Dopamine modulates diurnal and circadian rhythms of protein phosphorylation in photoreceptor cells of mouse retina

Abstract

Many aspects of photoreceptor metabolism are regulated as diurnal or circadian rhythms. The nature of the signals that drive rhythms in mouse photoreceptors is unknown. Dopamine amacrine cells in mouse retina express core circadian clock genes, leading us to test the hypothesis that dopamine regulates rhythms of protein phosphorylation in photoreceptor cells. To this end we investigated the phosphorylation of phosducin, an abundant photoreceptor-specific phosphoprotein. In mice exposed to a daily light-dark cycle, robust daily rhythms of phosducin phosphorylation and retinal dopamine metabolism were observed. Phospho-phosducin levels were low during the daytime and high at night, and correlated negatively with levels of the dopamine metabolite 3,4-dihydroxyphenylacetic acid. The effect of light on phospho-phosducin levels was mimicked by pharmacological activation of dopamine D4 receptors. The amplitude of the diurnal rhythm of phospho-phosducin was reduced by > 50% in D4 receptor-knockout mice, due to higher daytime levels of phospho-phosducin. In addition, the daytime level of phospho-phosducin was significantly elevated by L-745,870, a dopamine D4 receptor antagonist. These data indicate that dopamine and other light-dependent processes cooperatively regulate the diurnal rhythm of phosducin phosphorylation. Under conditions of constant darkness a circadian rhythm of phosducin phosphorylation was observed, which correlated negatively with the circadian rhythm of 3,4-dihydroxyphenylacetic acid levels. The circadian fluctuation of phospho-phosducin was completely abolished by constant infusion of L-745,870, indicating that the rhythm of phospho-phosducin level is driven by dopamine. Thus, dopamine release in response to light and circadian clocks drives daily rhythms of protein phosphorylation in photoreceptor cells.

Fig. 1. Characterization of antibody AN519 (anti-Pdc-Ser⁷³)

A. Time course of purified bovine Pdc phosphorylation by PKA, analyzed by autoradiography and Western Blotting. B. Specificity of AN519 for phosphorylated bovine Pdc. Increasing amounts of phosphorylated (Pdc-pSer⁷³) or non-phosphorylated (Pdc) bovine phosducin were subjected to Western blot analysis with AN519. C. Phosphorylation of Pdc in mouse retinal homogenate. Mouse retinal extract (40 μg) was incubated with ATP/Mg, okadaic acid, IBMX with (lanes 1, 3, 5) or without cAMP (lanes 2, 4) followed by SDS/PAGE. The proteins were visualized either by Coomassie brilliant blue (CBB) staining or Western blot analysis with AN519 or anti-Pdc-pan. See Materials and methods for details.

Fig. 2. Diurnal changes of phosducin phosphorylation state and dopamine and DOPAC levels in mouse retina

Retinas of C57Bl/6J mice, kept on a 12 hr light – 12 hr dark (LD) cycle, were dissected at the times indicated. From each mouse, one retina was used to measure levels of pSer⁵⁴-Pdc, pSer⁷¹-Pdc, and total (pan) Pdc, and the other retina used to measure dopamine and DOPAC content. One-way ANOVA shows significant diurnal rhythms in pSer⁵⁴-Pdc, pSer⁷¹-Pdc (A, B) and DOPAC (D) (p<0.001) but not pan-Pdc (A) or dopamine (C); n = 4 per time point. A significant negative correlation exists between the degree of Pdc phosphorylation and retinal DOPAC content (r = −0.82 and −0.84 for pSer⁵⁴-Pdc/DOPAC and pSer⁷¹-Pdc/DOPAC pairs, respectively; Pearson correlation, p<0.001).

Fig. 3. Agonists of dopamine D4 receptors induce dephosphorylation of Pdc in the retina and promote Pdc/ G_tβ interaction

C57Bl/6J mice, which had been dark adapted for 14 h beginning at ZT 12, were injected intraperitoneally (ip) with PD168077 (A, B), quinpirole (B, C), vehicles, or were exposed to light for 30 min (100 μW/cm²) (A, C). Retinas were dissected 30 min after injection or the beginning of light exposure. A. PD168077 (1 mg/kg of body weight) caused dephosphorylation of both Ser⁵⁴ and Ser⁷¹ of Pdc mimicking the effect of light (n=3; ANOVA p<0.001). B. The effects of PD168077 and quinpirole on Ser⁷¹-Pdc were dose dependent (n= 4; ANOVA p<0.001 for both drugs) C. Proteins from dark-adapted (D), dark-adapted quinpirole-treated (Q, 1 mg/kg b.w. for 30 min) or light-treated (L, 30 min, 100μW/cm²) retinas were subjected to immunoprecipitation with anti-G_tβ antibody (G_β1, C-16) or non-immune rabbit IgG. Precipitated proteins were analyzed by immunoblotting (anti-Pdc-pan, 1:20,000 and G_β1, 1:1,000 for Pdc and G_tβ, respectively). Results shown are representative of 3 independent experiments. Quinpirole and light promoted the interaction of Pdc with G_tβ.

