Loss of FBXO7 (PARK15) results in reduced proteasome activity and models a parkinsonism-like phenotype in mice

Abstract

Mutations in the FBXO7 (PARK15) gene have been implicated in a juvenile form of parkinsonism termed parkinsonian pyramidal syndrome (PPS), characterized by Parkinsonian symptoms and pyramidal tract signs. FBXO7 (F-box protein only 7) is a subunit of the SCF (SKP1/cullin-1/F-box protein) E3 ubiquitin ligase complex, but its relevance and function in neurons remain to be elucidated. Here, we report that the E3 ligase FBXO7-SCF binds to and ubiquitinates the proteasomal subunit PSMA2. In addition, we show that FBXO7 is a proteasome-associated protein involved in proteasome assembly. In FBXO7 knockout mice, we find reduced proteasome activity and early-onset motor deficits together with premature death. In addition, we demonstrate that NEX (neuronal helix-loop-helix protein-1)-Cre-induced deletion of the FBXO7 gene in forebrain neurons or the loss of FBXO7 in tyrosine hydroxylase (TH)-positive neurons results in motor defects, reminiscent of the phenotype in PARK15 patients. Taken together, our study establishes a vital role for FBXO7 in neurons, which is required for proper motor control and accentuates the importance of FBXO7 in proteasome function.

Keywords: FBXO7; PARK15; PSMA2; parkinsonism; ubiquitination.

Figure 1. Characterization of FBXO7 ^−/− mice

1. A

   Lysates of the indicated tissues isolated from P30 rat were immunoblotted with FBXO7 and γ‐tubulin antibodies. The latter served as a loading control.

2. B

   Schematic of the genetically modified PARK15 gene locus located on chromosome 22.

3. C

   Brain lysates from P5 FBXO7 ^+/+, FBXO7 ^+/−, or FBXO7 ^−/− mice were immunoblotted with FBXO7, βgal or γ‐tubulin antibodies. The latter served as a loading control.

4. D

   LacZ staining of P18 FBXO7 ^−/− sagittal brain section. Ctx = cortex, Hpc = hippocampus, Cb = cerebellum, Olf = olfactory bulb, CC = corpus callosum, Str = striatum, Th = thalamus. Scale bar = 1 mm.

5. E, F

   Body weight of P5 or P18 FBXO7 ^+/+, FBXO7 ^+/−, or FBXO7 ^−/− mice. n = 17, 46, 16 (E) and n = 16, 25, 11 (F), respectively (ANOVA, ***P < 0.001, mean ± s.e.m.).

6. G

   Image of representative P18 FBXO7 ^+/+, FBXO7 ^+/− or FBXO7 ^−/− mice.

7. H

   Kaplan–Meier survival curve of male and female FBXO7 ^−/− mice. n = 14 and 16, respectively.

8. I, J

   P18 FBXO7 ^+/+, FBXO7 ^+/−, or FBXO7 ^−/− littermates were tested in the open field (I) and on the rotarod (J). n = 20, 23, 16, respectively (ANOVA, ***P < 0.001, mean ± s.e.m.).

9. K

   Average weight of 6‐month‐ and 12‐month‐old FBXO7 ^+/+ and FBXO7 ^+/− littermates. n = 15 and 18, respectively (ANOVA, mean ± s.e.m.).

10. L

    6‐month‐ and 12‐month‐old FBXO7 ^+/+ and FBXO7 ^+/− littermates were examined on the rotarod. n = 15 and 18, respectively (ANOVA, mean ± s.e.m.).

Source data are available online for this figure.

Figure EV1. Characterization of FBXO7 ^−/− mice

1. A–C

   Lysates of cortices (A), hippocampi (B), and cerebella (C) of the indicated age isolated from rat were immunoblotted with FBXO7, pan 14‐3‐3, or γ‐tubulin antibodies. The latter two served as loading controls.

2. D

   Genotyping of FBXO7 ^+/+, FBXO7 ^+/−, and FBXO7 ^−/− mice. Detection of wild‐type and mutant FBXO7 alleles.

3. E

   Quantitative PCR of FBXO7 ^+/+, FBXO7 ^+/− and FBXO7 ^−/− brain tissue. n = 4, 2, and 4, respectively (mean ± s.e.m.).

