Toll-like receptors in the CNS: implications for neurodegeneration and repair

J. Verhaagen et al. (Eds.) Progress in Brain Research, Vol. 175 ISSN 0079-6123 Copyright r 2009 Elsevier B.V. All rights reserved

CHAPTER 9

Toll-like receptors in the CNS: implications for neurodegeneration and repair Johannes M. van Noort and Malika Bsibsi Delta Crystallon BV, Leiden, The Netherlands

Abstract: The role of Toll-like receptors (TLRs) in the CNS is only starting to be uncovered. As in peripheral organs, multiple TLRs are dynamically expressed. They are involved in mounting a host-defense response against microbial invasion of the CNS. The many different TLRs expressed on microglia are likely the most important first line of defense in this respect. Intriguingly, microglial TLR tend to trigger a very standard cytokine and chemokine response, irrespective of the type of TLR agonist they meet. The main purpose of this standardized response by microglia may be to recruit the assistance by other cells rather than to immediately mount a destructive response toward invaders. As is generally the case for microglial responses, TLR-mediated responses can also work out in either beneficial or detrimental ways, depending on the strength and timing of the activating signal. Yet, the role of TLRs in the CNS extends well beyond controlling host-defense responses alone. Other cells in the CNS, including astrocytes, neurons, and oligodendrocytes, can also express multiple functional TLRs upon activation. These play important roles in tissue development, cellular migration, and differentiation; in limiting inflammation; and in mounting repair processes following trauma. The TLRmediated reactions of these other neural cells to TLR agonists is highly cell specific and does not necessarily resemble that of microglia at all. It appears likely that endogenous agonists for TLRs are particularly relevant to activate these endogenous TLR functions on neural cells, also during development when microbial invaders have not yet entered the stage. In this chapter, current data are reviewed to highlight the emerging variety of functional roles of TLRs in the CNS. Keywords: toll-like receptors; microglia; astrocytes; neurons; cytokines cellular responses to a large variety of structures, often of microbial origin. Upon activation by typically microbial structures including lipopolysaccharides, teichoic acids, peptidoglycans, and various forms of nucleic acids, TLR family members trigger innate inflammatory responses that are usually dominated by the NF-kBmediated production of cytokines such as TNF-a and IL-1b, and chemokines including IL-8 and CCL5. It is no surprise, therefore, that the

Introduction and scope of this chapter TLRs comprise a family of at least 11 conserved pattern-recognition receptors that mediate diverse

Corresponding author.

Tel.: +31 71 518 1541; Fax: +31 71 518 1901; E-mail: [email protected] DOI: 10.1016/S0079-6123(09)17509-X

139

140

TLR-mediated cellular response — also in the CNS — is often understood to be a pro-inflammatory host-defense response, primarily designed to eliminate invading pathogens. Yet, TLRs activate many more types of cellular responses than only host-defense responses. TLRs also appear on neural cells under conditions where no foreign invaders are obvious, for example, in the CNS of people with Alzheimer’s disease (Walter et al., 2007) or multiple sclerosis (Bsibsi et al., 2002). Evidence is accumulating that especially in response to endogenous molecular ligands, TLRs also play a role in tissue development, cellular migration and differentiation, and repair processes. Especially these more subtle and beneficial roles of TLR functions may well be very relevant to the CNS. After all, even in neural cell types including neurons and oligodendrocytes, cells that do not play an active role in destructive inflammatory responses, functional TLRs, are indeed expressed. Even on astrocytes, TLRs can activate production of a wide range of neuroprotective and antiinflammatory mediators rather than merely stimulating pro-inflammatory factors. A one-dimensional view of the role of TLRs in the CNS would therefore be underestimating their clearly multifaceted functional roles. In this chapter, current evidence for the expression and functional roles of TLRs on major cell types in the CNS will be reviewed. It will be separately discussed what is currently known about the role of TLRs in microglia, astrocytes, and neurons. This will illustrate how variable TLR-mediated responses can be depending on the cell type. It must be kept in mind, however, that research on the role of TLRs in the CNS has only just begun, and many questions are still unanswered. Oligodendrocytes, for example, are known to express TLRs at least in cell culture (Bsibsi et al., 2002) but functional data on oligodendroglial TLRs remain lacking. Yet, the evidence that has become available over the past years begins to paint an exciting picture of TLRs as a dynamically expressed family of receptors that play important roles not only in defending the CNS against pathogens but also in shaping the CNS itself during development, and repairing it after damage.

