The Journal of Neuroscience
http://jneurosci.msubmit.net
JN-RM-5283-14
SUMOYLATION OF LYS590 OF THE f-LOOP OF NCX3 BY SUMO1
PARTICIPATES IN BRAIN NEUROPROTECTION INDUCED BY ISCHEMIC
PRECONDITIONING
Lucio Annunziato, Federico II University of Naples
Ornella Cuomo, Federico II University of Naples
Giuseppe Pignataro, Federico II University of Naples
Rossana Sirabella, Fondazione IRCCS SDN
Pasquale Molinaro, Federico II University of Naples
Serenella Anzilotti, SDN IRCCS Via Gianturco
Antonella Scorziello, Department of Neuroscience, Physiology Unit, School
of Medicine, Federico II University of Naples
Maria Josè Sisalli, Federico II University of Naples
Claudia Savoia, Federico II University of Naples
Gian Franco Di Renzo,
Commercial Interest: No
This is a confidential document and must not be discussed with others, forwarded in any
form, or posted on websites without the express written consent of The Journal for
Neuroscience.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Title:
SUMOYLATION OF LYS590 OF THE f-LOOP OF NCX3 BY SUMO1 PARTICIPATES IN
BRAIN NEUROPROTECTION INDUCED BY ISCHEMIC PRECONDITIONING
Abbreviated title:
NEUROPROTECTIVE EFFECT OF NCX3 SUMOYLATION
Author names and affiliation, including postal codes:
Ornella Cuomo1, Giuseppe Pignataro1, Rossana Sirabella2, Pasquale Molinaro1,
Serenella Anzilotti2, Antonella Scorziello1, Maria Josè Sisalli1, Claudia Savoia1,
Gianfranco Di Renzo1, Lucio Annunziato1
1
Division of Pharmacology, Department of Neuroscience, Reproductive and Dentistry
Sciences, School of Medicine, Federico II University of Naples, Via Pansini 5, 80131
Naples, Italy; 2Fondazione IRCCS SDN, Via Gianturco, 113, 80143, Naples, Italy
Corresponding author with complete address, including an email address and
postal code
Lucio Annunziato 1Division of Pharmacology, Department of Neuroscience, Reproductive
and Dentistry Sciences, School of Medicine, Federico II University of Naples, Via Pansini
5, 80131 Naples, Italy
Tel +390817463325, Fax +390817463323; email:
lannunzi@unina.it
Number of pages: 20
Number of figures, tables, multimedia and 3D models (separately): 8 Figures
Number of words for Abstract: 248
Number of words for Introduction: 500
Number of words for Discussion: 883
Conflict of Interest: The authors declare no competing financial interests
Acknowledgements:
We thank Dr. Paola Merolla for the editorial revision, Prof. Hallenbeck J.M. from
National Institute of Neurological Disorders and Stroke (NINDS) and National Institutes
of Health (NIH), Bethesda, Maryland, USA for kindly providing us SUMO1 antibody and
Vincenzo Grillo and Carmine Capitale for their technical assistance.
This work was supported by Grant COFIN 2008, Ricerca-Sanitaria (Grant RFFSL352059), Ricerca finalizzata (2006), Progetto-Strategico (2007), Progetto Ordinario
(2007), Ricerca finalizzata (2009), Ricerca-Sanitaria progetto Ordinario (2008) from
Ministero della Salute (all L. A.), and Progetto Giovani Ricercatori GR-2010-2318138
from Ministero della Salute to A.S.
42
43
44
45
46
1
47
Abstract
48
The small ubiquitin-like modifier (SUMO), a ubiquitin-like protein involved in post-
49
translational protein modifications, is activated by several conditions, such as heat stress,
50
hypoxia,
51
neuroprotection against stressful stimuli by regulating the function and the fate of proteins
52
involved in stress signalling pathways. Sumoylation enzymes and substrates are
53
expressed not only in the cytoplasmic and nuclear compartments, but also at the plasma
54
membrane level. Among the numerous plasma membrane proteins controlling ionic
55
homeostasis during cerebral ischemia, NCX3, one of the three Sodium/Calcium
56
exchangers expressed in the CNS, exerts a protective role during ischemic
57
preconditioning. In this study we evaluated whether NCX3 is a target for sumoylation and
58
whether this post-translational modification participates in ischemic preconditioning-
59
induced neuroprotection. To test these hypotheses, we analyzed (1)SUMO1 conjugation
60
pattern after ischemic preconditioning; (2)the effect of SUMO1 knockdown on the ischemic
61
damage after transient middle artery occlusion (tMCAO) and ischemic preconditioning;
62
(3)the possible interaction between SUMO1 and NCX3; and (4)the molecular determinants
63
of NCX3 sequence responsible for sumoylation.
64
We found that (1)SUMO1 knockdown worsened ischemic damage and reduced the
65
protective effect of ischemic preconditioning; (2)SUMO1 bound to NCX3 at the lysine
66
residue 590 of the f-loop, and its silencing increased NCX3 degradation; and (3)NCX3
67
sumoylation took part to SUMO1 protective role during ischemic preconditioning. Thus, our
68
results demonstrate that NCX3 sumoylation by SUMO1 confers additional neuroprotection
69
in ischemic preconditioning.
70
Finally, this study suggests that NCX3 sumoylation might be a new potential target to
71
enhance ischemic preconditioning-induced neuroprotection.
and
hibernation.
Recent
evidence
72
73
74
75
76
77
78
79
80
2
indicates
that
sumoylation
confers
81
Introduction
82
Small Ubiquitin-like Modifier (SUMO) conjugation is an enzymatic process that is
83
biochemically analogous to, but functionally distinct from, the ubiquitination process. It
84
involves the covalent attachment of SUMO to substrate proteins, however, unlike
85
ubiquitination, sumoylation does not lead to degradation of the target substrate but may
86
modulate intracellular protein localization, activity, and stability (Pichler and Melchior,
87
2002; Gill, 2003; Seeler and Dejean, 2003; Hay, 2005; Mukhopadhyay and Dasso, 2007;
88
Zhao, 2007). So far, four SUMO proteins have been identified, SUMO1–4. SUMO 1-3 are
89
present in the brain, whereas SUMO4 is mainly localized in the kidney (Bohren et al.,
90
2004). Interestingly, sumoylation is activated after different stress conditions, including
91
anoxia, hypothermia, and hypoxia (Yang et al., 2008c; Yang et al., 2008a, b; Lee et al.,
92
2009). Changes in brain sumoylation pattern have also been reported after focal cerebral
93
ischemia (Cimarosti and Henley, 2008), where it may represent a protective response.
