Two novel mouse models mimicking minor deletions in 22q11.2 deletion syndrome revealed the contribution of each deleted region to psychiatric disorders

22q11.2 deletion syndrome (22q11.2DS) is a disorder caused by the segmental deletion of human chromosome 22. This chromosomal deletion is known as high genetic risk factors for various psychiatric disorders. The different deletion types are identified in 22q11.2DS patients, including the most common 3.0-Mb deletion, and the less-frequent 1.5-Mb and 1.4-Mb deletions. In previous animal studies of psychiatric disorders associated with 22q11.2DS mainly focused on the 1.5-Mb deletion and model mice mimicking the human 1.5-Mb deletion have been established with diverse genetic backgrounds, which resulted in the contradictory phenotypes. On the other hand, the contribution of the genes in 1.4-Mb region to psychiatric disorders is poorly understood. In this study, we generated two mouse lines that reproduced the 1.4-Mb and 1.5-Mb deletions of 22q11.2DS [Del(1.4 Mb)/+ and Del(1.5 Mb)/+] on the pure C57BL/6N genetic background. These mutant mice were analyzed comprehensively by behavioral tests, such as measurement of locomotor activity, sociability, prepulse inhibition and fear-conditioning memory. Del(1.4 Mb)/+ mice displayed decreased locomotor activity, but no abnormalities were observed in all other behavioral tests. Del(1.5 Mb)/+ mice showed reduction of prepulse inhibition and impairment of contextual- and cued-dependent fear memory, which is consistent with previous reports. Furthermore, apparently intact social recognition in Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice suggests that the impaired social recognition observed in Del(3.0 Mb)/+ mice mimicking the human 3.0-Mb deletion requires mutations both in 1.4-Mb and 1.5 Mb regions. Our previous study has shown that Del(3.0 Mb)/+ mice presented disturbance of behavioral circadian rhythm. Therefore, we further evaluated sleep/wakefulness cycles in Del(3.0 Mb)/+ mice by electroencephalogram (EEG) and electromyogram (EMG) recording. EEG/EMG analysis revealed the disturbed wakefulness and non-rapid eye moving sleep (NREMS) cycles in Del(3.0 Mb)/+ mice, suggesting that Del(3.0 Mb)/+ mice may be unable to maintain their wakefulness. Together, our mouse models deepen our understanding of genetic contributions to schizophrenic phenotypes related to 22q11.2DS. Supplementary Information The online version contains supplementary material available at 10.1186/s13041-021-00778-7.

Forty-five protein cording genes are located within the 3.0-Mb deleted region and 37 of them are conserved in the mouse chromosome 16qA13, making it possible to generate a mouse model of 22q11.2DS. To date, five lines of 22q11.2DS mouse model have been generated [15][16][17][18][19], four of which mimicked the 1.5-Mb deletion (LCR22A-B deletion). Thus, the contributions of the genes located on the remaining 1.4-Mb region (LCR22B-D in 1.4-Mb deletion) to this syndrome are poorly understood. Although this 1.4-Mb deletion is included in 3.0-Mb deletion of Del(3.0 Mb)/+ mouse model mimicking the most common 3.0-Mb deletion in the human 22q11.2 locus, the influence of 1.4-Mb deletion alone on psychiatric phenotypes has not been studied in this model [19]. Individuals with this 1.4-Mb deletion including LCR22B-D and LCR22C-D deletion have been diagnosed as ADHD, anxiety disorder, developmental delay and intellectual disability [6]. Therefore, it is possible that key genes located in this 1.4-Mb deleted region are responsible for the development of psychiatric disorders. Furthermore, repeated behavioral analyses of the other 22q11.2DS mouse models show contradicting results depending on the experimental paradigms. In this regard, it is known that a difference in the genetic background gives rise to variable phenotypic expression, making direct comparisons among different models challenging [20,21]. For example, previous 1.5-Mb deletion mouse models, except for Df(h22q11)/+, were generated by introducing the deletion into the donor chromosome from different strains. Even though they were backcrossed with C57BL/6 strain afterwards, the donor chromosome from original strains remains, leaving skepticism about direct comparisons of phenotypes and accurate phenotypic evaluation of each gene deletion. The advent of CRISPR/Cas9 system enabled the generation of mutants in a pure genetic background by CRISPR/Cas9mediated genome rearrangement [22][23][24][25]. In the present study, we generated two mouse lines that reproduced the 1.4-Mb deletion and 1.5-Mb deletion in 22q11.2DS and characterized their behavioral phenotypes. Furthermore, our previous study has shown the deficits of PPI and disturbance of behavioral circadian rhythm in Del(3.0 Mb)/+ mice [19]. For further analysis, we evaluated the rescuing effects of the anti-psychotic drug on the reduced PPI and sleep/wakefulness cycles related to circadian rhythm using Del(3.0 Mb)/+ mice.

