Agmatine improves liver function, balance performance, and neuronal damage in a hepatic encephalopathy induced by bile duct ligation

Abstract Introduction In the current study, we investigate whether oral administration of agmatine (AGM) could effectively reduce motor and cognitive deficits induced by bile duct ligation (BDL) in an animal model of hepatic encephalopathy (HE) through neuroprotective mechanisms. Methods The Wistar rats were divided into four groups: sham, BDL, BDL+ 40 mg/kg AGM, and BDL+ 80 mg/kg AGM. The BDL rats were treated with AGM from 2 weeks after the surgery for 4 consecutive weeks. The open field, rotarod, and wire grip tests were used to assess motor function and muscle strength. The novel object recognition test (NOR) was performed to evaluate learning and memory. Finally, blood samples were collected for the analysis of the liver markers, the animals were sacrificed, and brain tissues were removed; the CA1 regions of the hippocampus and cerebellum were processed to identify apoptosis and neuronal damage rate using caspase‐3 immunocytochemistry and Nissl staining. Results The serological assay results showed that BDL severely impaired the function of the liver. Based on histochemical findings, BDL increased the neuronal damage in CA1 and Purkinje cells, whereas apoptosis was significantly observed only in the cerebellum. AGM treatment prevented the increase of serum liver enzymes, balance deficits, and neuronal damage in the brain areas. Apoptosis partially decreased by AGM, and there were no differences in the performance of animals in different groups in the NOR. Conclusions The study suggests AGM as a potential treatment candidate for HE because of its neuroprotective properties and/or its direct effects on liver function.


INTRODUCTION
Hepatic encephalopathy (HE) is a cerebral dysfunction syndrome secondary to acute or chronic liver diseases. The syndrome is characterized by a wide spectrum of neuropsychiatric manifestations ranging from mild subclinical alterations to coma (Dhanda et al., 2013). Various grades of mental and motor disturbances, including personality changes, poor concentration, attention, and judgment, learning and memory deficits, anxiety, fatigue, slow movements, confusion, and lethargy, were associated with HE. These changes affect the quality of life in patients and decrease their ability to perform work and daily life activities (Méndez et al., 2011;Monfort et al., 2009;Munoz, 2008).
Despite considerable experimental efforts, the pathophysiology of HE is still largely unknown. Hyperammonemia is the main cause of HE, which disrupts the blood-brain barrier and exerts a prominent toxic effect on brain tissue (Braissant et al., 2013;Jiang et al., 2013). In the liver, ammonia is converted to urea through the urea cycle, which is water soluble and can be excreted by the kidneys. The first step of the urea cycle is converting NH4 + to carbamoyl phosphate by mitochondrial carbamoyl phosphate synthetase I (CPS-1). N-acetyl glutamate (NAG) is an essential allosteric cofactor of CPS-1, and its availability increases ureagenesis. Previous studies showed that high-protein diets and arginine could increase NAG synthesis and promote the urea cycle.
It was even reported that agmatine (AGM), a metabolite of arginine, can increase NAG more than arginine. AGM was introduced as an effective therapeutic adjunct in some urea cycle disorders and toxic hyperammonemia (Dhanda & Sandhir, 2018;Nissim et al., 2002;Savlan et al., 2014).
Oxidative stress and inflammation are other causes of developing liver disease and HE; antioxidants and anti-inflammatories have been extensively suggested as treatment strategies. The anti-inflammatory, antioxidative, and neuroprotective effects of AGM have been approved frequently. A recent study has reported the therapeutic effects of AGM on hepatic and renal injury in an experimental model of obstructive jaundice and has related these positive effects to the antioxidant properties of AGM (Ommati et al., 2020).
NMDA-receptor (N-methyl-D-aspartate) activation contributes to the pathogenesis of HE and has been shown that NMDA-receptor antagonists are of potential therapeutic value in the treatment of this syndrome. AGM is an endogenous NMDA antagonist, and a neurotransmitter/neuromodulator with some cognitive protective effects, such as anxiolytic, antidepressant, and memory-enhancing activity (Aglawe et al., 2021;Ahmadi et al., 2015;Kim et al., 2016;Nissim et al., 2002;Su et al., 2009;Taksande et al., 2014).
Based on the mentioned hypotheses in the etiology of HE and the positive functions of AGM in this approach, the present study was designed to explore the effects of AGM in bile duct ligated (BDL) rats, an animal model of HE. For these purposes, the motor and cognitive behaviors of the experimental animals were assessed along with providing histopathological evidence.

