Sage Journals: Discover world-class research

Abstract

Ischemic preconditioning is a phenomenon that describes how a sublethal ischemic insult can induce tolerance to subsequent ischemia. This phenomenon has been observed after focal or global ischemia in different animal models. However, the hypothesis that bacterial infection might lead to neuronal tolerance to injury has not been investigated. To mimic cerebral bacterial infection, we injected bacterial lipopolysaccharide (LPS) in the right dorsal hippocampus, followed 24 hours later by an excitotoxic lesion using kainic acid in the mouse model. Quantification of lesion size after cresyl violet counterstaining revealed that LPS pretreatment afforded neuroprotection to CA3 neurons against KA challenge. To investigate the events underlying this protection, we studied the cytokine profile induced after LPS injection. Interleukin (IL)-1β and transforming growth factor beta 1 (TGF-β1) were the main cytokines expressed at 24 hours after LPS injection. Because IL-1β has been described as deleterious in acute injury, we decided to investigate the function of TGF-β1. An adenovirus expressing a constitutively active form of TGF-β1 was injected intracerebrally 1 week before the induction of excitotoxic lesion, and neuronal protection was observed. To confirm the neuroprotective role of TGF-β1, the TGF-β1 adenovirus was replaced by recombinant human TGF-β1 protein and total neuroprotection was observed. Furthermore, the antibody-mediated blocking of TGF-β1 action prevented the protective effect of pretreatment with LPS. We have demonstrated in vivo that the cerebral tolerance phenomenon induced by LPS pretreatment is mediated by TGF-β1 cytokine.

Keywords

Bacterial infection Neuroprotection TGF-β1 Adenovirus

Ischemic preconditioning, a phenomenon describing how a brief episode of sublethal ischemia can induce tolerance to a subsequent ischemia, was first described in the heart (Murry et al., 1986). Using a canine model of myocardial infarction, the authors showed that the size of an infarct could be reduced to 25% when four episodes of 5 minutes of circumflex occlusions were followed by a sustained 40-minute occlusion. The tolerance or preconditioning phenomenon has also been observed in different organs including the brain. Cerebral ischemic preconditioning was first demonstrated in vivo by use of a global ischemia model in gerbils (Kitagawa et al., 1990). When two short sublethal ischemic episodes of 2 minutes were applied more than 24 hours before a longer ischemic insult of 5 minutes, protection of the CA1 hipoccampal neurons was observed, thus showing that a sublethal ischemia may confer ischemic tolerance in the brain. The acquisition of tolerance is a transient phenomenon, which requires at least 24 hours to be induced (Kitagawa et al., 1991). Cerebral ischemic tolerance was observed not only in the gerbil model but also in the C57BL/6J mouse (Wu et al., 2001) and in the rat (Liu et al., 1992; Nishi et al., 1993; Simon et al., 1993) after forebrain or focal cerebral ischemia.

The precise mechanism of cerebral tolerance is still not fully understood. However, many candidates have been suggested as possible inductors, transduction, and effectors of ischemic tolerance. The delay and the observed time window of brain preconditioning may reflect the underlying changes in cellular metabolism and gene expression. The suggested mechanisms of ischemic tolerance are as follows: (1) membrane stabilization and inhibition of excitability by an action on the ATP-sensitive K⁺ channel or the glucose receptor, therefore reducing the excitability in the hippocampus; (2) cell protection and inhibition of apoptosis by protein phosphorylation, nitric oxide production, cytokines, and glial cell proliferation; and (3) stress response (for review, see Kirino, 2002).

In an attempt to understand the molecular mechanisms of this neuroprotection, several groups have been able to mimic the preconditioning in vivo and in vitro with molecules such as ATP-sensitive K⁺ channels agonists or NMDA (Kasischke et al., 1996; Riepe et al., 1997), with agonists and antagonists of the adenosine A1 receptor (Heurteaux et al., 1995), electrically via the application of 2M KCl (Kawahara et al., 1995; Kobayashi et al., 1995) and by anoxia/hypoxia in various models (Pohle and Rauca, 1994; Simon et al., 1993), therefore inducing a cross-tolerance (for review, see Kirino, 2002). However, the hypothesis that a bacterial infection might also lead to cerebral tolerance has not been investigated extensively. When injected systematically to mimic peripheral infection, bacterial lipopolysaccharide (LPS) has been shown to induce tolerance to focal brain ischemia (Tasaki et al., 1997) through the production of the cytokines interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α). We have previously shown that the intracerebral (i.c.) injection of LPS leads to glial cell activation without inducing neuronal death (Andersson et al., 1992) and to local cytokine synthesis at 4 and 24 hours (Cunningham et al., 2002; Walsh et al., 2001). To test the hypothesis that LPS injection and glial activation could be protective against acute neuronal degeneration, we used a model of intracerebral endotoxin injection and investigated whether TGF-β1 might confer neuroprotection. Twenty-four hours after the i.c. injection of LPS, an excitotoxic lesion was induced by the injection of kainic acid (KA). The protective effect of LPS against the acute neurodegeneration was examined by measuring the lesion area. We have previously shown that transforming growth factor (TGF)-β1 is the principal cytokine expressed 24 hours after LPS injection in the brain (Cunningham et al., 2002). TGF-β1 is known to be multifunctional in many pathophysiologic processes, such as cell growth and differentiation (Massague et al., 1992), inflammation (Ruscetti and Palladino, 1991), and cell repair (Sporn et al., 1986). TGF-β1 is minimally expressed in the intact brain, but its expression increases after human stroke (Ata et al., 1997; Krupinski et al., 1996) and experimental ischemic lesions (Lindholm et al., 1992; Logan et al., 1992; Morgan et al., 1993; Wang et al., 1995; Wiessner et al., 1993). To evaluate the role of TGF-β1 in the neuroprotective process, we used an adenovirus expressing the active form of TGF-β1 gene (Sime et al., 1997).

