Th

GFMa

b

c

d

e

f

a

ARRA

KSAIHHM

Cacsifwts

sbh[T

0d

Neuroscience Letters 492 (2011) 33–38

Contents lists available at ScienceDirect

Neuroscience Letters

journa l homepage: www.e lsev ier .com/ locate /neule t

in chloride enhances parvalbumin-positive interneuron survival by modulatingeme metabolism in a model of cerebral ischemia

iovanni Li Volti a,b, Agata Zappalàc, Gian Marco Leggiod, Carmen Mazzolad, Filippo Dragod,rancesco La Deliac, Maria Francesca Serapidec, Rosalia Pellitteri e, Ignazio Giannonef,ichela Spatuzzae, Valentina Ciciratac, Federico Ciciratac,∗

Department of Drug Sciences, Section of Biochemistry, University of Catania, Viale A. Doria 6, 95125 Catania, ItalyOasi Institute for Research on Mental Retardation and Brain Aging, Troina 94018, ItalyDepartment of Physiological Science, University of Catania, Viale A. Doria 6, 95125 Catania, ItalyDepartment of Clinical and Molecular Biomedicine, Section of Pharmacology, University of Catania, Viale A. Doria 6, 95125 Catania, ItalyInstitute of Neurological Sciences - Section of Catania, National Research Council, Viale Regina Margherita 6, 95123 Catania, ItalyDepartment of Pharmaceutical Sciences, University of Catania, Viale A. Doria 6, 95125 Catania, Italy

r t i c l e i n f o

rticle history:eceived 15 October 2010eceived in revised form 19 January 2011ccepted 19 January 2011

eywords:trokentioxidants

a b s t r a c t

SnCl2 has been reported to increase the expression of heme-oxygenase 1 (HO-1), a major antioxidantenzyme, and to decrease ischemic injury, in non-nervous tissues. This study examined the neuroprotec-tive effect of SnCl2 in the hippocampus of rats submitted to cerebral ischemia. SnCl2 was administered 18 hbefore bilateral carotids obstruction. Changes in HO-1 expression and activity, heme content, induciblenitric oxide synthase (iNOS) expression and parvalbumin positive interneuron survival were studied.Thereafter both behavior and memory recovery were tested. The administration of SnCl2 increased theexpression of HO-1 protein and HO activity in the hippocampus and concomitantly decreased heme con-

schemia/reperfusioneme oxygenaseippocampusemory

tent at both mitochondrial and nuclear level. Furthermore, ischemized animals showed a strong increasein iNOS expression in the hippocampus, where a loss of parvalbumin positive interneurons also occurred.Pre-treatment with SnCl2, decreased both iNOS expression in ischemized rats and increased cell survival.The beneficial effects of SnCl2 were prevented by concomitant treatment with SnMP, a strong inhibitorof HO activity. SnCl2 also caused an improvement in short term memory recovery. Our results showedthat following SnCl2 administration, HO-1 is strongly induced in the hippocampus and modulate iNOS

stron

expression, resulting in a

erebral ischemia leads to the increase of extracellular glutamatend activation of its receptors, with a subsequent increase in intra-ellular calcium and overproduction of nitric oxide (NO) by NOynthase enzymes (NOS) [9]. The role of iNOS, the inducible NOSsoform, is provided by the observation that mice lacking the geneor iNOS have significantly reduced infarct volumes compared withild-type controls [11]. Among the cellular protective mechanisms

riggered by oxidative stress, the heme oxygenase (HO) systemeems to play a major role.

The neuroprotective effect of HO is related to the conver-ion of free heme, a pro-oxidant molecule, into antioxidant end

yproducts: carbon monoxide and biliverdin [20]. Two HO isoformsave been shown to be catalytically active in heme degradation19]. HO-2 is constitutively expressed, whereas HO-1 is inducible.his latter isoform, was recently implicated in basal [6] and

∗ Corresponding author. Tel.: +39 095 7384036; fax: +39 095 7384220.E-mail address: cicirata@unict.it (F. Cicirata).

304-3940/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.neulet.2011.01.048

g neuroprotective effect.© 2011 Elsevier Ireland Ltd. All rights reserved.

inducible [5] resistance to NO toxicity in mouse motor neuronsand astroglial cells [1] and in various experimental models, includ-ing ischemia/reperfusion (I/R) injury [13]. Various pharmacologicalagents have been used so far in order to upregulate HO-1 in differ-ent organs. In particular, metals have been shown to be particularlyeffective in increasing HO-1 expression and activity [15].

