SUMO2/3 is Associated with Ubiquitinated Protein Aggregates in the Mouse Neocortex after Middle Cerebral Artery Occlusion

Abstract

Protein modifications cooperatively act to protect the proteome from cellular stress. Focal cerebral ischemia increases protein ubiquitination, resulting in formation of ubiquitin-rich aggregates. A concurrent elevation in small ubiquitin-related modifier (SUMO)-conjugated proteins has also been reported, but a potential connection to ubiquitin remains unexplored. Here we show that SUMO2/3 conjugates are present in postischemic ubiquitin-rich aggregates, physically associated with ubiquitin. The coaggregation of SUMO2/3 and ubiquitin is induced rapidly after ischemia, depends on reperfusion, and is also observed in the absence of ischemic damage. The association between SUMO and ubiquitin suggests overlapping functional roles after ischemia/reperfusion.

Introduction

Ischemic stroke caused by occlusion of major cerebral arteries produces rapid neuronal death in the center of the ischemic territory, while cell death is delayed in peripheral regions, the so called ischemic penumbra. ¹ The slow evolution of the penumbra toward irreversible damage renders the tissue potentially salvageable, and elucidating the factors that influence the survival of penumbral tissue is of great importance for the development of new stroke therapies. One such factor might be the modification of proteins with ubiquitin and small ubiquitin-related modifier (SUMO), which is dramatically increased in the penumbral area after transient focal cerebral ischemia and not in the ischemic core.^{2, 3} Ubiquitin and SUMO attachment to proteins has the potential to influence protein stability, interaction and localization, and may therefore represent an effective way to rapidly adapt the cellular proteome to the changing, stressful environment after stroke. Indeed, elevated SUMO conjugation has been associated with neuroprotection,^{4, 5, 6} while the role of increased ubiquitin accumulation in stroke outcome has not been clarified.

Ubiquitin and SUMO were long thought to have separate targets and functions, but increasing evidence suggests that a crosstalk and interdependence exists between these two modifications.^{7, 8} Along those lines, the temporal profile and conjugation pattern of SUMO and ubiquitin after focal ischemia were found to be strikingly similar,^{2, 3} which warrants further examination of a potential connection between the two modifications in this setting. Ubiquitin is enriched in protein aggregates after focal ischemia,^{2, 9} however, it remains unclear if also SUMO is present within such inclusions. Therefore, the aim of the current study was to investigate whether SUMO is associated with postischemic ubiquitin aggregates in a mouse model of middle cerebral artery occlusion.

Materials and Methods

Animals

Experiments were performed in 6- to 8-week-old male C57Bl/6J mice (mean weight 22.8 g; Jackson Laboratory, Bar Harbor, ME, USA). All experimental procedures were approved by the IACUC (Institutional Animal Care and Use Committee) of Weill Cornell Medical College and were performed according to the IACUC, NIH, and ARRIVE guidelines (http://www.nc3rs.org/ARRIVE).

Middle Cerebral Artery Occlusion

Middle cerebral artery occlusion was induced by the intraluminal suture method as previously reported. ² Briefly, mice were anesthetized with isoflurane (1.5% to 2%) and a 6-0 surgical nylon monofilament was introduced into the right internal carotid artery through the external carotid artery to occlude the origin of the middle cerebral artery. For transient ischemia, the filament was left in place for the indicated time points and then withdrawn. The filament was not retracted in mice subjected to permanent ischemia. Cerebral blood flow was measured in the ischemic hemisphere using transcranial laser-Doppler flowmetry (Periflux System 5010, Perimed, PA, USA). Only animals with 85% reduction in cerebral blood flow during middle cerebral artery occlusion and 80% cerebral blood flow recovery after 10 minutes of reperfusion were included in the study. In all mice, rectal temperature was kept at 37.0°C±0.5°C during surgery and in the recovery period until the animals regained consciousness. Out of a total of 76 animals used in this study 10 had to be excluded because they either did not meet cerebral blood flow criteria (n=8) or died prematurely (n=2) during surgery.

