Introduction

Stroke still constitutes a major cause of extensive health problems and mortality throughout the world (Donnan et al. 2008). Ischemic stroke is characterized by a sudden interruption of the blood flow to distinct brain region that is sufficient to alter normal cell functions with reversible or permanent neurological deficit. Although reperfusion therapy is the cornerstone of treatment in acute ischemic stroke, post-reperfusion lesion occurs due excessive generation of reactive oxygen species (ROS), leading to oxidative damage and apoptotic neuronal death (Fukuda and Badaut 2012).

Studies have shown that the cerebral ischemic injury is associated with a highly complex pathological process in which inflammation, oxidative stress and apoptosis are the major contributory pathways (Shirley et al. 2014).

Thus, therapeutic strategies that are aimed at preventing the ongoing oxidative and apoptotic neural damage and promoting neural regeneration offer an opportunity for neuroprotection. Evidence suggests that the use of antioxidant herbal medicines or their bioactive constituents may be an effective therapy for treating ischemic injuries (Wang et al. 2014; Xu et al. 2016; Pan et al. 2018).

Crocus sativus L. (saffron) is a herbaceous plant used in folk medicine for many centuries. Saffron and its metabolites have gained an increasing importance in the modern pharmacological studies owning to its health promoting potential (José Bagur et al. 2018). Studies over the past two decades have demonstrated the immense therapeutic properties of C. sativus as anti-inflammatory, anxiolytic, anti-tumor, anti-oxidant, antidepressant, anti-neurodegenerative, as well as learning and memory improving properties (Hosseinzadeh and Noraei 2009; Ramadan et al. 2012; Sadeghnia et al. 2013; Festuccia et al. 2014; Soeda et al. 2016).

Evidence obtained in experimental studies has suggested that saffron, which can counteract the neurochemical and neurobehavioral changes and improve neurological deficit, has a neuroprotective potential in ischemia/reperfusion (I/R) brain injury. This effect is probably through preserving the antioxidant capacity and regulating the apoptotic cell death (Saleem et al. 2006; Vakili et al. 2014; Sadeghnia et al. 2017).

Recent insight into the mechanisms involved in stroke indicates an important role of angiogenesis and neurogenesis as possible processes associated with stroke recovery (Xu et al. 2016). Vascular endothelial growth factor (VEGF) is a key angiogenic and neurotrophic factor in neural regeneration and angiogenesis (Xu et al. 2016; Pan et al. 2018). Accumulating data suggest that VEGF protects the brain against ischemic injury and mediates neurological recovery by reducing infarct volume and enhancing neurogenesis and neuromigration in rats (Herz et al. 2012; Xu et al. 2016; Pan et al. 2018). Therefore, VEGF has received attention as a potential target for angiogenic therapy after stroke (Navaratna et al. 2009; Xu et al. 2016).

Against this background, the current study aimed to investigate the antioxidant and antiapoptotic effects of saffron administration in rats with transient focal cerebral I/R brain injury and whether a modulation of VEGF expression is involved in the underlying mechanisms of saffron neuroprotection.

Materials and methods

Plant material

Crocus sativus (saffron) was procured from the local market (Agricultural Seeds, Spices and Medicinal Plants Co., Abd El-Rahman M. Harraz), and kept at 2–4 °C.

Preparation of saffron extract

Three hundred g of saffron stigmas were soaked for 3 days in 3 l of 80% ethanol at room temperature, and then the extract was filtered using Whatman® filter paper. Afterward, the filtrate was concentrated under vacuum using the rotatory evaporator (40 °C), and percolated several times till exhaustion (El-Alfy et al. 2012). The yield was about 120 g of dark red residues. The extract was later scanned for the active principals previously reported in saffron; dereplication, using LC-DAD/ESI-MS technique.

Sample preparation for LC/MS analysis

One mg of the bioactive extract of Crocus sativa was dissolved in 2 ml water. After membrane filtration (0.45 mm) an aliquot of the solution was subjected to the LC/MS system. The whole procedure was performed in dim light.

LC-DAD/ESI-Ms analysis

A XEVO-TQD#QCA423 HPLC system coupled via an ESI interface, consisting of an automatic sampler injector, a binary pump, a continuous vacuum degasser and a column heater-cooler. All the operations, the acquiring and analysis of data were controlled by MSD Trap Control Version 1.60.1897 software.