METHODS

Animals

Male and female C57BL/6J mice aged 6 to 8 weeks were anesthetized with Avertin (2,2,2 tribromoethanol in tertiary amyl alcohol) at a dose of 0.1mL/5g body weight and injected stereotactically in the dorsal hippocampus via a pulled glass micropipette (coordinates: bregma, −1.9 mm; lateral, −1.5 mm; depth, −1.37 mm). The experiments were performed under Home Office Licence (UK).

Acute injury

The control groups were injected with a dose of 0.25 nM of KA acid (KA) (n = 6) or 1 μg/μL LPS (n = 3). A further group was injected with a dose of 1 μg/μL LPS, followed 24 hours later with a dose of 0.25 nM KA (n = 6). Another group was injected with a mixture of 1 μg/μL LPS and 0.25nM KA (n = 6). After surgery, the mice were allowed to recover under a warm lamp and then housed in IVC racks (Techniplast UK, Kettering). Five days after the KA injection or LPS alone, the mice were deeply anesthetized and transcardially perfused with fixative, and the brains were removed and embedded in OCT as previously described (Andersson et al., 1991a).

TGF-β1 cytokine

Protein

Recombinant TGF-β1 human (RhTGF-β1) protein (4 ng; R&D Systems Europe Ltd, Abingdon, UK) was injected 1 hour before the KA injection in one group of six mice.

Anti-human TGF-β1 RII antibody

Anti-human TGF-β1 type II receptor (αTGFβ1-RII) (500 ng; R&D Systems Europe Ltd, Abingdon, UK) was injected 24 hours after the injection of 1 μg/μL LPS in one group of eight mice. Another group of six mice were injected 24 hours after the LPS injection with 250 ng αTGFβ1-RII, followed immediately by the KA injection.

Quantification of the lesion

Sections measuring 10 μm were cut on a cryostat, and sections were collected every 50 μm throughout the dorsal hippocampus. The sections were stained with a Nissl stain to reveal the area of neuronal degeneration. The anterior limit of the lesion was defined as the first section presenting a minimum of 10 pyknotic neurons, probably undergoing apoptosis, in one microscopic field. Similarly, the caudal limit of the lesion was defined as the last section presenting a maximum of 10 pyknotic neurons within a field. The edges of the lesions were determined by the border between the healthy neurons and the pyknotic neurons (Fig. 2B). The extent of the neuronal loss was quantified by serial reconstruction of the dorsal hippocampus pyramidal fields with LEICA Qwin image analysis software (Leica Imaging Systems Ltd., Cambridge, UK). The lesion was always restricted to the CA3 region of the hippocampus. The area of neuronal loss between two sections was calculated according to the following mathematical formula: area = (a +b)×d ÷ 2, as illustrated in Fig. 1. The total area of neuronal loss of entire right dorsal hippocampus is the result of the sum of the lesion area calculated as described above.

FIG. 1.

Representative diagram of the lesion measurement. The mathematical formula used to calculate the lesion between two sections is as follows: (a +b)×d ÷2, where d represents the distance between two sections (50 μm), and a and b represent the length of the lesion of two successive sections.

FIG. 2.

Detection of the neuronal injury by Cresyl Violet counterstaining of coronal sections in the region of the CA3 neurons. Animals were injected with (A) LPS, (B) KA, (C) LPS followed 24 hours by KA, (D) AdTGFβ1^223/225 adenovirus followed 7 days later by KA, (E) rhTGFβ1 followed 1 hour later by KA, and (F) LPS followed 24 hours later by αTGFβ1-RII antibody and KA. Scale bar = 50 μm

Immunohistochemistry

Frozen 10-μm coronal serial sections were cut on a cryostat and mounted on 3-aminopropyltriethoxysilane-coated slides. The sections were air-dried and fixed in ice-cold absolute ethanol for 10 minutes before staining. The primary antibodies, anti-CD68 mAb (FA-11) and anti-CD3 mAb (KT3) (Serotec, Oxford, UK), were incubated on the sections for 2 hours and detected using a biotinylated secondary antibody and the ABC Elite detection kit from Vector laboratories (Peterborough, UK) with 3,3 diaminobenzidine (DAB) as previously described (Walsh et al., 2001). Sections used for photographic purposes were counterstained with cresyl violet.