The aim of this study was identifying SnCl2 as a critical mod-ulator of the HO system in the brain and elucidating its role as aneuroprotectant in an experimental model of cerebral ischemia inrat.

Male Wistar rats (Harlan, Milan, Italy) weighing 270–290 gwere used. The animals were kept under standardized tempera-ture, humidity and light conditions with free access to food andwater. Animal care and use followed the directives of the Council

of the European Community (86/609/EC). SnCl2 (10 mg/100 g b.w.)was administered 18 h before induction of cerebral ischemia. Tinmesoporphyrin (SnMP) (Frontier Scientific Europe, Carnforth, UK)a potent HO activity inhibitor, was used at a dose of 10 mg/kg. Ratswere anesthetized by an intraperitoneal injection of ethyl urethane

3 ience Letters 492 (2011) 33–38

(vcncai7t

aa1pbi(iaprce

dir5

mwHd

(paf(u�

Bfiolii

[wiscd

itmcrct

fd

Fig. 1. (A) ELISA detection of HO-1 protein in the hippocampus of vehicle and SnCl2

4 G. Li Volti et al. / Neurosc

1.2 g/kg b.w.) and then subjected to transient ischemia as pre-iously described [4]. A midline neck incision was made and theommon carotid arteries were exposed, separated from the vagalerves, occluded with 8 to 0 silk sutures, and ligated under a surgi-al microscope. Each animal’s rectal temperature was maintainedt 37.5 ± 0.5 ◦C throughout the experiment using an electric heat-ng pad. Following 20 min occlusion reperfusion was allowed for2 h. The wound was sutured and animals were then placed withheir dams on a warm heating pad (35–37 ◦C) for recovery.

Cryostat sections were washed three times in PBS, blockedt room temperature for 1 h in 1% normal goat serum in PBS,nd then incubated overnight with rabbit polyclonal anti-HO-(1:200) (Enzo Life Sciences, Plymouth Meeting, PA) or rabbit

olyclonal anti-iNOS (1:200) (Chemicon, Temecula, CA). After incu-ating with the primary antibodies, the slides were washed and

ncubated with anti-rabbit HRP-conjugated secondary antibody1:1000, Chemicon). In a separate set of experiments, a doublemmunohistochemistry was performed using mouse monoclonalnti-parvalbumin (1:200) (Abcam, Cambridge, UK) and rabbitolyclonal anti-iNOS (1:200) antibodies followed by anti-mousehodamine-conjugated (1:1000, Chemicon) and anti-rabbit FITC-onjugated (1:1000, Chemicon) secondary antibodies in order tovaluate specific hippocampal interneuron following ischemia.

The intensity of iNOS staining was calculated as previouslyescribed [18]. Digitally fixed images of parvalbumin positive

nterneurons at 100× magnification were analyzed under a fluo-escence microscope equipped with an image analyzer and optimas.2 software (Optimas Corp., CA, USA).

Mitochondrial and nuclear fractions were obtained using a com-ercially available kit (Pierce Biotechnology, Rockford, IL). Hemeas extracted from 40 �l of organelles fraction with 300 �l ofCl/acetone (250 �l of HCl in 10 ml of acetone) as previouslyescribed [3].

A commercially available enzyme linked immunosorbent assayELISA) kit (Stressgen, Victoria, Canada) was used to measure HO-1rotein concentration in the hippocampus. HO activity was assayeds previously described [17] by extracting bilirubin with chloro-orm and determining its concentration spectrophotometricallyPerkin-Elmer Dual UV/VIS Beam Spectrophotometer Lambda 25)sing the difference in absorbance at a wavelength from � 460 to530 nm with an absorption coefficient of 40 mmol−1 cm−1.

Open Field test was performed as previously described [21].riefly, spontaneous motor activity was scored in a circular openeld arena. The behavior of each animal was observed for a periodf 5 min and the occurrence of these conditions was recorded:ocomotion, rearing and grooming. Open Field test was performedn order to evaluate animal motor abilities following cerebralschemia/reperfusion.

Passive avoidance test was performed as previously described16]. The apparatus for the step-through passive-avoidance testas an automated shuttle-box (Ugo Basile, Milan, Italy) divided

nto an illuminated compartment and a dark compartment of theame size by a wall with a guillotine door, and the floor was a gridapable of delivering a controlled foot-shock (0.5 mA intensity, 2 suration).

