Introduction
Chronic neuropathic pain is often stubbornly resistant to analgesics and persists long after the initial injury has healed1. Although it is clear that neuropathic pain has a central component, the complex interplay between the disparate cellular subpopulations of the central nervous system (CNS) has only recently emerged2, 3.
In this issue, Kawasaki et al.4 hone our understanding of these central mechanisms. They show that two matrix metalloproteases, MMP2 and MMP9, mediate changes involved in pain hypersensitivity after nerve injury. MMPs can now be added to the list of extracellular molecules capable of amplifying and transforming primary afferent input to the CNS. The MMPs seem to have no role in acute pain, but blocking their activity dampens chronic pain in a mouse model of peripheral neuropathic pain.
MMPs are better known for their role in extracellular matrix remodeling and cleaving bioactive molecules5. But, recently, they have been implicated as key mediators of the cellular responses to inflammation and degeneration.
Using a battery of techniques, Kawasaki et al.4 show that MMP2 and MMP9 are responsible for two parallel but temporally, spatially and cellularly distinct mechanisms that occur after damage to a peripheral nerve—in this case, spinal nerve ligation (Fig. 1). Within 6 hours after ligation, MMP9 was upregulated, but this increase was only transient, declining to the pre-injury level by the third day. Several days after ligation, MMP2 activity increased and remained elevated for up to 3 weeks. Using selective pharmacological blockers, siRNA, and MMP2- or MMP9-null mice, the authors established that pain hypersensitivity behaviors depend on MMP9 during the first few days after injury and thereafter become dependent on MMP2 (ref. 4).
Figure 1: Pain players.
Kawasaki et al.4 reveal two parallel mechanisms involved in neuropathic pain behaviors in rodents, both mediated by extracellular MMPs. After damage to a peripheral nerve, there is an early but transient upregulation of MMP9 (red) in the injured dorsal root ganglion (DRG). The protease activity of the MMP9 induces IL-1
cleavage to its active form, which subsequently activates spinal microglia, known mediators of spinal changes involved in neuropathic pain2, 3. MMP9 activity declines after a few days, but this decline is concomitant with an increase in activity of MMP2 (blue), which also acts through IL-1. Inhibiting MMP9, on early days, or MMP2, on later days, suppresses pain hypersensitivity in the spinal nerve ligation model of neuropathic pain.
Katie Ris-Vicari
Full size image (78 KB)Immunocytochemical analysis revealed distinct and separate cellular locations for MMP9 and MMP2. In the dorsal root ganglion, MMP9, but not MMP2, was increased in the cell bodies of primary sensory neurons, and MMP2, but not MMP9, was increased in non-neuronal satellite cells. The cellular distinction continued in the spinal dorsal horn. Here, the action of MMP9 was mediated by activation of microglia, whereas MMP2 was itself upregulated in astrocytes. Furthermore, the signaling pathways within the dorsal horn glial cells were different. In the microglia, the p38 mitogen-activated protein kinase pathway was involved in MMP9 signaling, whereas in astrocytes the key signaling pathway for MMP2 involved ERK.
Despite all these differences between MMP9 and MMP2 activation, it seems that both MMPs require a common molecular player, the cytokine interleukin-1 (IL-1). The authors showed that both MMP9 and MMP2 cleaved IL-1 into its active form. Moreover, in line with previous findings6, IL-1–neutralizing antibody blocked the pain hypersensitivity caused by administering either MMP9 or MMP2. Pain hypersensitivity after peripheral nerve injury is known to be inhibited by blocking IL-1 (ref. 6). What constrains the apparent selectivity of MMP9 to microglia and MMP2 to astrocytes is not currently understood. Whether peripheral nerve injury involves matrix remodeling in the CNS, or whether the MMPs are acting in a more conventional signaling role, is also unclear.
MMPs, which now form a 23-member family in humans7, were originally described as effectors of tissue remodeling on the basis of their proteolytic actions on the extracellular matrix during the development of peripheral tissues. Subsequently, MMPs were found to be essential for cancer and peripheral inflammatory diseases such as rheumatoid arthritis. More recently, MMPs have been shown to be crucial in the CNS for inflammatory and immune disorders.
The finding that the two MMPs operate in two classes of glial cells within the CNS emphasizes the growing awareness of the vicious circle of interactions between neurons and glia in the processing of spinal cord pain and the generation of neuropathic pain. The role of spinal cord glial cells in pain hypersensitivity seems nearly undeniable8. But, ultimately, glial cells act through the enhanced and aberrant output of the nociceptive neurons in the dorsal horn that project to pain networks in the brain. Therefore, understanding the signaling between these neurons and the glial cells is crucial. Microglia-to-neuron signaling is mediated, in part, by release of brain-derived neurotrophic factor (BDNF) from dorsal horn microglia, leading to disinihibition of neurons in lamina I9, but the nature of astrocyte-to-neuron signaling in the dorsal horn remains elusive.
To inhibit MMP9 and MMP2, Kawasaki et al.4 used powerful experimental tools—the proteins TIMP1 and TIMP2, respectively. TIMP1 and TIMP2 are endogenous proteins, raising the possibility that they may be involved in MMP signaling after peripheral nerve injury. Finding out whether nerve injury affects the expression or activity of TIMP1 or TIMP2 would be a step toward addressing the larger question of how MMP9 and MMP2 are activated after nerve injury.
The findings of Kawasaki et al.4 have potentially fruitful implications for the clinic. Early interventions in the case of nerve damage, such as during surgery, could target MMP9 or TIMP1 activity, and treatment for already established neuropathic pain conditions could potentially be directed at the later-acting MMP2 or TIMP2. The mechanisms outlined by Kawasaki et al.4 provide multiple points for potential intervention, adding to the growing array of therapeutic targets for neuropathic pain.