 Hz) was significantly decreased in Fmr1-KO rats. Absolute and relative power at delta (0.5-4 Hz) and theta frequency bands (4-8 Hz) were not significantly different between Fmr1-KO and WT rats.

Gamma synchronization of ITC and ERSP is reduced in Fmr1-KO rats. ITC analysis showed that
auditory click-train stimuli elicited frequency-dependent responses from 10 to 80 Hz in both Fmr1-KO and WT rats (Fig. 3A). In the gamma frequency band (30-80 Hz) for ASSR, ITC reached a maximum at around 40-50 Hz in both Fmr1-KO and WT rats (Fig. 3B). ITC was significantly decreased at 30, 40, 50, 60, and 70 Hz in Fmr1-KO rats compared with WT rats. Meanwhile, there were no significant differences in ITC at 10, 20, 80 Hz between genotypes (Fig. 3C). In ERSP analysis, auditory click-train stimuli similarly elicited frequency- Data represent mean ± SEM (n = 13-14). **P < 0.01, significant differences between groups; unpaired t-test.  (Fig. 4A). Fmr1-KO rats showed lower ERSP than WT rats, although a peak response was observed at 40-50 Hz within the gamma frequency band in both Fmr1-KO and WT rats (Fig. 4B). ERSP was significantly decreased at 30, 40, 50, and 60 Hz in Fmr1-KO rats compared with WT rats. In contrast, there was no significant difference in ERSP at 10, 20, 70 and 80 Hz between Fmr1-KO and WT rats (Fig. 4C).

Discussion
While the behavioral and neurophysiological characterizations of Fmr1 KO mice have been extensively replicated, those of Fmr1-KO rats remain limited. The present study is the first to describe the neurophysiological phenotypes of Fmr1-KO rats. First, we assessed basal EEG profiles in Fmr1-KO rats. Fmr1-KO rats displayed significantly augmented baseline gamma power compared to WT but decreased alpha power. This pattern of baseline EEG power in Fmr1-KO rats is consistent with findings in subjects with FXS, who show augmented power in the gamma frequency band, but reduced power in the alpha frequency band 8,13 . This augmented gamma power is also consistent with studies in Fmr1 KO mice 22,33 , which display augmented neuronal excitability related to alterations in input to fast spiking inhibitory interneurons synchronized to the gamma frequency band 34,35 . Alpha activity reportedly produces bouts of inhibition that repeat every 100 ms, and these alpha oscillations modulate gamma activity driven by GABAergic inhibitory activity in sensory information processing [36][37][38] . This pattern of increased gamma and decreased alpha power suggests the presence of impaired gamma activity with involvement from GABAergic inhibitory neurons in Fmr1-KO rats. In this study, we did not observe any alterations in power in the theta frequency band, although power in the beta frequency band was reduced in Fmr1-KO rats. Subjects with FXS  (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30-80 Hz) frequency bands in WT and Fmr1-KO rats. (C) Relative power at delta, theta, alpha, beta, and gamma frequency bands in WT and Fmr1-KO rats. Data represent mean ± SEM (n = 13-14). *P < 0.05, **P < 0.01, significant differences between groups; unpaired t-test.

