|Home | About | Journals | Submit | Contact Us | Français|
Mesencephalic astrocyte-derived neurotrophic factor (MANF), also known as Arginine-rich, Mutated in Early Stage of Tumors (ARMET), is a secreted protein which reduces endoplasmic reticulum (ER) stress. Previous studies have shown that MANF mRNA expression and protein levels are increased in the cerebral cortex after brain ischemia, a condition which induces ER stress. The function of MANF during brain ischemia is still not known. The purpose of this study was to examine the protective effect of MANF after ischemic brain injury. Recombinant human MANF was administrated locally to the cerebral cortex before a 60-min middle cerebral artery occlusion (MCAo) in adult rats. Triphenyltetrazolium chloride (TTC) staining indicated that pretreatment with MANF significantly reduced the volume of infarction at two days after MCAo. MANF also attenuated TUNEL labeling, a marker of cell necrosis/apoptosis, in the ischemic cortex. Animals receiving MANF pretreatment demonstrated a decrease in body asymmetry and neurological score as well as an increase in locomotor activity after MCAo. Taken together, these data suggest that MANF has neuroprotective effects against cerebral ischemia, possibly through the inhibition of cell necrosis/apoptosis in cerebral cortex.
Mesencephalic astrocyte-derived neurotrophic factor (MANF), an endoplasmic reticulum (ER) stress protein (Apostolou et al, 2008), was first purified from the conditioned medium of a ventral mesencephalic cell line 1 (VMCL1) (Petrova et al., 2003). MANF was named initially as Arginine-rich, Mutated in Early Stage of Tumors (or ARMET) as it was thought to contain an arginine-rich region at its N-terminus (Shridhar et al., 1996a; Shridhar et al., 1996b). Subsequent studies have shown that MANF protein does not contain this arginine-rich region (Mizobuchi et al., 2007; Petrova et al., 2003). MANF belongs to the same family of evolutionarily conserved proteins as the conserved dopamine neurotrophic factor (CDNF) (Lindholm et al., 2007). MANF has been found to have trophic effects on embryonic dopaminergic neurons in vitro (Petrova et al., 2003). CDNF protects and repairs mecencephalic dopamine neurons in vivo in rat 6-hydroxydopamine model of Parkinson’s disease (Lindholm et al., 2007).
MANF is highly expressed in the hippocampus and cortex as well as in the midbrain, including some dopaminergic neurons (Lindholm et al., 2008). The expression of MANF is enhanced after endoplasmic reticulum (ER) stress in cell lines (Apostolou et al., 2008; Mizobuchi et al., 2007) and after brain injury in vivo. Immunoreactivity for MANF is increased in the ischemic cerebral cortex after brain ischemia (Apostolou et al., 2008). Similarly, MANF mRNA is increased after brain ischemia and epileptic insults in the hippocampus and in the cerebral cortex (Lindholm et al., 2008). Thus, it is possible that the upregulation of MANF after injury may be a result of activation of endogenous neuroprotective processes during insults.
Previous studies have indicated that MANF has selective protective/trophic effects in cultured dopaminergic neurons (Petrova et al., 2003), which may involve an increase in GABAergic neurotransmission (Zhou et al., 2006). MANF-induced protection was also found in non-dopaminergic cells in sympathetic ganglion (Petrova et al., 2003). In cultured U2OS cells, overexpression of MANF by lentivirus-mediated transfection increases resistence to glucose-free conditions and to ER stress induced by tunicamycin (Apostolou et al., 2008). Similarly, knockdown of MANF by siRNA transfection increased ER-stress –induced cell death and caspase-3 activation in HeLa cells (Apostolou et al, 2008). These data suggest that the expression of MANF can modulate cell degeneration and death induced by ER stress and low glucose in cultured cells. Since MANF mRNA and protein are up-regulated after cerebral ischemia, an insult that triggers ER stress, it is of importance to know the role of MANF during stroke.
