Neuroinflammation in the medullary visceral zone exert a powerful impaction on the systemic inflammation in sepsis through cholinergic anti-inflammatory pathway

Abstract

To investigate whether sepsis-induced neuroinflammation of medulla visceral zone (MVZ) predominates the systemic inflammation through cholinergic anti-inflammatory pathway (CAP), and to explore the effect of central anti-inflammation on systemic inflammation. 112 adult Sprague–Dawley male rats were randomly divided into sepsis experimental group (n = 56) and neuroinflammation experimental group (n = 56). The two experimental groups were individually randomly divided into control group (n = 8), model group (n = 16), central anti-inflammatory group (n = 16) and vagus transection group (n = 16). Rats in two control groups were administered with saline at the dose of 6 mL/kg intraperitoneally or with 25 μL artificial cerebrospinal fluid injected into forth ventricle once a day for 3 days. Rats in two model groups were administered with Lipopolysaccharide (LPS) at the dose of 6 mg/kg intraperitoneally or with 25 μg/25 μL LPS injected into forth ventricle once a day for 3 days. Rats in two central anti-inflammatory groups were fed with 10 mg/mL minocycline sucrose solution as the only water source for 4 days prior to be treated as the model groups of their own, and feeding style was continued until the end of the experiment. Rats in the two vagus transection groups were undergone right vagotomy and 7 days of adaptive feeding prior to be treated as the same as those in the central anti-inflammatory group of their own. The Murine Sepsis Score (MSS), mortality rate and heat rate variability (HRV) were recorded during the last 3 days of intervention. Then the rats were sacrificed and blood samples were collected for ELISA analysis to detect the serum level of inflammatory cytokines such as TNF-α, IL-6, and IL-10. The expression of TNF-α and IL-6 in medulla oblongata were analyzed by Western blot. The correlation and regression analysis among the expression levels of cytokines in medulla oblongata, HRV indexes and serum inflammatory cytokines were performed. The mortality rate and MSS of the sepsis model group and the MVZ’s neuroinflammation model group were significantly higher than those of their own control group, and the central anti-inflammation reduced the mortality rate and MSS scores of the two model groups, while the right vagotomy abolished the effect of central anti-inflammatory. In the sepsis model group and the MVZ’s neuroinflammation model group, the levels of TNF-α, IL-6, and other cytokines in serum and MVZ were significantly increased, and HRV indexes (SDNN, RMSSD, LF, HF, LF/HF) were significantly decreased (P = 0.000). Central anti-inflammatory treatment reversed the above changes. However, right vagotomy abolished the central anti-inflammatory effect. Correlation and regression analysis showed that there was a significant linear correlation among the expression of inflammatory factors in MVZ, the indexes of HRV and the levels of serum cytokines. Our study shows that sepsis-induced MVZ’s neuroinflammation exert a powerful influence on the systemic inflammation through CAP in sepsis. Central anti-inflammation effectively improves systemic inflammation through inhibiting MVZ’s neuroinflammation in sepsis. The time domain and frequency domain indexes of HRV can reflect the regulatory effect of CAP and the degree of inflammation of MVZ, which may be potentially used to monitor the condition and treatment effectiveness of sepsis patients.

Introduction

Sepsis is an uncontrolled systemic inflammatory response of the host to infection, leading to life-threatening organ dysfunction^1,2. Sepsis exhibits significant heterogeneity in both severity and clinical outcomes. Multiple factors such as genetic susceptibility, site and cause of infection, host response, and related interventions greatly influence patients’ outcomes³. In addition, the pathological mechanism of sepsis involves abnormal inflammatory reaction to infection, including the immune reaction activation of both innate and adaptive immunity, participation of endothelial cells and complement system, coagulation abnormalities, and other complex interactions⁴. The heterogeneity on outcome and complexity on mechanism have led to much debate on the bundled therapy for sepsis⁵, such as fluid resuscitation and steroid use. As a result, future research directions for sepsis remain unclear. Currently, it remains one of the leading causes of death worldwide. Therefore, a recognition of sepsis from foundation to clinical is needed⁶.

The failure of early goal directed therapy in large randomized controlled clinical trials indicates that these treatments did not change the underlying pathological changes of sepsis^7,8. Chinese scholars have utilized Xuebijing Injection to treat sepsis effectively because of its suppressing the excessive release of inflammatory mediators produced by monocytes or macrophages stimulated by endotoxins, thereby Xuebijing Injection modulates hypersensitive immune responses through various mechanisms and significantly reduces the mortality rate in sepsis⁶. This suggests that modulating immunity and inhibiting abnormal inflammatory responses should be still key issues for research and treatment of sepsis.

Sepsis induces activation of microglia and excessive production of pro-inflammatory cytokines in central nervous system (CNS), thus it promotes neuroinflammation, neurodegeneration, and autonomic dysfunction^9–11. In fact, the central and peripheral nervous systems are the first systems to be impaired in septic patients¹². Based on the theory of cholinergic anti-inflammatory pathway (CAP), the medullary visceral zone (MVZ) plays an important regulatory role in systemic inflammation and neuroinflammation^13,14. Therefore, maintaining normal structure and function of MVZ is crucial in sepsis. Previous studies have also demonstrated that sepsis-induced apoptosis and suppression of cholinergic and catecholaminergic neurons in MVZ may be associated with functional inhibition of CAP and uncontrolled systemic inflammation¹⁵. However, Whether the uncontrolled systemic inflammation caused by sepsis is primarily an inflammatory storm caused by the disordered self-activation of peripheral cytokines, or it is mainly an inflammation-induced dysfunction of MVZ leading to the loss control of systemic inflammation still remains unclear, as there is currently no relative research on this puzzle.