4. F

   LacZ staining of P18 FBXO7 ^+/+ and FBXO7 ^−/− sagittal brain sections. Ctx = cortex, Hpc = hippocampus, Cb = cerebellum, Olf = olfactory bulb, CC = corpus callosum, Str = striatum, Th = thalamus. Scale bar = 1 mm.

5. G–I

   Higher magnification of LacZ‐stained P18 FBXO7 ^−/− sagittal brain: cortex (G), cerebellum (H), and hippocampus (I). CC = corpus callosum, PCL = Purkinje cell layer, IGL = internal granule cell layer, WM = white matter, DG = dentate gyrus, CA1, 2, 3 = cornus ammonis 1, 2, 3. Scale bars = 200 μm.

6. J, K

   Cultured rat cerebellar granule neurons (J) or mouse cortical tissue (K) was subjected to subcellular fractionation analyses. Nuclear fraction and postnuclear supernatant (PNS) were immunoblotted with FBXO7, SP1, and pan 14‐3‐3 antibodies. The latter two served as quality control for the nuclear fraction and PNS, respectively.

7. L, M

   Brain weight of P5 and P18 FBXO7 ^+/+, FBXO7 ^+/−, or FBXO7 ^−/− mice. n = 8, 20, and 7 (L) and n = 6, 11, and 8 (M), respectively (ANOVA, **P < 0.01, mean ± s.e.m.).

8. N

   P18 littermates were tested on the hanging wire. n = 20, 23, and 16, respectively (ANOVA, ***P < 0.001, mean ± s.e.m.).

Source data are available online for this figure.

Figure 2. Histological analyses of the FBXO7 ^−/− mice

1. A–D

   Striatal tissue from P18 FBXO7 ^+/+ and FBXO7 ^−/− mice was subjected to HPLC analyses to determine dopamine (A), 3,4‐dihydroxyphenylacetic acid (DOPAC) (B), and homovanillic acid (HVA) (C) concentrations along with the metabolite ratio (D). n = 12 and 13, respectively (unpaired t‐test, *P < 0.05, mean ± s.e.m.).

2. E

   Representative images from FBXO7 ^+/+ and FBXO7 ^−/− midbrains. Insets display higher magnification of TH⁺ neurons of substantia nigra. Scale bar = 500 μm. Stereological analysis of the ratio of TH⁺Nissl⁺ cells/Nissl⁺ cells in the substantia nigra. Three independent litter pairs were analyzed (paired t‐test, mean ± s.e.m.)

3. F

   Sagittal paraffin sections of brains from P18 FBXO7^+/+ and FBXO7^−/− mice were subjected to immunohistochemistry using GFAP antibody. Five independent litter pairs were analyzed (paired t‐test, *P < 0.05, mean ± s.e.m.). Scale bar = 40 μm.

4. G

   Representative images of sagittal paraffin sections of cortices from P18 FBXO7 ^+/+ and FBXO7 ^−/− mice that were subjected to TUNEL staining. Three independent litter pairs were analyzed (paired t‐test, *P < 0.05, mean ± s.e.m.). Arrows indicate TUNEL⁺ cells. Scale bar = 40 μm.

5. H

   Analyses of TUNEL staining of cerebellum and hippocampus. Three independent litter pairs were examined (paired t‐test, mean ± s.e.m.).

6. I

   Cortical neurons were transfected at day in vitro (DIV) 3 with control vector, a functional FBXO7 shRNA plasmid, or a non‐functional FBXO7 shRNA plasmid together with a transfection marker. At DIV6, apoptotic neurons were counted. Four independent experiments were included in the analysis (ANOVA, **P < 0.01, mean ± s.e.m.).

Figure 3. FBXO7 interacts with the proteasomal subunit PSMA2 and binds to the proteasome

1. Lysates of HEK293T cells, transfected with the indicated plasmids, were subjected to immunoprecipitation (IP) with FLAG antibody (PSMA2), followed by immunoblotting (IB) with myc antibody (FBXO7).

2. Recombinant PMSA2 together with glutathione bead‐coupled GST‐FBXO7 or GST only was incubated in Co‐IP buffer. Precipitated proteins were subjected to immunoblotting with PSMA2 antibody. 2% input was used as a loading control for PSMA2. GST‐coupled proteins were visualized by Coomassie staining.