94
Furthermore, SUMO-1 knockdown decreases cell resistance to oxygen–glucose
95
deprivation (OGD), whereas SUMO-1 overexpression renders neurons less susceptible to
96
OGD-induced cell damage (Lee et al., 2009). Intriguingly, sumoylation is strongly activated
97
in the brain of hibernating animals during the torpor phase, thus shielding neurons from
98
damage induced by reduced blood flow and substrate deprivation (Lee et al., 2009).
99
To date, no information is available regarding a potential neuroprotective role of protein
100
sumoylation during endogenous neuroprotective events like ischemic preconditioning in
101
vivo. Ischemic preconditioning, a phenomenon whereby a subliminal injurious stimulus
102
applied before a longer harmful ischemia produces neuroprotection (Kirino, 2002; Dirnagl
103
et al., 2003; Gidday, 2006), shares several trasductional pathways with hibernation, a
104
condition in which sumoylation is activated. Several proteins involved in the
105
neuroprotection mediated by ischemic preconditioning, have been proposed as
106
sumoylation substrates (Rajan et al., 2005). Recently, convincing evidence has revealed
107
that sumoylation enzymes and substrates are also present at the plasma membrane level,
108
wherein they regulate plasma membrane protein expression. This finding led us to
109
hypothesize that sumoylation might influence ionic homeostasis by regulating the stability
110
and the expression of those transmembrane proteins involved in brain ischemia,
111
specifically NCX3.
112
The reason for investigating NCX3 as a possible target for SUMO-mediated
113
neuroprotective
114
role in cerebral ischemia(Pignataro et al., 2004; Molinaro et al., 2008), since both NCX3
effect
is
that
NCX3
3
plays
a
relevant
neuroprotective
115
knockdown and NCX3 knockout exacerbate the ischemic damage induced by permanent
116
and transient middle cerebral artery occlusion (pMCAO and tMCAO) (Pignataro et al.,
117
2004; Molinaro et al., 2008). More intriguingly, we have also identified NCX3 as a new
118
additional effector of ischemic preconditioning (Pignataro et al., 2012). Given these
119
premises, the aim of the present study was to evaluate whether NCX3 is a target for
120
sumoylation and whether this post-translational modification participates in promoting
121
ischemic preconditioning-induced neuroprotection. For these aims, we analyzed: (1)
122
SUMO1 conjugation pattern after ischemic preconditioning, (2) the effect of SUMO1
123
knockdown on ischemic damage after tMCAO and ischemic preconditioning, (3) the
124
possible interaction between SUMO1 and NCX3, and, finally, (4) the molecular
125
determinants of NCX3 sequence responsible for sumoylation.
126
4
127
128
METHODS
129
Experimental Groups
130
Male Sprague–Dawley rats (n=110; 250-300 g; Charles River) were housed under diurnal
131
lighting conditions (12 h darkness/light). Experiments were performed according to the
132
international guidelines for animal research and approved by the Animal Care Committee
133
of “Federico II” University of Naples (Naples, Italy).
134
135
Focal Ischemia
136
Transient focal ischemia was induced, as previously described (Pignataro et al., 2008), by
137
suture occlusion of the middle cerebral artery (MCA) in male rats anesthetized with 1.5 %
138
sevofluorane, 70% N2O, and 28.5% O2. In brief, a 2-O surgical monofilament nylon suture
139
(Doccol, CA, USA) was inserted from the external carotid artery into the internal carotid
140
artery and advanced into the circle of Willis until it reached the branching point of the MCA,
141
thereby occluding the MCA (Longa et al., 1989). Achievement of ischemia was confirmed
142
by monitoring regional cerebral blood flow in the area of the right MCA. Cerebral blood
143
flow (CBF) was monitored through a disposable microtip fiber optic probe (diameter
144
0.5mm) connected through a Master Probe to a laser Doppler computerized main unit
145
(PF5001; Perimed, Sweden) and analyzed using PSW Perisoft 2.5 (Pignataro et al., 2008).
146
Animals not showing a cerebral blood flow reduction of at least 70%, as well as those
147
dying after ischemia induction (5%), were excluded from the study. Rectal temperature
148
was maintained at 37±0.5°C with a thermostatically controlled heating pad and lamp. All
149
surgical procedures were performed under an operating stereomicroscope. After 100 min
150
MCA occlusion, rats were re-anesthetized and the filament was withdrawn to restore blood
151
flow. Animals were randomized into the different experimental groups in a blind manner.
152
153
Preconditioning Experimental Protocol
154
Ischemic preconditioning was induced as previously described (Pignataro et al., 2008). In
155
brief, preconditioning was achieved by subjecting rats to 30 min of MCAO, followed by 72
156
h of reperfusion. At the end of the reperfusion period, the MCA was re-occluded for 100
157
min. Animals were then recovered for 24 h. The success of the experimental procedures
158
was confirmed by measuring CBF in all the experimental steps.
159
160
Evaluation of the Infarct Volume
5
161
For the analysis of ischemic damage,
rats were sacrificed 24h after tMCAO or
162
preconditioning+tMCAO. The ischemic volume was evaluated with 2,3,5-triphenyl
163
tetrazolium chloride (TTC) staining. In particular, the brains were cut into 1-mm coronal
164
slices with a vibratome (Campden Instrument, 752M) and the sections were incubated in
165
2% TTC for 20 min and in 10% formalin overnight. The infarction area was calculated with
166
image analysis software (Image-Pro Plus) (Bederson et al., 1986). The total infarct volume
167
was expressed as percentage of the volume of the hemisphere ipsilateral to the lesion in
168
order to correct brain edema. Ischemic damage was evaluated in a blind manner.
169
170
Intracerebroventricular administration of siRNA
171
In rats positioned on a stereotaxic frame, a 23-g stainless steel guide cannula was
172
implanted into the right lateral ventricle using the stereotaxic coordinates from the bregma:
173
0.4mm caudal, 2mm lateral and 2mm below the dura (Franklin and Paxinos, 1997;
174
Pignataro et al., 2004).