This study is to explore the activation of MVZ’s neuroinflammation by sepsis models and its impact on CAP’s modulation and systemic inflammation. It is also to investigate the central anti-inflammatory effects and whether these effects were abolished after disconnection of CAP, thus confirming that whether or not the neuroinflammation in MVZ is a main mechanism underlying the loss control of the systemic inflammation in sepsis. Additionally, we also utilized a neuroinflammation model in MVZ induced by lipopolysaccharide injection into the fourth ventricle to investigate whether or not the MVZ’s neuroinflammation exert a significant influence on CAP’s modulation and systemic inflammation. And the same to the sepsis models, we also investigate the effects of central anti-inflammation and disconnection of CAP in MVZ’s neuroinflammation models, thereby validating whether or not neuroinflammation in MVZ is a critical pathological mechanism for the uncontrolled systemic inflammation in sepsis.

Materials and methods

Rats’ grouping and treatment

Specific Pathogens Free (SPF) Sprague–Dawley male rats were purchased from the Hubei Provincial Experimental Animal Research Center, with body weight of 180–220 g. License number: SCXK (Hubei) 2020-0018. The animals were housed in the experimental animal center of Guizhou Medical University and provided with access to food and water ad libitum. The light/dark cycle was set at 12/12 h (lights on during 06:00–18:00) and temperature maintained 21 ± 0.5 °C. The experiments commence After 7 days’ adaptation feeding. Rats were randomly divided into two experimental groups: sepsis experimental group and MVZ neuroinflammation experimental group, each experimental group was further divided into Control Group (8 rats), Model Group (16 rats), Central Anti-inflammatory Group (16 rats), and CAP disconnection + central anti-inflammatory group, here, we called it Vagus Transection Group (16 rats). All the maneuvering measures were complied with the Committee on Protection, Welfare and Ethics of Experimental Animals¹⁶, this research was approved by the Animal Care Welfare Committee of Guizhou Medical University (no. 2100819).

According to the different processing methods of the rats in different groups, the rats underwent 11–21 days of treatment before sampling and analysis. Treatment of rats in each group are as follows:

(1) Control Group: The sepsis control group was administered with 6 mg/kg of normal saline, intraperitoneal injection, once a day for 3 days.

The control group of MVZ neuroinflammation was injected with 25 μL of artificial cerebrospinal fluid (ACSF) into the fourth ventricle once a day for three consecutive days. ACSF is composed of sodium chloride (6.279 g), potassium chloride (0.216 g), calcium chloride (0.353 g), magnesium chloride (0.488 g), sodium bicarbonate (1.932 g), glucose (0.6 g), disodium hydrogen phosphate (0.358 g), add purified water to make up to 1000 mL in volume¹⁷.

(2) Model Group: sepsis model group: LPS (6 mg/kg) was intraperitoneally injected once a day for 3 days¹⁸. MVZ neuroinflammation model group: LPS 25 μg/25 μL was injected into the fourth ventricle once a day for 3 days.

In strict accordance with ethical requirements, whether the rats went through the fourth ventricle injection or being sacrificed should be under anesthetization by inhalation of isoflurane. Preparation of MVZ neuroinflammation model including such steps: After anesthesia, the rats were fixed with a stereotaxic instrument. According to the Paxions and Watson atlas, set the anterior fontanel as the coordinate, locate the surface injection point of the fourth ventricle as 11.6 mm posterior to the anterior fontanelle on the median line. A small hole was made at this point with a microsyringe, and the needle was inserted vertical into the skin 7.5–8.5 mm to reach the fourth ventricle. The pale yellow clear cerebrospinal fluid was extracted without resistance while gently withdrawing the plunger of the syringe, which was the proof of entry into the fourth ventricle. LPS dissolved in ACSF (25 μg/25 μL) was injected vertically into the fourth ventricle slowly. After the injection, the needle was retained for 5 min to allow LPS to be fully absorbed, then the needle was withdrawn slowly.

(3) Central Anti-inflammatory Group: after 4 days of feeding with 10 mg/mL minocycline phosphate (dissolved in 5% sucrose solution) as the only water source, the sepsis model rats were administered with intraperitoneally injection of LPS (6 mg/kg), and the MVZ neuroinflammation model rats were accepted injection of LPS 25 μg/25 μL by fourth ventricle as their own Model Groups, once a day for 3 days. Minocycline feeding was continued until the day of sampling.

(4) CAP Transection Group, rats undergone both CAP transection and central anti-inflammatory treatment. After isolation and transection of the right cervical vagus nerve under anesthetization by inhalation of isoflurane, rats went through adaptive feeding for 7 days, the rest of the treatment was the same as the Central Anti-inflammatory Group as mentioned above. The rats’ processing procedure is shown in the Fig. 1.

Figure 1.

The rats’ grouping and processing procedure.

During the treatment period, there are separately 7, 3, 4 rats died in the Sepsis Group, Central Anti-inflammatory Group and Vagus Transection Group in the sepsis experiment; simultaneously, there are separately 10, 4, 7 rats died in the Neuroinflammation Group, Central Anti-inflammatory Group and Vagus Transection Group in the neuroinflammation experiment. All the survival rats were involved with the next experiment.

Murine Sepsis Score (MSS) of rat

According to MSS scoring system¹⁹ (Table 1), the higher the score, the more severe the rats. According to our previous experiment¹⁵, MSS > 15 means success of preparing sepsis model. Three experimenters evaluated every experimental rats separately, the average score was taken for evaluation of the severity of the rats.

Table 1.

Murine sepsis score.