Animals
Animals were housed under a 12 h light/dark cycle (light on at 8:00, off at 20:00) with maintained temperature (23 ± 1 °C) and humidity (55 ± 10%) and ad libitum access to food and water. Animal experiments were approved by the Institutional Animal Care and Use Committee of The University of Tokyo and conducted in accordance with the guideline of The University of Tokyo.

Preparation of single-guide RNAs (sgRNAs), Cas9 mRNA and single-stranded oligodeoxyribonucleotide (ssODN)
To design sgRNA with smaller number of off-target sites, CRISPRdirect software was used (http:// crispr. dbcls. jp; [26]). Two pairs of sgRNAs positioned on either endpoint of the deletion were designed. The protospacer sequences of sgRNAs were listed in Additional file 1 (Table S1). These protospacers were cloned into BsaI-pDR274 vector (Addgene #42250). The sgRNAs were transcribed in vitro using the DraI-digested pDR274 vectors as a template and the MEGAshortscript T7 kit (Ambion, CA, USA) and purified using MEGAclear kit (Ambion). The Cas9 mRNA was transcribed in vitro using a Message-Max T7 ARCA-Capped Message mRNA transcription kit (Cellscript, WI, USA), polyadenylated with an A-plus Poly(A) Polymerase Tailing kit (Cellscript) and purified using MEGAclear kit (Ambion). The ssODNs designed to bridge the deletion endpoints were 120 nucleotides in length and positioned directly adjacent to the most external sgRNA site. The sequences of ssODNs were listed in Additional file 2 (Table S2).

Genotyping of genome engineered mice by PCR assay
Genomic DNA was extracted from mouse tail tips to serve as templates for genotyping PCR assay. PCR products were amplified with Ex Taq DNA polymerase (Takara Bio Inc., Shiga, Japan). The PCR primers for genotyping of Del(1.4 Mb)/+ mice were as follows: 5′-AAC CAC AGG GGT GGA AAG TC-3′ and 5′-TGC TAG CCT GCA TCG TAA GG-3′ (552 bp product, the deletion band); 5′-CAC TTG TCA ACT GAC TAC TGT TTG -3′ and 5′-GAG CCA GGT CTA GGA ACG TC-3′ (654 bp product, the internal control). The following primers were used for genotyping of Del(1.5 Mb)/+ mice: 5′-CTA AGG AAT GGT TCC GGC CA-3′ and 5′-TTT CAC GGA GGC GGT ATT CA-3′ (424 bp product, the deletion band); 5′-GAG AAA GTG GGT GGG AAG GC-3′ and 5′-GTC CCT CGC CAC AGT CAT AA-3′ (532 bp product, the internal control). The reaction for detection of 1.4-Mb deletion was conducted under the following conditions: initial denaturation at 98 °C for 2 min, 30 cycles of melting at 98 °C for 30 s, annealing at 64 °C for 30 s, and extension at 72 °C for 30 s, with additional extension at 72 °C for 2 min at the end. The condition of 1.5-Mb deletion was as follows: 98 °C for 2 min, 30 cycles of 98 °C for 30 s, 65 °C for 30 s, and 72 °C for 30 s, with additional extension at 72 °C for 2 min at the end. PCR products were analyzed in 2% agarose gel electrophoresis and the sequences were confirmed by Sanger sequencing (Fasmac Co., Kanagawa, Japan).

Array comparative genomic hybridization (array CGH) analysis
Array CGH analysis was performed according to previously reported methods [10]. In briefly, an Agilent Sure-Print G3 Mouse CGH 4 × 180 K Microarray (Agilent, Santa Clara, CA, USA) was used according to the manufacturer's instructions. Copy number variation (CNV) calls were made with Nexus Copy Number software v9.0 (BioDiscovery, El Segundo, CA, USA) using the Fast Adaptive States Segmentation Technique 2 (FASST2) algorithm, which is a hidden Markov model-based approach. The log 2 ratio threshold for copy number loss (or deletion) was set at − 0.4. The significance threshold to adjust the sensitivity of the segmentation algorithm was set at 1 × 10 −6 , and at least five contiguous probes were required for CNV calls. Genomic locations are reported in NCBI Build 37/UCSC mm9 coordinates.

