Introduction
Stroke is one of the leading causes of adult death and disability [
16]. Despite this, intravenous thrombolysis remains the only well-established pharmacological treatment available. Inflammation is a major pathophysiological event in the post-stroke brain, initiated by activation of brain resident immune cells, microglia, and infiltration of peripheral leukocytes [
30]. Post-stroke inflammation can further exacerbate injury but also serves tissue repair. Thus, more profound knowledge of the mechanisms of ischemic injury and post-ischemic inflammation is still urgently needed to kindle novel therapeutic strategies.
Mesencephalic astrocyte-derived neurotrophic factor (MANF) is an 18 kDa protein widely expressed in different tissues, including the brain, and has cytoprotective properties [
31,
39,
41,
45,
54]. In an uninjured brain, MANF protein is expressed mainly in neurons [
14,
40] and its expression levels are increased upon acute ischemia [
7,
48,
80]. Interestingly, at 24 h after ischemia, MANF protein expression has been reported also in microglia/macrophages of the ischemic region [
60,
77], but the protein expression of MANF in later post-ischemic timepoints has not been characterized in animal models of stroke. No previous studies have reported MANF expression in the human stroke brain. We and others have observed that in naïve, non-injured, rodent brains endogenous MANF protein expression is primarily localized in neurons. However, mRNA levels are relatively high in all brain cell types.
Endogenous MANF is localized in the endoplasmic reticulum (ER) lumen but can also be secreted from cells, especially after ER Ca
2+ depletion [
7,
21,
26,
65], and
Manf gene is induced upon the activation of unfolded protein response of the ER [
33,
65]. Intracellular MANF significantly maintains ER protein folding homeostasis and reduces ER stress-induced apoptosis [
7,
25,
38]. MANF has been shown to directly interact with several ER luminal proteins, including the chaperone protein GRP78 (a.k.a. BiP), [
15,
21,
76] and the ER stress sensors IRE1α, ATF6, and PERK [
32]. ER stress sensors are also known to modulate inflammation, and MANF has been shown to affect inflammation processes. Endogenous neuronal MANF protects against cerebral ischemia since embryonic
Manf deletion from neuronal lineage cells leads to larger infarcts in
NestinCre/+::Manfflox/flox(fl/fl) mice compared to wild type mice [
48].
Exogenous MANF protects brain tissue from ischemic damage in rat models of ischemic stroke when delivered intracranially as a protein or via viral vector [
2,
3,
23,
50,
72,
79] and, more importantly, alleviates functional deficits when injected intracranially several days after cerebral ischemia [
6,
48]. The mode of action of exogenous MANF remains unclear, but immunomodulatory effects have been suggested as a putative mechanism for MANF’s cytoprotective properties. Administration of recombinant MANF into mouse brain has been shown to downregulate inflammation by decreasing NF-κB-mediated pro-inflammatory cytokine production in vivo in an ischemic stroke model in aged mice [
23] and in vitro after oxygen–glucose deprivation [
23,
83].
We have previously shown that viral-vector mediated overexpression of MANF in the peri-infarct region increases the number of phagocytic microglia/macrophages after ischemic stroke [
48] and downregulates proteins S100A8 and S100A9 related to innate immunity [
66]. Furthermore, MANF has been shown to increase the pro-regenerative and anti-inflammatory activation, known as alternative activation, of innate immune cells in an ischemic stroke model [
77] and the damaged retina of mouse and fruit fly [
51]. However, MANF does not penetrate the blood–brain barrier. So far, MANF therapy has been administered by intracranial delivery, which is highly invasive and thus not a realistic approach for therapeutic use. Intranasal delivery of several other proteins is neuroprotective in rat and mouse transient middle cerebral artery occlusion models [
12,
22,
43,
78]. In principle, intranasally delivered molecules can bypass the blood–brain barrier (BBB) and access the central nervous system via several pathways, including the olfactory and trigeminal nerves, vascular and cerebrospinal fluid pathways, and the lymphatic and glymphatic systems [
24,
44,
67].
Since the knowledge of post-stroke MANF protein expression after cerebral ischemia is limited to acute time-points (24–48 h) and data from human patients is lacking, the initial aim of our study was to characterize the evolution of endogenous MANF protein expression in the ischemic human brain. The emphasis on human stroke brain samples is crucial as it not only provides a clinically relevant perspective but allows identifying potential species-specific differences in MANF protein expression, which need to be taken into account for translational research and therapeutic development. We then studied whether we observe a similar expression pattern in rodent models of ischemic stroke, and identified which cells express MANF protein. Secondly, we conducted a proof-of-concept study using non-invasive intranasal delivery of recombinant human MANF (rhMANF) for neuroprotection in a rat model of distal middle cerebral artery occlusion (dMCAo) model. The findings from our intranasal rhMANF study led us to the third aim: demonstrate the effects of intravenously administered rhMANF in the dMCAo model.
We show for the first time how MANF protein expression evolves temporally in the post-stroke human brain and similarly in the rat brain. We demonstrate that after stroke, activated microglia/macrophages prominently express MANF in both species and that stroke causes a clear transition of MANF protein expression pattern towards microglia/macrophages. Moreover, we show that systemic delivery of rhMANF reduces ischemic cerebral injury. The protective effect can be mediated via the downregulation of pro-inflammatory cytokine and upregulation of anti-inflammatory cytokine production in the infarcted cortex. These data strengthen the translational relevance of MANF as an endogenous cytoprotective mediator, both in neurons and in phagocytic microglia/macrophage cells post-stroke. Systemic administration of MANF presents a novel therapeutic approach to utilize these cytoprotective and anti-inflammatory properties in ischemic stroke.
Materials and methods
Patients
We studied MANF immunoreactivity in 7 acute ischemic stroke patients treated at our hospital and who had died early after stroke onset. The rapid autopsy procedures and sample collection have been described in detail previously [
42,
56,
57]. The study had the approval of the local research ethics committee, and next of kin gave informed consent for the study. Patient details are given in Table
2.
