The Study on Neural Remodeling of Medullary Visceral Zone in Early Sepsis and Interfered by Cholinergic Anti-Inammatory Pathway

including our have shown that the Medullary Visceral Zone (MVZ) can effectively regulate systemic inammation and immunity through the Cholinergic Anti-inammatory Pathway (CAP). Sepsis usually causes neuroinammation in the Central Nervous System (CNS), which will inevitably affect the structure and function in related brain areas such as MVZ, whether the intervention of CAP can affect the structure and function of the MVZ in sepsis needs to be further veried.


Abstract
Background Studies including our own have shown that the Medullary Visceral Zone (MVZ) can effectively regulate systemic in ammation and immunity through the Cholinergic Anti-in ammatory Pathway (CAP). Sepsis usually causes neuroin ammation in the Central Nervous System (CNS), which will inevitably affect the structure and function in related brain areas such as MVZ, whether the intervention of CAP can affect the structure and function of the MVZ in sepsis needs to be further veri ed.
Methods 64 adults, speci c pathogens free Sprague-Dawley male rats were used in this study. The septic models were prepared by cecum ligation and puncture (CLP) method, GTS-21 (a selective α7 nicotinic acetylcholine receptor agonist which can mimic CAP's activation) and MLA (a powerful and selective nicotine acetylcholine receptor antagonist which can mimic CAP's blocking) were used to interfere CAP.
The pathological changes, apoptosis, the expressions of Tyrosine Hydroxylase (TH) and Choline acetyltransferase (CHAT), the expression levels of GAP-43 mRNA, Olig-2 mRNA, VEGF mRNA, GFAP mRNA, MMP-9 mRNA in MVZ were analyzed among different groups.

Results
In this study, we found that sepsis induced apoptosis and functional suppression of catecholaminergic and cholinergic neurons and gliosis in MVZ, up-regulated key genes' expressions such as GAP-43 mRNA GFAP mRNA VEGF mRNA MMP-9 mRNA, down-regulated the expression of Olig-2 mRNA. GTS-21, a selective α7 nicotinic acetylcholine receptor agonist, obviously mitigated the above changes; whereas, methyllycaconitine (MLA), a powerful and selective nicotine acetylcholine receptor antagonist, signi cantly aggravated these changes.

Conclusions
Our research shows that activating CAP can effectively mitigate the neural remodeling and neuronal suppression induced by early sepsis in MVZ, the mechanism may involve with its control of systemic and local in ammation. This study reveals that MVZ and CAP may be potential targets to curb the in ammatory storm in early sepsis.

Background
The uncontrolled systemic in ammation storm induced by early sepsis is closely related to high mortality [ 1 ]. Anti-in ammatory therapy can evidently improve the mortality and prognosis of septic patients [ 2 ]. Therefore, seeking the body's own in ammation and immune regulatory pathways to suppress the early uncontrolled in ammation storm of sepsis has signi cant theoretical innovation meaning.
The Medullary Visceral Zone (MVZ) is a relaying station to transmit peripheral in ammatory information to the high-level nervous center [ 3 ]. Previous studies have found that MVZ effectively regulates systemic in ammation and immune strength through Cholinergic Anti-in ammatory Pathways (CAP) [ 4 ]. GTS-21 a selective α7 nicotinic acetylcholine receptor agonist can evidently down-regulate the serum levels of in ammatory mediators such as soluble CD14 (Presepsin), High Mobility Group Box-1 (HMGB-1), interleukin (IL)-10, Tumor Necrosis Factor (TNF)-α, IL-1α, IL-6 and CD4+CD25+Treg, CD4+IL-17+TH17 lymphocyte percentage; On the contrary, methyllycaconitine (MLA), a powerful and selective nicotine acetylcholine receptor antagonist, greatly aggravates systemic in ammation and immunity of sepsis. It is well known that both GTS-21 and MLA can pass through the blood-brain barrier and play multiple central roles [ 5 -6 ]. Recent studies also suggested that intervention with the central CAP had a wide range of effects such as improving the impaired cognitive function [ 7 ] and regulating systemic in ammation [ 8 ]; then, a problem emerges, besides peripherally mimicking or antagonizing CAP to in uence systemic in ammation, whether GTS-21 and MLA also affect the structure and function of septic MVZ through which further affect systemic in ammation and immunity is still unclear.
Studies have con rmed that while MVZ regulates systemic immunity and in ammation, systemic in ammation also induces MVZ's neuro-in ammation and even leads to its disorder of autonomic regulation [ 9 - [10][11] ]. In Clinic once patients in Intensive Care Unit were diagnosed with septic encephalopathy, the mortality rate is signi cantly higher than those without encephalopathy [ 12 ]. From these researches we reasonably believe that MVZ is de nite to be attacked by systemic in ammation in the early stage of sepsis, which may underlie the disorder of CAP's regulatory function and the in ammatory storm in the early stage of sepsis. Based on the theory that various constitutions of nerves have different functional states, which may be an important mechanism for individuals survival adaptation and even modifying the social role they played [ 13 ], it is necessary to explore the MVZ pathology in sepsis and the interfered effect by CAP activation or blocking to provide ideas of central intervention to combat septic in ammation storm.