Quantitative RT-PCR
Total RNA was isolated from hippocampi of male mice using miRNeasy Mini Kit (QIAGEN, MD, USA). For quantification of mRNA, first strand cDNA was synthesized using PrimeScript RT reagent Kit (Perfect Real Time) (Takara Bio Inc.). Briefly, 300 ng of total RNA was reverse transcribed using 25 pmol of oligo dT primer and 50 pmol of random 6 mer in 10 μL reaction. The resultant cDNA was diluted at 1:30 ratio in TE. All samples within an experiment were reverse transcribed at the same time. All real-time PCR reactions were performed using the PowerUp SYBR Green Master Mix (Applied Biosystems, CA, USA) and the StepOnePlus Real-Time PCR system (Applied Biosystems). The experiments were conducted in triplicate for each data point. The relative quantification in gene expression was determined using ∆∆Ct method [27]. The sequences of the primers are listed in Additional file 3 (Table S3).

Behavioral tests
All behavioral tests were carried out with male mice at the age of 8-11 weeks. We used Del(1.4 Mb)/+ and Del(1.5 Mb)/+ lines which were backcrossed to C57BL/6N for 3 generations. Del(3.0 Mb)/+ mice was backcrossed to C57BL/6N for 4 generations. Prior to each experiment, mice were placed in the testing room for at least 1 h to acclimate to the experimental environments. The experimenter was blind to genotype throughout the experimental procedures. All apparatus used in behavioral tests were cleaned with 70% ethanol and wiped with paper towels between each session. For the characterization of behavioral phenotypes in Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice, we conducted a series of behavioral tests consisted of the following testing paradigms: (1) open field test, (2) five-trial social interaction test, (3) prepulse inhibition, and (4) contextual and cued fear conditioning test. We set at least 1 day of rest between each testing paradigm.

Open field test
Mice were placed in the center of the open field arena (diameter: 75 cm, height: 35 cm) and were allowed to explore freely the arena for the following 10 min under moderately light conditions (15 lx). Their movement was recorded with a camera mounted above the arena, and their activity was measured automatically using Smart V3.0 tracking software (Panlab, Barcelona, Spain). The open field was divided into an inner circle (diameter: 50 cm), and an outer area surrounding the inner circle. Measurements included total distance moved, time spent in the outer zone and number of transitions between inner and outer section.

Five-trial social interaction test
Five-trial direct social interaction test was performed as described previously [19]. Subject mice were placed individually into home cage (45 cm × 28 cm × 16 cm) for 1 h before starting test under the moderately light conditions (15 lx). A juvenile intruder mouse (5-week-old) was introduced into the subject mouse's home cage. The subject mouse was exposed to the same intruder mouse for 5 min over 4 trials with an inter-trial-interval of 30 min. During the fifth trial, the subject mouse was exposed to a novel intruder mouse (5-week-old). The time spent in social interaction (close following, inspection, anogenital sniffing and other social body contacts) was recorded.

Prepulse inhibition (PPI)
The prepulse inhibition (PPI) test was carried out by using SR-Lab system (San Diego Instruments, San Diego, CA, USA) as described previously [19]. After each mouse was placed in an enclosure (12.7 cm, 3.8 cm inner diameter) under moderately bright light condition (180 lx), they were acclimated for 10 min in the presence of background white noise (65 dB). The movement of the animal in the startle chamber was measured by a piezoelectric accelerometer mounted under the enclosure at the sampling rate of 1 kHz. Individual mouse received 20 startle trials, 10 no-stimulus trials and 40 PPI trials. The intertrial interval was between 10 and 20 s. The startle trial consisted of a single 120 dB white noise burst lasting 40 ms. PPI trials consisted of a prepulse (20 ms burst of white noise at 69, 73, 77 or 81 dB intensity) followed by the startle stimulus (120 dB, 40 ms white noise) 100 ms later. Each of the four prepulse trials (69, 73, 77 or 81 dB) was carried out 10 times. Five consecutive startle trials were presented at the beginning and end of the session. The remainder of sixty different trials were performed pseudorandomly to ensure that each trial was done 10 times and that no two consecutive trials were identical. During the session, 65 dB background white noise was continually present. The largest amplitude in the recording window was taken as the startle amplitude for the trial. Basal startle amplitude was determined as the mean amplitude of the 10 startle trials. PPI (%) was calculated as follows: 100 × (pulse-alone response − prepulse-pulse response)/pulse-alone response, in which prepulse-pulse response was the mean of the 10 PPI trials (69, 73, 77 or 81 dB) and pulse-alone response was the basal startle amplitude.