3. Schematic depicts full‐length FBXO7 and FBXO7 deletion mutants used in mapping analyses. UbRD = ubiquitin‐related domain, FP = FBXO7/PI31 domain, F‐box = F‐box domain, PRR = proline‐rich region.

4. The indicated FBXO7 expression plasmids were used to transfect HEK293T cells together with the FLAG‐PSMA2 plasmid. Lysates were then subjected to IP with myc antibody followed by IB with FLAG antibody.

5. HEK293T cells were transfected with full‐length myc‐FBXO7 or ΔUbRD and subjected to fractionation using a 10–40% linear glycerol gradient. Fractions were immunoblotted for myc (FBXO7), Rpt6 (26S/30S proteasome), and PSMB5 (20S proteasome).

6. Cortical lysates were subjected to fractionation using a 10–40% linear glycerol gradient. Fractions were immunoblotted for FBXO7, Rpt6 (26S/30S proteasome), and PSMA2 (20S proteasome).

7. GST or GST‐Rad23 bound to glutathione–Sepharose were used to precipitate proteasomes from brain lysates followed by IB with FBXO7, PI31, or PSMA2 antibody.

8. HEK293T control lysates and FLAG‐FBXO7 cells were subjected to immunoprecipitation, followed by digestion. The resulting peptide mixtures were analyzed by MS; ratio FBXO7 versus control; mean of three independent measurements based on the intensity of the relevant protein.

9. STRING analysis of the proteins listed in (H) (string‐db.org). PSMA2, PSMA6, PSMB1, PSMB3, PSMB4, PSMB6 are members of the gene ontology term (GO) “proteasome core complex” (GO: 0005839).

Source data are available online for this figure.

Figure EV2. FBXO7 interacts with the proteasomal subunit PSMA2 and binds to the proteasome

1. Lysates of HEK293T cells, transfected with the indicated plasmids, were subjected to IP with myc antibody for FBXO7, followed by IB with FLAG antibody for PSMA2.

2. Input from HEK293T cells, transfected with full‐length myc‐FBXO7, ΔUbRD, or ΔFP (panel C and Fig 3E), was immunoblotted with myc (FBXO7) or GAPDH antibody. The latter served as a loading control.

3. HEK293T cells were transfected with myc‐FBXO7 ΔFP and subjected to fractionation using a 10–40% linear glycerol gradient. Fractions were immunoblotted for myc (FBXO7), Rpt6 (26S/30S proteasome), and PSMB5 (20S proteasome).

4. HEK293T cells were subjected to fractionation using a 10–40% linear glycerol gradient. Fractions were immunoblotted for FBXO7, Rpt6 (26S/30S proteasome), and PSMB5 (20S proteasome).

Source data are available online for this figure.

Figure 4. FBXO7 ubiquitinates PSMA2

1. A

   Lysates of HEK293T cells, transfected with empty control vector, myc‐FBXO7, or myc‐FBXO7 ΔF‐box plasmids together with the FLAG‐PSMA2 expression plasmid, were immunoblotted with FLAG, myc, or pan 14‐3‐3 antibody.

2. B

   Lysates of HEK293T cells, transfected with the indicated plasmids, were subjected to boiling, followed by IP with GFP antibody (PSMA2), followed by IB with ubiquitin (DAKO) antibody (upper panel). Inputs were immunoblotted with GFP or myc antibody (lower panels). Three independent blots were quantified and band intensity normalized to control average (ANOVA, *P < 0.05, **P < 0.001, mean ± s.e.m.)

3. C, D

   Lysates of HEK293T cells, transfected as in (B), were first subjected to a denaturing protocol, then to IP with GFP antibody, followed by IB with K63‐specific ubiquitin antibody (C) or K48‐specific antibody (D). Inputs were immunoblotted with GFP or myc antibody (lower panel). Three independent blots were quantified and band intensity normalized to control average (ANOVA, *P < 0.05, mean ± s.e.m.).

4. E

   Lysates of HEK293T cells, transfected with the indicated HA–ubiquitin, GFP‐PSMA2, or myc‐FBXO7 plasmids, were subjected to boiling, followed by IP with GFP antibody (PSMA2), followed by IB with HA antibody (upper panel). Inputs were immunoblotted with GFP, myc or HA antibody (lower panels). Three independent blots were quantified and band intensity normalized to control (lane 2) average (ANOVA, Dunnett's multiple comparison ***P < 0.001, mean ± s.e.m.).

