The Combination of Human Urinary Kallidinogenase and Mild Hypothermia Protects Adult Rats Against Hypoxic-Ischemic Encephalopathy-Induced Injury by Promoting Angiogenesis and Regeneration

Objectives: Human Urinary Kallidinogenase (HUK) is a tissue kallikrein that plays neuroprotective role in ischemic conditions via different mechanisms. Mild hypothermia (MH) is another robust neuroprotectant that reduces mortality but does not profoundly ameliorate the neurological outcome in hypoxic-ischemic encephalopathy (HIE) patients. However, whether the combination of HUK and MH can be used as a promising neuroprotective treatment in HIE is unknown. Methods: One-hundred and forty-four adult Wistar rats were randomly divided into five groups: Sham, HIE, HUK, MH and a combination of HUK and MH treatment. The HIE rat model was established by right carotid dissection followed by hypoxia aspiration. The survival curve was created within 7 days, and the neurological severity scores (NSS) were assessed at days 0, 1, 3, and 7. Nissl staining, Terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL), immunofluorescent staining and western blotting were used to evaluate neuronal survival, apoptosis and necrosis, tight-junction proteins Claudin-1 and Zonula occludens-1 (ZO-1), vascular endothelial growth factor (VEGF), doublecortex (DCX), bradykinin receptor B1 (BDKRB1), BDKRB2 and Ki67 staining. Results: The combined treatment rescued all HIE rats from death and had a best survival curve compared to HIE. The Combination also reduced the NSS scores after HIE at days 7, better than HUK or MH alone. The combination of HUK and MH reserved more cells in Nissl staining and inhibited neuronal apoptosis and necrosis as well as significantly attenuated HIE-induced decreases in claudin-1, ZO-1, cyclin D1 and BDKRB1/B2 in comparison to HUK or MH treatment alone. Moreover, the combined treatment increased the expression of VEGF and DCX as well as the number of Ki67-labeled cells. Conclusions: This study demonstrates that both HUK and MH are neuroprotective after HIE insult; however, the combined therapy with HUK and MH enhanced the efficiency and efficacy of either therapy alone in the treatment of HIE, at least partially by promoting angiogenesis and regeneration and rescuing tight-junction loss. The combination of HUK and MH seems to be a feasible and promising clinical strategy to alleviate cerebral injury following HIE insult. Highlights: - The combination of HUK and MH distinctly reduces neurological dysfunction in HIE rats.- HUK enhances the neuroprotective effects of MH in HIE.- MH attenuates tight-junction disruption, upregulates the BDKR B1/2, DCX and cyclin D1.- The combination of MH and HUK enhances the expressions of MH/HUK mediated-BDKR B1/2, DCX, cyclin D1 and Ki67 positive cells.


INTRODUCTION
Hypoxic-ischemic encephalopathy (HIE) is a quite common and severe disorder in neonates, children and adults and is also a leading cause of mortality and morbidity. The neuropathogenesis of brain injury in patients with HIE remains largely elusive. The molecular mechanisms may include brain-bloodbarrier (BBB) breakdown, tight junction and microvascular disruptions, edema and cell death (Guardia Clausi et al., 2016;Kossatz et al., 2016;Thornton et al., 2017). A better understanding of the pathogenesis of HIE is needed, and it is necessary to develop various novel and effective treatments for HIE brain injury.