Contextual and cued fear conditioning test
The fear conditioning test was conducted using ImageJ FZ1 (O'Hara & Co., Ltd., Tokyo, Japan) as described previously [19]. The conditioning chamber was a square arena (10 cm × 10 cm × 10 cm) with clear Plexiglas walls and a metal grid floor connected to a circuit board that delivered electric shocks to the metal grid. A video camera was set in front of the cage to record the behavior. In the conditioning session, mice were individually placed into the conditioning chamber and allowed to explore freely for 3 min. After 3 min exploratory period, each mouse was exposed to two tone-footshock pairings (tone, 30 s; footshock, 2 s, 0.8 mA at the termination of the tone; separated by 1 min intertrial interval). One min after the second footshock, the mouse was returned to its home cage. Twenty-four h after conditioning, the context-dependent test was performed, in which each mouse was placed back into the conditioning chamber, and the freezing response was measured for 6 min in the absence of the conditioned stimulus. Forty-eight h after the footshock, each mouse was tested for auditory (tone) fear conditioning in a novel opaque chamber. Different environmental cues (e.g. light condition and background noise) were provided in the novel chamber. Mice were tested in the novel chamber for a 3 min baseline period (pre-tone) followed by another 3 min for the conditioning tone during which the tone was presented persistently for 3 min. Total freezing rate was measured as an index of fear memory. Motionless bouts lasting more than 2 s were considered as freeze.

Electroencephalogram (EEG) and electromyogram (EMG) analysis
Sleep analysis was performed as described previously [28]. Eight-to ten-week-old mice were subjected to EEG/ EMG electrode implantation surgery. The surgery was performed under isoflurane anesthesia (4% for induction, 2% for maintenance). The scalp was incised along the midline to expose the cranium. Four holes were generated on the skull using 1.0-mm drill bits (anteroposterior: 0.50 mm, lateral: ± 1.27 mm and anteroposterior: − 4.53 mm, lateral: ± 1.27 mm from bregma). The four electrode pins were lowered onto the dura under stereotaxic control (David Kopf Instruments, #940/926) and fixed using dental cement (3 M ESPE, Ketac Cem Aplicap). Subsequently, two EMG wires were inserted into the neck extensor muscles and covered with dental cement. A 6-pin header dissected from 2 × 40 pin header (Useconn Electronics Ltd., #PH-2 × 40SG) was inserted to the top of the EEG/EMG electrode into close the holes of the insulator.
Seven days after surgery, the mice were attached to a tether cable and singly housed in a recording cage (19.1 × 29.2 × 12.7 cm). The tether cable was hung by a counterbalanced lever arm (11.4 cm-long, Instech Laboratories, #MCLA) that allows the mice to move freely. All mice were allowed at least 5 days of recovery from surgery and habituation to the recording conditions for at least 4 days. The floor of the cage was covered with aspen chips and nest materials. To examine sleep/wake behavior under baseline conditions, the EEG/EMG signal was recorded for three consecutive days from the onset of the light phase.
EEG/EMG signals were amplified, filtered (EEG: 0.5-100 Hz; EMG: 0.5-300 Hz) with a multi-channel amplifier (NIHON KODEN, #AB-611J), and digitized at a sampling rate of 250 Hz using an analogue-to-digital converter (National Instruments, #PCI-6220) with LabVIEW software (National Instruments). The EEG/ EMG data were visualized and semi-automatically analyzed by MATLAB-based software followed by visual inspection. Each 20-s epoch was staged into wakefulness, non-rapid eye movement (NREM) sleep and REM sleep. Wakefulness was scored based on the presence of low amplitude, fast EEG activity and high amplitude, variable EMG activity. NREM sleep was characterized by high amplitude, delta (1-4 Hz)-frequency EEG waves and a low EMG tonus, whereas REM sleep was staged based on theta (6-9 Hz)-dominant EEG oscillations and EMG atonia. The total time spent in wakefulness, NREM sleep, and REM sleep were derived by summing the total number of 20-s epochs in each state. Mean episode durations were determined by dividing the total time spent in each state by the number of episodes of that state. Epochs that contained movement artifacts were included in the state totals but excluded from the subsequent spectral analysis. EEG signals were subjected to fast Fourier transform analysis from 1 to 30 Hz with 1-Hz bins using MATLABbased custom software. The EEG power density in each frequency bin was normalized to the sum of 16-30 Hz in all sleep/wake state.