Material And Methods
Rats management and septic models preparation The septic models were prepared by cecum ligation and puncture (CLP) way [ 14 ], the main processes were included: Rats were deeply anesthetized with iso urane inhalation [ 15 -16 ], the right lower abdomen were shaved and disinfected prior to be cut open to abdominal cavity. The cecum were dissociated and ligated at the middle with a 5-0 suture, the ligated pare was pierced with a 21G needle twice and gently squeezed out a small amount of intestinal contents from the puncture holes. At last the cecum was returned into the abdominal cavity, the inner layer was sutured with 5-0 sutures and the outer layer with 3-0 sutures ( Fig  1). After operation rats were accepted intraperitoneal injection of Piperacillin (50 mg/Kg, i.p. tid×3d).
After operation, rats were delayed to recover to their consciousness, their bodies were sluggish and often curled up, they were looked inactive, eat and drink less, they breathed quickly and di cultly. These appearances suggested that the model is successful. After 12 hours of operation, some rats began to die. As time went by, most rats showed low skin temperature, weakened muscle strength, bloody secretions in the eyes, stopped eating and drinking and began to diarrhea. Further development, they appeared shortness of breath, no resistance to passive supine. Some rats excreted large amount of mucus watery stool and eventually died.
Rats grouping and treatment After 7 days of adaptive feeding, rats were divided into three groups according to the random number table: Control Group: Rats (n=8) were feed as usually without any treatment; Sham Group: Rats (n=8) were subjected to open and suture the abdominal cavity without CLP manipulation, afterwards they were accepted intraperitoneal injection of Piperacillin (50 mg/Kg, i.p. tid×3d); Sepsis Group: The sepsis rats (n=48) were prepared with CLP method as mentioned above. One hour after the septic rats return to their consciousness, they were randomly divided into 3 groups, 16 rats in each group. a: Model Group: accepting intraperitoneal injection of Piperacillin (50 mg/Kg, i.p. tid×3d) and saline (1 mL/100 g, i.p. Dosage and duration: 4.8 mg/Kg, i.p. tid×3d) was given to each rat [ 18 ]. After 3 days, the rats were sacri ced to collect the medullary tissue for analysis under anesthesia with iso urane inhalation.

