The Transcription Factor Interferon Regulatory Factor 1 Is Expressed after Cerebral Ischemia and Contributes to Ischemic Brain Injury

All animal procedures were approved by the Animal Care Committee of the University of Minnesota. Mice with a targeted mutation in the IRF-1 gene (homozygous [−/−] mice and their heterozygous [+/−] littermates) were originally obtained from Dr. T.W. Mak (Amgen Institute, Toronto, Canada) and had been back-crossed to C57BL/6 mice three to five times. Mice were subsequently bred in a virus antibody-free facility at the Uniformed Services University of the Health Sciences. The IRF-1^−/− colony was maintained by mating IRF-1^−/− mice to either IRF-1^−/− or IRF-1^+/− mice. An IRF-1 wild-type (+/+) colony was maintained by mating IRF-1^+/+ mice that were derived from heterozygous mating of either IRF-1^+/+ or IRF-1^+/− mice. To prevent the background of the IRF-1^−/− and IRF-1^+/+ colonies from straying, all new breeding pairs were the progeny of IRF-1^+/− to IRF-1^+/− matings. The genotype of all IRF-1 mice was determined by PCR using primers and methods described previously (37). The IRF-1 primers amplify a 300-bp sequence from IRF-1^+/+ and IRF-1^+/− genomic DNA, whereas neo primers amplify a product of 525 bp, present only in IRF-1^+/− and IRF-1^−/− mice. C57BL/6 mice were used as controls in some experiments and were obtained from The Jackson Laboratory.

Focal cerebral ischemia was produced by occlusion of the MCA as previously described (20). Mice were anesthetized with 2% halothane in 100% oxygen. Body temperature was maintained at 37 ± 0.5°C by a thermostatically controlled infrared lamp. A 2-mm hole was drilled in the inferior portion of the temporal bone to expose the left MCA. The MCA was elevated and cauterized distal to the origin of the lenticulostriate branches. Mice in which the MCA was exposed but not occluded served as sham-operated controls. Wounds were sutured and mice were allowed to recover and returned to their cages. Rectal temperature was controlled until mice regained full consciousness. Mice were killed at different time points after MCA occlusion for determination of IRF-1 mRNA or measurement of infarct size (see below).

C57BL/6 and IRF-1^+/+, IRF-1^+/−, and IRF-1^−/− mice were killed various times after MCA occlusion (n = 6–12/time point). As described in detail elsewhere (20), a 2-mm-thick coronal brain slice was cut at the level of the optic chiasm and the infarcted cortex was dissected using the corpus callosum as a ventral landmark. The homotopic region of the contralateral cortex was also sampled. Total RNA was extracted, and the integrity of the RNA was determined on denaturing formaldehyde gels. IRF-1 and hypoxanthine-guanine phosphoribosyl transferase (HPRT) mRNA were detected by reverse transcription PCR as described in detail previously (37, 38). The primer and probe sequences for IRF-1 and HPRT have been published previously (37–39). Cycle numbers for IRF-1 and HPRT were 29 and 23, respectively. The IRF-1 primers amplify a 148-bp fragment from cDNA. After agarose gel electrophoresis, amplified products were transferred to Hybond N+ membranes (Amersham) by standard Southern blotting techniques. Blots were hybridized with an internal oligonucleotide probe. Labeling of the probe and subsequent detection was carried out with the enhanced chemiluminescence system (ECL; Amersham) according to the manufacturer's instructions. Chemiluminescent signals were quantified using a scanner and imaging software (NIH Image; National Institutes of Health). The relative gene expression in test samples was assessed by linear regression analysis of a standard curve generated from a cDNA sample known to be positive for the gene of interest. Data were normalized to the housekeeping gene HPRT.

Mice were killed 4 d after MCA occlusion. Brains were removed and frozen in cooled isopentane (−30°C). Coronal forebrain sections (30-μm-thick) were serially cut in a cryostat, collected at 150-μm intervals, and stained with thionin for determination of infarct volume by an image analyzer (MCID, Imaging Research Inc.; reference 20). To factor out the contribution of ischemic edema to the total volume of the lesion, infarct volume in cerebral cortex was corrected for swelling (see reference 20 for details). The correction method is based on the determination of ischemic swelling by comparing the volume of ischemic and nonischemic hemispheres (40).

Neurological deficits were assessed by a neurological scoring system widely used in mice (for examples see references 41, 42) that has been described previously (20). The examiner was not aware of the identity of the mice. The neurological scores were as follows: 0, normal motor function; 1, flexion of torso and contralateral forelimb upon lifting the mouse by the tail; 2, circling to the contralateral side when holding the mouse by the tail on a flat surface, but normal posture at rest; 3, leaning to the contralateral side at rest; 4, no spontaneous motor activity (20). Mice were evaluated before MCA occlusion and at 24-h intervals up to 4 d after MCA occlusion.