175
Two different siRNAs were tested (Quiagen siRNA-A #SI01910034 and siRNA-B
176
#SI01910027). siRNAs (10l, 2M or 10M) and siCTL were icv administered twice a day
177
starting from 72 hours before preconditioning or tMCAO induction. Since their
178
effectiveness was similar (about 50% reduction), all the experiments were performed with
179
siRNA A (10l, 10M), which was administered twice a day, starting from 72 hours before
180
preconditioning or tMCAO induction.
181
182
Western Blotting Analysis
183
Cortical samples were harvested from ischemic brains of rats subjected to 100 min of
184
MCAO,
185
preconditioning+tMCAO. Ipsilateral cortex and striatum were obtained at four different
186
reperfusion time intervals after the last occlusion, at 3, 5, 24, and 72 hours. The same
187
sample groups were obtained from brains of sham-operated animals.
188
Rat brain samples were first homogenized with an 18-gauge needle in a lysis buffer (100
189
mM Tris-HCl, pH 6.8, 2% SDS, 50 mM EDTA) containing a protease inhibitor cocktail, 1
190
mM PMSF, and 10 mM NEM. Then, they were sonicated for 10 sec, heated at 99 °C for
191
10 min, and centrifuged for 15 min at 14,000 rpm. Protein concentration was estimated
192
using the Bradford reagent (Bio-Rad Laboratories, Segrate, Milan, Italy). Twenty
193
micrograms or 50 g of proteins were mixed with a Laemmli sample buffer. Next, the
194
samples
from
were
brains
of
separated
preconditioned
on
8%
sodium
6
rats,
and
dodecyl
from
sulfate
rats
subjected
polyacrylamide
to
gel
195
electrophoresis,
and
transferred
onto
Amersham™
Hybond™-ECL
nitrocellulose
196
membranes (GE Healthcare, Milan, Italy). The non-specific binding sites were blocked with
197
an incubation of 5% non-fat dry milk in Tris buffered saline (TBS) 0.1% Tween 20 (Sigma-
198
Aldrich, Milan, Italy) for 2 hours at room temperature. After that, blots were probed with
199
SUMO-1 (1:1000) antibody (kindly given by Hallenbeck J.M. from National Institute of
200
Neurological Disorders and Stroke (NINDS) and National Institutes of Health (NIH),
201
Bethesda, Maryland, USA), with NCX3 antibody (kindly given by Philipson K. from David
202
Geffen School of Medicine at UCLA, Los Angeles, CA), and with β-actin (1:1000, Sigma-
203
Aldrich, Milan, Italy) overnight at 4 °C. Finally, detection was achieved using a horseradish
204
peroxidase-conjugated secondary anti-rabbit antibody (1:2000; Cell Signaling; 60min at
205
room temperature in 5% non-fat milk) and an enhanced luminescence kit (GE Healthcare,
206
Milan, Italy) (Cuomo et al., 2008).
207
208
Generation and Stable Expression of Wild-Type and Mutant NCX3FLAG cDNAs
209
Murine NCX3 cDNA with a 3XFLAG epitope at the C-terminus (NCX3FLAG) was cloned in
210
pReceiver-M13 expressing vector by Genecopoeia. Site-directed and deletion mutants of
211
NCX3FLAG were generated by means of QuikChange strategy (Stratagene, La Jolla, CA).
212
In particular, the amino acid regions 292-710 and 528-676 were deleted from NCX3FLAG
213
obtaining a mutant lacking the entire f loop, NCX3292-710 (NCX3f), and another lacking
214
a portion of f loop, NCX3528-676, respectively. Two site-directed mutants were obtained
215
by substituting lysine 456 with glutamate (NCX3K456E), and
216
lysine 590 with leucine and glutamate, respectively (NCX3FK589LE).
217
All mutant exchangers obtained were verified by sequencing both DNA strands (Primm,
218
Milan, Italy). To stably express wild-type, site-directed and deletion mutants of NCX3FLAG,
219
pReceiver-M13 plasmids carrying cDNAs were transfected into wild-type BHK cells by
220
Lipofectamine 2000 (Life Technologies, San Giuliano Milanese, Italy) protocol. Stable cell
221
lines were selected by G418 resistance and by a Ca2+-killing procedure (Molinaro et al.,
222
2013).
223
224
225
Immunoprecipitation assay
226
BHK cells expressing NCX3 wild type and mutant forms, and brain tissue from control rats
227
and rats exposed to tMCAO or preconditioning followed by tMCAO were homogenized in
228
lysis buffer containing (mM) 50 mM Hepes, pH 7.4, 100 mM NaCl, 1.5 mM MgCl2, 0.2 %
229
NP40 with protease inhibitor cocktail, 1 mM PMSF, and 20 mM NEM. One milligram of
7
phenylalanine 589 and
230
lysate was precleared using protein A/G plus (Santa Cruz) for 1 h at 4C with constant
231
rotation and centrifuged for 2 min at 8000 rpm. Precleared lysates from BHK cells were
232
immunoprecipitated with anti-SUMO1 antibody (1:100; monoclonal anti-SUMO1 antibody,
233
Quiagen). In particular, lysates were incubated overnight with anti-SUMO1; then protein
234
A/G plus was added to the lysates for 2 h at 4°C with constant rotation. Finally,
235
immunoprecipitates were washed 4 times, and the total cell lysates or immunoprecipitates
236
were resolved by SDS-PAGE gel and transferred to a nitrocellulose membrane.
237
Immunoblot analysis was performed using an anti-flag antibody. As for the lysates from
238
brain tissue, 1 mg of precleared lysate was immunoprecipitated with an anti-NCX3
239
antibody (1:100 polyclonal anti-NCX3 antibody, Swant) using the same experimental
240
procedure described above. Finally, total lysates or immunoprecipitates were resolved by
241
SDS-PAGE gel and transferred to a nitrocellulose membrane. Immunoblot analysis was
242
performed using anti-SUMO1 and anti-NCX3 antibodies.
