Elevated global SUMOylation in Ubc9 transgenic mice protects their brains against focal cerebral ischemic damage

Abstract

We have previously shown that a massive increase in global SUMOylation occurs during torpor in ground squirrels, and that overexpression of Ubc9 and/or SUMO-1 in cell lines and cortical neurons protects against oxygen and glucose deprivation. To examine whether increased global SUMOylation protects against ischemic brain damage, we have generated transgenic mice in which Ubc9 is expressed strongly in all tissues under the chicken β-actin promoter. Ubc9 expression levels in 10 founder lines ranged from 2 to 30 times the endogenous level, and lines that expressed Ubc9 at modestly increased levels showed robust resistance to brain ischemia compared to wild type mice. The infarction size was inversely correlated with the Ubc9 expression levels for up to five times the endogenous level. Although further increases showed no additional benefit, the Ubc9 expression level was highly correlated with global SUMO-1 conjugation levels (and SUMO-2,3 levels to a lesser extent) up to a five-fold Ubc9 increase. Most importantly, there were striking reciprocal relationships between SUMO-1 (and SUMO-2,3) conjugation levels and cerebral infarction volumes among all tested animals, suggesting that the limit in cytoprotection by global SUMOylation remains undefined. These results support efforts to further augment global protein SUMOylation in brain ischemia.

Figure 1. Ubc9 expression levels in WT and Ubc9 transgenic mice.

(A) Ubc9 message levels in tail sample from WT and Ubc9 transgenic founder lines (first generation) were measured in triplicate by real-time PCR. The average Ubc9 mRNA level in each animal was normalized by its tubulin mRNA level and plotted relative to the WT average level. (B) Ubc9 protein levels in tail samples. Protein (20 µg) isolated from tail samples from WT and Ubc9 transgenic mice was analyzed by Western blot (upper panel). Intensities of Ubc9 protein bands were normalized to actin and plotted relative to WT average level (lower panel). n = 2–10 (mean±SD). *P<0.05, **P<0.01 (note: G3 n = 2 with large SD). (C) Ubc9 protein levels in brain samples. Proteins (20 µg each) isolated from brains of WT and transgenic mice were analyzed by Western blot (upper panel). Intensities of Ubc9 protein bands were normalized to actin and plotted relative to WT average level (lower panel). n = 3–7 (mean±SD). *P<0.05, **P<0.01. (D) Ubc9 expression levels in brain tissue (cerebral cortex, Layer III-external pyramidal region) of WT and Ubc9 transgenic mice (H3, G3). Ubc9 was shown in red. (E) Ubc9 expressing cells in the cerebral cortex of Ubc9 transgenic mouse (line H3). Upper panels show all layers of cerebral cortex, and lower panels show an enlarged region of the Layer III-External pyramidal area (dashed rectangle in the upper merged panel). NeuN (green), a pan-neuronal marker; parvalbumin (purple), a marker for interneurons; GFAP (yellow), a marker for astrocytes; CNPase (blue), a marker for oligodendrocytes. Ubc9 is shown in red.

Figure 2. Ubc9 transgenic mice are more tolerant to ischemia than WT mice.

(A) Examples of cresyl violet stained coronal brain sections from WT and Ubc9 transgenic mice (N2, H3, K5 and G3) after 24 h pMCAO. The inset shows an enlarged section with a clear infarcted area for a precise measurement. (B) Scatter plot of brain infarction volumes in WT (n = 12) and Ubc9 transgenic mice, N2 (n = 8), H3 (n = 10), K5 (n = 8) and G3 (n = 4) after 24 h pMCAO. Average infarction volume in each animal group is shown as a short horizontal line. *P<0.05, **P<0.01, ***P<0.001. (C) The brain infarction volumes (y-axis) after 24 h pMCAO are plotted against their Ubc9 protein levels (x-axis). A correlation is seen among animals whose Ubc9 levels are under 5-fold of wild type. (Pearson correlation coefficient) r = −0.713, p<0.0001.

Figure 3. SUMO-1 and SUMO-2,3 conjugation levels in brains of WT and Ubc9 transgenic mice with and without MCAO surgery.

(A) Top panel: A representative immunoblot of high molecular weight (HMW) (100∼300 kDa) conjugates of SUMO-1 and SUMO-2,3 along with Ubc9 and actin in brain samples from WT and Ubc9 transgenic (N2, H3, N3, K5, G3) mice without MCAO surgery. Bottom panel: Quantitative analysis of Ubc9, SUMO-1 HMW-conjugates and SUMO-2,3 HMW-conjugates. Intensities were measured, normalized to actin in each sample, and plotted relative to WT average level. n = 3–5 per group (mean±SD; except N3, n = 1). *P<0.05, **P<0.01. (B) Comparable immunoblot (top panel) and quantitative analysis (bottom panel) of Ubc9, SUMO-1 and SUMO-2,3 HMW-conjugates from mice after MCAO surgery. Data are shown as the mean with standard deviation of 4–10 samples in each animal group. *P<0.05, **P<0.01. (C) Immunofluorescence analysis of SUMO-1 and SUMO-2,3 protein distributions in infarcted brains. Top row, H&E staining of coronal brain sections from WT and Ubc9 transgenic (H3) mice after 24 h pMCAO. CX, cerebral cortex; CC, corpus callosum; LV, lateral ventricle; ST, striatum; TH, thalamus; HT, hypothalamus. The second row from the top, comparable coronal brain sections from WT and H3 mice immunostained for SUMO-1 (red), SUMO-2,3 (yellow), NeuN (green) and parvalbumin (purple). con, contralateral side; ips, ipsilateral side. Bottom three sets of paired panels show comparable regions of the layer III-external pyramidal area (enlarged from dashed areas in the upper panels).

Figure 4. Correlation of SUMO-1 HMW-conjugates or SUMO-2,3 HMW- conjugates with either Ubc9 level or infarction volume.

(A) The brain SUMO-1 HMW-conjugates level (y-axis) after 24 h pMCAO was plotted against the Ubc9 level (x-axis) in the same sample. Animals whose Ubc9 levels were under 5-fold the wild type show a significant correlation. (n = 42; r: Pearson correlation coefficient = 0.713, p<0.0001). (B) The brain SUMO-2,3 HMW-conjugates level (y-axis) was plotted against the Ubc9 level (x-axis) in the same sample. Animals whose Ubc9 levels were under 5-fold the wild type show only a trend toward correlation. (n = 42; r = 0.321, p = 0.068). (C) The brain infarction volumes (y-axis) after 24 h pMCAO were plotted against their SUMO-1 HMW-conjugates levels (x-axis). (n = 42; r = −0.682, p<0.0001). (D) The brain infarction volumes (y-axis) after 24 h pMCAO were plotted against their SUMO-2,3 HMW-conjugate levels. (n = 42; r = −0.426, p = 0.0133).