General features of TLRs The family of TLRs share structural properties not only in the extracellular leucine-rich repeat structures designed to register the presence of multiple ligands but also in their intracellular domains which interact with intracellular adaptor proteins that relay the agonist engagement signal. Currently, five of such adaptors are known. The dominant and founding member of the family of these adaptors is MyD88 (or MyD88-1), which relays the signal for most TLR family members and tends to predominantly induce NF-kBmediated activation of genes, including those encoding TNF-a, CCL5 (RANTES), IL-1b, and CXCL8. This dominant pathway has been particularly well characterized in myelomonocytic cells, which explains why TLR-mediated responses are generally portrayed as pro-inflammatory. Yet, signaling by TLR is much more complex than this. The intracellular domain of TLR3, for example, is slightly different from that of the others and unable to interact with MyD88-1. It can only bind to a variant member of the adaptor family called MyD88-3, also referred to as TRIF/TICAM-1. Activation of this signaling route by TLR3 leads to the dominant induction of interferon response factors instead of NF-kB and subsequent activation of genes encoding type-I interferons including interferon-a and -b. Also TLR4 can interact with MyD88-3, but TLR3 is particularly dominant in doing so. The way TLRs interact with each of the five different adaptors is not just controlled by structural preferences embedded in the structure of the intracellular domain of TLRs, but also by the mere availability of adaptors. Not every cell expresses the same set. For example, it is rapidly becoming clear that selective expression of the less frequently used adaptor MyD88-5 in neurons renders these cells uniquely sensitive to TLR-mediated activation of the JNK pathway to apoptosis, instead of the NF-kB pathway to inflammatory mediators. In this way, selective expression of adaptors strongly influences the quality of the response mounted by different types of cells to a given TLR agonist. Very often, activation of TLR-mediated signaling by various agonists does not involve a

141

straightforward key-and-lock mode of ligand– receptor binding. That extracellular domains of all TLRs share important structural features, yet mediate responses to widely different agonists, already clarifies that more complex interactions are involved. Recognition of double-stranded RNA by TLR3 is relatively simple, involving two separate TLR3 extracellular domains acting in concert to bind one double-stranded RNA helix. In general, however, many different additional proteins are required for the activation of TLRmediated signaling by their agonists, including coreceptors and docking molecules on the cell surface and binding catalysts, such as CD14 and heat shock proteins, that promote certain interactions (Bsibsi et al., 2007; Jou et al., 2006; Lehnardt et al., 2008). This interplay with several additional proteins is probably crucial in allowing structurally conserved TLRs to respond to myriad different agonists that share no obvious structural similarity. In addition to the essential contribution by co-receptors and accessory interaction partners, most TLRs also operate as homo- or heterodimers. In short, activation of TLR-mediated signaling is far more complex than simply having one agonist molecule bound to the extracellular domain of one TLR. For many of the currently known TLR-signaling scenarios, different partners have already been identified. For many other scenarios, details remain to be clarified.

Expression and function of TLRs in microglia Much like other macrophage-like cells, microglia can express essentially all different TLR family members. While TLR expression is hardly detectable in resting microglia in a healthy CNS, multiple TLRs rapidly appear upon activation of the cells. As in macrophages, TLRs are exclusively found within endosomal vesicles of microglia, illustrating their prime role in probing anything taken up by phagocytosis. While microglia express all TLRs at readily detectable levels (Bsibsi et al., 2002), TLRs 1–4 are the most dominant ones and, in particular, TLR2 is induced at somewhat higher levels than other family

