To generate Vip ^tTA mice, we inserted a tTA2-polyA cassette at the start codon of Vip gene in its second exon by the CRISPR/Cas9-mediated targeting strategy (Figure 1A). The donor DNA was synthesized, containing tTA2 cDNA (Urlinger et al., 2000), SV40 polyA signal, and 300 bp sequences of the mouse Vip gene (NCBI Gene: 22353) 5′ and 3′ to the start codon. C57BL/6J mice were purchased from Japan SLC, Inc. (Shizuoka, Japan). One crRNA targeting exon 2 (5′- TCT​TTT​CAG​AGG​CAC​CGA​GA -3′) of the mouse Vip gene was designed using the online sgRNA design tool available at https://crispr.mit.edu/ and purchased from Integrated DNA Technologies (Coralville, IA). The fertilized pronuclear-stage embryos were prepared by in vivo fertilization in human tubal fluid medium (ARK Resource; Kumamoto, Japan) with sperms from two C57BL/6J males and oocytes from five superovulated females injected with anti-inhibin serum and human chorionic gonadotropin. Next, the ssODN (40 ng/μl) and the complex of crRNAs (0.61 μM), tracrRNA (0.61 μM), and Cas9 protein (30 ng/μl) (Integrated DNA Technologies) in TE were injected into the nucleus of approximately two hundred pronuclear-stage embryos by microinjection. Embryos were then washed and cultured in potassium simplex optimization medium (ARK Resource) overnight. The obtained ninety-three 2-cell embryos were transferred to recipient mice. Fourteen mice were born and tested for the correct targeting of the tTA2-polyA cassette by genomic PCR and sequencing. Six out of 14 mice (F0) had a correct insertion of the cassette, and two F0 mice were bred with C57BL/6J mice to obtain F1 generation. These two lines backcrossed to C57BL/6J mice at least twice were used for further analyses. We did not discriminate between these two lines in this manuscript because they were indistinctive regarding the tTA2 expression and circadian behavior. To evaluate the specific expression of tTA2, Vip ^tTA mice were crossed to Actb-tetO-EGFP reporter mice. Actb-tetO-EGFP reporter mice were generated from Actb-tetO-FLEX-EGFP mice (Aida et al., 2015) by crossing with germ cell-specific Prdm1-Cre mice (JAX 008827) to eliminate the Cre-dependency of EGFP expression.

Avp-Cre mice were reported previously (Mieda et al., 2015) and used in hemizygous condition. This line is a transgenic mouse harboring a modified BAC transgene, which has an insertion of codon-improved Cre recombinase gene immediately 5′ to the translation initiation codon of exogenous Avp gene in the BAC, but without manipulation of the endogenous Avp loci in the mouse.

All mice were maintained under a strict 12 h light/12 h dark (LD) cycle in a temperature- and humidity-controlled room and fed ad libitum. All experimental procedures involving animals were approved by the appropriate institutional animal care and use committees of Kanazawa University and Tokyo Medical and Dental University.

Immunostaining was performed as described previously (Mieda et al., 2015). Mice were sacrificed at approximately ZT5 by transcardial perfusion of PBS followed by 4% paraformaldehyde fixative. To examine the specificity of tTA2 expression by fluorescent immunostaining (in Figures 1C,D), Vip ^wt/tTA ;Actb-tetO-EGFP mice were pretreated with intracerebroventricular colchicine injections (40 µg in 1 µl saline) for 48 h before perfusion fixation. Serial coronal brain sections (30 µm thick) were prepared with a cryostat (CM 1860, Leica) and collected in 4 series—one of which was further immunostained. The antibodies used were: rabbit anti-GFP (Thermo Fisher Scientific, 1:1,000), rabbit anti-AVP (Millipore, 1:4,000); rabbit anti-VIP (Immunostar, 1:1,000); biotinylated goat anti-rabbit IgG antibody (Vector Lab, 1:1,000), and Alexa 488-conjugated goat anti-rabbit IgG (Molecular Probes, 1:1,000). The expression of EGFP was detected by its native fluorescence for fluorescent immunostaining (Figures 1C,D). The VIP expression levels were quantified by Photoshop (Adobe) as follows (Figure 2B). First, the images were transformed to grayscale. Then, the mean intensities of pixels within the SCN were calculated. Finally, the values of the region lateral to the SCN were regarded as background and were subtracted from those of the SCN. For immunohistochemistry using diaminobenzidine (DAB) reactions (Figure 1B), color development was performed using the VECTASTAIN Elite ABC-HRP Kit (PK6100, Vector Lab) and DAB Substrate Kit (SK4100, Vector Lab). For Figures 1A,C,D representative optical section was imaged from each stained section by laser-confocal microscopy (Olympus, FluoView FV10i), then fluorescent cells in the images were counted. For Figures 1B, 2A, immunostaining of EGFP or VIP in sections was observed by epifluorescence or bright-field microscopy (KEYENCE, BZ-9000E).