Variable	Score and description
Appearance	0—Coat is smooth
	1—Patches of hair piloerected
	2—Majority of back is piloerected
	3—Piloerection may or may not be present, mouse appears “puffy”
	4—Piloerection may or may not be present, mouse appears emaciated
Level of consciousness	0—Mouse is active
	1—Mouse is active but avoids standing upright
	2—Mouse activity is noticeably slowed. The mouse is still ambulant
	3—Activity is impaired. Mouse only moves when provoked, movements have a tremor
	4—Activity severely impaired. Mouse remains stationary when provoked, with possible tremor
Activity	0—Normal amount of activity. Mouse is any of: eating, drinking, climbing, running, fighting
	1—Slightly suppressed activity. Mouse is moving around bottom of cage
	2—Suppressed activity. Mouse is stationary with occasional investigative movements
	3—No activity. Mouse is stationary
	4—No activity. Mouse experiencing tremors, particularly in the hind legs
Response to stimulus	0—Mouse responds immediately to auditory stimulus or touch
	1—Slow or no response to auditory stimulus; strong response to touch (moves to escape)
	2—No response to auditory stimulus; moderate response to touch (moves a few steps)
	3—No response to auditory stimulus; mild response to touch (no locomotion)
	4—No response to auditory stimulus. Little or no response to touch. Cannot right itself if pushed over
Eyes	0—Open
	1—Eyes not fully open, possibly with secretions
	2—Eyes at least half closed, possibly with secretions
	3—Eyes half closed or more, possibly with secretions
	4—Eyes closed or milky
Respiration rate	0—Normal, rapid mouse respiration
	1—Slightly decreased respiration (rate not quantifiable by eye)
	2—Moderately reduced respiration (rate at the upper range of quantifying by eye)
	3—Severely reduced respiration (rate easily countable by eye, 0.5 s between breaths)
	4—Extremely reduced respiration (> 1 s between breaths)
Respiration quality	0—Normal
	1—Brief periods of laboured breathing
	2—Laboured, no gasping
	3—Laboured with intermittent gasps
	4—Gasping

Evaluation of CAP’s activity

Based on the previous study¹⁵, the time-domain parameters such as standard deviation of all RR intervals (SDNN), root mean square difference of successive RR intervals (RMSSD), frequency domain index such as low frequency (LF, 0.2–0.75 Hz) and high frequency (HF, 0.75–2.5 Hz Hz) power were analyzed by 5-min short-duration electrocardiogram. HRV was measured by BL-420F biological signal acquisition and analysis system manufactured by Chengdu Tailong Software Co., LTD.

Blood collection and medulla oblongata sampling

After MSS evaluation and HRV analysis, all the survival rats were anesthetized by inhalation of isoflurane, their chests were opened to expose the hearts. Blood samples (8 mL) were collected from the right ventricle and remained for 1 h at room temperature, then they were centrifuged at 3000r for 10 min to obtain serum for detection. After blood collection, the rats were perfused with normal saline until the internal organs turned pale, then the medulla oblongata of the rats were got and frozen for further treatment.

Enzyme linked immunosorbent assay (ELISA) test (kit)

The serum cytokines detection was performed according to the instructions of the ELISA kit. 10 μL rat’s serum was added into the sample well of the plate, and the corresponding standard well was set up at the same time. The plate was gently shaken and then incubated at 37 °C for 2 h. After washing and drying, 100 μL biotin-labeled antibody working solution was added, including Rat Tumor Necrosis Factor Alpha (TNF-α), batch number: MM-0180R1, manufacturer: elabscience; Rat Interleukin 6 (IL-6), batch number: MM-0190R1, Manufacturer: elabscience; Rat IL-10, batch number: MM-0195R1, Manufacturer: elabscience. Afterwards, the plates were incubated at 37 °C for 1 h prior to be dried and washed for 3 times. At last, 50 μL of chromogenic A and B were added to each well, and the optical density (OD) value of each well was measured at 450 nm wavelength by a spectrophotometer reader within 10 min after stopping reaction. The concentrations of various inflammatory molecules were calculated.

Western blot assay

According to Paxions and Watson's atlas, samples of the medulla oblongata should contain the dorsal vagus motor nucleus (DVMN), the nucleus of the solitary tract (NTS), and (rostral ventrolateral medulla, RVLM), Fig. 2. A small amount of medulla oblong tissue was placed in 2 mL EP tubes, and 200 μL of tissue lysate was added to each tube to lyse cells. Then the total protein was extracted by centrifugation, and protein concentration was determined by BCA method. Equal amount of protein was loaded, denatured and electrophorized on the polyacrylamide gel and then were transferred to polyvinylidene difluoride membranes, followed by blocking with 8% skim milk at room temperature for 1 h and washing. The membranes were cut into strips and put into the corresponding primary antibodies solution, including: rabbit polyclonal antibody TNF-a (21KD), manufacturer: Biyuntian, batch number: AF8208, dilution ratio: 1:1000; Rabbit polyclonal antibody IL-6 (24KD), manufacturer: Affinity, batch number:DF6087, dilution ratio: 1:1000; Rabbit polyclonal antibody IL-10 (21KD), manufacturer: Affinity, batch number: DF6894, dilution ratio: 1:100, the solution was incubated overnight at 4 °C on a shaker. The next day, the strips were washed and put into HRP labeled sheep anti-rabbit secondary antibody (manufacturer: Wuhan Boster, batch number: BA10541, dilution ratio: 10,000). The PVDF membrane was immersed in the secondary antibody solution and incubated at room temperature for 2 h on a shaker. Finally, ECL luminescent solution was used for imaging and exposure on a gel imager, and the film gray value was analyzed by ipp6.0 software to calculate the expression level of proteins.

Figure 2.

Mapping of the sampling site of the medulla oblongata.

According to Paxions and Watson’s atlas, set the anterior fontanel as the coordinate, locate the central site of medulla oblongata sampling as 12.8 mm posterior to the anterior fontanelle. Samples of the medulla oblongata should contain the dorsal vagus motor nucleus (DVMN), Marked with red box; the nucleus of the solitary tract (NTS), Marked with bule box; and (rostral ventrolateral medulla, RVLM), Marked with green box.

Statistical analysis

All the data were expressed as mean ± standard deviation (M$\overline{\text{X}}$ ± S). The experimental data were statistically processed by SPSS 22.0 software package. Measurement data were analyzed by analysis of variance. Leven’s homogeneity test was performed first, and Bonferroni test was used to determine homogeneous data, otherwise, they were determined by Tamhane’s test. The counting data were analyzed by χ² test. P < 0.05 suggests the results significant, P < 0.01 suggests the results very significant.

