Neuron-glia communication: metallothionein ... - Wiley Online Library

Journal of Neurochemistry, 2004, 88, 454–461

doi:10.1046/j.1471-4159.2003.02193.x

Neuron-glia communication: metallothionein expression is specifically up-regulated by astrocytes in response to neuronal injury Roger S. Chung,* Paul A. Adlard, Justin Dittmann,* James C. Vickers,* Meng Inn Chuah* and Adrian K. West* *NeuroRepair Group, School of Medicine, University of Tasmania, Hobart, Tasmania, Australia Institute for Brain Aging and Dementia, Gillespie Neuroscience Research Facility, University of California Irvine, Gillespie, Irvine, California, USA

Abstract Recent data suggests that metallothioneins (MTs) are major neuroprotective proteins within the CNS. In this regard, we have recently demonstrated that MT-IIA (the major human MT-I/-II isoform) promotes neural recovery following focal cortical brain injury. To further investigate the role of MTs in cortical brain injury, MT-I/-II expression was examined in several different experimental models of cortical neuron injury. While MT-I/-II immunoreactivity was not detectable in the uninjured rat neocortex, by 4 days, following a focal cortical brain injury, MT-I/-II was found in astrocytes aligned along the injury site. At latter time points, astrocytes, at a distance up to several hundred microns from the original

injury tract, were MT-I/-II immunoreactive. Induced MT-I/-II was found both within the cell body and processes. Using a cortical neuron/astrocyte co-culture model, we observed a similar MT-I/-II response following in vitro injury. Intriguingly, scratch wound injury in pure astrocyte cultures resulted in no change in MT-I/-II expression. This suggests that MT induction was specifically elicited by neuronal injury. Based upon recent reports indicating that MT-I/-II are major neuroprotective proteins within the brain, our results provide further evidence that MT-I/-II plays an important role in the cellular response to neuronal injury. Keywords: brain injury, metallothionein, neuronal injury. J. Neurochem. (2004) 88, 454–461.

Metallothionein (MT) is a small, zinc-binding protein, expressed throughout the body. In the adult central nervous system (CNS), MT is mainly expressed in astrocytes, although there are some reports of neuronal expression (for an excellent review, see Hidalgo et al. 2001). Within humans, there are two main classes of MT found within the CNS: MT-I and MT-II (MT-I/-II), and MT-III (Hidalgo et al. 2001). While MTs are often considered in the context of zinc metabolism, or protection from free radical damage (see commentary by Palmiter 1998), more recently there have been reports indicating that this protein confers a protective effect following brain injury. Indeed, MT-I and MT-II knockout mice are susceptible to physical, chemical and ischemic brain injury (Penkowa et al. 1999a, 1999b; Carrasco et al. 2000; Trendelenburg et al. 2002), while mice that overexpress MT isoforms in the brain are comparatively more resistant to injury (Campagne et al. 1999; Giralt et al. 2002; Penkowa et al. 2002).

The mechanism by which MT is protective is currently unknown. Possibilities include the zinc-binding properties of the protein (seven zinc ions per protein molecule) or its ability to scavenge free radicals. An intriguing possibility, which has recently arisen from the work of Hidalgo and colleagues, is that MT is able to reduce inflammation associated with CNS injury, leading to enhanced recovery (Penkowa et al. 2000; Penkowa and Hidalgo 2001). Interestingly, we have recently discovered that human MT-IIA has a previously unsuspected ability to enhance neuronal

454

Received April 6, 2003; revised manuscript received September 29, 2003; accepted September 29, 2003. Address correspondence and reprint requests to Roger S. Chung, NeuroRepair Group, School of Medicine, University of Tasmania, PO Box 252–58, Hobart, Tasmania 7001, Australia. E-mail: [email protected] Abbreviations used: FCS, fetal calf serum; GFAP, glial fibrillary acidic protein; MT, metallothionein; MT-I/-II, metallothionein-I and –II; PBS, phosphate-buffered saline; PI, post injury.

2003 International Society for Neurochemistry, J. Neurochem. (2004) 88, 454–461

Neurons induce astrocytic MT in CNS injury

recovery following injury by directly acting upon neurons (Chung et al. 2003). This data suggests that MT promotes neuronal recovery by two distinct mechanisms: by decreasing the inflammatory response associated with injury; and by directly promoting neuronal recovery. Taken together, this work suggests that MT plays an important role in the cellular response to brain injury. To elucidate this, MT-I/-II expression was investigated in several different experimental models of cellular injury; focal rat cortical brain injury, and scratch wound injury in primary rat astrocyte and neuron/astrocyte co-cultures.

