The actions of adenosine are mediated by 4 types of G protein-coupled membrane receptors, A1, A2a, A2b, and A3 adenosine receptors (A1AR, A2aAR, A2bAR, A3AR). All types of adenosine receptors are expressed in the kidney, both in vascular and epithelial cells. We have focused on the role of adenosine 1 receptors (A1AR) which are predominantly found in afferent arterioles and at lower expression levels in proximal tubules and other nephron segments. Adenosine causes a long-lasting and dose-dependent vasoconstriction of afferent arterioles that is most pronounced in the terminal portion of the vessel. In perfused mouse afferent arterioles adenosine-induced vasoconstriction is caused by Gi-protein-dependent activation of PLC and subsequent release of Ca from SR stores through an IP3-dependent pathway. Vasoconstriction is maintained by activation of Ca-activated Cl channels, depolarization, and Ca influx through voltage-dependent Ca channels. To cause afferent arteriolar vasoconstriction, adenosine had to be added to the bath. Inclusion of adenosine in the perfusate caused vasoconstriction only after inhibition of NO synthases indicating that intravascular adenosine leads to the release of endothelial dilatators that counteract the A1AR-mediated smooth muscle cell activation. Studies in A1AR knockout mice have shown that the vasoconstriction that is normally elicited by an increase in NaCl concentration at the macula densa site of the nephron (tubuloglomerular feedback, TGF) is completely absent in A1AR-deficient animals suggesting that adenosine is a critical component of the TGF signaling pathway. Fluid and electrolyte absorption along proximal tubules and loops of Henle were largely normal in A1AR-/- mice in contrast to the inhibitory effect of acute administration of A1AR antagonists on proximal fluid reabsorption. Thus, while the role of adenosine in TGF is non-redundant, its effect on proximal transport is fully compensated during chronic A1AR deficiency. The recent demonstration of NaCl-dependent ATP release by MD cells (Bell et al. PNAS 2003) raises the possibility that adenosine in the juxtaglomerular interstitium may be formed from nucleotide precursors by the action of ecto-ATPases with the final step from 5-AMP to adenosine being mediated by ecto-5-nucleotidase/CD73 (e-5-NT/CD73). In order to study the possible role of this pathway in TGF, we generated e-5-NT/CD73-deficient mice by gene targeting. Wild type mice, and mice heterozygous and homozygous for the e-5-NT/CD73 null mutation were born in a near Mendelian ratio. E-5-NT/CD73 -/- mice were viable and showed no gross anatomical abnormalities. RT-PCR and immunohistochemistry confirmed the absence of e-5-NT/CD73 mRNA transcripts and protein in the kidneys of e-5-NT/CD73 knockout mice. Blood pressure measured by the tail cuff method and blood chemistry with the exception of a significant increase in alkaline phosphatase appeared normal. E-5-NT/CD73-/- mice showed a marked decrease in the bradycardic response of i.v. injected 5-AMP, an effect that is mediated by adenosine and absent in A1AR-/- mice. Micropuncture experiments to address TGF function showed that e-5-NT/CD73-/- mice had a significantly reduced change in stop flow pressure (Psf) to an increase of tubular perfusion flow from 0 to 30 nl/min of -5.5 ±.8 mmHg (n=24) compared to their wild type littermates (-9.8 .±9 mmHg, n=13, p=.002). Similarly, the fall of early proximal flow rate to the same saturating loop flow increase was significantly reduced (-5.3 ±0.5 nl/min in wild type, and -1.6 ±0.5 nl/min in e-5-NT/CD73-/-; p<.0001). Furthermore, whereas TGF responses did not change significantly during repeated changes in perfusion flow (up to 5 times) in e-5-NT/CD73+/+ mice, a gradual decline in the magnitude of the residual TGF response was noted in e-5-NT/CD73-/- mice with TGF responses falling from the initial decrease of -5.5 mmHg (n=24) to only -1.3 ±.5 mmHg during the fifth response (n=10, p=.004 compared to first response; p<.0001 compared to wild types). These data are in agreement with earlier pharmacological evidence supporting a role of e-5-NT/CD73 in TGF (Thomson et al., JCI 2000). Preliminary observations in P2X1 knockout mice indicate that the presence of this particular P2 receptor subtype is not necessary for TGF responsiveness. We conclude that adenosine has a unique vasoconstrictor role in the kidney that is the result of a highly selective expression of A1AR in a small portion of the afferent arteriole. These receptors are activated in a paracrine fashion by adenosine that is generated in the juxtaglomerular interstitial space and reaches its binding sites from the outside of the vessel. The formation of adenosine in the JG interstitium depends to a large extent on e-5-NT/CD73-mediated hydrolysis of 5-AMP that is presumably generated from released ATP.