Isolation of Ubiquitin/SUMO-Containing Soluble and Aggregate Fractions and Detection of Proteins by Western Blotting

Soluble and aggregate fractions were obtained from the ipsilateral neocortex. The cortical sampling and isolation procedure is described in detail in Hochrainer et al. ² Briefly, cytosolic soluble proteins were extracted by mechanical lysis in buffer without detergent (15 mmol/l Tris-HCl, pH 7.6; 250 mmol/l sucrose, 1 mmol/l MgCl₂, 2.5 mmol/l ethylenediaminetetraacetic acid, 1 mmol/l ethylene glycol tetracetic acid, 1 mmol/l Na₃VO₄, 5 mmol/l NaF, 20 mmol/l phenylphosphate, 20 mmol/l N-ethylmaleimide, 1 mmol/l dithiothreitol, and 1x protease inhibitors (Roche Applied Sciences, Indianapolis, IN, USA)), followed by high-speed centrifugation. Residual pellets containing insoluble cytoplasmic material as well as intact nuclei were sonicated and incubated in the same buffer substituted with 2% Triton X-100 and 150 mmol/l KCl. Aggregates, defined by insolubility under these conditions, were obtained by centrifugation.

Equal amounts of proteins as determined by protein concentration measurement of the soluble fraction were loaded and separated by 4% to 20% (Life Technologies, Grand Island, NY, USA) or 7% Tris-glycine-sodium dodecyl sulfate (SDS) gel electrophoresis. After transfer of proteins to polyvinylidene difluoride membranes (EMD Millipore, Billerica, MA, USA), total ubiquitin, lysine 48-linked ubiquitin, lysine 63-linked ubiquitin, SUMO1, and SUMO2/3 were detected by western blotting with anti-ubiquitin (clone ubi-1, Life Technologies), lysine 48-specific anti-ubiquitin (clone Apu2; EMD Millipore), lysine 63-specific anti-ubiquitin (clone D7A11, Cell Signaling Technology, Danvers, MA, USA), anti-SUMO1, and anti-SUMO2/3 (both Cell Signaling Technology) antibodies, respectively. Protein bands from free ubiquitin/SUMO as well as protein band smear in the higher molecular weight range representing conjugated ubiquitin/SUMO (indicated by brackets in the figures) were quantified with Kodak 1D V3.6 image software (Kodak, Rochester, NY, USA). Densitometric data were obtained in the linear range of blot exposure. The background optical density of each blot was determined in an empty lane and the obtained value was subtracted from each specific signal. Ubiquitin and SUMO levels in experimental groups were expressed relative to levels in sham animals, which were set to 1. Free ubiquitin and SUMO were normalized to glyceraldehyde 3-phosphate dehydrogenase levels. Data are expressed as mean±s.d.

Coimmunoprecipitation of Ubiquitin and SUMO from Aggregate Fractions

Aggregate fractions were solubilized as specified previously in Shimada et al. ¹⁰ Briefly, pellets were denaturated in 2% SDS followed by renaturation through removal of SDS and addition of 0.6% alpha-cyclodextrin/0.6% Tween-20. Immunoprecipitation was performed with anti-ubiquitin antibody (clone FK2; EMD Millipore), anti-SUMO2/3 antibody (clone 8A2; Abcam, Cambridge, MA, USA), or anti-mouse monoclonal IgG₁ isotype control antibody (Cell Signaling Technology) bound to protein-G sepharose (GE Healthcare Life Sciences, Pittsburgh, PA, USA). Precipitates were loaded on 7% Tris-glycine SDS polyacrylamide gels and western blot procedures were executed as described above.

Statistical Analysis

Sample size was determined based on previous published work on ubiquitination in ischemia by our laboratory. ² Animals were randomly assigned to sham and stroke groups and analysis was performed in a blinded manner. Two group comparisons were statistically analyzed by the two-tailed Student's t-test. Multiple comparisons were performed by one-way ANOVA and Bonferroni post hoc test. Differences were considered significant for P<0.05.

Results

Transient Middle Cerebral Artery Occlusion Leads to Coaggregation of Ubiquitin and SUMO2/3 in the Ipsilateral Neocortex

To address whether SUMO conjugates found after ischemia are of soluble or, like ubiquitin conjugates, of insoluble nature we compared their levels in soluble cytosolic and Triton X-100 insoluble aggregate fractions from mouse neocortices after transient focal ischemia. SUMO2/3 conjugates were, like ubiquitin, significantly enriched in the aggregate fraction as opposed to its soluble counterpart (Figure 1A). At the same time, conjugated as well as free ubiquitin was depleted from the soluble fraction, while SUMO2/3 levels remained constant (Figure 1A). No significant changes were observed in either fraction for free and conjugated SUMO1 (Figure 1A). To evaluate the nature of ubiquitin conjugates after focal cerebral ischemia we probed blots with lysine 48- and lysine 63-specific antibodies. Interestingly, detection with both antibodies resulted in a strong signal (Figure 1B), indicating that aggregates contain ubiquitin chains linked via both lysine 48 and 63.