Retention of the passive avoidance response was measured dur-ng the retention tests performed 1 and 7 days after the learningrial. Each animal was again placed into the illuminated compart-

ent, the door was opened, and the latency to re-enter the darkompartment was recorded. No foot-shock was delivered whenetention tests were performed. If the rat failed to enter the dark

ompartment within 300 s during retention testing, the trial waserminated and the latency was recorded as 300 s.

One-way analysis of variance (ANOVA) followed by Bon-erroni’s t-test was performed in order to estimate significantifferences among groups. Regarding behavioural tests, the

treated animals. (B) HO activity measurement in the hippocampus of vehicle andSnCl2 treated animals. (C) Nuclear and mitochondrial (D) levels of heme content inthe hippocampus of vehicle and SnCl2 treated animals. Data are presented as themean ± S.E.M. (*p < 0.01 vs vehicle).

results were expressed as mean ± S.E.M. The data for eachexperiment was analyzed using one-way ANOVA followed byTukey’s post hoc procedure for multiple comparisons. Differ-ences between groups were considered to be significant atp < 0.05.

ELISA test showed that hippocampal HO-1 protein contentincreased up to 3.2 ± 0.4 folds in SnCl2 non-ischemized animalwhen compared to vehicle group (Fig. 1A). Consistently, SnCl2 treat-ment increased HO activity up to 3.8 ± 0.2 folds compared to controlbasal activity (Fig. 1B). Furthermore, we observed a concomitantreduction of heme content in both nuclear (−50%, Fig. 1C) and mito-chondrial (−60%, Fig. 1D) compartments following SnCl2 treatment.

These results were further confirmed by immunohistochemi-cal analysis showing that SnCl2 increased HO-1 expression in allhippocampal regions (Fig. 2A and B).

The hippocampus of ischemized animals showed increasediNOS staining (Fig. 2D and G) compared to sham operated animals(Fig. 2C and G). Strikingly, SnCl2 treatment lowered the expressionof iNOS (Fig. 2E and G) thus suggesting that SnCl2 plays a stronginhibitory role on the expression of iNOS. To further confirm therole of the HO system in such regulation, in a separate set of exper-iments, animals were also treated with SnMP, a strong inhibitorof HO activity. These set of experiments showed that SnMP com-pletely reversed the inhibitory effect of SnCl2 on iNOS expression(Fig. 2F and G). Double immunofluorescence analysis confirmedthat the iNOS positive cells in the hippocampus were parvalbuminpositive interneurons (Fig. 3).

We further tested the cellular density of the hippocampus, 72 hafter transient ischemia in both vehicle and SnCl2 treated animals.Vehicle treated animals submitted to transient ischemia showeda dramatic decrease of PV-positive neuronal cells throughout thehippocampus (number of hippocampal interneurons 85 ± 2, 62%

of PV positive cells) (Figs. 4B and 5B) when compared to shamoperated animals (number of hippocampal interneurons 137 ± 3)(Figs. 4A and 5A). The decrease was more evident in the CA1, wherePV-positive cells of transient ischemia animals corresponded to38.1% (number of hippocampal interneurons 29 ± 3) in respect of

G. Li Volti et al. / Neuroscience Letters 492 (2011) 33–38 35

Fig. 2. Photomicrographs of HO-1 distribution in hippocampus of vehicle (A) and SnCl2 treated (B) rats. Transient ischemia injury (C) resulted in a significant increase ofiNOS positive neurons when compared to sham operated animals (D). Our data also showed that SnCl2 (10 mg/kg b.w. i.p.) treatment (E) significantly reduced iNOS positiveneurons. This effect was abolished by SnMP (10 mg/kg) (F). Integrated optical density showing iNOS expression quantification in different experimental groups (G) (*p < 0.01when compared to sham operated; **p < 0.01 when compared to vehicle; ***p < 0.01 when compared to SnCl2).