Statistical analysis
The significance of differences (p < 0.05) was assessed by two-tailed Welch's t-test for comparison of two groups. In multiple comparison, the nonparametric Kruskal-Wallis test was used because the homogeneity of variance was violated by Bartlett test. For the analysis of variance (ANOVA) with two factors (two-way repeated measure ANOVA), normality of sample distribution was assumed, and followed by Bonferroni post hoc analysis. Differences in frequency distribution were evaluated using the Kolmogorov-Smirnov test. All data are expressed as the mean ± standard error of the mean (SEM). The detailed statistics methods and values were described at Additional file 4 (Table S4).

Generation of Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
To generate two mutant lines with partial deletions in 22q11.2DS (Fig. 1a), we conducted CRISPR/Cas9mediated chromosome editing using C57BL/6N mouse zygotes as previously described [19,22]. To introduce the deletion between Phosphatidylinositol 4-kinase alpha (Pi4ka) and DiGeorge syndrome critical region 2 (Dgcr2) on mouse chromosome 16, we designed a pair of sgR-NAs on each target locus of Pi4ka intron 47 and Dgcr2  (Fig. 1b). Then, two pairs of sgRNAs, hCas9 mRNA and a bridging ssODN were co-injected into the cytoplasm of 838 C57BL/6N zygotes, and 432 viable 2-cell embryos were transplanted into recipient ICR female mice (Additional file 5: Table S5). We obtained a total of 23 pups from manipulated embryos. Nine out of 23 pups died before weaning. Genotyping PCR assay of 14 viable founder candidates were conducted (Fig. 1d) and the deletion were confirmed by direct sequencing of the PCR products (Fig. 1h). Six offspring harbored the desired structural variants (Additional file 5: Table S5). We used founder #3 to establish a mutant line [hereinafter referred to as Del(1.4 Mb)/+]. The deletion allele of the founder was successfully transmitted through the germline (Fig. 1e).
To generate a mouse with a deletion in the targeted region from Dgcr2 to Histone Cell Cycle Regulator (Hira) [hereinafter referred to as Del(1.5 Mb)/+], we used a pair of sgRNAs on each locus of Dgcr2 intron 4 and Hira intron 23 and an ssODN bridging the deleted region (Fig. 1c). We obtained a total of 10 pups from manipulated embryos and 2 pups harbored the desired mutation (Fig. 1f, i and Additional file 5: Table S5). The deletion allele of the founder #2 was successfully transmitted through the germline (Fig. 1g).

The decrease in genomic copy number and mRNA expression of genes within target regions in Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
To confirm a genomic copy number loss on chromosome 16qA13 in our two mutant lines, we performed array CGH analysis. The expected decrease in genomic copy number of target region was confirmed in both Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice ( Fig. 2a and b).
Next, to confirm the decreased expression of the deleted genes in the two 22q11.2DS mouse lines, we performed quantitative RT-PCR. We found that expression of the genes in the deleted region between Pi4ka and Dgcr2 was significantly reduced to approximately 50% in Del(1.4 Mb)/+ mice compared with WT littermates (Fig. 2c). In Del(1.5 Mb)/+ mice, expression of the genes, except for Dgcr8, in the deleted region between Dgcr2 and Hira was reduced to approximately 50% of the expression in WT mice (Fig. 2d). Expression of Dgcr8 was reduced to 67.6%, possibly due to negative feedback regulation by the Microprocessor complex containing DGCR8 itself [29]. In addition, expression of Carbonic anhydrase 15 (Car15) that is adjacent to Dgcr2 was reduced to 43.1% in Del(1.5 Mb)/+ mice, which is consistent with the previous expression analysis of Df(h22q11)/+ mice [30].

General locomotor activity in Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
Several studies have conducted ethological analysis of the 22q11.2DS mouse models. However, it is not clear whether genes located in 1.4-Mb region contribute to the disease development. Therefore, we characterized the behavioral phenotypes of our two mutant lines. Firstly, to examine the general locomotor activity of Del(1.4 Mb)/+ mice, we conducted an open field test for 10 min. The total distance moved of Del(1.4 Mb)/+ mice was significantly shorter than that of WT littermates (Fig. 3a), indicating the hypoactivity in Del(1.4 Mb)/+ mice. This result is consistent with previous finding of Del (3.0 Mb)/+ mice [19]. There was no significant difference in the zone preference between two genotypes (Fig. 3b), indicating that the anxiety-like behavior of Del(1.4 Mb)/+ mice was normal. In Del(1.5 Mb)/+ mice, we found the decreased mean value of total distance moved as compared to WT littermates, although there is no statistically significant difference between Del(1.5 Mb)/+ and WT littermates (Fig. 3c). Also, there is no significant difference in time in outer zone as compared to WT littermate controls (Fig. 3d). Our finding of apparent normal locomotor activity in Del(1.5 Mb)/+ mice is consistent with Df1/+ mice which are one of the 22q11.2DS mouse models with the 1.5-Mb deletion [31] and inconsistent with other 1.5-Mb deletion models, Df(16)A +/− and Lgdel/+ mice, which have shown hyperactivity [17,32,33].

