Diminished iron concentrations increase adenosine A(2A) receptor levels in mouse striatum and cultured human neuroblastoma cells

Abstract

Brain iron insufficiency has been implicated in several neurological disorders. The dopamine system is consistently altered in studies of iron deficiency in rodent models. Changes in striatal dopamine D(2) receptors are directly proportional to the degree of iron deficiency. In light of the unknown mechanism for the iron deficiency-dopamine connection and because of the known interplay between adenosinergic and dopaminergic systems in the striatum we examined the effects of iron deficiency on the adenosine system. We first attempted to assess whether there is a functional change in the levels of adenosine receptors in response to this low iron. Mice made iron-deficient by diet had an increase in the density of striatal adenosine A(2A) (A(2A)R) but not A(1) receptor (A(1)R) compared to mice on a normal diet. Between two inbred murine strains, which had 2-fold differences in their striatal iron concentrations under normal dietary conditions, the strain with the lower striatal iron had the highest striatal A(2A)R density. Treatment of SH-SY5Y (human neuroblastoma) cells with an iron chelator resulted in increased density of A(2A)R. In these cells, A(2A)R agonist-induced cyclic AMP production was enhanced in response to iron chelation, also demonstrating a functional upregulation of A(2A)R. A significant correlation (r(2)=0.79) was found between a primary marker of cellular iron status (transferrin receptor (TfR)) and A(2A)R protein density. In conclusion, the A(2A)R is increased across different iron-insufficient conditions. The relation between A(2A)R and cellular iron status may be an important pathway by which adenosine may alter the function of the dopaminergic system.

Fig. 1

Dietary iron deficiency increases levels of A_2AR and TfR and decreases D2R in the striatum of mice. A, Representative Western blots from two controls (C, C) and two ID mice striata. Global protein density is controlled with actin immunoblotting B. Densitometric analysis of the results of panel A; data are given as means±SEM (n=8 per group). a, p<0.001.

Fig. 2

Effects of genetic variability in striatal iron concentration on A_2aR levels. (A) Representative Western blots from striata of two strain-40 and two strain-1 animals probed for A_2AR and actin. (B) Densitometric analysis (means±SEM, n=8 per group). a, p<0.001.

Fig. 3

Iron chelation increases levels of A_2AR and TfR in cultured human neuroblastoma cells. (A) Representative Western blots of cells treated with 0-100 μM DFO for 24 h. (B) DFO (25-100 μM) concentrations significantly increased both the A_2AR and TfR levels after a 24-h exposure; no changes were seen in the levels of A₁R, HSP-70 or SAPK/JNK. Values are mean+SEM, with 4 animals per group. a, p<0.001 as compared to 0 μM DFO. (C) Representative confocal images of non-permeabilized SH-SY5Y cells, treated with 50 μM DFO for 24 h. Cells were stained with the antibody, which only recognizes the extracellular domain of A_2AR, (D) Mean fluorescence intensity±SEM (n=40 cells) treated with DFO.

Fig. 4

Time course analysis of the effects of iron chelation on A_2AR and TfR levels in human neuroblastoma cells. Cells were treated with 50 μM DFO for 2, 6, 12,or 24 h. (A) Representative Western blots for A_2AR and TfR density with protein control blotting for actin (a, p<0.001). (B) Results from densitometric analysis are shown as means±SEM. (n=4 per group, a, p<.001 (as compared with 0 μM DFO)). (C) Correlation analysis between A_2AR and TfR in cells treated with 50 μM DFO for 24 h. (D) DFO treatment induced a trend for alterations in A_2AR mRNA levels but results were not statistically significant (mean±SEM, n=6).

Fig. 5

Acute iron chelation sensitizes human neuroblastoma cells to selective A_2AR ligands. Confluent SH-SY5Y cells were serum-deprived in the presence of adenosine deaminase (2 IU/ml) for 16 h before experimentation. At the same time cells were either treated with a vehicle control or with 50 μM DFO for 16 h before experimentation. CGS-21680 was applied at each dose specified for 10 min, and cells were then permeabilized, and intracellular cAMP was extracted and assessed by fluorescent immunoassay. Control concentration response curve (black squares) and DFO-treated (open triangles) values represent the mean±SEM from three separate experiments. The curve fit was generated using Graph Pad Prism (sigmoidal concentration-response relationship) analysis of the mean responses to three individual experiments comparing CGS-21680 concentration-response series in the presence or absence of DFO pre-treatment.