A critical piece of evidence, however, missing to date, would be data showing the actions of MANF protein in an established preclinical model of stroke. The purpose of this study was, therefore, to examine if there is a protective effect of exogenous MANF in cerebral ischemia. Our data suggest that pretreatment with MANF protein reduces cerebral infarction and DNA fragmentation as well as promotes behavioral recovery in stroke rats.
Adult male Sprague-Dawley rats (250-350 g, purchased from the Charles River Laboratory) were used. Animals were subjected to intracerebral administration of MANF or vehicle (PBS) and a 60-min MCAo (see below). Recombinant human MANF protein was dissolved in PBS, at pH 7.4, at various concentrations.
The MANF open reading frame, excluding the signal peptide, was cloned into a T7lac based vector containing a His-tag fusion (Peränen and Furuhjelm, 2001; Peränen et al., 1996). Expression was done in Origami B cells (Novagen, Darmstadt, Germany) in the presence of IPTG for 3 h at 37°C. The cells were lysed in L-buffer (20 mM Tris HCl pH 8.0, 0.5% TX-100, 0.4 mM PMSF) by sonication. Then NaCl and imidazole were added to final concentrations of 0.5 M and 0.02M respectively. The lysate was centrifugated (15.000×g, 15 min at 4°C) and the obtained supernatant was passed through a 0.45 μM filter. The MANF protein was purified by the HisTrap kit according to the manufacturer (GE Healthcare, Buckinghamshire, UK). The buffer of the eluted protein was exchanged (20 mM phospahate buffer; pH 8.0, 150 mM NaCl) by a PD-10 column. The his-tag was cleaved by AcTEV (Invitrogen) in the absence of DTT. The cleaved products were passed through the HiTrap Chelating column to get rid of the his-tag and AcTEV, while the cleaved MANF was obtained in the flow through. The final purification of MANF was done by the HiTrap Q column (GE Healthcare, Buckinghamshire, UK). Buffer exchange and concentration of MANF were carried out using the Amicon Ultra-4 filter device (Millipore), and aliquots of MANF were stored at -80°C.
Animals were anesthetized with chloral hydrate (0.4 g/kg, i.p.). MANF protein (0, 1, 6, 12, 24 μg) or vehicle was given intracerebrally into three cortical sites as previously described at about 20 min before MCAo (Chiang et al., 1999). Two microliters of MANF solution (0.17-4 μg/μl) or PBS was injected at a rate of 1 μl/min at each site. The needle was retained in place for 5 min after injection. Ligation of the right middle cerebral artery (MCA) and bilateral common carotids (CCAs) was performed with methods previously described (Chen et al., 1986). Briefly, the bilateral CCAs were identified and isolated through a ventral midline cervical incision. Rats were placed in stereotaxic apparatus and a craniotomy was made in the right hemisphere. After the last injection the right (MCA) was ligated with a 10-0 suture and bilateral common carotids (CCA) were ligated with nontraumatic arterial clamps for 60 minutes. After sixty minutes of ischemia the suture around the MCA and arterial clips on CCAs were removed to introduce a reperfusional injury. After recovery from anesthesia, the rats were returned to their home cage. Body temperatures during and after surgery were maintained at 37 °C.
Body asymmetry was analyzed using an elevated body swing test (Borlongan and Hida, 1998). Rats were examined for lateral movements/turning when their bodies were suspended 20 cm above the testing table by lifting their tails. The frequency of initial turning of the head or upper body contralateral to the ischemic side was counted in 20 consecutive trials. The maximum impairment in body asymmetry in stroke animals is 20 contralateral turns/20 trials. In normal rats, the average body asymmetry is 10 contralateral turns/20 trials (i.e. the animals turn in each direction with equal frequency).
Neurological deficits were evaluated using Bederson’s score (Bederson et al. 1986). In a postural reflex test, rats were examined for the degree of abnormal posture when suspended by 20-30 cm above the testing table. They were scored according to the following criteria.
Locomotor activity was measured using an infrared activity monitor (Columbus, OH). Animals were individually placed in a 42×42×31 cm plexiglass open box which contained horizontal infrared sensors spaced 2.5 cm apart. Motor activity (total distance traveled) was calculated every 10 min for 60 min.
