- Top of page
- Materials and methods
Activated microglia are implicated in the pathogenesis of disease-, trauma- and toxicant-induced damage to the CNS, and strategies to modulate microglial activation are gaining impetus. A novel action of the tetracycline derivative minocycline is the ability to inhibit inflammation and free radical formation, factors that influence microglial activation. Minocycline is therefore being tested as a neuroprotective agent to alleviate CNS damage, although findings so far have yielded mixed results. Here, we showed that administration of a single low dose of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) or methamphetamine (METH), a paradigm that causes selective degeneration of striatal dopaminergic nerve terminals without affecting the cell body in substantia nigra, increased the expression of mRNAs encoding microglia-associated factors F4/80, interleukin (IL)-1α, IL-6, monocyte chemoattractant protein-1 (MCP-1, CCL2) and tumor necrosis factor (TNF)-α. Minocycline treatment attenuated MPTP- or METH-mediated microglial activation, but failed to afford neuroprotection. Lack of neuroprotection was shown to be due to the inability of minocycline to abolish the induction of TNF-α and its receptors, thereby failing to modulate TNF signaling. Thus, TNF-α appeared to be an obligatory component of dopaminergic neurotoxicity. To address this possibility, we examined the effects of MPTP or METH in mice lacking genes encoding IL-6, CCL2 or TNF receptor (TNFR)1/2. Deficiency of either IL-6 or CCL2 did not alter MPTP neurotoxicity. However, deficiency of both TNFRs protected against the dopaminergic neurotoxicity of MPTP. Taken together, our findings suggest that attenuation of microglial activation is insufficient to modulate neurotoxicity as transient activation of microglia may suffice to initiate neurodegeneration. These findings support the hypothesis that TNF-α may play a role in the selective vulnerability of the nigrostriatal pathway associated with dopaminergic neurotoxicity and perhaps Parkinson's disease.
Although predominantly viewed as scavenger cells, microglia also are known to initiate tissue repair and regeneration through secretion of growth and neurotrophic factors, thereby exerting a beneficial/neuroprotective role (Kreutzberg 1996; Gonzalez-Scarano and Baltuch 1999; Stoll and Jander 1999; Streit et al. 1999). On the other hand, activation of microglia may also initiate inflammation and exacerbate degeneration owing to release of cytotoxic products, such as reactive oxygen and nitrogen species, pro-inflammatory cytokines and proteases (Colton and Gilbert 1987; Boje and Arora 1992). Thus, microglial activation may exert neurotrophic and/or cytotoxic effects, observations indicative of a dual role for this cell type in response to neural injury. The conditions dictating whether microglia exhibit a beneficial or deleterious role are not completely understood. Nevertheless, certain features that seem to influence their role include the density, distribution and the morphological or functional state of these cells across various brain regions (Lawson et al. 1990; Ren et al. 1999).
Given the above considerations, interventions designed to modulate microglial responses could be viewed as a means of achieving neuroprotection against diverse injuries of the CNS. To address this possibility, several studies have attempted to use candidate drugs, including antioxidants and anti-inflammatory agents, to demonstrate that inhibition of microglial activation can afford protection against neuronal injuries (Bruce-Keller et al. 2000; Li et al. 2001; Asanuma et al. 2003; Liu et al. 2003; Yan et al. 2003; Zawadzka and Kaminska 2005). Among the compounds investigated is the antimicrobial agent minocycline, a broad-spectrum semisynthetic tetracycline derivative with anti-inflammatory properties that are distinct from its antimicrobial actions (Yrjanheikki et al. 1999; Wu et al. 2002).
Minocycline has been shown to afford protection against brain ischemia (Yrjanheikki et al. 1998, 1999), excitotoxicity (Tikka and Koistinaho 2001; Tikka et al. 2001), β-amyloid neurotoxicity (Ryu et al. 2004) and spinal cord injury (Stirling et al. 2004), and to delay disease onset in a murine model of amyotrophic lateral sclerosis (Kriz et al. 2002; Van Den Bosch et al. 2002). Similarly, it has been reported to be neuroprotective against dopaminergic neurotoxicity caused by 6-hydroxydopamine and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) (Du et al. 2001; He et al. 2001; Wu et al. 2002). The neuroprotection afforded by minocycline is thought to be associated with its ability to inhibit microglial activation, thereby reducing the levels of cytotoxic factors released by microglia. However, recent studies have demonstrated that minocycline treatment worsened or exacerbated brain injury (Yang et al. 2003; Diguet et al. 2004; Tsuji et al. 2004), indicating that minocycline may have actions that depend on the specific brain injury in question. The possible causes for such differential effects exhibited by minocycline remain unclear. In this study, we evaluated the effects of minocycline on the neurotoxicity of two dopaminergic neurotoxicants, MPTP and methamphetamine (METH). We show that, although minocycline treatment attenuates microglial activation, it fails to afford protection against dopaminergic neurotoxicity. Furthermore, we present evidence to demonstrate that this lack of neuroprotection results from the inability of minocycline to completely abolish the induction of tumor necrosis factor (TNF)-α and its receptors (TNFRs), thereby failing to modulate TNF signaling.