36 G. Li Volti et al. / Neuroscience Letters 492 (2011) 33–38

F neuroc nd, th

tpSp1ocwiPp(

wwSii

Fpc

ig. 3. Photomicrographs of double immunostaining of the CA1 hippocampal interence, B). Merge (C). (For interpretation of the references to color in this figure lege

he cells of the CA1 of sham operated animals (number of hip-ocampal interneurons 76 ± 3). Interestingly, rats pretreated withnCl2 (Figs. 4C and 5C), showed a less severe depletion of hip-ocampal PV positive cells (number of hippocampal interneurons01 ± 4). In particular, CA1 of SnCl2 treated animals showed 75%f PV positive cells (number of hippocampal interneurons 57 ± 2)ompared to untreated animals. The beneficial effects of SnCl2ere abolished by SnMP treatment (number of CA1 hippocampal

nterneurons 30 ± 4, 39.5% of PV positive interneuronal cell) and ofV positive cells throughout hippocampus (number of hippocam-al interneurons 80 ± 3, 58.4% of PV interneuronal cell survival)Figs. 4D and 5D).

Transient ischemia induced low locomotion level (Fig. 6A),hich was significantly increased up to +254% by SnCl2 (p < 0.05)

hen compared to vehicle. Consistent with previous results,

nCl2/SnMP co-treatment in transient ischemia rats prevented thencrement of locomotion activity mediated by SnCl2 (Fig. 6A). Rear-ng and grooming behavior in transient ischemia rats showed no

ig. 4. Photomicrographs of parvalbumin positive interneurons in CA1 hippocampal regioositive cells in the CA1 region. (C) Prevented hippocampal interneurons cell loss followonfirmed by the co-administration of SnCl2 and SnMP, a potent HO activity inhibitor, wh

ns showing the co-localization of iNOS (green fluorescence, A) and PV (red fluores-e reader is referred to the web version of the article.)

significant differences in all animals groups (Fig. 6B and C). Eventhough, animal showed different locomotion activity, no motordeficit was observed in all experimental groups.

In the Passive Avoidance test, SnCl2 treatment did not affect thelatency time of rats to re-enter the dark compartment in both Rt1and Rt2 compared to vehicle treated rats (Fig. 6D). In sham operatedrats, SnCl2 treatment increased up to +71% (p < 0.05) the perfor-mance, while SnCl2/SnMP co-administration prevented (+29.3%)the improved behavior induced by SnCl2 when compared to vehicle(Fig. 6E).

Following transient ischemia, passive avoidance test in vehi-cle treated rats showed a significant reduction (−64.2%) (p < 0.05)of latency time when compared to the corresponding sham oper-ated group. By contrast, SnCl2 dramatically improved (+467%;

p < 0.01) the performance with respect to vehicle treated rats. Con-sistent with our previous biochemical data, the co-administrationof SnCl2/SnMP strongly antagonized the positive effects of SnCl2(Fig. 6H).

n following transient ischemia injury. (A) Sham operated animals. (B) Parvalbumining transient ischemia. The involvement of the HO system in this mechanism wasich completely prevented the effects of SnCl2 (D).

G. Li Volti et al. / Neuroscience Letters 492 (2011) 33–38 37

Fig. 5. Drawings representing parvalbumin positive interneuron distribution in different hippocampal regions in sham operated animals (A) and following transient ischemiainjury (B). Our data showed that SnCl2 (C) selectively attenuated CA1 hippocampal interneuron loss following transient ischemia injury. The beneficial effects of SnCl2 wereprevented by the concomitant treatment with SnMP (10 mg/kg) (D).

Fig. 6. SnCl2 improved locomotion of transient ischemia rats compared to vehicle group (*p < 0.05) (A). Similarly, no significant changes were observed for rearing (B) andgrooming (C). Effects of SnCl2 (10 mg/kg b.w. i.p.) on intact rats in the passive avoidance test. No significant changes were observed following SnCl2 treatment (D). Effectsof SnCl2 on ischemic rats tested in the passive avoidance test (E) (***p < 0.01) and its effect was prevented by SnMP (10 mg/kg) (#p < 0.05). In the second retention test,SnCl2 + SnMP transient ischemia group showed a significantly decreased latency compared to vehicle and SnCl2 transient ischemia groups (##p < 0.01).

3 ience L

ttrfismraetr

eif

motCitthvitsisbciea

rhemrImsapitipdHti

e

[

[

[

[

[

[

[

[

[

[

[

8 G. Li Volti et al. / Neurosc

Regarding Rt2, vehicle treated rats of both sham operated andransient ischemia groups showed quite similar scores. The latencyime of transient ischemia rats was higher compared to the cor-esponding group of transient ischemia rats in Rt1 (Fig. 6H). Thisnding showed that the retention of long term memory improvedpontaneously up to normal values after 12 days of recovery. Treat-ent with SnCl2 in both sham operated and transient ischemia

ats did not change significantly the latency time of the passivevoidance test compared to the respective controls (Fig. 6H). How-ver, co-treatment with SnCl2 and SnMP in both sham operated andransient ischemia groups strongly reduced the latency time withespect to both vehicle and SnCl2 treated rats (Fig. 6H).