Ethics statement

The study was approved by Ethics Committee of Guizhou Medical University (no. 2100819). All the maneuvering measures were complied with the guidelines and regulations of Committee on Protection, Welfare and Ethics of Experimental Animals. This study was carried out in compliance with the ARRIVE guidelines.

Results

Murin sepsis score (MSS) and mortality among different groups

Although there was no significant difference in body weight among groups, or between pretreatment and after treatment, both sepsis and MVZ neuroinflammation led to significant increase in MSS and significant decrease in 3-day survival rate in two model rats. Central anti-inflammation by minocycline not only significantly reduced MSS of both model rats, but also significantly improved the survival rate in the neuroinflammation model rats, and showed a trend of improving the survival rate in the sepsis model rats. However, right vagotomy abolished the reduction of MSS caused by central anti-inflammation, it also significantly reduced the survival rate in Central Anti-inflammatory Group of the neuroinflammation experimental group, and the increasing survival rate by central anti-inflammatory in the sepsis experimental group also showed a decreasing trend in the Vagus Transection Group, see Fig. 3.

Figure 3.

The MSS and Kaplan–Meier Survival Curve among different groups. (A) In the first day after treatment, there were separately 8, 12, 14 and 13 rats survived in Control Group, Sepsis Group, Central Anti-inflammation Group and Vagus Transection Group, and In the second day after treatment, there were separately 8, 10, 13 and 12 rats survived in Control Group, Sepsis Group, Central Anti-inflammation Group and Vagus Transection Group, In the third day after treatment, there were separately 8, 9, 13 and 11 rats survived in Control Group, Sepsis Group, Central Anti-inflammation Group and Vagus Transection Group. All the survival rats were involved with the MSS, HRV, Western blot and Elisa test. The MSS of the Sepsis Model Group was significantly higher than that of the Control Group (24.29 ± 2.21 vs 0, P = 0.000), and central anti-inflammation significantly reduced the MSS score in the model rats (5.43 ± 0.98 vs 24.29 ± 2.21, P = 0.000). The MSS score of the Vagus Transection Group was significantly higher than that of the Central Anti-inflammatory Group (17.29 ± 0.95 vs 5.43 ± 0.98, P = 0.000), and the difference between any two groups was extremely significant (P = 0.000). (B) The survival rate of the Sepsis Group was significantly lower than that of the Control Group (82.1% vs 100%, P = 0.029), and the survival rate of the Central Anti-inflammatory Group was higher than that of the Sepsis Group (93.2% vs 82.1%, P = 0.123). The survival rate of the Vagus Transection Group has a trend to descend when compared to that of the Central Anti-inflammatory Group (87.8% vs 93.2%, P = 0.401). (C) In the first day after treatment, there were separately 8, 11, 14 and 12 rats survived in Control Group, Neuroinflammation Group, Central Anti-inflammation Group and Vagus Transection Group, and In the second day after treatment, there were separately 8, 8, 13 and 10 rats survived in Control Group, Neuroinflammation Group, Central Anti-inflammation Group and Vagus Transection Group, In the third day after treatment, there were separately 8, 6, 12 and 9 rats survived in Control Group, Neuroinflammation Group, Central Anti-inflammation Group and Vagus Transection Group. All the survival rats were involved with the MSS, HRV, Western blot and Elisa test. The MSS score of the Neuroinflammation Group was significantly higher than that of the Control Group (26.43 ± 1.99 vs 0, P = 0.000), and the Central Anti-inflammatory Group significantly reduced the MSS score of the Model Group (14.29 ± 2.87 vs 26.43 ± 1.99, P = 0.000). The MSS score of the Vagus Transection Group was significantly higher than that of the Central Anti-inflammatory Group (18.71 ± 2.69 vs 14.29 ± 2.87, P = 0.006). (D) The survival rate of the Neuroinflammation Group was significantly lower than that of the Control Group (71.4% vs 100%, P = 0.004), and the survival rate of the Central Anti-inflammatory Group was significantly higher than that of the Neuroinflammation Group (90.7% vs 71.4%, P = 0.022). The survival rate of the Vagus Transection Group tended to decrease compared to that of the Central Anti-inflammatory Group (81.6% vs 90.7%, P = 0.220). Note: **P < 0.01, *P < 0.05.

HRV analysis in rats among different groups

HRV indexes including SDNN, RMSSD, LF, HF and LF/HF were all significantly decreased in both sepsis and MVZ neuroinflammation model rats, and central anti-inflammation by minocycline significantly improved these HRV indexes in two model rats. However, right vagotomy basically abolished the central anti-inflammatory effect, resulting in a significant decrease in these indexes, refer to Fig. 4.

Figure 4.

The indexes of HRV among different groups. Note: **P < 0.01; *P < 0.05.

HRV indexes in the Sepsis Group were significantly lower than those in the Control Group [SDNN (ms): 1.28 ± 0.22 vs 3.40 ± 0.35, P = 0.000; RMSSD (ms): 0.86 ± 0.18 vs 2.83 ± 0.37, P = 0.000; LF (ms²): 126.00 ± 9.17 vs 306.45 ± 19.76, P = 0.000; HF (ms²): 24.85 ± 1.81 vs 42.94 ± 3.61, P = 0.000; LF/HF: 5.07 ± 0.07 vs 7.15 ± 0.29, P = 0.000]; HRV indexes in the Central Anti-inflammatory Group were significantly higher than those in the Model Group [SDNN (ms): 2.31 ± 0.17 vs 1.28 ± 0.22, P = 0.000; RMSSD (ms): 1.93 ± 0.19 vs 0.86 ± 0.18, P = 0.000; LF (ms²): 205.94 ± 22.17 vs 126.00 ± 9.17, P = 0.000; HF (ms²): 31.15 ± 3.19 vs 24.85 ± 1.81, P = 0.001; LF/HF: 6.61 ± 0.13 vs 5.07 ± 0.07, P = 0.000]; HRV indexes in the Vagus Transection Group were significantly lower than those in the Central Anti-inflammatory Group [SDNN (ms): 2.31 ± 0.17 vs 1.28 ± 0.22, P = 0.000; RMSSD (ms): 1.93 ± 0.19 vs 0.86 ± 0.18, P = 0.000; LF (ms²): 205.94 ± 22.17 vs 126.00 ± 9.17, P = 0.000; HF (ms²): 31.15 ± 3.19 vs 24.85 ± 1.81, P = 0.001; LF/HF: 6.61 ± 0.13 vs 5.07 ± 0.07, P = 0.000]. There was no significant difference in LF and HF between the Model Group and the Vagus Transection Group (P = 0.116; P = 0.464).

