Background and Purpose: Hypoxic-ischemic brain injury, due to reduced supply of oxygen to brain, is a major cause of death and disability. There is no exclusive treatment available so far. Glycine-2-Methylproline-Glutamate (G-2MePE, NNZ 2566), an analogue of Glycine-Proline-Glutamate reduces neuronal injury after focal ischemia in adult rats. The current study investigated into the neuroprotective effects of G-2MePE after global hypoxic-ischemic brain injury in adult rats.Methods: Adult male rats received a single sub cutaneous injection of G-2MePE (1.2mg/kg) 3h post hypoxic-ischemic brain injury or the same volume of normal saline. Brains were extracted 5 days after the treatment. Tissue damage in the cortex, hippocampus and striatum was assessed. Neuronal survival, glial reactions, caspase-3 activity and TNF-α cytokine activity were also assessed.Results: The treatment with G-2MePE was associated with a significant reduction of tissue damage, improvement in neuronal survival, reduction in reactive microglia, TNF-α positive cells and caspase-3 positive cells in hippocampus and cortex but an elevation of astrocytosis.Conclusions: Neuroprotection with G-2MePE after hypoxic-ischemic brain injury in adult rats is associated with reduced neuronal necrosis, apoptosis, modulated inflammatory responses and augmented astrocytosis.
Stroke is a leading cause of death and disability in developed countries affecting 1 in 250 persons. Ischemic stroke is the most common type which occurs when blood supply to the brain is reduced, causing neurons starving for oxygen, resulting in neuronal death or dysfunction, culminating in corresponding loss of neurologic function (Dirnagl et al. , 1999, Chan, 2001, Hayashi and Abe, 2004). A range of promising drugs have been tested in vitro and in vivo in the past; but the majority have failed to translate to the clinic (Gleichmann et al. , 2000, Kawano et al. , 2001, Domanska-Janik et al. , 2004, Mohammad et al. , 2004, Wahlgren and Ahmed, 2004, Kilic et al. , 2006, Switzer and Hess, 2006). One obvious reason is the lack of animal models that represent the pathogenesis and pathology of human stroke. Thus testing the potential compound using multiple animal models may provide a more objective view for their preclinical development. The only therapy with proven benefits for acute ischemic stroke patients is the intravenous thrombolytic therapy with recombinant tissue-type plasminogen activator (rt-PA) within 3h of the onset of stroke (Fisher and Schaebitz, 2000, Acampa et al. , 2014), the window which has been recognised as critical for preventing neuronal death after stroke .
N-terminal glycine-proline-glutamate (GPE) is a neurobioactive tri peptide, naturally cleaved from IGF-1 by an acid protease (Sara et al. , 1989). GPE crosses the blood brain barrier (BBB) and can prevent brain injury after embolic-stroke in aged rats (Guan et al. , 2004, Shapira et al. , 2009) and after hypoxic-ischemic (HI) injury in developing adult brain (Sizonenko et al. , 2001). GPE is small and does not activate the IGF-1 receptor (Alonso De Diego et al. , 2006). GPE has a short half-life (<2min) in plasma due to a rapid peptidase-mediated breakdown (Batchelor et al. , 2003).
G-2MePE is a structural analogue of GPE designed to improve protease resistance by alpha methylation of the proline moiety (Harris and Brimble, 2006). It is known that G-2MePE can reduce injury after focal brain injury and trauma (Bickerdike et al. , 2009) and HI injury in neonatal rat (Svedin et al. , 2007). Reduced inflammation has been reported among the protective effects of G-2MePE (Svedin et al. , 2007, Lu et al. , 2009, Wei et al. , 2009, Cartagena et al. , 2013). There are various rodent models of stroke designed for testing the efficacy of neuroprotectants for clinical use with some of them reproducing the pathological changes and others mimicking the pathophysiological events of human stroke. Focal injury models are often severe with limited windows of opportunity to intervene and short outcome evaluation at 24h (Zhou et al. , 2013, Gheibi et al. , 2014). Therefore testing G-2MePE in different animal models of focal brain injury is desirable. HI induces focal brain damage through a global hypoxic insult with local reduction in blood flow.This approach allows progressive brain damage and provides the opportunity to evaluate longer term outcome. Thus the aim of the study was to examine the treatment effects of G-2MePE after HI brain injury in adult rats.