Sociability and social memory in Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
Next, to assess sociability and social memory in the two mutant mouse lines, we performed five-trial social interaction test. In this test, a test mouse was exposed to the same intruder mouse for four successive trials (trial 1-4). On the fifth trial, the test mouse was exposed to a novel intruder mouse (trial 5). There was no significant difference in sociability (trial 1-4) and social memory (trial 5) between Del(1.4 Mb)/+ mice and WT littermates (Fig. 4a). Similarly, we found no significant difference in the interaction time to intruders between Del(1.5 Mb)/+ mice and WT littermates (Fig. 4b).

Sensorimotor gating of Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
Prepulse inhibition (PPI) of the auditory startle response is often used as a translatable measure of the sensorimotor gating that is considered to be the endophenotypic marker of psychiatric conditions such as schizophrenia [34][35][36][37]. Moreover, previous reports performing a PPI analysis showed that reduced PPI is observed in 3.0-Mb and 1.5-Mb deletion mouse models [17, 19, 30-32, 38, 39]. To investigate whether the deletion of genes on the 1.4-Mb region is responsible for the impairment of sensorimotor gating, we conducted auditory PPI analysis in Del(1.4 Mb)/+ mice. Del (1.4 Mb)/+ mice showed a tendency to decrease PPI, but there was no statistically significant difference between Del(1.4 Mb)/+ mice and WT littermates (Fig. 5a). The amplitude of startle response to the loud sound was comparable between Del(1.4 Mb)/+ mice and WT littermates (Fig. 5b). Next, we evaluated PPI of Del(1.5 Mb)/+ mice. In line with the previous studies, PPI was significantly reduced in Del(1.5 Mb)/+ mice compared with WT littermates (Fig. 5c). There was no significant difference in the amplitude of startle response between Del(1.5 Mb)/+ mice and WT littermates (Fig. 5d).

Effects of acute haloperidol administration on PPI deficits in Del(3.0 Mb)/+ mice
In our previous study, Del (3.0 Mb)/+ mice have exhibited the PPI deficit [19]. In the present study, we investigated whether the PPI deficit observed in Del(3.0 Mb)/+ mice was rescued by the typical anti-psychotic drug haloperidol, dopamine D2 receptor antagonist. We found no significant effect of acute haloperidol administration on PPI deficits in Del(3.0 Mb)/+ mice (Additional file 6: Figure  S1a). Likewise, there was no significant effect of haloperidol on startle response between Del(3.0 Mb)/+ groups (Additional file 6: Figure S1b).

Associative learning and memory of Del(1.4 Mb)/+ and Del(1.5 Mb)/+ mice
A fear-conditioning test has repeatedly revealed a dysfunction of associative learning and memory in 22q11.2DS mouse models [17, 31-33, 40, 41]. To assess the function of learning and memory in our 22q11.2DS models, we conducted contextual-dependent and auditory cued-dependent fear-conditioning test. There was no significant difference in the freezing time both in contextual-dependent and cued-dependent fear conditioning tests between Del(1.4 Mb)/+ and WT littermates (Fig. 6a), indicating the apparently normal learning and memory in Del(1.4 Mb)/+ mice. On the other hand, as shown in the previous 1.5-Mb deletion models, Del(1.5 Mb)/+ mice showed decreased freezing time both in contextual-dependent and cued-dependent tests (Fig. 6b).