The chromatographic separation was performed on ACQUITY UPLC-BEH C18 1.7 μm (4.6 mm × 150 mm, 5 μm) column at a temperature of 30 ̊C. The mobile phase consisted of mixtures of water +0.1% formic acid (A) and methanol+0.1% formic acid (B), starting with 5 min isocratic at 90% A, followed by a gradient to obtain 70% A at 15 min, 70% B at 22 min, 90% B at 25 min and isocratic 100% B from 29 min to 32 min at a flow rate of 0.2 ml/min. All solvents were filtered through a 0.45 μm nylon filter prior to use. The mobile phase flow-rate was kept constantly at 0.2 mL/min.

The MS conditions were as follows: collision energy (Ampl), 0.6–1.0 V; collision gas, He; HV capillary voltage, 3.5 kV; nebulizer/ drying gas N2, 10 L/min; temperature, 350 °C; pressure of nebulizer gas, 40 psi; HV voltage, 3.5 kV; flow rate, 1.0 mL/min.

Tandem mass spectrometry (MS–MS)

Precursor ions were selected and fragmented in the collision cell applying collision energies in the range of 10–30 eV. Argon was used as collision gas. Product ions were detected using the following parameter settings: pulser frequency, 10 kHz; spectra rate, 1.5 Hz. For CID of in-source fragment ions, in-source CID energy was increased from 0 to 100 V. ESI source (electrospray voltage 4.0 kV, sheath gas: nitrogen; capillary temperature: 275 °C) in positive ionization modes.

The peaks and spectra were processed using the Maslynx 4.1 software and tentatively identified by comparing its retention time (Rt) and mass spectrum with reported data.

Compounds were characterized by their retention times, mass spectra and comparison to Mass Bank database and reference literature.

Experimental animals

Three-months-old male Wistar rats weighing 250–280 g were utilized. Rats were purchased from the Animal House Colony of the National Research Center (Dokki, Giza, Egypt). Rats were accommodated in a room with a 12-h light/dark cycle with free access to food and water all over the experimental period. Animals were allowed to adapt to the laboratory environment for one week before experimentation. The study was accomplished in accordance with the ethical procedures and policies approved by the Animal Care and Use Committee of National Research Centre (approval number: MREC-16-314) and following the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978).

Rat model of cerebral ischemia/reperfusion

Overnight fasted rats were anesthetized with 10% chloral hydrate (40 mg/kg, ip) and subjected to either left cerebral ischemia-reperfusion or sham operation. The left middle cerebral artery (MCA) was be occluded as described previously (Longa et al. 1989) with slight modifications. Briefly, the left carotid region was exposed through a midline neck incision; the external carotid artery (ECA) and the common carotid artery (CCA) were exposed. A non-traumatic micro vascular clip was introduced from the carotid bifurcation into the internal carotid artery (ICA); thereby occluding the origin of the MCA and a focal cerebral ischemia was conducted in the ICA. The body temperature of rats was maintained at 36.5–37.5 °C throughout the surgery using a heating lamp. After 60 min, recirculation was established by gentle withdrawal of the clamps. For the rats in the sham-operated controls, the left carotid region was only exposed. Finally, the incision was sutured; the animal was allowed to recover from anesthesia.

Blood pressure during ischemia and reperfusion was monitored as previously described (Yang et al. 1994). After blood reinfusion, clamp removal and wound closure, animals were placed in recovery cages with ambient temperature maintained at 22 °C.

Experimental design

Forty Wistar rats were divided into 4 groups, 10 rats each. The 1st two groups were served as sham (operated without cerebral artery occlusion and received only normal saline), and left brain I/R (operated with cerebral artery occlusion and treated with normal saline) groups. In the other 2 groups, Crocus sativus (saffron) extract, at doses of 100 and 200 mg/kg, respectively, were administered to rats through ip route for 3 successive weeks before subjected to left brain I/R and then administered four times (60 min before surgery, during the surgery, and on 1 day, 2 days after the I/R). After 72 h of brain I/R, animals were euthanized and their brains were removed. The left-brain hemispheres were homogenized in ice-cold saline (20% w/v) and were kept at −80 °C for biochemical analysis. Another left-brain hemispheres were kept in 10% formalin for further histopathological and immunohistochemical analysis.