Material and MethodsAnimals
All experimental protocols in this study were approved by the Auckland University Ethics committee. Adult male Wistar rats (280-310g) were obtained from the Animal Resources Unit of the University of Auckland. These rats were housed with 12h dark: 12h light cycle at room temperature with 60% relative humidity and free access to food and water ad libitum.
HI brain injury
The procedure has been previously described (Guan et al. , 2004). Animals were anesthetised with 3% isoflurane/oxygen and the right common carotid artery was permanently ligated with 0/3 silk thread. Rats were left at room temperature for 2h and then placed in an incubator at 90±5% relative humidity and temperature 31±0.5oC for 30 min. Hypoxia was induced by quickly reducing the oxygen level to 6% by introducing nitrogen gas and then maintained at 6% ± 0.2% level by adjusting the flow of nitrogen and air. Hypoxia continued for 15 min under close monitoring.
Drug preparation and administration
G-2MePE (Neuren Pharmaceuticals, Auckland) was dissolved in 0.9% saline solution. The experimenter was masked to the treatment groups throughout the experiments. Rats were randomly divided into two treatment groups for receiving either vehicle (0.3 ml saline, n=20) or an equal volume of G-2MePE dissolved in saline (1.2mg/ kg, n=18) subcutaneously (SC), 3h after the HI. After the SC injection, animals were housed in groups of 3 in the holding room and weighed daily to ensure their well-being.
Brain histological preparation
Five days after the experiments rats were euthanized by an intraperitoneal injection of pentobarbital (125mg/kg) and the brains were transcardially perfused with normal saline followed by fixation with 10% neutral buffered formalin, by gravity feed. The brains were left in formalin solution for 2 days for continued fixation, coronally sliced to 3mm blocks, processed and paraffin embedded. 8μm thick coronal sections were cut at 3 levels i.e., at the level of striatum, at the level of cerebral cortex & dorsal hippocampus and at the level of cortex & ventral horn of the hippocampus (Guan et al. , 2000, Shapira et al. , 2009) and mounted on albumen-chrome alum pre-coated slides.
Brain damage assessment
The degree of tissue damage was assessed using a scoring system after thionin-acid fuchsine staining. Briefly, deparaffinised and hydrated sections were stained with aqueous solutions of thionin and acid fuchsine. Three coronal sections from each brain region were used for scoring in the ipsilateral (ischemic) hemisphere. Tissue damage was assessed at three levels anterior from inter-aural reference zero plane, (A-P 6, 4.5 and 4.2mm) for the lateral cortex; at two levels (A-P 4.5 and 4.2mm) for the hippocampus & the dentate gyrus and at one level (A-P 6mm) for the striatum respectively (Paxinos and Watson, 1982). A 4 point scoring system was used to assess the histopathological damage (0 = no damage; 1 = few cells damaged; 2 = < half of brain tissue damaged; 3 = > half of the tissue damaged and 4 = almost all neurons died and no tissue survived) (Rochester et al. , 2005). Brain tissue with selective neuronal death, pan-necrosis and cellular reaction was treated as damaged (Guan et al. , 2000). Neuronal death was confirmed by the presence of acidophilic (pink) cytoplasm and contracted nuclei (Brown and Brierley, 1972, Auer et al. , 1985, Guan et al. , 1993). Scoring was done by an experimenter masked to the study. Average tissue damage scores in different regions were statistically analysed.
The extent of neuronal survival in different parts of the brain, caspase-3 expression and inflammatory responses in the cortex & hippocampus were evaluated by immunohistochemical staining in brain paraffin preparations from experiment 1. 10 sections each from the experimental groups were taken for each antibody labelling. Primary and secondary antibodies used and their dilutions were as follows. (i) monoclonal mouse anti-NeuN for labelling surviving neurons (Chemicon International, Temecula, CA, USA, 1:200) (ii) goat polyclonal anti-ionised calcium-binding adaptor molecule-1(Iba-1) for reactive microglia (Abcam, via Sapphire Bioscience, Hamilton, New Zealand , 1:200) (iii) polyclonal rabbit anti-glial fibrillary acidic protein (GFAP) for reactive astrocytes (Chemicon International, Temecula, CA, USA, 1:500), (iv) monoclonal mouse anti-TNF-α for pro-inflammatory cytokine TNF-α (AbD Serotec, Auckland, 1:200) and (v) rabbit polyclonal anti-cleaved caspase-3 ASP 175 for cells undergoing caspase-3 mediated apoptosis (Cell Signalling Technology, U.K; 1:200) (vi) biotin conjugated anti-mouse, anti-goat and anti-rabbit secondary bodies (Vector Laboratories, Burlingame, CA, 1:200).