Materials and methods

455

in ice cold buffer (20 mM Tris-HCl, 1 mM DTT, pH 7.6) and centrifuged at 2000 g for 5 min. A sample of the supernatant (equivalent to 50 lg total protein) was electrophoresed on a 10% Bis-Tris Nu-PAGETM. gel (Invitrogen, Carlsbad, CA, USA) under reducing conditions (Nu-PAGETM. antioxidant buffer was added to the upper electrophoretic chamber), and western blotting performed as described previously (Chung et al. 2002). Briefly, the proteins were electro-transferred to a 0.45-lm nitrocellulose membrane. The membrane was blocked with 5% skim milk in phosphate buffered saline-Tween 20 (PBS-T), followed by incubation with the primary antibody (MT-I/-II, Dako) at a dilution of 1 : 500 in PBS-T. This was followed by incubation with the secondary antibody (rabbit anti-mouse IgG conjugated with HRP; AP160P, Chemicon, Melbourne, Australia), and immunodetection using the Lumi-light immunofluorescence kit (Roche, Mannheim, Germany).

Rat focal cortical brain injuries All procedures involving animals were approved by the Animal Experimentation Ethics Committee of the University of Tasmania, and are consistent with the Australian Code of Practise for the Care and Use of Animals for Scientific Purposes. Focal injuries to the rat neocortex were made as reported previously (King et al. 1997). Briefly, 250-g male rats were anaesthetised with Nembutal (0.1 mL/ 100 g, sodium pentobarbitol), the head shaved, and the rat then immobilised in a Stoelting stereotaxic frame. An incision was then made down the midline of the skull, and a hole then drilled through the skull above the Par 1 region of the somatosensory cortex. A Hamilton syringe with an attached 25-gauge bevelled needle was aligned with the hole, and then lowered to a depth of 1.5 mm into the brain. After 10 min, the syringe was slowly removed and the area filled with gel-foam. Antiseptic powder was then applied, and the skin stapled closed.

Neuron/astrocyte co-cultures Cortical tissue (including meningeal layers) was removed from embryonic day 19 (sperm positive day ¼ E1) Hooded Wistar rat embryos and incubated in sterile 10 mM HEPES buffer (37C). This was followed by trypsin digestion (0.25%; Sigma, St Louis, MO, USA), followed by addition of trypsin inhibitor (40 BTEE units/mg protein; Gibco, Auckland, New Zealand). Three gentle washes of the cell pellet were made using fresh HEPES buffer. The cell suspension was then triturated carefully using a 1-mL pipette. Cells were then plated onto glass coverslips (132 mm2) pre-coated overnight with 0.01% poly L-lysine (Sigma), at a cell density of 1 · 106 cells/well. Cultures were maintained at 37C in humidified air containing 5% CO2 for 21 days before injury. The culture medium consisted of NeurobasalTM. medium (Gibco), supplemented with 0.1% (f/c) B-27 supplement (Gibco), 0.1 mM (f/c) L-glutamine (Sigma), and 200 U/mL gentamicin (Sigma). Scratch wounds were made on an inverted microscope (Leitz Fluovent) using a fine goniotomy knife.

Fluorescent immunohistochemistry of rat brain sections At the appropriate time (1, 4, 7 or 14 days post injury), rats were re-anaesthetised and transcardially perfused with 0.01 M phosphatebuffered saline (PBS) followed by 0.01 M phosphate-buffered 4% paraformaldehyde. Brains were removed and post-fixed for 6 h in 0.01 M phosphate-buffered 4% paraformaldehyde at 4C. They were then sectioned by vibratome at a thickness of 50 lm. For immunohistochemistry, the sections were incubated with a combination of two primary antibodies; anti-MT-I/-II/anti-ferritin (1 : 500 Dako, Glostrup, Denmark; 1 : 10 000 ICN, Costa Mesa, CA, USA), or anti-MT-I/-II/anti-GFAP (1 : 500 Dako; 1 : 2000 Dako). Anti-ferritin was used as a marker of activated microglial cells (King et al. 2001). The antibodies were diluted in 0.1% PBS, 0.03% tritonX-100TM. Sections were then incubated with two secondary antibodies (goat anti-rabbit IgG conjugated to alexafluor 488, 1 : 250 dilution; Molecular Probes, Eugene, OR, USA, and horse anti-mouse rat adsorbed IgG, 1 : 250 dilution; Vector Laboratories, Burlingame, CA, USA) applied in 0.1% PBS. Brain sections were mounted using Permafluor mounting medium (Immunotech). Specimens were viewed on the Olympus BX-60 fluorescence microscope, or using confocal microscopy.