Neurological assessment

Neurological deficit were assessed 24 h after surgery on a four-point scale (0 = no observable neurological deficit, 1 = failure to extend left forepaw fully, 2 = circling to the left, 3 = falling to the left, and 4 = no spontaneous walking with a depressed level of consciousness (Longa et al. 1989; El-Marasy et al. 2018). Rats scoring 1–3 points indicated successful model establishment. The person doing the scoring was blind to the treatment groups. Rats that did not show neurological deficits immediately after reperfusion (neurological score < 1) were excluded from the study.

Behavioral assessments/sensorimotor tests

Two behavioral assessment tests were performed after 24 h of the last drug administration. During behavioral assessments, unnecessary disturbance of animals was avoided. Besides, all animals were treated gently with avoiding tough maneuver.

Initiation of walking

Rats were placed on a flat surface, and the time for the animal to move 1 body length was recorded (Hattori et al. 2000). This test was conducted 3 times.

Locomotor balance and coordination

Motor coordination of rats was assessed using an accelerating rotarod (Model No. 7750; Ugo Basile), according to the procedure described before by (El-Marasy et al. 2018). Three training sessions at fixed speed of 4 rotations per minute (rpm) were done for 3 successive days before drug administration to rats. On the fourth day, the rats were placed on the testing rod and the speed of the rotarod started at 4 rpm and then increased gradually to reach 40 rpm over 300 s. The basal falling time for each rat was recorded using a cut off limit of 300 s. Twenty-four hours after the last administration of the tested extract, each rat was then re-placed on an accelerating rotarod for 300 s test session and the final falling time was recorded.

Biochemical investigation

Left brain hemispheres were homogenized in ice-cold saline (20% w/v) and kept at −80 °C for further measurement of malondialdehyde (MDA), reduced glutathione (GSH), nitric oxide (NO), brain naturetic peptide (BNP) and vascular endothelial growth factor (VEGF).

Thiobarbituric acid reactive species measurement

Brain lipid peroxides formation were measured as malondialdehyde (MDA), which is the end product of lipid peroxidation and reacts with thiobarbituric acid (TBA) as a TBA reactive substance (TBARS) to produce a pink colored complex which has peak absorbance at 535 nm as described previously (Ruiz-Larrea et al. 1994).

In brief, 0.5 ml of homogenate sample were mixed with 4.5 ml of TCA-TBA reagent (20% TCA, and 0.8% TBA, 3:1) and heated for 20 min in a boiling water bath. After cooling, the mixture was centrifuged at 3000 rpm for 10 min. The supernatants were collected, and the absorbance was read against blank (distilled water instead of sample), at 535 nm. The amount of MDA produced was calculated, using a molar absorption coefficient of 1.56 × 105 M−1 cm−1 and expressed as nmol/g tissue.

Determination of reduced glutathione level

The level of reduced glutathione (GSH) was determined as described previously (Ellman 1959) with modification (Bulaj et al. 1998).

In brief, brain homogenate was centrifuged at 4000 rpm for 5 min at 4 °C in a cooling centrifuge, and then 0.5 ml of the supernatant was added to 0.5 ml of trichloro-acetic acid (TCA) 10%. The mixture was vortex mixed and then centrifuged at 4000 rpm/5 min in a cooling centrifuge. In a clean test tube, 1.8 ml of phosphate buffer pH 8.0 was added to 0.1 ml of the supernatant and 0.1 ml of Ellman reagent. The absorbance was measured at 412 nm after exactly 5 min against blank (distilled water instead of sample). Glutathione level was calculated using the extinction coefficient of 1.36 × 104 M–1 cm–1. The results were expressed in μmol GSH/g tissue.

Determination of nitric oxide level

Nitric oxide (NO) was determined in rat brain homogenate according to the method described previously (Miranda et al. 2001), using commercial kit (Biodiagnostic, Egypt).

