CNS expression of murine fragile X protein (FMRP) as a function of CGG-repeat size

Abstract

Large expansions of a CGG-repeat element (>200 repeats; full mutation) in the fragile X mental retardation 1 (FMR1) gene cause fragile X syndrome (FXS), the leading single-gene form of intellectual disability and of autism spectrum disorder. Smaller expansions (55-200 CGG repeats; premutation) result in the neurodegenerative disorder, fragile X-associated tremor/ataxia syndrome (FXTAS). Whereas FXS is caused by gene silencing and insufficient FMR1 protein (FMRP), FXTAS is thought to be caused by 'toxicity' of expanded-CGG-repeat mRNA. However, as FMRP expression levels decrease with increasing CGG-repeat length, lowered protein may contribute to premutation-associated clinical involvement. To address this issue, we measured brain Fmr1 mRNA and FMRP levels as a function of CGG-repeat length in a congenic (CGG-repeat knock-in) mouse model using 57 wild-type and 97 expanded-CGG-repeat mice carrying up to ~250 CGG repeats. While Fmr1 message levels increased with repeat length, FMRP levels trended downward over the same range, subject to significant inter-subject variation. Human comparisons of protein levels in the frontal cortex of 7 normal and 17 FXTAS individuals revealed that the mild FMRP decrease in mice mirrored the more limited data for FMRP expression in the human samples. In addition, FMRP expression levels varied in a subset of mice across the cerebellum, frontal cortex, and hippocampus, as well as at different ages. These results provide a foundation for understanding both the CGG-repeat-dependence of FMRP expression and for interpreting clinical phenotypes in premutation carriers in terms of the balance between elevated mRNA and lowered FMRP expression levels.

Figure 1.

Scatter plots of mouse CGG repeat versus: (A) Fmr1 mRNA level (normalized to the endogenous control Gusb), (B) FMRP protein level (normalized to Gapdh) and (C) the ratio of FMRP to Fmr1 mRNA, an indication of Fmr1 translation efficiency. Each graph includes solid black regression lines and gray 95% confidence interval lines. Open circles indicate animals that were bred as F2 crosses of animals from the initial breading. One outlier is shaded gray. FMR1 mRNA measurements were performed in triplicate, and FMRP measurements were in duplicate. R-squared values are as follows: Fmr1 = 0.80, FMRP = 0.52 and Fmr1/FMRP = 0.80.

Figure 2.

Box-plot comparison of human and mouse brain FMRP levels. FMRP levels in the frontal cortex of normal (n = 7) and premutation (n = 17) human individuals and wt (n = 57) and expanded-CGG (n = 48) mice. Solid bisecting lines are medians, dotted lines are means and individual points represent outliers. Error bars are standard errors of the mean (SEM). Significant differences were seen between the following groups: P < 0.001 for mouse wt versus mouse expanded; mouse wt versus human expanded; human normal versus human expanded; and human normal versus mouse expanded, using one-way ANOVA. The P-value between mouse expanded and human expanded is 0.26.

Figure 3.

Dependence of FMRP expression on mouse age. FMRP levels in (A) frontal cortex, (B) hippocampus and (C) cerebellum of wt (open circles) and expanded-repeat (solid circles) mice at five different ages. Each point represents the average of three to seven mice, with the exception of P35 expanded-CGG hippocampus, which included two mice. Error bars correspond to SEMs for each age group.

Figure 4.

FMRP immunohistochemistry in mouse brain. FMRP staining of brain regions of mice with wt (∼10), low (82, 89 and 96) and high (175, 181 and 222) CGG-repeat lengths.

Figure 5.

Quantification of FMRP levels in the mouse tissues shown in Figure 4. Error bars are SEM of four or five separate experiments.