Similarly, HRV parameters in the MVZ Neuroinflammation Group were much lower than those in the Control Group [SDNN (ms): 1.10 ± 0.22 vs 3.52 ± 0.35, P = 0.000; RMSSD (ms): 0.74 ± 0.18 vs 2.87 ± 0.37, P = 0.000; LF (ms²): 114.76 ± 9.17 vs 312.76 ± 19.76, P = 0.000; HF (ms²): 21.70 ± 1.81 vs 41.32 ± 3.49, P = 0.000; LF/HF: 5.29 ± 0.08 vs 7.58 ± 0.23, P = 0.000]; HRV parameters in the Central Anti-inflammatory Group were significantly higher than those in the Model Group [SDNN (ms): 2.13 ± 0.17 vs 1.10 ± 0.22, P = 0.000; RMSSD (ms): 1.81 ± 0.19 vs 0.74 ± 0.18, P = 0.000; LF (ms²): 194.70 ± 22.17 vs 114.76 ± 9.17, P = 0.000; HF (ms²): 28.00 ± 3.19 vs 21.70 ± 1.81, P = 0.001; LF/HF: 6.95 ± 0.14 vs 5.29 ± 0.08, P = 0.000]; HRV parameters in the Vagus Transection Group were obviously lower than those in the Central Anti-inflammatory Group [SDNN (ms): 1.44 ± 0.08 vs 2.13 ± 0.17, P = 0.000; RMSSD (ms): 1.18 ± 0.15 vs 1.81 ± 0.19, P = 0.000; LF (ms²): 133.68 ± 21.06 vs 194.70 ± 22.17, P = 0.000; HF (ms²): 21.16 ± 3.31 vs 28.00 ± 3.19, P = 0.000; LF/HF: 6.34 ± 0.53 vs 5.29 ± 0.08, P = 0.001]. There was no significant difference in LF and HF between the Model Group and the Vagus Transection Group (P = 0.071; P = 0.741).

Serum inflammatory cytokines comparation among different groups

The serum levels of TNF-a, IL-6 and IL-10 in both sepsis models and MVZ neuroinflammation models were significantly increased, and central anti-inflammation by minocycline significantly reduced the serum levels of TNF-a, IL-6 and IL-10 in these two model rats. However, right vagotomy abolished the central anti-inflammatory effect, resulting in significant increase of serum TNF-a, IL-6 and IL-10 in the rats of Vagus Transection Group, refer to Fig. 5.

Figure 5.

The serum concentration of inflammatory cytokines among different groups.

The serum concentration of inflammatory cytokines in the Sepsis Group were significantly higher than those in the Control Group [TNF-a (ng/L): 32.84 ± 1.39 vs 9.51 ± 1.50, P = 0.000; IL-6 (ng/L): 36.94 ± 2.50 vs 9.94 ± 1.44, P = 0.000; IL-10 (ng/L): 19.01 ± 1.21 vs 5.67 ± 1.07, P = 0.000]; They decreased significantly in the Central Anti-inflammatory Group compared to the Model Group [TNF-a (ng/L): 16.46 ± 1.55 vs 32.84 ± 1.39, P = 0.000; IL-6 (ng/L): 25.01 ± 1.13 vs 36.94 ± 2.50, P = 0.000; IL-10 (ng/L): 9.44 ± 1.03 vs 19.01 ± 1.21, P = 0.000], whereas, in the Vagus Transection Group, they had marvelous rebounds compared to the Central Anti-inflammatory Group [TNF-a (ng/L): 27.36 ± 1.48 vs 16.46 ± 1.55, P = 0.000; IL-6 (ng/L): 30.30 ± 1.91 vs 25.01 ± 1.13, P = 0.007; IL-10 (ng/L): 14.08 ± 0.78 vs 9.44 ± 1.03, P = 0.001].

Similarly, the concentrations of serum inflammatory cytokines in the Neuroinflammation Group were significantly higher than those in the Control Group [TNF-a (ng/L): 67.29 ± 3.49 vs 14.16 ± 2.00, P = 0.000; IL-6 (ng/L): 35.24 ± 3.54 vs 7.15 ± 2.13, P = 0.000; IL-10 (ng/L): 11.89 ± 0.54 vs 6.31 ± 0.49, P = 0.000]; They significantly decreased in the Central Anti-inflammation Group compared to the Model Group [TNF-a (ng/L): 26.24 ± 4.43 vs 67.29 ± 3.49, P = 0.000; IL-6 (ng/L): 21.76 ± 1.30 vs 35.24 ± 3.54, P = 0.001; IL-10 (ng/L): 8.00 ± 0.53 vs 11.89 ± 0.54, P = 0.000]; the same to the results in the sepsis models, The serum levels of inflammatory cytokines in the Vagus Transection Group increased significantly compared to the Central Anti-inflammatory Group [TNF-a (ng/L): 47.20 ± 4.78 vs 26.24 ± 4.43, P = 0.000; IL-6 (ng/L): 29.45 ± 1.28 vs 21.76 ± 1.30, P = 0.019; IL-10 (ng/L): 10.35 ± 0.69 vs 8.00 ± 0.53, P = 0.006]. There was no significant difference in the serum concentrations of IL-6 and IL-10 between the Vagus Transection Group and the Neuroinflammation Group (P = 0.083; P = 0.065).