Determination of brain natriuretic peptide level

Rat Brain Natriuretic Peptide (BNP) was measured in brain homogenates using commercial ELISA kit, Elabscience Biotechnology Co., Ltd-USA.

Determination of VEGF level

Rat VEGF (vascular endothelial cell growth factor) was measured in brain homogenates using commercial ELISA Kit, MyBioSource, Inc., San Diego, USA.

Histopathological examination

The brain tissues from the different groups were fixed in 10% neutral buffered formalin and processed for paraffin embedding to obtain 4 μm sections. The sections were stained with hematoxylin and eosin (H&E) and examined microscopically (Bancroft and Gamble 2008; Mostafa et al. 2017).

Immunohistochemistry for caspase 3, Bax proteins and VEGF

The immunohistochemistry was performed following the methods described (Ogaly et al. 2018; Abdel-Rahman et al. 2017). Brain tissue sections were deparaffinized in xylene and rehydrated in graded alcohol. The tissues were pretreated with 10 mM citrate buffer, pH 6.0 in microwave oven at 500 W for 10 min for antigenic retrieval. The slides were washed with PBS, and blocked with ultra V Blocking solution (Thermo scientific, USA) for 5 min. Sections were incubated for 2 h at 4 °C in a humidified chamber with one of the following primary antibodies: rabbit anti-caspase-3 polyclonal antibody at 1:50 dilution (ab13847; Abcam, Cambridge, UK) and rabbit anti-Bax monoclonal antibody (ab32503; Abcam, Cambridge, UK) at a 1:250 dilution. The sections were incubated with a biotinylated goat anti rabbit antibody (Thermo scientific, USA) for 10 min. Finally, sections were incubated with Streptavidin peroxidase (Thermo scientific, USA). To visualize the reaction, slides were incubated for 10 min with 3,3′-diaminobenzidine tetrahydrochloride (DAB, Sigma). The slides were counterstained with Mayer’s haematoxylin then dehydrated and mounted. Primary antibodies were omitted and replaced by PBS for negative controls.

The quantitative immunoreactivity of caspase-3, Bax and VEGF was evaluated in neurons in different areas of the brain tissue. Caspase-3and Bax stained with dark brown color in the cytoplasm of the cell was considered an apoptotic cell. Percentage of positive stained area (%) was calculated as mean of 10 fields/slide by using Lieca Qwin 500 Image Analyzer (Leica, Cambridge, England).

Statistical analysis

The results are expressed as mean ± S.E. of seven animals, and all statistical comparisons were made by means of one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison tests. The data were analyzed with GraphPad Prism v. 5.0 (GraphPad Software, Inc., CA, USA). Difference was considered significant when p value is ≤0.05.

Results

LC-MS analysis

According to LC-MS/MS analysis, four compounds were tentatively identified in the active saffron extract using their mass characteristics. Crocin and crocetin derivatives (picrocrocin; Crocetin gentiobiosyl, glucosyl; Crocetin+2 Glu) were detected in saffron extract (Table 1 and Fig. 1).

Table 1 Identified compounds in saffron extract and their relative percentage obtained from LC-MS.
Full size table
Fig. 1
figure 1

ESI mass spectra of compound 1 (a), 2 (b), 3 (c), and 4 (d)

Full size image

Effect of saffron on neurological deficits

Animals subjected to 60 min of cerebral ischemia revealed a significant increase in the latency to move one body length compared with the sham group. Animals treated with saffron revealed significant decrease in latency to move their bodies compared to rats in the control untreated group. Moreover, there was a significant reduction in the latency to fall in the left brain I/R group as compared to the sham group. While saffron treated groups revealed latency to fall nearly as animals in the sham group (Table 2).

Table 2 Effect of saffron on sensorimotor behavior in I/R rats
Full size table

Effect of saffron on MDA, NO and GSH content in I/R injured brain

Table 3 shows the level of MDA (nmol/g tissue), GSH (μg/g tissue) and nitric oxide (NO) in I/R rats’ brains. I/R injury resulted in significant increase in MDA and NO content and significantly decreased brain tissue content of GSH. In saffron treated group, the brain content of MDA and NO were significantly decreased compared to I/R group. However, treatment of rats with saffron extract significantly increased the GSH content.