5. F

   Brain lysates from FBXO7 ^+/+ and FBXO7 ^−/− mice immunoblotted with PSMA2 or γ‐tubulin antibody. The latter served as a loading control. Ten brains from each genotype were included in the quantitative analyses (paired t‐test, mean ± s.e.m.).

6. G

   cDNA generated from cortex isolated from FBXO7 ^+/+, FBXO7 ^+/−, and FBXO7 ^−/− mice, was analyzed with quantitative PCR using FBXO7‐ and PSMA2‐specific primers along with the housekeeping gene β‐actin. n = 4, 2, and 4, respectively (unpaired t‐test, mean ± s.e.m.)

7. H

   For the in vitro ubiquitination assay, purified PSMA2, tagged and immunopurified FBXO7, and SCF core components from HEK293T cells, E1 enzyme, ubiquitin, and ATP together with the indicated E2 enzymes were incubated for 1 h at 37°C. The reactions were immunoblotted with PSMA2, ubiquitin, or FBXO7 antibodies.

Source data are available online for this figure.

Figure EV3. FBXO7 ubiquitinates PSMA2 and regulates proteasome assembly

1. Control of knockdown for Fig 5A. Lysates of HEK293T cells transfected with empty control vector or PSMA2 shRNA plasmids were immunoblotted with PSMA2 or γ‐tubulin antibody. The latter served as a loading control.

2. Control of knockdown for Fig 5B. Lysates of HEK293T cells transfected with empty control vector, FBXO7 shRNA, or non‐functional FBXO7 shRNA plasmids, were immunoblotted with FBXO7 or γ‐tubulin antibody. The latter served as a loading control.

3. Control of knockout for Fig 5C. Brain lysates from FBXO7 ^+/+ or FBXO7 ^−/− mice were immunoblotted with βgal or pan 14‐3‐3 antibody. The latter served as a loading control.

4. Whole brain lysates from FBXO7 ^+/+ , FBXO7 ^+/−, and FBXO7 ^−/− mice at age P5 and P16 were subjected to immunoblotting with βgal and PI31 antibodies. Pan 14‐3‐3 was used as a loading control.

5. Lysates from HEK293T cells, transfected with empty control vector, functional PI31 shRNA, or non‐functional PI31 shRNA, were analyzed for chymotrypsin‐like proteasome activity assay (LLVY‐AMC). Three independent experiments were carried out (ANOVA, mean ± s.e.m.).

6. Input from (E) was immunoblotted with PI31 or γ‐tubulin antibody. The latter served as a loading control.

7. Control of knockdown and loading for Fig 5E. Lysates of HEK293T cells transfected with empty control vector or FBXO7 shRNA plasmids were immunoblotted with FBXO7, PSMA2, or γ‐tubulin antibody. The latter served as a loading control.

8. Control of knockdown and loading for Fig 5F. Lysates of HEK293T cells transfected with empty control vector or FBXO7 shRNA plasmids were immunoblotted with FBXO7, Rpt6, or pan14‐3‐3 antibody. The latter served as a loading control.

9. HEK293T cells transfected with empty control vector or the FBXO7 RNAi plasmid were subjected to fractionation using a 10–40% linear glycerol gradient. Fractions were subjected to immunoblotting with FBXO7, Rpt6 (26S/30S proteasome), or PSMB5 (20S proteasome) antibody.

10. Control of knockout and loading for Fig 5H. Brain lysates from FBXO7 ^+/+ or FBXO7 ^−/− mice were immunoblotted with βgal, FBXO7, or γ‐tubulin antibody. The latter served as loading control.

Source data are available online for this figure.

Figure 5. FBXO7 ubiquitinates PSMA2 and regulates proteasome assembly

1. A

   Lysates from HEK293T cells, transfected with empty control vector, PSMA2 shRNA A, PSMA2 shRNA B, or non‐functional PSMA2 shRNA, were subjected to chymotrypsin‐like proteasome activity assay. Three independent experiments were included in the analysis (ANOVA, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

2. B

   Lysates from HEK293T cells, transfected with control vector, FBXO7 shRNA, or non‐functional FBXO7 shRNA, were subjected to a chymotrypsin‐like proteasome activity assay. Three independent experiments were included in the analysis (ANOVA, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

3. C

   Brain lysates from wild‐type or homozygous FBXO7 mice were analyzed for chymotrypsin‐like proteasome activity. Three independent experiments were included in the analysis (paired t‐test, *P < 0.05, **P < 0.01, mean ± s.e.m.).

