Genetic background affects human glial fibrillary acidic protein promoter activity

Abstract

The human glial fibrillary acidic protein (hGFAP) promoter has been used to generate numerous transgenic mouse lines, which has facilitated the analysis of astrocyte function in health and disease. Here, we evaluated the expression levels of various hGFAP transgenes at different ages in the two most commonly used inbred mouse strains, FVB/N (FVB) and C57BL/6N (B6N). In general, transgenic mice maintained on the B6N background displayed weaker transgene expression compared with transgenic FVB mice. Higher level of transgene expression in B6N mice could be regained by crossbreeding to FVB wild type mice. However, the endogenous murine GFAP expression was equivalent in both strains. In addition, we found that endogenous GFAP expression was increased in transgenic mice in comparison to wild type mice. The activities of the hGFAP transgenes were not age-dependently regulated. Our data highlight the importance of proper expression analysis when non-homologous recombination transgenesis is used.

Figure 1. hGFAP promoter controlled transgene expression in five different mouse lines.

(A) Transgenic constructs used for oocyte injection. (B) Widespread expression of ECFP in FVB(hGFAP-ECFP)_GCFD mice with high levels in the cerebellum. Scale bar indicates 1 mm. (C) Abundant fluorescent signals from Bergmann glia of FVB(hGFAP-ECFP)_GCFD, FVB(hGFAP-EGFP)_GFEA/C, B6N(hGFAP-AmCyan)_GCYM and FVB(hGFAP-CT2_GCFT × R26tdTom). Transgene copy numbers are indicated below the respective mouse lines. Scale bars indicate 100 µm.

Figure 2. Immunohistochemical analysis of reporter protein expression in different transgenic mouse lines showed lower expression in the B6N background when compared to FVB.

Cerebellar vibratome slices (cb) of 8-week-old mice were immunolabeled with anti-GFP (A and C) and anti-S100β antibodies (A, C and E), endogenous fluorescence of tdTomato in E. Upper panels depict transgene expression in B6N, lower panels in FVB. The S100β staining indicates all Bergmann glia. Results of comparative analysis in B6N and FVB mice are presented as percentage of transgene expressing Bergmann glia (S100β positive cells) (B, D and F). ***: p<0.001, **: p<0.01. Scale bars indicate 50 µm.

Figure 3. Comparative Western blot analysis of endogenous GFAP and transgenic proteins in B6N and FVB mice.

Cerebellar homogenates of transgenic and wild type mice (1 w and 8 w) were probed with anti-GFP (to detect ECFP or EGFP), anti-human estrogen receptor α (ER α, recognizing CT2), and anti-GFAP and anti-α-tubulin antibodies. (A) Western blot analysis of transgene expression. (B) Western blot analysis of endogenous GFAP expression in WT and transgenic mice (hGFAP-ECFP)_GCFD. (C) Western blot analysis of endogenous GFAP expression in WT and five transgenic mouse lines (hGFAP-ECFP_GCFD; hGFAP-EGFP_GFEA; hGFAP-EGFP_GFEC; hGFAP-CT2_GCTF; hGFAP-AmCyan_GCYM) in both FVB and B6N background.

Figure 4. Quantitative RT-PCR analysis of transgene and endogenous GFAP mRNA levels in FVB and B6N mice.

(A) Cerebellar GFAP mRNA levels in wild type B6N and FVB mice (1 w and 8 w). (B-D) Transgenic mRNA levels compared to endogenous GFAP mRNA levels in the cerebellum of B6N and FVB mice (1 w and 8 w). (B) hGFAP-ECFP_GCFD. (C) hGFAP-EGFP_GFEC. (D) hGFAP-CT2_GCTF. Relative expression is normalized to GFAP mRNA level in 1 w B6N mice. *: p<0.05, **: p<0.01, ***: p<0.001. Data are obtained from three independent experiments with samples from three mice (n = 3) in every experiment.

Figure 5. Backcrossing of B6N (hGFAP-ECFP)GCFD mice to FVB for a single generation re-activated transgenic ECFP expression.

(A and B) Cerebellar slices of 8-week-old mice were immunostained with anti-GFP and anti-S100β antibodies and analyzed. Only single ECFP expressing Bergmann glia (S100β positive cells) were detected in B6N(hGFAP-ECFP)_GCFD mice (A, upper panel), while ∼91.5% of Bergmann glia were ECFP positive in FVB(hGFAP-ECFP)_GCFD mice (A, lower panel). Backcrossing of B6N(hGFAP-ECFP)_GCFD for one generation with FVB WT mouse led to increased ECFP expression in B6NxFVB1 littermates (A, middle panel). (C) GFAP and ECFP mRNA levels in B6N, FVB and B6NxFVB1 mice (8 w). Relative expression is normalized to GFAP mRNA level in B6N mice. *: p<0.05 and ***: p<0.001. Data are obtained from three independent experiments with samples from three mice (n = 3) in every experiment. Scale bars indicate 100 µm.

Figure 6. Backcrossing of B6N(hGFAP-AmCyan)GCYM mice to FVB increased transgenic AmCyan expression.

Cerebellar (A) and cortical (B) brain slices of 8-week-old mice were immunostained with anti-S100β antibodies. Backcrossing of B6N(hGFAP-AmCyan)_GCYM (upper panels in A and B) for one generation to FVB (lower panels in A and B) did not significantly enhance AmCyan expression in cerebellum, but caused higher levels in the cortex of B6NxFVB1 littermates when compared to B6N. Results of comparative analysis (right panels in A and B) in B6N and B6NxFVB1 mice are provided as percentage of transgene expressing Bergmann glia (A) and cortical astrocytes (B) (S100β positive cells). (C) GFAP and AmCyan mRNA levels in 1-week-old (B6N, B6NxFVB1 and B6NxFVB2) and 8-week-old (B6N and B6NxFVB1) mouse cerebellum. A second backcrossing resulted in enhanced transgene levels. Relative expression is normalized to GFAP mRNA level in 1-week-old B6N mice. *: p<0.05. Data are obtained from three independent experiments with samples from three mice (n = 3) in every experiment. Scale bars indicate 50 µm.