Xylene-deparaffinised slides were hydrated in decreasing concentrations of ethanol and washed with 0.1m/L phosphate buffered saline except in the case of Iba-1 where Tris-buffered saline (TBS) was used. Antigen retrieval was done by heating the slides in citrate buffer (pH 6.0) at near boiling temperature (95oC) in a water bath for 20 minutes. Quenching of endogenous peroxidase was done by treating with 1% H2O2 in methanol for 30 min. After washing, blocking was done with appropriate 5% serums prepared in 0.1m PBS. The slides were then incubated with respective primary or secondary antibodiesdiluted in 2.5% serums,each overnight, at 40C. PBS washed slides were then incubated with ExtrAvidin®-peroxidase (Sigma-Aldrich Pty. Ltd) for 3h at room temperature. Finally sections were treated with SIGMA FAST (™) 3, 3’- diaminobenzidine tetrahydrochoride (Sigma-Aldrich Pty. Ltd). Negative control sections were processed in the same way except that primary antibody was excluded from the incubating solution. Caspse-3 immunolabelled slides were counterstained with weak thionin solution to distinguish the area for counting.Alcohol dehydrated sections were then cleared in xylene and permanently mounted with DPX.
All immunohistochemical assessments were done in a masked fashion in both ipsilateral and contralateral hemispheres. Neuronal survival was assessed by estimating NeuN positive cells in the lateral cortex, CA1-2, CA3, CA4 sub regions of the hippocampus, dentate gyrus and the striatum. Iba-1, TNF-α, GFAP and Caspase-3 positive cells were estimated in cortex and hippocampus. Immunopositive cell counts were done using Stereo Investigator software (Ver. 10, MBF Bio Science, Vermont, USA) by light microscopy (Nikon eclipse 80i, Tokyo, Japan). A contour was drawn at 2x magnification covering the region of interest. The cells were counted after random sampling by Stereo Investigator at 40x magnification by applying a fractionator probe consisting of a counting frame for object inclusion/exclusion. The grid and counting frame sizes applied for cortex and striatum were 400×400μm and 100×100μm respectively and for hippocampal sub regions were 100×100μm and 75×75μm. Cells touching the bottom and left redlines of the counting frame were excluded whereas those touching the top and right green lines were included. Immunopositive cells thus estimated by the Stereo Investigator software were converted to cell density (cells/mm2) by applying the equation: Estimated number of immunopositive cells/ region of interest (µm2)] X 106.
The treatment effects of G-2MePE on the tissue damage scores (thionin/acid fuchsine), neuronal survival (NeuN), immunomodulation (Iba-1, GFAP, TNF-α) and apoptosis (Caspase-3) on the ipsilateral and contralateral sides of the brain were analysed using two way ANOVA followed by Bonferroni post-hoc tests, with brain regions and treatments as dependant factors. Data are presented as mean ± SEM. Statistical significance was assumed at P < 0.05.
Results G-2MePE treatment attenuated brain tissue damage after HI injury
HI induced unilateral damage in the middle cerebral artery territoryipsilateral to the ligated carotid arteries including the lateral cortex (Figure 1A), hippocampus (Figure 1 C, E G) and striatum (Figure 1I). The insult was associated with selective neuronal loss, acidophilic staining in cytoplasm, tissue pan-necrosis, cellular reactions and tissue infarction.