Two days after reperfusion rats were decapitated. The brains were removed and sliced into 2.0-mm-thick sections. The brain slices were incubated in a 2% TTC solution (Sigma, St. Louis) for 15 min at room temperature and then transferred into a 4% paraformaldehyde solution for fixation. The area of infarction in each slice was measured with a digital scanner and Imagetools programs (University of Texas Health Sciences Center). The volume of infarction in each animal was obtained from the product of average slice thickness (2 mm) and sum of infarction areas in all brain slices examined.
Rats were decapitated 24 hours after reperfusion. The brains were taken out, frozen with isopentane and cut into sections (25 μm) in a cryostat. The sections were mounted on microscope slides, air-dried, then fixed with 4% paraformaldehyde for 30 minutes at 4°C. A standard terminal deoxynucleotidyl transferase mediated dUTP nick-end labeling (TUNEL) procedure for frozen tissue sections, with minor modifications, was performed (In Situ Cell Death Detection Kit, POD, Roche Applied Science. Briefly, slide-mounted sections were rinsed in 0.5% Triton X-100 in 0.1 M phosphate buffer for 20 minutes at 80°C. Sections were rinsed three times with 0.1 M phosphate buffer. To label damaged nuclei, 100 μl of TUNEL reaction mixture was added onto each sample followed by a 60-minute incubation at 37°C in a humidified chamber. Procedures for controls were carried out as described in the manufacturer’s manual. Sections were rinsed three times with 0.1 M phosphate buffer, rinsed with H2O, air dried and coverslipped. Histological images were acquired using an Infinity 3 camera, NIKON 80i microscope and QCapture Pro 5.0 software. Pixel density of TUNEL-positive nuclei was quantified from six sections of each brain area and averaged using NIS Elements 2.3 software. For the illustration (Fig. 2) images were adjusted equivalently for brightness and contrast for optimal visualization of signal with Adobe Photoshop 7.0.
A real time cortical blood flow was measured continuously with a laser Doppler flowmeter (PF-5010, Periflux system) in anesthetized rats. A blood flow probe (1 mm diameter) was strereotaxically (in proximity to 2.1 mm posterior and 4.4 mm lateral to the bregma) placed in the cortex within 1 mm to dura. The probe was sometimes moved within 1 mm to this location to avoid large vessels on the cortical surface. Baseline was recorded for 10 min before three local PBS or MANF (6μg) injections were given. Cerebral blood flow was also measured from 15 min before MCAo to 15 min after the onset of reperfusion. Brain and body temperatures were monitored and maintained throughout the experiment at 37°C with a heating pad.
In anesthetized rats the femoral artery was cannulated with polyethylene catheters (model PE-50; Dural Plastics an Engineering). Mean arterial pressure was recorded through strain gauge recorder (model P23 ID, Statham) and recorded on a strip-chart recorder (model RS 3600, Gould). Arterial blood (0.2 ml) was withdrawn from the femoral artery 15 min before and after three cortical MANF injections, total of 6μg, in the rats. Blood was heparinized, blood pH, and concentrations of CO2, O2, Na+, K+, Ca2+ and hemotocrit were analyzed with a blood gas analyzer (GEM Premier 3000, Instrumentation laboratory). Brain and body temperatures were monitored and maintained throughout the experiment at 37 °C with a heating pad.
Two tailed Student’s t-test and one or two way ANOVA were used for statistical comparison. Student Newman-Keuls and Dunnet test were used for post-hoc analysis. Data are presented as mean ± s.e.m. p-values of <0.05 were considered significant.
A total of 44 rats were used to examine cerebral infarction at 2 days after MCAo. Of these, 15 rats were pre-treated with PBS and the remaining rats were pre-treated with MANF at 1 μg (n=8), 6 μg (n=8), 12 μg (n=8), or 24 μg (n=5). An U-shaped relationship between the dose of MANF and volume of infarction was found. The maximal protection of MANF, or minimal infarction volume after MANF injection, was found at doses of 6 μg/rat (Fig 1A. p=0.029, one-way ANOVA+ post hoc Dunnett’s test). Similarly, the area of largest infarction in any brain slice was significantly reduced by MANF at dose of 6 μg (Fig. 1C. p=0.0053, one-way ANOVA). Doses higher or lower than 6 μg/rat showed less protection, (Fig. 1A & 1C).