Table 3 Effect of saffron on brain content of MDA (nmol/g tissue), GSH (μg/g tissue) and NO (μmol/g tissue) in I/R rats
Full size table

Effect of saffron on BNP level in I/R injured brain

The data depicted in Figs. 2 and 3 showed the effect of saffron on BNP and brain vascular endothelial growth factor (VEGF) contents, respectively. A significant increase in brain BNP was recorded in I/R group compared to sham group. While, treatment of ischemic rats with saffron extract resulted in a significant decrease of BNP in rats’ brains, as compared to I/R group.

Fig. 2
figure 2

Effect of saffron on brain naturetic peptide (BNP) content (pg/g tissue) in I/R rats. a significantly different from sham group at p ≤ 0.05. b significantly different from Left brain I/R group at p ≤ 0.05

Full size image
Fig. 3
figure 3

Effect of saffron on brain vascular endothelial growth factor (VEGF) content pg/g tissue) in I/R rats. a significantly different from sham group at p ≤ 0.05. b significantly different from Left brain I/R group at p ≤ 0.05

Full size image

Effect of saffron the brain histopathology of in I/R injured brain

Histological assessment of brain sections stained with H&E revealed no pathological features in the sham group (Fig. 4a). The brain sections of ischemic group stained with H&E revealed that 60 min ischemia and 48 h reperfusion caused marked lesions in brain tissue. As shown in Fig. 4b, multifocal areas of gliosis and clear apoptotic and degenerative changes in many neurons occurred in ischemic hemisphere. Some neuron appeared shrunken deeply basophilic with pyknotic or lysed nuclei. Perivascular edema was also observed (Fig. 4b). The 3rd and 4th saffron treated groups showed few numbers of apoptotic, degenerated and necrosed neurons (Fig. 4c, d) in brain cortex and pyramidal hippocampal cells.

Fig. 4
figure 4

Histopathological changes in the brain tissue of the different groups. a. 1st sham control group showing normal histological picture (H&E X400); b. 2nd ischemic brain group showing large focal area of gliosis (arrow), apoptosis, degeneration and necrosis of neurons (arrow) (H&E X400); c. 3rd saffron 100 mg/kg treated group showing reduced numbers of degenerated and necrosed neurons (arrow) (H&E X400); d. 4th saffron 200 mg/kg treated group showing few number of degenerated and necrosed neurons (H&E X400)

Full size image

Effect of saffron on caspase-3 and BAX protein expression in I/R injured brain

Figures 5 and 6 summarize the results of immunohistochemical evaluation of caspase-3 and Bax expression in the different experimental groups. Caspase-3 protein showed increased expression in the neuron of ischemic group (8.11 area %) compared to the control group (0.51 area %), saffron treated group (100 mg/kg) (4.04 area %) and saffron (200 mg/kg) treated group (2.02 area %). In addition, Bax protein expression in the neuron of brain ischemic group (4.99 area %) was significantly higher than the control group (0.59 area %), saffron (100 mg/kg) treated group (2.56 area %) and saffron (200 mg/kg) treated group (2.02 area %).

Fig. 5
figure 5

Representative caspase-3 immunohistochemistry in the brain cortex of different experimental groups (X400). a. group I (sham) showing very weak immunostaining reaction; b. group II (ischemic brain) showing strong immunopostive reaction in numerous cells (arrow); c. group III (saffron 100 mg/kg) showing reduction in the number of immunostained cells (arrow); d. group IV (saffron 200 mg/kg) showing weak immunostaining reaction (arrow); E. The bar chart represents immunopositive cells expressed as area %. Values with different superscripts are significantly different (p ≤ 0.05)

Full size image
Fig. 6
figure 6

Representative Bax immunohistochemistry in the brain cortex of different experimental groups (X400). a. group I (sham) showing little immunoreactivity; b. group II (ischemic brain) showing intense immunopositive staining in numerous cells (arrow); c. group III (saffron 100 mg/kg) showing reduced immunostaining reaction (arrow) in the neurons; d. group IV (saffron 200 mg/kg) showing weak immunostaining reaction (arrow); E. The bar chart represents immunopositive cells expressed as area %. Values with different superscripts are significantly different (p ≤ 0.05)

Full size image

Effect of saffron on VEGF level and protein expression in I/R injured brain

As shown in fig. 3, following reperfusion, the level of VEGF in the I/R group was markedly decreased, compared to the sham group, as detected by ELISA analysis. The levels of VEGF in the saffron treated groups were significantly increased compared to the I/R group.