Para n section preparation and pathological observation
Three days after operation, all survival rats in each group were deeply anesthetized with2-4% iso urane inhalation, brain tissues were quickly stripped and xed in 4% paraformaldehyde buffer solution (PBS) for 12 ~ 24 hrs. 3 rats' Medulla Oblongata from each group were cut off and dealt with conventional dehydration, para n embedding to prepare sections (30 μm), which were used for haematoxylin and eosin (HE) staining, TdT mediated dUTP Nick End Labeling (TUNEL) and Immuno uorescence analysis.
The rest Medulla Oblongata were preserved in the refrigerator at -80℃ for PCR detection. TUNEL 5 para n sections from ve groups were taken, they were dewaxed and immerged in Proteinase K solution.then the Terminal Deoxynucleotidyl Transferase (TdT) buffer was added to incubate. In the end, DAPI (Beyotime Biotechnology, lot number: C1002) and uorescence quencher (southernbiotech, lot number: 0100-01) were added to stain the nuclei and the apoptotic cells. Sections were observed with a uorescent microscope (Olympus BX53 biological microscope). The survival cell nuclei were stained blue, the apoptotic cell nuclei were stained red. Select 3 elds of view (the left side, the middle and the right side) to count labeling index (LI) under high magni cation(400 times) on every sections, LI = the number of positive cells in each eld/all cells in the eld, and the apoptosis index (AI) of each group is the average value of LI for each section Immuno uorescence double labeling for TH or CHAT combined with Caspase 3 After dehydration sections were blocked with 10% normal goat serum for 1 hour and then incubated for 24 h in a cocktail of primary antibodies for the labeling of caspase3 (produced by Wuhan Sanyan Bio. Co., China. Lot: 66470-2-IG, dilution: 1:50), Tyrosine Hydroxylase (TH, produced by Wuhan Boster Co., China. Lot: BM4568, dilution: 1:50) or Choline acetyltransferase (CHAT, produced by Wuhan Bioss Co., China. Lot: bs-2423R, dilution: 1:50). Afterwards, sections were incubated with two uorescent-labeled secondary antibodies FITC labeled goat anti-rabbit IgG, used to mark TH or CHAT, produced by Wuhan Boster Co., China. Lot: BA1105, dilution: 1:100, Cy3 labeled goat anti-mice IgG, used to mark caspase3, produced by Wuhan Boster Co., China. Lot: BA1031, dilution: 1:100) for 4 hours. Finally, the sections were mounted on gelatin-coated slides and covered with mounting medium with DAPI, for nuclear staining of all cells present in the slice [ 19 ]. Images from the different experimental groups were captured with Olympus BX53 Biological Microscope. The normal nuclei were stained blue, Cholinergic neurons expressing CHAT and catecholaminergic neurons expressing TH were stained green, apoptotic neurons expressing caspase3 were stained red. Three images with 400 folds' enlargement from every group were analyzed with imagepro (ipp6.0) software [ 20 ] and the average densities were acquired.
Quantitative RT-PCR All the rest rats' Medulla Oblongata specimens from each group were used for the target RNA detection.
RNA extraction accorded to the instruction of Reagant Trizol kit (produced by Aidlab. Co. Lot: 252250AX).
In a gene bank database standard cDNA sequences of related gene were obtained, DNA starting primers were designed, synthesized and supplied by Qingke Co. Ltd. The sequences are as follows (Table1).
Select Rat GADPH as inner reference to compute the expression levels of GAP-43 mRNA, Olig-2 mRNA, VEGF mRNA, GFAP mRNA, MMP-9 mRNA. 0.1g fresh sampling tissue of Medulla Oblongata from every group was taken and homogenized in 1mL Trizo reagent (Aidlab, Lot:252250AX) by bead mill (Retsch Tissue Lyzer II, Qiagen, Valencia CA, USA). RNA was isolated from medulla homogenates by addition of 10% BCP and standard phase separation, followed by overnight precipitation with isopropanol at -20°C. RNA was puri ed using the Qiagen RNeasy Mini kit (Qiagen), Start PCR ampli cation reaction in PCR instrument. The reaction conditions are as follows: Predenaturation: 50℃ for 2min.; denaturation: 95℃ for 10min; annealing: 95℃ for 30sec, extension 60℃ for 30sec, altogether for 40 cycles. Electrophoresis analysis: taking the product of 5ul PCR and 6 x DNA loading buffer 2ul to mix, then take the mixture into 2% agarose gel (containing EB) to electrophoresis at 150V voltage for 35min prior to being observed under ultraviolet lamp, after focus adjustment, the images were photographed with gel analysis system. With gray scale scanning software existing in the gel imaging system the electrophoresis target zone was analyzed, the intensity ratios of the target gene compared to the reference gene were acquired. Statistics Measurement data were expressed as mean ± standard deviation (x̅ ±SD). Data were statistically processed with SPSS 19.0 software package. Intergroup differences were analyzed with analysis of variance (ANOVA). Leven homogeneity test is performed rst. Comparing of homogeneous data is determined by Bonferroni Test, otherwise they were judged by Tamhane Test; P <0.05 was considered statistically signi cant.

Specimens observation
Observed from the specimens, there was no signi cant difference in the appearance of the brain tissues among ve groups of rats. Viewed from the ventral aspect, the anterior midlines of the brainstem of rats in Model Group and MLA Group were slightly shallower than the other groups, suggesting edema in the brainstem, see mRNA and had a tendency to increase the expression of Olig-2 mRNA. MLA, the antagonist of α7nAChR, is the opposite (Figure 7).