It has been documented that mild hypothermia (MH) is effective in the treatment of HIE. The potential mechanisms of MH may include the dramatic inhibition of autophagosomes and lysosomes; reduction of the activity of neuronal nitric oxide synthase (nNOS) and apoptotic caspase enzymes (Wood et al., 2016); downregulated expression of lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1), tumor necrosis factor-α (TNF-α) and oxidant species (Akamatsu et al., 2014); and induction of cold-shock proteins and heat shock proteins correlated with subsequent apoptosis (Lee et al., 2017). Clinically, although therapeutic MH alleviates cell damages and results in lower death rates in HIE (Mitra et al., 2017), the limited efficacy of hypothermia is insufficient and only slightly improves serious neurological impairments (Shankaran et al., 2012;Manley et al., 2017). Therefore, it is necessary to find a strategy to enhance the neuroprotective effects of MH in the treatment of HIE. These facts also inspired us to explore combined treatments of MH with certain neuroprotectants. In our previous study, we screened a series of medicines that might have synergistic effects or enhance the positive effects of MH (34 • C for 4.5 h) on primary cortical neurons following oxygen glucose deprived (OGD/R) treatment and identified the potential neuroprotectant human urinary kallidinogenase (HUK; Gao et al., 2014Gao et al., , 2016. Whether the combination of HUK and MH could also provide enhancing effects in the HIE adult model is of interest. HUK is a tissue kallikrein originally isolated from human urine that cleaves low molecular weight kininogen to release bradykinin (Stankowski and Gupta, 2011). HUK can mediate a complex series of physiological actions through its receptor signaling pathways, and suppress inflammatory mediators such as interleukin-1(IL-1), interleukin-6 (IL-6), or TNF-α, selectin E (ELAM-1), monocyte chemoattractant protein-1 (MCP-1), and Nuclear Factor-κB (NF-κB; Chen et al., 2010); increase erythrocyte deformability; prolong the re-calcification time; lower blood viscosity (Chao and Chao, 2006); and inhibit platelet aggregation, apoptosis, hypertrophy and fibrosis (Chao et al., 2010). Accumulated clinical evidence has suggested that HUK is an effective therapy for acute ischemic stroke (Zhang et al., 2012;Li et al., 2015;Miao et al., 2016;Ni et al., 2017). Recent studies by Han and Li indicated that HUK promotes post-ischemic angiogenesis and cerebral perfusion via induction of vascular endothelial growth factor (VEGF) and activation of bradykinin B1/B2 receptors (BDKRB1/B2), which further enhanced the expression of VEGF (Han et al., 2015;Li et al., 2015). In spite of the similar pathophysiological characteristics of HIE and ischemic stroke, both of which are caused by low blood and oxygen support and have ischemic penumbra (Gao et al., 2014), few studies have shown how HUK functions under HIE insult and whether HUK functions via mediating VEGF and BDKRB1/B2 receptors. Since VEGF and doublecortex (DCX) are usually taken as the indices of angiogenesis and neurogenesis respectively, in this study we used them as the markers of angiogenesis and neurogenesis (Couillard-Despres et al., 2005;Garcia et al., 2014;Zhao et al., 2017). Whether HUK exerts beneficial effects on neurons under HIE insult and how the combination of HUK and MH are synergistic or enhance positive effects need to be explored.
Therefore, in the present study, we sought to determine the following: (1) whether HUK enhances the efficacy of MH in terms of the rat survival rate, cell death, tight junction preservation and restoration of neurological function in the HIE adult rat model; (2) how the expression of VEGF and BDKRB1/B2 receptors mediates the additional protection of combined treatment; and (3) whether the combination of HUK and MH provides enhanced neurogenesis by regulating the expression of DCX and Ki67 compared to either treatment alone in HIE adult rats.

Ethics Statement
The experiments in this study were approved and carried out in accordance with the Institutional Animal Care and Use Committee of the Laboratory Animals Center, Southern Medical University. The protocol was approved by the Institutional Animal Care and Use Committee (Animal Ethic Approval No: NYFF-2015-87). We certify that the rats in our study were treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals  revised 1996 guidelines. We attest that all efforts were made to minimize the number of animals used and their suffering.

