Quantitative Proteomics Analysis to Study the Protective Effects of Nerve Growth Factor on Hippocampal Mitochondria in VD Rats


 Background

Vascular dementia (VD) is a kind of clinical syndrome characterized with the impairment cognitive function caused by cerebrovascular disease. Genetics, biochemical, and morphological analyses of cell and animal models, reveal that mitochondria could have roles in this neurodegeneration.
Methods

We used Sprague-Dawley rats to establish VD model, and used the proteomics method based on relative quantification (iTRAQ) to identify the differentially expressed proteins in hippocampus mitochondria.
Results

A total of 33 differentially expressed proteins were identified between the VD rats and the VD rats treated with nerve growth factor groups. And five differentially expressed proteins (Rgs14, Slc7a14, Ppm1l, Kcnj10 and Syngr1) were identified after completing the sham-operate control, VD rats and VD rats treated with nerve growth factor groups, then successfully confirmed by western blot. Bioinformatics analysis suggested that the mitochondrial molecular mechanism of VD and the protective effect of nerve growth factor on mitochondrial function of VD rats may be due to different molecular mechanisms.
Conclusion

We estimated that mitochondrial dysfunction may be the onset of VD and key role in the pathological process of VD. This study not only has a deeper understanding of the mitochondrial molecular mechanism of VD, but also is helpful for the screening of drug targets.


Abstract
Background Vascular dementia (VD) is a kind of clinical syndrome characterized with the impairment cognitive function caused by cerebrovascular disease. Genetics, biochemical, and morphological analyses of cell and animal models, reveal that mitochondria could have roles in this neurodegeneration.

Methods
We used Sprague-Dawley rats to establish VD model, and used the proteomics method based on relative quanti cation (iTRAQ) to identify the differentially expressed proteins in hippocampus mitochondria.

Results
A total of 33 differentially expressed proteins were identi ed between the VD rats and the VD rats treated with nerve growth factor groups. And ve differentially expressed proteins (Rgs14, Slc7a14, Ppm1l, Kcnj10 and Syngr1) were identi ed after completing the sham-operate control, VD rats and VD rats treated with nerve growth factor groups, then successfully con rmed by western blot. Bioinformatics analysis suggested that the mitochondrial molecular mechanism of VD and the protective effect of nerve growth factor on mitochondrial function of VD rats may be due to different molecular mechanisms.

Conclusion
We estimated that mitochondrial dysfunction may be the onset of VD and key role in the pathological process of VD. This study not only has a deeper understanding of the mitochondrial molecular mechanism of VD, but also is helpful for the screening of drug targets. Background Vascular dementia (VD) is a kind of clinical syndrome characterized by cognitive impairment caused by cerebrovascular factors. At present, the objective reality of the medical profession is that with the progress of medical conditions and technology, the mortality rate of cerebrovascular diseases is decreasing year by year. However, the incidence of cognitive dysfunction due to cerebrovascular disease is gradually increasing. Therefore, it is of great signi cance to clarify the molecular mechanism of VD and provide reliable basis for the prevention and treatment of VD.
The pathological changes of VD involve the oxidative stress, mitochondrial lesions, abnormal genomic activity, synaptic dysfunction, protein metabolism disorders [1]. At present, the mechanisms about VD involve the cerebral ischemia and contradiction between supply and demand of oxygen metabolism exist in neurons. The cerebral ischemia and hypoxia lead to oxidative stress and increase of free radicals in brain cells, which lead to mitochondrial dysfunction [2,3]. Our previous studies also revealed abnormal expression of mitochondrial processing peptidase and mitochondrial precursor protein in the hippocampus of VD rats, suggesting that mitochondrial dysfunction functioned in the occurrence of neurodegenerative diseases [4].
It is well known that mitochondria are important organelles of human body. The position of mitochondria in nervous system is particularly important [5]. Neurons consume a lot of energy when they are active.
When the oxidative phosphorylation process of mitochondria is impeded, neurons cannot start glycolysis as other types of cells to compensate [6]. In addition, the dynamic process of mitochondrial division and fusion is known as the differentiation and redistribution in synapses [7]. Mitochondrial dynamics and mitochon-drial transporting mechanisms ensure the long-distance transport and energy average distribution of mitochondria in neuronal endings. Abnormal mitochondrial fusion and division hinder the transport of mitochondria in neurons, resulting in the disorder of mitochondrial distribution and the decrease of energy metabolism e ciency [8]. Finally, through feedback, the cell signaling pathway is disturbed, which further leads to the degeneration of synaptic function and neuron death. Recent studies have also shown that mitochondrial dysfunction and fusion disorder can lead to degenerative diseases of the nervous system [9]. Therefore, abnormal mitochondrial fusion and division may be one of the key factors of dysfunction in mitochondria.
The nerve growth factor (NGF) can maintain the growth, development and reproduction of neurons [10].
NGF can promote neurite outgrowth in normal and ischemic brain regions and then slows down and prevents nerve cell apoptosis [11]. The experimental study of cerebral ischemia also shows that the combination of NGF and receptor can activate the injured nerve cells, help the recovery of the injured brain cells, regulate the metabolic function of nerve cells, improve the living environment of brain nerve cells, and effectively inhibit the death of brain cells [12]. Another proteomics study con rmed the different proteome in the VD cortex, using the iTRAQ-2D-LC-MS/MS strategy [13].
In our study, we used the samples from hippocampal brain area of VD rats with or without NGF intervention, to detect and identify the mitochondrial proteins with altered expression levels. The bioinformatics tools David and STRING were used to analyze dysfunctional proteins and determine if functional relationships would occur. This method can not only identify the protein sequence, but also quantify the related protein expression changes, which is helpful to quantify the potential biomarkers and better understand the pathology at the molecular level. The proteome reported in this paper also provided a reference data set for the basic research or translation research of VD in the future.