Discussion
In the Central Nervous System (CNS), LPS can activate astrocytes through Toll-like receptor4 (TLR4) to produce pro-in ammatory cytokines and arouse neuro-in ammation [ 21 -22 ]. Moreover, cytokines elevated in the peripheral blood circulation can also be transported into the CNS to directly inspire neuroin ammation [ 23 -24 ].Neuro-in ammation boost nerve cells apoptosis, neurodegeneration and metabolic disorders through NF-kB/STAT3/ERK pathway and mitochondria-mediated apoptosis [ 25 -262 ]. Based on these testimonies, we reasonably concluded that in ammation occurs in MVZ in early sepsis, what's more, the con guration and function of MVZ is bond to change under in ammation and they should be affected by the intervention of CAP. In this study we found that although the appearance of the whole brain specimens from ve group didn't show signi cant difference in the early stage of sepsis, HE staining signi ed that MVZ neurons in three Sepsis Groups signi cantly lost and glial cells signi cantly proliferated. It is no doubt that neuro-in ammation should be responsible for these changes. The pathologic changes suggested that neuro-remodeling occurred in MVZ in early sepsis stage. Neuroremodeling involves changes in the activity and number of functional neurons and regulations in the expression of key genes. What will happen to these functional neurons and key genes in sepsis? When CAP is interfered what will happened to them next? Solving these problems may help to explain the suppression of CAP and uncontrolled systemic in ammation in sepsis. . In order to explore the activity and apoptosis of these two kinds of neurons, we conducted CHAT/Caspase 3 and TH/Caspase 3 immuno uorescence double labeling detection. The results tell us that sepsis leads to obvious apoptosis of these two kinds of neurons. Along with apoptosis of catecholaminergic neurons, the biosynthesis of TH declined obviously, GTS -21 protected catecholaminergic neurons from apoptosis and promoted the biosynthesis of TH in sepsis, MLA signi cantly worsen the apoptosis and further reduced the production of TH in sepsis; For the cholinergic neurons, GTS-21 has the tendency to facilitate the production of CHAT, while MLA signi cantly reduce the output of CHAT in sepsis. Our results are consistent with other studies that CAP activation in CNS has multiple functions such as anti-apoptosis, which depend on its reduction of central in ammation, preventing the brain from injury and improving the activity of functional neurons [ 35 -36 ]. Here, we found an interesting phenomenon that TH expression decreased simultaneously with the increasing apoptosis of catecholaminergic neurons, while CHAT expression had no signi cant reduction when there was evident apoptosis of cholinergic neurons, suggesting that sepsis promotes the functional activation of the cholinergic system of MVZ to increase the vagal output, which may be an important mechanism for its systemic anti-in ammatory effect [ 37 -38 ]. This study indirectly indicates that cholinergic neurons may be the major adjusting mediators of systemic in ammation in MVZ.