Primary astrocyte cultures Primary astrocyte cultures were prepared as described previously (Vincent et al. 2003). Briefly, cortical tissue was removed from neonatal rats (1–3 days), finely minced, and trypsinised. Ten per cent FCS (Sigma) was added to inactivate trypsin, and the cell suspension triturated and filtered through an 80-lm gauze filter. Cells were then plated into T-75 mL flasks pre-coated overnight with poly L-lysine (0.01%, Sigma) at a cell density of 1–2 · 107 with 10 mLs of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FCS and 1% penicillin/streptomycin. Cultures were maintained at 37C in humidified air containing 5% CO2 for 10 days. Flasks were then split into two T-75-mL flasks treated with poly L-lysine (Sigma) and 100 lL of Ara-C added (0.08391 g/10 mL, Sigma). When confluent, cells were removed from flasks by trypsinization (0.25%) and plated at a density of 2 · 104 cells/well on coverslips in 24-well plates. The purity of astrocyte cultures was determined by double labelling immunofluorescence, using an anti-GFAP primary antibody and nuclear yellow staining (to stain nuclei). Astrocyte cultures were at least 98% pure (results not shown).

Protein isolation, SDS–PAGE and western blotting A 7-mm wide cube of brain surrounding the lesion site was collected for each time point (three animals per group), pooled, homogenised

Fluorescent immunocytochemistry of primary cultures At the appropriate time, cells were fixed with 4% paraformaldehyde for 20 min. Coverslips with pure astrocyte cultures were then


456 R. S. Chung et al.

incubated with both anti-MT-I/-II (1 : 500; Dako) and anti-GFAP (1 : 500; Dako) antibodies diluted in 0.1% PBS, 0.03% triton X-100TM. Co-cultures received a combination of either anti-MT-I/-II and anti-GFAP or anti-MT-I/-II and anti-bIII-tubulin (1 : 10 000; Promega, Madison, WI, USA). Coverslips were then incubated with two secondary antibodies (horse anti-mouse IgG conjugated to alexafluor 488 and goat anti-rabbit IgG conjugated to alexafluor 594, 1 : 1000 dilution; Molecular Probes), applied in 0.1% PBS. Coverslips were mounted onto slides using Permafluor mounting medium (Immunotech, Marseilles, France). Analysis of MT-I/-II immunoreactivity All coverslips within an individual experiment underwent immunocytochemistry at the same time. Experiments were performed in triplicate. For analysis, 25 digital images were taken from three coverslips per group. In the case of coverslips with scratch wounds, images were captured adjacent to the injury site. All images within an individual experiment were captured at the same exposure time at 400 · magnification. Changes in MT-I/-II expression were qualitatively determined by mean intensity values, using a method based on that of Levadoux and colleagues (Levadoux et al. 1999). To validate the use of mean MT-I/-II intensity values as a method of observing changes in MT-I/-II expression at the immunocytochemical level, changes in MT-I/-II expression were observed following treatment of astrocyte cultures with zinc, a well-known inducer of MT-I/-II. Real-time PCR and western blotting results showed parallel changes in mean MT-I/-II intensity values (results not shown), indicating that this is a suitable method for qualitatively observing changes in MT-I/-II expression. Briefly, images were converted to black and white, and the mean intensity value of each image measured (Adobe Photoshop). Standard error and student’s t-test analysis were performed using SigmaStat (Jandel Scientific Software Corporation, San Rafael, CA, USA). Using this method, it was possible to directly compare different treatment groups within an individual experiment using quantitative tools. However, it was not possible to normalise intensity values between experiments performed in two or more different cultures (derived from separate fetuses). This is because the basal levels of MT-I/-II varied between different cultures, and immunocytochemistry and further analysis was not performed at the same time. In these cases, where results from different cultures were compared with each other, the overall changes in MT-I/-II expression (trendline) were qualitatively compared. Media exchange experiments In some cases, media exchange experiments were performed. In this case, media was collected from injured cultures (astrocyte only, neuron only or neuron/astrocyte co-cultures) and applied to uninjured astrocyte only or neuron/astrocyte co-cultures. Changes in MT-I/-II expression were analysed as discussed previously.

Results

MT-I/-II expression is up-regulated after focal injury to the rat neocortex We investigated the response of MT-I/-II following focal cortical brain injury, using a rat model previously characterised