Scientific Reports
| (2020) 10:18799 | https://doi.org/10.1038/s41598-020-75893-x www.nature.com/scientificreports/ show enhanced basal theta power and no change in power in the beta frequency band 8,13 . Resting power in the theta and beta frequency bands are not changed in Fmr1 KO mice 22 . The discordance in theta and beta frequency bands between rodents and subjects with FXS is important to note and requires further study. Nevertheless, our finding that basal power is enhanced in the gamma frequency band in a preclinical rat model of FXS is consistent with the EEG phenotypes observed in Fmr1 KO mice and subjects with FXS. Enhanced neuronal excitability is associated with behavioral symptoms such as increased anxiety and locomotor activity 9,39,40 , some of the most consistent behavioral symptoms observed in subjects with FXS 41 . The present study found that Fmr1-KO rats showed enhanced locomotor activity for the first 30 min after placement in a novel environment but not at 30-60 min or 60-90 min. Studies using Fmr1 KO mice have reported that the hyperactivity was a reaction to novelty 42,43 . Our results from Fmr1-KO rats suggest that the hyperactivity may be a reaction to novelty, because the significant augmentation in locomotor activity was only observed for the first 30 min after placement in a novel testing field. Interestingly, startle response to sounds of collision between pedestal and top cage was observed in Fmr1-KO rats during measurement of locomotor activity only in Fmr1-KO not wild type rats (data not shown). Sensory hypersensitivity is a common phenotype in subjects   www.nature.com/scientificreports/ demonstrated increased basal power in the gamma frequency band. A similar impairment of the phase-locking factor of ASSR at only 30-58 Hz was observed in subjects with FXS, with phase-locking abnormalities reported to be associated with increased gamma single-trial power 14 . Further, there was also clear impairment of ERSP at 30-60 Hz in Fmr1-KO rats. We used a previously developed ASSR paradigm that adopted click sounds as stimuli in freely moving rats 32 , with a similar method having been used in subjects with schizophrenia 49 . However, no robust impairment of ERSP in the gamma frequency band has been observed in subjects with FXS or Fmr1 KO mice. It is important to note that these studies used "chirp" stimuli to elicit ASSR 14,22 . Although the differences associated with using "click" and "chirp" stimuli for eliciting ASSR remain elusive, the click auditory stimulus has been reported to induce large effect sizes and deficits in ASSR evoked by 40-Hz click stimulation have been used as a translational biomarker for schizophrenia 15,50,51 . These findings may support the use of click-stimulievoked ASSR as a robust method for determining ITC and ERSP in Fmr1-KO rats. Taken together, the specific attenuation of ASSR at 30-60 Hz in Fmr1-KO rats as it is in Fmr1 KO mice and subjects with FXS suggests that ASSR in the gamma frequency band is a biomarker for FXS. Fmr1-KO rats exhibited ASSR attenuation and enhanced basal power specific to gamma frequency band, these similar EEG features were observed in Fmr1 KO mice and subjects with FXS 13,14,22 . Preclinical research in Fmr1 KO mice has shown that increased gamma excitability decreases excitatory drive in fast-spiking inhibitory interneurons, resulting in increased and poorly synchronized pyramidal cell firing in the gamma frequency range 52 . Fmr1-KO rats show deficits of switching process from an elevated gamma state to a reduced gamma state with insufficiently synchronization related to firing rate of fast-spiking interneurons in visual cortex, and disrupted cortical processing of auditory stimuli 25,53,54 . These findings suggest that gamma power abnormalities with gamma band-ASSR attenuation and augmented baseline gamma power could attribute to imbalance between excitation and inhibition of the neural network in underlying mechanism of FXS.

Methods
Animals. Five-week-old Fmr1-KO and wild type (WT) rats were purchased from Oriental Yeast Co., Ltd.
(Tokyo, Japan). Fmr1-KO rats were generated using the ZFN method 21 , and the lines were first generated on an outbred Sprague-Dawley background at Sage Laboratories, LLC.
Surgery. Surgical operations were conducted in 13 WT and 14 Fmr1-KO male rats at 7-8 weeks of age to implant 4 electrodes under anesthesia with 2-2.5% isoflurane. A recording electrode was embedded onto the surface of the cortex using a customized switchable pedestal and electrode cables (S.E.R. Corporation, Tokyo, Japan). This paradigm using a switchable pedestal was previously developed to enable ECoG recordings to be taken from the parietal and temporal cortex in freely moving rats 32 . Briefly, the recording electrodes were placed at AP − 1.0 mm/ML − 1.0 mm (parietal cortex) and AP − 4.5 mm/ML − 7.5 mm/VD − 4.0 mm (temporal cortex) relative to bregma. Reference and ground electrodes were placed at AP 8.0 mm/ML − 1.5 mm and AP − 10.0 mm/ML − 1.5 mm, respectively. The switchable pedestal with stainless steel wire was attached to the cranium using methacrylic resin (GC Corporation, Tokyo, Japan). Rats were allowed to recover for approximately 14 days before testing. The rats were housed in groups until electrode implantation and then alone after implantation in ventilated cages under a 12-h light/dark cycle with food and water available ad libitum. All animal experimental procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals, 8th edition and approved by the Institutional Animal Care and Use Committee of Astellas Pharma Inc. Furthermore, Astellas Pharma Inc., Tsukuba Research Center was awarded Accreditation Status by the AAALAC International.
Spontaneous locomotor activity. Locomotor activity was evaluated in rats implanted with a switchable pedestal at 9 weeks old. First, the rats were acclimated to the test room under 1 lx of red right. After 1 h of acclimation, test animals were moved from their home cages to new cages exposed to over 300 lx of light. Motor activity was measured for 90 min using a Supermex sensor (Muromachi inc., Tokyo, Japan) comprising paired infrared pyroelectric detectors that measure radiated body heat, and analyzed using CompACT AMS, ver. 3.82 (Muromachi Inc.). These data were divided into 3 time periods (0-30, 30-60, 60-90 min) and the total count was calculated for each time period. EEG recording. EEG experiments were performed according to previously reported methods 32 . Electrophysiological activity was recorded inside a Faraday cage with speakers attached to the top of the cage, using a high-impedance differential AC amplifier (model #1800, A-M Systems, Carlsbrog, WA, USA) and Spike 2 (Cambridge Electronic Design; CED, Cambridge, UK) with CED1401 (CED). All EEG experiments were performed in dim light (< 300 lx). All animals were implanted with electrodes connected to an amplifier via electrode cables during habituation and recording, and were allowed to move freely. Data from Spike2 were converted using EEGLAB in the MATLAB toolbox (Math Works, Natick MA, USA). The EEG measurements were conducted at 10 weeks of age in all animals.