Jennifer Whistler, PhD

GPCR Trafficking and the Regulation of Responsiveness to Drugs of Abuse

My laboratory studies the trafficking of GPCRs important in addiction, pain and neuropsychiatric disease.  We have active programs studying opioid, dopamine, and cannabinoid receptors and several collaborations and pilot projects with other GPCRs. Our most advanced project has focused on the mu opioid receptor (MOR), one of three members of the opioid receptor family. Like most GCPRs, the MOR undergoes a cascade of regulatory events following its activation by endogenous peptide ligands, such as enkephalin and endorphin. These include phosphorylation by GPCR kinases (GRKs), recruitment of arrestin, and endocytosis.  Following its endocytosis, the MOR is then dephosphorylated and recycled to the plasma membrane where it is available for the next round of signal transduction (Whistler et al., Science 2002).  Several opioid agonist drugs behave like endogenous peptide ligand with regards to receptor trafficking, including methadone, etorphine and, at high concentrations, fentanyl. However, the prototypical, and clinically most common, opioid agonist drugs, including morphine and oxycodone (and heroin), fail to promote endocytosis of the MOR. My laboratory has hypothesized that the failure of morphine to promote endocytosis of the MOR contributes to the high liability of this drug to promote the side of effects of analgesic tolerance and withdrawal symptoms indicative of physical dependence.  This hypothesis is supported by our observations that, at equi-analgesic doses, agonists such as methadone and enkephalin have a reduced liability, compared to morphine, to promote tolerance and dependence in rodent models. However, each time a different drug is used in these paradigms, not only is endocytosis of the receptor altered, but also multiple other pharmacological properties including ligand affinity, potency, efficacy and bioavailability. Thus, it has been difficult, using pharmacology alone, to examine the precise role of receptor endocytosis in tolerance and dependence.

To directly assess the role of MOR trafficking in the development of tolerance and dependence to morphine, we therefore generated mutant receptors that are phosphorylated, recruit arrestin and are endocytosed in response to morphine.  One of these receptors, RMOR, is endocytosed and recycled with morphine. In cell-based models we demonstrated that the RMOR receptors show opioid pharmacology undistinguishable from wild-type MORs (affinity, efficacy, potency) and reduced cAMP superactivation in response to chronic morphine, a cellular hallmark of tolerance and dependence.  We also have developed a MOR, AMOR, that is not endocytosed even upon activation by endogenous peptide or methadone, due to mutation of the GRK phosphorylation sites. In our cell-based work, we demonstrated that cAMP superactivation was enhanced in response to methadone in cells expressing the AMOR receptors (Finn and Whistler, Neuron 2001).  This strong cell-based data, demonstrating that enhancing endocytosis decreases, while preventing endocytosis increases, a cellular hallmark of tolerance and dependence, inspired us to generate a knock-in mouse that expresses the RMOR receptor.  Pharmacological and histological analysis showed that receptor expression, number, affinity and signaling of the MOR and RMOR receptors were indistinguishable. In addition, analgesia to morphine was enhanced in RMOR mice compared to their wild-type littermates.  Thus, desensitization and endocytosis of the receptor does not decrease the efficacy of morphine, presumably because the receptor is rapidly recycled and thus resensitized following endocytosis. Analgesia in the RMOR mice is selectively enhanced in response to morphine, as methadone analgesia is indistinguishable between WT and RMOR mice. Together these data imply that it is specifically the endocytosis of the receptor in response to morphine that is enhancing analgesia. Importantly, mice expressing the RMOR receptor also develop substantially reduced tolerance and dependence to morphine despite the fact that it is a better analgesic in these mice (Kim et al, Current Biology, 2008).  Ongoing experiments are examining whether biochemical, pharmacological and electrophysiological hallmarks of tolerance and dependence are altered in the RMOR mice compared to their MOR littermates. Ongoing experiments are also designed to model addiction in the MOR and RMOR mice to determine whether endocytosis contributes not only to the adaptive changes responsible for analgesic tolerance and dependence, but also to the changes responsible for drug reinforcement, reward, craving and relapse in a mouse model of operant self-administration. In short, these mice will provide an outstanding model to dissect the specific changes in gene expression and neuronal plasticity that are responsible for tolerance and dependence.

Taken together our existing data imply that opioid drugs with the pharmacological properties of morphine, but the ability to promote endocytosis of the MOR would be outstanding analgesics with a reduced propensity for promoting tolerance and dependence.  We have developed one such opioid that takes advantage of the observation that the MOR appears to form dimers/oligomers. A cocktail consisting of an analgesic dose of morphine plus a small, non analgesic dose of methadone (or enkephalin) promotes endocytosis of the MOR, presumably because a single methadone-occupied protomer of the MOR is sufficient to promote endocytosis of an oligomeric complex containing both methadone and morphine-occupied receptors (He et al, Cell 2002, He et al Current Biology, 2005). Rats receiving morphine alone develop tolerance and dependence within 5 days.  However, rats receiving a cocktail consisting of the same dose of morphine spiked with a small dose of methadone, that promotes MOR endocytosis, do not develop tolerance and dependence.  Lower doses of methadone, that do not promote morphine-induced endocytosis, do not prevent tolerance and dependence (He et al, Current Biology 2005). The long-term goal is to test the morphine-methadone cocktail in human pain trials for analgesic tolerance.