Inflammatory cytokines expression in MVZ among different groups

The expression levels of TNF-a and IL-6 in MVZ induced by sepsis or by MVZ neuroinflammation were all significantly increased, and central anti-inflammation by minocycline significantly reduced the expression levels of TNF-a and IL-6 in MVZ induced either by sepsis or by MVZ neuroinflammation. However, right vagotomy abolished the central anti-inflammatory effect by minocycline, resulting in apparently increased expression of TNF-a and IL-6 in MVZ in both Central Anti-inflammation Groups, see Fig. 6.

Figure 6.

The expression of inflammatory cytokines in MVZ among different groups. (A) The western blotting scanning images of inflammatory cytokines among different groups. The display of cropped gels and blots was used in order to improve the clarity and conciseness of the presentation, the original blots/gels are presented in Supplementary Fig. 1. (B) The histograms show that The expressions of MVZ inflammatory cytokines in the Sepsis Group were significantly higher than those in the Control Group (TNF-a: 0.551 ± 0.044 vs 0.083 ± 0.039, P = 0.000; IL-6: 0.885 ± 0.108 vs 0.229 ± 0.113, P = 0.000). Whereas, in the Central Anti-inflammatory Group, these cytokines were significantly dampened compared to those of the Model Group (TNF-a: 0.181 ± 0.054 vs 0.551 ± 0.044, P = 0.000; IL-6: 0.454 ± 0.091 vs 0.885 ± 0.108, P = 0.000). The declination of cytokines by minocycline’s anti-inflammation in MVZ was obviously reversed by the right vagotomy (TNF-a: 0.370 ± 0.069 vs 0.181 ± 0.054, P = 0.002; IL-6: 0.664 ± 0.026 vs 0.454 ± 0.091, P = 0.023). The expression of TNF-a in the Central Anti-inflammatory Group was not significantly different from that in the Control Group (P = 0.051). (C) Similarly, the expressions of MVZ inflammatory cytokines in the Neuroinflammation Group were significantly higher than those in the Control Group (TNF-a: 0.761 ± 0.051 vs 0.087 ± 0.023, P = 0.000; IL-6: 0.974 ± 0.036 vs 0.150 ± 0.017, P = 0.000); These cytokines decreased significantly in the Central Anti-inflammatory Group compared to those in the Model Group (TNF-a: 0.328 ± 0.074 vs 0.761 ± 0.051, P = 0.000; IL-6: 0.505 ± 0.056 vs 0.974 ± 0.036, P = 0.001). The expressions of MVZ inflammatory cytokines in the Vagus Transection Group were significantly higher than those in the Central Anti-inflammatory Group (TNF-a: 0.618 ± 0.083 vs 0.328 ± 0.074, P = 0.003; IL-6: 0.806 ± 0.090 vs 0.505 ± 0.056, P = 0.001). There was no significant difference in the expression of TNF-a between the Neuroinflammation Group and the Vagus Transection Group (P = 0.138).

Correlation and regression analysis

There was a strong positive correlation between the expression of inflammatory cytokines in MVZ and the concentrations of inflammatory cytokines in serum. The Pearson Correlation Coefficient between MVZ’s expression of TNF-a in and serum concentration of TNF-a, IL-6, IL-10 was 0.928, 0.829, 0.607, respectively. The Pearson Correlation Coefficient between MVZ’s expression of IL-6 and serum concentration of TNF-a, IL-6, IL-10 was 0.853, 0.931, 0.724, respectively. All P values reached 0.000.

Both the expression of inflammatory cytokines in MVZ and the levels of inflammatory cytokines in serum were negatively correlated with the indexes of HRV (P = 0.000), Table 2.

Table 2.

Coefficient of correlation between HRV indicators and cytokines level in serum or MVZ.

	TNF-a (MVZ)	IL-6 (MVZ)	TNF-a (Serum)	IL-6 (Serum)	IL-10 (Serum)
SDNN	− 0.903	− 0.930	− 0.808	− 0.950	− 0.771
RMSSD	− 0.896	− 0.929	− 0.800	− 0.943	− 0.769
LF	− 0.857	− 0.908	− 0.778	− 0.935	− 0.768
HF	− 0.815	− 0.831	− 0.767	− 0.853	− 0.664

All the P values were 0.000.

The regression analysis were made including TNF-a expression in MVZ to SDNN in HRV, SDNN in HRV to serum TNF-a concentration, and TNF-a expression in MVZ to serum TNF-a concentration, the results showed that the residual values of the three pairs of variables were normal distribution, and the normal P–P plot of the standard residual also showed that the three pairs of variables were linearly correlated, with a high degree of fitting and a very significant correlation with each other (Fig. 7).

Figure 7.

The regression analysis between expression of inflammatory cytokines in MVZ, SDNN of HRV indexes and the concentrations of TNF-a in serum. The histograms in the upper row shows that the standard residuals of the three pairs of regression variables are normally distributed, the lower row of P–P plots show that the three sets of variables are linear regressions with good fitting. (A) Regression analysis was performed between the expression of TNF-a in MVZ and serum TNF-a level, β value = 0.928, T value = 11.654, P = 0.000; (B) Regression analysis was performed between the expression of TNF-a in MVZ and SDNN in HRV, β value = − 0.903, T value = − 9.884, P = 0.000; (C) The regression analysis was performed between SDNN in HRV and serum TNF-a level showed that βvalue = − 0.808, T value = − 6.431, P = 0.000.

Discussion

Our previous studies have confirmed that sepsis-induced neuroinflammation in MVZ is involved in the functional suppression of CAP and systemic inflammatory storm¹⁵. Therefore, it is necessary to confirm that whether MVZ’s neuroinflammation play a predominant role on the systemic inflammation in sepsis through the cholinergic anti-inflammatory pathway. Here, we prepared sepsis experiment and neuroinflammation experiment, and interfere with central anti-inflammation or CAP transection + central anti-inflammation to confirm our hypothesis.