Immunohistochemistry on human brain sections
On autopsy, 1 cm
3 cortical samples including white matter were dissected, formaldehyde fixed and embedded in paraffin. Tissue sampling sites including infarcted brain tissue and contralateral healthy tissue were decided based on the individual topography of each brain infarction, using macroscopic examination of the parenchyma and cerebrovasculature, and comparing it to the most recent computed tomography scans [
42,
56,
57]. Briefly, on autopsy, the infarcted brain areas were identified during the macroscopic examination of the brain parenchyma and cerebrovasculature in comparison with the most recent computed tomography scans. Since the localization and size of the infarcts were unique in each case, we preferred to target the tissue sampling on the basis of the individual infarct topography rather than standard localizations. Samples from the corresponding areas of the contralateral or non-infarcted hemispheres were processed in a similar way. Examination of the hematoxylin–eosin–stained sections to grade the severity of ischemic neuronal changes was performed by a neuropathologist without information of the sample localization. Focusing on the integrity of the nucleus, we ascribed scores for signs of ischemic neuronal changes to each tissue section as follows: 1, largely normal morphology but scattered neurons had nuclear abnormalities such as pyknosis, low nuclear cytoplasmic contrast, or smearing of nuclear border (similar to type III neurons); 2, a large proportion of neurons had nuclear abnormalities; 3, a large proportion of neurons had nuclear abnormalities while scattered ones exhibited signs of irreversible damage such as shrunken cytoplasm with irregular borders and invisible nuclei (similar to type IV neurons); and 4, a large proportion of neurons showed irreversible changes. These characteristic morphological neuronal changes are well appreciated in light microscopy after > 24 h following the ischemia onset in large, fatal infarctions. The ischemic neuronal changes were scored in each sample with a focus on the integrity of the nucleus (Table
2).
The formalin-fixed, paraffin-embedded samples were cut into 4 µm sections and mounted on microscope slides. The sections were deparaffinized and heated in 10 mM citrate buffer, pH 6.0 for antigen retrieval. For anti-MANF immunostaining, endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide in methanol (Sigma Aldrich), and the non-specific antibody binding was blocked with 10% normal goat serum (cat#PK-6101, Vector Laboratories, Burlingame, CA, USA) in 0.1% Tween-20 (Sigma Aldrich) in TBS (TBS-T). To further reduce non-specific staining, the sections were blocked with avidin and biotin (Vector Laboratories, cat# SP-2001) followed by incubation with rabbit anti-MANF (1:800, cat#HPA011175, Atlas antibodies, Bromma, Sweden) in 1.5% normal goat serum in 0.1% TBS-T at 4 °C overnight. The next day, sections were incubated with secondary antibody (biotinylated goat anti-rabbit 1:200, cat#PK-6101, Vector Laboratories) followed by incubation with avidin–biotin complex (ABC kit, cat#PK-6101, Vector Laboratories). The color was developed using a peroxidase reaction with 3´,3´-diaminobenzidine (DAB; cat#SK-4100, Vector Laboratories), the sections were counterstained with hematoxylin, dehydrated, and coverslipped. Rabbit IgG (cat#NI01, Merck Millipore, Temecula, CA, USA) was used as a negative control instead of the primary antibody using the same protein concentration as with the primary antibody. For anti-CD68 staining, the Novolink™ Polymer Detection System kit (cat#RE7140-K, Leica Biosystems, Newcastle Upon Tyne, UK) was used according to manufacturer’s instructions. Briefly, after antigen retrieval, the sections were incubated for 5 min with peroxidase block, washed, incubated for 5 min with protein block, washed, and incubated for 1 h at room temperature with mouse anti-human CD68 (1:200, clone PG-M1, cat#M0876, Dako, Glostrup, Denmark). After washing, the sections were incubated for 30 min with post primary, washed, incubated for 30 min with Novolink™ polymer, washed, and incubated with DAB chromogen. Mouse IgG3 isotype control (cat#MAB007, R&D Systems, Minneapolis, MN, USA) was used as a negative control instead of the primary antibody using the same protein concentration as with the primary antibody.
Animals
All the animal experiments comply with current Finnish and Taiwanese laws.
Finland A total of 140 male Sprague Dawley rats (age 7–8 weeks, weight 200–270 g, Envigo, Netherlands) and 32 C57BL/6NHsd male mice (age 8–9 weeks, weight 22–28 g, Envigo, Netherlands) were used for the experiments. Gene-modified
NestinCre/+::
Manffl/fl male mice (n = 4) were used to investigate post-stroke MANF expression after MANF deletion from neuronal lineage cells.
Manffl/fl male littermates (n = 5) were used as controls. The generation of
Manffl/fl and
NestinCre/+::
Manffl/fl mice has been described in detail before [
38].
Manf exon 3 was conditionally removed by crossing the
Manffl/fl mice with Nestin-Cre transgenic mice (B6.Cg-Tg[Nes-cre]1Kln/J, a gift from Edgar Kramer). Congenic
Manffl/fl or
Manffl/+ mice in C57BL/6JRcc background were crossed with either
NestinCre/+::Manffl/fl or
NestinCre/+::Manffl/+ mice to generate
NestinCre/+::Manffl/fl and
Manffl/fl mice. All animal experiments were conducted according to the 3R principles of EU directive 2010/63/EU on the care and use of experimental animals, and according to local laws and regulations, and were approved by the national Animal Experiment Board of Finland (protocol approval number ESAVI/5459/04.10.03/2011 and ESAVI/7812/04.10.07/2015). The animals were housed in groups of 4–5 with ad libitum access to food and water under a 12 h/12 h dark–light cycle and the well-being of the animals was monitored daily.