Moreover, VEGF protein expression was detected in the neurons of both cerebral cortex and hippocampus of the different experimental groups using immunohistochemistry. The expression of VEGF protein was markedly decreased I/R group (5.02 area %) compared with the sham group (11.43 area %). Interestingly, VEGF showed a significant increase in the groups receiving saffron (200 and (100 mg/kg) to 12.48 and 12.96 area % compared to untreated I/R group (Fig. 7).

Fig. 7
figure 7

Representative VEGF immunohistochemistry in the brain cortex of different experimental groups (X400). a. group I (sham) showing positive immunostaining in neurons; b. group II (ischemic brain) showing weak immunopositive staining in degenerated neurons and astrocytes; c. group III (saffron 100 mg/kg) and d. group IV (saffron 200 mg/kg) showing strong positive immunostaining reaction in both neurons and astrocytes; E. The bar chart represents immunopositive cells expressed as area %. Values with different superscripts are significantly different (p ≤ 0.05)

Full size image

Discussion

In the present study, saffron treatment significantly reduced neurological deficit scores at the two examined dose levels. Moreover, saffron significantly promoted the antioxidative capacity and reduced the expression of the apoptotic markers; caspase-3 and BAX in neurons. Results of the brain histopathology were also consistent with the neurological and biochemical examination. The most important finding of the current study is that administration of saffron effectively increased VEGF in the I/R brain tissues which may have played an important role in the mechanism of saffron neuroprotection. To our knowledge, this is the first study that shows a role of VEGF in the protective effect of saffron against cerebral I/R injury.

The brain is the most vulnerable organ to ischemia compared to other organs. This prominent vulnerability to ischemic damage, in part, attributed to its high metabolic rate with a near exclusive dependence on glucose as an energy substrate, and the limited glycogen stores (Vidale et al. 2017; Lee et al. 2000).

The most upstream consequence of reduced blood flow during cerebral ischemia is the energetic failure due to the energy substrate limitation and insufficient synthesis of adenosine triphosphate (ATP) causing a drop in the total ATP level, lactate acidosis and loss of neuronal ionic homeostasis (Xing et al. 2012). Following this initial step, an ischemic cascade takes place involving neurotransmitter release and calcium overload with concomitant degradation of essential membranes, excessive sodium and water influx, and edema (Lipton 1999). This integrated response in ischemic cells stimulates the mitochondrial ROS production which directly damage proteins, lipids, carbohydrates, and nucleic acid (Lo et al. 2003).

Downstream of free radicals overproduction, other intercellular and intracellular signaling mechanisms are stimulated and finally mediate the neuronal cell death mechanisms (Zheng and Yenari 2004; Kalogeris et al. 2012; Mostafa et al. 2017).

To inspect whether saffron could cause functional improvement after I/R, the behavioral and sensorimotor performances were evaluated (Table 2). According to our results, Rats subjected to 60 min occlusion causes deficits in the sensorimotor behavior with longer latency to initiate movement compared with the sham animals. This behavioral effect is supposed to be due to the subcortical injury resulting from the ischemia (Hattori et al. 2000). Saffron treated rats significantly decreased the latency to move their bodies compared to I/R injured rats. Additionally, skilled motor movement was affected by cerebral ischemia induction in rats. Rats with I/R revealed short latency to fall from the rotarod pole, denoting deficits in ability to balance after occlusion compared to sham group. These findings are in accordance with previous studies, as damage to the striatum has been associated with alterations in generalized locomotor activity, skilled motor, and sensorimotor control (Kirik et al. 1998; Hattori et al. 2000). Rats that were administered saffron revealed improvement in locomotor balance and coordination ability compared to I/R injured rats (Table 2).

Accumulating evidences indicated that transient cerebral I/R injury significantly increased the generation of ROS, and lipid peroxidation (LPO), as well as marked depletion of the endogenous antioxidant (Collino et al. 2006).