Experimental Design
A total of 144 adult male Wistar rats weighing 300-350 g were obtained from the Experimental Animal Center of Southern Medical University. The rats were randomly divided into five groups: Sham group, which underwent a sham operation; HIE group, which was subjected to right carotid dissection followed by hypoxia aspiration (8% O 2 with 92% N 2 for 3 h; Edwards et al., 2017) to simulate a medium to severe hypoxia-ischemic encephalopathy neurological severity scores (NSS; 7-18 scores); HUK group, which was injected with 0.0016 PNAU/100 g HUK (Techpool Bio-pharma, Guangzhou, China) via the tail vein twice a day in successive 2 days, starting from 1 h after inducing HIE; MH group, which was treated with MH (34 • C) for 4.5 h started from 1 h after inducing HIE; and combined treatment group (HUK+MH), which was administered with HUK and MH treatment as preceding methods. The 7-day survival curve and NSS (at days 0, 1, 3, 7) were determined to evaluate the effectiveness of HIE model. Nine rats in each group were euthanized, the brains were obtained from rats at 48 h after HIE to do the paraffin sections and were used for assessment using Nissl, Terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL), immunofluorescent staining and Western Blotting (Figure 1).

HIE Animal Model
The Wistar adult male rats were anesthetized using 5% isoflurane and maintained under 1.5% isoflurane anesthesia with 70% N 2 and 30% O 2 , retaining a left femoral artery intubation while monitoring the rectal temperature. HIE model was set up and modified following previous studies (Harding et al., 2016;Yang et al., 2016;Edwards et al., 2017). Briefly, the neck hair was shaved, and the skin was disinfected first. Then a 15 mm-long incision was made above the sternum and at the midline of the neck. Next, the subcutaneous tissue, sternohyoid and sternocleidomastoid muscles were bluntly separated. When the right common carotid artery pulse could be seen, we dissociated the common carotid artery and its two branches of the internal and external carotid artery, ligated the proximal and distal end of the internal carotid artery and mutilated the internal carotid artery in the middle of the two ligations. After confirming no bleeding, we finally sutured the muscles and skin layer by layer and disinfected the incision. Afterward, the rats were placed into a chamber under 8% hypoxia with 92% N2 inhalation for 3 h while maintaining the rectal temperature at 38 • C with a thermostatic pad. By using Transcranial Doppler (TCD) to verify that there was no blood flow in the right carotid, the model was considered to be successfully established.

7-day Survival Curves and Neurological Severity Scores (NSS)
Seven-day survival curves were made according to the time and quantity of the rat deaths in each group. Neurologic deficiency was assessed on the basis of modified NSS assessment. The NSS evaluation consisted of motor (0-6 scores), sensory (0-2 scores), reflex (0-4), and balance (0-6 scores) tests, with the results measured on a scale of 0-18. Score 0 indicates normal neurologic function, score 1-6 indicates mild deficiency, score 7-12 indicates moderate deficiency, score 13-18 indicates severe deficiency (score 18 indicates death). All the mice subjected to neurological severity assessments were evaluated by three blinded, trained investigators after inducing HIE (day 0) and were reassessed at days 1, 3, and 7.

Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick End-Labeling (TUNEL) Staining
Five sections of the dentate gyrus (DG) of right hippocampa in each rat were detected. In each section, five areas were randomly chosen to analyze. TUNEL staining was conducted using the in situ Cell Death Detection Kit (Roche, 11684795910, Germany) following the manufacturer's instructions. Briefly, we prepared 5 µm thick paraffin sections that were then immersed in dimethylbenzene I and II (Sigma Aldrich, 296333, USA) for 15 min and then sequentially dewaxed at 100%, 85%, 75% ethanol (Sigma Aldrich, 49836, USA) and distilled water for 5 min. Next, we incubated the sections with diluted protease K (DAKO, Cat. #S3020, USA) at 37 • C for 30 min and with Permeabilization Buffer (Aviva, OOMB00004, USA) for 20 min at room temperature. Finally, we incubated the sections with a mixed TUNEL working solution (TdT: dUTP = 2:29) at 37 • C for 2 h, stained the nuclei with DAPI (Sigma Aldrich, D9542, USA) at room temperature for 10 min, and sealed the sections with the anti-fluorescence quenching sealing agent Fluoroshield TM (Sigma Aldrich, F6182, USA). The prepared sections were photographed with a NIKON ECLIPSE TI-SR inverted fluorescence microscope (NIKON, Japan) with a UV excitation wavelength 330-380 nm, emission wavelength 420 nm (for DAPI); and FITC excitation 465-495 nm, emission 515-555 nm (for TUNEL). The images were quantified using Image-Pro Plus 6.0 software (Media Cybernetics Inc., Rockville, MD, USA). The ratio of the number of TUNEL-positive cells to the number of total nuclei (DAPI staining) was used to present the intensity of the TUNEL.