Taiwan A total of 42 male Sprague–Dawley rats (age 12 weeks; weight 250–300 g, BioLASCO Taiwan Co., Ltd) were used for the experiments. Animals were housed in groups of 3–4 rats under a 12-h light/dark cycle at a controlled temperature (22 °C) and humidity (50%) with free access to food and water. The animal experiments were carried out in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. Experimental protocols were approved by the Institutional Animal Care and Use Committee of the National Defense Medical Center (approval number IACUC 16258), and Approval of Animal Use Protocol Board Buddhist Tzu Chi General Hospital (approval number IACUC 107-04), Taiwan, R.O.C.
All the experiments were performed in a blinded manner and are reported according to the ARRIVE guidelines. Blinding procedures were such that the person performing the stroke surgery, behavioral testing, and data analysis did not know treatment allocation.
Distal middle cerebral artery occlusion
Finland For rats, cortical cerebral ischemia was induced by occluding the distal middle cerebral artery (dMCA) together with a bilateral common carotid artery (CCA) occlusion [
13]. The rats were anesthetized with 4% chloral hydrate (Sigma Aldrich) intraperitoneally (i.p.; 400 mg/kg) and lidocaine (Orion Pharma, Espoo, Finland) was used as a local anesthetic. The surgery was performed as described previously [
2,
3,
5,
48]. Briefly, the CCAs were isolated through a cervical incision. A small craniotomy was made on the right side of the skull and the right dMCA was ligated directly with a 10-0 suture. CCAs were simultaneously occluded with non-traumatic arterial clips. After 60 or 90 min, the dMCA and CCAs were reopened to allow reperfusion. For mice, a permanent dMCAo was performed similarly by ligating the dMCA permanently with the 10–0 suture, but without occluding the CCAs. The body temperature of the animals was maintained at 37°C throughout the procedures until recovery from anesthesia when the animals were returned to their home cages. The animals received one dose of carprofen (Rimadyl Vet 50 mg/ml, Zoetis Animal Health, Copenhagen, Denmark) subcutaenously (5 mg/kg) for post-operative pain. A total of 12 rats (6 in the vehicle group, 6 in the rhMANF group) and 3 mice did not survive the surgery. There was no further mortality.
Taiwan The procedures were the same as in Finland, except before the reopening of the dMCA, the rats were anesthetized with 5% isoflurane in 30% O2 / 70% N2O using the V-10 Anesthesia system (VetEquip, Inc., Pleasanton, CA). Following induction of anesthesia, the level of isoflurane was maintained at 1.5% during the procedure. After dMCAo surgery, ketoprofen was administered once at a dose of 2.5 mg/kg per rat for pain relief.
Intranasal administration
The rats were assigned to different treatment groups in a random manner. RhMANF (P-101-100, Icosagen, Estonia) in phosphate-buffered saline (PBS) or only PBS was administered to the nasal cavity of the rats as described previously [
5,
46]. Briefly, the rats were anesthetized with isoflurane (4.5%) or, when administered during dMCAo surgery, 4% chloral hydrate, and 10 µl of rhMANF or PBS was pipetted into each nostril. To test the neuroprotective effect of intranasal rhMANF delivery, rhMANF or PBS was administered at three different time points: 12 h before dMCAo, immediately before dMCAo, and immediately after reperfusion, equaling a total of 20 µg or 60 µg of rhMANF. There was no significant difference between the doses in infarct volume (
p = 0.49, Student’s
t-test). Therefore, the 20 µg and 60 µg groups were combined for statistical analysis when compared to the vehicle group.
Intravenous administration
The rats were assigned to different treatment groups in a random manner. RhMANF in 0.9% NaCl or 0.9% NaCl (vehicle control) were given as an intravenous bolus (500 µl per dose). Either one dose of 1.5 µg rhMANF was injected into the tail vein approximately 15 min after dMCAo reperfusion or three doses of 1.5 µg rhMANF were injected into the femoral vein with 10 min intervals starting 10 min after the dMCAo reperfusion.
Behavioral tests
The body asymmetry test, modified Bederson’s neurological test and analysis of locomotor activity were assessed in rats as described previously [
2,
3,
5,
48]. Briefly, body asymmetry was analyzed from 20 consecutive trials by lifting the rats above the testing table by the tails and counting the frequency of initial turnings of the head or upper body contralateral to the ischemic side (the maximum impairment in stroke animals is 20 contralateral turns whereas naïve animals turn in each direction with equal frequency resulting in 10 contralateral turns) [
11]. In the modified Bederson’s score the neurological deficits were scored according to the following criteria: 0 = no observable deficit; 1 point = rats show decreased resistance to lateral push; 2 points = rats keep the contralateral forelimb to the breast and extend the other forelimb straight when lifted by the tail in addition to behavior in score 1; 3 points = rats twist the upper half of their body towards the contralateral side when lifted by the tail in addition to behavior in other scores [
9]. Locomotor activity was measured using an infrared activity monitor for one hour (Med Associates, St. Albans, VT, USA).
Laser Doppler flowmetry
The effect of intranasal rhMANF (10 µg) on cerebral blood flow (CBF) was measured with Laser Doppler flowmetry (LDF; moorVMS-LDF, Moor Instruments, Axminster, UK) during and after dMCAo (n = 19). The skull was thinned by drilling a small hole on the cortex [A/P − 2.0; L/M 4.0 relative to bregma [
53]] and the MoorVMS-LDF1 optical fibre probe was attached on the skull with a PH-DO single fibre holder (Moor Instruments) and dental cement. The baseline CBF was monitored for 10 min before the occlusion of CCAs and dMCA, and monitoring was continued for 10 min after reperfusion. Body temperature was maintained at 37°C with an automated heating pad. One animal was excluded due to insufficient (< 65%) reduction of CBF after stroke induction.