Techniques used for monitoring cerebral blood flow (CBF) in mice have been described in detail previously (20). Mice were anesthetized with halothane (maintenance 1%) and the femoral artery and trachea were cannulated. Mice were artificially ventilated with an oxygen-nitrogen mixture by a mechanical ventilator (SAR-830; CWI Inc.). The inspiration time was set at 0.1 s, the respiratory rate at 120/min, and the inspiratory flow at ∼250 ml/min. The oxygen concentration in the mixture was adjusted to maintain arterial pO₂ between 150 and 170 mmHg. End-tidal CO₂ was continuously monitored using a CO₂ analyzer (Capstar-100; CWI Inc.) and maintained at 2.6–2.7%, which corresponds to a pCO₂ of 33–35 mmHg (20).

For monitoring changes in CBF produced by MCA occlusion, two laser-Doppler flow probes (Vasamedic) were placed through burr holes placed in the center (3.5 mm lateral to the midline and 1 caudal to bregma) and the periphery (1.5 mm lateral to the midline and 1.7 mm rostral to lambda) of the ischemic territory (20, 43). The location of the probe was selected in preliminary experiments to correspond to the region of brain that is spared from infarction in the IRF-1–deficient mice. After placement of the probes, the MCA was occluded and CBF was monitored for 90 min. CBF data are expressed as percentage of the preocclusion value.

Techniques for testing cerebral vascular reactivity to hypercapnia in mice were similar to those previously described (20). Mice were anesthetized and instrumented as described above. CBF was continuously monitored in the frontoparietal cortex with a laser-Doppler probe. After stabilization of arterial pressure and blood gases, CO₂ was introduced into the circuit of the ventilator and the increase in CBF produced by hypercapnia was monitored. CO₂ administration was discontinued after the CBF increase reached a plateau (usually 2–3 min).

Data in text and figures are expressed as means ± SEM. Multiple comparisons were evaluated statistically by the analysis of variance and Tukey's test. Two-group comparisons were analyzed by the two-tailed Student's t test for independent samples. Neurological scores were analyzed by the Kruskal-Wallis analysis of variance followed by the Tukey-Kramer test (Systat) (20). For all procedures, probability values of <0.05 were considered statistically significant.

We first sought to determine if focal cerebral ischemia enhances IRF-1 mRNA expression. In C57BL/6 mice, MCA occlusion was associated with pronounced upregulation of IRF-1 mRNA in the postischemic brain (Fig. 1). IRF-1 mRNA expression was increased by 12 h and remained elevated 1–7 d after MCA occlusion (P < 0.05 from sham-operated mice; analysis of variance and Tukey's test). IRF-1 mRNA expression did not increase in the contralateral (nonischemic) cortex (P > 0.05) (Fig. 1). IRF-1 mRNA expression was reduced in IRF-1^+/− mice and absent in IRF-1^−/− mice (Fig. 2).

In these experiments we used mice with a null mutation of IRF-1 to determine whether IRF-1 contributes to cerebral ischemic injury. In C57BL/6 mice (n = 6), MCA occlusion produced reproducible infarcts involving mainly the cerebral cortex (Figs. 3 and 4). Size and regional distribution of the infarct were comparable to those previously reported in mice from this and other laboratories (20, 41). The volume of the infarct in IRF-1^+/+ mice did not differ from that of C57BL/6 mice (P > 0.05; Fig. 4 A). Infarct volumes in IRF-1 knockout mice were smaller than those observed in C57BL/6 mice (Figs. 3 and 4). The reduction was less pronounced in IRF-1^+/− (23 ± 3%; P < 0.05; n = 5), than in IRF-1^−/− mice (46 ± 9%; P < 0.05; n = 6), and involved the infarct border throughout the entire rostrocaudal extent of the ischemic lesion (Figs. 3 and 4). The volume of postischemic brain swelling did not differ between C57BL/6 (6.6 ± 0.6 mm³) and IRF-1^+/− mice (5.7 ± 0.8; P > 0.05), but was significantly reduced in IRF-1^−/− mice (3.5 ± 0.8; P < 0.05).

To determine whether the reduction in infarct volume was associated with a better functional outcome, the neurological deficits produced by MCA occlusion were scored in C57BL/6 and IRF-1^−/− mice. 24 h after MCA occlusion, neurological deficits did not differ between C57BL/6, IRF-1^+/−, and IRF-1^−/− mice (Fig. 5). However, although the deficits in C57BL/6 remained stable throughout the monitoring period, the deficit in IRF-1^−/− mice improved over time (Fig. 5). By 4 d after MCA occlusion, the motor impairment was less pronounced in the IRF-1^+/− and IRF-1^−/− mice than in C57BL/6 (P < 0.05), the improvement being more marked in IRF^−/− than in IRF-1^+/− mice (Fig. 5). These data indicate that the functional outcome of cerebral ischemia is improved in IRF-1^+/− and IRF-1^−/− mice.