members; this appears to apply to microglia in rodents as well as in humans. There is ample evidence to indicate that microglial TLRs are crucial as a first line of defense against bacterial or viral infection. In response to the appearance of multiple bacterial or viral TLR agonists, TLR-mediated signaling promotes production of a variety of inflammatory mediators (reviewed by Konat et al., 2006; Block et al., 2007). Also, phagocytosis is stimulated by TLR activation, which may be particularly relevant to clearance of bacteria, as well as of aggregated or abnormal proteins such as amyloid fibers from the CNS (Chen et al., 2006; Tahara et al., 2006). As in most cases of microglial activation, a balance exists between a mild TLRmediated response by microglia, which is beneficial in combating infection, clearing debris, and promoting repair, and an exaggerated or persistent response that contributes to tissue damage and neurodegenerative processes. Thus, microglial activation via TLRs has been found to exert beneficial effects as well as deleterious effects depending on the context and strength of the challenge applied. In studies of experimental cerebral ischemia, for example, such apparently contradictory effects have been clearly highlighted. Upon induction of experimental ischemia in mice, both TLR2 and 4 can markedly contribute to tissue damage as mice deficient for either of these TLRs develop much less brain damage and display increased neuronal survival (Lehnardt et al., 2007; Ziegler et al., 2007; Kilic et al., 2008; Caso et al., 2008). Yet, exposure of experimental animals to agonists for TLR2, 4, or 9 prior to induction of ischemia surprisingly increases their resistance to injury caused by cerebral ischemia and/or reperfusion (Hua et al., 2008; Stevens et al., 2008). Clearly, conditioning of the CNS by prior TLR activation limits the detrimental effects of subsequent inflammatory challenge. This effect is also well known in other situations where macrophage activation is key to disease development, and relies on TLR-induced downregulation of many different genes including those controlling TLR functions themselves (Mages et al., 2007). Thus, different timing of TLR activation relative to

142

induction of ischemia can lead to radically different outcomes. The TLR-mediated response after experimental nerve damage tends to work out more beneficially in most cases. Absence of either TLR2- or TLR4-mediated signaling after peripheral nerve injury results not only in delayed recruitment of macrophages and clearance of cellular debris but also in delayed axonal regeneration and locomotor recovery (Boivin et al., 2007). Conversely, activation of TLR signaling in this model promotes these regenerative effects. Also in the CNS these post-trauma effects of at least TLR2 and 4 are apparent. Their activation is beneficial after spinal cord injury (Kigerl et al., 2007), or surfactant-mediated experimental cytolysis (Glezer et al., 2006). Under such experimental conditions where pathogens are not in play, both TLRs play important roles in regulating gliosis and tissue repair processes and promote mobilization of oligodendrocyte progenitors. The full range of such microglia-mediated effects may still not be uncovered. Apart from controlling microbial invasion and neurodegeneration in response to various types of cellular injury including Wallerian degeneration (Lehnardt et al., 2008; Boivin et al., 2007), microglial TLR are also important for control of painful neuropathies (Tanga et al., 2005; Kim et al., 2007a, b), and for migration and differentiation of neural precursors (Aarum et al., 2003). Intuitively, one might perhaps assume that the response by microglia is diverse and tailored to meet the unique defensive demands to counteract each structural challenge. Indeed, it has been suggested that viral agonists would trigger different types of microglial responses from bacterial stimuli and different responses would again be observed in responses to endogenous TLR agonists (Jack et al., 2005). Yet, other data indicate that differential responsiveness may very well not be a dominant trait of the microglial response. At least in culture, human microglia are not very sensitive at all to the type of stimulus they meet, and just respond to the mere fact that there is a stimulus. As illustrated in Fig. 1, activation of human microglia by agonists for either TLRs 3, 4, 8, or 9 produces hardly any differences in the

patterns of cytokines and chemokines that are induced. In all of these four cases examined in detail, IL-13, TNF-a, CCL5 (RANTES), CXCL10 (IP-10), IL-12p40, IL-15, IL-6 sR, TGF-b, and IL-10 are induced at rather similar levels with only few exceptions. In no case there is a unique mediator induced by a single TLR agonist, which is also not induced by other agonists. Clearly, these first results should be extended to other response pathways as well. Yet, they do suggest that the prime role of microglia is not to combat a variety of microbial threats by sophisticated and differentiated response profiles tailored to fight off different challenges in different ways. Instead, microglia tend to ‘‘raise the red flag’’ and appear to simply mount a single common response to structurally different types of insults, for other mechanisms to take over the actual tailored host-defense role. Given the prominent presence of both TNF-a (as a prime activator of blood–brain barrier endothelial cells) and several chemokines that will help attract peripheral blood leukocytes (including CCL2, CCL5, and CXCL10) in the set of TLR-induced mediators, microglia appear to be particularly inclined to recruit the assistance by the peripheral immune repertoire to help resolve inflammatory insults or microbial invasion. On the other hand, many cells in the CNS including glial cells themselves also express receptors for the various chemokines secreted by activated microglia in response to TLR agonists. Therefore, TLRmediated activation of microglia is also likely to trigger a wave of migratory activity by glial cells themselves to the site of insult or injury. Microglia do not seem to be particularly well equipped to solve any insult themselves, but rather call upon other cells to do this.