The AAV-2 ITR-containing plasmid pAAV-TRE-ChR2-EYFP (Addgene #110339, a gift from Dr. Hyungbae Kwon) was modified to construct pAAV-TRE-jGCaMP7s by replacing a BamHI-HindIII fragment containing hChR2-EYFP cDNA with a BamHI-HindIII fragment containing jGCaMP7s cDNA amplified by PCR from the plasmid pGP-AAV-CAG-FLEX-jGCaMP7s-WPRE (Addgene #104495, a gift from Dr. Douglas Kim & GENIE Project), using the following primers: 5′- atagga​tccgcc​acc​ATG​ggt​tct​cat​ca -3′ and 5′- gcgaag​ctTCAc​ttc​gct​gtc​atc​att​tg -3′. pAAV-EF1a-DIO-mCherry was a gift from Dr. Bryan Roth (Addgene #50462).

Recombinant AAV vectors (AAV2-rh10), AAV-TRE-jGCaMP7s and AAV-EF1α-DIO-mCherry, were produced using a triple-transfection, helper-free method and purified as described previously (Mieda et al., 2015). The titers of recombinant AAV vectors were determined by quantitative PCR: AAV-TRE-jGCaMP7s, 6.3 × 10¹¹; and AAV-EF1α-DIO-mCherry, 5.2 × 10¹² genome copies/ml. Stereotaxic injection of AAV vectors into the SCN of Vip ^wt/tTA ;Avp-Cre ⁺ mice was performed as described previously (Maejima et al., 2021). Two weeks after surgery, coronal brain slices (100 µm thick) were prepared, and fluorescence of jGCaMP7s and mCherry was observed as described above.

Male and female, Vip ^wt/wt , Vip ^wt/tTA , and Vip ^tTA/tTA littermates, aged 8–17 weeks, were individually housed in a polycarbonate cage placed in a light-tight box. Spontaneous movements in the homecage were monitored by infrared sensors (O’hara) in 1-min bins as described previously (Mieda et al., 2015). Actogram, activity profile, and χ² periodogram analyses were performed via ClockLab (Actimetrics). The free-running period and amplitude (Qp values) were calculated for the last 10 days in constant darkness (DD) by periodogram. The activity onset was calculated from the daily activity profile (average pattern of activity) of the last 7 days in LD using the mean activity level as a threshold.

All results are expressed as mean ± SEM. For comparison of three or four groups in Figures 2, 4, one-way ANOVA followed by Tukey post hoc tests were performed. Probability (p) values less than 0.05 were considered to be statistically significant. Only relevant information from the statistical analysis was indicated in the text and figures.

We generated knock-in mice that express tTA2 (Urlinger et al., 2000) specifically in VIP neurons. To do this, we employed the CRISPR/Cas9-mediated homologous recombination to target the Vip gene of the mouse genome to insert a tTA2-polyA cassette at the start codon of Vip gene in its exon 2 (Vip ^tTA ) (Figure 1A). To localize tTA2 activity, we crossed Vip ^wt/tTA mice to Actb-tetO-EGFP reporter mice, which express EGFP in the presence of tTA. EGFP expression was mostly restricted in the SCN and the cerebral cortex in these mice (Figure 1B). The cerebral cortex contains a population of GABAergic interneurons that expresses VIP (Taniguchi et al., 2011). In contrast, there were few EGFP + cells in other regions where some neurons express VIP, such as the hippocampus and amygdala. In the SCN, EGFP expression was almost completely colocalized with VIP immunoreactivity (83.76 ± 0.91% of VIP + cells were also EGFP+, 87.11 ± 1.47% of EGFP + cells were also VIP+; counts of three SCN slices each from three mice) (Figure 1C). In contrast, there was almost no overlap with AVP immunoreactivity (2.45 ± 0.59% of AVP + cells were also EGFP+, 5.21 ± 1.16% of EGFP + cells were also AVP+) (Figure 1D). Thus, the expression of tTA2 occurred specifically and efficiently in VIP neurons within the SCN, confirming that Vip ^tTA mice are a useful tool for VIP-neuron-specific genetic manipulations.