Nissl Staining
Nissl staining was used to observe neurons in the right hippocampal dentate gyrus (DG) of the insulted hemisphere with Toluidine Blue Staining Reagent (Sigma Aldrich, 89640, USA). Briefly, we dewaxed the paraffin sections (as for TUNEL staining) and stained them with 5% toluidine blue at room temperature for 10 min. Sections were then soaked in 95% ethanol (Sigma Aldrich, 49836, USA) for 2 min, dimethylbenzene (Sigma Aldrich, 296333, USA) for 3 min, and sealed with neutral balsam (Thermo Fisher Scientific Inc., B10100, USA). The Nissl staining slides were observed using a Nikon microscope (Nikon, Eclipse Ci-E, Japan). We used neuronal number per mm 2 to measure the survival neurons in Nissl staining. We counted the number of the neurons in 50 × 50 µm 2 of the initial images, then calculated the average neuron number per mm 2 .

Immunofluorescence
Five sections of the DG regions of each hippocampus were processed for the immunofluorescence detection. Immunofluorescence detection was conducted at 48 h after HIE. The paraffin sections were put into dimethylbenzene I and II (Sigma Aldrich, 296333, USA) for 15 min and then sequentially dewaxed at 100%, 85%, 75% ethanol (Sigma Aldrich, 49836, USA) and distilled water for 5 min, respectively. For antigen retrieval, we incubated the sections with citrate buffer antigen retrieval solution (pH 6.0; Sigma Aldrich, C9999, USA) in a microwave oven on medium-temperature heat for 8 min and medium-low-temperature heat for 7 min. After blocking with a 5% BSA (Sigma Aldrich, B2064, USA) solution for 1 h, we incubated the sections with primary antibody, anti-VEGF (a marker of angiogenesis, mainly in vascular endothelial cells; 1:1000, AbCam, ab46154, UK), anti-doublecortin (a marker of neurogenesis; 1:1000, AbCam, ab18723, UK), or anti-Ki67 (a non-specific proliferation of cells; 1:1000, AbCam, ab15580, UK) antibody for 2 h at room temperature. After washes, the sections were incubated with secondary antibody, goat anti-rabbit IgG (Cy3; 1:5000, AbCam, ab6939, UK) or goat anti-rabbit IgG (FITC; 1:5000, AbCam, ab6717, UK) for 1 h, and then DAPI (Sigma Aldrich, D9542, USA) at room temperature for 5 min. Finally, the sections were sealed with the anti-fluorescence quenching sealing agent Fluoroshield TM (Sigma Aldrich, F6182, USA). The images were collected with a NIKON Fluorescent Microscope (NIKON, ECLIPSE C1, Japan) at excitation wavelength 330-380 nm and emission wavelength 420 nm (blue); FITC excitation 465-495 nm (green) and emission 515-555 nm; and CY3 excitation 510-560 nm and emission 590 nm (red). The images were analyzed using Image-pro plus 6.0 (Media Cybernetics Inc., Rockville, MD, USA). Five areas were randomly chosen to analyze in each image. The signal densities of VEGF, DCX and Ki67 were normalized by the density of the signals of nuclei.