Analysis of TUNEL
As mentioned above, pathological studies suggest that sepsis in early stage leads to apoptosis of MVZ neurons and glial cell proliferation [ 39 ], manifesting neuro-remodeling occurs. In order to clarify the tendency and regulating mechanism of neuro-remodeling in MVZ in sepsis, we sequentially detected the expression levels of GAP-43 mRNA, Olig-2 mRNA, VEGF mRNA, GFAP mRNA, and MMP-9 mRNA in each group. Results show that compared with the control group, the expressions of GAP-43 mRNA, VEGF mRNA, GFAP mRNA and MMP-9 mRNA were signi cantly up-regulated, while the expression of Olig-2 mRNA was signi cantly down-regulated in the Sepsis Group; GTS-21 signi cantly reduced the expressions of GAP-43 mRNA, VEGF mRNA, GFAP mRNA and MMP-9 mRNA, increased the expression of Olig-2 mRNA in sepsis; MLA was completely the opposite.
Gap-43 is involved in the regeneration of neurons, extension of axons, and regulation of synaptic development and reconstruction to restore impaired nerve function [ 40 -413 ], it dynamically adjusts its expression level according to the severity degree of nerve damage [ 43 -44 ]. In this study, it can be seen that sepsis leads to signi cant up-regulation of Gap-43 mRNA expression, which is down-regulated by GTS-21 through inhibiting in ammation, and further promoted by MLA through exacerbating in ammation. MMP-9 is a kind of protease capable of refactoring extracellular matrix [ 45 -46 ]. The content, activity and gene expression of MMP-9 arouse after a variety of physiological stimulation and pathological damage in the nervous tissue. By shaping the dendritic spines and altering the function of the excitatory synapses, MMP-9 participates in the dynamic adjustment of synaptic plasticity [ 47 -48 ]. Our study con rmed that through regulating central in ammation levels by CAP interfering, the gene expression level of MMP -9 changed signi cantly, suggesting the need for neuro-remodeling. GFAP can be seen as a biomarker of neuro-in ammation [ 49 -50 ], its increasing expression represents astrocyte activation [ 51 ]. The hyperplasia of glial cells in MVZ induced by sepsis is apparently related to the up-regulation of GFAP mRNA expression. The down-regulation of GFAP mRNA expression by GTS-21 shows that the activating of CAP suppress neuro-in ammation and gliosis, on the contrary, the up-regulation of GFAP mRNA expression by MLA shows that the blocking of CAP intensify neuro-in ammation and gliosis in MVZ. Vascular endothelial growth factor (VEGF) promotes angiogenesis by stimulating endothelial cell migration, proliferation and formation, which is related to nerve recovery and neuroprotection [ 52 -534 ]. VEGF mRNA up-regulation manifests that neuro-in ammation induced by sepsis results in damage to blood vessels and nerves in MVZ and the urgent need to restore. The down-regulation of its expression by GTS-21 should contribute to its central anti-in ammation, while MLA is the opposite. Olig-2 is a speci c marker of oligodendrocytes. The decreased expression of Olig-2 mRNA indicates destruction of myelin [ 55 ], when the myelin is compensatory to regenerate, the Olig-2 expression level increases [ 56 ]. This study signi ed that neuro-in ammation caused by sepsis led to destruction of myelin, GTS-21 prompted myelination through anti-in ammatory, whereas MLA further worsened the damage to myelin through aggravating the in ammation. It can be seen that the decrease of CAP output in sepsis from our past research may be related to the destruction of myelin sheath in MVZ.
The expression levels of GAP-43 mRNA, VEGF mRNA, GFAP mRNA and MMP-9 mRNA are consistently upregulated in sepsis in MVZ, they also have almost the same reaction to the intervention of CAP, indicating that MVZ has an integral internal regulatory mechanism on in ammation, injury and restoration, which jointly promotes the recovery of neurovascular unit and its functions. These key genes' expression levels can re ect the severity of MVZ in ammation, the degree of damage and the tendency of repairing and remodeling. At the same time, the consistency of key gene expression also re ects the complex communication connection of CNS at the cellular and even molecular level. Even if the nerve tissue is damaged, there is still communication help to promote the recovery of damaged nerve function [ 57 -58 ].
Here, an issue emerged, that is if these key genes is subjected to the in ammatory level, or directly controlled by CAP through a speci c mechanism needed to be answered by further research.

Conclusion
This study preliminarily con rmed that sepsis causes functional neurons' apoptosis and inactivity in MVZ, key genes' expressions related to glial activation, nerve repairing and remodeling signi cantly increased, indicating the tendency of MVZ restoring and remodeling in sepsis; CAP can signi cantly affect these process. Agonists of α7nAChR can improve the remodeling of MVZ and facilitate MVZ to "normalize", which may be one of mechanism that GTS-21 can effectively curb the septic in ammatory storm, while antagonists of α7nAChR do the opposite. Therefore, anti-in ammation in MVZ may be a prospective way to treat the early stage of sepsis. Of course, this study still has some limitations, for example, whether the effect of α7nAChR intervention on Neuro-remodeling of MVZ is through some speci c pathways, or is it secondary to the adjustment of the level of systemic or central in ammation, or include both. These questions need to be studied in the future research.

Availability of data and materials
Because this study is supported by the Guizhou Provincial Science and Technology Foundation, the data will not make known to public until the research project is checked and accepted. Therefore, we declare that all the datasets used and/or analysed during the current study are available from the author on reasonable request. The link email is mrbright789@sina.com.

Competing interests
The authors declare that they have no competing interests.  Brain specimens of ve groups viewed from the ventral aspect