Fig 1.
Design of the Ai162-nCTG mouse line and its transmission to progeny.
(A) Schematic representation of the procedure used to generate Ai162-nCTG using the PITCh system. Ai162-nCTG expresses nuclear-localized mCherry (nuc-mCherry), the TVA receptor, and RG in a Cre-dependent manner. (B) Top: Schematic showing the targets of genotyping primers. Bottom: Representative results of gel electrophoresis. (C) Ratio of each mouse genotype. Males heterozygous for Cre (+/Cre) and females heterozygous for nCTG (+/nCTG) were crossed. The dotted line indicates the expected Mendelian ratio. n = 6 mothers for ChAT-Cre, OT-Cre, vGluT2-Cre, vGAT-Cre, and OTR-iCre. n = 12 mothers for AVP-Cre. (D) Ratio of males and females for the mice harboring both Cre and nCTG shown in (C). The dotted line indicates the expected ratio.
Fig 2.
Specificity and efficiency of Ai162-nCTG and polydipsia and polyuria exhibited by mice double-positive for AVP-Cre and Ai162-nCTG.
(A) Representative coronal image of the PVH from a double heterozygous mouse harboring AVP-Cre and nCTG. Green and magenta represent AVP mRNA and mCherry mRNA, respectively. Blue, DAPI. Scale bar, 50 μm. (B) Distribution of mean pixel intensity of RFP channel in the PVH of +/AVP-Cre; +/nCTG mice. Each ROI corresponds to a cell identified by DAPI signal. Note that each brain was analyzed without staining with antibodies. Despite the variability of mCherry expression level, mCherry+ neurons (magenta ROIs) can be identified from their fluorescence (see the interval between the two arrows). n = 140 ROIs from 7 mice. (C) Cumulative probability of mean pixel intensity of RFP channel. “AAV” corresponds to the mice harboring only AVP-Cre (+/AVP-Cre) that received AAV-FLEx-TVA-mCherry (n = 70 ROIs from 7 mice), while “nCTG” was calculated from the data shown as magenta in (B). The p-value is shown in the panel (Kolmogorov–Smirnov test). (D) The number of AVP-expressing (AVP+) neurons in the PVH was smaller in +/AVP-Cre; +/nCTG mice compared with mice harboring only AVP-Cre (n = 5 mice each for P7 and P30, n = 9 mice each for 8-weeks-old). Significant difference was found in P30 and 8-week-old mice (*p < 0.05, **p < 0.01, ***p < 0.001, two-tailed Mann-Whitney U-test). (E) Approximately half of the AVP+ neurons were labeled with mCherry in +/AVP-Cre; +/nCTG mice, while almost all the mCherry+ neurons co-expressed AVP (n = 9 mice each). We did not find significant difference between males and females (two-tailed Mann-Whitney U-test). (F) Representative images of the urine-marked area (white). Scale bar, 2 cm. (G) Both male and female +/AVP-Cre; +/nCTG mice displayed a larger fraction of the area covered by urine (n = 8 each and 7 each for males and females, respectively; *p < 0.05, ***p < 0.001, two-tailed Welch’s t-test). (H) Water intake over 24 h was significantly larger in +/AVP-Cre; +/nCTG mice (n = 7 mice each; **p < 0.01, two-tailed Welch’s t-test). (I) Food intake did not show a significant difference (n = 7 mice each). Error bars, SEM.
Fig 3.
Comparison of the distributions of presynaptic neurons visualized by the nCTG- or AAV-based approach.
(A, B) Schematic of virus injection (A) and experimental time line (B). Female mice were used in this experiment. Mice double-positive for AVP-Cre and Ai162-nCTG (+/AVP-Cre; +/nCTG) received only rabies virus (RV) injection. In the AAV-based approach (AAVs), a mixture of AAVs expressing TVA and RG was injected into the PVH of AVP-Cre (+/AVP-Cre; +/+) mice prior to RV injection. (C) Representative coronal sections containing starter cells, defined as the overlap of mCherry (magenta) and GFP (green), in the PVH. Blue, DAPI. Scale bar, 30 μm. (D) The number of starter cells in the PVH was significantly larger in the conventional AAV-based approach (**p < 0.01, two-tailed Welch’s t-test). (E) Representative coronal sections containing starter cells in the SO and SCH. Scale bar, 30 μm. (F) The number of starter cells in each region of +/AVP-Cre; +/nCTG mice (**p < 0.01, one-way ANOVA with repeated measurements with post hoc Tukey’s HSD). (G) Representative coronal sections of presynaptic cells revealed by rabies-GFP. Abbreviations follow the Allen Mouse Brain Atlas [36]. Scale bars, 50 μm. (H) Normalized inputs from various nuclei to the PVH AVP neurons. The convergence index was defined as the number of labeled presynaptic neurons in each brain region normalized to the number of starter cells. The convergence index of nCTG-based approach was not significantly different from that of AAV-based approach (p > 0.34, two-way ANOVA). n = 6 and 7 mice for nCTG and AAVs, respectively. Error bars, SEM.
Fig 4.
Cell-type characterization of presynaptic neurons by in situ hybridization.
(A–C) Left, representative coronal sections showing GFP expression by rabies virus (green) and Calcr (A), Pmch (B), or Agrp (C) expression (magenta) using AAV-based approach. Blue, DAPI. Scale bars, 50 μm. Right, the fraction of GFP-positive cells co-expressing each marker gene. For Calcr, n = 6 each; for Pmch, n = 6 and 5 for nCTG and AAVs, respectively; for Agrp, n = 6 and 7 for nCTG and AAVs, respectively. No significant differences were observed (p = 0.09, 0.79, and 0.23 for Calcr, Pmch, and Agrp, respectively, by a two-tailed Mann–Whitney U-test). Error bars, SEM.
Fig 5.
Comparison of input to PVH AVP neurons by injecting RV at P7, P30 and 8-week-old females.
(A, B) Schematic of the virus injection. Female mice were used in this experiment. Rabies virus was injected into the unilateral PVH at postnatal day (P) 7, 30 or 8 weeks of age. Data for 8-week-old mice correspond to the “nCTG” data in Fig 3. (C) Representative coronal sections containing starter cells, defined as the overlap of mCherry (magenta) and GFP (green), in the PVH. Blue, DAPI. Scale bar, 30 μm. (D) The number of starter cells in the PVH did not differ significantly between the three groups (one-way ANOVA). n = 5, 5 and 6 mice for P7, P30 and 8 weeks groups, respectively. (E) Representative coronal sections of presynaptic cells revealed by rabies-GFP. Scale bars, 50 μm. (F) Normalized inputs from the nuclei to the PVH AVP neurons. In the brain regions that we examined, no significant differences were found between ages (p > 0.48, two-way ANOVA). Error bars, SEM.