Collagen VI deficiency causes behavioral abnormalities and cortical dopaminergic dysfunction

Abstract

Mutations of genes coding for collagen VI (COL6) cause muscle diseases, including Ullrich congenital muscular dystrophy and Bethlem myopathy. Although COL6 genetic variants were recently linked to brain pathologies, the impact of COL6 deficiency in brain function is still largely unknown. Here, a thorough behavioral characterization of COL6-null (Col6a1-/-) mice unexpectedly revealed that COL6 deficiency leads to a significant impairment in sensorimotor gating and memory/attention functions. In keeping with these behavioral abnormalities, Col6a1-/- mice displayed alterations in dopaminergic signaling, primarily in the prefrontal cortex. In vitro co-culture of SH-SY5Y neural cells with primary meningeal fibroblasts from wild-type and Col6a1-/- mice confirmed a direct link between COL6 ablation and defective dopaminergic activity, through a mechanism involving the inability of meningeal cells to sustain dopaminergic differentiation. Finally, patients affected by COL6-related myopathies were evaluated with an ad hoc neuropsychological protocol, revealing distinctive defects in attentional control abilities. Altogether, these findings point towards a previously undescribed role for COL6 in the proper maintenance of dopamine circuitry function and its related neurobehavioral features in both mice and humans. This article has an associated First Person interview with the first author of the paper.

Keywords: Central nervous system; Cognitive function; Collagen VI; Dopamine; Mouse model; Prefrontal cortex.

Fig. 1.

Col6a1^–/– mice display alterations in sensorimotor gating, social behavior and working memory. Results of different behavioral analyses conducted in 3- to 6-month-old wild-type and Col6a1^–/– mice. (A) Latency to fall in each trial of the rotarod test during the habituation and the test phases, according to the protocol shown in the top panel (minute, m; trial, T). The average value among the three trials for each phase is reported in the bottom panel. (B) Ambulatory distance displayed during exposure to an empty open-field arena (inset shows the total distance traveled in 60 min). (C) Left: startle response amplitude (arbitrary units) displayed by animals following the presentation of no stimulus or a 120 dB acoustic stimulus (Startle). Right: percentage of inhibition of the acoustic startle response displayed by the same mice after the presentation of a 70, 75, 80, 85 and 90 dB prepulse stimulus. (D) Time spent in the investigation of the same unfamiliar mouse during each of four successive 1-min trials and towards the stimulus mouse (new unfamiliar mouse) presented in the fifth trial, according to the protocol shown in the top panel (inter-trial interval, ITI). (E) Total exploration time in each phase of the temporal order object recognition task. Lower panel: discrimination ratio displayed by mice during the test phase (sample 3). Although no significant differences were detected in general locomotor activity, as measured by rotarod (A), nor in an open-field arena (B), Col6a1^–/– mice showed reduced prepulse inhibition and startle amplitude reaction (C), loss of preference between familiar and non-familiar conspecific animals (D), and defective temporal order object recognition (E). Error bars indicate s.e.m. **P<0.01; ***P<0.001; two-way ANOVA with Newman–Keuls post hoc test; WT, n=11; Col6a1^–/–, n=9.

Fig. 2.

LC-MS/MS analysis of wild-type and Col6a1^–/– mouse brains reveals altered neurotransmitter levels. Levels of dopamine (A) and serotonin (5-HT) (B) in Col6a1^–/– brains are reduced compared with their levels in WT brains. The amount of the main excitatory (glutamate) (C) and inhibitory (GABA) (D) neurotransmitters does not change between WT and Col6a1^–/– brains. Values represent nanograms or micrograms of neurotransmitter per gram of brain tissue. The box plots show the 25-75th percentiles, solid black lines in the middle indicate the median, red dotted lines indicate mean values, and whiskers show the minimum and maximum values. *P<0.05; **P<0.01; unpaired two-tailed Student's t-test; n=6 mice for each genotype).

Fig. 3.

Drd2 expression is altered in Col6a1^–/– mouse brains. (A-C) qRT-PCR analysis of Htra1, Gabbr1, Grin2a, Grin2b and Drd2 mRNA levels in the medial PFC (mPFC) (A), striatum (B) and hippocampus (C) of WT and Col6a1^–/– mice. Data normalized to both Gapdh and Actb (β-actin) are shown. Error bars indicate s.e.m. *P<0.05; ***P<0.001; unpaired two-tailed Student's t-test; n=12 mice for each genotype.

Fig. 4.

Dopaminergic signaling is impaired in the medial PFC of Col6a1^–/– mice. (A) qRT-PCR analysis for Th, Slc6a3 and Nr4a2 mRNA levels in the medial PFC of wild-type and Col6a1^–/– mice. *P<0.05; unpaired two-tailed Student's t-test; WT, n=11; Col6a1^–/–, n=12. (B) Representative western blot (left) and densitometric analysis (right) of TH and DAT in the medial PFC of WT and Col6a1^–/– mice. *P<0.05; unpaired two-tailed Student's t-test; WT, n=9; Col6a1^–/– n=8. Error bars indicate s.e.m.

Fig. 5.

Meningeal cells from Col6a1^–/– mice are not able to induce dopaminergic differentiation of SH-SY5Y cells. (A) Immunohistochemistry for COL6 on WT and Col6a1^–/– brain tissues. COL6 labeling is abundant at the level of the meninges, leptomeninges and the adventitia of large blood vessels, as indicated by GFAP staining (left panels), which marks astrocytic endfeet at the pial surface, and Mmr2 staining (middle panels), which marks endothelial cells of leptomeningeal vessels. COL6 is not detectable in the brain parenchyma or in neurons, stained by the anti-NeuN antibody (right panels). Scale bars: 50 µm. (B) Representative fluorescence microscopy images of SH-SY5Y cells cultured alone (top panels) or co-cultured with WT (middle panels) or Col6a1^–/– (bottom panels) meningeal cells, stained for nuclei (Hoechst 33528, blue, left panels), TH (red, middle panels) and merged (right panels). The insets show higher magnifications of merged images. Increased TH signal was observed when SH-SY5Y cells were co-cultured with WT meningeal cells, but not with Col6a1^–/– meningeal cells. Three experimental replicates were performed. Scale bar: 100 µm. (C) qRT-PCR analysis for TH transcript levels in SHSY5Y cells in the absence of meningeal cells and after co-culture with WT or Col6a1^–/– meningeal cells. Error bars indicate s.e.m. *P<0.05; unpaired two-tailed Student's t-test; n=3 different experiments.

Fig. 6.

Neuropsychological assessment in patients affected by COL6-related disorders. Percentage of patients for each score class in each test according to the equivalent scores, reported in Table 1. Executive functions are the most impaired. In the Stroop Test (errors), 40% of patients had a pathological score, 10% a borderline score and 40% a score at the lower limits of normal. Modified Wisconsin Card Sorting Tests, MCST; Rey Auditory Verbal Learning Test, RAVLT.