Statistical Analyses
Continuous data are presented as the mean ± standard deviation (SD). For the survival curve, Log-Rank Test (Kaplan-Meier) was used. For the repeated measurement data, Mauchly's test The total profile of the animals which were subjected to the experiment. A total of 144 adult male Wistar rats were used in this experiment. The rats were randomly divided into five groups: Sham, Hypoxic-ischemic encephalopathy (HIE), Human urinary kallidinogenase (HUK), Mild hypothermia (MH), and Combination group (n = 24). Before treatment, 24 rats died after HIE was induced, and eight died after treatment, two in HUK group, two in MH group and four in HIE group. The 7-day survival curve and neurological function assessment (neurological severity scores, NSS) were performed within 7 days (n = 12). Rats were euthanized and detected at the 48 h after HIE (n = 3 for every detection). (B) The division, treatments and time sequence of the experiment. The rats were randomly divided into five groups: Sham, HIE, HUK, MH and HUK + MH (combination) group. Rats in Sham group accepted sham operation and no treatments. Rats in HIE received operation with no treatment; rats in HUK were injected with 0.0016 PNAU/100 g HUK via the tail vein twice a day in successive 2 days, starting from 1 h after inducing HIE; rats in the MH group were treated with MH (34 • C) for 4.5 h starting at 1 h after inducing HIE; and rats in the combined treatment group (HUK+MH) accepted HUK and MH treatment as prescribed. The 7-day survival curve and neurological function assessments (at day 0, 1, 3, 7) were performed. Terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling (TUNEL), Nissl, immunofluorescent staining and Western Blot were performed at the 48-h time point after HIE.
of sphericity was used to determine whether they fit the spherical distribution. When the data did not fit the spherical distribution, the MANOVA test was made first, followed by LSD test being used to compare multiple groups. For the two-factor design, two-way ANOVA was used to analyze the interaction of two factors, and the main or single effect of each factor was analyzed by a General Linear Model. For multiple comparisons, analyses were conducted by two-way ANOVA followed by Tukey's test. The significance level was set at P < 0.05, two-tailed. All statistical analyses were performed using IBM SPSS Version 20 (SPSS Statistics V24, IBM Corporation, Armonk, NY, USA).

The Combination of HUK and MH Improved the Survival Curve and Increased the Neurological Function
The survival curve of HIE group was the lowest in the survival curve chart, 4 out of 12 rats died, and three were at day 1 and one at day 4 (Figure 2A). HUK or MH ended with 2 deaths among 12 rats but had no difference of survival with the HIE group (Log-Rank Test (Kaplan-Meier), χ 2 = 0.848, P = 0.357, Figure 2A). The best survival (0 out of 12 died, Figure 2A) was noted when HUK was combined with MH, which was the same as the Sham FIGURE 2 | Survival curves and the analysis of NSS. (A) 7-day survival curves: the 7-day survival curves were created according to the deaths in each group at different time points (n = 12). The survival curve of HIE group was the lowest in the survival curve chart, 4 out of 12 died, three were at day 1 and one at day 4. HUK or MH ended with two deaths of 12. The best survival curve (0 out of 12 died) was noted when HUK was combined with MH, which was the same as the Sham Group (0 out of 12 died). The survival curves of HUK and MH groups have no difference with that of the HIE group (Log-Rank Test (Kaplan-Meier), χ 2 = 0.848, P = 0.357), but the survival curve of combination is better than that of the HIE group (Log-Rank Test (Kaplan-Meier), χ 2 = 4.615, * P = 0.032). (B) The analysis of NSS scores: Mauchly's test of sphericity shows the NSS scores does not fit the spherical distribution (Mauchly's test of sphericity, χ 2 = 85.722, * * * P < 0.001). There is an interaction between the different treatments and the time points (MANOVA, F = 6.990, * * * P < 0.001). (C) NSS analyses in groups: among the Sham, HIE, HUK, MH and combination of HUK and MH groups (n = 12), NSS assessments were carried out at day 0, day 1, day 3 and day 7. The bar shows the NSS score of every group. The NSS scores are statistically different at different time points (MANOVA, F = 61.585, * * * P < 0.001). At day 0, there is no statistical difference among different treatments (MANOVA, F day 0 = 0.006, P day 0 = 0.999). At day 1 and day 3, only combination, not HUK or MH alone reduced the NSS scores after HIE (MANOVA, LSD test, * P day 1: Combi-HIE = 0.013, * P day 3: Combi-HIE = 0.049). At day 7, the combination reduces the NSS scores compared to HIE (MANOVA, LSD test, * * P Combi-HIE = 0.002), and better than HUK or MH (MANOVA, LSD test, * P Combi-HUK = 0.043, * P Combi-MH = 0.026). But neither HUK nor MH were different to the HIE group (MANOVA, LSD test, P HUK-HIE = 0.233, P MH-HIE = 0.331). * P < 0.05, * * P < 0.01, * * * P < 0.001.