243
244
Primary cortical neurons and Immunocytochemistry
245
Mixed cultures of cortical neurons from 2-4-day old Wistar rat pups were prepared by
246
modifying a previously described method (Abramov et al., 2007; Scorziello et al., 2013). In
247
brief, the tissue was minced, trypsinized (0.1% for 15 min at 37°C), triturated, plated on
248
poly-D-lysine-coated coverslips, and cultured in Neurobasal medium (Life Technologies,
249
San Giuliano Milanese, Italy) supplemented with B-27 (Life Technologies, San Giuliano
250
Milanese, Italy) and 2 mM L-glutamine. The cells were plated at a concentration of
251
1.8x106 on 25-mm glass coverslips. Cultures were maintained at 37°C in a humidified
252
atmosphere of 5% CO2 and 95% air, fed twice a week, and maintained for a minimum of
253
10 days before experimental use.
254
For the immunocytochemistry experiments, cortical neurons were rinsed twice in cold 0.01
255
M phosphate buffered saline (PBS) at pH 7.4 and fixed at room temperature in 4% (w/v)
256
paraformaldehyde (Sigma-Aldrich, Milan, Italy) for 20 minutes. Following three washes in
257
PBS, cells were blocked in PBS containing 10% FBS and the following antibodies: anti-
258
NCX3 rabbit (kindly supplied by Dr. Philipson, dilution 1:4000), anti-SUMO monoclonal
259
mouse (dilution 1:1000, Santa Cruz). Neurons were then incubated overnight at 4°C. Next,
260
slides were washed in PBS, incubated with anti-rabbit cy2 antibody (dilution 1:200,
261
Jackson Immunoresearch Laboratories) and anti-mouse cy3 antibody (dilution 1:200,
262
Jackson Immunoresearch Laboratories) for 1 hour at room temperature (25°C) under dark
263
conditions, and washed again with PBS. Finally, they were mounted onto Slow
8
264
fadeTM antifade (Invitrogen-Molecular Probes) and images were observed using a Zeiss
265
LSM510 META/laser scanning confocal microscope. Single images were taken with an
266
optical thickness of 0.7 µm with a size of 1024×1024 pixels.
267
268
Tissue processing, immunostaining and confocal immunofluorescence
269
For the histological examination, animals were transcardially perfused under deep
270
anesthesia with saline solution containing 0.01 ml heparin, followed by 60 ml of 4%
271
paraformaldehyde in saline solution. The brains were removed and post-fixed overnight at
272
+4 °C and cryoprotected in 30% sucrose in 0.1 M phosphate buffer (PB) with sodium azide
273
0.02% for 48 h at 4 °C. The brains were sectioned and frozen on a sliding cryostat at 40
274
μm thickness. The sections were then incubated with PB triton x 0,3% blocking solution
275
(0,5%milk, 10%FBS, 1%BSA) for 2 hours. Rabbit anti-NCX3 antibodies (kindly provided
276
by Dr. Philipson, dilution 1:4000) and mouse anti-SUMO monoclonal antibodies (dilution
277
1:500, Santa Cruz) were used as primary antibodies. The sections were then incubated
278
with the corresponding fluorescent-labelled secondary antibodies (Alexa 488/Alexa 594-
279
conjugated antimouse/antirabbit IgGs). Nuclei were counterstained with Hoechst. Images
280
were observed with a Zeiss LSM510 META/laser scanning confocal microscope. Single
281
images were taken with an optical thickness of 0.7 m with a size of 1024x1024 pixels. In
282
double-labelled sections, the pattern of immunoreactivity for both antigens was identical to
283
that seen in single-stained material. In control groups primary antisera were replaced with
284
normal serum. A possible cross-reactivity between IgGs in double immunolabelling
285
experiments was assessed by processing sections through the same immunocytochemical
286
sequence except that the primary antisera were replaced either with normal serum or with
287
only one primary antibody; the full complement of the secondary antibodies was
288
maintained. In addition, the secondary antibodies used in this study are highly pre-
289
adsorbed to the IgGs of numerous species. Tissue labelling without primary antibodies
290
was also tested to exclude autofluorescence. No specific staining was observed under
291
these control conditions, thus confirming the specificity of the immunosignals.
292
293
294
Statistic Analysis
295
Values are expressed as means ± S.E.M. Statistical analysis was performed with 2-Way
296
ANOVA, followed by Newman-Keuls test. Statistical significance was accepted at the 95%
297
confidence level (p < 0.05).
298
9
299
RESULTS
300
301
SUMO1 conjugation increases in the temporoparietal cortex after ischemia,
302
preconditioning, and preconditioning+ischemia
303
The role played by SUMO1 conjugation during cerebral ischemia and ischemic
304
preconditioning was assessed by evaluating the sumoylation pattern in the ipsilesional
305
temporoparietal cortex and striatum of rats subjected to ischemia, preconditioning, and
306
preconditioning+ischemia and later sacrificed at different reperfusion time intervals: 3, 5,
307
24, and 72 hours. In the ipsilesional temporoparietal cortex, SUMO1 protein conjugation
308
significantly increased at 5 and 24 hours after tMCAO, at 3, 5, 24, and 72 hours after
309
preconditioning, and at 5 hours after preconditioning+tMCAO (Fig. 1A, 1B and 1C).
310
However, in the striatum no changes were observed under all the experimental conditions
311
(Fig.2A, 2B and 2C). The results obtained by Western blot were confirmed by confocal
312
immunofluorescent experiments. In particular, 5 hours after tMCAO, preconditioning, or
313
preconditioning+tMCAO, SUMO1 immunolabelling increased in the temporoparietal cortex
314
(Fig. 3), whereas no variations were assessed in the striatum (Fig. 4).
315
316
317
SUMO1 colocalizes with NCX3 in primary cortical neurons, in temporoparietal
318
cortex, and in striatum of ischemic or preconditioned rats and immunoprecipitates
319
with the antiporter isoform in NCX3 stably transfected BHK cells.