- Top of page
- Materials and methods
Microglia and astrocytes play a major role in brain inflammatory responses (Raivich et al. 1996; Ransohoff et al. 1996). Brain injury-related activation of microglia, in particular, is associated with enhanced expression of cytokines, chemokines and growth factors (De Bock et al. 1996; Botchkina et al. 1997; Sriram et al. 2002; Sriram and O'Callaghan 2005). Many of these factors have been implicated in the pathogenesis of neurodegenerative disorders such as PD (Boka et al. 1994; Mogi et al. 1996), Alzheimer's disease (Bauer et al. 1991), multiple sclerosis (Merrill 1992) and stroke (Sairanen et al. 2001). Pharmacological suppression of microglial activation therefore represents a potential intervention strategy to modulate or alleviate a variety of CNS injuries. In the present study, we demonstrated that suppression of multiple indices of microglial activation with the tetracycline antibiotic minocycline fails to protect against the striatal dopaminergic neurotoxicity caused by MPTP or METH. Data obtained using mice deficient in specific cytokines, chemokines or their receptors, combined with our previous observations on the role of TNF-α in striatal dopaminergic neurotoxicity, suggest that a complete blockade of the striatal effects of microglial-associated TNF-α may be required to achieve dopaminergic neuroprotection with minocycline or related interventions designed to inhibit microglial activation.
Although minocycline treatment significantly attenuated F4/80, a marker of microglial activation, it did not completely block MPTP- or METH-mediated induction of mRNAs encoding pro-inflammatory cytokines and chemokines such as TNF-α, IL-1α, IL-6 and CCL2. Thus, neurotoxicologically significant levels of these cytokines and chemokines were potentially available to initiate dopaminergic neurodegeneration. These findings suggest that even transient activation of microglia and/or elaboration of microglia-derived factors may be sufficient to elicit neurodegeneration in the striatum. To determine which of these microglia-derived factors play a key role in the neurodegenerative process, we examined the effects of MPTP in transgenic mice lacking IL-6, Ccl2 or Tnfr1/2 genes. MPTP neurotoxicity was not altered in mice lacking either the Il-6 or Ccl2 gene, but deficiency of both Tnfr genes protected against the dopaminergic neurotoxicity of MPTP. TNF-α is known to be associated with degenerating dopaminergic neurons in patients with PD (Boka et al. 1994) and our demonstration of its early induction in our MPTP-dosing model (Sriram et al. 2002) is suggestive of a participatory role for this cytokine in damage to the nigrostriatal dopaminergic pathway. Furthermore, TNFRs have been found to be expressed on dopaminergic neurons in PD (Boka et al. 1994). Their localization on the very cell type selectively vulnerable to damage in PD is consistent with the region-specific nature of the detrimental effects of TNF-α. Thus, consistent with our previous findings (Sriram et al. 2002), TNF-α appears to be a critical factor in striatal dopaminergic neurotoxicity. However, contrasting findings on the involvement of TNFRs in dopaminergic neurotoxicity have been reported. Rousselet et al. (2002) failed to observe neuroprotection of nigral dopaminergic neurons in TNFR-deficient mice, following a multiple-dose MPTP regimen. Similarly, Leng et al. (2005) reported absence of neuroprotection in mice lacking TNFRs, following chronic dosing with MPTP. An earlier report from the same group (Ferger et al. 2004), however, indicated that mice lacking the Tnf-α gene exhibited significant reductions in the loss of dopaminergic markers in the striatum, but not in the nigra. Thus, TNF-α appears to exhibit region-specific actions in the CNS. Indeed, we have recently observed such regional heterogeneity in the action of TNF-α in response to MPTP (K. Sriram, unpublished observations). Another important point to note is that neither of the above studies evaluated the effects of multiple or chronic doses of MPTP on glial cells. It is well known that high doses of MPTP can damage astrocytes (Di Monte et al. 1992; Wu et al. 1992), which might contribute to alterations in glial–neuronal signaling. Such an effect would modulate cellular responses of glial cells and alter the pattern of pro-inflammatory cytokine expression. The repertoire of pro-inflammatory cytokines expressed following different dosing paradigms may vary and the outcome of such varied responses could have very different or even opposing effects. We cannot, however, rule out the contribution of IL-1α, as we did not examine the role of this cytokine using genetic manipulations.