Animals and Ethics Statement
Adult male Sprague-Dawley rats (220±20g; No. 00010751; Experimental Animal Center, Wuhan First Hospital, Wuhan, Hubei Province, China; Laboratory animal facility license number: 00014834) were kept in the laboratory with temperature of 22±2 ℃ and humidity of 65±2%, 4 rats per cage. Animals were fed in a 12 h light/dark cycle and were given food and water at will. All operations were performed in accordance with the guidelines approved by the Animal Care Committee of Inner Mongolia People's Hospital. We made every effort to relieve pain and discomfort.

Surgery procedure
The animals were weighed accurately and anesthetized by intraperitoneal injection of 10% chloral hydrate according to 0.3ml/100g. After intraperitoneal injection of chloral hydrate for 3-5 minutes, the righting re ex disappeared in animals. The animals were xed on the operating table in supine position to ensure smooth breathing and disappeared with normal neck skin. The skin and subcutaneous tissue were cut through the middle of the neck, then the digastric muscle and sternocleidomastoid muscle were separated bluntly by tweezers. The carotid sheath was exposed. The bilateral common carotid artery and vagus nerve were separated carefully by glass needle, and the operation line 0 was inserted under the common carotid artery for reserve. Bilateral silk thread was drawn and bilateral common carotid artery of rats was clamped with non-invasive arterioles to block blood ow for 10 minutes, then removed and restored perfusion for 10 minutes for twice. Finally, the skin was sutured and the surgical site was disinfected. The gentamicin sulfate (0.2 ml) was injected intraperitoneally for anti-in ammation.
Preparation and detection of pathological specimens Fresh brain tissue was xed in 4 % paraformaldehyde and sectioned routinely. Mitochondria in hippocampus were labeled by Mito-Tracker GreenFM490/560, and morphological structure of mitochondria was observed by confocal laser microscopy.

Mitochondrial preparation
The isolated hippocampus was homogenated with glass homogenizer in the medium with ice residue. The mitochondria were extracted by differential centrifugation and gradient density centrifugation 10 times. The supernatant was centrifuged at 2000g for 3 minutes and repeated once; the supernatant was centrifuged at 12500g for 8 minutes and the supernatant was discarded; the precipitate was dissolved in 0.8ml 3% Ficoll medium and carefully spread on 3.2ml 6% Ficoll medium and centrifuged at 11050g for 30 minutes and the supernatant was discarded; the precipitate was suspended in the separation medium and centrifuged at 12500g for 8 minutes; nally, the precipitate was suspended with an appropriate amount of separation medium and the protein concentration was adjusted to a certain level About 10-20mg / ml. The protein concentration was determined by Lowry ` s method with bovine serum albumin as the standard protein. Fresh mitochondrial suspension was taken to measure respiratory activity. All the above operations were completed at 0-4 ℃.