Sleep/wakefulness analysis of Del(3.0 Mb)/+ mice by EEG/ EMG recording
In our previous study, Del (3.0 Mb)/+ mice have shown lower behavioral activity in their subject night, suggesting the disturbed intrinsic circadian behavioral rhythm [19]. To investigate the details of their circadian rhythm changes, we conducted electrophysiological analysis of sleep/wakefulness cycles by EEG/EMG recording using Del(3.0 Mb)/+ mice. We found that Del(3.0 Mb)/+ mice showed no distinct changes in the hourly pattern during the light phase in all sleep and wake stages. However, there was reduction in wakefulness and increased NREMS time during the first few hours immediately after the start of dark period with statistical significance at Zeitgeber time (ZT) 14 ( Fig. 7a and b), but no difference in the REMS between Del(3.0 Mb)/+ mice and WT (Fig. 7c). These changes in the hourly pattern of wakefulness and NREMS did not reflect the cumulative time spent in each stage during light and dark phases (Fig. 7d). Episode duration of each stage was not significantly different between Del(3.0 Mb)/+ and WT littermates ( Fig. 7e). There were no significant differences in the REM latency and interval ( Fig. 7f and g) and no significant difference in the transition among wakefulness, NREMS and REMS (Fig. 7h). In EEG spectral analysis, EEG power densities during wakefulness and NREMS did not significantly differ between Del(3.0 Mb)/+ and WT littermates ( Fig. 7i and j). Meanwhile, EEG power density in the theta frequency range (6-12 Hz) during REMS was significantly reduced in Del(3.0 Mb)/+ mice as compared with WT littermates (Fig. 7k).