320
To identify new targets of SUMO1 that might contribute to the neuroprotective effect of
321
sumoylation during ischemic conditions, we verified whether NCX3 might be a potential
322
candidate for the sumoylation process. Immunohistochemical analysis performed on rat
323
brain
324
colocalization between NCX3 and SUMO1 signals in the temporoparietal cortex and
325
striatum (Fig. 3 and 4). To further confirm the colocalization between NCX3 and SUMO1,
326
we performed an immunocytochemical analysis in primary rat cortical neurons. We found
327
that these two proteins colocalized in neuronal cell bodies (Fig. 5). Then, by using the
328
SUMO plot analysis program (http://www.abgent.com/sumoplot), we analyzed the
329
probability of NCX3 sumoylation. NCX3 sequence contains nine sequences that might be
330
recognized by the SUMO conjugation enzyme; however one of these regions is localized
331
extracellularly (Fig. 6A) according to the newly proposed NCX3 topology (Ren and
332
Philipson, 2013). To assess the interaction between SUMO1 and NCX3, we performed
slices
after
tMCAO,
preconditioning,
10
and
preconditioning+tMCAO
showed
333
immunoprecipitation experiments on wild-type BHK cells and on BHK cells stably
334
transfected with NCX3FLAG (NCX3F) and we found immunoprecipitation between NCX3 and
335
SUMO1 (Fig. 6B). Furthermore, to better determine a possible sumoylation site, we
336
performed experiments on an NCX3FLAG mutant lacking the f-loop (NCX3Ff), since this
337
region contained most of the lysine residues with the highest probability of SUMO1
338
binding. Wild-type and stably transfected BHK cell lysates were immunoprecipitated with
339
SUMO1 antibody and then immunoblotted with an anti-flag antibody. Results showed the
340
presence of an immunoprecipitate containing NCX3F and SUMO1 in NCX3F-transfected
341
BHK cells. By contrast, no immunoprecipitate was observed in NCX3Ff-transfected BHK
342
cells, thus restricting the putative sumoylation site in the f-loop of the antiporter (Fig. 6B).
343
The substitution of lysine456, a residue with high probability of conjugation, with glutamate
344
(NCX3K456E) did not prevent co-immunoprecipitation between NCX3 and SUMO1 (Fig.
345
6C). To restrict the analysis of the putative sumoylation sites, we removed the Ca2+
346
binding domain 2 (CBD2), a subregion of the f-loop 528-676 (NCX3F528-676) that
347
contains three of the seven sequences present in the f-loop. This NCX3 deletion mutant
348
showed no immunoprecipitation with SUMO1, thus suggesting that the lysine responsible
349
for sumoylation was present in this region. To determine the single amino acid residue
350
involved in SUMO1 attachment, we performed immunoprecipitation on a site-directed
351
mutant obtained by substituting phenylalanine589 and lysine590 with leucine and
352
glutamate, respectively (NCX3FK589LE). In this case as well, no immunoprecipitation was
353
found, thus demonstrating that lysine590 is required for sumoylation (Fig. 6C).
354
The biochemical interaction between SUMO1 and NCX3 was also confirmed in vivo.
355
Indeed, results from immunoprecipitation experiments performed on rat cortex from
356
ischemic and preconditioned animals showed that NCX3 immunoprecipitated with SUMO1
357
(Fig. 7A).
358
359
360
SUMO1
knockdown
downregulates
361
preconditioning+tMCAO and reduces the neuroprotective effect elicited by ischemic
362
preconditioning in rats.
363
The effectiveness of siRNAs against SUMO1 was previously evaluated in non-ischemic
364
rats intracerebroventricularly injected with the silencing agents. Western Blot analysis
365
revealed a 50% reduction of SUMO1 both in cortex and in striatum (Fig. 8A). Two different
11
NCX3
protein
levels
after
366
siRNAs were tested as indicated in materials and method section, and, since their efficacy
367
was equivalent, all the experiments were performed only with siRNA-A.
368
To correlate the neuroprotection induced by preconditioning+tMCAO with the interaction
369
between NCX3 and SUMO1, we verified whether SUMO1 knockdown was able to
370
modulate NCX3 expression after preconditioning+tMCAO. As shown in Fig. 7, in the
371
temporoparietal cortex of rats subjected to preconditioning+tMCAO and previously treated
372
with siRNA targeting SUMO1, NCX3 protein levels significantly decreased both 24 and 72
373
hours after tMCAO (Fig. 7B) compared to control. By contrast, in the temporoparietal
374
cortex of rats receiving siRNA control and exposed to the same protocol of
375
preconditioning+tMCAO, NCX3 expression significantly increased 72 hours after ischemia
376
(Fig. 7B), confirming our previous results (Pignataro et al., 2012). Finally, to evaluate the
377
effect of SUMO1 silencing on the protection induced by ischemic preconditioning, rats
378
were
379
preconditioning, tMCAO, or preconditioning followed by tMCAO. SUMO1 knockdown
380
induced a significant increase in ischemic volume after tMCAO (Fig. 8B) and, more
381
interestingly, it partially prevented the neuroprotective effect induced by ischemic
382
preconditioning (Fig. 8C) (% infarct volume: 50.9 ±3 in rats treated with siRNA control and
383
subjected to tMCAO vs 63±1.3 in rats treated with siRNA-A SUMO1 and subjected to
384
tMCAO;
385
preconditioning+tMCAO vs 39.1±9 in rats treated with siRNA-A SUMO1 and subjected to
386
preconditioning+tMCAO).
intracerebroventricularly
15.3±1.65
in
rats
injected
treated
with
with
387
12
siRNA-A
siRNA
targeting
control
and
SUMO1
before
subjected
to
388
DISCUSSION
389
The results of the present paper demonstrate that SUMO1 is involved in the
390
neuroprotective mechanisms elicited by in vivo ischemic preconditioning, and, more
391
interestingly, propose for the first time NCX3 as a new putative effector of SUMO1
392
neuroprotective action.
393
Several proteins are known to participate in the pathophysiology of brain ischemia and in
394
ischemic preconditioning-induced neuroprotection. Although some of these proteins have
395
been proposed as sumoylation substrates (Rajan et al., 2005), so far, no correlation has
396
yet been found between them and SUMO levels.
397
Recent evidence indicates that the sumoylation enzymes and substrates are present not
398
only in the nuclear compartment but also at the plasma membrane level (Rajan et al.,
399
2005). In particular, sumoylation emerged also as a mechanism able to regulate
400
plasmamembrane proteins involved in ionic homeostasis regulation (Silveirinha et al.,
401
2013). In the present paper we hypothesized that a new substrate for sumoylation process
402
may be represented by the isoform 3 of the plasmamembrane sodium calcium exchanger.