To further examine the topographic relationship of protection, the area of infarction in each slice was compared in animals treated with 6 μg MANF or vehicle. A significantly lower infarction was found in animals receiving 6 μg MANF treatments (Fig. 1B, p=0.0008, two-way ANOVA). Post hoc analysis revealed that significant differences between PBS and 6 μg MANF treatments was mainly localized to rostral slices (Fig. 1B, p<0.01).
Eleven rats were sacrificed for TUNEL labeling analysis at one day after MCAo. Enhanced TUNEL activity was found in the ischemic cortex in all stroke rats. TUNEL activity was analyzed, using optical pixel density, in two brain regions based on the distance from the site of injection (1: distal to the injection site, at level +2.2 mm from bregma. 2. proximal the injection site, at level −0.20 mm from bregma). TUNEL optical density in all animals was normalized to the mean of TUNEL density in stroke rats pretreated with vehicle. We found that MANF did not alter TUNEL activity at the level distal to the MANF injection site (PBS = 100 ± 14%, N=6, MANF = 103 ± 11%, n=5, p=0.8652, t test, Fig. 3A). In contrast, at a level proximal to the site of MANF injection, MANF significantly reduced TUNEL pixel density (Fig. 2B, p=0.0163, Student’s t-test, PBS = 100 ± 8%, MANF = 55 ± 13%). These data suggest that MANF may have anti-apoptotic activity close to the site of injection. Representative photomicrographs for TUNEL activity are shown for PBS (Fig. 2C) and MANF pretreatment (Fig. 2D).
A total of 23 rats were used to examine behavioral recovery after cerebral ischemia on days 2, 7 and 14. Of these animals, 13 were treated with MANF 6 μg, the optimal dose for reduced infarct volume, and 10 were treated with vehicle. Three behavioral/neurological tests were used.
There was a spontaneous increase in total distance traveled from days 2 to 14 after MCAo (Fig. 3A, p<0.001, two-way ANOVA). The distance traveled in one hour was significantly increased on day 7 by MANF (Fig. 3A, p<0.05, 2-way ANOVA + post hoc Student-Newman-Keuls test). The data on locomotor movement on day 7 was further analyzed every 10 min. MANF significantly increased total distance traveled (Fig. 3B, p=0.011, F(1,126)=6.706, 2-Way ANOVA).
There was significant reduction in body asymmetry between day 2 and day 14 after MCAo (Fig. 3C, p<0.001, two-way ANOVA). Administration of MANF significantly incremented the reduction in body asymmetry (p=0.002, two-way ANOVA). Post-hoc test revealed that MANF significantly incremented the reduction in body asymmetry on day 7 after MCAo (p<0.001).
Bederson’s score was significantly reduced from day 2 to day 14 after MCAo (Fig. 3D, p<0.001, two-way ANOVA). MANF significantly augmented the attenuation in Bederson’s score (p=0.003). Posthoc test revealed that differences between PBS and MANF treatments occurred on day 7 (p<0.011).
A total of 11 rats were used to examine the effects of MANF (6μg) or PBS on cerebral blood flow (Fig. 4). Baseline CBF was measured for 10 min before administration of MANF or PBS. The change of CBF after MANF or PBS was calculated using the following formula. Δ CBF = (CBF at post-injection − basal CBF) / basal CBF × 100%. We found that local administration of MANF did not alter the basal CBF in cerebral cortex. (Fig. 4A, PBS: 15.2±8.7%, MANF: 13.8±14.1%, p=0.9393, t-test). The CBF was also compared before and during MCAo [Δ CBF = (CBF during MCAo − CBF before MCAo) / CBF before MCAo × 100%]. We found that there is a significant reduction of CBF during MCAo in both animals treated with MANF or PBS. No difference in the decrease of CBF was found between MANF and PBS treatment (Fig.4B, PBS: -81.2±4.6%, MANF: -80.7±3.5%, p=0.9268, t-test). A typical tracing of cerebral blood flow during MCAo is shown in Fig 4C.
