The Journal of Neuroscience, June 25, 2008 • 28(26):6537–6538 • 6537

Journal Club

Editor’s Note: These short, critical reviews of recent papers in the Journal, written exclusively by graduate students or postdoctoral fellows, are intended to summarize the important findings of the paper and provide additional insight and commentary. For more information on the format and purpose of the Journal Club, please see http://www.jneurosci.org/misc/ifa_features.shtml.

The Utility of a New In Vitro Model of the Penumbra

Trisha A. Dwyer,1 Damien E. Earl,2 and Liping Wang1 Departments of 1Neurology and 2Physiology and Pharmacology, Graduate Program in Neurosciences and Neurological Disorders, University of Toledo College of Medicine, Toledo, Ohio 43614 Review of Kaushal and Schlichter (http://www.jneurosci.org/cgi/content/full/28/9/2221)

Ischemic stroke in the brain is produced by tion and the effects of on the (TNF-␣) and NO, which are downstream arterial occlusion, resulting in dramatically stroke penumbra are inconclusive. products of NF-␬B and p38 MAPK signal- reduced levels of oxygen and glucose to the Highlighting the importance of micro- ing cascades, respectively. Interestingly, region supplied. In the stroke core, where glial activity in the stroke penumbra, a re- NF-␬B activation and TNF-␣ production is completely absent, irreversible cent study by Kaushal and Schlichter (2008) significantly increased, whereas neither p38 loss of tissue () occurs within min- published in The Journal of Neuroscience MAPK nor NO production were affected utes. The surrounding tissue, known as the uses an innovative three-stage in vitro exper- [Kaushal and Schlichter (2008), their Fig. 2 stroke penumbra, has limited perfusion and imental paradigm of the stroke penumbra (http://www.jneurosci.org/cgi/content/full/ impaired function, which may either re- [Kaushal and Schlichter (2008), their Fig. 1 28/9/2221/F2)]. cover or progress to infarction over time. http://www.jneurosci.org/cgi/content-nw/ Kaushal and Schlichter (2008) next ex- Penumbral damage might be mediated by full/28/9/2221/F1]. In the first stage, neu- amined which soluble factors might be ac- glutamate-induced excitotoxicity, because ron and astrocyte cocultures were placed tivating the NF-␬B signaling cascade. Be- Na ϩ-dependent glutamate transporters re- ϩ under oxygen and glucose deprivation cause in vivo evidence suggested that verse uptake after loss of ATP and Na (OGD) for various times, simulating the extracellular glutamate increased during pump function. However, damage might stroke core. In the second stage, these (Graham et al., 1990), and in also be caused by immune responses, which OGD-stressed cocultures (OGD-SCs) vitro evidence suggested that glutamate can persist for days after an ischemic insult were coincubated for 24 h with microglia activated microglia via group II metabo- (Dirnagl et al., 1999). cultured on Transwell inserts allowing tropic glutamate receptors (mGluRIIs) The initial immune response after a soluble factors released from OGD-SCs to (Taylor et al., 2005), the authors investi- stroke involves activation of microglia, resi- activate the microglia. In the third and fi- gated whether glutamate activated the mi- dent CNS mononuclear phagocytes. After nal stage, microglia were washed and in- croglia resulting in production of TNF-␣ activation, microglia retract their processes, cubated with naive /astrocyte co- in their model of the stroke penumbra. migrate to the site of injury, and release cultures for 48 h. This approach allowed Direct activation of microglia using a se- proinflammatory cytokines, nitric oxide the authors to examine whether microglia lective mGluRII agonist (DCG-IV (NO), neurotrophic factors, and chemo- were activated by OGD-SCs and whether [(2S,2ЈR,3ЈR)-2-(2Ј,3Ј-dicarboxycyclo- kines that attract other microglia and im- activated microglia produced soluble sig- propyl)glycine]) proved sufficient to in- mune cells (Lai and Todd, 2006). Although nals that subsequently damaged naive co- duce release of TNF-␣, but not NO activation of microglia after stroke has been cultures. Thus, this new method provided [Kaushal and Schlichter (2008), their well documented (Gerhard et al., 2005), a systematic approach to isolate and study Fig. 4 (http://www.jneurosci.org/cgi/ current in vitro models of microglial activa- key signaling factors that may contribute content/full/28/9/2221/F4)]. Addition- to cellular deterioration within the stroke ally, incubating naive cocultures with penumbra. agonist-treated microglia revealed