403
NCX3 has been recognized as a key player in the evolution of the ischemic brain damage
404
(Molinaro et al., 2008) and, more interestingly, as a new additional effector of ischemic
405
preconditioning (Pignataro et al., 2012; Sisalli et al., 2014). Indeed, we found that, SUMO1
406
conjugation pattern significantly increased after ischemia in the ipsilateral temporoparietal
407
cortex, whereas no changes were observed in the striatum. The observation that this
408
increase occurred in a brain area that is far less damaged by ischemic injury than the
409
striatum supports our hypothesis that sumoylation constitutes a cell protective response to
410
stress associated to ischemic injury. More important, evidence showing that SUMO1
411
conjugation pattern also increased in the ipsilateral temporoparietal cortex during ischemic
412
preconditioning, where a reduction of the damage occurred, thus suggesting a role of
413
SUMO1 in ischemic preconditioning-induced neuroprotection .
414
In further support of a possible interaction between NCX3 and SUMO1, confocal analysis
415
revealed that NCX3 and SUMO1 colocalized in the same neurons of the temporoparietal
416
cortex under physiological conditions, after preconditioning, tMCAO and preconditioning
417
followed by tMCAO. More interestingly, SUMO1 and NCX3 signals increased in the same
418
neurons after preconditioning and preconditioning followed by tMCAO. Colocalization
419
results were further reinforced by the co-immunoprecipitation study, showing a direct
420
binding between NCX3 and SUMO1 in BHK cells stably expressing NCX3.
13
421
To investigate the molecular determinants involved in NCX3 sumoylation we identified, by
422
using a bioinformatic approach, nine putative consensus sequences of the NCX3 protein
423
that could be recognized by SUMO conjugating enzymes. Eight sequences were localized
424
in the cytosolic side, seven of which in the f-loop region, which plays a key regulatory
425
function in the antiporter activity. Interestingly, the removal of the whole f-loop prevented
426
the binding between NCX3 and SUMO1, thus indicating that this region contains the
427
molecular determinants responsible for sumoylation. Similarly, when we removed the
428
CBD2 domain, corresponding to the subregion 528-676 of the f-loop, immunoprecipitation
429
between NCX3 and SUMO1 was also hampered. This finding thus indicated that the
430
putative sumoylation site is indeed localized in the CBD2 domain containing the three
431
identified putative binding sites. Moreover, since the FKND sequence, present at the level
432
of 589-592 aa, showed the highest score for sumoylation, we hypothesized that lysine590
433
was the sumoylation site for NCX3. Indeed, the substitution of phenylalanine589 and
434
lysine590
435
unrecognizable for the sumoylation enzymes, prevented NCX3 and SUMO1 co-
436
immunoprecipitation.
437
Results showing that downregulation of NCX3 protein levels occurred in SUMO1-silenced
438
rats exposed to preconditioning+tMCAO suggested that this post-translational modification
439
may increase NCX3 stability, thus preventing NCX3 from the degradation following tMCAO
440
(Pignataro et al., 2004; Pignataro et al., 2012).
441
According to previous studies, suggesting a neuroprotective role for SUMO1, we found
442
that SUMO1 silencing significantly increased the ischemic damage induced in rats by
443
tMCAO. More relevant, our findings demonstrated that SUMO1 silencing partially reverted
444
the neuroprotection mediated by ischemic preconditioning.
445
The involvement of SUMO1 in preconditioning was also supported by previous in vitro
446
observations, showing that preconditioned neurons maintained elevated SUMO-1
447
conjugation levels and SUMO-1-silenced neurons showed a reduced survival rate after
448
oxygen and glucose deprivation and an attenuated protective response to preconditioning
449
(Lee et al., 2009). Similarly, in vivo studies have shown that transgenic mice
450
overexpressing SUMO conjugating enzyme, Ubc9, are more resistant than wild type mice
451
to brain ischemia (Lee et al., 2011). Moreover, it has been demonstrated that a massive
452
conjugation of SUMO1 is highly activated in the torpor phase of hibernating ground
453
squirrels (Lee et al., 2009), which are naturally tolerant to oligemic conditions. Indeed,
with
leucine
and
glutamate,
14
respectively,
rendering
this
sequence
454
hibernation torpor has been shown to provide neuroprotection against a superimposed
455
lethal ischemic insult (Frerichs and Hallenbeck, 1998).
456
NCX3 downregulation during ischemic preconditioning in rats silenced for SUMO1 along
457
with the observation that SUMO1 silencing significantly increased the ischemic damage
458
and partially reverted the neuroprotection exerted by ischemic preconditioning support the
459
hypothesis that sumoylation of NCX3 may play a key role in the neuroprotection during
460
ischemic preconditioning. In fact, sumoylation of NCX3 might be considered a protective
461
mechanism to preserve NCX3 from degradation during ischemic conditions.
462
Collectively, our results show that SUMO1 plays a fundamental role in the neuroprotection
463
elicited by ischemic preconditioning and that its protective role might be, at least in part,
464
mediated by NCX3 conjugation. Finally, this post-translational process may represent a
465
potential
466
preconditioning.
pharmacological
target
to
enhance
467
468
469
470
471
472
15
the
beneficial
effect
of
ischemic
473
474
REFERENCES
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
Abramov AY, Scorziello A, Duchen MR (2007) Three distinct mechanisms generate oxygen free
radicals in neurons and contribute to cell death during anoxia and reoxygenation. J
Neurosci 27:1129-1138.
Bederson JB, Pitts LH, Germano SM, Nishimura MC, Davis RL, Bartkowski HM (1986)
Evaluation of 2,3,5-triphenyltetrazolium chloride as a stain for detection and quantification
of experimental cerebral infarction in rats. Stroke 17:1304-1308.
Bohren KM, Nadkarni V, Song JH, Gabbay KH, Owerbach D (2004) A M55V polymorphism in a
novel SUMO gene (SUMO-4) differentially activates heat shock transcription factors and is
associated with susceptibility to type I diabetes mellitus. J Biol Chem 279:27233-27238.
