Biomedical Sciences Graduate Program | University of California, San Francisco

Ethanol, adenosine and PKA signaling.
Adenosine transport. Adenosine mediates many acute intoxicating responses to ethanol (Diamond and Gordon, 1997). We focused our initial studies on the role of A2R in mediating responses to ethanol. Using several cell lines in vitro, we have shown that A2R is required for ethanol-induced changes in cAMP signaling (Gordon et al., 1986). We discovered that ethanol inhibits adenosine uptake via a facilitative nucleoside tranporter, ENT1, leading to the accumulation of extracellular adenosine (Nagy et al., 1990). We provided evidence that ethanol inhibition of adenosine uptake appears to be regulated by cAMP-dependent phosphorylation of the transporter or a nearby protein (Nagy et al., 1991). We have cloned and expressed ENT1 (Handa et al., 2001; Choi et al., 2000) and are characterizing its regulation by phosphorylation. Robert Messing has generated mice lacking ENT1 and collaborative studies are planned to characterize cellular responses to ethanol in primary neurons in culture.

Ethanol-dependent cAMP signaling and PKA translocation.
Ethanol inhibition of adenosine uptake leads to increases in extracellular adenosine and a cascade of molecular events. Increased levels of adenosine activate A2R to stimulate cAMP production via Gas (Sapru et al., 1994). cAMP binds to the regulatory subunits of PKA, thereby releasing catalytic subunits (PKA Ca) to change cell function and gene expression. Our first observation was that ethanol induced sustained PKA Ca translocation into the nucleus (Dohrman et al., 1996). There are two phases of ethanol-induced PKA Ca translocation (Dohrman et al., 2002). The first phase occurs in a few minutes and requires A2R activation by adenosine. The second phase extends over hours and requires protein synthesis (Dohrman et al., 2002). PKA Ca remains in the nucleus as long as ethanol is present (Dohrman et al., 1996). Thus, ethanol-induced activation of A2R-dependent PKA signaling and sustained localization of PKA Ca in the nucleus appear to be due to two distinct pathophysiologic events. Studies are underway to identify a peptide or protein which binds PKA Ca and keeps it in the nucleus.

Ethanol-induced PKA subunit translocation, CREB phosphorylation and CRE-mediated gene expression. We have shown that ethanol causes CREB phosphorylation (Constantinescu et al., 1999) and CRE-mediated gene transcription in NG108-15 cells (Asher et al., 2002). Ethanol-induced CRE-mediated gene transcription also exhibits two phases. The first phase requires A2R; the second phase does not. However, little is known about which PKA isotypes regulate this process. We have found that, under basal conditions, NG108-15 cells contain type I PKA (CbRIb) primarily in cytosol and type II PKA (CaRIIb) in the particulate and nuclear fractions. Antagonists of both type I and type II PKA inhibit ethanol-induced CRE-mediated gene transcription. However, only the type II PKA antagonist inhibits ethanol-induced PKA Ca translocation to the nucleus and CREB phosphorylation; the type I antagonist is without effect (Constantinescu et al., 2002). Our data suggest that ethanol-induced CREB phosphorylation and gene activation are differentially mediated by the two types of PKA. It is likely that each type of PKA regulates different molecular mechanisms contributing to sustained drinking.

Synergy of ethanol action on neural cells.
The NAc is populated by medium spiny neurons which express the Gi-coupled D2R and Gs-coupled A2R on the same cells. Wild-type NG108-15 cells express A2R. In order to study neuronal responses to ethanol in the presence of A2R and D2R, we developed an NG108-15/D2 cell line that expresses D2R (Asai et al., 1998). We confirmed that the D2R is functionally active in these cells (Gordon et al., 2001; Yao et al., 2001). D2R agonists are known to inhibit adenylyl cyclase activity (AC). However, D2R activation can sometimes increase cAMP. We asked whether D2 could activate the PKA pathway in NG108-15/D2 cells. We found that a D2 agonist activates PKA signaling ranging from increases in cAMP to CRE-mediated gene expression. The mechanism involves D2 coupling to Gi/o, release of bg dimers, activation of AC II and/or IV, translocation of PKA Ca to the nucleus and subsequent PKA-dependent increases in gene expression (Yao et al., 2002). Most importantly, we find a remarkable synergy between D2 and ethanol-induced activation. Subthreshold concentrations of NPA or ethanol, which have no effect alone, when added together induce maximal translocation of PKA and activation of CRE-mediated gene expression (Yao et al., 2002). This makes neurons with D2R and A2R, like those in the NAc, hypersensitive to ethanol. Synergy requires Gi bg stimulation of AC II and/or IV concomitant with ethanol/A2R activation of AC via Gas (Yao et al., 2002). Another example of the importance of bg interaction comes from a collaborative study with Antonello Bonci. Here we find that bg dimers and Gas mediate cooperative increases in spike firing in NAc neurons caused by D1 and D2 agonists (Hopf et al., submitted).

Release of Gi bg dimers in NAc mediates voluntary alcohol consumption.
Because about 90% of medium spiny neurons in the NAc express D2 and A2 receptors on the same cells, our data suggested that bg dimers released in NAc by dopamine might play a role in sustaining drinking behavior. In collaboration with Patricia Janak, we designed experiments to determine whether release of bg dimers is required for continual ethanol consumption in a two-bottle choice paradigm where rats are allowed continuous access to ethanol. We used an adenovirus expressing the carboxyl terminus of bARK (Ad5bARK minigene) which binds and inactivates bg dimers. We found that bilateral microinjection of the AD5bARK minigene into the medial NAc decreased voluntary intake of ethanol at day 7 after injection (Yao et al., 2002). Water intake was unaffected despite maximal viral expression in neurons at this time point. Rats recovered their pre-injection levels of ethanol intake by day 14, a time when we find diminished viral expression. Our results strongly suggest that expression of a bg inhibitor peptide in the medial NAc reduces voluntary consumption of ethanol. The exact mechanism whereby bg inhibition reduces ethanol consumption is an exciting observation that requires further study. Our hypothesis is that the decrease in ethanol consumption we observe after inhibiting bg function in vivo may reflect a disruption of D2R- and ethanol/A2R-induced synergistic PKA activation.

Pathophysiologic significance of A2R and D2R synergy. The functional significance of PKA translocation induced by subthreshold levels of NPA and ethanol is suggested directly by synergistic increases in CRE-mediated gene expression and indirectly by decreases in ethanol consumption caused by expression of a dominant negative bg inhibitor peptide in the NAc (Ad5bARK minigene). Synergy between dopamine and ethanol requires activation of D2R with release of Gi bg subunits together with activation of A2R/Gas by adenosine. It is tempting to speculate, therefore, that the simultaneous expression of D2 and A2 receptors in the NAc may account for the central role of this brain region in regulating ethanol intake. It is possible that antagonists which target adenosine and dopamine receptors alone or in combination might prevent, attenuate or reverse excessive drinking by blocking CRE-mediated gene expression.