This study showed that SnCl2 strongly increased HO-1xpression in neuronal cells, prevented PV-positive hippocampalnterneuron loss and improved performance of short term memoryollowing ischemia/reperfusion injury.

HO-1 induction was associated with significant variations initochondrial and nuclear heme pool. In fact, significant decrement

f heme content at mitochondrial level was observed in SnCl2-reated animals. This finding is consistent with previous studies ofonverso et al. [8] showing that administration of hemin, a potent

nducer of HO-1 expression and activity, results in significantranslocation of HO-1 protein to mitochondrial compartment, andhis accounted for the increased heme degradation. Also nucleareme pool was decreased following SnCl2 treatment. This obser-ation is noteworthy since nuclear HO-1 has been shown to bencreased under various experimental conditions including exci-otoxic injury [12] and therefore it has been suggested that it mayerve as a regulator of important genes involved in cell survival. Thenvolvement of the HO system in iNOS regulation following tran-ient ischemia is corroborated by the prevention of SnCl2 effectsy SnMP, a potent inhibitor of HO activity. These results are alsoonsistent with recent observations of Li Volti et al. [14], show-ng that SnCl2 treatment results in a significant decrease of iNOSxpression following renal I/R injury, reduction of oxidative stressnd apoptosis via the p38MAPK pathway.

We further evaluated whether SnCl2 mediated iNOS down-egulation resulted also in a significant increase in PV-positiveippocampal interneuron survival. PV is a cytosolic calcium bufferxpressed in GABAergic interneurons [2,10] and it is a superiorarker for a subpopulation of GABAergic interneurons [7]. Our

esults showed that SnCl2 pretreatment significantly decreased/R mediated loss of this interneuron subpopulation and that this

echanism is related to increased HO activity. Finally, our resultshowed that increased neuronal cell survival also translated inton improved functional performance of ischemized animals. Inarticular, we observed that SnCl2 prevents short term memory

mpairment (Rt1) and also in this case the mechanism is relatedo HO-1 induction. Interestingly, as far as long term memory (Rt2)s concerned, we did not observe any further recovery of memoryerformance following SnCl2 treatment, thus suggesting that HO-1id not further improve long term memory. However, inhibition of

O activity by SnMP impaired the physiological recovery of long

erm memory in ischemized rats, suggesting that HO-1 is requiredn the physiological mechanism underlying such recovery.

Taken all together, our data showed that SnCl2 regulates thexpression and functional activity of HO-1 in the brain and this

[

etters 492 (2011) 33–38

effect leads to a down-regulation of iNOS, thus preventing CA1PV-positive hippocampal interneuron loss and improving mem-ory retention. Furthermore, our results may provide the molecularbasis for the neuroprotective mechanism/s of important drugs usedin daily clinical practice to attenuate or prevent cerebral injury dueto transient ischemia.

References

[1] R. Acquaviva, A. Campisi, P. Murabito, G. Raciti, R. Avola, S. Mangiameli,I. Musumeci, M.L. Barcellona, A. Vanella, G. Li Volti, Propofol attenuatesperoxynitrite-mediated DNA damage and apoptosis in cultured astrocytes: analternative protective mechanism, Anesthesiology 101 (2004) 1363–1371.

[2] C. Andressen, I. Blumcke, M.R. Celio, Calcium-binding proteins: selective mark-ers of nerve cells, Cell Tissue Res. 271 (1993) 181–208.

[3] H. Atamna, K. Boyle, Amyloid-beta peptide binds with heme to form a perox-idase: relationship to the cytopathologies of Alzheimer’s disease, Proc. Natl.Acad. Sci. U.S.A. 103 (2006) 3381–3386.

[4] S.S. Baliga, K.M. Jaques-Robinson, N.M. Hadzimichalis, R. Golfetti, G.F. Mer-rill, Acetaminophen reduces mitochondrial dysfunction during early cerebralpostischemic reperfusion in rats, Brain Res. 1319 (2010) 142–154 (Epub 2010January 14).