within our laboratory (King et al. 1997, 2001). MT-I/-II and ferritin labelling were not evident in the normal uninjured neocortex. GFAP immunoreactivity however, was extensive and uniform throughout normal cortex, labelling glial cell bodies and their processes (results not shown). At one day post injury (PI), MT-I/-II labelling was absent (Fig. 1a). Furthermore, GFAP immunoreactivity remained extensive and uniform, similar to uninjured brains (results not shown). At 4 days PI, ferritin labelling was sparse (results not shown). MT-I/-II labelling was localised to glial-like cell bodies and their processes, both adjacent to the injury site and surrounding blood vessels (Fig. 1b). The pial surface, up to 100 lm to either side of the lesion tract, was also immunoreactive for MT-I/-II, with labelling confined to fibrous-like processes. Interestingly, at this and all later time points, all MT-I/-II labelled structures were immunoreactive for GFAP, but not all GFAP positive cells were labelled with MT-I/-II (Fig. 1b). At this time, GFAP immunoreactivity was more extensive throughout the cortex than at earlier time points, with many GFAP positive cells displaying the morphology of reactive astrocytes (results not shown). By 7 days PI, ferritin immunoreactivity was greatest, localised in cells in a bridge of tissue between either side of the original lesion tract, and remained distinct from the margins of the lesion. The lesion tract itself was demarcated by an intense band of MT-I/-II immunoreactivity (Fig. 1c). This MT-I/-II immunoreactivity was localised primarily in fibrous like processes, which were often thickened and aligned towards the tract (Fig. 1d). MT-I/-II immunoreactivity was also found in cell bodies along the lesion boundary. However, the most distinct glial (cell body and processes) labelling was evident up to 2 mm from the injury site. GFAP immunoreactivity was more extensive than at 4 days PI, and co-localised with MT-I/-II primarily in fibrous-like processes, which were often thickened and orientated towards the lesion tract (Fig. 1d). MT-I/-II labelling was found in a small number of ferritin immunoreactive microglia (Fig. 1e), in accordance with other reports in the literature (see review by Hidalgo et al. 2001). By 14 days PI, ferritin labelling had declined markedly (results not shown). MT-I/-II immunoreactivity was also reduced, confined largely to fibrous-like processes present at the pial surface, with very little labelling at the edges of the tract (results not shown). Western blotting analysis confirmed these immunohistochemical observations over the entire time course of experiments (Fig. 2). MT-I/-II is up-regulated following scratch wound in neuron/astrocyte co-cultures To investigate mechanisms underlying the induction of MTI/-II expression in response to neuronal injury, we performed physical injury (in the form of scratch wound injury) to cortical neuron/astrocyte co-cultures. In uninjured



457

Fig. 1 Immunohistochemical analysis following focal cortical brain injuries. At 1 day PI, MT-I/-II labelling is absent (a). Arrows indicate the location of the needle tract. At 4 (b) and 7 (c) days post injury, MT-I/-II immunoreactivity is localised to glia-like cells (arrowheads), cells and processes surrounding blood vessels (arrows in b), and processes along the pial surface (arrow in c). In (d) (confocal microscopy), cells labelled for MT-I/-II (green; arrowheads) have a cell body and/or processes labelled for GFAP (red). The pial surface is indicated (arrow). In contrast, very ferritin immunoreactive microglia (E; red) were co-labelled for MT-I/-II (green). The border of the needle tract is indicated (arrow).

1.

2.

3.

4.

5.

Fig. 2 Western blot analysis was performed for rat cortex proteins (1, uninjured; 2, 1 day PI; 3, 7 days PI; 4, 14 days PI, 5, sheep liver MT-I/-II). Immunoreactivity for MT-I/-II was present in a single band for the sheep liver sample as well as at 7 days PI in the rat brain.

co-cultures, MT-I/-II was detected at a basal level in some astrocytes (Fig. 3a) and most MTI/II immunoreactivity was nuclear. However, most astrocytes did not exhibit MT-I/-II immunoreactivity and it was interesting that a MT-I/-II positive astrocyte was often found beside an MT-I/-II negative astrocyte indicating that, in this context, MT expression is probably responding to intracellular factors (Fig. 3a). At 1 h post injury (PI), a number of MT-I/-II immunoreactive astrocytes were observed aligning along the scratch wound boundary (Fig. 3b). MT-I/-II labelling was found in both the cell body, as well as astrocytic processes, but was noticeably absent from the nucleus. Only astrocytes in close proximity to the scratch wound exhibited elevated MT-I/-II labelling.

By 6 h PI, a number of broad, flat, fibrous astrocytic processes were observed entering the scratch wound area, which were both GFAP and MT-I/-II immunoreactive (Fig. 3c). The distribution of astrocytes with elevated MT-I/II levels beside the scratch wound was similar to that observed at 1 h PI, although there appeared to be more of these cells at 6 h PI. Many of the astrocytes within 100 lm of the scratch wound exhibited increased MT-I/-II immunoreactivity (Fig. 3c). MT-I/-II immunoreactivity had significantly increased by 24 h PI, and was found in almost all astrocytes across the entire coverslip (Fig. 3d). MT-I/-II staining was found throughout the cell body and fine processes. Measurements of the mean intensity of MT-I/-II immunoreactivity (Fig. 4) in co-cultures indicated a statistically significant (p < 0.05), increasing trend from uninjured (average mean intensity value ¼ 37.57 ± 0.60) to 24 h PI (96.60 ± 2.38; p < 0.01, student’s t-test). MT-I/-II expression is unaltered following injury in primary astrocyte cultures To investigate whether the changes in MT-I/-II expression observed in neuron/astrocyte co-cultures could be attributed to physical injury to astrocytes, primary neonatal astrocyte cultures were physically injured by performing scratch wound injuries. No significant change in MT-I/-II expression was observed following injury, at up to 24 h PI (Fig. 5), either immediately adjacent to the injury site, or at up to 200 lm away.