Physiological parameters were examined in 6 rats. Blood samples and blood pressure were monitored 15 min before and 45 min after MANF administration. We found that administration of MANF did not alter systemic blood pressure, blood gases, and electrolytes (Table 1, p>0.05, t test). There was also no change in body or brain temperature.
In this study, we found that pre-stroke delivery of MANF protein, at a dose of 6 μg, reduced cerebral infarction, suppressed DNA fragmentation, and facilitated motor recovery in stroke rats. Local administration of MANF did not alter blood pressure, body temperature, serum electrolytes or blood gases. Taken together, these are the first data suggesting that recombinant MANF protein has neuroprotective effects against cerebral ischemia. This neuroprotective effect of MANF is not secondary to the changes in physiological parameters.
As seen in Fig 1, the largest infarction area was found in slice #3. Administration of MANF significantly reduced the size of infarction in this slice. In this study, CBF was measured through a fine probe placed in the ischemic cortex in slices #3 or #4. The distance between the CBF recording site and the injection sites was 1-2 mm. We found that administration of MANF did not alter cerebral blood flow before or during MCAo in this region. These data suggest that MANF -mediated protective responses are probably not due to changes in cerebral blood flow near the injection site during ischemia. It is still not known if local circulation may be altered by MANF at sites distal to the injection or at a later time point after reperfusion.
There is a “U” shaped relationship between the dose of MANF and size of infarction. The maximal protection of MANF was found at doses of 6 μg/rat. Doses higher or lower than 6 μg/rat showed less protection. The mechanisms underlying of such a dose-response relationship are not known. However, similar responses are seen with other trophic factors. For example, bone morphogenetic protein 7 (BMP7) at lower doses reduced H2O2 toxicity in cortical neuronal culture and 6-OHDA-mediated toxicity in dopaminergic neurons, while at higher doses increased toxicity (Cox et al., 2004; Harvey et al., 2004). Similar dose-response patter has been previously reported also with persephin, a trophic factor in the GDNF-family, in cerebral ischemia (Tomac et al., 2002)
Pretreatment with MANF into cerebral cortex reduced TUNEL labeling, suggesting that the protective effect of MANF is mediated, at least in part, through the inhibition of DNA fragmentation. There is a topographic relationship for this protection after local MANF injection. Less TUNEL labeling was found at sites proximal to the injection site. Similarly, the reduction in cerebral infarction was found closest the injection sites, i.e. the rostral brain slices. These data may suggest that locally applied MANF, unlike other trophic factors such as GDNF or BMP7, is less diffusible in the cortex. Alternatively, the effective dose of MANF may act primarily at the site of the lesion.
We found that MANF reduces TUNEL labeling and cerebral infarction at 24 hours and 2 days after MCAo, respectively. The time points for these histological analyses was chosen based on the peak expression of various neurodegenerative markers. For example, TTC staining was used for the detection and quantification of cerebral infarction at 2 days after MCAo as indicated in many studies (Wang et al., 2003). Measurement of the size of infarction at earlier time points (i.e. 1 day) can be affected by brain edema (Slivka et al., 1995), while at later time points can be confounded by liquefication of the infarcted tissue (Chou et al., 2006). TUNEL was analyzed at 24 hours after MCAo (Zhang et al., 2002). There is less TUNEL labeling at 8 hours after MCAo (Zhang et al., 2002). Our recent time course study after focal ischemia also indicates that TUNEL activity returned to basal levels (compared to sham surgery) on day 6 in the ischemic cortex (Luo et al., 2008, in press). These data suggest that MANF reduces the progression of neurodegeneration after cerebral ischemia.