Determination of IL-1β, IL-10, TNF-α and TGF-β1 in brain homogenates by ELISA

The concentrations of IL-1β, IL-10, TNF-α and TGF-β1 in brain tissue were determined by ELISA. Animals were transcardially perfused with heparinized saline. Brains were removed and a thick (2mm) coronal section taken through the hippocampus. The hippocampus was then punched from this section, snap frozen in liquid nitrogen and homogenized using disposable pestle and mortars in phosphate buffered saline (PBS) containing the protease inhibitor cocktail Complete. Homogenates (0.1 g/mL) were centrifuged at 10,000g for 5 minutes at 4°C, and the supernatants were assayed according to the manufacturers instructions for IL-1β, IL-10, and TNF-α ELISA kits (Quantikine). All standards and samples were diluted in the extraction buffer and assayed in duplicate. For the determination of TGF-β1, homogenates were subjected to a pretreatment for the conversion of latent TGF-β1 to active TGF-β1. The conversion was effected by acidification with 2.5 M acetic acid and 10 M urea for 10 minutes, followed by neutralization with 2.7 M NaOH and 1 M HEPES and a final 1:25 dilution in wash buffer. After centrifugation, the samples were added to plates coated overnight with recombinant human TGF-β sRII/Fc chimera. After 3 hours incubation, the plates were washed, and biotinylated polyclonal anti-TGF-β1 antibody at 200 ng/mL was added. Bound TGF-β1 was detected as described previously (Cunningham et al., 2002).

To determine the in vivo time course of TGF-β1 expression by the adenovirus expressing the active form of TGF-β1 (AdTGFβ1^223/225), six naive mice were injected in the right dorsal hippocampus at the concentration of 10⁵ pfu/μL. Two animals were killed at each of 3 time points (3, 7 and 10 days), and the TGF-β1 was determined as described above. The limits of reliable quantitation of the assays for IL-10, TNF-α and TGF-β1 were less than 5 pg/mL, 4 pg/mL, and 15 pg/mL, respectively.

Statistics

The data are presented as the mean ± standard deviation of the mean. Where statistical analysis has been used, ANOVA test with Fischer's post hoc test was applied.

Reagents

Adenovirus

The construction of the adenoviral vector has been described previously in detail (Bett et al., 1994). We have used the AdTGFβ1^223/225 adenovirus, expressing the active form of TGF-β1 and the AdDL70-3, the control virus with no exogenous cDNA insert. They were used at the concentration of 10⁵ pfu/μL. The production of TGF-β1 by the AdTGFβ1^223/225 construct in the brain was confirmed by ELISA.

LPS

LPS (Equine abortus) was from Sigma Chemicals (Poole, Dorset).

ELISA

IL-1β and TNF-α ELISA kits (Quantikine), recombinant human TGF-β sRII/Fc chimera, recombinant TGF-β1, and biotinylated polyclonal anti TGF-β1 antibodies used for the TGF-β1 ELISA were obtained from R&D Systems Europe Ltd. (Abingdon, UK).

RESULTS

LPS injection

An intracerebral injection of 1 μg/μL LPS was made into the right dorsal hippocampus of five C57BL/6J mice. The animals were killed 5 days later. To identify and quantify the lesion, the sections were counterstained with Cresyl Violet. Examination of the brain sections revealed no neuronal death (Fig. 2A).

Kainic acid injection

KA acid (0.25 nM) was injected into the right dorsal hippocampus of six mice. The animals were killed 5 days later. A clear lesion was evident with a specific neuronal loss in CA3 as previously reported (Andersson et al., 1991b). The lesion is illustrated in Fig. 2B. The extent of neuronal loss was quantified by serial reconstruction of the lesion in the dorsal hippocampus pyramidal fields (Fig. 3).

FIG. 3.

Quantification of the size of the lesion in animals injected with KA (n = 6), LPS followed 24 hours by KA (n = 6), and a mix of LPS and KA (n = 6) after Cresyl Violet counterstaining. ∗ P < 0.0001 for LPS injection followed 24 hours later by the KA versus unique KA injection by Fisher's test. ∗∗ P = 0.0017 for LPS and KA injected together versus unique KA injection by Fisher's test.

LPS/Kainic acid

An intracerebral injection of 1 μg/μL LPS in the right dorsal hippocampus was followed 24 hours later by KA acid injection. The KA injection was made in exactly in the same site as the LPS injection, using the same stereotaxic coordinates. Six mice were injected. The mice were killed 5 days later, and the lesions were quantified. A restricted area of neuronal death was present in the CA3 field of one animal, but in the other five animals, no lesion was visible (Fig. 1C). The difference between the two animal groups is significant (P < 0.0001) (Fig. 3).

Six mice were injected with a mixture containing 1 μg/μL LPS and 0.25 nM KA. The injection was made into the right dorsal hippocampus. The animals were killed 5 days after this injection. The combination of both LPS and KA acid induces a lesion that was smaller (7.73 × 10⁴ ± 1.48 μm²) than that seen with KA acid injection alone (17.72 × 10⁴ ± 3.18 μm²). The difference in lesion area between the injection of LPS and KA and the KA acid injection alone is significant (P = 0.0017). However, if we compare the lesion size following injection of the mixture of LPS and KA with the lesion following the injection of the LPS and KA with a delay of 24 hours, the lesion as a consequence of the single mixture injection of LPS and KA is seven fold bigger than the lesion resulting of the injection of LPS followed 24 hours later by KA. This difference between the two protocols is significant (P = 0.0299) (Fig. 3). There is clearly a difference in the neuroprotection in the group that received LPS 24 hours before the KA injection and the group that received the mixture of LPS and KA.

LPS-evoked cytokine profile

Because LPS appears to be protective, we were interested to know which cytokines were expressed in LPS-injected brains. The profile of cytokines in brains injected with LPS was determined by ELISA kit detection. The brains were injected in the right dorsal hippocampus with 2.5μg/μL LPS. The injected sites were dissected and prepared for cytokine detection by ELISA 24 hours after the LPS injection. Four cytokines were studied: two proinflammatory cytokines, IL-1β and TNF-α, and two antiinflammatory cytokines, TGF-β1 and IL-10. The results are illustrated in Fig. 4. All cytokines show significant differences between LPS and saline intra-cerebral injections (IL-1βP = 0.001; TNF-αP = 0.0001; TGF-β1, P = 0.0016; IL-10, P = 0.0010).