Fig. 4 The mean intensity of MT-I/-II immunoreactivity surrounding the scratch wound site in co-cultures was measured up to 24 h PI. Uninjured astrocytes exhibited basal expression of MT-I/-II. Scratch wound injury resulted in a significant increase in MT-I/-II staining intensity over time. Error bars represent standard error values. Each trendline and corresponding symbol represents a single replicate experiment.

Fig. 3 MT-I/-II and GFAP double-immunocytochemical analysis following scratch wound injuries in neuron/astrocyte co-cultures. In uninjured cultures (a), MT-I/-II immunoreactivity was primarily nuclear. While all MT-I/-II immunoreactive cells were GFAP-positive astrocytes, not all astrocytes were MT-I/-II positive (arrows). At 1 h PI (b), MT-I/-II was significantly increased in some astrocytes aligned along the injury border (arrows), and was found throughout the cell body and associated processes. At 6 h PI (c) a number of astrocytes exhibited MT-I/-II immunoreactivity throughout the cell body and processes. While some MT-I/-II immunoreactive astrocytes were still aligned along the injury border, the majority of MT-I/-II-positive astrocytes were distributed dispersly up to 100 lm from the injury site. At 24 h PI (d), MT-I/-II immunoreactivity had significantly increased, and was most prominent in the cell body, but also in some processes. Scale bars represent 100 lm.

MT-I/-II expression increases in response to media from injured neurons Based on the observation that only astrocytes co-cultured with neurons up-regulate MT-I/-II expression in response to cellular injury, it was hypothesised that the induction of MT was neuron dependant. To investigate this hypothesis, a series of media exchange experiments were conducted. Briefly, culture medium from injured astrocyte, neuron or neuron/astrocyte cultures, all at 24 h PI, was applied to uninjured co-cultures. Based upon changes in MT-I/-II

Fig. 5 The mean intensity of MT-I/-II immunoreactivity following scratch wound injury in primary astrocyte cultures. No changes in MTI/-II expression were observed either immediately bordering the injury site (r), or at up to 200 lm from the injury border (j). Error bars represent standard error values.

Table 1 Culture media from injured cultures (astrocytes, neurons and co-cultures) were applied to uninjured co-cultures Injured culture media source

Average MT immunoreactivity value

No media change Injured astrocytes Injured neurons Injured co-cultures

1 0.96 ± 0.08 2.13 ± 0.07* 0.99 ± 0.13

The average MT immunoreactivity (as assessed by mean intensity values measured from digital images) was determined, and normalised to cultures from the same experiment that had received no media change. Only culture media from injured neuron cultures resulted in an increase in MT immunoreactivity. *p < 0.01; student’s t-test.

immunoreactivity (Table 1), only the culture medium from injured neuron cultures caused induction of MT-I/-II expression. In further, preliminary experiments, the same culture medium was applied to uninjured astrocyte cultures. In all cases, there was no observable change in MT-I/-II immunoreactivity after 24 h (results not shown).



459

Fig. 6 The mean intensity of MT-I/-II immunoreactivity following 1 mM glutamate treatment in primary astrocyte cultures. No increase in MT-I/-II expression was observed at up to 24 h post treatment. Error bars represent standard error values.

MT-I/-II is rapidly produced in response to glutamate treatment The requirement of a neuronal component for up-regulation in MT-I/-II expression in astrocytes following injury suggests communication between injured neurons and surrounding astrocytes. One of the major communication systems between neurons and astrocytes is based upon glutamate signalling (for review, see Vesce et al. 1999). To investigate whether glutamate could be involved in mediating the previous observations, uninjured astrocyte only and neuron/ astrocyte cocultures were treated with 1 mM glutamate. While MT-I/-II expression in astrocytes cultured alone did not increase following glutamate treatment at up to 24 h (Fig. 6), in neuron/astrocyte co-cultures this resulted in rapid MT-I/-II induction within astrocytes. Un-treated co-cultures exhibited no MT staining, or staining confined to the nucleus (Fig. 7a). Within an hour of glutamate treatment, a number of astrocytes exhibited strong MT-I/-II staining, although not all astrocytes exhibited MT-I/-II immunoreactivity. Indeed, MT-I/-II immunoreactive astrocytes were often found next to non-immunoreactive astrocytes (Fig. 7b). By 6 and 24 h post treatment, almost all astrocytes were strongly MT-I/-II immunoreactive, throughout the cytoplasm and associated processes (Figs 7c and d). To further investigate the possibility that glutamate is involved in the astrocytic up-regulation of MT-I/-II, we applied several glutamate receptor blockers (100 lm of either CNQX or MK801; Sigma) to neuron/astrocyte co-cultures 1 h prior to scratch injury. At 24 h PI, treatment with glutamate receptor blockers significantly reduced the increase in mean intensity of MT-I/-II immunoreactivity following scratch injury by 55.4% ± 3.2% (CNQX, p < 0.01) and 63.2% ± 2.7% (MK801, p < 0.01), respectively. Discussion