The spontaneous recovery of motor functions after stroke in rodents receiving vehicle or no treatment has been reported in many studies (Chang et al., 2003). These include neurological scores (Chou et al., 2006), step through latency (Yonemori et al., 1999) paw placing (Ren et al., 2000), and locomotor activity (Chang et al., 2003), etc. In this study, we also found body asymmetry and neurological scores were reduced, while total distance traveled was increased, over 2 weeks, in stroke animals receiving vehicle. Pretreatment with MANF further reduced these behavioral deficits, mainly within 2 weeks after MCAo. These data suggest that pretreatment with MANF enhances the early recovery of motor function in stroke animals.
MANF mRNA and protein are widely expressed in the brain, but also, highly expressed in lungs, stomach and testis (Lindholm et al., 2008). The expression of MANF, both in neuronal tissue and non-neuronal cell lines, is altered after injury. Expression of MANF mRNA in dentate gyrus, piriform cortex, parietal cortex, retrosplenial cortex and thalamic reticular nucleus is increased after epileptic insults and in dentate gyrus and in the CA1-region after ischemia (Lindholm et al., 2008). In addition, MANF immunoreactivity is increased after MCAo (Apostolou et al., 2008). In non-neuronal cells, MANF expression is increased in cell lines derived from bone tissue (U2OS), kidneys (HEK293) and neuroblastoma (SHSY-5Y) after ER stress induced by tunicamycin, thapsigargin and lactatystin (Apostolou et al., 2008). The upregulation of MANF after injury may be a result of activation of endogenous neuroprotective processes during insults, which is further supported by the findings that upregulation of MANF expression by lentivirus-mediated transfection increases resistence to glucose-free conditions and ER stress induced by tunicamycinn in cultured U2OS cells (Apostolou et al., 2008). Although MANF and CDNF are structurally similar with almost a 60% amino acid sequence identity (Lindholm et al., 2007), only MANF, but not CDNF, is upregulated by tunicamycin in U2OS-cells (Apostolou et al., 2008). These data may suggest that MANF, but not CDNF, is involved in ER stress responses.
There are some similarities and differences in neuroprotective responses, against brain ischemia, induced by MANF and other trophic factors, such as GDNF or BMP7. For examples, the expressions of MANF, BMP7 and GDNF mRNA were increased after cerebral ischemia (Arvidsson et al., 2001; Chang et al., 2003; Lindholm et al., 2008; Miyazaki et al., 2001). The increase of their expression could represent an endogenous neuroprotective response to reduce ischemic insults. MANF may induce protection through an anti-apoptotic pathway, a similar mechanism was reported for GDNF and BMP7 (Wang et al., 2001; Yu et al., 2008). On the other hand, specific receptors were found for GDNF or BMP7 –mediated neuroprotective and trophic effects (for reviews, see (Airaksinen and Saarma, 2002; Harvey et al., 2005). The receptor for MANF has not yet been discovered. MANF can reduce ER stress in cell culture after injury. Such a response was not reported for GDNF or BMP7.
GDNF (Harvey et al., 2003) and BMP7 can reduce behavioral deficits in stroke animals when given systemically or intraventricularly (Chang et al., 2003; Lin et al., 1999). We found that MANF, given intraventricularly, has less protective effects (unpublished observations). Since the neuroprotection induced by MANF is close to the injection site, we propose that MANF may act relatively locally. The potential advantage of this property is that MANF may be useful for discrete brain damage, such as focal ischemia.
In conclusion, our data suggest that MANF has neuroprotective effects against CNS injury in a transient focal ischemia model. The protective response of MANF may involve the inhibition of apoptosis/necrosis. The effectiveness of MANF pretreatment in stroke may be clinically useful for patients susceptible to ischemic events, for example, for those suffering from transient ischemic attacks. Furthermore, since MANF expression is enhanced after brain ischemia, and over-expression of MANF reduces ER stress, (Apostolou et al., 2008) it is possible that MANF may be involved in endogenous protection against ischemia in patients. Beside ischemic injury (DeGracia and Montie, 2004), several diseases including type II diabetes, Parkinson’s disease, and Alzheimer’s disease (Lindholm et al., 2006; Ron and Walter, 2007) are thought to involve ER stress. MANF may also play important roles in these diseases.
This research was supported by the IRP of NIDA, NIH, DHHS, Sigrid Jusélius Foundation, the Academy of Finland