FIG. 4.

Quantification of IL-1β, TNF-α, IL-10, and TGF-β1 protein in animals injected with saline (n = 3) and LPS (n = 3) and killed 24 hours later by ELISA. ∗ P < 0.0001 for saline versus LPS injection by Fisher's test. Open box, saline-injected animals; Closed box, LPS-injected animals.

Adenovirus TGF-β1

The most likely cytokine to be protective was TGF-β1 (Henrich-Noack et al., 1996; Prehn et al., 1993a; Prehn et al., 1993b). We therefore examined whether the delivery of TGF-β1 was neuroprotective either when delivered as the cytokine alone or with an adenoviral vector system. The determination of the TGF-β1 expression by the AdTGF-β1^223/225 adenovirus was performed after injection of the vector in the right dorsal hippocampus at the concentration of 10⁵ pfu/μL. At 7 days postinoculation, a peak of TGF-β1 expression was observed by ELISA (564 pg/mg) (data not shown). At this time, the microglia stained by a CD68 marker shows a very weak activation of microglia without lymphocyte recruitment as observed by the CD3 staining (data not shown).

Seven days after the adenovirus injection, during the peak of cytokine expression, KA was injected in the same site in the dorsal hippocampus, and the animals were killed 5 days later. For the control animals, the AdTGFβ1^223/225 was replaced by AdDL70-3. The brains from mice injected with AdDL70-3 followed 7 days later by the KA showed clear lesions in the CA3 area (9.37 × 10⁴ ± 1.33 μm²). These mice showed a significant difference in lesion size compared with the animals injected with KA alone (17.72 × 10⁴ ± 3.18 μm²) (Fig. 5) and with the animals injected with LPS followed 24 hours later by KA (7.73 × 10⁴ ± 1.48 μm²) (P = 0.0088).

FIG. 5.

Quantification of the size of the lesion in animals injected with KA (n = 6), with LPS followed 24 hours by KA (n = 6), with a recombinant human TGFβ1 protein followed 1 hour later by KA (n = 5), with AdTGFβ1^223/225 adenovirus followed 7 days later by KA (n = 7), with TGF-β1 RII antibody followed by KA and 24 hours after the LPS (n = 6), with Ad70-3 control adenovirus (n = 5) followed 7 days later by KA, and with TGF-β1 RII antibody 24 hours after the LPS (n = 8). ∗ P < 0.0001 for KA versus LPS 24 hours and KA, rhTGF-β1 and KA, and AdTGFβ1^223/225 and KA. ∗∗ P = 0.048 for KA versus LPS and αTGFβ1-RII and KA.

The group with the AdTGFβ1^223/225 shows a total neuroprotection for three mice and a clear reduction in lesion size for the other four animals (Fig. 2D). The mean size of the lesion in the seven mice is 4.35 × 10⁴ ± 1.86 μm². The difference in the size of the lesion between the animals with KA acid injection and the animals that have received AdTGFβ1^223/225 adenovirus 7 days before the injury is significant (P < 0.0001), as shown in Fig. 5.

TGF-β1 cytokine

To further investigate the role of TGF-β1, two other experiments were conducted. Five animals were injected with 4 ng of human TGF-β1 recombinant protein in the right dorsal hippocampus 1 hour before KA acid injection. In four mice, there was no neuronal loss, and in one mouse there was a small lesion. However, the structure of the CA3 neurons in the animals with the absence of obvious lesion shows a disruption between the neuronal cells in the injected side by comparison with the noninjected side (Fig. 2E), but there were no apoptotic neurons. This group shows a significant difference of the size of the lesion in comparison to the five other groups (P < 0.0001) (Fig. 5). Additional experiments were performed to block the TGF-β1 neuroprotection by using a soluble TGF-β1 receptor II antibody (αTGFβ1-RII) as an antagonist. Six mice were injected 24 hours after the LPS injection with 250 ng of the αTGFβ1-RII, followed immediately by the 0.25 nM of KA. The animals were killed 5 days later. As a control, we injected eight mice with 500 ng of the αTGFβ1-RII 24 hours after the LPS injection. αTGFβ1-RII that was injected 24 hours after the LPS induced a small lesion (1.70 × 10⁴ ± 0.9 μm²) (Fig. 5). The lesion induced by the injection of αTGFβ1-RII and KA 24 hours after the LPS pretreatment is greater than the lesion observed with KA alone (32.44 × 10⁴ ± 13.19 μm² versus 17.72 × 10⁴ ± 3.18 μm²) (P = 0.047) (Fig. 5), and it is observed in the CA3 or CA1 area. This lesion is also significantly different to the lesion induced by KA after LPS activation (P < 0.0001).

DISCUSSION

The tolerance phenomenon and neuronal sparing induced by a sublethal ischemic lesion have been demonstrated in vivo by Kitagawa et al. (1990). However, the mechanisms leading to neuronal protection against an excitotoxic lesion have not been elucidated. The majority of animal models investigating tolerance concern sequential ischemic insults: (1) global-global ischemia, (2) focal-global ischemia, and (3) global-focal ischemia. No investigation on the effects of cerebral bacterial infection followed by a subsequent insult has been carried out. The mechanisms of ischemic tolerance or tolerance to excitotoxicity and the resulting preservation of neurons is likely to be important in a variety of neuropathologic diseases.

