Figure 1. Expression of phosphate-activated glutaminase (PAG) in mouse ventral midbrain DA neurons.

(A) Confocal mosaic z-projected image of the ventral midbrain showing TH (green, left) and PAG (magenta, right) immunoreactivity. Merged image (center) shows that some TH⁺ DA neurons co-express PAG (white). The specificity of the PAG antibody was verified in GLS1 KO mice; see Figure 1—figure supplement 1A. (B) Magnified confocal images in the VTA (left) and SNc (right) showing TH⁺ only (thin blue arrow), PAG⁺ only (blue arrow head) and TH⁺/PAG⁺ cells (thick blue arrow). (C) Stereological counts of TH⁺ only (green), PAG⁺ only (magenta) and TH⁺ / PAG⁺ (white) cells in the VTA and SNc of juvenile (P25) wild type mice (n = 4). Cell numbers in the VTA (TH⁺only = 4681, PAG⁺only = 3411, TH⁺ / PAG⁺=3673) were greater than in the SNc (TH⁺only = 2564, PAG⁺only = 2909, TH⁺ / PAG⁺=2595) (two-way ANOVA: main effect of brain region, F_(1,18)= 18.36; p<0.001; effect size (ES) partial η² = 0.51), but the relative proportions of cell types did not differ between regions(main effect of cell type, F_(2,18)= 1.22; p=0.318; cell type X brain region interaction, F_(2,18)= 2.70; p=0.094). (D) Single-cell RT-PCR analysis of cells expressing TH mRNA, in the VTA and SNc of juvenile mice (P25-37), showing the percentage of cells that co-expressed PAG and VGLUT2 mRNA. In the VTA, most cells were either TH⁺ only (7/22) or TH⁺/PAG ⁺/VGLUT2⁺(8/22); there were also TH⁺/PAG⁺ cells (5/22) and rarely TH⁺/VGLUT2⁺ (2/22). In the SNc, most cells were either TH⁺ only (5/12) or TH⁺/PAG⁺ cells (6/12); and rarely TH⁺/PAG⁺/VGLUT2⁺ (1/12). No TH⁺ cells expressed GAD mRNA. For the full coexpression analysis, including GAD mRNA, see Figure 1—figure supplement 1B and C. (E) Comparison of the relative number of TH⁺ / PAG⁺ cells in juvenile (P25) and adult (P60) mice. In both the VTA and SNc, there was a significant increase in the number of TH⁺ / PAG⁺ cells. # indicates a significant main effect of age (two-way ANOVA, F_(1,10)= 8.26; p=0.017, ES partial η² = 0.45); there was no significant region effect (F_(1,10)= 2.154; p=0.173), nor interaction, (F_(1,10)= 0.846; p=0.379). See Figure 1—source data 1.xlsx for source data and all statistical analysis.

DOI: http://dx.doi.org/10.7554/eLife.27566.003

Figure 1—figure supplement 1. Expression of PAG in dopamine neurons.

(A) Validation of the phosphate-activated glutaminase (PAG) antibody in GLS1 KO mice (neonates were used since KOs survive only for a few hours). Immunoreactivity was absent in GLS1 KO brain. Sagittal sections are shown. Abbreviations: Ctx, cortex; Hipp, hippocampus. (B) Sample gel images of single-cell reverse transcription (RT) PCR from the VTA (top) and the SN (bottom). For each region, the upper gel shows the multiplex result for glutamate decarboxylase (GAD67, 702 bp), tyrosine hydroxylase (TH, 377 bp), vesicular glutamate transporter (VGLUT2, 250 bp); the lower gel shows PAG (512 bp). Numbers on the top of each image are cell numbers; each lane in the multiplex gel (top) and PAG gel (bottom) was from the same cell. (C) Euler diagrams showing RT-PCR results in the VTA (top) and the SN (bottom). Numbers inside each square indicate the number of cells expressing the gene or combination of genes. TH⁺ cells are grouped (green square) in the diagram on the left; TH^— cells are divided into those expressing GAD67 (gray squares) and VGLUT2 (blue squares), on the right. PAG expressing cells are indicated by magenta-filled magenta squares. Cells expressing TH, VGLUT2 and PAG are indicated by yellow filled squares. There was no overlap of TH and GAD.

DOI: http://dx.doi.org/10.7554/eLife.27566.005