HUK Upregulated VEGF-Labeled Cells, Tight Junctions and BDKRB1/B2, and the Combination of HUK and MH Enhanced the Effects of HUK or MH Alone
Since HUK or combined treatment can prevent neural apoptosis, we investigated if HUK, MH or the combination can promote angiogenesis since vascular endothelial cells constitute of the critical part of BBB construction. The neuroprotective effect of HUK has been shown to be correlated with the BDKRB1 and the BDKRB2 (Han et al., 2015). We also examined if the protein levels of BDKRB1 and BDKRB2 were altered in the rat right hippocampus. Results showed that HUK had a synergistic effect with MH on VEGF-labeled cells (twoway ANOVA, F = 158.140, P < 0.001, Figure 4B). The VEGF-labeled cells were detected by anti-VEGF antibody immunofluorescent staining (VEGF, a marker of angiogenesis, mainly expressed in vascular endothelial cells, Beck and Plate, 2009). The ratios of the signal of VEGF-positive cell to cell nuclei (DAPI staining) increased in the HUK (0.063 ± 0.002) and MH groups (0.048 ± 0.004) compared to the HIE group (0.022 ± 0.002; two-way ANOVA, Tukey's test, * * * P HUK-HIE < 0.001, * * * P MH-HIE < 0.001, Figures 4A,C). In the combined treatment group (0.098 ± 0.007), the ratio was higher than the other treatment groups (two-way ANOVA, Tukey's test, * * * P Combi-HUK < 0.001, * * * P Combi-MH < 0.001, Figures 4A,C).

DISCUSSION
This study shows that HUK may be a potential neuroprotectant, functioning to induce angiogenesis and promote neuronal regeneration in an adult HIE rodent model. In this study, we investigated the effects of MH and HUK in the HIE model and noted four principal findings. First, we observed that the combination of HUK and MH rescued all adult HIE rats from death and profoundly improved neurological dysfunction at days 7. Second, MH significantly attenuated the tight-junction disruptions and increased the expressions of VEGF, BDKR B1/2, cyclin D1, Ki67 and doublecortin (DCX). Third, HUK enhanced these neuroprotective effects of MH. Lastly, the combination of MH and HUK may have their angiogenesis and neuronal regeneration effects at least partially via upregulating the expression of BDKR B1/2, and doublecortin (DCX). To our knowledge, this is the first study to explore the neuropathogenesis of combined MH and HUK therapy in the HIE model.