Expression and function of TLRs in astrocytes Astrocytes from both human and rodent origin express TLRs1–4 in readily detectable amounts, but much lower levels of the other TLRs. Yet, the TLR profile of astrocytes is not a fixed quality, but a highly dynamic feature. In a healthy human CNS, TLR expression on astrocytes is very

143

second stimulus

isolation and culture

16

MICROGLIA

TLR3 IL-6 CXCL8 CCL2 CCL4 CCL8 CCL3

12 8 4

TLR4

0 16 12 8 4

TLR8

0 16 12 8 4

TLR9

0 16 12 8 4

IL-10

β TGF-β

IL-6sR

IL-15

IL-12p40

CXCL10

CCL5

TNF-α

IL-13

0

Fig. 1. Production of a standard set of cytokines and chemokines by cultured human microglia in response to different TLR agonists. When human microglia are isolated and cultured in the presence of GM-CSF, marked production of several chemokines occurs, as can be monitored by antibody arrays that allow simultaneous evaluation of 40 different soluble mediators. Following subsequent activation with selective TLR agonists, strikingly similar patterns of cytokines and chemokines are induced in all cases.

difficult to detect by immunohistochemistry, but when inflammation develops, TLRs readily emerge on the cell surface of astrocytes (Bsibsi et al., 2002). The data that have accumulated on the expression of TLRs on astrocytes of either rodent or human origin have consistently revealed a striking preference of astrocytes to express high levels of TLR3 (Jack et al., 2005; Farina et al., 2005; Bsibsi et al., 2006). While other TLR family members, notably including TLR2 and 4, are clearly detectable on astrocytes both in cell culture models and in vivo during trauma or inflammation, especially activated astrocytes produce much more TLR3 than any other TLR.

This intriguing preference of astrocytes to express up to 200-fold elevated levels of TLR3 upon activation is somewhat puzzling since the only currently known agonist for TLR3 is doublestranded RNA, which is believed to emerge as an intermediate during viral replication. Yet, doublestranded RNA is generally inside cells rather than secreted into the microenvironment, and its detectable presence in the microenvironment of CNS cells is rare. It is therefore difficult to envisage that the consistently dominant expression of TLR3 on the surface of astrocytes under a wide variety of conditions is designed solely to detect the very infrequent presence of extracellular

144

double-stranded RNA. The quality of the astroglial response to TLR3-mediated signaling only adds to this apparent confusion since it is primarily neuroprotective rather than pro-inflammatory. As dissected using microarray-based transcript analysis, the TLR3-mediated response in human astrocytes is far more comprehensive than the TLR4-mediated response. Also, it is striking that in response to TLR3-mediated activation astrocytes produce a variety of factors that are either well-known mediators of neuroprotection, or antiinflammatory mediators, as is illustrated in Fig. 2. Examples of the first group include cilliary neurotrophic factor, neurotrophin-4, leukemia inhibitory factor, vascular endothelial growth factor, embryonic growth/differentiation factor, neuregulin,

neurite growth-promoting factor (pleiotrophin), brain-derived neurotrophic factor, glial growth factors 1 and 2, and erythroid differentiation protein. An impressive number of studies have highlighted the neuroprotective qualities of all of these mediators. Several have been or still are under scrutiny as candidate therapeutic agents. Examples of the second group include tumor necrosis factor-inducible protein 6, TGF-b, IL-9, IL-10, and IL-11. Together, the mediators that are all selectively induced at least two-fold by TLR3 do not appear to represent a traditional proinflammatory host-defense response. Indeed, when poly I:C as an activator of such TLR3mediated responses in astrocytes is added to organotypic human brain slice cultures, survival

ASTROCYTE

TLR3

TSG-6 CNTF NT-4 LIF GM-CSF VEGF GDF-1 GGF-1/2 BDNF TGF-β2 FGF-receptor Ephrin type B-receptor