an Received April 25, 2008; revised May 23, 2008; accepted May 28, 2008. Two key signaling pathways known for increase in death, which was We thank Drs. Marthe Howard and Joseph Margiotta for their construc- tive criticism and guidance in the Journal Paper Review course in the Neu- their production of proinflammatory sig- blocked by either an mGluRII antago- rosciences Graduate Program and for their advice in preparing this nals, nuclear factor ␬B (NF-␬B) and p38 nist [2-(S)-␣-ethylglutamic acid] or a manuscript. mitogen-activated protein kinase (MAPK), soluble TNF-␣ receptor (TNFR) (sT- CorrespondenceshouldbeaddressedtoTrishaA.Dwyer,Departmentof were assayed in microglia to determine NFR1). Together, these data suggest Neurology, University of Toledo College of Medicine, Mail Stop 1195, 3000 whether their activity levels increased after that activation of microglial mGluRIIs Arlington Avenue, Toledo, OH 43614. E-mail: [email protected]. ␣ DOI:10.1523/JNEUROSCI.1841-08.2008 exposure to OGD-SCs. Also assayed was the induces release of TNF- , contributing Copyright©2008SocietyforNeuroscience 0270-6474/08/286537-02$15.00/0 production of tumor necrosis factor ␣ to the death of naive /astro- 6538 • J. Neurosci., June 25, 2008 • 28(26):6537–6538 Dwyer et al. • Journal Club cytes. Although TNF-␣ release caused microglial inducible NO synthase is ele- alone. Nevertheless, the model could also be by activation of microglial mGluRIIs vated in the penumbra after transient used with purified neuronal cultures for di- has already been reported (Taylor et al., middle cerebral artery occlusion (Van- rect comparison with the results of Lai and 2005), Kaushal and Schlichter (2008) nucchi et al., 2007). Hence, there seems to Todd (2008). If the findings are the same as are the first to show that this occurs after be a disconnect between the current study in their current paper, then the investigators activation of microglia by OGD-SCs. and in vivo evidence regarding the pro- may want to examine versus OGD. Demonstrating this point, naive neu- duction of NO. The difference may be at- The penumbra model will also prove useful rons/astrocytes coincubated with mi- tributable to other soluble factors that for investigating neuron–neuron as well as croglia exposed to OGD-SCs in the come from blood supply or other cell astrocyte–neuron interactions and their presence of mGluRII antagonists exhib- types not present in Kaushal and Schlich- contributions to damage in the penumbra. ited reduced cell death [Kaushal and ter’s (2008) model, although the new In summary, this new model should provide Schlichter (2008), their Fig. 5 (http:// model as of yet is the closest to simulate valuable information about the soluble fac- www.jneurosci.org/cgi/content/full/28/ the stroke penumbra in vitro. tors and signaling pathways involved in 9/2221/F5)]. Contrasting the results of Kaushal and ischemia and guide researchers toward a One question that could be addressed Schlichter (2008), a concurrent in vitro suitable therapeutic target to decrease the in future studies is which signaling cas- study (Lai and Todd, 2008) found increased progressive damage occurring in the stroke cades TNF-␣ activates and through which production of NO and TNF-␣ in microglia penumbra. TNFRs. TNF-␣ is known to activate incubated for 24 h in medium that was con- TNFR types I (TNFRI) and II (TNFRII). ditioned by neurons exposed to 30 min hyp- References TNFRI activation is linked to neurode- oxia, but not neuron-conditioned medium Dirnagl U, Iadecola C, Moskowitz MA (1999) generation, whereas TNFRII mediates (NCM) from neurons subjected to longer Pathobiology of ischaemic stroke: an inte- survival during retinal ischemia (Fontaine durations of hypoxia (2, 4, and 6 h). The grated view. Trends Neurosci 22:391–397. Fontaine V, Mohand-Said S, Hanoteau N, Fuchs et al., 2002). The study by Kaushal and studies used similar methods: both exposed C, Pfizenmaier K, Eisel U (2002) Neurode- Schlichter (2008) supports TNFRI stimu- microglia to soluble factors from stressed generative and neuroprotective effects of tu- lation on neurons/astrocytes indicated by cultures and then incubate naive cells with mor necrosis factor (TNF) in retinal ischemia: caspase-8 activation [Kaushal and Schli- soluble factors produced by the microglia. opposite roles of TNF receptor 1 and TNF re- chter (2008), their Fig. 3 http://www. Yet, Lai and Todd (2008) did not find in- ceptor 2. J Neurosci 22:RC216. jneurosci.org/cgi/content/full/28/9/2221/ creased microglial TNF-␣ production after Gerhard A, Schwarz