Although several lines of evidence demonstrate that HUK and MH may play important roles against ischemic stroke (Chen et al., 2010;Gao et al., 2014;Han et al., 2015), the efficacy of either single or combined treatment in HIE still needs to be explored and the neuro-pathogeneses involved in the neuroprotection need to be confirmed. From day 1 to 7, although neither HUK nor MH attenuated the neurological deficit, their combination profoundly ameliorated behavioral impairment, strongly suggesting that HUK enhanced the beneficial effect of MH. The neurological behavioral improvement continued for all observations on days 1, 3 and 7 after HIE, indicating that a short period of administration of exogenous HUK can persistently improve neurological dysfunction over a relatively long time. This result is similar to Han's study showing that HUK could improve neurological dysfunction in ischemic stroke rats (Han et al., 2015). We found that HUK or MH prevented neurons from undergoing necrosis and apoptosis, while this combination led to a more pronounced neuronal survival, indicating that this combination of HUK and MH could provide more effective neuroprotection and this combination enhanced the neuroprotective effects of either alone.
It has been well documented that ZO-1 (Tight junction protein-1) and claudin-1, a group of integral membrane proteins, are closely associated with BBB integrity, representing one mode of cell-to-cell adhesion in endothelial cell sheets and responsible for the barrier function of brain vessels (Luissint et al., 2012;Ren et al., 2015;Li et al., 2016;Zanotto et al., 2017;Bhargavan and Kanmogne, 2018). Dysfunction of BBB mechanisms is well established after ischemia (Fernández-López et al., 2012;Oakley and Tharakan, 2014;Liu et al., 2017); however, we know less about how tight junctions respond to HIE insult and if HUK/MH influences the tight junctions under HIE stimulus. Our study showed that the expression of ZO-1 and claudin-1 in the cortex were significantly downregulated on day 2 following HIE insult and that MH attenuated this downregulation. This finding is consistent with previous studies indicating a similar BBB disruption with changes in tight-junction proteins in neonatal HIE and ischemic stroke (Ek et al., 2015;Shi et al., 2017). Our result strongly indicated that HIE insult may significantly disrupt the integrity of the BBB, while HUK or MH prevent the injury of endothelial cells and decrease the loss of tight junctions. In addition, the combination of HUK and MH exerted a much larger benefit on tight junction expression in HIE rats, and this attenuation of the destruction of tight junctions may contribute to the protective effects of HUK and MH on BBB integrity. These findings (Figures 4A-C,F,G) suggest that the synergistic neuroprotective effects of HUK and MH on BBB may at least partially result from endothelial cell stabilization and decreases in the loss of tight junctions.
Accumulated evidence has shown that HUK may be associated with angiogenesis in ischemic stroke (Ek et al., 2015). We propose that in HIE rats, HUK may exert neuroprotection partially via modulating angiogenesis and regeneration. To investigate this hypothesis, we detected VEGF, BDKR B1/B2, Doublecortin (DCX), Cyclin D and Antigen Ki-67 (Ki-67) along the hippocampus by immunohistochemistry and western blotting.
In the current study, the increased VEGF at day 2 following HIE along the hippocampus suggests that angiogenesis may occur after HIE onset. This result is similar to previous studies (Schoch et al., 2002;Han et al., 2015;Shi et al., 2017), indicating that hypoxic insult may lead to a significant increase in the levels of VEGF in mouse brain that correlated with the hypoxic stimulus. This upregulated VEGF expression induced by HIE in the brain may play both neurotoxic and neuroprotective roles: (1) leading to vasogenic edema and the leakage of blood-borne substances into the brain parenchyma and increased BBB permeability; and (2) stimulating and sustaining regeneration and neovascularization. The neuroprotective effect of upregulated VEGF occurs from several hours to several weeks following hypoxic-ischemia (Sun and Guo, 2005;Kaur and Ling, 2009;Baburamani et al., 2012). Since VEGF has been well documented as playing a role in angiogenesis (Doeppner et al., 2017;Zhao et al., 2017;Zou et al., 2017), our finding demonstrates that either HUK or MH promotes angiogenesis, but the combination gave the highest value. In addition, we also noted the downregulated expression of the BDKRB1/2 after HIE insult, while HUK, MH and the combination restored the downregulation of BDKRB1/2. It has been documented that the activation of BDKRB1/2 is involved in the angiogenesis process (Spinetti et al., 2011;Mousa et al., 2014). Our observation that HUK mediated angiogenesis following HIE insult is similar to previous studies (Han et al., 2015;Li et al., 2015). Xu's group found that in ischemic experimental models and stroke patients, HUK promoted post-ischemic angiogenesis and cerebral perfusion via increasing vessel density, enhancing VEGF and apelin/APJ expression, and inducing ERK1/2 phosphorylation (Han et al., 2015;Li et al., 2015). Taken together, these data strongly indicate that the BDKRB1/B2 receptors were involved in the angiogenetic effect of HUK in HIE rats, and HUK may promote angiogenesis in HIE stimulus via upregulating the expression of VEGF and BDKRB1/2.