Our results revealed that saffron significantly reversed the increased LPO, measured in terms of MDA, and the reduction GSH content induced by the transient focal cerebral ischemia (Table 3). These observations confirm the potential antioxidant role of saffron in brain I/R (Premkumar et al. 2003). The antioxidant activity of saffron could be attributed to the synergistic action of its bioactive carotenoids constituents and their cleavage apocarotenoids products, safranal, crocin and picrocrocin (Rahaiee et al. 2014). Both picrocrocin and crocin were identified as components of the studied saffron extract (Table 1, Fig. 1). Previously, safranal (Hosseinzadeh and Sadeghnia 2005) and crocin (Akbari et al. 2018; Zheng et al. 2007) were reported to effectively reduce the oxidative damage induced by I/R injury in rats.

Additionally, we observed that NO was excessively generated in I/R injured group (Table 3). This excessive NO production could be partially attributed to its overproduction by neuronal nitric oxide synthase (nNOS) in brain neurons that upregulated after reperfusion (Gursoy-Ozdemir et al. 2004; Abdel-Salam et al. 2015). Additionally, another source of NO is the activation of endothelial NOS (eNOS) during reperfusion. NO diffuses and reacts with superoxide radicals in cerebral endothelia forming highly reactive cytotoxic radicals (Gursoy-Ozdemir et al. 2000). Interestingly, treating I/R group with saffron (100 and 200 mg/kg) resulted in significant reduction of brain NO level compared to I/R untreated group (Table 3). Supporting this view, previous reports suggested that suppression of NO generation just before reperfusion preserves BBB. Furthermore, activation of the antioxidant enzyme SOD by saffron components also prevents decomposition of NO into peroxynitrite (Mark et al. 2004; Zheng et al. 2007). In line with these findings, Nam et al. (2010) revealed that crocin and crocetin have been shown to reduce inflammation and nitric oxide release from microglia following LPS administration.

Brain natriuretic peptide (BNP) is a peptide-structured compound classified as one of the natriuretic hormones. BNP acts as a potent natriuretic, diuretic, and vasodilator by inducing sodium excretion and so participates in regulation of sodium and water homeostasis, and blood pressure (BP) (De Vito 2014; Hodes and Lichtstein 2014). In addition to this systemic effects that can decrease cerebral blood flow (Sviri et al. 2006). In the brain, BNP acts as neurotransmitters or neuromodulators (Hodes and Lichtstein, 2014). Previous clinical and pre-clinical observations concluded that BNP is a part of an endogenous mechanism to the brain response to injury or damage. BNP plasma concentrations are continuously elevated during acute brain injuries (Sviri et al. 2006). Thereby, plasma BNP levels can be used for estimating the severity of the stroke and the clinical progress of stroke patients (Gonick and Buckalew 2015). High BNP correlates with a poor clinical outcome in trauma and stroke patients and can be a useful biomarker reflecting the severity of acute ischemic stroke and the infarct volume (Sviri et al. 2006; Montaner et al. 2012). In the same context, our results demonstrated a significant increase in BNP level in brain of I/R group compared sham group (Fig. 2).

Elevated BNP levels are involved in the pathogenesis of brain edema in either ischemic or hemorrhagic strokes (Modrego et al. 2008). Previous reports provide evidence that the ischemic area in the brain could be a potential source of circulating BNP and that the occlusion of MCA upregulates BNP mRNA expression in rat brain (Goetze et al. 2004; Tomita et al. 2008).

In the present investigation; saffron administration before, during and after I/R exerted a significant reduction in BNP level brain tissue (Fig. 2). In consistence with previous studies (Bayram et al. 2005; Xu et al. 2014), this BNP lowering effect of saffron may be attributed to its antioxidant activity.

At histological evaluation, neuropathological changes were observed in the infarct area of the control ischemic group with apoptotic features in the periphery of the ischemia (Fig. 4b). Assessment of ischemic region in treated groups showed that saffron, at dose 200 mg/kg and 100 mg/kg, attenuated the histopathological alterations in ischemic regions (Fig. 4c and d). In accordance with previous studies, these findings indicated that saffron or its active component, crocin (Sarshoori et al. 2014), effectively reduced brain I/R-induced injury and improved the neurological outcomes.