Blood pressure measurement
The rats (n = 13) were anesthetized with 4% chloral hydrate (400 mg/kg i.p.) and lidocaine (Orion Pharma) was used as a local anesthetic. The trachea was intubated and the measurement probe was inserted into the left common carotid artery. The femoral vein was cannulated for rhMANF or saline administration. The mean arterial blood pressure and heart rate were recorded using PowerLab 8/30 Channel Recorder (ADInstruments) for 40 min. Baseline blood pressure was measured for 10 min and a 500 µl bolus of increasing rhMANF dose (1.5 µg, 15 µg and 150 µg) or saline was injected every 10 min into the femoral vein. The body temperature was maintained at 37°C during all procedures. The animals were euthanized immediately after the measurement.
Analysis of blood gases and electrolytes
Under chloral hydrate (4%, 400 mg/kg i.p.) anesthesia, the femoral artery was cannulated with a PE-50 polyethylene tube for monitoring of blood gasses and electrolytes (n = 15). Body temperature was automatically maintained at 37.5 ± 0.5°C by a rectal temperature sensor and a heating pad (CMA-150, Sweden). A 100 µl arterial blood sample was withdrawn from the femoral artery and immediately injected into the epoc® Blood Analysis System (Epocal) for analysis of the blood gasses and other physiological parameters including pH, partial pressure of oxygen (pO2), partial pressure of carbon dioxide (pCO2), sodium, potassium, glucose, lactate, and hemoglobin 10 min before the dMCAo (baseline), at dMCAo reperfusion, and at 30 min and 2 h after the reperfusion.
Immunohistochemistry on rat and mouse brain sections
The animals were deeply anesthetized with pentobarbital (90 mg/kg i.p., Mebunat, Orion Pharma) and transcardially perfused with saline followed by 4% paraformaldehyde. Brains were post-fixed in 4% paraformaldehyde for 2 days, dehydrated in a series of ethanol and xylene, and embedded in paraffin. Brains were cut into 5 µm sections using a Leica HM355S microtome and mounted on Labsolute microscope slides (Th. Geyer, Renningen, Germany).
For chromogenic anti-MANF (1:800) immunostaining, the same protocol was used as for human sections, except with 1.5% goat serum for blocking, and no avidin–biotin blocking was used. Double immunofluorescence stainings with the rabbit anti-MANF antibody (1:200) were performed with goat anti-Iba1 (1:250, cat#ab5076, Abcam, Cambridge, UK), mouse anti-CD68 (1:500, cat#MCA341R, AbD Serotec, Kidlington, UK), and mouse anti-NeuN (1:200, cat#MAB377, Millipore). Goat anti-rabbit Alexa488 (cat#A11034, Life Technologies, Paisley, UK) and goat anti-mouse Alexa568 (cat#A11004, Life Technologies) or donkey anti-rabbit Alexa488 (cat#A21206, Life Technologies) and donkey anti-goat Alexa568 (cat#A11057, Life Technologies) were used as secondary antibodies (1:500), the slides were coverslipped with Vectashield Hardset Antifade Mounting Medium with DAPI (cat#H-1500, Vector Laboratories) and imaged with a Zeiss LSM 700 confocal microscope.
The specificity of the anti-MANF antibody was confirmed with additional pre-adsorption controls and NestinCre/+::Manffl/fl knockout tissue.
Analysis of infarction size
The rats were euthanized 2 days after dMCAo, the brains were sliced into seven 2 mm coronal sections, and stained with 2% 2,3,5-triphenyltetrazolium chloride (Sigma Aldrich) in PBS for 15 min at room temperature. The sections were transferred into 4% paraformaldehyde (Sigma Aldrich) for fixation, scanned and analyzed with open-source ImageJ software. To calculate the infarction volume, the infarct area of each section was first corrected for brain swelling by subtracting the area of non-infarcted ipsilateral hemisphere from the total area of contralateral hemisphere [
37], then multiplied by the thickness of the section, and finally the infarct volume of each section was summed up to provide a total infarct volume for each animal.
In NestinCre/+::Manffl/fl knockout mice and Manffl/fl control mice, the infarction area was analyzed from hematoxylin stained paraffin sections 14 days post-stroke. The infarction area was delineated in Pannoramic Viewer programme and the infarction size was calculated as percentage of the total brain area of each section. The average infarction size was calculated from three sections per each animal (2 striatal and 1 thalamic section).
Cytokine ELISA
The ipsilateral cortical tissues and serum were collected 1 day after dMCAo from rats. The tissue samples were homogenized in lysis buffer (PRO-PREPTM, iNtRON Biotechnology, Korea), centrifuged at 12 000 g for 30 min, and the supernatants were collected and stored at -80 °C. The cytokine (TNFα, IL-1β, and IL-6, IL-10) levels were quantified using commercial ELISA kits (DY510; DY501; DY506; DY522; R&D Systems Minneapolis, MN, USA) according to manufacturer's instructions. The cytokine levels were normalized to the total protein concentration of the sample.
Mouse MANF ELISA
Endogenous MANF levels were quantified from mouse serum after permanent dMCAo with an in-house double antibody sandwich mouse MANF (mMANF) ELISA. Terminal blood samples were taken by cardiac puncture 1 h, 6 h, 24 h and 48 h after permanent dMCAo. The blood was let to coagulate for at least 30 min and serum was separated by centrifugation with 2,000 g at room temperature for 10 min and stored at −80°C until analysis. Sera were diluted 1:40 for quantitation. The development and validation of mMANF ELISA were described in detail elsewhere [
20]. The dynamic range of mMANF ELISA is 31.25–1000 pg/ml, and sensitivity 29 pg/ml. The assay detects both mouse and human MANF and does not give a signal from MANF knockout tissues. In brief, 96-well MaxiSorp plates were coated with goat anti-human MANF antibody (AF3748, R&D Systems) in 50 mM carbonate buffer, pH 9.6. After blocking with 1% casein-PBST (PBS; 0.05% Tween 20), mouse sera and standard samples (mMANF; CYT-827, ProSpec, Rehovot, Israel) diluted in blocking buffer were applied to the plate and incubated overnight at +4°C. After washing, bound MANF in the wells was detected using rabbit anti-MANF antibody (LS-B2688, LSBio, Seattle, WA, USA), followed by incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG secondary antibody (NA9340V, GE Healthcare, Chicago, IL, USA). The color signal was developed using the DuoSet ELISA Development System and absorbance was read at 450 nm and 540 nm using a VICTOR3 plate reader (PerkinElmer, Waltham, MA, USA).