In this study, to mimic a bacterial infection, we injected the proinflammatory agent LPS into the dorsal hippocampus of mice. We observe that LPS affords protection of approximately 93% of CA3 neurons killed by this particular dose of KA. LPS gives the maximum neuroprotection when it is injected 24 hours before KA. The observed delay is necessary to obtain the protection; it has been described in the different models (Barone et al., 1998; Kitagawa et al., 1991; Perez-Pinzon et al., 1997a, 1997b) and is correlated with protein synthesis. To investigate the possible role of cytokines after the endotoxin injection, we have examined the cytokine profile in the hippocampus 24 hours after i.c. LPS injection. In this study, we have examined two proinflammatory cytokines, IL-1β and TNF-α, and two antiinflammatory cytokines, TGF-β1 and IL-10, because the balance between pro- and antiinflammatory cytokines in the brain may have a profound impact on the preservation of neuronal elements. We have shown that the levels of both IL-1β and TGF-β1 are significantly increased 24 hours after i.c. LPS injection, and this is associated with a protection of CA3 neurons.

To characterize the role of TGF-β1 in neuroprotection, we have used an adenovirus expressing the active form of TGF-β1 (Sime et al., 1997). The AdTGFβ1^223/225 adenovirus does not induce an acute inflammation 7 days after the intracerebral injection; no leukocyte recruitment was detected, and only weak microglia activation was observed. The excitotoxic lesion produced by injection of KA 7 days after the AdTGF-β1^223/225 adenovirus injection induces a neuronal death that is significantly smaller. To confirm that this neuroprotection can be attributed to TGF-β1, we have injected recombinant human TGF-β1 protein 1 hour before the KA injection. A total neuroprotection against excitotoxic lesion was observed. It is of interest that the control adenovirus AdDL70-3 also afforded small protection. An undetectable inflammation may occur after the i.c. AdDL70-3 injection and which may, in turn, induce protection. We have used a human TGF-β1 RII antibody as an antagonist of TGF-β1. The i.c. injection of TGF-β1 RII antibody 24 hours after the LPS injection and followed by the KA injection has produced a lesion that is twofold greater than that induced by KA alone. It is of interest to note that the observed lesion included both CA3 and CA1 subfields. It is possible that the blockage of endogenous TGF-β1 production can in some way sensitize CA1 neurons to the very low concentrations of KA used in this study. The blockage of TGF-β1 pathway inhibits the protection observed after the LPS injection. It is also of interest to note that the injection of the TGF-β1 RII antibody alone 24 hours after the LPS injection induces a small lesion. These two results confirm the neuroprotective role of TGF-β1 during an excitotoxic injury.

The expression of IL-1β, IL-10, and TNF-α is known to be rapidly inducible by experimental and clinical insults to the central nervous system (CNS), and it is clear that IL-1β and TNF-α exacerbate neurodegeneration or impaired neuronal function (Allan and Rothwell, 2001; Rothwell, 1999). In rodents, IL1-β mRNA cerebral expression is increased within 15 to 30 minutes, and protein is increased within 1 hour, of experimental cerebral ischemia (stroke), brain injury, or infusions of excitotoxins (Minami et al., 1992; Yabuuchi et al., 1994), all of which lead to neuronal death. Injection of recombinant IL-1β protein into the brain parenchyma of the rat exacerbates neuronal damage induced by cerebral ischemia, traumatic injuries, or excitotoxins (Lawrence et al., 1998; Loddick and Rothwell, 1996). TNF-α, like IL-1β, is also known to exacerbate neuronal death when delivered after cerebral ischemia (Barone et al., 1997) or traumatic brain injury (Arvin et al., 1996). In the current study, we also detected a small quantity of the antiinflammatory cytokine IL-10, which might participate in the neuroprotective effects. A drastic reduction of the infarct size 24 hours after middle cerebral artery occlusion has been reported after central administration of this cytokine (Spera et al., 1998), and we cannot exclude the participation of IL-10 in the neuroprotection observed after delivery of LPS. IL-10 might explain the neuroprotection observed by the injection of the mixture of LPS and KA before the TGF-β1 effect.

In contrast to the effect of IL-1β, TGF-β1 has been described as acting as a neuroprotectant in in vitro models of cerebral ischemia. Prehn et al. describe the neuroprotective effect of TGF-β1 on rat cortical neurons against neuronal degeneration caused by hypoxia (Prehn et al., 1993a). The same group observed neuroprotection mediated by an i.c. injection of TGF-β1 protein in vivo after transient forebrain ischemia (Henrich-Noack et al., 1996; Prehn et al., 1993b), although the observed neuroprotection was less significant than in the present study. TGF-β1 has also been shown to reduce infarct size in mice (Prehn et al., 1993b) and rabbits (Gross et al., 1993) subjected to focal cerebral ischemia. The use of the extracellular domain of the TGF-β1 type II receptor as a soluble TGF-β1 antagonist in a model of excitotoxic or ischemic reverses the neuroprotective effect of TGF-β1 (Ruocco et al., 1999).