CXCL8 CXCL2 CCL3 IL-9 CCL-4 TNF-α CXCL6 IL-11 IL-6 CCL5 IL-10 CCL2

TLR4

TSG-6 CNTF HGF activator TNF-α

CXCL8 CXCL2 CCL3 CXCL 9

Fig. 2. Production of cytokines, chemokines, and growth factors by human astrocytes activated either via TLR3 or 4. Cultured human astrocytes were stimulated either with poly I:C or with ultrapure LPS as agonists for TLR3 and 4, respectively. After 1–2 days, gene transcript levels were evaluated by microarrays for several hundred genes encoding cytokines, chemokines, growth factors, and their receptors. While the TLR3-triggered response includes marked induction of a wide variety of anti-inflammatory or neuroprotective mediators, this is much less the case after TLR4-mediated activation (for details, see Bsibsi et al., 2006).

145

of neurons in such cultures significantly improves. When LPS is added, no such improvement is observed (Bsibsi et al., 2006). These data strongly suggest that TLR3 on astrocytes may be an inducible regulator to activate tissue repair responses and limit inflammatory processes that develop in response to stress or trauma. As many different types of activators including pro-inflammatory cytokines and oxidative stress and TLR agonists themselves selectively induce TLR3 on astrocytes, the subsequent TLR3-mediated response is likely to represent a common secondary repair response designed to protect the CNS microenvironment after a first wave of host-defense activity. The dynamics of the glial response to stimuli, at least in cell culture models, is in line with such a view. Astrocytes take much longer than microglia to either upregulate TLRs or produce cytokines and growth factors in response to TLR activation. Also, TLR3-mediated activation of astrocytes leads to strong induction of indoleamine 2,3-dioxygenase (Suh et al., 2007). This enzyme converts extracellular tryptophan into kynurenine, thereby reducing its concentrations in the microenvironment, which in turn markedly enhances the sensitivity of any nearby T cell for Fas-ligand-induced apoptosis (Kwidzinksi et al., 2003). In this way, the TLR3-mediated induction of indoleamine 2,3-dioxygenase in astrocytes contributes to the elimination of any activated T cell or, in other words, acts as a local immune-suppressive factor. It appears unlikely that the potent antiinflammatory and tissue repair responses that are apparently mediated by TLR3 would be uniquely dependent on an exogenous microbial signal like double-stranded viral RNA. Instead, a still unknown endogenous agonist for astroglial TLR3 to control this pathway in the absence of infection appears to be much more likely.

Expression and function of TLRs in neurons After a first report in 2005 on the expression of TLR3 in cultured human neurons following viral infection (Prehaud et al., 2005), the expression of TLR3 on neurons in human brain tissue samples

was confirmed in cases of rabies or herpes simplex virus infection (Jackson et al., 2006). Indeed, convincing evidence has now accumulated that neurons can express different functional TLRs, including TLR2, 3, 4, and 8 (Lafon et al., 2006; Ma et al., 2006; Tang et al., 2007; Cameron et al., 2007; Kim et al., 2007a, b; Acosta and Davies, 2008). As in other cells, levels of expression are dynamic, and influenced by soluble mediators including interferon-g, or by energy deprivation. A very interesting aspect of TLR-mediated signaling in neurons appears to be the role of a peculiar member of the MyD88 family, viz., MyD88-5 (previously known under the acronym SARM for the inconvenient ‘‘sterile alpha and HEAT/Armadillo motifs containing protein’’) which is preferentially expressed by neurons (Kim et al., 2007a, b). Most TLRs except TLR3 signal via the founding family member of the MyD88 family, which predominantly activates NF-kBmediated responses. The neuronal MyD88-5, on the other hand, associates with mitochondria, microtubules, and JNK3, and regulates neuronal death during deprivation of oxygen and glucose. Preferred expression of MyD88-5 in neurons confers a different quality of TLR responsiveness to these cells as compared to cells such as glial cells that do not express MyD88-5, but utilize other adaptors to relay TLR-mediated signaling. As a consequence, TLR3, which is concentrated in the growth cones of neurons, triggers growth cone collapse (Cameron et al., 2007), TLR2 and 4 induce apoptotic death (Tang et al., 2007), and TLR8 inhibits neurite outgrowth and triggers apoptosis (Ma et al., 2006). In all these cases, signaling pathways operate independently from NK-kB. Clearly, by introducing MyD88-5 as the dominant adaptor for TLR-mediated intracellular signaling, neurons turn most TLR-mediated signals into negative signals for growth, development, and even survival. Yet, additional functions for neuronal TLR remain. For example, engagement of TLR4 on neurons induces the expression of nociceptin, an opioid-related neuropeptide (Acosta and Davies, 2008). Again, however, this response is somewhat different from TLR4mediated responses in many other cells in that neurons apparently use the co-receptor MD-1

