Endothelium-specific amyloid precursor protein deficiency causes endothelial dysfunction in cerebral arteries

Abstract

The exact physiological function of amyloid-β precursor protein (APP) in endothelial cells is unknown. Endothelium-specific APP-deficient (eAPP^-/-) mice were created to gain new insights into the role of APP in the control of vascular endothelial function. Endothelium-dependent relaxations to acetylcholine were significantly impaired in basilar arteries of global APP knockout (APP^-/-) and eAPP^-/- mice ( P < 0.05). In contrast, endothelium-independent relaxations to nitric oxide (NO)-donor diethylamine-NONOate were unchanged. Western blot analysis revealed that protein expression of endothelial nitric oxide synthase (eNOS) was significantly downregulated in large cerebral arteries of APP^-/- mice and eAPP^-/- mice as compared to respective wild-type littermates ( P < 0.05). Furthermore, basal levels of cyclic guanosine monophosphate (cGMP) were also significantly reduced in large cerebral arteries of APP-deficient mice ( P < 0.05). In contrast, protein expression of prostacyclin synthase as well as levels of cyclic adenosine monophosphate (cAMP) was not affected by genetic inactivation of APP in endothelial cells. By using siRNA to knockdown APP in cultured human brain microvascular endothelial cells we also found a significant downregulation of eNOS mRNA and protein expressions in APP-deficient endothelium ( P < 0.05). These findings indicate that under physiological conditions, expression of APP in cerebral vascular endothelium plays an important protective function by maintaining constitutive expression of eNOS .

Keywords: Amyloid precursor protein; amyloid-β; cerebral arteries; endothelial nitric oxide synthase; endothelium.

Figure 1.

(a) Representative genotyping analysis of endothelium-specific amyloid precursor protein-deficient mice (eAPP^−/− mice). Mice carrying loxP sites of APP gene (floxed APP) were bred with mice carrying the Tie2-Cre transgene (Tie2-Cre Tg) to generate eAPP^−/− mice and wild-type littermates (control). The PCR products were 503 bp for floxed allele and 443 bp for wild-type allele (upper blot) as well as 100 bp for presence of Cre recombinase (Cre⁺, lower blot) Cre⁻ indicates absence of Cre recombinase. (b) PCR analysis of APP mRNA in the aorta of wild-type littermates (APP^flox/flox;Tie2-Cre⁻) and eAPP^−/− (APP^flox/flox;Tie2-Cre⁺) mice. Endothelial cells (ECs) were isolated from mouse aortas and subjected to PCR analysis. The remaining aortas without endothelium were also analyzed. Please note that the PCR product was 168 bp for the exon 3-deleted APP allele and it was absent in eAPP^−/− mice ECs. GAPDH was used as a loading control (168 bp).

Figure 2.

Endothelial function in isolated basilar arteries of APP^−/− and eAPP^−/− mice. Endothelium-dependent relaxations to acetylcholine were significantly reduced in APP^−/− mice compared to wild-type littermates (a; n = 6 APP^−/−; n = 8 wild-type; group-by-concentration interaction P < 0.05) and eAPP^−/− mice compared to wild-type littermates (b; n = 7 eAPP^−/−; n = 8 wild-type; group-by-concentration interaction P < 0.05). In contrast, endothelium-independent relaxations to NO-donor DEA-NONOate were unaffected in APP^−/− mice (c; n = 6 APP^−/−; n = 8 wild-type; group-by-concentration interaction P > 0.05) and eAPP^−/− mice (d; n = 5 eAPP^−/−; n = 6 wild-type; group-by-concentration interaction P > 0.05). All results are shown as mean ± SD and relaxations are expressed as percentage of the increase in intraluminal diameter from contraction with U46619 (10⁻⁸ to 3 × 10⁻⁸mol/L). *Differences with respect to single concentration between wild-type littermates and APP^−/− mice or between wild-type littermates and eAPP^−/− mice are statistically significant (P < 0.05; unpaired Student’s t-test).

Figure 3.

Representative Western blot and qRT-PCR analyses of large cerebral arteries. (a) Protein expressions of APP and eNOS in wild-type littermates and APP^−/− mice (n = 5 for each group). *P < 0.05 vs. wild-type (WT) littermates (unpaired Student’s t-test). (b) qRT-PCR analysis of APP mRNA in large cerebral arteries of wild-type littermates (APP^flox/flox;Tie2-Cre⁻) and eAPP^−/− (APP^flox/flox;Tie2-Cre⁺) mice in the presence and absence of endothelial cells (n = 3 independent experiments for each group). Please note that APP mRNA was significantly reduced in wild-type mice without endothelium (E-) while it was unchanged in eAPP^−/− mice without endothelium. *P < 0.05 vs. wild-type (WT) littermates; ^†P < 0.05 vs. wild-type littermates with endothelium (E+); n.s. indicates not significant (two-way ANOVA). (c) Protein expressions of APP and eNOS in wild-type littermates and eAPP^−/− mice (n = 6 for each group). *P < 0.05 vs. wild-type (WT) littermates (unpaired Student’s t-test). (d) Protein expression of prostaglandin I₂ (PGI₂) synthase in wild-type littermates and APP^−/− mice (n = 5 for each group). (e) Protein expression of PGI₂ synthase in wild-type littermates and eAPP^−/− mice (n = 5 for each group). All western blot results are the relative densitometry compared with β-actin protein. All results are represented as box plots with whiskers showing the median, the 25th to 75th percentiles, and min–max range.

Figure 4.

Basal levels of cGMP and cAMP in large cerebral arteries. (a) cGMP levels were significantly reduced in APP^−/− mice as compared to wild-type (WT) littermates (n = 7 per group). (b) cGMP levels were significantly reduced in eAPP^−/− mice as compared to wild-type (WT) littermates (n = 11 per group). (c) cAMP levels in wild-type (WT) littermates and APP^−/− mice (n = 6 per group). (d) cAMP levels in wild-type (WT) littermates and eAPP^−/− mice (n = 7 per group). Results were normalized against tissue protein levels, and box plots with whiskers are showing the median, the 25th to 75th percentiles, and min–max range. *P < 0.05 vs. wild-type (WT) littermates (unpaired Student’s t-test).

Figure 5.

Expression of eNOS mRNA and protein in human BMECs. Cells were transduced with siRNA against APP (APP-siRNA) or control siRNA (Ct-siRNA) for three days. Exposure of HBMECs to APP-siRNA resulted in significantly attenuated expression of eNOS mRNA (A; n = 5 per group) and protein (B; n = 4 per group). Please note that APP expression is absent in APP-siRNA treated BMECs. eNOS expression was normalized to β-actin. Box plots with whiskers are showing the median, the 25th to 75th percentiles, and min–max range. Bars in dot plot are representing the mean. *P < 0.05 vs. Ct-siRNA (unpaired Student’s t-test).

Figure 6.

Concentration-dependent contractions to endothelin-1 in isolated basilar arteries of wild-type littermates and eAPP^−/− mice. The contractions to endothelin-1 were expressed as percentage of the decrease in the basal intraluminal diameter. Results are mean ± SD (n = 6 for wild-type (WT) littermates and n = 7 for eAPP^−/− mice). *Difference with respect to single concentration between wild-type littermates and eAPP^−/− mice is statistically significant (P < 0.05; unpaired Student’s t-test).