TGF-β1 is likely to antagonize the effects of IL-1β through interactions with its transduction pathway. TGF-β1 exerts its biologic effects through a heteromeric complex of two receptor subunits, TGFβRI and TGFβRII. This complex in turn transduces the signal through activation of the SMAD family (Massague, 1998). However, TGF-β1 also uses other signal transduction pathways including MAPK (Allan and Rothwell, 2001), a pathway shared with IL-1β signaling. In vitro, IL-1β has been shown to be a potent inducer of TGF-β1 synthesis and release in vascular smooth muscle cells (da Cunha and Vitkovic, 1992; Yue et al., 1994). Furthermore, TGF-β1 has been shown to directly downregulate proinflammatory cytokine production such as IL-1β and TNF-α (Bogdan et al., 1992; Fadok et al., 1998); thus its presence before the excitotoxic insult may ensure a less aggressive acute inflammatory response.

TGF-β1 is also known to play a role in the regulation of the plasmin cascade through its upregulation of the serine protease inhibitor PAI-I (plasminogen activator inhibitor). PAI-1 has been shown in vitro in astrocyte culture to be overexpressed after TGF-β1 treatment (Buisson et al., 1998) and to mediate TGF-β1 associated neuroprotection. Indeed, the PAI-I inhibits tissue type plasminogen activator (t-PA), a serine protease that is involved in neuronal plasticity and cell death induced by excitotoxin. Mice deficient for t-PA are resistant to neuronal loss after the intrahippocampal injection of excitotoxins (Tsirka et al., 1997) or ischemic insult (Wang et al., 1998). TGF-β1 inhibits t-PA activity through expression of the inhibitor PAI-1 and thus prevents the degradation of the extracellular matrix (Docagne et al., 1999). In vitro, TGF-β1 can also acutely stabilize neuronal Ca²⁺ homeostasis under conditions of pathophysiologic Ca²⁺ overload, such as ischemia, and also produces a rapid neuronal induction of the survival gene Bcl2 (Prehn et al., 1994) that might suppress the detrimental effects of activated microglia on neighboring cells.

In conclusion, we have demonstrated that bacterial infection, mimicked by i.c. LPS injection, protects hippocampal neurons against an excitotoxic insult. This neuroprotection is strongly associated with the elevation of the anti-inflammatory TGF-β1. The in vivo injection of TGFβ1 protein or of an adenovirus expressing the active form of TGF-β1 afforded complete neuroprotection, whereas the blockage of TGF-β1 pathway by the injection of TGF-β1 RII antibody results in a dramatic enhancement of the area of the lesion. TGF-β1 is demonstrated in vivo as being an important neuronal survival factor. This protection may be through its downregulation of proinflammatory genes such as IL-1β and TNF-α or by its activation of PAI-1 and consequent inhibition of the plasmin cascade. TGF-β1 has been detected in necrotic human brain lesions (Ata et al., 1997) and also after ischemic stroke in human (Krupinski et al., 1996). These data suggest that TGF-β1 may play a role in antiinflammatory processes and in tissue remodeling after ischemic brain lesion and may be used for therapeutic purposes.

Footnotes

Abbreviations used:

References

Allan

Rothwell

(2001) Cytokines and acute neurodegeneration. Nat Rev Neurosci 2:734–744

Andersson

Perry

Gordon

(1991a) The kinetics and morphological characteristics of the macrophage-microglial response to kainic acid-induced neuronal degeneration. Neuroscience 42:201–214

Andersson

Perry

Gordon

(1991b) The CNS acute inflammatory response to excitotoxic neuronal cell death. Immunol Lett 30:117–182

Andersson

Perry

Gordon

(1992) The acute inflammatory response to lipopolysaccharide in CNS parenchyma differs from that in other body tissues. Neuroscience 48:169–186

Arvin

Neville

Barone

Feuerstein

(1996) The role of inflammation and cytokines in brain injury. Neurosci Biobehav Rev 20:445–452

Ata

Funa

Olsson

(1997) Expression of various TGF-beta isoforms and type I receptor in necrotizing human brain lesions. Acta Neuropathol (Berl) 93:326–333

Barone

White

Spera

Ellison

Currie

Wang

Feuerstein

(1998) Ischemic preconditioning and brain tolerance: Temporal histological and functional outcomes, protein synthesis requirement, and interleukin-1 receptor antagonist and early gene expression. Stroke 29:1937–1950; discussion 1950–1931

Barone

Arvin

White

Miller

Webb

Willette

Lysko

Feuerstein

(1997) Tumor necrosis factor-alpha. A mediator of focal ischemic brain injury. Stroke 28:1233–1244

Bett

Haddara

Prevec

Graham

(1994) An efficient and flexible system for construction of adenovirus vectors with insertions or deletions in early regions 1 and 3. Proc Natl Acad Sci U S A 91:8802–8806

10.

Bogdan

Paik

Vodovotz

Nathan

(1992) Contrasting mechanisms for suppression of macrophage cytokine release by transforming growth factor-beta and interleukin-10. J Biol Chem 267:23301–23308

11.

Buisson

Nicole

Docagne

Sartelet

Mackenzie

Vivien

(1998) Up-regulation of a serine protease inhibitor in astrocytes mediates the neuroprotective activity of transforming growth factor beta 1. Faseb J 12:1683–1691

12.

Cunningham

Boche

Perry

(2002) Transforming growth factor β1 (TGFβ1), the dominant cytokine in murine prion disease: Influence on inflammatory cytokine synthesis and alteration of vascular extracellular matrix. Neuropatol Appl Neurobiol 28:107–119

13.

da Cunha

Vitkovic

(1992) Transforming growth factor-beta 1 (TGF-beta 1) expression and regulation in rat cortical astrocytes. J Neuroimmunol 36:157–169

14.