Figure 1.

SUMO2/3, but not SUMO1, coaggregates and exhibits a direct interaction with ubiquitin in Triton X-100 insoluble aggregates in the ipsilateral neocortex after transient MCAO. (A) Ipsilateral neocortical tissue was obtained from mice subjected to either sham-surgery or 30-minute MCAO and 3 hours of reperfusion. Aggregated as well as soluble conjugated and free ubiquitin, SUMO1, and SUMO2/3 were visualized by western blotting with respective antibodies. Optical densities of ubiquitin-, SUMO1-, SUMO2/3-, and GAPDH-stained bands were quantified. Brackets indicate areas used for quantification of conjugated ubiquitin and SUMO. ∗P<0.05 from sham; n=3 per group. (B) Aggregates were isolated as in A and evaluated for the presence of lysine 48- and lysine 63-linked ubiquitin chains by western blotting with respective antibodies. Areas marked with brackets were quantified. ∗P<0.01 from sham; n=3 per group. (C) Ubiquitin-containing aggregates were isolated from ipsilateral and contralateral cortices of mice treated with 30-minute ischemia and 1-hour reperfusion. Aggregates were solubilized by denaturation in 2% SDS-containing buffer and renaturated as described in the Materials and Methods section. Ubiquitin was immunoprecipitated from the SDS-soluble supernatant and presence of SUMO2/3 was tested by western blotting. Pull down with mouse IgG₁ isotype control antibody was executed as precipitation control. The shown blots are representative of three independent experiments. (D) The same assay as in C was performed with immunoprecipitation of SUMO2/3. Ubiquitin coprecipitation was evaluated by western blotting with anti-ubiquitin antibody. Blots are representative for three experiments. Conj, conjugated; contra, contralateral; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; h, hours; IgG, immunoglobulin G; IP, immunoprecipitation; ipsi, ipsilateral, kDa, kilodalton; MCAO, middle cerebral artery occlusion; SDS, sodium dodecyl sulfate; SUMO, small ubiquitin-related modifier; S, sham.

SUMO2/3 and Ubiquitin Interact within Postischemic Aggregates

To examine the potential association between SUMOylation and ubiquitination after ischemia, we asked whether SUMO2/3 and ubiquitin could physically interact within postischemic aggregates. To this end, we solubilized aggregate fractions by SDS-mediated denaturation and ensured successful solubilization by exclusive detection of ubiquitin conjugates in the supernatant (data not shown). When ubiquitin was pulled down from SDS-solubilized contents we observed significant coprecipitation of SUMO2/3 (Figure 1C). Only minimal ubiquitin or SUMO was detected in the contralateral cortex or by pull down with an unrelated IgG control antibody from the ipsilateral cortex, attesting to the specificity of the interaction. To further confirm our data obtained with ubiquitin pull down we performed reciprocal immunoprecipitation with SUMO2/3 and detection of ubiquitin by western blotting. Analogous to the result with ubiquitin pull down, we found marked coprecipitation of ubiquitin with SUMO2/3 (Figure 1D). These results suggest a direct association between ubiquitin and SUMO2/3 within Triton X-100 insoluble protein aggregates after focal cerebral ischemia.

SUMO2/3 Accumulation within Ubiquitin-Rich Aggregates is Induced Rapidly with Reperfusion Independent of the Duration of Ischemia

Next, we investigated whether SUMO2/3 aggregates are formed under the same conditions established earlier for ubiquitin aggregation. ² Since reperfusion was absolutely necessary for ubiquitin accumulation, ² we first examined SUMO aggregation levels after transient ischemia with different periods of reperfusion or times of permanent ischemia. Reperfusion readily induced SUMO2/3 accumulation peaking after 1 and 3 hours, while permanent middle cerebral artery occlusion failed to increase SUMOylation levels at any time point (Figure 2A). SUMO2/3 as well as ubiquitin aggregation occurred rapidly and was observed as early as 5 minutes after reperfusion (Figure 2B). Finally we determined SUMO2/3 aggregation levels after ischemic episodes of different durations. As also observed for ubiquitin (Figure 2C and ref. 2), middle cerebral artery occlusion lasting 10 or 15 minutes, which produces either no or minimal tissue damage in our model, ² was sufficient to trigger SUMO2/3 aggregation (Figure 2C).

Figure 2.