Fig. 4. Role of dopamine D4 receptors in the regulation of Pdc Ser⁷¹ phosphorylation

Pdc Ser⁷¹ phosphorylation was studied in dark-adapted wild type (Drd4^+/+) and dopamine D4 receptor knock-out (Drd4^−/−) mice on a C57Bl/6 background. A. Mice were injected ip with quinpirole (1 mg/kg body weight [b.w.]) and retinas were dissected in darkness 15, 30 and 60 min after drug administration. Two-way ANOVA indicated significant effects of quinpirole (p<0.001), genotype (p<0.001), and a significant interaction of quinpirole and genotype (p=0.011); n=4 / group. B. pSer⁷¹-Pdc/Pdc was measured in Drd4^+/+ and Drd4^−/− mice, maintained on a regular 12 h light/12h dark schedule, at ZT6 (middle of the day) and ZT18 (middle of the night). Both time of day (p<0.001) and genotype (p=0.02) significantly affected phosphorylation of Pdc Ser⁷¹, and a significant interaction of time and genotype was observed (p=0.005); n=4 / group. C. The dopamine D4 receptor antagonist L-745,870 (1 mg/kg b.w.) increases the amount of pSer⁷¹-Pdc during the daytime (ZT6). n=4 / group, t-test, * -p<0.05). Representative immunoblots and densitometry data are shown for each experiment.

Fig. 5. Diurnal changes of retinal dopamine and DOPAC in Drd4^+/+ and Drd4^−/− mice

Retinas from wild type and mutant mice were dissected at ZT6 during the daytime (in light) and at ZT18 at night and dopamine and DOPAC levels were measured by HPLC with electrochemical detection. n = 5 / group. No significant differences in dopamine or DOPAC levels between genotypes were found (two-way ANOVA, p>0.05).

Fig. 6. Circadian rhythms of pSer⁷¹-Pdc/Pdc, dopamine, and DOPAC in the retinas of C57Bl/6J and C3H/f^+/+ mice

Retinal concentrations of pSer⁷¹-Pdc, total Pdc, dopamine and the dopamine metabolite DOPAC were measured in two strains of mice, C57Bl/6J (A–D) and C3H/f^+/+ (E–H), on the second/third day of constant (24h/day) darkness. Two-factor ANOVA shows a significant effect of “circadian time” for C3H/f^+/+ mice but not C57Bl/6J mice for pSer⁷¹-Pdc/Pdc (n=4–5; p<0.001), dopamine (n=5; p<0.05) and DOPAC (n=5; p<0.001); and significant interactions between factors circadian time and genotype (p<0.05 for pSer⁷¹-Pdc/Pdc ratio and dopamine, p<0.001 for DOPAC). A significant negative correlation existed between retinal DOPAC content and pSer⁷¹-Pdc/Pdc ratio (Pearson correlation coefficient r = −0.77, p<0.001).

Fig. 7. Abolition of the circadian rhythm of Pdc phosphorylation state by the dopamine D4 receptor antagonist L-745,870

C3H/f^+/+ mice were implanted with osmotic pumps releasing L-745,870 or vehicle and transferred to constant darkness. On the second day of constant darkness, retinas were dissected in the middle of subjective day (circadian time 6 (CT6)) and the middle of subjective night (CT18). From each mouse, one retina was used to measure levels of pSer⁵⁴-Pdc, pSer⁷¹-Pdc and total (pan) Pdc and the other retina to measure the content of L-745,870. The levels of L-745,870 was 156±26 and 129±21 pg / retina at CT6 and CT18, respectively (n=5, t-test, p=0.47) in drug treated mice. The degree of phosphorylation of both Ser⁵⁴ and Ser⁷¹ exhibited circadian rhythms in vehicle treated mice (two-way ANOVA, Student-Newman-Keuls test, p<0.001, n=5) that were eliminated by the dopamine D4 receptor antagonist (p>0.05).

Fig. 8. A working model for the diurnal and circadian control of photoreceptor protein phosphorylation state

See text for details.