DNA methyltransferase contributes to delayed ischemic brain injury

Abstract

DNA methylation is important for controlling the profile of gene expression and is catalyzed by DNA methyltransferase (MTase), an enzyme that is abundant in brain. Because significant DNA damage and alterations in gene expression develop as a consequence of cerebral ischemia, we measured MTase activity in vitro and DNA methylation in vivo after mild focal brain ischemia. After 30 min middle cerebral artery occlusion (MCAo) and reperfusion, MTase catalytic activity and the 190 kDa band on immunoblot did not change over time. However, [(3)H]methyl-group incorporation into DNA increased significantly in wild-type mice after reperfusion, but not in mutant mice heterozygous for a DNA methyltransferase gene deletion (Dnmt(S/+)). Dnmt(S/+) mice were resistant to mild ischemic damage, suggesting that increased DNA methylation is associated with augmented brain injury after MCA occlusion. Consistent with this formulation, treatment with the MTase inhibitor 5-aza-2'-deoxycytidine and the deacetylation inhibitor trichostatin A conferred stroke protection in wild-type mice. In contrast to mild stroke, however, DNA methylation was not enhanced, and reduced dnmt gene expression was not protective in an ischemia model of excitotoxic/necrotic cell death. In conclusion, our results demonstrate that MTase activity contributes to poor tissue outcome after mild ischemic brain injury.

Fig. 1.

Time-dependent changes of MTase activity inDnmt^+/+ (A) andDnmt^S/+ mice (B) after 30 min middle cerebral artery occlusion and reperfusion. MTase activity was measured in ischemic cortex (COR) and striatum (STR) along with the respective contralateral tissue and sham (S)-operated animals. MTase activity was measured in nuclear extracts in vitro using a hemimethylated oligo-probe. Levels in cerebellum (CER) are shown for comparison. n = 2 animals per time point. Mean ± SE. p > 0.05.

Fig. 2.

A, MTase protein levels in olfactory bulb (OB), striatum (STR), and hippocampus (HIP) in brain lysates from normal (wt) and Dnmt^S/+ mice (s/+). Brain lysates were subjected to SDS-PAGE and immunoblot analysis performed using a polyclonal MTase antibody. The experiment was repeated three times; a representative experiment is shown. B, Time-dependent changes in MTase protein expression in striatum during reperfusion after 30 min middle cerebral artery occlusion. MTase protein was present in normal brain (sham) and did not change over time (0, 1, 3, 6, 12, 18, 24, 48, and 72 hr) between left (L = ischemic) and right (R = non-ischemic) hemispheres. The experiment was repeated three to four times per time point; a representative experiment for the time points 3, 6, and 18 hr is shown.

Fig. 3.

Methyl group incorporation into brain DNAin vivo in Dnmt^+/+(A) and Dnmt^S/+mice (B) after 30 min middle cerebral artery occlusion. On reperfusion,l-[methyl-³H]methionine was administered and [³H]methyl group incorporation was measured after 12 hr. DNA methylation was significantly increased in ischemic cortex (COR) and striatum (STR) compared with the contralateral side inDnmt^+/+ (A) but not Dnmt^S/+ mice (B). Levels in cerebellum (CER) are shown for comparison. n = 4 and 5 animals. Mean ± SE. *p < 0.05. Paired Student'st test.

Fig. 4.

A, Infarct volume was 29% smaller in Dnmt^S/+ mice compared withDnmt^+/+ mice after 30 min filamentous middle cerebral artery occlusion (MCAo) and 72 hr reperfusion. Brain lesion volume was determined on serial coronal hematoxylin and eosin-stained cryostat sections (20 μm).B, Dnmt^S/+ mice had significantly higher numbers of viable cells in ischemic striatum after 30 min MCAo/72 hr reperfusion compared with controls. Viable cells with neuronal appearance were counted on coronal sections (20 μm) through the anterior commissure. Mean ± SE. n = 14 and 15 animals per group. *p < 0.05 compared withDnmt^+/+ mice. Student'st test.

Fig. 5.

Treatment with the MTase inhibitor 5-aza-2′-deoxycitide (5-aza-dC; 20 μg) or the specific deacetylation inhibitor trichostatin A (TSA; 1 or 10 μg) reduced lesion size after 30 min of filamentous middle cerebral artery occlusion and 72 hr of reperfusion compared with vehicle in 129/V mice. Drugs or vehicle was administered intracerebroventricularly 10 min before ischemia onset. Lesion volume was determined quantitatively. Data are presented as mean ± SE.n = 5–9 animals per group. *p< 0.01 versus vehicle. Student's t test.

Fig. 6.

A, Methyl group incorporation into brain DNA in vivo after 2 hr versus 30 min MCAo in 129/SV mice. On reperfusionl-[methyl-³H]methionine was administered, and [³H]methyl group incorporation into brain DNA was measured after 12 hr reperfusion. Unlike after 30 min MCAo (also see Fig. 3), DNA methylation did not increase in ischemic cortical tissue from the MCA territory compared with the contralateral side after 2 hr MCA occlusion. n = 5 (2 hr MCAo) and 4 (30 min MCAo) animals. Mean ± SE. *p < 0.05. Paired Student's t test. B, WhenDnmt^+/+ andDnmt^S/+ mice were subjected to 2 hr MCAo/24 hr reperfusion and infarct volume was quantitated, no difference in indirect infarct volume was noted between groups. This result was confirmed by a direct method to measure infarct volume (137 ± 19.6 vs 121.7 ± 11.5 mm³ forDnmt^S/+ vsDnmt^+/+ mice, respectively). Mean ± SE. n = 4 and 5. p> 0.05 compared with wild type. Student's ttest.