To investigate the role that MT plays in the cellular response to neuronal injury, MT-I/-II expression was examined in

Fig. 7 MT-I/-II and GFAP double-immunocytochemical analysis following 1 mM glutamate treatment in neuron/astrocyte cocultures. In un-treated cultures (a), MT-I/-II immunoreactivity was primarily nuclear. While all MT-I/-II immunoreactive cells were GFAP-positive astrocytes, not all astrocytes were MT-I/-II immunoreactive. At 1 h post treatment (PT), a number of astrocytes exhibited strong MT staining (b), but not all astrocytes were MT immunoreactive. By 6 h PT, almost all astrocytes were strongly MT immunoreactive (c), and this was similarly observed at 24 h PT (d). Scale bars ¼ 100 lm.

several different experimental models of neuronal injury. We observed a strikingly similar profile in MT-I/-II induction in both culture and in vivo models, differentiated only in their temporal expression profiles. Intriguingly, scratch wound injury in pure astrocyte cultures resulted in no change in MT-I/-II expression. This suggested that MT induction was specifically elicited by neuronal injury. This was further supported by media exchange experiments. Astrocytes are positioned within the CNS to make contact not only with neurons, but also with a number of neighbouring astrocytes (see review by Araque et al. 2001). Such association has been demonstrated to facilitate quick intracellular communication between cells (Giaume and McCarthy 1996; Venance et al. 1997). Furthermore, astrocytes have been demonstrated to respond physically to such intracellular signals (Lin et al. 1998) as well as via released



factors (Hassinger et al. 1996). In this regard, we observed a gradient in the number of MT-I/-II immunoreactive astrocytes which extended over time, beginning from alongside the injury site to include a number of astrocytes some distance from the original injury site. This would suggest that the astrocytes in the immediate vicinity of the original injury receive the signal to up-regulate MT expression, which is subsequently transferred to neighbouring cells. Surprisingly, it was noted that only uninjured co-cultures, and not uninjured astrocytes, responded to media from injured neuron cultures. A possible explanation for the medium exchange experiments discussed above is that neurons rapidly communicate with neighbouring astrocytes via an extracellular signalling factor. In preliminary experiments we investigated the possible involvement of glutamate in this injury-mediated cellular signalling process, as it is well known that glutamatergic signalling is involved in rapid cellular communication between neurons and astrocytes (for review see Vesce et al. 1999). We found that uninjured co-cultures treated with glutamate rapidly up-regulate expression of MT-I/-II in a similar temporal pattern as scratch injured co-cultures. Furthermore, the addition of glutamate receptor blockers partially but significantly inhibited the astrocytic up-regulation of MT-I/-II. These lines of evidence suggest that glutamate released from injured cells following scratch injury might be involved in the astrocytic up-regulation of MT-I/-II that we have observed. We also found that uninjured astrocyte cultures did not up-regulate MT-I/-II in response to glutamate treatment. As it has been reported that such cultures exhibit very low levels of glutamate receptor expression (Schlag et al. 1998), this further supports the hypothesis that glutamate is involved in the cell-to-cell signalling that regulates MT-I/-II levels within astrocytes following injury. However, we are unable to rule out the possibility that other factors may also be involved in up-regulating MT-I/-II expression following injury, as it has been clearly demonstrated that these proteins can be induced by a number of factors that could conceivably be active following neuronal injury, including nitric oxide, zinc and free radicals. As evidence for this, we noted that glutamate receptor blockers did not completely inhibit MT-I/-II up-regulation following injury, suggesting that other signalling factors may also be involved. Recently, Montoliu et al. (2000) demonstrated that exogenous MT-III is able to protect neurons from glutamate neurotoxicity (1 mM). Based on the high structural similarity between MT-I/-II and MT-III, it seems possible that MT-I/-II exhibits similar protective properties. As an indication of this, it has been demonstrated that both MT-I/-II and MT-III transgenic knockout mice are highly susceptible to kainic acid (a chemical glutamate analogue) -induced seizures (Erickson et al. 1997; Carrasco et al. 2000, respectively), while MT-III over expressing mice are protected from such