Cimarosti H, Henley JM (2008) Investigating the mechanisms underlying neuronal death in
ischemia using in vitro oxygen-glucose deprivation: potential involvement of protein
SUMOylation. Neuroscientist 14:626-636.
Cuomo O, Gala R, Pignataro G, Boscia F, Secondo A, Scorziello A, Pannaccione A, Viggiano D,
Adornetto A, Molinaro P, Li XF, Lytton J, Di Renzo G, Annunziato L (2008) A critical role
for the potassium-dependent sodium-calcium exchanger NCKX2 in protection against focal
ischemic brain damage. J Neurosci 28:2053-2063.
Dirnagl U, Simon RP, Hallenbeck JM (2003) Ischemic tolerance and endogenous neuroprotection.
Trends Neurosci 26:248-254.
Franklin KBJ, Paxinos G (1997) The mouse brain in stereotaxic coordinates. San Diego: Academic
Press.
Frerichs KU, Hallenbeck JM (1998) Hibernation in ground squirrels induces state and speciesspecific tolerance to hypoxia and aglycemia: an in vitro study in hippocampal slices. J Cereb
Blood Flow Metab 18:168-175.
Gidday JM (2006) Cerebral preconditioning and ischaemic tolerance. Nat Rev Neurosci 7:437-448.
Gill G (2003) Post-translational modification by the small ubiquitin-related modifier SUMO has big
effects on transcription factor activity. Curr Opin Genet Dev 13:108-113.
Hay RT (2005) SUMO: a history of modification. Mol Cell 18:1-12.
Kirino T (2002) Ischemic tolerance. J Cereb Blood Flow Metab 22:1283-1296.
Lee YJ, Castri P, Bembry J, Maric D, Auh S, Hallenbeck JM (2009) SUMOylation participates in
induction of ischemic tolerance. J Neurochem 109:257-267.
Lee YJ, Mou Y, Maric D, Klimanis D, Auh S, Hallenbeck JM (2011) Elevated global
SUMOylation in Ubc9 transgenic mice protects their brains against focal cerebral ischemic
damage. PLoS One 6:e25852.
Longa EZ, Weinstein PR, Carlson S, Cummins R (1989) Reversible middle cerebral artery
occlusion without craniectomy in rats. Stroke 20:84-91.
Molinaro P, Cuomo O, Pignataro G, Boscia F, Sirabella R, Pannaccione A, Secondo A, Scorziello
A, Adornetto A, Gala R, Viggiano D, Sokolow S, Herchuelz A, Schurmans S, Di Renzo G,
Annunziato L (2008) Targeted disruption of Na+/Ca2+ exchanger 3 (NCX3) gene leads to a
worsening of ischemic brain damage. J Neurosci 28:1179-1184.
Molinaro P, Cantile M, Cuomo O, Secondo A, Pannaccione A, Ambrosino P, Pignataro G, Fiorino
F, Severino B, Gatta E, Sisalli MJ, Milanese M, Scorziello A, Bonanno G, Robello M,
Santagada V, Caliendo G, Di Renzo G, Annunziato L (2013) Neurounina-1, a novel
compound that increases Na+/Ca2+ exchanger activity, effectively protects against stroke
damage. Mol Pharmacol 83:142-156.
16
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
Mukhopadhyay D, Dasso M (2007) Modification in reverse: the SUMO proteases. Trends Biochem
Sci 32:286-295.
Pichler A, Melchior F (2002) Ubiquitin-related modifier SUMO1 and nucleocytoplasmic transport.
Traffic 3:381-387.
Pignataro G, Meller R, Inoue K, Ordonez AN, Ashley MD, Xiong Z, Gala R, Simon RP (2008) In
vivo and in vitro characterization of a novel neuroprotective strategy for stroke: ischemic
postconditioning. J Cereb Blood Flow Metab 28:232-241.
Pignataro G, Boscia F, Esposito E, Sirabella R, Cuomo O, Vinciguerra A, Di Renzo G, Annunziato
L (2012) NCX1 and NCX3: two new effectors of delayed preconditioning in brain ischemia.
Neurobiol Dis 45:616-623.
Pignataro G, Gala R, Cuomo O, Tortiglione A, Giaccio L, Castaldo P, Sirabella R, Matrone C,
Canitano A, Amoroso S, Di Renzo G, Annunziato L (2004) Two sodium/calcium exchanger
gene products, NCX1 and NCX3, play a major role in the development of permanent focal
cerebral ischemia. Stroke 35:2566-2570.
Rajan S, Plant LD, Rabin ML, Butler MH, Goldstein SA (2005) Sumoylation silences the plasma
membrane leak K+ channel K2P1. Cell 121:37-47.
Ren X, Philipson KD (2013) The topology of the cardiac Na(+)/Ca(2)(+) exchanger, NCX1. J Mol
Cell Cardiol 57:68-71.
Scorziello A, Savoia C, Sisalli MJ, Adornetto A, Secondo A, Boscia F, Esposito A, Polishchuk EV,
Polishchuk RS, Molinaro P, Carlucci A, Lignitto L, Di Renzo G, Feliciello A, Annunziato L
(2013) NCX3 regulates mitochondrial Ca(2+) handling through the AKAP121-anchored
signaling complex and prevents hypoxia-induced neuronal death. J Cell Sci 126:5566-5577.
Seeler JS, Dejean A (2003) Nuclear and unclear functions of SUMO. Nat Rev Mol Cell Biol 4:690699.
Silveirinha V, Stephens GJ, Cimarosti H (2013) Molecular targets underlying SUMO-mediated
neuroprotection in brain ischemia. J Neurochem 127:580-591.
Sisalli MJ, Secondo A, Esposito A, Valsecchi V, Savoia C, Di Renzo GF, Annunziato L, Scorziello
A (2014) Endoplasmic reticulum refilling and mitochondrial calcium extrusion promoted in
neurons by NCX1 and NCX3 in ischemic preconditioning are determinant for
neuroprotection. Cell Death Differ 21:1142-1149.
Yang W, Sheng H, Warner DS, Paschen W (2008a) Transient global cerebral ischemia induces a
massive increase in protein sumoylation. J Cereb Blood Flow Metab 28:269-279.