Our observations on the effect of minocycline on MPTP or METH neurotoxicity are at least partially inconsistent with those of earlier studies that reported neuroprotective effects of minocycline against the dopaminergic neurotoxicity caused by MPTP (Du et al. 2001; Wu et al. 2002). The disparity between the results of the present study and those reported previously is probably due to differences in the dosing regimen employed. The previous studies (Du et al. 2001; Wu et al. 2002) employed a multiple-dose MPTP regimen and varying doses of minocycline (ranging from 1.4 to 120 mg/kg) administered before, along with or after MPTP, to demonstrate neuroprotective effects on dopaminergic cell loss in the substantia nigra. Although the minocycline regimens used in the present study were similar, in that low- and high-dose paradigms were used before, during and after administration of MPTP or METH, we administered only a single low dose of MPTP or METH that does not cause dopaminergic cell body loss (O'Callaghan and Miller 1994; Sriram et al. 2004) or microglial and astroglial activation in the substantia nigra (shown in this study). Our intention was to evaluate very early changes associated with dopaminergic neurotoxicity in the striatum, the onset of which precedes the degeneration of dopaminergic cell bodies in the substantia nigra (Borit et al. 1975; Bradbury et al. 1986; Ichitani et al. 1991; Sauer and Oertel 1994; Calabresi et al. 2000). The single-dose MPTP or METH regimens used in this study have the advantage of making post-dosing time points easier to establish and analyze than multidose regimens. Finally, our single-dose regimens of MPTP or METH were designed to achieve partial denervation of dopaminergic nerve terminals, which decreases morphological and neurochemical markers of neurotoxicity in the striatum by approximately 50%. By doing so, we could readily evaluate the potential neuroprotective as well as neurodegenerative effects of a given intervention by the convenient evaluation of multiple markers of dopaminergic neurotoxicity and the associated microglial and astroglial responses.
Because our results clearly demonstrated dose-dependent albeit incomplete suppression of microglial markers by minocycline, without achieving neuroprotection against MPTP or METH neurotoxicity in the striatum, it seemed likely that other key factors distinguish the neurotoxic effects of the single- versus multiple-dosing models of these compounds. Prostanoids and nitric oxide represent two likely candidates because increased COX2 (Teismann et al. 2003; Feng et al. 2003) and iNOS (Itzhak et al. 1999; Liberatore et al. 1999; Dehmer et al. 2000) have been reported following a multiple-dosing regimen of MPTP or METH, effects that may not be associated with our single-dose model. Indeed, neither MPTP nor METH altered the striatal expression of Cox2, Nos1 or Nos2 (iNos) mRNAs. These data stand in contrast to those of other studies in which a multiple-dose MPTP regimen was shown to cause robust expression of iNOS in the substantia nigra linked to dopaminergic cell loss in this region. Furthermore, mice deficient in the iNos gene were more resistant to MPTP (Liberatore et al. 1999). However, in agreement with our findings, no induction of iNOS was observed in the striatum and iNOS deficiency did not protect against striatal dopaminergic neurotoxicity caused by MPTP (Liberatore et al. 1999; Dehmer et al. 2000). Thus, regional differences in the expression of microglia and their elaboration of cytotoxic factors may determine the susceptibility of brain regions to neurotoxicity, and such differences may serve as the basis for dopaminergic neurotoxicity models resulting in nigral cell loss versus those that damage striatal dopaminergic nerve terminals while sparing the cell bodies in the nigra. Indeed, we (K. Sriram, unpublished observations) and others (Lawson et al. 1990; Ren et al. 1999) have observed that the levels of microglial markers and microglia-derived factors are differentially expressed across brain regions. Therefore, factors produced by activated microglia that may be specific to various brain regions may influence the fate of the neurons and/or nerve terminals in that particular region, and the effects may vary depending on the dose or dosing paradigm followed.
In summary, attenuation of microglial activation is insufficient to protect against striatal dopaminergic neurotoxicity. We attribute the failure of minocycline to afford neuroprotection against MPTP- and METH-induced dopaminergic neurotoxicity in striatum to its inability to completely suppress the induction of TNF-α and its receptors, leaving signaling via TNFRs at least partially active. As TNF-α is an obligatory component of dopaminergic neurotoxicity (Sriram et al. 2002), the failure to abolish TNF signaling reflects the lack of efficacy of minocycline in completely abolishing microglial activation. On the other hand, deficiency of TNFRs afforded neuroprotection against striatal dopaminergic neurotoxicity, suggesting that TNF-α signaling must be completely abolished to achieve protection against dopaminergic neurotoxicity in the striatum. The present observations support our earlier hypothesis that TNF-α is a key player in striatal dopaminergic neurotoxicity. Finally, our findings also strengthen the notion that TNF-α may be responsible for the selective vulnerability of the nigrostriatal pathway to degenerative changes associated with dopaminergic neurotoxicity and perhaps PD.