Genome-wide screen identifies drug-induced regulation of the gene giant axonal neuropathy (Gan) in a mouse model of antiretroviral-induced painful peripheral neuropathy.

Nearly 40 million children and adults worldwide are living with HIV/AIDS (UNAIDS, 2007). In the United States, it is estimated that over 1 million persons are currently infected with HIV and an average of 50,000–60,000 new cases are diagnosed each year (Centers for Disease Control [CDC], 2008; Hall et al., 2008). The highly active antiretroviral therapy (HAART) used to treat HIV is often initiated early in the infection before patients develop symptoms of AIDS and has been credited with extending the lives of HIV patients. HAART involves the combined use of multiple anti-HIV drugs, usually including a protease inhibitor and one or more of the nucleoside reverse transcriptase inhibitors ([NRTIs] Simpson, 2002; White, 2001). Each of these drugs has a different mechanism of action and they work synergistically to interrupt the life cycle of the HIV. HAART is extremely effective in reducing viral load and maintaining and/or improving immune function. However, HAART is also associated with significant side effects including diarrhea, pancreatitis, liver failure, and peripheral neuropathic pain. The symptom burden of these side effects leads one-in-four HIV-infected patients to limit or stop treatment to improve their comfort and quality of life (d’Arminio Monforte et al., 2000). Thus, a clear understanding of the mechanisms underlying the development and persistence of the HAART side effects is of utmost importance. Painful peripheral neuropathy is one of the most prevalent side effects of HIV treatment and is a complication of HIV infection, itself, occurring in approximately 35% of patients (Moore, Wong, Keruly, & McArthur, 2000; Schifitto et al., 2002; So, Holtzman, Abrams, & Olney, 1988). The drug-induced painful peripheral neuropathy resulting from HAART is most closely associated with the dideoxynucleoside (d-drug) family of NRTIs, including stavudine, lamivudine, and zalcitabine (Dalakas, 2001; Moore et al., 2000; White, 2001). The mechanism of action for the d-drugs is to block the function of reverse transcriptase, an enzyme that is necessary for the HIV virus to build new DNA from RNA. However, the d-drugs also strongly inhibit g-polymerase, resulting in a dose- and time-dependent decrease in levels of intracellular mitochondrial DNA, especially in the liver, skeletal muscle, and peripheral nerves (Walker, 2003; for review see Dalakas, 2001 and Simpson, 2002). The mitochondrial toxicity occurring in the long sensory nerves of the lower extremities destabilizes peripheral nerve health, leading to the dieback of distal small sensory fibers and epidermal denervation and possibly contributing to the development of neuropathic pain (McCarthy et al., 1995; Polydefkis et al., 2002; for review see Dorsey & Morton, 2006). Patients with NRTI-induced painful peripheral neuropathy present with increasingly debilitating and intractable pain that begins bilaterally in the feet. The symptoms include painful responses to normally innocuous stimuli such as the touch of socks and bed sheets (allodynia), exaggerated painful symptoms to noxious stimuli (hyperalgesia) and significant spontaneous pain (Brinley, Pardo, & Verma, 2001; Cornblath & McArthur, 1988; Wulff, Wang, & Simpson, 2000). Further, these patients frequently alter their gait to avoid the pain that results from placing pressure on the soles of the feet, limiting their mobility and increasing their risk for falls (Verma, 2001). While painful symptoms associated with peripheral neuropathies of all etiologies are difficult to treat, NRTI-induced neuropathic pain is particularly difficult to manage, as few drugs used to manage neuropathic pain, alone or in combination, are effective in relieving HIV patients’ pain (Simpson, 2002). Although withdrawal from HAART can significantly improve neuropathic symptoms within 16 months, in some cases withdrawal can also exacerbate painful symptoms and not all patients will recover from this disorder after the drugs are stopped (Berger et al., 1993). Moreover, cessation of therapy is not usually a viable option because drug treatment is necessary to maintain virologic control and a functional immune system (Moyle & Sadler, 1998; Simpson, 2002; Verma, 2001). Thus, there is a great need for new interventions and discovery-based research to identify novel therapeutic opportunities that can improve pain management and quality of life for patients with NRTI-associated peripheral neuropathic pain. Several research groups have examined mechanisms underlying antiretroviral-induced neuropathic pain in rodent models. These studies have linked persistent allodynia to microgliosis (Wallace et al., 2007), sensory C-fiber axonal “dieback” (Wallace et al., 2007), altered intracellular calcium buffering secondary to drug-induced mitochondrial damage in dorsal root ganglion neurons (Joseph, Chen, Khasar, & Levine, 2004), enhanced neurotransmitter release and increased chemokine signaling in the spinal dorsal horn (Bhangoo et al., 2007; Joseph et al., 2004). These findings are based on tissue sampling and analysis that occurred relatively late in the course of symptoms, suggesting that the drug effects linger or representing evolved nerve pathology. They do little, however, to illuminate the mechanisms underlying the development of allodynia. To explore the developmental mechanisms of this disorder in more detail, we generated a mouse model of NRTI-induced painful peripheral neuropathy by giving a single weight-based intravenous injection of stavudine that produced a robust tactile allodynia within 24 hr of drug treatment. Using this mouse model, we examined molecular changes that occurred in the dorsal horn of the spinal cord, which is the principal region where integration of ascending pain transmission and descending pain modulatory effects occur. We chose to exploit the power of an unbiased, whole-genome approach to examine these changes. In contrast to a more traditional candidate approach where prior research and theory drives hypothesis testing of known molecules of interest, use of whole-genome microarray technology allows exploration of gene changes at specific time points following drug treatment and increases the likelihood of discovering novel molecular targets involved in either known or novel pain pathways.