We demonstrate that propentofylline (PPF), a methylxanthine derivative that exhibits anti-allodynic properties in a neuropathic pain rodent model, inhibits astrocytic activation and modulates spinal astrocytic promoter activation for glutamate transporters, GLT-1 and GLAST in vivo
. Using a newly engineered line of transgenic reporter mice, we specifically observed glutamate transporter alterations in the spinal cord dorsal horn and extended our previous findings with propentofylline to mice. Of note, previous work from our laboratories indicates that in primary astrocyte cultures PPF induces a similar suppression of the activated astrocytic phenotype along with enhancement of GLT-1 and GLAST protein (Tawfik et al., 2006
) and that there is a close correlation between GLT-1 and GLAST promoter activation and transporter expression and function in spinal cord (Regan et al., 2007
). Taken together, these findings suggest that glial activation and glutamate clearance capabilities are inextricably linked and may be an important target for future pain therapeutics.
Glutamate is the primary excitatory neurotransmitter in the CNS and, as such, participates in the majority of crucial brain physiological functions including synaptic transmission and CNS development (Danbolt, 2001
). In order to maintain efficient signaling via glutamate receptors and prevent excitotoxicity, the extracellular concentration of glutamate is exquisitely regulated by a series of sodium-dependent glutamate transporters. Five subtypes of excitatory amino acid transporters have been cloned (Kanai and Hediger, 1992
; Pines et al., 1992
; Storck et al., 1992
; Fairman et al., 1995
; Arriza et al., 1997
). Antisense and genetic knockout studies have identified GLT-1 as the most crucial in maintenance of low extrasynaptic glutamate levels (Rothstein et al., 1996
; Tanaka et al., 1997
) while GLAST and possibly EAAC1 play secondary roles.
Excessive accumulation of glutamate can lead to synaptic dysregulation. For example, it has previously been demonstrated that following nerve injury, there is excessive release of excitatory amino acids at the level of the spinal cord. In addition, N-methyl-D-aspartate (NMDA) receptor antagonists, such as MK-801, are capable of reversing injury-induced pain behaviors (Mao et al., 1992
; Garrison et al., 1993
). It follows that impaired uptake of synaptic glutamate via a decrease in astrocytic glutamate transporters may be, in part, responsible for neuropathic pain-related central sensitization.
Previously, Sung et al. (2003)
reported that after chronic constriction injury (CCI) of the sciatic nerve, there was a biphasic alteration in GLT-1 and GLAST protein levels; at days 1 and 4 both transporter proteins were increased, and at days 7 and 14 both were decreased. Another recent study demonstrated that following facial nerve axotomy, GLT-1 protein is decreased ipsilateral to the injury on day 3 (Lopez-Redondo et al., 2000
). In the current study, we observed a decrease in both GLT-1 and GLAST promoter activation at day 12, which is consistent with these findings. However, the magnitude of the change was approximately 35% for both GLT-1 and GLAST of the sham surgery control group in our study whereas Sung et al. (2003)
report decreases of 48% and 41%, respectively, at day 14. The sciatic nerve receives inputs from L4–L6 and therefore CCI affects a wider range of lumbar spinal cord levels. This may result in a greater neurochemical and metabolic response than does a more proximal L5 spinal nerve transection. In addition, CCI has a marked local inflammatory component (Clatworthy et al., 1995
) not present in the spinal nerve transection model which suggests that spinal cord responses may differ. Importantly, the spinal cord has several-fold less GLT-1 than cortex and therefore, it may be more sensitive to perturbations in transporter levels (Regan et al., 2007
). This suggests that even a subtle decrease in transporter levels may dysregulate glutamate uptake.
In the current study, we made use of double transgenic reporter mice which allowed us to specifically determine alterations in GLT-1 and GLAST promoter activation levels. An important limitation of our study is that we did not determine transporter function directly; however, previous work in this mouse model has demonstrated that transporter protein and uptake levels correlate well with promoter expression (Regan et al 2007
). With this tool, we determined that spinal nerve transection leads to a significant decrease in GLT-1 puncta ipsilateral to the nerve injury. The exact physiological difference between the observed puncta and diffuse expression of GLT-1 and GLAST is not clear. One can speculate that given the perinuclear location of the puncta, these represent sites of newly formed transporter, while the cytoplasmic expression is suggestive of active, membrane localized proteins. It would therefore be interesting to determine, at a later time point, whether the diffuse cytoplasmic staining after injury persists, or whether protein turnover would eventually lead to a reduction in all GLT-1 expression. Propentofylline, in contrast, enhanced GLT-1 and GLAST puncta ipsilateral to the nerve injury. In previous in vitro
work (Tawfik et al., 2006
), we showed that propentofylline enhanced mRNA for GLT-1, which is consistent with our current findings of increased promoter activation.
Importantly, the ultimate decrease in glutamate transporters, regardless of the cause, has been shown to have functional significance in the spinal cord. Recently, it was observed that glutamate uptake was reduced by 72% in the ipsilateral dorsal horn, compared to sham, 4–6 weeks following spinal nerve ligation (SNL) (Binns et al., 2005
). In addition, Liaw et al. (2005)
demonstrated that inhibition of glutamate transport with dl
-TBOA), produced spontaneous nociceptive behaviors as well as thermal hyperalgesia and mechanical allodynia in rats. Our data indicate that GLT-1 promoter activation is enhanced with PPF treatment. This transporter is responsible for over 90% of synaptic glutamate clearance (Tanaka et al., 1997
) giving functional relevance to our observations. Together, these data suggest that the suppression of glial glutamate transporter protein and/or function may be a mechanism common to sensitization, regardless of the model used, as L5 spinal nerve transection or ligation, chronic constriction injury (CCI) and facial nerve axotomy all resulted in ultimate decreases in transporter protein or function.
In this study we demonstrate a possible mechanism for the observed anti-allodynic effect of propentofylline following peripheral nerve injury, namely enhancement of the principally astrocytic glutamate transporter GLT-1. These data may direct future development of drugs that act via modulating glial function. While effective treatment of neuropathic pain remains a challenge for the clinician, the complex nature of this syndrome suggests a multimodal mechanism that may require a multi-target solution. Dissection of key players in central sensitization will result in novel targets for therapy that may provide patients with much needed alternatives and adjunct therapies to relieve their pain.