125I-labeled MANF
The rats (n = 10) underwent 60 min dMCAo together with CCA occlusion. Immediately after reperfusion, a mixture of unlabeled rhMANF (1 µg/µl, Icosagen) and
125I-rhMANF (approximately 1.4 ng/µl; 54058 CPM/µl), labeled by lactoperoxidase
-catalyzed radioiodination [
10], in 80 mM Na-phosphate buffer (pH 7.5) containing 1% bovine serum albumin (Sigma Aldrich), was administered intranasally (10 µl per each nostril). After 60 min the animals were perfused transcardially with 200 ml of saline. Blood samples were collected by cardiac puncture before perfusion. Liver samples were collected after perfusion and used for radiolabel quantification without homogenization. The whole brains were collected and homogenized in ELISA lysis buffer as described below. Half of the brain lysate was used for measuring the amount of radioactivity (counts per minute) and the signal was normalized to the original brain weight. Radioiodine content was quantified using the Perkin Elmer-Wallac Wizard 1480 Gamma Counter. Background radioactivity was subtracted from the gamma counts.
Sample processing for hMANF ELISA
Blood samples were taken either from the tail vein or by cardiac puncture before perfusion. Sera were prepared as described above and stored at −80°C until analysis. Before the brains were collected, the animals were perfused transcardially with 200 ml of saline to remove blood. The whole brain was snap-frozen in isopentane on dry ice and stored in −80°C until homogenization. The whole brains of rats treated with intranasal 125I-labeled rhMANF were homogenized by grinding in liquid nitrogen and lysed with ELISA lysis buffer (137 mM NaCl; 20 mM Tris–HCl, pH 8.0; 2.5 mM EDTA; 1% NP40; 10% glycerol; 660 mg tissue/ml) containing protease inhibitors (Complete, Mini, EDTA-free Protease Inhibitor Cocktail, Roche, Mannheim, Germany). Half of the brain lysate was further processed for hMANF ELISA and half of the lysate was used as such to measure gamma counts as described above. From the brains of rats treated with i.v. rhMANF, the infarcted cortex and corresponding contralateral cortex were dissected out and homogenized with ELISA lysis buffer containing protease inhibitors. The brain lysates were incubated on ice for at least 20 min, centrifuged 15 000 g at 4°C for 20 min, and the supernatants were collected.
Human MANF ELISA
The rhMANF protein levels from rat brain supernatants and serum were analyzed with an in-house double antibody sandwich ELISA specific for human MANF (hMANF) as described previously except without heterophilic antibody blocker [
18]. The dynamic range of hMANF ELISA is 62.5–2000 pg/ml, and sensitivity 45 pg/ml. The sera samples were diluted at 1:10, 1:20, 1:75 or 1:100 and brain homogenates at 1:2 in blocking buffer [1% casein in PBS; 0.05% Tween 20 (PBST)] for quantitation. Briefly, 96-well MaxiSorp (Nunc, Fischer Scientific, Waltham, MA, USA) were coated with goat anti-human MANF antibody (AF3748, R&D Systems, Minneapolis, MN, USA) in 50 mM carbonate buffer, pH 9.6. Wells were blocked using blocking buffer, and diluted samples and standards (hMANF; P-101-100, Icosagen) were applied to wells for duplicate measurements, and incubated overnight at +4°C. After washing with PBST, HRP-conjugated mouse anti-human MANF antibody (4E12, Icosagen) was applied to the wells and incubated for 5 h at room temperature. HRP signal was generated using DuoSet ELISA Development System (R&D Systems) according to manufacturer’s instructions, and the absorbance was read at 450 nm and 540 nm (for wavelength correction) using a VICTOR3 plate reader (Perkin Elmer). The rhMANF concentration in brain supernatant was normalized to the total protein concentration of the sample determined with the Lowry method (DC Protein Assay, Bio-rad Laboratories, Hercules, CA, USA).
The integrity of blood–brain barrier (BBB) was evaluated with Evans blue extravasation 2 days after dMCAo [
36]. The rats (n = 13) were anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and infused via the right femoral vein with 37°C Evans blue dye (2% in 0.9% normal saline, 4 ml/kg) over 5 min. Two hours later, the rats were perfused with 300 ml of normal saline to wash out any remaining dye in the blood vessels and then the brains were removed and sectioned to 2 mm thickness with a rodent brain matrix. Coronal brain sections were taken starting at + 2 mm and ending at − 2 mm from bregma. BBB permeability was evaluated in the contralateral non-ischemic cortex, ipsilateral ischemic cortex, and in the cerebellum. The cerebellum was used as an internal control. Each tissue sample was weighed immediately and placed in 0.5 ml of 0.9% normal saline for homogenization of the sample. For protein precipitation, 0.5 ml of 60% trichloroacetic acid solution was added and vortexed for 2 min. The mixture was subsequently cooled down at 4°C for 30 min and centrifuged (1500 g at 4°C) for another 30 min. The absorbance of Evans blue in the supernatant was then measured with a spectrophotometer (Molecular Devices OptiMax, USA) at 610 nm. The amount of Evans blue dye in the sample was calculated from a standard curve obtained from known amounts of the dye and was expressed as μg/g of the net tissue weight.