treatment (Erickson et al. 1997). Combined with the numerous observations that MT-I/-II is significantly up-regulated following neuronal injury, this is suggestive that these proteins are important in the brain’s response to glutamate neurotoxicity caused by either neuronal injury or prolonged synaptic activity. It was noted that there was a temporal difference in up-regulation in MT-I/-II expression in response to injury between the in vitro and rat brain injury models. This suggests that, following cortical brain injury, a number of factors might act together in a complex manner to regulate MT-I/-II expression in response to injury. In this regard, it is possible that, within the brain, a number of factors that are known to induce MT-I/-II expression in culture, including glutamate, zinc and cytokines, act together to regulate MTI/-II expression in response to injury. However, differences both spatially and temporally in these factors might allow for quite precise, controlled regulation of MT-I/-II expression. This would be in contrast to the prevailing thought that these proteins are regulated as a generalised response to a variety of cellular stresses. Recently, transgenic MT mice experiments have indicated that MTs have a protective role within the brain following injury. For instance, MT-I and MT-II knockout mice are susceptible to physical, chemical and ischemic brain injury (Penkowa et al. 1999a,b; Carrasco et al. 2000; Trendelenburg et al. 2002), whilst mice which over-express MT isoforms in the brain are comparatively more resistant to injury (Campagne et al. 1999; Penkowa et al. 2002). While it is unclear how MT acts to protect neurons following injury, some reports suggest the potential for an extracellular action (Giralt et al. 2002; Penkowa et al. 2002; Chung and West 2003; Chung et al. 2003). This is intriguing because the prevailing dogma is that MTs are only found, and act, inside cells (Palmiter 1998). We have demonstrated that MT-I/-II is specifically up-regulated by astrocytes, the main cellular source of MTs in the brain, in the proximity to neuronal injury, both in culture and in vivo. In view of recent literature, this could be viewed as a process facilitating MT-I/-II release into the extracellular environment in response to neuronal injury. Indeed, there are numerous reports of its detection in the extracellular environment in vivo (Garvey 1984; Bremner et al. 1987; Hidalgo et al. 1988), and in culture (Trayhurn et al. 2000). Elucidation of extracellular MT release mechanisms will improve our understanding of the role that these proteins play in the cellular response to neuronal injury. In summary, we have demonstrated that MT-I/-II expression is rapidly up-regulated in astrocytes in response to neuronal injury, both in culture and following cortical brain injury, in part mediated by rapid communication between neurons and astrocytes. In parallel to a number of other studies, this is a further indication of the recently identified importance of metallothioneins in the cellular response to neuronal injury.



Acknowledgement This work was funded by the National Health and Medical Research Council (NHMRC) of Australia.

References Araque A., Carmignoto G. and Haydon P. G. (2001) Dynamic signaling between astrocytes and neurons. Annu. Rev. Physiol. 63, 795–813. Bremner L., Mehra R. K. and Sato M. (1987) Metallothionein in blood, bile and urine. Toxicology 108, 129–140. Campagne M. V., Thibodeaux H., van Bruggen N., Cairns B., Gerlai R., Palmer J. T., Williams S. P. and Lowe D. G. (1999) Evidence for a protective role of metallothionein-1 in focal cerebral ischemia. Proc. Natl Acad. Sci. USA 96, 12870–12875. Carrasco J., Penkowa M., Hadberg H., Molinero A. and Hidalgo J. (2000) Enhanced seizures and hippocampal neurodegeneration following kainic acid-induced seizures in metallothionein-I + IIdeficient mice. Eur. J. Neurosci. 12, 2311–2322. Chung R. S. and West A. K. (2003) A role for extracellular metallothioneins in CNS injury and repair. Invited commentary in Neuroscience in press. Chung R. S., Holloway A. F., Eckhardt B. L., Harris J. A., Vickers J. C., Chuah M. I. and West A. K. (2002) Sheep have an unusual variant of the brain-specific metallothionein, metallothionein-III. Biochem. J. 365, 323–328. Chung R. S., Vickers J. C., Chuah M. I. and West A. K. (2003) Metallothionein-IIA promotes initial neurite elongation and post-injury reactive neurite growth, and facilitates healing following focal cortical brain injury. J. Neurosci. 23, 3336–3342. Erickson J. C., Hollopeter G., Thomas S. A., Froelick G. J. and Palmiter R. D. (1997) Disruption of the metallothionein-III gene in mice: analysis of brain zinc, behavior, and neuron vulnerability to metals, aging, and seizures. J. Neurosci. 17, 1271–1281. Garvey J. S. (1984) Metallothionein: structure/antigenicity and detection/ quantitation in normal physiological fluids. Environ. Health Prosp. 54, 117–127. Giaume C. and McCarthy K. D. (1996) Control of gap-junctional communication in astrocytic networks. Trends Neurosci. 19, 319– 325. Giralt M., Penkowa M., Lago N., Molinero A. and Hidalgo J. (2002) Metallothionein-1+2 protect the CNS after a focal brain injury. Exp. Neurol. 173, 114–128. Hassinger T. D., Guthrie P. B., Atkinson P. B., Bennett M. V. and Kater S. B. (1996) An extracellular signaling component in propagation of astrocytic calcium waves. Proc. Natl Acad. Sci USA 93, 13268– 13273. Hidalgo J., Giralt M., Garvey J. S. and Armario A. (1988) Physiological role of glucocorticoids on rat serum and liver metallothionein basal and stress conditions. Am. J. Physiol. 254, E71–E78. Hidalgo J., Aschner M., Zatta P. and Vasak M. (2001) Roles of the metallothionein family of proteins in the central nervous system. Brain Res. Bull. 55, 133–145. King C. E., Jacobs I., Dickson T. C. and Vickers J. C. (1997) Physical damage to rat cortical axons mimics early Alzheimer’s neuronal pathology. Neuroreport 8, 1663–1665. King C. E., Canty A. J. and Vickers J. C. (2001) Alterations in neurofilaments associated with reactive brain changes and axonal