146

instead of the routinely used MD-2 along with CD14 as interaction partners for binding the TLR4 agonist LPS. This illustrates that neurons apparently modulate the TLR-signaling platform not only by introducing unusual intracellular adaptors to link TLRs to unique signaling pathways, but also by employing uncommon surface co-receptors which modulate the response. Concluding remarks and future perspectives As in other parts of the body TLRs are crucial to fight off microbial invasion also in the CNS. Yet, evidence is rapidly accumulating that TLRs fulfill functions well beyond orchestrating host-defense responses alone. TLRs are dynamically expressed on every type of neural cell, and influence inflammation and repair, also under sterile conditions (Bsibsi et al., 2002; Prinz et al., 2006; Tahara et al., 2006; Walter et al., 2007). In addition, they play a role in differentiation, migration, and development of neural cells. This suggests that a number of endogenous agonists must be available to allow for TLR-mediated processes to become activated also in the absence of microbial invaders. A few of such endogenous agonists are already known, including soluble CD14 that acts as an endogenous agonist for TLR2 (Bsibsi et al., 2007) and facilitates TLR4 activation by various ligands including brain gangliosides (Jou et al., 2006). Also heat shock proteins such as HSP60 that can be released by stressed or injured neural cells can promote TLR4-mediated responses in the absence of any microbial challenge (Lehnardt et al., 2008). Many other endogenous TLR agonists probably still await uncovering. When comparing expression and function of TLRs on different types of neural cells, striking differences emerge which emphasize that each type of cell has a different relationship with TLRs, and is likely to respond differently to any given TLR agonist. While microglia, for example, express many different TLRs and respond to their activation with a surprisingly standardized ‘‘danger’’ response, astrocytes clearly display differential responses to different TLR agonists. Neurons relay TLRmediated signaling to completely different signaling

pathways from other cells, by linking the TLR machinery to a unique intracellular adaptor. Along with effects of location, timing, and strength of TLR-mediated responses, and the fact that in most cases multiple TLRs will be activated at the same time, these aspects of TLR functioning in different cell types pose formidable challenges to the development of a coherent paradigm for TLR functions in the CNS as a whole. Both cell culture models and in vivo experiments will be required for a full clarification of these functions. A more profound understanding of TLR functions in the CNS, however, is very likely to make a crucial contribution to our understanding of the CNS itself, including issues of development and repair of brain functions, well beyond problems of infection and host-defense responses alone.

Abbreviations MyD88 TLR TRIF

myeloid differentiation primary response gene 88 toll-like receptor toll IL-1 receptor-domain-containing adapter-inducing interferon-b

Acknowledgements Our studies on TLRs in the CNS have been financially supported by the Foundation for the support of MS Research in The Netherlands. Human brain tissue samples used were kindly provided by the Netherlands Brain Bank, Amsterdam, The Netherlands.

References Aarum, J., Sandberg, K., Budd Haeberlein, S. L., & Persson, M. A. A. (2003). Migration and differentiation of neuronal precursor cells can be directed by microglia. Proceedings of the National Academy of Sciences of the United States of America, 100, 15983–15988. Acosta, C., & Davies, A. (2008). Bacterial lipopolysaccharide regulates nociceptin expression in sensory neurons. Journal of Neuroscience Research, 86, 1077–1086.