The studies with the opioid cocktail, led us to believe that opioid receptor dimers do, in fact, exist. Indeed, we have observed that opioid receptors exist as both homomeric and heterodimeric receptor complexes. We have identified an opioid receptor ligand that is an antagonist at homomeric delta opioid receptors (DORs) but a potent agonist at heterodimers of the DOR and kappa opioid receptor (KOR) (Waldhoer et. al,  PNAS 2005).  We found this ligand is a tissue-selective analgesic showing good antinociceptive potency spinally but not centrally. This was the first demonstration of functional relevance of an opioid receptor heterodimer. These data suggest that opioid receptor heterodimers do indeed exist in vivo and could represent novel and pharmacologically distinct targets, possibly with tissue-selective expression.

Opioid receptors are not only important targets for analgesia, but also one of the few targets currently available for the treatment of alcoholism.  Naltrexone, a non-selective opioid receptor antagonist, has shown limited efficacy for the treatment of alcoholism.  This may be due, in part, to the non-selective nature of this drug.  Current efforts in the laboratory are focused on identifying whether particular opioid receptor subtypes or heterodimers could be more effective targets for the treatment of alcoholism. We are also examining whether the composition or distribution of opioid receptor heterodimers or subtypes is altered in several disease models, including chronic pain, morphine tolerance, drug sensitization, and chronic alcohol consumption. To facilitate these studies, we have generated mice with a disruption of each opioid receptor gene. To facilitate our efforts to map the brain areas and circuits responsible for opioid receptor related behaviors we have also generate transgenic mice with a conditional “floxed” allele of each opioid receptor gene.  

Following their endocytosis, GPCRs can be either recycled to the plasma membrane, which serves to “resensitize” cells to the next agonist exposure, or they can be targeted for degradation, which leads to downregulation of receptor function. We have found that the MOR is a “recycling” receptor  while the DOR is targeted for degradation.  This degradation appears to be modulated through interaction with GPCR-associated sorting protein (GASP) (Whistler et. al, Science 2002). We have found that GASP interacts with members of several other GPCR families including the cannabinoid CB1 receptor (Martini et al, FASEB J. 2007), the D2 dopamine receptor (Bartlett et al, PNAS 2005), and the Bradykinin B1 receptor (Enquist et. al, Mol. Pharm. 2006) as well as several mutant GPCRs whose trafficking has switched from recycling to degradation (Thompson et. al, JBC 2007).
           
Importantly, we found that interaction of the CB1 receptor with GASP promotes downregulation of the receptor (Martini et. al, FASEB J. 2007).  Disrupting the interaction between GASP and the CB1 prevents agonist-induced receptor downregulation in vivo and the development of analgesic tolerance to spinal cannabinoids (Tappe-Theodore et. al, J. Neurosci. 2007).

In the dopamine receptor family, we have found that D2 but not D1 receptors bind GASP and are targeted for post-endocytic degradation (Bartlett et. al, PNAS 2005). This observation has significant implications towards our understanding of dopaminergic signaling, which is altered in several neuropsychiatric diseases including Parkinson’s disease, schizophrenia, attention deficit hyperactivity disorder and depression.  Both D1 and D2 dopamine receptors respond to the same agonist ligand, dopamine.  However, they are coupled to G proteins with profoundly different signaling outcomes.  Hence, if D1 and D2 receptors have different post-endocytic fates following activation by dopamine, a second dopamine exposure of a cell or a circuit that expresses both receptor types would be expected to have an altered signaling profile, with a D1 response predominating.  Current studies are aimed and elucidating the role of GASP-mediated GPCR degradation in several disease models using mice with a disruption of the GASP gene.

In summary, we have developed a unique set of tools, including cell lines and transgenic animals, that allow us to evaluate the role of GPCR trafficking in adaptations to chronic treatment with drug, and in animal models of neuropsychiatric disease. My laboratory integrates multiple approaches, including cell biology, pharmacology, electrophysiology and in vivo models of behavior to address the role of receptor trafficking in addiction and other neuropsychiatric diseases. Thus, for each project and receptor class, we have translated our in vitro and cell-based observations to preclinical animals models. We are also actively pursuing clinical partners for development of novel opioid receptor ligands or ligand cocktails for treatment of chronic pain without the development of opioid tolerance.