SUMO2/3 accumulation within ubiquitin-rich aggregates is rapidly induced dependent on reperfusion without a correlation to the duration and severity of ischemia. (A) Ipsilateral neocortices were harvested from sham-treated animals, after 30-minute MCAO followed by indicated reperfusion times, or after 1-, 3-, 6-, and 24-hour permanent MCAO. (B) Insoluble SUMO2/3 and ubiquitin aggregate levels were determined in ipsilateral neocortices of sham-operated animals, or mice, which underwent 30 minutes of ischemia and 5-, 15-, and 60-minute reperfusion. (C) Tissue was obtained from ipsilateral neocortices of sham-operated animals, or mice, which underwent 5, 10, 15, and 30 minutes of ischemia and 1- hour reperfusion. Optical densities of ubiquitin-, SUMO1- and SUMO2/3-stained bands were quantified from all experimental groups. Areas used for quantification are marked with a bracket. ∗P<0.05 from sham; n=3 per group. h, hours; kDa, kilodalton; MCAO, middle cerebral artery occlusion; min, minutes; S, sham; SUMO, small ubiquitin-related modifier.

Discussion

Both SUMO and ubiquitin conjugates are dramatically increased after focal cerebral ischemia,^{2, 3, 6, 9} and here we provide evidence that these events might be linked. We find that SUMO conjugation is not only globally elevated after ischemia/reperfusion, but that it also, like ubiquitin, accumulates in insoluble protein aggregates. The accumulation is specific for SUMO2/3 since we could not identify any induction of SUMO1 conjugation. While SUMO2 and SUMO3 are identical except for three amino acids in the amino-terminus and are functionally indistinguishable, SUMO1 shares only 50% identity and has both distinct and overlapping targets. ¹¹ Our finding of an induction of SUMO2/3, but not SUMO1 after ischemia is consistent with previous reports showing that SUMO2/3 constitutes a more abundant protein modification than SUMO1 in response to cellular stress.^{8, 12} Furthermore, our data in focal ischemia are in line with earlier observations in transient forebrain ischemia where no major changes in SUMO1 conjugation patterns were observed, while SUMO2/3 accumulation was significantly induced. ¹³

SUMO modification was first thought to be functionally distinct from the ubiquitin system. However, over the past decade it has become clear that in addition to their structural similarities, ubiquitin and SUMO have significant functional interactions.^{7, 14} These include (1) the competitive targeting of same attachment sites in common substrates, (2) the interdependent conjugation to distinct attachment sites in common targets, or (3) the sequential modification of different targets within the same signaling pathway, leading to antagonistic or synergistic effects on target protein properties. Our data of coprecipitation of SUMO2/3 and ubiquitin from SDS-denatured contents of the aggregate fraction strongly suggest a covalent, direct interaction between the two modifications. This could be in the form of SUMO2/3 and ubiquitin conjugated as monomers or mono-specific chains to different acceptor sites or as mixed chains to a common acceptor site within the same substrates. A potentially direct SUMO and ubiquitin interaction was suggested previously in a forebrain ischemia model where removal of SUMO conjugates led to a reduction in molecular weight of ubiquitinated proteins, indicating attachment to the same substrates. ¹⁵

An extensive crosstalk exists between SUMOylation and ubiquitination with potential implications for the functional consequences of the modification.^{7, 14} The significance of the accumulation of ubiquitin and SUMO2/3 conjugates for the outcome of cerebral ischemia, however, remains unknown. Of note, the induction of SUMO2/3 as well as ubiquitin conjugation is dependent on reperfusion, indicating that their presence might be associated with survival potential. Aggregation of SUMO and ubiquitin is rapid, transient, resolves before major tissue damage evolves and is not linked to cerebral infarction in our model, ² suggesting that they are not causative for ischemic injury. In fact, elevated SUMO conjugation has been associated with neuroprotection, since its induction and suppression resulted in reduced and increased neuronal damage and infarction, respectively.^{4, 5, 6} The implications of ubiquitin aggregation for the outcome of focal cerebral ischemia are not clear. However, our previous data suggest that the presence of ubiquitin aggregates may in fact indicate a greater potential for tissue viability in this setting. ² Considering the interaction of postischemic SUMO2/3 and ubiquitin modifications shown here, we speculate that SUMO2/3 and ubiquitin may work in concert to protect vital constituents of the cellular proteome from ischemic stress. Future studies must be directed at (1) understanding the contributing factors for the induction of ubiquitin/SUMO accumulation and (2) identifying modified substrates to evaluate the actual implications for brain survival after stroke.