Docagne

Nicole

Marti

MacKenzie

Buisson

Vivien

(1999) Transforming growth factor-beta1 as a regulator of the serpins/t-PA axis in cerebral ischemia. Faseb J 13:1315–1324

15.

Fadok

Bratton

Konowal

Freed

Westcott

Henson

(1998) Macrophages that have ingested apoptotic cells in vitro inhibit proinflammatory cytokine production through autocrine/paracrine mechanisms involving TGF-beta, PGE2, and PAF. J Clin Invest 101:890–898

16.

Gross

Bednar

Howard

Sporn

(1993) Transforming growth factor-beta 1 reduces infarct size after experimental cerebral ischemia in a rabbit model. Stroke 24:558–562

17.

Henrich-Noack

Prehn

Krieglstein

(1996) TGF-beta 1 protects hippocampal neurons against degeneration caused by transient global ischemia. Dose-response relationship and potential neuroprotective mechanisms. Stroke 27:1609–1614; discussion 1615

18.

Heurteaux

Lauritzen

Widmann

Lazdunski

(1995) Essential role of adenosine, adenosine A1 receptors, and ATP-sensitive K+ channels in cerebral ischemic preconditioning. Proc Natl Acad Sci U S A 92:4666–4670

19.

Kasischke

Ludolph

Riepe

(1996) NMDA-antagonists reverse increased hypoxic tolerance by preceding chemical hypoxia. Neurosci Lett 214:175–178

20.

Kawahara

Ruetzler

Klatzo

(1995) Protective effect of spreading depression against neuronal damage following cardiac arrest cerebral ischemia. Neurol Res 17:9–16

21.

Kirino

(2002) Ischemic tolerance. J Cereb Blood Flow Metab 22:1283–1296

22.

Kitagawa

Matsumoto

Kuwabara

Tagaya

Ohtsuki

Hata

Ueda

Handa

Kimura

Kamada

(1991) ‘Ischemic tolerance’ phenomenon detected in various brain regions. Brain Res 561:203–211

23.

Kitagawa

Matsumoto

Tagaya

Hata

Ueda

Niinobe

Handa

Fukunaga

Kimura

Mikoshiba

(1990) ‘Ischemic tolerance’ phenomenon found in the brain. Brain Res 528:21–24

24.

Kobayashi

Harris

Welsh

(1995) Spreading depression induces tolerance of cortical neurons to ischemia in rat brain. J Cereb Blood Flow Metab 15:721–727

25.

Krupinski

Kumar

Kaluza

(1996) Increased expression of TGF-beta 1 in brain tissue after ischemic stroke in humans. Stroke 27:852–857

26.

Lawrence

Allan

Rothwell

(1998) Interleukin-1beta and the interleukin-1 receptor antagonist act in the striatum to modify excitotoxic brain damage in the rat. Eur J Neurosci 10:1188–1195

27.

Lindholm

Castren

Kiefer

Zafra

Thoenen

(1992) Transforming growth factor-beta 1 in the rat brain: Increase after injury and inhibition of astrocyte proliferation. J Cell Biol 117:395–400

28.

Liu

Kato

Nakata

Kogure

(1992) Protection of rat hippocampus against ischemic neuronal damage by pretreatment with sublethal ischemia. Brain Res 586:121–124

29.

Loddick

Rothwell

(1996) Neuroprotective effects of human recombinant interleukin-1 receptor antagonist in focal cerebral ischemia in the rat. J Cereb Blood Flow Metab 16:932–940

30.

Logan

Frautschy

Gonzalez

Sporn

Baird

(1992) Enhanced expression of transforming growth factor beta 1 in the rat brain after a localized cerebral injury. Brain Res 587:216–225

31.

Massague

(1998) TGF-beta signal transduction. Annu Rev Biochem 67:753–791

32.

Massague

Cheifetz

Laiho

Ralph

Weis

Zentella

(1992) Transforming growth factor-beta. Cancer Surv 12:81–103

33.

Minami

Kuraishi

Yabuuchi

Yamazaki

Satoh

(1992) Induction of interleukin-1 beta mRNA in rat brain after transient forebrain ischemia. J Neurochem 58:390–392

34.

Morgan

Nichols

Pasinetti

Finch

(1993) TGF-beta 1 mRNA increases in macrophage/microglial cells of the hippocampus in response to deafferentation and kainic acid-induced neurodegeneration. Exp Neurol 120:291–301

35.

Murry

Jennings

Reimer

(1986) Preconditioning with ischemia: A delay of lethal cell injury in ischemic myocardium. Circulation 74:1124–1136

36.

Nishi

Taki

Uemura

Higashi

Kikuchi

Kudoh

Satoh

Nagata

(1993) Ischemic tolerance due to the induction of HSP70 in a rat ischemic recirculation model. Brain Res 615:281–288

37.

Perez-Pinzon

Dietrich

Rosenthal

Sick

(1997a) Rapid preconditioning protects rats against ischemic neuronal damage after 3 but not 7 days of reperfusion following global cerebral ischemia. J Cereb Blood Flow Metab 17:175–182

38.

Perez-Pinzon

Mumford

Dietrich

Rosenthal

Sick

(1997b) Rapid ischemic preconditioning protects rats from cerebral anoxia/ischemia. Adv Exp Med Biol 428:155–161

39.