In these studies we investigated whether the reduction in CBF produced by MCA occlusion was comparable in IRF-1^−/− mice and controls. As illustrated in Fig. 6 A, the reduction in CBF produced by MCA occlusion in IRF-1^−/− mice was not different from that observed in C57BL/6 (P > 0.05; n = 5–8/ group). Similarly, systemic arterial pressure did not differ between IRF-1^−/− mice and controls (Fig. 6 B). We then studied the reactivity of the cerebral circulation to systemic hypercapnia in IRF-1^−/− mice. The increase in CBF produced by hypercapnia (arterial pC0₂ = 50–60 mmHg) was 73 ± 12% of control in C57BL/6 and 61 ± 8% in IRF-1^−/− mice (P > 0.05; n = 5/group). Thus, the degree of CBF reduction produced by MCA occlusion and the reactivity of the cerebral circulation to hypercapnia are not significantly altered in IRF-1^−/− mice.

We sought to determine whether IRF-1, a transcription factor involved in tissue responses to inflammation, cell proliferation, and programmed cell death, is involved in the mechanisms of cerebral ischemic injury. We found that IRF-1 mRNA expression is selectively upregulated in the ischemic brain after occlusion of the mouse MCA. The upregulation is observed within 12 h of induction of ischemia and is still present at 7 d. To determine whether the IRF-1 mRNA expression contributes to ischemic brain damage, the infarct volume produced by MCA occlusion was assessed in mice with a null mutation of the IRF-1 gene. We found that the volume of brain injury produced by MCA occlusion is markedly reduced in mice with a null mutation of the IRF-1 gene. The reduction is more pronounced in IRF-1^−/− than in IRF-1^+/− mice and, importantly, is associated with an improvement of the neurological deficits produced by the ischemic insult. The findings demonstrate that IRF-1 is upregulated after induction of cerebral ischemia and that such upregulation plays a critical role in the mechanisms of ischemic brain injury.

Several transcription factors have been shown to be activated after cerebral ischemia. For example, activator protein 1, factors encoded by early genes, Stat3, hypoxia inducible factor 1, and nuclear factor κb are upregulated after focal or global cerebral ischemia (29–32). However, direct evidence for an involvement of these transcription factors in the mechanisms of ischemic brain injury is lacking (for example see reference 33). In the present study, we have demonstrated that deletion of IRF-1 reduces the magnitude of cerebral ischemic damage and improves neurological recovery. Therefore, IRF-1 is the first transcription factor that has been shown to be linked directly to the development of ischemic brain injury.

Like most mice generated by homologous recombination, the genetic background of IRF-1^−/− is heterogeneous, including both C57BL/6 and SV129 genes (44). However, such genetic heterogeneity is unlikely to contribute to the observed reduction in infarct volume. This conclusion is supported by the following observations. First, the infarct volume of IRF-1^+/+ mice, bred from the same heterozygotic stock from which the homozygous knockouts were derived, was not different from that of inbred wild-type C57BL/6 controls. Second, a clear gene dosage effect is observed when wild-type, IRF-1^+/−, and IRF-1^−/− mice are compared for IRF-1 mRNA or infarct volume. The reduction in infarct volume cannot be attributed to differences in the degree of cerebral ischemia between C57BL/6 and IRF-1^−/− mice: MCA occlusion produced comparable decreases in CBF in wild-type and IRF-1^−/− mice. Furthermore, arterial pressure and the reactivity of the cerebral circulation to the vasodilation produced by hypercapnia were not altered in IRF-1^−/− mice. Therefore, genetic factors affecting the susceptibility of the brain to cerebral ischemia or vascular factors influencing the degree of ischemia cannot be responsible for the protection observed in IRF-1^−/− mice.

After induction of cerebral ischemia, the temporal evolution of the ensuing brain damage is heterogeneous. In the center of the ischemic territory, where the ischemic insult is most severe, irreversible neuronal damage occurs rapidly (45, 46). Pathogenic factors involved in the rapid phase of the damage include energy failure, glutamate excitotoxicity, and reactive oxygen species (47–49). In the peripheral regions of the ischemic territory, the damage progresses at a slower pace, reaching a maximum 2–3 d after induction of ischemia (46, 50). Factors contributing to the delayed evolution of the damage include postischemic inflammation and apoptosis (1, 51). The observation that mice lacking the IRF-1 gene are relatively protected from ischemic infarction suggests that IRF-1 is involved in the mechanisms of ischemic brain injury. However, the delayed time course of its expression suggests that this transcription factor is involved mainly in events occurring in the late stages of cerebral ischemia. The observation that neurological deficits in IRF-1^−/− mice are not different from those of wild-type mice 24 h after ischemia and improve 3–4 d later is also consistent with the hypothesis that IRF-1 contributes to the progression, rather than the initiation, of tissue damage.