Yang W, Sheng H, Warner DS, Paschen W (2008b) Transient focal cerebral ischemia induces a
dramatic activation of small ubiquitin-like modifier conjugation. J Cereb Blood Flow Metab
28:892-896.
Yang W, Sheng H, Homi HM, Warner DS, Paschen W (2008c) Cerebral ischemia/stroke and small
ubiquitin-like modifier (SUMO) conjugation--a new target for therapeutic intervention? J
Neurochem 106:989-999.
Zhao J (2007) Sumoylation regulates diverse biological processes. Cell Mol Life Sci 64:3017-3033.
560
561
562
17
563
FIGURE LEGENDS
564
565
566
Figure
1:
567
preconditioning, and tMCAO+preconditioning in the ipsilateral temporoparietal
568
cortex
569
Western Blot analysis of SUMO1 conjugation pattern after tMCAO, preconditioning, and
570
tMCAO+preconditioning in ipsilateral temporoparietal cortex are represented. Data were
571
normalized on the basis of actin levels and expressed as percentage of sham-operated
572
controls, taken as 100%. Values represent meansSEM (n=6). *p<0.05, compared to
573
control animals. Representative blots are on the left.
574
Figure
575
preconditioning, and tMCAO+preconditioning in striatum
576
Western Blot analysis of SUMO1 conjugation pattern after tMCAO, preconditioning, and
577
tMCAO+preconditioning stimulus in striatum (B) are represented. Data were normalized on
578
the basis of actin levels and expressed as percentage of sham-operated controls, taken as
579
100%. Values represent meansSEM (n=6). *p<0.05, compared to control animals.
580
Representative blots are on the left.
581
Figure 3: SUMO1 and NCX3 expression following tMCAO, preconditioning, and
582
tMCAO+preconditioning in ipsilateral temporoparietal cortex by confocal analysis
583
Confocal microscopic images displaying SUMO1 (A-M) (green), NCX3 (B-N) (red),
584
HOECHST (C-O) (blue), and Merge (D-P) (yellow) in the temporoparietal cortex region of
585
rats exposed to tMCAO, preconditioning, or preconditioning+tMCAO after 5 hours of
586
reperfusion.
587
A representative brain slice cartoon indicating the area of interest at the top of the Figure.
588
Scale bars in A-P: 50 m
589
Figure 4: SUMO1 and NCX3 expression following tMCAO, preconditioning, and
590
tMCAO+preconditioning in striatum by confocal analysis
591
Confocal microscopic images displaying SUMO1 (A-M) (green), NCX3 (B-N) (red),
592
HOECHST (C-O) (blue), and Merge (D-P) (yellow) in striatum of rats exposed to tMCAO,
593
preconditioning, or preconditioning+tMCAO after 5 hours of reperfusion. A representative
594
brain slice cartoon indicating the area of interest is at the top of the Figure. Scale bars in
595
A-P: 50 m
2:
Time-course
Time-course
of
of
SUMO1
SUMO1
conjugation
conjugation
596
18
pattern
pattern
following
following
tMCAO,
tMCAO,
597
598
Figure 6: SUMO1-NCX3 colocalization in cortical neurons
599
Confocal microscopic images displaying SUMO1 (green), NCX3 (red), HOECHST (blue),
600
and merge (yellow) in primary cortical neurons. Scale bars: 10 m
601
Figure 7: SUMO1 and NCX3 interaction in BKH cells transfected with NCX3 mutants
602
(A) Schematic representation of NCX3 mutants used for the immunoprecipitation assay.
603
(B) Immunoprecipitation experiments on WT BHK cells and in BHK cells transiently
604
transfected with NCX3 flag (NCX3F), or NCX3F lacking f-loop (NCX3 f). Cell lysates were
605
immunoprecipitated with SUMO1 and immunoblotted for anti-flag antibody. Both total
606
extracts and immunoprecipates were divided into two parts and each part was run in 8 or
607
12% gel. (C) Immunoprecipitation experiments on WT BHK cells and in BHK cells
608
transiently transfected with NCX3F , NCX3F  NCX3F K456E, or NCX3F FK489LE.
609
Cell lysates were immunoprecipitated with SUMO1 and immunoblotted for anti-flag
610
antibody.
611
Figure 7: SUMO1 and NCX3 interaction in brain tissue and effect of SUMO1 silencing
612
on NCX3 expression after preconditioning+ischemia
613
(A)
614
Brain tissue from control and ischemic rats was immunoprecipitated for NCX3 and then
615
blotted for SUMO1. The same filter was then incubated with NCX3 antibody, to confirm the
616
results. (B) Western Blot analysis of NCX3 after preconditioning+tMCAO in cortex of rats
617
previously treated with siRNA-A targeting SUMO1 and in rats treated with siRNA control
618
and exposed to preconditioning+tMCAo. The data were normalized on the basis of actin
619
levels and expressed as percentage of sham-operated controls, taken as 100%. Values
620
represent meansSEM (n=6). *p< 0.05, compared to control animals.
621
Figure 8: Effect of siRNA targeting SUMO1 on ischemic damage induced in male
622
rats by tMCAO and by preconditioning+tMCAO
623
(A) Representative blots and densitometric analysis of SUMO1 protein levels after icv
624
infusion of siRNA control, siRNA-A, or siRNA-B in rat cortex and striatum. Data were
625
normalized on the basis of -actin levels and represented as % of SUMO1 expression in
626
control animals. Values represent meansSEM (n=3). *p<0.05 versus SUMO1 levels in
627
controls. (B) Effect of siRNA control or siRNA-A on infarct volume induced by tMCAO.
628
Each column represents the meanSEM of the percentage of the infarct volume. siRNA-A
629
(10l, 10M) or siCTL were icv administered twice a day starting from 72 hours before or
Immunoprecipitation experiments for the analysis of NCX3-SUMO1 interaction.
19
630
tMCAO induction. n=5-8 animals in each group. *p<0.05 versus siRNA control treated
631
group. (C) Effect of siRNA control or siRNA-A on neuroprotection mediated by
632
preconditioning+tMCAO. Each column represents the meanSEM of the percentage of the
633
infarct volume. siRNA-A (10l, 20M) or siCTL were icv administered twice a day starting
634
from 72 hours before preconditioning induction. n=5-8 animals in each group. *p<0.05
635
versus siRNA control treated group.
636
20