Statistical analysis
GraphPad Prism (version 9.2.0, GraphPad Software, San Diego, CA, USA) was used for statistical analysis. Normal distribution of each dataset was analyzed by Shapiro–Wilk test. Normally distributed data were analyzed with two-tailed Student’s
t-test, one-way ANOVA or two-way repeated-measures ANOVA. Data with non-normal distribution were analyzed with nonparametric Mann–Whitney U test and corrected for multiple comparisons with the Holm-Šídák method, when applicable. The multiple comparisons adjusted
p values are reported for the Mann-Whitney U test. Statistical significance was considered at
p < 0.05. The results are presented as mean ± standard deviation. All the used statistical tests, related to each figure, and their results are shown in the statistical table (Table
1). Exclusion criteria: In order to validate the effectiveness of the middle cerebral artery occlusion (MCAo) procedure, we meticulously assessed the accuracy of ligation. As exclusion criteria, animal that exhibited evidence of unsuccessful ligation resulting in an absence of stroke-induced lesions was eliminated from the study, and this was determined by the laser Doppler measurements.
Table 1
Statistical table
a | | Normal distribution (Shapiro–Wilk p = 0.115) | t-test | p = 0.641 |
b | | Normal distribution (Shapiro–Wilk p = 0.942) | One-way ANOVA | F(4,18) = 0.41; p = 0.801 |
c | | Normal distribution (Shapiro–Wilk p = 0.222) | t-test | p = 0.038 |
d | | Non-normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: Section 1: p = 0.900 Section 2: p = 0.900 Section 3: p = 0.664 Section 4: p = 0.195 Section 5: p = 0.121 Section 6: p = 0.201 Section 7: p = 0.900 |
e | | Non-Normal distribution (Shapiro–Wilk p = 0.016) | Mann–Whitney U | Adjusted p values: p ≥ 0.771 |
f | | Non-normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: d2: p = 0.750 d7: p = 0.051 d14: p = 0.0004 |
g | | Non-normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: d2: p = 0.715 d7: p = 0.058 d14: p = 0.0003 |
h | | Normal distribution (Shapiro–Wilk p = 0.109) | Two-way RM ANOVA | Time x Treatment interaction: F(2,54) = 0.13; p = 0.878 |
i | | Normal distribution (Shapiro–Wilk p = 0.172) | Two-way RM ANOVA | Time x Treatment interaction: F(2,54) = 0.01; p = 0.993 |
j | | Normal distribution (Shapiro–Wilk p = 0.241) | Two-way RM ANOVA | Time x Treatment interaction: F(3,81) = 0.39; p = 0.760 |
k | | Normal distribution (Shapiro–Wilk p = 0.674) | t-test | p = 0.006 |
l | | Non-normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney | Adjusted p values: Section 1: p = 0.244 Section 2: p = 0.244 Section 3: p = 0.209 Section 4: p = 0.209 Section 5: p = 0.113 Section 6: p = 0.209 Section 7: p = 0.244 |
m | | Normal distribution (Shapiro–Wilk p = 0.833) | t-test | p < 0.0001 |
n | | Normal distribution (Shapiro–Wilk p = 0.908) | t-test | p < 0.0001 |
o | | Normal distribution (Shapiro–Wilk p = 0.900) | t-test | p < 0.0001 |
p | | Normal distribution (Shapiro–Wilk p = 0.167) | t-test | p < 0.0001 |
q | | Non-Normal distribution (Shapiro–Wilk p = 0.025) | Mann–Whitney U | p = 0.005 |
r | | Non-Normal distribution (Shapiro–Wilk p = 0.032) | Mann–Whitney U | p = 0.511 |
s | | Normal distribution (Shapiro–Wilk p = 0.480) | t-test | p = 0.038 |
t | | Non-Normal distribution (Shapiro–Wilk p = 0.007) | Mann–Whitney U | p = 0.015 |
u | | Non-Normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: 30 min: p = 0.004 1 h: p = 0.004 |
v | | Non-Normal distribution (Shapiro–Wilk p = 0.004) | Mann–Whitney U | Adjusted p values: p ≥ 0.111 |
w | | Normal distribution (Shapiro–Wilk p = 0.099) | Two-way RM ANOVA (Mixed-effects model) | Time x Treatment interaction: F(3,38) = 2.66; p = 0.062 |
x | | Non-Normal distribution (Shapiro–Wilk p = 0.044) | Mann–Whitney U | Adjusted p values: p ≥ 0.151 |
y | | Non-Normal distribution (Shapiro–Wilk p = 0.018) | Mann–Whitney U | Adjusted p values: p ≥ 0.733 |
z | | Non-Normal distribution (Shapiro–Wilk p = 0.012) | Mann–Whitney U | Adjusted p values: p ≥ 0.340 |
aa | | Non-Normal distribution (Shapiro–Wilk p = 0.0005) | Mann–Whitney U | Adjusted p values: p ≥ 0.419 |
ab | | Normal distribution (Shapiro–Wilk p = 0.159) | Two-way RM ANOVA (Mixed-effects model) | Time x Treatment interaction: F(3,38) = 0.70; p = 0.555 |
ac | | Non-Normal distribution (Shapiro–Wilk p = 0.0006) | Mann–Whitney U | Adjusted p values: p ≥ 0.186 |
ad | | Non-Normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: p ≥ 0.999 |
ae | | Non-Normal distribution (Shapiro–Wilk p < 0.0001) | Mann–Whitney U | Adjusted p values: p ≥ 0.997 |
af | | Normal distribution (Shapiro–Wilk p = 0.715) | t-test | p = 0.093 |
Discussion
This is the first study to show that after stroke MANF protein expression is triggered in microglia/macrophages in the human brain parenchyma. Similar spatiotemporal changes in MANF protein expression are found in the ischemic human and rodent brain. During the first days after ischemic stroke, MANF protein expression was decreased in the infarct core in both patients and rodents. However, when microglia/macrophages are activated in the post-ischemic brain, MANF protein expression is intensely present in those cells. The post-stroke inflammatory response is more delayed in human patients than in rodents, and the number of microglia/macrophages in ischemic brain tissue has been shown to be highest around 2 weeks post-stroke in patients [
42], whereas in rats the peak is already at day 7 [
5]. We observed that MANF had a similar expression pattern than the phagocytic marker CD68 in both patients and experimental animals, thus making dMCAo a relevant model for studying the role of MANF in inflammation and ischemia. The most important finding of this study is that there is drastic change of MANF protein expression towards microglia/macrophages after stroke. Although, in non-injured brains, MANF mRNA levels are relatively high in all cell types, the immunoreactivity experiments with MANF antibodies show protein to be found primarily in neurons. However, after stroke, there is a drastic change in this, and brain microglia/macrophages express robustly MANF protein. We showed this with human stroke patients'
postmortem brains, rat brains and mouse MANF knockout brains where MANF was deleted from the neuronal lineage of cells (neurons, astrocytes, oligodendrocyte precursors and oligodendrocytes). Moreover, to bridge the gap between the translation of our findings from rodent models to humans, we explored more how systemic administration of MANF can enhance the recovery process after stroke using rat cortical ischemia–reperfusion model and found that enhancement of recovery correlates with inflammatory biomarkers.