J, Myers R, Wise R, Banati RB (2005) Evolution of microglial activation in pa- F3]. However, the neuroprotective prop- longer periods of hypoxic treatment in neu- tients after ischemic stroke: a [ 11C](R)- erties of TNFRII have not been elucidated ronal cultures, whereas Kaushal and Schli- PK11195 PET study. NeuroImage 24:591–595. in the new stroke penumbra model. Ex- chter (2008) showed increased microglial Graham SH, Shiraishi K, Panter SS, Simon RP, Fa- periments using the new model along TNF-␣ production with increasing OGD den AI (1990) Changes in extracellular amino with transgenic neuronal knock-outs of insult in the cocultures. In addition, Kaushal acid neurotransmitters produced by focal cere- TNFRI and TNFRII would provide a and Schlichter (2008) did not show an in- bral ischemia. Neurosci Lett 110:124–130. ␣ Kaushal V, Schlichter LC (2008) Mechanisms of clearer picture of the effects of TNF- re- crease in NO production from microglia microglia-mediated neurotoxicity in a new lease on neurons and astrocytes. with increasing OGD insult, whereas Lai model of the stroke penumbra. J Neurosci The new model may also establish us- and Todd (2008) reported an increase in 28:2221–2230. ing microglial activation by OGD-SCs as a microglia exposed to a shorter hypoxic Kaushal V, Koeberle PD, Wang Y, Schlichter LC ϩ ϩ more relevant in vitro ischemic model NCM (30 min). The discrepancy in TNF-␣ (2007) The Ca 2 -activated K channel than other microglial activators like lipo- and NO production between the two stud- KCNN4/KCa3.1 contributes to microglia ac- polysaccharide (LPS). Kaushal and Schli- ies could be attributable to the severity of the tivation and nitric oxide-dependent neurode- generation. J Neurosci 27:234–244. chter’s (2008) current results differ from treatment or the type of culture used. Spe- Lai AY, Todd KG (2006) Microglia in cerebral those studies using LPS to activate the mi- cifically, Lai and Todd (2008) used hypoxia ischemia: molecular actions and interactions. croglia. A previous in vitro model using (reduced oxygen only) on Ͼ98% neuronal Can J Physiol Pharmacol 84:49–59. LPS implicated p38 MAPK and NO as cultures, whereas Kaushal and Schlichter Lai AY, Todd KG (2008) Differential regulation mechanistic players in neurodegeneration (2008) used OGD (reduced oxygen and glu- of trophic and proinflammatory microglial ef- (e.g., stroke) (Kaushal et al., 2007). In the cose) on 70%/30% neurons/astrocytes. The fectors is dependent on severity of neuronal injury. Glia 56:259–270. current study, microglial activation by amount or types of soluble factors activating Seki Y, Feustel PJ, Keller Jr RW, Tranmer BI, Kimel- OGD-SCs did not induce p38 MAPK ac- the microglia could be different because of berg HK (1999) Inhibition of ischemia- tivity or production of NO. Thus, Kaushal the disparities between cultures and/or induced glutamate release in rat striatum by di- and Schlichter (2008) have highlighted treatment conditions. hydrokinate and an anion channel blocker. the importance of using OGD-SCs to ac- Despite these differences, the new model Stroke 30:433–440. tivate microglia in stroke models, because of the stroke penumbra provides a useful Taylor DL, Jones F, Kubota ES, Pocock JM (2005) Stimulation of microglial metabo- activation using other molecules like LPS approach for assessing soluble factors in- tropic glutamate receptor mGlu2 triggers tu- may not induce the same microglial sig- volved in ischemia. One advantage of the mor necrosis factor-induced neurotoxicity in naling cascades as an ischemic event. model is the fact that the cultures contained concert with microglial-derived Fas ligand. Although using OGD-SCs to activate both neurons and astrocytes. Astrocytes are J Neurosci 25:2952–2964. microglia might be a more relevant model known to contribute to glutamate excito- Vannucchi MG, Bizzoco E, Corsani L, Gianfriddo of stroke than LPS, there are still differ- toxicity (Seki et al., 1999) and likely add to M, Pedata F, Faussone-Pellegrini MS (2007) ences between the authors’ findings and microglial activation of mGluRIIs. There- Relationships between neurons expressing neuronal nitric oxide synthase, degree of mi- those of other studies using in vivo isch- fore, the representation of multiple cell croglia activation and animal survival. A study emia or in vitro hypoxia as models of types more accurately reflects the in vivo sit- in the rat cortex after transient ischemia. Brain stroke. First, there is in vivo evidence that uation than using neurons or astrocytes Res 1132:218–227.