[5] A. Bishop, J.C. Marquis, N.R. Cashman, B. Demple, Adaptive resistance to nitricoxide in motor neurons, Free Radic. Biol. Med. 26 (1999) 978–986.

[6] A. Bishop, S.F. Yet, M.E. Lee, M.A. Perrella, B. Demple, A key role for hemeoxygenase-1 in nitric oxide resistance in murine motor neurons and glia,Biochem. Biophys. Res. Commun. 325 (2004) 3–9.

[7] M.R. Celio, Parvalbumin in most gamma-aminobutyric acid-containing neuronsof the rat cerebral cortex, Science 231 (1986) 995–997.

[8] D.P. Converso, C. Taille, M.C. Carreras, A. Jaitovich, J.J. Poderoso, J. Boczkowski,HO-1 is located in liver mitochondria and modulates mitochondrial heme con-tent and metabolism, FASEB J. 20 (2006) 1236–1238.

[9] V.L. Dawson, T.M. Dawson, Free radicals and neuronal cell death, Cell DeathDiffer. 3 (1996) 71–78.

10] B. Hontanilla, A. Parent, H.S. de las, J.M. Gimenez-Amaya, Distribution of cal-bindin D-28k and parvalbumin neurons and fibers in the rat basal ganglia, BrainRes. Bull. 47 (1998) 107–116.

11] C. Iadecola, F. Zhang, S. Xu, R. Casey, M.E. Ross, Inducible nitric oxide syn-thase gene expression in brain following cerebral ischemia, J. Cereb. Blood FlowMetab. 15 (1995) 378–384.

12] G. Li Volti, R. Ientile, N.G. Abraham, A. Vanella, G. Cannavo, F. Mazza, M. Curro, G.Raciti, R. Avola, A. Campisi, Immunocytochemical localization and expressionof heme oxygenase-1 in primary astroglial cell cultures during differentiation:effect of glutamate, Biochem. Biophys. Res. Commun. 315 (2004) 517–524.

13] G. Li Volti, L.F. Rodella, C. Di Giacomo, R. Rezzani, R. Bianchi, E. Borsani,D. Gazzolo, R. Motterlini, Role of carbon monoxide and biliverdin in renalischemia/reperfusion injury, Nephron Exp. Nephrol. 104 (2006) e135–e139.

14] G. Li Volti, V. Sorrenti, P. Murabito, F. Galvano, M. Veroux, A. Gullo, R. Acqua-viva, A. Stacchiotti, F. Bonomini, L. Vanella, C. Di Giacomo, Pharmacologicalinduction of heme oxygenase-1 inhibits iNOS and oxidative stress in renalischemia-reperfusion injury, Transplant. Proc. 39 (2007) 2986–2991.

15] M.D. Maines, A. Kappas, Metals as regulators of heme metabolism, Science 198(1977) 1215–1221.

16] V. Micale, G.M. Leggio, C. Mazzola, F. Drago, Cognitive effects of SL65.0155 aserotonin 5-HT4 receptor partial agonist, in animal models of amnesia, BrainRes. 1121 (2006) 207–215.

17] R. Motterlini, A. Hidalgo, I. Sammut, K.A. Shah, S. Mohammed, K. Srai, C.J. Green,A precursor of the nitric oxide donor SIN-1 modulates the stress protein hemeoxygenase-1 in rat liver, Biochem. Biophys. Res. Commun. 225 (1996) 167–172.

18] L.F. Rodella, F. Ricci, E. Borsani, R. Rezzani, A. Stacchiotti, C. Mariani, R. Bianchi,Exposure to aluminium changes the NADPH-diaphorase/NPY pattern in the ratcerebral cortex, Arch. Histol. Cytol. 69 (2006) 13–21.

19] S.W. Ryter, J. Alam, A.M. Choi, Heme oxygenase-1/carbon monoxide: from basicscience to therapeutic applications, Physiol. Rev. 86 (2006) 583–650.

20] H.M. Schipper, W. Song, H. Zukor, J.R. Hascalovici, D. Zeligman, Heme

oxygenase-1 and neurodegeneration: expanding frontiers of engagement, J.Neurochem. 110 (2009) 469–485.

21] A. Tamburella, V. Micale, A. Navarria, F. Drago, Antidepressant properties ofthe 5-HT4 receptor partial agonist, SL65.0155: behavioral and neurochemi-cal studies in rats, Prog. Neuropsychopharmacol. Biol. Psychiatry 33 (2009)1205–1210.