The observation that the cerebral response to acute focal injury is reflected in the whole cerebrum is not new. We observed CD68 positive cells also in the contralateral hemispheres, especially in the white matter of infarcted human brains. In the same samples, elevated cyclo-oxygenase 2 and tumor necrosis factor α immunoreactivity, and increased density of intercellular adhesion molecule 1-expressing microvessels were found in the contralateral hemisphere when comparing to non-infarcted control brains [
42,
56,
57]. These data suggest that inflammatory changes occur also in the contralesional human brain after ischemic stroke. Grossly, the contralateral hemisphere is not always completely intact tissue since ischemic stroke can cause brain herniation also influencing the contralateral side. Also changes in electrical activity, cerebral blood flow, and metabolism are known to occur contralesionally after ischemic stroke possibly representing diaschitic or reparative effects after injury [
4], and may include inflammatory events as well. CD68 positive cells have been found in the healthy brain as well, particularly in the white matter, and to increase with age [
27,
28]. Therefore, it is also possible that the observed CD68 expression in the contralateral hemisphere of infarcted brains represents basal CD68 expression in the human brain.
Manf mRNA is highly expressed in mouse microglia in the healthy brain [
82] but microglial MANF protein expression has been reported to be very low [
60]. We observed some MANF positive microglial cells in the contralateral hemisphere of rat brains, but it seems that translation of MANF protein is strongly induced after ischemia in activated microglia/macrophages. There are several possibilities for why MANF translation could be induced in the reactive microglia/macrophages. MANF may be needed for the increased production and secretion of proteins such as cytokines, or for ER remodeling during the activation of microglia/macrophages. The increased protein production in the activated microglia/macrophages may induce unfolded protein response and subsequent MANF translation. Shen
et al. showed that the ER chaperone GRP78 and MANF were both induced in activated microglia after ischemia [
60]. They also showed that microglial MANF upregulation is not specific for ischemia as the ER stress inducer tunicamycin triggered MANF expression as well and caused microglial activation in primary cultures. Moreover, activated microglia/macrophages are highly migratory and MANF may be needed in the ER for enabling motility as ER needs to be flexible in motile cells. It has been shown that endogenous and exogenous MANF is important for neuronal migration [
69,
70]. Since there is large demand to ER in phagocytic cells, we postulate that MANF protein expression in microglia/macrophages is induced because of the load that the changing morphology of the ER in phagocytic cells induces. Moreover, we have not been able to distinguish between microglia and macrophages, and whether there are differences in MANF protein expression in the two cell types. It is also possible that MANF immunoreactivity in microglia/macrophages after stroke is originating from circulating MANF, and in this case these cells would take it up and therefore become MANF-positive for immunoreactivity. We have previously shown that in
Manffl/fl control mice and in
NestinCre/+::Manffl/fl mice
Manf mRNA expression is increased in the ischemic cortex after stroke [
48]. In addition, we have previously shown that
NestinCre/+::Manffl/fl mice have increased lesion volume 2 days after stroke [
48], indicating that endogenous neuronal MANF is neuroprotective. Moreover, the microglia field is developing promptly, and new microglia phenotypes have been identified [
62,
63], that remain to be studied in relation to MANF or stroke.
MANF may also be important for frank phagocytosis as it is expressed particularly in the round, most reactive state of microglia/macrophages, or MANF may be needed for immune cell recruitment. It has been shown that AAV-MANF increases the number of CD68 positive cells in the peri-infarct area 4 days after dMCAo [
48] and that intravitreal rhMANF injection increases the number of CD11b positive cells in the damaged retina [
51].
Additionally, MANF may be secreted from immune cells and could help to restore the homeostatic environment in the injured tissue. At least in vitro, endogenous MANF is known to be secreted from non-neuronal cells and the secretion is greatly enhanced upon ER Ca
2+ depletion [
7,
21,
65]. In vivo, MANF has been shown to modulate microglia/macrophage activity toward the regenerative M1 type in a paracrine manner [
51,
77]. It would be highly interesting to generate a mouse line with microglia/macrophage specific MANF deletion and study whether the recovery from stroke is hindered in these mice compared to wild type.