Pohle

Rauca

(1994) Hypoxia protects against the neurotoxicity of kainic acid. Brain Res 644:297–304

40.

Prehn

Backhauss

Krieglstein

(1993b) Transforming growth factor-beta 1 prevents glutamate neurotoxicity in rat neocortical cultures and protects mouse neocortex from ischemic injury in vivo. J Cereb Blood Flow Metab 13:521–525

41.

Prehn

Peruche

Unsicker

Krieglstein

(1993a) Isoform-specific effects of transforming growth factors-beta on degeneration of primary neuronal cultures induced by cytotoxic hypoxia or glutamate. J Neurochem 60:1665–1672

42.

Prehn

Bindokas

Marcuccilli

Krajewski

Reed

Miller

(1994) Regulation of neuronal Bcl2 protein expression and calcium homeostasis by transforming growth factor type beta confers wide-ranging protection on rat hippocampal neurons. Proc Natl Acad Sci U S A 91:12599–12603

43.

Riepe

Esclaire

Kasischke

Schreiber

Nakase

Kempski

Ludolph

Dirnagl

Hugon

(1997) Increased hypoxic tolerance by chemical inhibition of oxidative phosphorylation: “chemical preconditioning”. J Cereb Blood Flow Metab 17:257–264

44.

Rothwell

(1999) Annual review prize lecture cytokines–killers in the brain? J Physiol 514:3–17

45.

Ruocco

Nicole

Docagne

Ali

Chazalviel

Komesli

Yablonsky

Roussel

MacKenzie

Vivien

Buissson

(1999) A transforming growth factor-β antagonist unmasks the neuroprotective role of this endogenous cytokine in excitotoxic and ischemic brain injury. J Cereb Blood Flow Metab 19:1345–1353

46.

Ruscetti

Palladino

(1991) Transforming growth factor-beta and the immune system. Prog Growth Factor Res 3:159–175

47.

Sime

Xing

Graham

Csaky

Gauldie

(1997) Adenovector-mediated gene transfer of active transforming growth factor-beta 1 induces prolonged severe fibrosis in rat lung. J Clin Invest 100:768–776

48.

Simon

Niiro

Gwinn

(1993) Prior ischemic stress protects against experimental stroke. Neurosci Lett 163:135–137

49.

Spera

Ellison

Feuerstein

Barone

(1998) IL-10 reduces rat brain injury following focal stroke. Neurosci Lett 251:189–192

50.

Sporn

Roberts

Wakefield

Assoian

(1986) Transforming growth factor-beta: Biological function and chemical structure. Science 233:532–534

51.

Tasaki

Ruetzler

Ohtsuki

Martin

Nawashiro

Hallenbeck

(1997) Lipopolysaccharide pre-treatment induces resistance against subsequent focal cerebral ischemic damage in spontaneously hypertensive rats. Brain Res 748:267–270

52.

Tsirka

Rogove

Bugge

Degen

Strickland

(1997) An extracellular proteolytic cascade promotes neuronal degeneration in the mouse hippocampus. J Neurosci 17:543–552

53.

Walsh

Betmouni

Perry

(2001) Absence of detectable IL-1beta production in murine prion disease: A model of chronic neurodegeneration. J Neuropathol Exp Neurol 60:173–182

54.

Wang

Yue

White

Barone

Feuerstein

(1995) Transforming growth factor-beta 1 exhibits delayed gene expression following focal cerebral ischemia. Brain Res Bull 36:607–609

55.

Wang

Tsirka

Strickland

Stieg

Soriano

Lipton

(1998) Tissue plasminogen activator (tPA) increases neuronal damage after focal cerebral ischemia in wild-type and tPA-deficient mice. Nat Med 4:228–231

56.

Wiessner

Gehrmann

Lindholm

Topper

Kreutzberg

Hossmann

(1993) Expression of transforming growth factor-beta 1 and interleukin-1 beta mRNA in rat brain following transient forebrain ischemia. Acta Neuropathol (Berl) 86:439–446

57.

Zhan

Fujihara

Taga

Shimoji

(2001) A forebrain ischemic preconditioning model established in C57Black/Crj6 mice. J Neurosci Methods 107:101–106

58.

Yabuuchi

Minami

Katsumata

Yamazaki

Satoh

(1994) An in situ hybridization study on interleukin-1 beta mRNA induced by transient forebrain ischemia in the rat brain. Brain Res Mol Brain Res 26:135–142

59.

Yue

Wang

Olson

Feuerstein

(1994) Interleukin-1 beta (IL-1 beta) induces transforming growth factor-beta, (TGF-beta 1) production by rat aortic smooth muscle cells. Biochem Biophys Res Commun 204:1186–1192

Transforming Growth Factor-β1-Mediated Neuroprotection against Excitotoxic Injury in Vivo

Abstract

Keywords

METHODS

Animals

Acute injury

TGF-β1 cytokine

Protein

Anti-human TGF-β1 RII antibody

Quantification of the lesion

Immunohistochemistry

Determination of IL-1β, IL-10, TNF-α and TGF-β1 in brain homogenates by ELISA

Statistics

Reagents

Adenovirus

LPS

ELISA

RESULTS

LPS injection

Kainic acid injection

LPS/Kainic acid

LPS-evoked cytokine profile

Adenovirus TGF-β1

TGF-β1 cytokine

DISCUSSION

Footnotes

Abbreviations used:

References