4. D

   Proteasomes purified from FBXO7 ^+/+ or FBXO7 ^−/− brains were analyzed by native PAGE, followed by in‐gel chymotrypsin‐like activity assay and immunoblotting with PSMA2 antibody.

5. E, F

   Lysates from HEK293T cells, transfected with empty control vector, FBXO7 shRNA, or non‐functional FBXO7 shRNA, were analyzed by native PAGE followed by in‐gel chymotrypsin‐like proteasome activity assay (LLVY‐AMC), in the presence of SDS, and followed by immunoblotting with PSMA2 (E) or Rpt6 (F) antibody. In (E), ratio of free CPs to RP₁ + RP₂ − CP was calculated for each lane and normalized to control average. Quantification includes four (activity) and three (IB) independent experiments (ANOVA, *P < 0.05, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

6. G

   HEK293T cells were transfected with empty control vector or the FBXO7 shRNA plasmid and subjected to fractionation using a 10–40% linear glycerol gradient and fractions were analyzed for chymotrypsin‐like proteasome activity.

7. H

   Brain lysates from FBXO7 ^+/+ and FBXO7 ^−/− mice were analyzed on native PAGE followed by in‐gel chymotrypsin‐like proteasome activity assay and by IB with PSMA2 antibody. Quantification includes six wild‐type and seven FBXO7 ^−/− brains (activity) and five wild‐type and five FBXO7 ^−/− brains (IB) (unpaired t‐test, *P < 0.05, mean ± s.e.m.).

Source data are available online for this figure.

Figure EV4. Deletion of FBXO7 in the forebrain triggers early‐onset motor impairment

1. A

   Average body weight of 12‐month‐old FBXO7 ^+/+ or FBXO7 ^fl/fl mice. n = 13 and 10, respectively (t‐test, mean ± s.e.m.).

2. B

   Hind limb clasping of 12‐month‐old FBXO7 ^+/+ or FBXO7 ^fl/fl mice. n = 13 and 10, respectively (Mann–Whitney U‐test, mean ± s.e.m.).

3. C–F

   Open field test (C), balance beam test (time to cross and coordination score were measured, D), rotarod (E), and pole test (F) of 12‐month‐old FBXO7 ^+/+ or FBXO7 ^fl/fl mice. n = 13 and 10, respectively (unpaired t‐test or Mann–Whitney U‐test (coordination score), mean ± s.e.m.).

4. G

   Sagittal paraffin sections from NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice cortices were subjected to TUNEL staining. Three mice per genotype were included in the analysis (ANOVA, mean ± s.e.m.). Scale bar = 40 μm.

Figure 6. Deletion of FBXO7 in the forebrain triggers early‐onset motor impairment

1. A

   Genotyping PCR of NEX‐Cre, NEX‐Cre;fl/+, or NEX‐Cre;fl/fl mice using primers for floxed allele, wild‐type FBXO7 and Cre.

2. B

   Cortical and cerebellar lysates from NEX‐Cre, NEX‐Cre;fl/+, or NEX‐Cre;fl/fl mice were immunoblotted with FBXO7 or γ‐tubulin antibody. The latter served as a loading control.

3. C

   Body weight of 2‐ and 4‐month‐old NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice. n = 11, 14, and 15, respectively (ANOVA, **P < 0.01, mean ± s.e.m.).

4. D

   Hind limb clasping of NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice. 0 = normal, 3 = worst manifestation of clasping. n = 11, 14, and 15, respectively (Kruskal–Wallis test, Dunn's multiple comparison, ***P < 0.001, mean ± s.e.m.). Representative images of Nex‐Cre and Nex‐Cre;fl/fl mice.

5. E

   Open field test of NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice. n = 11, 14, and 15, respectively (ANOVA, **P < 0.01, mean ± s.e.m.).