Our understanding of the role of exogenous MANF and how it mediates its cytoprotective effects is still limited [
45]. Neuroplastin has been suggested to function as a plasma membrane receptor for MANF with modest binding [
75] and sulphatide-mediated cellular uptake has been postulated as the mechanism of how extracellular secreted/exogenous MANF enters cells [
8]. Endogenous MANF seems to be important primarily in maintaining ER homeostasis [
38,
52]. However, the role of exogenous MANF may be different, and exogenous MANF may interact with multiple target proteins intracellularly. Interestingly, the N-terminal RTDL amino acid sequence of MANF, which functions as an ER retention signal, is not required for the in vivo neuroprotective effect of recombinant MANF protein in cerebral ischemia [
50], implying that the neuroprotective effect of exogenous MANF may not be directly related to ER homeostasis. Furthermore, increasing amount of data shows that MANF has immunomodulatory effects (recently reviewed in [
45].)
We measured endogenous MANF levels from mouse serum at different time points during the first 2 days after ischemic stroke but found no difference compared to naïve animals. The level of endogenous MANF in serum was 4.6 ng/ml in mouse. In humans, serum MANF concentration has been reported to be between 3.5 and 6 ng/ml in healthy adults [
18‐
20,
61,
71] and less, approximately 2.5 ng/ml, in the aged [
61]. Increased serum MANF protein levels have been reported in patients with Parkinson’s disease diagnosed on average 6 years ago [
19], indicating that circulating MANF levels may be altered in chronic CNS disease. However, we were interested in the potential of MANF in acute diagnostics of stroke, but our data do not support the use of free circulating MANF as a potential biomarker in ischemic stroke during the first 2 days after stroke. Furthermore, it is unclear from where the free circulating MANF originates. In human blood cells, MANF protein has been detected mostly in platelets, to some extent in leukocytes, and very little in red blood cells but at least in the case of Parkinson’s disease patients the increased serum MANF levels were not originating from the blood cells [
19]. However, more relevant biomarker for stroke could be detection of cytokines from the serum. The amount of pro-inflammatory cytokines was downregulated, and anti-inflammatory cytokines upregulated in the infarcted cortex and serum 24 h after intravenous rhMANF treatment. Similar results were found by Han
et al. [
23] where intraventricular rhMANF therapy decreased pro-inflammatory cytokine levels in the infarction area after MCAo in aged mice. The anti-inflammatory effect of rhMANF was shown to be dependent on the TLR4/MyD88/NF-κB pathway in vitro [
23]. A vast amount of data from different disease models has shown that MANF regulates the NF-κB pathway [
45].
Intracranial delivery of MANF is neuroprotective but not feasible in ischemic stroke patients due to increased risk of hemorrhage caused by thrombolytic treatment and the infarct itself, and therefore it is vital to explore the possibility for non-invasive delivery. There is evidence that intranasally delivered peptides reach the central nervous system in humans [
55]. As an attempt to develop a non-invasive method for administering MANF, we show for the first time that intranasally delivered rhMANF reduced infarction size and promoted post-stroke recovery in rats. However, only 0.003% of the
125I-rhMANF dose was detected in the brain 1 h after intranasal administration and would theoretically result in approximately 20 pM brain concentrations. Most rhMANF after intranasal administration was found in the blood. The calculated bioavailability of
125I-rhMANF in the blood (0.4%) was higher compared to unlabeled rhMANF in the serum (0.05%) which may suggest either counting of free radioactivity, binding of rhMANF to a carrier protein masking the epitopes for ELISA, or binding of rhMANF to blood cells that are removed during sera sample preparation. The low levels of
125I-rhMANF detected in the brain may also have originated from the systemic circulation after stroke-induced disruption of the BBB. It has been reported that in the intraluminal MCAo rat model there is a leakage of large molecules through the BBB already 25 min after reperfusion [
1,
64]. Intranasal rhMANF therapy increased the MANF serum concentration about 0.85 ng/ml i.e., 20% compared to the endogenous level. However, interindividual variation in the rhMANF serum levels was high, likely reflecting differences in absorption in individual rats, which is a known problem in the intranasal delivery route [
44]. The interindividual variation after intranasal rhMANF delivery is reflected in the stroke outcome as the variation in the infarction volume was high as well. It is possible that the neuroprotective effect of intranasal rhMANF was due to systemic effects or effects on the brain endothelium. As the concentrations in the brain and serum were low after intranasal MANF administration, it is likely that the neuroprotective effect is mediated by other than direct effect in the brain. It remains speculative since we do not know it and since the mechanism of action of exogenously added MANF still remains mostly elusive, it is difficult to speculate further. Therefore, we chose to administer rhMANF i.v. and found neuroprotective effects on lesion volume also after i.v. therapy. Surprisingly, interindividual variation in rhMANF serum levels was still high. Furthermore, the half-life of rhMANF in the serum was short, about 10 min, and significantly shorter than the half-life in the brain parenchyma, which is expected to be 5.5 h [determined for the MANF homolog, rhCDNF [
49]. The fast clearance of rhMANF from blood circulation could reflect rapid proteolytic degradation, renal excretion, hepatic metabolism, tissue distribution, or binding to blood cells. Proteins smaller than 70 kDa are known to eliminate via renal clearance [
81]. The amount of rhMANF in the infarcted cortex after intravenous administration was small (average 32 pg/mg of total protein) compared to the amount of endogenous MANF in the mouse brain [250 ng/mg of total protein [
14]], only 0.01% of the endogenous amount, indicating that the brain levels acquired after systemic administration of rhMANF are not biologically significant. Therefore, we hypothesize that the therapeutic effect of rhMANF in ischemic stroke derives from the systemic circulation, possibly via modulation of immune cell phenotype. Additionally, in stroke and traumatic brain injury models, intracranial MANF treatment has been shown to decrease brain edema and BBB leakage [
23,
34,
35,
74]. However, we found no statistically significant effect on the BBB integrity after i.v. rhMANF delivery in ischemic stroke.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.