In stroke, activation apoptotic pathways occur mainly in the ischemic brain and result in excessive ROS generation, death receptor ligation, DNA damage, and lysosomal proteases activation (Chen et al. 2011). The mitochondrial apoptotic pathway occurs as the predominant pathway including activation cytochrome c release from the mitochondria, suppression of the antiapoptotic Bcl-2 protein expression, overexpression of the proapoptotic Bax protein, translocation of Bax into mitochondrial membrane, and activation of caspase-3 activity, which plays the central role in triggering apoptosis (Wang et al. 2014; Abdel-Salam et al. 2015; Ibrahim et al. 2015). According to the current investigation, I/R injury evoked a marked increase in the expression of both BAX and caspase-3 proteins as detected by immunostaining (Figs. 5 and 6). The antiapoptotic effect of saffron extract was evidenced by its ability to counteract the I/R-induced overexpression of BAX and caspase-3 proteins (Figs. 5 and 6). Our data agreed with those of other studies (Soeda et al. 2001; Ochiai et al. 2004; Boussabbeh et al. 2015a) in that saffron or its components were able to combat ischemic neuron death by improving the redox status and increasing GSH levels. Accordingly, the antiapoptotic action of saffron components was attributed to their ability to restore the mitochondrial membrane potential to inhibit caspase-3 activation (Boussabbeh et al. 2015b).

VEGF is a potent mitogen for angiogenesis in the ischemic boundary with a pivotal role in the post-ischemic neuroprotection by reducing infarct volume, improving enhancing neurogenesis, and functional recovery (Li et al. 2010; Herz et al. 2012; Yang et al. 2010). VEGF exerts its survival effects on endothelial and neuronal cells through activation of its specific VEGFR-2 receptor that then triggers the phosphatidylinositol 3- kinase (PI3 K)/Akt signaling pathway (Gerber et al. 1998).

In our model, we found that VEGF was suppressed in I/R injured brain as indicated by its lower level in the injured brain (Fig. 3) and its brain tissue immunopositivity compared to sham group (Fig. 7). Based on our finding, saffron administration (200 and 100 mg/kg) effectively increase VEGF protein expression and subsequently showed increased content in the ischemic brain tissues compared to I/R group (Figs. 3 and 7). These findings come in line with those of Ghorbanzadeh et al. (2017) who reported the efficacy of crocin in improving cardiac angiogenesis through enhancement of VEGF signaling pathway (Akt and ERK1/2) in heart.

Previous studies also suggested that VEGF is endowed with neuroprotective effect through inhibition of neuron apoptosis (Zhang et al. 2016). VEGF protects cells from apoptotic response by increasing Bcl-2 expression which then inhibits the translocation of Bax into mitochondrial membrane and thus prevents mitochondrial apoptotic pathway from being initiated (Nör et al. 1999; Neelam et al. 2013). Vezzani (2008) and Sun and Guo (2005) reported that PI3 K/Akt pathway activated by VEGF enhances Bcl2 and inhibits caspase 9 to mediate cell survival. Our results indicate that there is association between the inhibitory effect of saffron on the expression of apoptotic proteins and the upregulation of VEGF (Figs. 5, 6 and 7).

According to the observations of the dose-dependent effects of saffron, the current study demonstrated that the dose of 200 mg/kg produced the optimal neuroprotective effects on the rats’ brains following transient focal ischemic injury. The obtained results support the previously published data for the neuroprotective potential of saffron in I/R brain injury (Saleem et al. 2006; Sadeghnia et al. 2017; Vakili et al. 2014).

Conclusion

The data from the current study provide insight into the efficiency of saffron in reducing neural damage. The study reveals that prevention of oxidative stress, down-regulation of apoptotic proteins caspase-3 and Bax, as well as exertion of a vascular protective role by reducing NO and BNP, are among the main mechanisms mediate the neuroprotective effect of saffron. Moreover, our study is the first to show the modulation of VEGF by saffron as another mechanism arbitrating its neuroprotective and antiapoptotic potential. As a result, it can be concluded that saffron supplementation could provide a favorable strategy to protect against ischemia/reperfusion brain injury.