147 Block, M. L., Zecca, L., & Hong, J. S. (2007). Microgliamediated neurotoxicity: uncovering the molecular mechanisms. Nature Reviews Neuroscience, 8, 57–69. Boivin, A., Pineau, I., Barrette, B., Filali, M., Vallie`res, N., Rivest, S., et al. (2007). Toll-like receptor signaling is critical for Wallerian degeneration and functional recovery after peripheral nerve injury. The Journal of Neuroscience, 27, 12565–12576. Bsibsi, M., Bajramovic, J. J., Van Duijvenvoorden, E., Persoon, C., Ravid, R., van Noort, J. M., et al. (2007). Identification of soluble CD14 as an endogenous agonist for Toll-like receptor 2 on human astrocytes by genome-scale functional screening of glial cell derived proteins. Glia, 55, 473–482. Bsibsi, M., Persoon-Deen, C., Verwer, R. W. H., Meeuwsen, S., Ravid, R., & van Noort, J. M. (2006). Toll-like receptor 3 on adult human astrocytes triggers production of neuroprotective mediators. Glia, 53, 688–695. Bsibsi, M., Ravid, R., Gveric, D., & van Noort, J. M. (2002). Broad expression of Toll-like receptors in the human central nervous system. Journal of Neuropathology and Experimental Neurology, 61, 1013–1021. Cameron, J. S., Alexopoulou, L., Sloane, J. A., DiBernardo, A. B., Ma, Y., Kosaras, B., et al. (2007). Toll-like receptor 3 is a potent negative regulator of axonal growth in mammals. The Journal of Neuroscience, 27, 13033–13041. Caso, J. R., Pradillo, J. M., Hurtado, O., Leza, J. C., Moro, M. A., & Lizasoain, I. (2008). Toll-like receptor 4 is involved in subacute stress-induced neuroinflammation and in the worsening of experimental stroke. Stroke, 39, 1314–1320. Chen, K., Iribarren, P., Hu, J., Chen, J., Gong, W., Cho, E. H., et al. (2006). Activation of Toll-like receptor 2 on microglia promotes cell uptake of Alzheimer Disease-associated amyloid b peptide. The Journal of Biological Chemistry, 281, 3651–3659. Farina, C., Krumbholz, M., Giese, T., Hartmann, G., Aloisi, F., & Meinl, E. (2005). Preferential expression and function of Toll-like receptor 3 in human astrocytes. Journal of Neuroimmunology, 159, 12–19. Glezer, I., Lapointe, A., & Rivest, S. (2006). Innate immunity triggers oligodendrocyte progenitors reactivity and confines damages to brain injuries. The FASEB Journal, 20, 750–752. Hua, F., Ma, J., Ha, T., Kelley, J., William, D. L., Kao, R. L., et al. (2008). Preconditioning with a TLR2-specific ligand increases resistance to cerebral ischemia/reperfusion injury. Journal of Neuroimmunology, 199, 75–82. Jack, C. S., Arbour, N., Manusow, J., Montgrain, V., Blain, M., McCrea, E., et al. (2005). TLR signaling tailors innate immune responses in human microglia and astrocytes. Journal of Immunology, 175, 4320–4330. Jackson, A. C., Rossiter, J. P., & Lafon, M. (2006). Expression of Toll-like receptor 3 in the human cerebellar cortex in rabies, herpes simplex encephalitis, and other neurological diseases. Journal of Neurovirology, 12, 229–234. Jou, I., Lee, J. H., Park, S. Y., Yoon, H. J., Joe, E.-H., & Park, E. J. (2006). Gangliosides trigger inflammatory responses via