6. F

   Rotarod test of NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice. n = 11, 14, and 15, respectively (ANOVA, *P < 0.05, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

7. G

   Balance beam test of NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice. Coordination score and time to cross were measured. n = 11, 14, and 15, respectively (ANOVA (time), Kruskal–Wallis test, Dunn's multiple comparison (coordination score), *P < 0.05, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

8. H, I

   Sagittal paraffin sections from NEX‐Cre, FBXO7 ^fl/fl, or NEX‐Cre;fl/fl mice cortices were subjected to immunohistochemistry with GFAP (H) or Iba1 (I) antibody. Three mice per genotype were included in the analysis (ANOVA, *P < 0.05, **P < 0.01, mean ± s.e.m.). Scale bar = 40 μm.

Source data are available online for this figure.

Figure 7. Loss of FBXO7 in TH ⁺ cells results in late‐onset motor deficits

1. A

   Genotyping PCR of TH‐Cre, TH‐Cre;fl/+, or TH‐Cre;fl/fl mice using primers for floxed allele, wild‐type FBXO7 and Cre.

2. B

   Average body weight of 6‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 16 and 15, respectively (unpaired t‐test, **P < 0.01, mean ± s.e.m.).

3. C

   Average body weight of the weight‐corrected cohort of 6‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 11 and 10, respectively (unpaired t‐test, mean ± s.e.m.).

4. D–F

   Open field (D), balance beam (E) or pole test (F) of 6‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 11 and 10, respectively (unpaired t‐test or Mann–Whitney U‐test (coordination score), mean ± s.e.m.).

5. G

   Rotarod test of 6‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 11 and 10, respectively (unpaired t‐test, *P < 0.05, **P < 0.01, mean ± s.e.m.).

6. H

   DigiGait analysis of 8‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 7 for both groups (unpaired t‐test, *P < 0.05, **P < 0.01, mean ± s.e.m.). Data were visualized using a partial least squares (PLS) regression with orthogonal signal correction model.

7. I

   Average body weight of 12‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 16 and 9, respectively (unpaired t‐test, mean ± s.e.m.).

8. J–M

   Open field (J), balance beam (K), pole test (L), rotarod (M) of 12‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 16 and 9, respectively (unpaired t‐test or Mann–Whitney U‐test (coordination score), *P < 0.05, **P < 0.01, ***P < 0.001, mean ± s.e.m.).

Figure EV5. Loss of FBXO7 in TH ⁺ cells results in late‐onset motor deficits

1. A

   Average body weight of 2‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 15 and 9, respectively (unpaired t‐test, mean ± s.e.m.).

2. B–D

   Open field test (B), balance beam test (time to cross and coordination score were measured, C), and rotarod (D) of 2‐month‐old TH‐Cre and TH‐Cre;fl/fl mice. n = 15 and 9, respectively (unpaired t‐test or Mann–Whitney U‐test (coordination score), mean ± s.e.m.).

3. E

   Parameters tested in DigiGait analysis (unpaired t‐test, *P < 0.05, **P < 0.01, mean ± s.e.m.).

Figure 8. Loss of FBXO7 in TH ⁺ cells leads to reduced striatal dopamine

1. A

   Images of representative 12‐month‐old TH‐Cre and TH‐Cre;fl/fl midbrains immunostained with TH antibody (black) and Nissl stain (purple). Scale bar = 500 μm. Scale bar of inset = 10 μm.

2. B, C

   Stereological analysis of ratio of TH⁺Nissl⁺ cells/Nissl⁺ cells in the substantia nigra (A) from 2‐month‐old female (B) and 12‐month‐old male (C) TH‐Cre and TH‐Cre;fl/fl mice. Three animals from each genotype were analyzed (unpaired t‐test, mean ± s.e.m.).

3. D–G

   Striatal tissue from P18 TH‐Cre and TH‐Cre;fl/fl mice was subjected to HPLC analyses to determine dopamine (D), 3,4‐dihydroxyphenylacetic acid (DOPAC) (E), and homovanillic acid (HVA) (F) concentrations along with the metabolite ratio (G). N = 7 and 9 (2 months) and n = 15 and 12 (12 months) (unpaired t‐test, *P < 0.05, ***P < 0.001, mean ± s.e.m.)

4. H

   Images of representative TH‐Cre and TH‐Cre;fl/fl striatal sections immunostained with TH antibody. Scale bar = 1 mm.

5. I

   Immunohistochemistry of midbrain section TH‐Cre and TH‐Cre;fl/fl mice using TH and GFAP antibody. Quantitative measurements of GFAP stainings from four animals per genotype (unpaired t‐test, ***P < 0.001, mean ± s.e.m.).