Next, we examined whether Vip ^tTA mice are useful to simultaneously apply different genetic manipulations to VIP and AVP neurons nearby within the SCN. To do so, we crossed Vip ^wt/tTA mice to hemizygous Avp-Cre mice (Avp-Cre ⁺ ) that express improved Cre recombinase specifically in AVP neurons (Mieda et al., 2015). Then, we focally injected two AAV vectors in the SCN of Vip ^wt/tTA ;Avp-Cre ⁺ mice: AAV-TRE-jGCaMP7s and AAV-EF1α-DIO-mCherry that express a green fluorescent Ca²⁺ indicator protein jGCaMP7s (Dana et al., 2019) and a red fluorescent protein mCherry in a tTA- and Cre-dependent manner, respectively. As expected, jGCaMP7s was expressed specifically in the SCN shell, where AVP neurons locate, whereas mCherry expression was restricted in the SCN core, where VIP neurons locate (Figure 1E). Furthermore, there was almost no overlap in the expression of these two proteins. Thus, we successfully labeled VIP and AVP neurons differentially within the SCN local circuit.

The Vip coding sequence was interrupted by a tTA2-polyA sequence in the Vip ^tTA allele, resulting in the expression of tTA2 under the Vip promoter. In other words, the Vip ^tTA allele should act equivalently to a Vip knockout allele. Therefore, we next examined the VIP expression in heterozygous Vip ^wt/tTA and homozygous Vip ^tTA/tTA mice by immunostaining. We found that VIP-immunoreactivity in Vip ^wt/tTA mice was slightly reduced (∼20%) compared to wild-type mice (Vip ^wt/wt ) but was completely absent in Vip ^tTA/tTA (Figure 2). These results suggested that the insertion of a tTA2-polyA sequence at the translation initiation site completely blocks the expression of VIP peptide.

Previous studies of locomotor activity rhythm in Vip ^−/− mice have shown significantly attenuated circadian rhythmicity in constant darkness (DD), either with a shortened free-running period or multiple circadian periods (Colwell et al., 2003; Aton et al., 2005). A small number of Vip ^−/− mice even demonstrate arrhythmicity. Therefore, we next recorded the daily rhythm of spontaneous locomotor activity of Vip ^wt/wt , Vip ^wt/tTA , and Vip ^tTA/tTA mice. In these experiments, mice were individually housed in cages, first in a 12 h/12 h light/dark (LD) cycle for 10–15 days. Under this condition, all groups were entrained to the LD cycle and exhibited nocturnal rhythms in their activity (Figure 3).

Mice were then placed into DD, and their free-running rhythms were recorded. Under these conditions, several striking differences emerged in Vip ^tTA/tTA mice compared to Vip ^wt/wt and Vip ^wt/tTA mice. All Vip ^wt/wt and Vip ^wt/tTA mice free-ran with a single, stable circadian period (Figures 3A,B). The free-running periods were comparable between Vip ^wt/wt (23.89 ± 0.03 h) and Vip ^wt/tTA mice (23.82 ± 0.02 h) (Figure 4B). In contrast, among Vip ^tTA/tTA mice, one-half (5 of 12) exhibited multiple circadian periods or no statistically significant circadian period (type-A) (Figure 3C). The remaining half (7 of 12) free-ran with a single circadian period (type-B) (Figure 3D). However, the period of type-B Vip ^tTA/tTA mice (23.30 ± 0.16 h) was significantly shorter than that of Vip ^wt/wt mice (Figure 4B).

Type-A Vip ^tTA/tTA mice (Qp 1997.28 ± 61.92) also showed circadian amplitude significantly lower than that of Vip ^wt/wt mice (Qp 3,733.60 ± 306.41) in DD (Figure 4A). In contrast, there was no significant difference between type-B Vip ^tTA/tTA (Qp 3,840.01 ± 443.14) and Vip ^wt/wt mice. In addition, the circadian amplitude of Vip ^wt/tTA (Qp 3,706.82 ± 193.27) mice was not significantly different from Vip ^wt/wt mice (Figure 4A). The mean daily profiles of locomotor activity rhythm also demonstrated the attenuation of circadian rhythm of type-A Vip ^tTA/tTA mice both in LD and DD (Figures 4C,D). In contrast, type-B Vip ^tTA/tTA mice increased their locomotor activity during CT12∼18 (evening component) while they reduced that during CT21∼3 (morning component) in DD (Figure 4D). In LD, the activity onset was ∼1.5 h earlier in type-A and ∼0.8 h in type-B Vip ^tTA/tTA mice compared to other groups (Figures 3C,D and Figures 4C,D).

Based on the above data, we concluded that the circadian behavior rhythm of Vip ^wt/tTA mice was not significantly different compared to Vip ^wt/wt mice. However, Vip ^tTA/tTA mice were deficient in VIP expression and exhibited a variety of impairments in their circadian behavioral rhythms. The circadian phenotypes of Vip ^tTA/tTA mice were all consistent with the previous reports concerning Vip ^−/− mice (Colwell et al., 2003; Aton et al., 2005).