TLR4 in brain glia. The American Journal of Pathology, 168, 1619–1630. Kigerl, K. A., Lai, W., Rivest, S., Hart, R. P., Satoskar, A. R., & Popovich, P. G. (2007). Toll-like receptor (TLR)-2 and TLR4 regulate inflammation, gliosis, and myelin sparing after spinal cord injury. Journal of Neurochemistry, 102, 37–50. Kilic, U., Kilic, E., Matter, C. M., Bassetii, C. L., & Hermann, D. M. (2008). TLR4-deficiency protects against focal cerebral ischemiia and axotomy-induced neurodegeneration. Neurobiology of Disease, 31, 33–40. Kim, D., Kim, M. A., Cho, I.-H., Kim, M. S., Lee, S., Jo, E.-K., et al. (2007a). A critical role of Toll-like receptor 2 in nerve injury-induced spinal cord glial cell activation and pain hypersensitivity. The Journal of Biological Chemistry, 282, 14975–14983. Kim, Y., Zhou, P., Qian, L., Chuang, J.-Z., Lee, J., Li, C., et al. (2007b). MyD88-5 links mitochondria, microtubules, and JNK3 in neurons and regulates neuronal survival. The Journal of Experimental Medicine, 204, 2063–2074. Konat, G. W., Kielian, T., & Marriott, I. (2006). The role of Toll-like receptors in CNS responses to microbial challenge. Journal of Neurochemistry, 99, 1–12. Kwidzinksi, E., Bunse, J., Kovac, A. D., Ulrich, O., Zipp, F., Nitsch, R., et al. (2003). IDO (indolamine 2,3-dioxygenase) expression and function in the CNS. Advances in Experimental Medicine and Biology, 527, 113–118. Lafon, M., Megret, F., Lafage, M., & Prehaud, C. (2006). The innate immune facet of brain: human neurons express TLR-3 and sense viral dsRNA. Journal of Molecular Neuroscience, 29, 185–194. Lehnardt, S., Lehmann, S., Kaul, D., Tschimmel, K., Hoffmann, O., Cho, S., et al. (2007). Toll-like receptor 2 mediates CNS injury in focal cerebral ischemia. Journal of Neuroimmunology, 190, 28–33. Lehnardt, S., Schott, E., Trimbuch, T., Laubisch, D., Kreueger, C., Wulcyn, G., et al. (2008). A vicious cycle involving release of heat shock protein 60 from injured cells and activation of Toll-like receptor 4 mediates neurodegeneration in the CNS. The Journal of Neuroscience, 28, 2320–2331. Ma, Y., Li, J., Chiu, I., Wang, Y., Sloane, J. A., Lu¨, J., et al. (2006). Toll-like receptor 8 functions as a negative regulator of neurite outgrowth and inducer of neuronal apoptosis. The Journal of Cell Biology, 175, 209–215. Mages, J., Dietrich, H., & Lang, R. (2007). A genome-wide analysis of LPS tolerance in macrophages. Immunobiology, 212, 723–737. Prehaud, C., Megret, F., Lafage, M., & Lafon, M. (2005). Virus infection switches TLR-3-positive human neurons to become strong producers of interferon beta. Journal of Virology, 79, 12893–12904. Prinz, M., Garbe, F., Schmidt, H., Mildner, A., Gutcher, I., Wolter, K., et al. (2006). Innate immunity mediated by TLR9 modulates pathogenicity in an animal model of multiple sclerosis. The Journal of Clinical Investigation, 116, 456–464.

148 Stevens, S. L., Ciesielki, T. M., Marsh, B. J., Yang, T., Homen, D. S., Boule, J. L., et al. (2008). Toll-like receptor 9: a new target of ischemic preconditioning in the brain. Journal of Cerebral Blood Flow and Metabolism, 28, 1040–1047. Suh, H.-S., Zhao, M.-L., Rivieccio, M., Choi, S., Connoly, E., Zhao, Y., et al. (2007). Astrocyte indoleamine 2,3-disoxygenase is induced by the TLR3 ligand poly (I:C): mechanism of induction and role in antiviral response. Journal of Virology, 81, 9838–9850. Tahara, K., Kim, H. D., Jin, J. J., Maxwell, J. A., Li, L., & Fukuchi, K. (2006). Role of Toll-like receptor signaling in Abeta uptake and clearance. Brain, 129, 3006–3019. Tang, S.-C., Arumugam, T. V., Xu, X., Cheng, A., Mughal, M. R., Jo, D. G., et al. (2007). Pivotal role for neuronal Toll-like receptors in ischemic brain injury and functional deficits.

Proceedings of the National Academy of Sciences of the United States of America, 104, 13798–13803. Tanga, F. Y., Nutile-McMenemy, N., & DeLeo, J. A. (2005). The CNS role of Toll-like receptor 4 in innate neuroimmunity and painful neuropathy. Proceedings of the National Academy of Sciences of the United States of America, 102, 5856–5861. Walter, S., Letiembre, M., Liu, Y., Heine, H., Penke, B., Hao, W., et al. (2007). Role of Toll-like receptor 4 in neuroinflammation in Alzheimer’s disease. Cellular Physiology and Biochemistry, 20, 947–956. Ziegler, G., Harhausen, D., Schepers, C., Hoffmann, O., Ro¨hr, C., Prinz, V., et al. (2007). TLR2 has a detrimental role in mouse transient focal cerebral ischemia. Biochemical and Biophysical Research Communications, 359, 574–579.