461

sprouting following acute physical injury to the rat neocortex. Neuropathol. Appl. Neurobiol. 27, 115–126. Levadoux M., Mahon C., Beattie J. H., Wallace H. M. and Hesketh J. E. (1999) Nuclear import of metallothionein requires its mRNA to be associated with the perinuclear cytoskeleton. J. Biol. Chem. 274, 34961–34966. Lin J. H., Weigel H., Cotrina M. L., Liu S., Bueno E., Hansen A. J., Hansen T. W., Goldman S. and Nedergaard M. (1998) Gap-junction-mediated propagation and amplification of cell injury. Nat. Neurosci. 1, 494–500. Montoliu C., Monfort P., Carrasco J., Palacios O., Capdevila M., Hidalgo J. and Felipo V. (2000) Metallothionein-III prevents glutamate and nitric oxide neurotoxicity in primary cultures of cerebellar neurons. J. Neurochem. 75, 266–273. Palmiter R. D. (1998) The elusive function of metallothioneins. Proc. Natl Acad. Sci. USA 95, 8428–8430. Penkowa M. and Hidalgo J. (2001) Metallothionein treatment reduces proinflammatory cytokines IL-6 and TNF-a and apoptotic cell death during experimental autoimmune encephalomyelitis (EAE). Exp Neurol. 170, 1–14. Penkowa M., Carrasco J., Giralt M., Moos T. and Hidalgo J. (1999a) CNS wound healing is severely depressed in metallothionein I- and II-deficient mice. J. Neurosci. 19, 2535–2545. Penkowa M., Giralt M., Moos T., Thomsen P. S., Hernandez J. and Hidalgo J. (1999b) Impaired inflammatory response to glial cell death in genetically metallothionein-I- and –II-deficient mice. Exp. Neurol. 156, 149–164. Penkowa M., Giralt M., Carrasco J., Hadberg H. and Hidalgo J. (2000) Impaired inflammatory response and increased oxidative stress and neurodegeneration after brain injury in interleukin-6-deficient mice. Glia 32, 271–285. Penkowa M., Giralt M., Camats J. and Hidalgo J. (2002) Metallothionein 1+2 protect the CNS during neuroglial degeneration induced by 6-aminonicotinamide. J. Comp. Neurol. 444, 174–189. Schlag B. D., Vondrasek J. R., Munir M., Kalandadze A., Zelenaia O. A., Rothstein J. D. and Robinson M. B. (1998) Regulation of the glial Na+-dependent glutamate transporters by cyclic AMP analogs and neurons. Mol. Pharmacol. 53, 355–369. Trayhurn P., Duncan J. S., Wood A. M. and Beattie J. H. (2000) Regulation of metallothionein gene expression and secretion in rat adipocytes differentiated from preadipocytes in primary culture. Horm. Metab. Res. 32, 542–547. Trendelenburg G., Prass K., Priller J. et al. (2002) Serial analysis of gene expression identifies metallothionein-II as major neuroprotective gene in mouse focal cerebral ischemia. J. Neurosci. 22, 5879–5888. Venance L., Stella N., Glowinski J. and Giaume C. (1997) Mechanism involved in initiation and propagation of receptor-induced intercellular calcium signaling in cultured rat astrocytes. J. Neurosci. 17, 1981–1992. Vesce S., Bezzi P. and Volterra A. (1999) The highly integrated dialogue between neurons and astrocytes in brain function. Sci. Prog. 82, 251–270. Vincent A. J., West A. K. and Chuah M. I. (2003) Morphological plasticity of olfactory ensheathing cells is regulated by cAMP and endothelin-1. Glia 41, 393–403.