Hippocampal neuron staining
The hippocampal neuron of the sham-operate control, VD and VD rats treated with NGF groups were taken for 5 mm and xed with 10 % neutral buffer at 4 ℃, then embedded with conventional wax blocks. Slice thickness was 5um. The slices were immersed in dimethylbenzene I, II and III of 90%, 80% and 70% concentration for 5 minutes. The slices were then immersed in 5g Nissl solution and soaked at 37 ℃ for 20 minutes. After incubation, the slices were washed with distilled water and then separated with 95% ethanol. Microscopically, there are obvious hippocampal neurons. The slices were dehydrated with anhydrous ethanol, transparent with xylene, and xed with neutral rubber. Image analysis was carried out by using Image-Pro plus 4.5 software of Silver Spring Company in the United States to determine the cumulative integral optical density (IOD sum) and area value (area sum) of the positive product and calculate the average density (IOD / area) [14].

Extraction of total protein from mitochondrial cells in hippocampus
Total mitochondrial cell protein extraction and TMT quantitative proteomic experiment in hippocampus were commissioned by Majorbio Company in Shanghai. The instrument used was Triple TOF 5600 (AB Sciex, USA), and the analysis software was Proteinpilot (Applied Biosystems MDS Sciex, USA), Q-Exactive, EASY-nLC 1200, Orbitrap Fusion Tribrid (Thermo Fisher, USA) and Proteome Discoverer (Thermo Fisher, USA).

Total mitochondrial cell protein extraction
Samples were taken out under freezing condition and frozen on ice. The samples were added 10% SDS (Final concentration 1 %) and protease inhibitor PMSF (7.5 mM) to lyses for 30 minutes, and mixed every 5 minutes for 10 seconds. Protein supernatant was centrifuged at 12000g for 8 mins. BCA quantitative analysis was used, and then detected by SDS-PAGE. Protein was measured using the BCA method.

TMT Sample Preparation
Take 100 mg of above protein sample and supplement the volume with lysate to 100 µl. The nal concentration of 10 mM TCEP was added and the reaction time was 60 min at 37 ℃. Add the iodoacetamide (40 mM) at room temperature for 40 minutes. Pre-cooled acetone (6:1 w/w) was added to each tube, and precipitated at -20℃ for 4 hours. The Protein samples were centrifuged at 10000g for 20 mins and then the precipitations were collected. The precipitations were dissolved with 100 µl TEAB (100 mM). Finally, the trypsin solution (1:50 w/w) was added for enzymatic hydrolysis at 37 ℃ overnight.

High pH Reverse Phase Separation and Nano-LC−MS/MS Analysis
The Sep-Pack columns were used for desalination of TMT-labeled peptide mixtures. The high-pH reversed-phase column peptide fractionation kit (Thermo Fisher, USA) was used to classify the peptide fractions. A total of 10 distillates were charged according to the time. The distillates were concentrated by vacuum centrifugation and dissolved in a mass spectrometry sample buffer for the second dimensional analysis. Mass spectrometry parameters were set as follows: ion spray voltage of the inlet, 2.3 kV; mass spectra range, 350-1300 m/z; accumulation time per spectrum, 100 ms; and dynamic exclusion time, 18 s [15].

Proteomic Data Analysis and Bioinformatics
Proteomic data analysis and bioinformatics methods were described previously [15]. The Proteome Discoverer Software 2.1 (Thermo Fisher, USA), Protein pilot software (Applied Biosystems MDS Sciex) were used. All reported data were based on 95% con dence interval for protein and peptide identi cation as determined by Protein Pilot (Prot Score ≥1.3) [15]. Meanwhile, a cut off of 1% FDR (false discovery rate) and 40% isolation interference for peptide spectrum matches (PSMs) were required for all reported proteins. The Gene Ontology (GO) annotation was using the database (http://www.uniprot.org/uniprot/? query=organism:10116).

Veri cation of Differentially Expressed Proteins Using Western Blotting
Western blot was used to verify the quantitative proteomic results of TMT. The test method is described previously [16]. All the antibodies were diluted with 1:1000, and the secondary antibodies were diluted with 1:10000. The proteins were quanti ed by Image-J software. Each Western blot strip was measured three times and the average optical density was calculated.

Statistical Analysis
The data were expressed as mean ± SD. A Student' s t test in R language was used to calculate the P value of signi cant difference between samples. In this project, the screening criteria for signi cant difference expression proteins were: P < 0.05 & (FC<0.83 or FC>1.20).