Although there are many consensuses on the molecular mechanisms of abnormal systemic inflammation and immunity underlying the severe sepsis, such as uncontrolled and self-activating inflammatory storms, relative or absolute insufficiency of adrenocortical hormone to curb the systemic inflammation and so on, which lead to septic shock, multiple organ dysfunction, or even death²⁰. When sepsis patients are under the stress of cytokine storm, the increased cortisol level secreted by the adrenal cortex plays an important role to dampen the systemic inflammation, in addition, reflexive activation of the cholinergic anti-inflammatory pathway (CAP) is a key mechanism to inhibit sepsis-induced inflammatory storm, which facilitates preventing central and peripheral organs from severe damage^21–23. Studies have confirmed that sepsis induces excessive production of pro-inflammatory cytokines in the central nervous system, promotes neuroinflammation and autonomic nervous dysfunction^9,10. MVZ, including many important neuronal nucleus involving with inflammatory modulation such as the nucleus of the solitary tract (NTS), the rostral ventrolateral medulla (RVLM) and the dorsal motor nucleus of vagus (DMNV), is the primary center for the perception and modulation of systemic inflammation. NTS, with its reciprocal polysynaptic connections to the medulla oblongata and hypothalamic centers, is crucial for the central processing of peripheral inflammatory signals²². RVLM is interconnected with forebrain, dorsal motor nucleus of vagus (DMNV) and paraventricular nucleus of hypothalamus, it is also involved in the processing of inflammatory signals and is associated with pathological behaviors related to sepsis^24,25. Therefore, sepsis-induced MVZ neuroinflammation is inevitable to affect the function of MVZ in regulating systemic inflammation.

At present, the relationship between MVZ neuroinflammation and systemic inflammation is still unclear, that is, is it possible that the uncontrolled systemic inflammation induced by sepsis is predominated mainly by the inflammation and dysfunction of MVZ, meanwhile, the self-activation and enlargement of inflammatory cytokines are not the main mechanism for the cytokines storm. Further exploration is needed.

Peripherally inflammatory cytokines can induce neuroinflammation through multiple pathways, for example, cytokines can pass through periventricular organs which lack the blood–brain barrier (BBB), or bind to their receptors expressed on the endothelial cells to deliver systemic inflammation to CNS²⁶. In addition, sepsis is often accompanied by the disruption of BBB, which facilitates the entry of immune cells into brain and activation of glial cells. Microglia are resident immune cells in the central nervous system (CNS). They continuously monitor the CNS microenvironment and respond rapidly to stimuli²⁷. When they were stimulated by inflammatory cytokines such as TNF- a and IL-1, their bodies become enlarged and their processes become shortened, which signify they turn into “activated type” and participate in amplifying inflammatory signals^28,29. Researches show that sepsis induces massive proliferation of microglia in periventricular organs, hypothalamus, medulla oblongata, and limbic systems, leading to pathological behavior and emotion manifestations in rats³⁰. In the present study, the Sepsis group had much higher mortality and MSS scores than the Control Group. Our study is conformed to the previous investigation^31,32, that is, the pathological fatigue of rats with high MSS score is closely related to the activation of glial cells in the central nervous system and the increase of serum IL-1 and other cytokines, which induces strong systemic inflammation and local inflammation in the brain, especially in the medulla oblongata, leading to the disorder of autonomic nervous on modulation of systemic inflammation.

WB test showed that LPS-induced sepsis model also led to significant increased expression of TNF-a and IL-6 in MVZ, which suggests neuroinflammation in MVZ. The much higher levels of inflammatory cytokines in serum and MVZ underlie the significantly higher MSS score and mortality in Sepsis Group compared to the Control Group. It can be seen that on the one hand sepsis induces systemic inflammation, on the other hand it activates inflammation in the medulla oblong, leading to an increase in MSS score. Minocycline, which can be used to inhibit microglia and astrocytes to suppress the neuroinflammation³³. Our results showed that minocycline significantly reduced inflammatory cytokines in MVZ and also improved MSS score in septic rats, indicating that neuroinflammation in MVZ is closely related to pathological behavior in rats.

Here, we may wonder whether the pathological behavior of sepsis rats is related to organ dysfunction or hemodynamic abnormalities, such as hypotension³⁴, resulted from the systemic inflammation caused by sepsis. Therefore, we designed a model of MVZ neuroinflammation induced by LPS injection into the fourth ventricle to observe whether or not neuroinflammation in MVZ leads to sepsis-like pathological behavior or pathological changes in rats, such as high mortality and MSS score, changes in HRV indexes and in peripheral serum inflammatory cytokines. To our surprise, by preparing the model of MVZ neuroinflammation, we almost completely replicated these changes induced by sepsis. First, TNF-a and IL-6 in the MVZ of the Neuroinflammation Group were significantly increased compared to the Control Group. Secondly, although screening out the influence from peripheral organs dysfunction, the Neuroinflammation Group still showed significantly higher MSS scores as the Sepsis Group (24.29 ± 2.21 vs 26.43 ± 1.99), thirdly, adding to the mystique, the levels of TNF-a, IL-6 and IL-10 in peripheral serum in the Neuroinflammation Group were also significantly increased, and the survival rate was also significantly decreased compared to the Control Group. Finally, the central anti-inflammatory intervention by minocycline not only significantly reduced the expression of TNF-a and IL-6 in MVZ, but also significantly reduced the MSS score, serum levels of IL-6 and IL-10, and significantly increased the survival rate in the Neuroinflammation Group. The results of the two models (sepsis and MVZ neuroinflammation) and their interventions were nearly identical, therefore, we reasonably think that it is central MVZ’s neuroinflammation that dominates systemic inflammation and immune disorders in sepsis.