Discussion
22q11.2DS is known to be implicated in the onset of various psychiatric disorders. Since 22q11.2 deletion has the highest risk of schizophrenia, this syndrome has been considered as one of the most important genetic risk factors of schizophrenia. In 22q11.2DS, different deletion types have been identified. Approximately 90% of 22q11.2DS patients have the deletion of 3.0-Mb range, and the deletions of 1.5-Mb and 1.4-Mb regions are much less common. To date, most of animal studies associated with 22q11.2DS have focused on the 1.5-Mb deletion. Several genes encoded in the 1.5-Mb region (e.g., Tbx1, Dgcr8, Comt, Sept5, and Prodh) have been identified as potentially associated with psychiatric disorder-like behavioral phenotypes [42]. The contribution of the 1.4-Mb deletion for the onset of psychiatric disorders has been suggested in clinical studies [7,[9][10][11], and Crkl (CRK-like Proto-Oncogene, Adaptor Protein) heterozygous-deficient mice (Crkl +/− ) have been analyzed to investigate the contributions of Crkl gene to behavioral phenotypes relevant to psychiatric disorders [43]. However, the functional role of other genes located in 1.4-Mb region was poorly understood. In the present study, we generated two mouse lines with different deletion Our open field test showed that Del(1.4 Mb)/+ mice exhibited lower locomotor activity in a novel environment. This hypoactive phenotype is consistent with that of Del(3.0 Mb)/+ mice [19]. On the other hand, the general locomotor activity of Del(1.5 Mb)/+ mice was normal. These results suggest that the deletion of 1.4-Mb region, but not the deletion of 1.5-Mb region, is responsible for the hypoactive phenotype. However, the Lgdel/+ and Df(16)A +/− mice which are 22q11.2DS mouse models with the 1.5-Mb deletion displayed hyperactivity. This phenotypic discrepancy between our Del(1.5 Mb)/+ model and other 1.5-Mb deletion models may derive from different murine genetic backgrounds. The Lgdel/+ and Df(16)A +/− mice are not congenic models [17,33], while our Del(1.5 Mb)/+ model is a coisogenic strain generated by CRISPR/Cas9 system using C57BL/6N embryos. In line with this, the Df(h22q11)/+ model established in C57BL/6N congenic strain has displayed normal locomotion in the dark and bright conditioned open field task [44]. Regarding the anxiety-related behavior, Crkl +/− mice spent less time in the margin area of the open field than Crkl +/+ mice for 30 min test period, suggesting that Crkl heterozygous mutation act as a protective factor for phobia-like behavior [43] The impairment in basal sociability or facial recognition has been observed in 22q11.2DS patients [45][46][47][48][49][50]. In our experiments, sociability and social memory were apparently intact in Del(1.4 Mb)/+ mice, suggesting that social behavior is not affected by the deletion of genes within 1.4-Mb region. Del(1.5 Mb)/+ mice also showed apparently normal sociability and social memory. Interestingly, our previous study showed that the social recognition is impaired in Del(3.0 Mb)/+ mice [19]. Together, these results suggest that the impairment of social recognition requires multiple mutations both in 1.4-Mb and 1.5-Mb regions.
Our PPI analysis showed that the sensorimotor gating of Del(1.4 Mb)/+ mice were apparently intact. This result suggests that the 1.4-Mb deletion does not affect the sensorimotor gating. On the contrary, the reduced PPI was observed in Del(1.5 Mb)/+ mice, which is consistent with Df1/+, Df (16)A +/− , Df(h22q11)/+ and Del(3.0 Mb)/+ mice [17, 19, 30-32, 38, 39]. Therefore, genes located within the 1.5-Mb region is highly likely to be responsible for the defective sensorimotor gating observed in 22q11.2DS mouse models. In fact, Dgcr8 located in 1.5-Mb region is a strong candidate gene for the impaired PPI. Dgcr8 +/− mice showed decreased PPI, which was rescued by administration of haloperidol [51]. However, in our study, haloperidol did not rescue the PPI impairment in Del(3.0 Mb)/+ mice. This inconsistent result could be attributable to different drug administration period. While our Del(3.0 Mb)/+ mice were treated with single-shot haloperidol 30 min before the test, Dgcr8 +/− mice have been continuously treated with haloperidol for 3-4 weeks [51], suggesting that continuous administration of haloperidol is required to rescue the PPI impairment instead of acute administration. On the other hand, the hearing loss arising from otitis media was observed in Df1/+ mice [52]. Therefore, the reduction of PPI might be caused by a hearing loss, and haloperidol administration was ineffective on PPI deficits in Del(3.0 Mb)/+ mice.
In our fear-conditioning test, contextual-and cueddependent learning of Del(1.4 Mb)/+ mice was apparently normal, indicating that learning and memory are not affected by the 1.4-Mb deletion. On the other hand, Del(1.5 Mb)/+ mice, as expected, displayed the impairment in contextual-and cued-dependent learning, which is consistent with Df1/+ , Df(16)A +/− and Lgdel/+ mice [17, 31-33, 40, 41]. These findings support the theory that fear memory impairment is caused by the deletion of 1.5-Mb region but not the deletion of 1.4-Mb region. Consistent with this, mice with the conditional deletion of Dgcr8 within 1.5-Mb region in thalamic neurons have exhibited the associative fear memory deficits and the disrupted synaptic transmission at thalamic input to the lateral amygdala (LA) [41]. Hence, it is possible that the fear memory deficits of Del(1.5 Mb)/+ mice were caused by decreased Dgcr8 expression.
EEG/EMG recording revealed that wakefulness and NREMS patterns were significantly disturbed in Del(3.0 Mb)/+ mice immediately after the start of dark period; wakefulness reduced and NREMS increased at ZT14 compared with WT littermates. Del(3.0 Mb)/+ mice managed to wake up at ZT12 just like WT littermates, indicating that their response to light stimulation during light-to-dark transition is intact, but Del(3.0 Mb)/+ mice are unable to maintain their wakefulness after that. On the other hand, this disturbance of hourly patterns during the first few hours immediately after the start of dark phase has not observed in behavioral activity-based circadian rhythm study of Del(3.0 Mb)/+ mice [19]. This discrepancy may arise from the differences of the recording environment and detectors between the EEG/EMG-based and behavioral activity-based experiments. Therefore, it is difficult to simply compare these results. In contrast to wakefulness and NREMS stages, there was no significant difference in the REMS pattern between Del(3.0 Mb)/+ and WT littermates. As for REMS regulation, Goosecoid-like (Gscl), a homeobox transcription factor, and Dgcr14 are located within the 3.0-Mb region, and their absence is known to result in REMS disturbances [53]. Gscl −/− mice concurrently lacked the expression of Dgcr14 in the interpeduncular nucleus and showed abnormal REMS pattern [53]. Given that our Del(3.0 Mb)/+ mice have a heterozygous deletion of Gscl and Dgcr14 gene with the other allele intact, it is reasonable to consider that these genes are haploinsufficient in the context of REMS regulation, thus REMS of Del(3.0 Mb)/+ mice stayed unaffected. On the other hand, our EEG spectral analysis showed reduction in theta power during REMS of Del(3.0 Mb)/+ mice, which is congruous with Gscl −/− mice [53], suggesting that Gscl and Dgcr14 are haploinsufficient in the context of theta power regulation during REMS.
In summary, we generated two novel mouse models mimicking 1.4-Mb or 1.5-Mb deletions seen in 22q11.2DS patients on pure C57BL/6 genetic background. These mice were comprehensively analyzed to determine the deletion regions responsible for psychiatric disorder-like phenotypes observed in Del(3.0 Mb)/+ mice. The hypoactive phenotype is caused by deletion of the genes in 1.4 Mb region, while genes within 1.5 Mb regions are responsible for impairment of sensorimotor gating and fear memory. The impairment of social recognition requires mutations both in 1.4 Mb and 1.5 Mb regions. We also found the disturbed wakefulness and NREMS cycles in Del(3.0 Mb)/+ mice. In conclusion, this study using our model mice shed light on the genetic contribution to the diverse symptoms of 22q11.2DS.