G-2MePE treatment improved neuronal survival
Photomicrographs show the distribution and morphological changes of neuronal survival in experiment 1(Figure 3A-J). Morphologically, the surviving neurons in the injured areas were smaller in size, with fewer processes and dendrites. Compared to the vehicle treated group, G-2MePE treated neurons showed almost normal neuronal morphology in all brain regions. There was no difference in number of neurons in the vehicle treated and G-2MePE treated experimental groups in the contralateral cortex, CA1-2, CA3, CA4, dentate gyrus or striatum (Figure 4A-F). HI injury caused significant loss of NeuN positive neurons in the ipsilateral cortex, sub regions of hippocampus and the striatum, with more severe loss in the cortex (p<0.001), CA3, CA4 areas of the hippocampus (p<0.001) and striatum (p<0.001). Sub cutaneous treatment with G-2MePE significantly improved the neuronal density (cells/mm2) associated with HI injury on the ipsilateral cortex (598.75±93.45 vs 882.52 ±60.52; p<0.05), CA1-2 (975.89±245.40 vs 1733.36 ±112.77; p<0.01), CA3 (371.55±78.62 vs 1082.64 ±134.36; p<0.001), CA4 (229.38±39.34 vs 414.57±17.67; p<0.001) areas of the hippocampus, the dentate gyrus (1229.52±309.56 vs 2897.94 ±155.04; p<0.001) and striatum (544.22±109.23 vs 1023.74 ±50.23; p<0.001) to a level similar to the counts in contra lateral counterpart area (Figure 4A-F).
G-2MePE SC treatment augmented astrocytosis
Astrocytes were present throughout hippocampus and cortex. HI injury was associated with induction of hypertrophic astrocytes with thicker cell bodies (Figure 5A, B). GFAP positive cell counts (cells/mm2) in the hippocampus and cortex were similar between the two hemispheres of the vehicle treated group hippocampus and cortex. Treatment with G-2MePE was associated with a significant increase in reactive astrocytes in the ipsilateral hippocampus (184.52±29.19 vs 302.79±47.73, p<0.05) and cortex (144.04±23.78 vs 291.18±42; p<0.001; (Figure 6A, B).
G-2MePE treatment inhibited microglial activation
HI injury was associated with marked increase in reactive microglia with amoeboid morphology, throughout the ipsilateral hemisphere. There was intense infiltration in the necrotic areas (Figure 5C). Compared to the contralateral hemisphere there was approximately a 9 fold increase in reactive microglia population in the ipsilateral hippocampus and 8.5 fold increase in the lateral cortex. Treatment with G-2MePE significantly reduced the density of microglia (cells/mm2) in the ipsilateral hippocampus (867.08±195.49 vs 201.37±88.98; p<0.001) and cortex (711.41±174.80 vs 105.65±20.24; p<0.001; Figure 6C, D).
G-2MePE treatment inhibited TNF-α expression
HI injury caused infiltration and release of pro-inflammatory cytokines. TNF-α positive cells were diffusely distributed in all sub-regions of the hippocampus and the cortex (Figure 5E, F). HI injury was associated with a 2.4 fold increase in TNF-α positive cells in the ipsilateral hippocampus and 2.7 fold increase in cortex. Treatment with G-2MePE significantly reduced TNF-α positive cell counts (cells/mm2) in the ipsilateral hippocampus (603.5±104 vs 297.71±36.5 p<0.01) and cortex (585.3±96.7 vs 281.4 ±28.95; p<0.01), almost to the vehicle control levels (Figure 6E, F).
G-2MePE treatment reduced caspase -3 mediated apoptosis
Hypoxic-ischemic injury was associated with induction of caspase-3 mediated apoptotic cells in the ipsilateral side of the hippocampus and cortex. Morphologically there was a gradient of caspase-3 labelled cells; most recently cleaved caspase being densely stained (Figure 5G, H). G-2MePE treatment significantly reduced the apoptotic cell counts (cells/mm2) in the ipsilateral hippocampus (132.63±17.72 vs 68.62±11.22; p< 0.01) and cortex (86.97 ±12.54 vs 53.60±8.02; p<0.05; Figure 6G, H).
G-2MePE is neuroprotective in adult rats when administered 3 h after HI injury sub-cutaneously. The mechanisms of neuroprotection may include blocking the necrotic or apoptotic pathways of cell death as well as modulating the immune response and augmenting astrocytosis.