According to previous studies¹⁴, MVZ’s neuroinflammation affects systemic inflammation and immunity through CAP. Peripheral vagus nerve stimulation (pVNS) resulted in bradycardia, more expressions of c-Fos and choline acetyltransferase in brainstem nuclei, and a significant reduction of serum TNF-α level in LPS-induced sepsis³⁵, all of which suggested that systemic inflammation was inhibited through CAP’s activation. Other than the cecum punctured and ligated (CLP) sepsis models, intraperitoneal injection of lipopolysaccharide (LPS) induced sepsis models are almost free from influence by minocycline, which acts as an antibiotics, should play an antibacterial role in CLP models. Therefore, minocycline inhibits the systemic inflammation induced by sepsis probably through central pathway. The reasonable explanation is that minocycline inhibits activation of the MVZ’s glial cells to ameliorate MVZ’s neuroinflammation and thus to restore CAP’s modulation, resulting in significant reduction of the level of serum inflammation cytokines. In the Vagus Transection Group, the regulation of CAP was interrupted, thereby the central anti-inflammatory effect of minocycline was canceled, the levels of serum inflammatory cytokines were significantly increased, the MSS score was significantly increased, and the survival rate was decreased compared to the Central Anti-inflammation Group. These results verified that minocycline inhibits the systemic inflammation induced by sepsis through dampened the MVZ’s neuroinflammation.

Then how to quantify the regulatory activity of CAP? Numerous studies have confirmed that neuroinflammation in MVZ can result in nerve injury and autonomic dysfunction, lead to abnormal cardiovascular and respiratory regulation and disorder of autonomic nervous system, such as decreased blood pressure, decreased time domain and frequency domain index of HRV^36,37. The decrease in BP was paralleled by a significant decrease in time domain (SDNN) and frequency domain parameters (total and LF power)³⁸. It suggested that HRV’s indexes may reflect the regulation effect of autonomic nervous system in sepsis. CAP, which acts as a pathway of inflammatory reflex and executes MVZ’s instruction, and originates from vagus nerve, its regulatory activity must be reflected by HRV.

In this study, both sepsis and MVZ neuroinflammation resulted in significant decreases in HRV indexes such as SDNN, RMSSD, LF, HF, and LF/HF, indicating that both sepsis and MVZ neuroinflammation caused a decrease in vagal tone. Central anti-inflammation with minocycline significantly improved these indexes of HRV and significantly inhibited serum inflammatory cytokines in both model rats, while right cervical vagotomy abolished the improvement on HRV and systemic inflammation by central anti-inflammation with minocycline, suggesting that neuroinflammation in MVZ may exert great impaction on the systemic inflammation through CAP.

To confirm the hypothesis that MVZ’s neuroinflammation underlies the systemic inflammatory storm through CAP, we conducted a correlation analysis between any of the two kinds of variables, including the expression of inflammatory cytokines such as TNF-a and IL-6 in MVZ, HRV indexes such as SDNN, RMSSD, LF, HF and serum level of inflammatory cytokines such as TNF-a, IL-6, IL-10. Both the expression of inflammatory cytokines in MVZ and the levels of inflammatory cytokines in serum were strongly negatively correlated to the indexes of HRV, and the expression of inflammatory cytokines in MVZ is strongly positive correlated to the levels of serum inflammatory cytokines. At the same time, we conducted the regression analysis between three pairs of variables including TNF-a expression in MVZ and SDNN, SDNN and serum TNF-a level, and TNF- a expression in MVZ and serum TNF-a level. Results showed that the residuals of the three pairs of variables were normal distribution, and the normal P–P plot of the standard residual also showed that the three pairs of variables were linearly correlated, with a high degree of fitting and an extremely significant correlation. Thereby, we reasonably confirm that the neuroinflammation of MVZ underlies the changes of HRV and systemic inflammatory storm in sepsis. That is to say that the neuroinflammation of MVZ exert a critical influence on the systemic inflammation disorder.

Certainly, this study has limitations, LPS is a key component of the outer membrane of Gram-negative bacteria, which can cause systemic inflammation and neuroinflammation³⁹. The LPS model is considered an imperfect model of sepsis because it is characterized by a high, rapid, and transient elevation of cytokines⁴⁰, which differs from the inflammatory response of human sepsis. In addition, LPS induced inflammatory storm is mainly mediated through the Toll-like receptor 4 (TLR4) dependent inflammatory cascade⁴¹, which does not accurately reflect the multiple cytokines mediated inflammatory storm in human sepsis. However, this study required central intervention with minocycline, which is an antibiotic and certainly brings about disinfection effect in CLP model sepsis, therefore, CLP models don’t fit this study. In addition, the hemodynamic state and organ dysfunction of the rats were not monitored and observed in this study, so the related effects of MVZ neuroinflammation model on MSS, HRV and cytokines could not be evaluated objectively. At last, the neuroinflammation in the hypothalamus and cortex was not explored in this study, which may also affect HRV and systemic inflammation in the model rats.

Conclusion

This study conducted an investigation of the relationship between MVZ’s neuroinflammation and systemic inflammation, not only from the perspective that peripheral sepsis can induce MVZ’s neuroinflammation, but also from the perspective that MVZ’s neuroinflammation can impact on the peripheral systemic inflammation. It confirmed that sepsis-induced MVZ neuroinflammation exert a powerful impaction on the systemic inflammation through CAP, and central anti-inflammation effectively dampened the systemic inflammation which may become a valuable treatment for early inflammatory storm in sepsis. The time domain and frequency domain indexes of HRV not only can reflect the regulatory effect of CAP, but also can reflect the degree of inflammation and injuries of MVZ, which should be studied further for the clinic translation.

Author contributions

Xian Liu and Cheng Zhang equally contributed to this research, they performed the experiments, analyzed the data. Hongbing Li was responsible for the conception and design of the study and manuscript writing. Cheng Zhang processed the experimental data. Xian Liu assisted Cheng Zhang to perform the experiment and collect the data. All authors agree to publish this manuscript.

Funding

This study was supported by the Guizhou Provincial Science and Technology Foundation, NO: Key Projects [2023] 001; Science and Technology Fund of Guizhou Provincial Health Commission, No: gzwkj2021-001.

Data and materials availability

The data and materials will not make known to public until the research project supported by the Guizhou Provincial Science and Technology Foundation is checked and accepted. At the same time, we declare that all the data analyzed during the current study are available from the corresponding author on reasonable request. The link email is mrbright789@sina.com.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-67531-7.