Fig 2: DAGLB-mediated 2-AG biosynthesis in the nigral DANs is involved in locomotor skill learning.a Schematic diagram of eCB2.0 fiber photometry recoding while mice undergo rotarod locomotor skill training. Inset shows the average eCB2.0 (green), tdT (red), and normalized eCB2.0 (blue) signals during each trial from one representative mouse on day 2 of the 6-day training paradigm (n = 1). Data were presented as mean ± SD. b Normalized eCB2.0 signal intensity during the locomotor learning. n = 8 (4M/4F). Data were presented as mean ± SD. c Box and whiskers plot (the minimum, the 25th percentile, the median, the 75th percentile, and the maximum) of maximal daily increase of eCB2.0 signal intensity. n = 8 (4M/4F). Tukey’s multiple comparisons adjusted, *p = 0.02, **p = 0.001, ***p = 0.0003, ****p < 0.0001. d Performance of rotarod motor skill training. n = 8 (4M/4F). Data were presented as mean ± SEM. e Box and whiskers plot (the minimum, the 25th percentile, the median, the 75th percentile, and the maximum) of correlation coefficient of eCB2.0 signal intensity and rotarod performance. n = 8 (4M/4F). Tukey’s multiple comparisons adjusted, ***p = 0.0002 to 0.0006, ****p < 0.0001.

Fig 3: Identification of homozygous DAGLB mutations in affected families.a Pedigrees are shown for the four affected families. A double bar represents parental consanguinity. Slash indicates deceased individuals. Males are represented by squares, females by circles, and affected individuals by shading. b Schematic view of DAGLB gene structure and encoded protein domains. The four transmembrane segments are shown in blue, and the catalytic domain in maroon. Within the catalytic domain, a regulatory loop is colored in green. Variant sites are indicated by arrows. c–e Representative western blot (c) and quantification of DAGLB protein (d) and mRNA (e) levels in fibroblasts derived from Family 1 II-3, Family 2 II-4 and three age-matched HCs. N = 3-4 independent fibroblast cultures. Data were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) protein or mRNA levels as appropriate and represent mean ± SEM. One-way ANOVA, F = 33.09, ****p < 0.0001.

Fig 4: Daglb-deficiency in nigral DANs compromises the early dynamic release of 2-AG and impairs rotarod locomotor skill learning.a Schematic diagram of eCB2.0 fiber photometry recoding while DAN-Ctrl or DAN-Daglb KD mice undergo rotarod locomotor skill training. b Normalized eCB2.0 signal intensity over the course of locomotor learning of DAN-control and DAN-Daglb KD (n = 5 M per genotype) mice. Data were presented as mean ± SEM. Two-way ANOVA, genotype: F(1, 8) = 13. 68, p = 0.006. c Rotarod locomotor skill training of a different cohort of DAN-control (n = 16, 8M/8F) and DAN-Daglb KD (n = 16, 8M/8F) mice. Data were presented as mean ± SEM. Two-way ANOVA, genotype: F(1, 30) = 9.3, p = 0.0047. d Box and whiskers plot (the minimum, the 25th percentile, the median, the 75th percentile, and the maximum) of rotarod performance in the first and 10th trial on each day with the same cohorts of DAN-control (n = 16, 8 M/8F) and DAN-Daglb KD (n = 16, 8M/8F) mice. Multiple unpaired t test, two-tailed, *p < 0.05, **p < 0.005, ***p < 0.001.

Fig 5: DAGLB mediates the main 2-AG synthesis in the nigral DANs.a Diagram of AAV-mediated CRISPR/saCas9 gene targeting vector and the sequence of Dagla and Daglb sgRNAs. b Western blot of DAGLB in cultured cortical and hippocampal neurons transfected with control, Dagla, or Daglb KD AAV vectors in combination with Cre-expressing AAV vectors. Actin was used as a loading control. Four independent experiments were performed. c, d Bar graph quantifies DAGLB (c) and DAGLA (d) protein levels in four independent experiments. Data were presented as mean ± SEM. One-way ANOVA, multiple comparisons, adjusted ****p < 0.0001. e Schematic illustrates fiber photometry recording of 2-AG signals in the SN of DAN-control and DAN-Daglb KD mice. DLS: dorsolateral striatum. f Cartoon of eCB sensor eCB2.0. g Co-staining of eCB2.0, tdT and TH in the midbrain sections of DAN-Daglb KD mice. Scale bar: 100 μm. More than five independent experiments were performed. h Sample photon counts of eCB2.0 (wavelength: 500–540 nm) and tdT (wavelength: 575–650 nm) emission immediately before and 120 min after JZL184 (16 mg/kg) administration. i Time course of eCB2.0 signals in the SN of DAN-control [n = 4, 2 Male(M)/2 Female(F)] and DAN-Daglb KD (n = 4, 2M/2F) mice before and after JZL184 (16 mg/kg) treatment. Data were presented as mean ± SEM. Sidak’s multiple comparisons, adjusted ***p = 0.0001. ****p < 0.0001. j Dose response of eCB2.0 signals 120 min after JZL administration at 0 (vehicle only), 4, 16, and 40 mg/kg. n = 4 mice per group. Data were presented as mean ± SEM. Sidak’s multiple comparisons, adjusted ***p = 0.0003. ****p < 0.0001. k LC-MS/MS quantification of 2-AG in the SNc of DAN-control (n = 9) and DAN-Daglb KD (n = 12) mice. Data were presented as mean ± SEM. Unpaired t test, two-tailed, *p = 0.02.