Identi cation of hippocampal neurons by Nissl staining
Under the light microscope, neurons stained blue could be seen. Rats in the sham-operate control group on Nissl staining section, the number of neurons was more and the color was dark. The dark blue is visible Nissl body, and the light blue was indicated the nucleus ( Figure 1A). But in the same area in the VD group, the number of neurons in dentate gyrus decreased and the staining was shallow. Nissl bodies were decreased or dissolved in cytoplasm, dendrites swelled and broken ( Figure 1A). When the VD rats were treated with the nerve growth factor, the neurons started restocking and the color became more dark ( Figure 1A). Furthermore, the mean density (IOD/AREA) was calculated and the results were consistent with the observation results ( Figure 1B). The number of hippocampal neurons decreased, the Nissl body decreased or even disappeared, and the optical density decreased signi cantly. Because the light density was directly proportional to the color depth, it indicated that the neurons in this area become light stained. And, the color of neurons was determined by the number of Nissl bodies. In the absence of this substance in the cytoplasm, the light density decreased and the color became lighter.

TMT coupled with Nano-LC−MS/MS analysis of hippocampal mitochondria
Three hippocampal mitochondria were included in each group, and TMT proteomic analysis was carried out to identify different proteins that may be involved in the protection of NGF. The protein digestibility of the three samples was higher than 99%. The labeling e ciency was 95.9%. Through comparative analysis, 1572 proteins were identi ed as differentially expressed in the VD group vs sham-operate control group (Figure 2A, Table 1) and 1434 proteins were identi ed as differentially expressed in the VD rats treated with NGF group vs sham-operate control group ( Figure 2B, Table 1). Similarly, the comparison between VD group and VD rats treated with NGF group showed that 17 proteins were up-regulated and 16 proteins were down regulated ( Figure 2C, Table 1). And the GO annotation analysis showed that these proteins were major participants in cellular process, growth, response to stimulus, membrane part, metabolic process, multicellular organismal process and catalytic activity ( Figure S1-S3). GO annotation analysis was used to classify these differentially expressed proteins according to cell composition, molecular function and biological process (Figure 3 A−C). As an example, the MS/MS spectrum and relative reported ion strength of abnormal expression protein A0A0G2K1G8 (slc7a14) were shown in Figure 4.
Analysis of differentially expressed protein by KEGG shown that the differentially expressed proteins were involved the following pathways: Endocytosis, Thermogenesis, Pathways in cancer, Huntington's disease, Protein processing in endocannabinoid signaling, Parkinson's disease and so on ( Figure 5). After comparing the results Table 1, only ve differentially expressed proteins were got including Rgs14, Slc7a14, Ppm1l, Kcnj10 and Syngr1 (Figure 6).

Western Blot Validation of Quantitative Proteomic Analysis
The next stage was to focus on the above ve proteins, whose expressions were changed in three different comparisons (VD vs SC, VD vs VDN and VDN vs SC). From Figures 7A and 7B, we could see that the relative expression levels of Rgs14 was increased 2.1 fold in VD rats but signi cantly decreased in VD rats treated with NGF groups (1.5 fold) compared with the sham-operate control group. These results were consistent with the results based on iTRAQ ( Figure 7C). In addition, the same comparison method was used to observe the relative expression levels of Ppm1l and Kcnj10, and we found that the levels of these two proteins were reduced in VD rats but raised in VD rats treated with NGF groups compared with the sham-operate control group ( Figure 7A and B). This part of the results could not correspond to the results based on iTRAQ ( Figure 7C). The remaining two proteins Slc7a14 and Syngr1 showed the same trend in the expression levels, which were increased both in VD rats and VD rats treated with NGF groups compared with the sham-operate control group ( Figure 7A and B). At the same time, it was consistent with the iTRAQ results.