To further explore whether the combined effects of HUK and MH on the HIE insult were associated with neurogenesis or regeneration of all kinds of cells, DCX, which is an important predictor of neurogenesis, and cyclin D1 and Ki-67, which are predictors of non-specific regeneration, were evaluated. Cyclin D1 is a member of the cyclin protein family, plays essential roles in regulating cell cycle progression and proliferation and is usually recognized as a biomarker of regeneration of cells (Bouchard-Cannon et al., 2013;Cheyuo et al., 2015). Ki67 labels the G and S phase cells and highly expresses in proliferative cells (Yang et al., 2018). In the current study, our observation that cyclin D1 and Ki67 decreased with adult HIE insult implies that hypoxic-ischemic cells undergo death and regeneration is markedly reduced. Impressively, either MH or HUK treatment significantly reversed the HIE-induced downregulation of cyclin D1 and Ki67, strongly indicating that MH (Yenari and Han, 2013;Wood et al., 2016) or HUK following HIE onset could enhance regeneration by positively influencing regeneration of neurons or other cell. The mechanisms of MH or HUK enhancing regenerative properties in the current study could be explained, in part, by its effect on reducing apoptosis (Figure 3) and enhancing angiogenesis in a crucial time window, as indicated by the upregulation of VEGF. This MH-induced neurogenesis in the current work is consistent with Silasi and Colbourne's (2011) study, indicating that hypothermia promoted CA neurogenesis and cell regeneration in global ischemia. To further identify the effects of MH and HUK on neurogenesis, we examined the expression of doublecortin (DCX), a candidate marker for adult neurogenesis, by immunohistochemical staining in the hippocampus, and observed a change similar to cyclin D1 (Figures 5D,E). More interestingly, our study showed that the combination of MH and HUK led to more pronounced increase in the expression of DCX when compared to MH or HUK alone, indicating greater efficacies of neurogenesis. To our knowledge, this is the first study showing that HUK exerts neurogenesis functions by regulating DCX in adult HIE insult.
In addition to mediating angiogenesis and promoting proliferation, other mechanisms, like inhibiting vasodilation, suppressing inflammation and activating the innate immune response, are found to be involved in HUK's effects against ischemia (Ponticelli and Meroni, 2009;Chen et al., 2010;Sharma and Narayanan, 2014). In addition, Chen et al. (2010) found that HUK not only suppresses cerebral inflammation and downregulates NF-κB, but also activates the MAPK/ERK pathway in an experimental ischemic model. Whether those molecular mechanisms play important roles in HIE insult needs to be further explored.
In summary, for the first time, this study demonstrates the beneficial effects of a combination therapy of MH and HUK for treating adult HIE rats. The combination of MH and HUK was associated with improvement in functional recovery, combined cellular neuroprotection, attenuation of tight junction reduction, promotion of angiogenesis and regeneration in HIE rats. This finding is clinically meaningful because it strongly implies that an early combination of MH and HUK may correlate with improved outcomes and reduced mortality in patients with HIE. These data support the strategy of combining MH with pharmacological HUK to increase the clinical feasibility, efficacy and safety of the treatment of HIE patients. The combined treatment could represent a promising approach to create more effective and safer hypothermia therapies for patients with brain injuries.