IRF-1 deletion could attenuate the progression of ischemic brain injury by different mechanisms. One possibility is that IRF-1 deletion interferes with the inflammatory response associated with cerebral ischemia. Cerebral ischemia induces an inflammatory reaction in the injured brain (15– 17). Cytokines produced in the ischemic tissue are one of the factors that induce the expression of adhesion molecules in endothelial cells (9–12). Blood-borne neutrophils adhere to the cerebral endothelium, cross the blood vessel wall, and migrate into the brain parenchyma (13, 14, 17). Neutrophils contribute to ischemic brain injury because tissue damage is reduced if their invasion of the brain parenchyma is prevented (52–55; for review see reference 1). After cerebral ischemia, neutrophils express iNOS, an enzyme that produces toxic amounts of NO (18–20). Evidence suggests that NO produced by iNOS participates in ischemic brain injury. Administration of inhibitors of iNOS reduces ischemic damage, whereas iNOS^−/− mice have smaller infarcts and a better neurological outcome after MCA occlusion (19, 20, 56). Because IRF-1 plays a role in iNOS transcription (37, 57, 58), it is possible that iNOS is not expressed in IRF-1^−/− mice. Therefore, reduced iNOS expression could be one of the mechanisms by which IRF-1^−/− mice are protected from cerebral ischemia. On the other hand, the possibility that postischemic iNOS expression is driven by nuclear factor κB and hypoxia inducible factor 1, transcription factors that are also upregulated after cerebral ischemia (30, 32), cannot be ruled out.

Another potential mechanism for the reduction of ischemic injury in IRF-1^−/− mice is inhibition of programmed cell death. There is increasing evidence that programmed cell death occurs after focal cerebral ischemia (for review see reference 51). Genes involved in the apoptotic cascade are upregulated after cerebral ischemia (23–26), and internucleosomal DNA fragmentation, one of the hallmarks of apoptosis, occurs (59–62). Several lines of evidence suggest that apoptosis plays a pivotal role in the mechanisms of brain injury. Transgenic mice that overexpress the anti-apoptotic protein bcl2 or mice with a null mutation of p53, a protein that promotes apoptosis, are relatively protected from the effects of focal ischemia (63, 64). In addition, inhibition of the caspase family of cysteine proteases, one of the effectors of apoptotic cell death, reduces cerebral ischemic damage (65–67). There is evidence that IRF-1 is involved in the molecular mechanisms of apoptosis. For example, DNA damage-induced apoptosis is attenuated in T-lymphocytes and in macrophages lacking IRF-1 (68, 69). Therefore, it is conceivable that postischemic apoptotic neuronal death is attenuated in IRF-1^−/− mice resulting in less brain damage.

The mechanisms of IRF-1 expression after cerebral ischemia and its cellular localization have not been elucidated. In other systems, IRF-1 is induced by a wide variety of factors including IFNs and cytokines (34). Some of these factors, such as IL-1, IL-6 and TNF-α, are also expressed after cerebral ischemia (4–6) and could conceivably trigger IRF-1 mRNA expression. Furthermore, intracellular second messengers, such as calcium and protein kinase C, may also induce IRF-1 expression (70). These pathways are activated after cerebral ischemia (for examples see references 71, 72) and could contribute to trigger IRF-1 expression. Therefore, multiple mechanisms are likely to contribute to IRF-1 upregulation. Further studies will be required to address these issues.

In conclusion, we have demonstrated that the mRNA encoding for the transcription factor IRF-1 is markedly upregulated after focal cerebral ischemia. We have also shown that IRF-1^−/− mice are substantially protected from the brain damage and neurological deficits produced by MCA occlusion. The protection develops 2–3 d after induction of ischemia. The findings suggest that IRF-1–inducible genes contribute to the late stages of tissue damage after cerebral ischemia. Thus, IRF-1 is a critical transcription factor in the molecular mechanisms of ischemic brain injury and, as such, might be a new target for therapeutic interventions in ischemic stroke.

Supported by National Institutes of Health grants NS34179 and NS35806 (to C. Iadecola and M.E. Ross), by the Neuroscience Strategic Research Initiative, Academic Health Center, University of Minnesota (C. Iadecola and M.E. Ross), and by the Office of Naval Research (S.N. Vogel). C. Iadecola is an Established Investigator of the American Heart Association.

CBF

cerebral blood flow

HPRT

hypoxanthine-guanine phosphoribosyl transferase

IRF-1

interferon regulatory factor 1, iNOS, inducible nitric oxide synthase

MCA

middle cerebral artery

NO

nitric oxide