Discussion
In this study, we have extracted total mitochondria in hippocampus and detected by technology iTRAQ and veri ed by western blot. Speci cally, we identi ed several differentially expressed proteins of mitochondria in hippocampus neuros, to our knowledge this is the rst report of the difference expression protein in mitochondria in hippocampus. We also observed the morphological changes of neurons. The data of proteomic study from Veen map revealed that the expression of ve proteins have changed signi cantly, including regulator of G protein signaling 14 (Rgs14), solute carrier family 7 member 14 (Slc7a14) protein phosphatase Mg 2+ /Mn 2+ -dependent, IL (Ppm1l), ATP-sensitive inward recti er potassium channel 10 (Kcnj10) and Synaptogyrin-1 (Syngr1). We also veri ed these proteins by western blot.
Speci cally, the expression of Rgs14 present increased dramatically after interfered by NGF compared with control group, revealing Rgs14 has the striking regulation to the function of hippocampus mitochondria. Rgs14 is a kind of 60-kDa protein belonging to the RGS 12 family (R12) which is mainly expressed in the area CA2 of hippocampus pyramidal neurons and enriched post-synaptically in dendrites and spines in central nervous system [17][18][19]. Accumulating evidence has revealed that Rgs14 may negatively regulates the capacity of learning and memory in the CA2 region of the hippocampus by integrating G protein and ERK/MAPK signaling. [20][21][22][23][24] Furthermore, some researchers demonstrated Rgs14 knockout (KO) rats possess an unusually robust capacity for long-term potentiation (LTP) in CA2, which is absent in wild-type (WT) rats, and markedly enhanced spatial learning [19]. Taken together, these results show that Rgs14 acts as a critical factor for hippocampal based learning and memory and it belongs to a negative regulator. We have previously demonstrated Bilateral common carotid artery ligation (2-VO) could bring about the decline of capacity in spatial learning via Morris water maze on rats. And our current WB data showed the relative expression levels of Rgs14 VD rats and VD rats treated with NGF groups respectively are 2.1 and 1.5 compared with sham-operate control group, suggesting that NGF plays an important neuro-protection role. We assumed that the levels of Rgs14 might be speci c increasing in VD rats, but partial recovery treated by NGF. This nding provides interesting clues towards the molecular mechanism about mitochondrial dysfunction simultaneously. We also inferred that Rgs14 proteins may be a kind of novel potential drug targets to dementia.
Similar to this protein are Ppm1l and Kcnj10. It was shown that Ppm1l inhibited in ammatory responses by directly dephosphorylating IKKβ and suppressing NF-κB activation in cardiac macrophages, subsequently prevented excessive scar formation and improved cardiac function [20]. Moreover, it was reported that Ppm1l was highly expressed in the central nervous system during mouse development and played an important role in brain development [21]. To date, Ppm1l should be discovered and reported by proteomics techniques in mitochondrial hippocampus for the rst time. Our ndings established that Ppm1l expression of the mitochondrial hippocampus in VD was decreased signi cantly comparing with the sham-operate control group, whereas recovered after intervention by NGF (Fig. 7). We speculated Ppm1l was associated with the mitochondrial endoplasmic reticulum stress and displayed a lower expression in VD rats. But, the NGF could elevate the level of Ppm1l by promoting the growth, development, differentiation and maturity of central and peripheral neurons, maintaining the normal function of the nervous system, and accelerating the repair of the injured nervous system. Similarly, the NGF might elevate the level of Kcnj10 to protect the injured nervous system and there were few reports on the relationship between Kcnj10 and mitochondrial hippocampus.
With regard to Syngr1 and Slc7a14, the levels were found increased both in VD rats and VD rats treated with NGF groups. These results might indicate that Syngr1 and Slc7a14 played an important role in pathogenesis of vascular dementia, but the nerve growth factor did not affect the levels of these two proteins.

Conclusion
Proteomics analysis of the relative expression of mitochondrial proteins in hippocampus of Vascular dementia rats could nd potential biological targets. Rgs14, Ppm1l and Kcnj10 might be good potential biological targets for drugs treating Vascular dementia. Nerve growth factor might be able to treat Vascular dementia.

Declarations
Ethics approval and consent to participate This study was approved by the Animal Care Committee of Inner Mongolia People's Hospital. Consent for publish I give my consent for publication on behalf of all authors involved.

Availability of data and materials
All data, models, and code generated or used during the study appear in the submitted article.

Competing interests
The authors declare no competing interests in regards to this project.   It can be seen that the point on the left is the protein with down-regulated differential expression, and the point on the right is the protein with up-regulated differential expression.
The more points on the left, right and top are, the more signi cant the difference is.   Top 20 pathways with the largest number of proteins. From left to right, it is arranged according to the number of protein contained from high to low. The higher the column is, the more active the biological pathway is in the measured sample.

Figure 6
Differential proteins Veen map.

Figure 7
Western blot validation of Rgs14, Slc7a14, Ppm1l, Kcnj10 and Syngr1 identi ed by the iTRAQ-based proteomics approach. A and B. β-Actin was a loading control. Experiments were repeated three times. C. Differentially expressed proteins in hippocampal mitochondria in SC, VD and VDN groups Identi ed by iTRAQ based quantitative proteomics approach.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.