Bile duct ligation (BDL) surgery
Animals were anesthetized by intraperitoneal administration of ketamine (100 mg/kg) and xylazine (20 mg/kg). The abdominal wall was incised in the middle line, and common bile duct was seen and ligated with a nonabsorbable suture by two ligatures. The first ligature was tied below the junction of the hepatic ducts, and the second was above the entrance of the pancreatic ducts. The common bile duct was cut completely between the ligatures (BDL rats). In sham operation group, animals underwent operation without BDL ligation and cutting (Ahmadi et al., 2015). The percentage of mortality before the start of treatment (2 weeks after BDL surgery) was about 50%.

Drug treatments
Two weeks after surgery, the BDL animals were randomized into three groups that were gavaged with two different doses of AGM ( Each rat was hung vertically with its front legs on a horizontal steel wire with a length of 80 cm and a diameter of 7 mm, and latency to fall was recorded. The test was done for three trials for each animal, and the average time was reported; the intertrial rest interval was 10 min (Haghani et al., 2013).

Novel object recognition test
A NOR test, which assesses the non-rewarded, nonspatial memory, was chosen to evaluate cognitive impairment induced by BDL and the possible protective effects of AGM. The test was performed in a plexiglass box (40 × 60 × 60 cm 3 ). A camera connected to a computer equipped with a video tracking system recorded the animal's behavior. The task was performed during three stages: habituation, training, and retention. The day before the test, each animal was placed in the empty NOR box without objects for habituation and allowed to freely explore it for 5 min. In the training stage, two objects (A: green and B: black) with the same volume and size were placed in the corners of the NOR box at a certain distance from the wall, and the rats were allowed to explore items for 5 min. After 45 min, in the retention stage, one of the objects (B: black) was replaced by a new object (C: red). The animal's exploratory behaviors, including pointing the nose to an object at a distance of 2 cm or touching it with the nose or forehead, were recorded for 3 min. The objects and box were cleaned with alcohol between the experiments. A preference index (in the retention stage) was calculated as the ratio between the time spent exploring the novel object (object C) to the total time spent in exploring both the objects (objects A and C) and similar for object B in training phase was calculated (Faatehi et al., 2019).

Liver function tests
After the behavioral test, the animals were anesthetized, and blood was drawn from the apex of the heart. Blood samples were centrifuged, and plasma was separated and stored frozen at −20 • C until assayed. Liver enzymes, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total and direct bilirubin, are measured to determine the hepatic injury. We also calculated AST to ALT ratio (AST/ALT) because AST and ALT are not in parallel during liver damage and clinical studies suggest AST/ALT as a useful predictor for liver injury. Standard kits (Pars Azmun, Tehran, Iran) and an autoanalyzer Selectra XL (Vital Science, the Netherlands) were used to determine liver function markers (Botros & Sikaris, 2013).

Histological studies
Immediately after collecting the blood samples, transcranial perfusion was conducted, first with isotonic saline followed by 4% paraformalde-

Assessment of jaundice following BDL
The general appearances of rats (after 6 weeks of the surgery) and their livers (the end of the study) were assessed and shown in Figure 2. The BDL rats had yellow liver, ears, sclera, skin, paws, and tails, as well as an enlarged abdomen.

Assessment of liver function tests
As demonstrated in Figure 3, the plasma level of liver enzymes and bilirubin (total and direct) showed a significant increase in the BDL group compared with the sham group (ALP: p < .001, Figure 3a, AST: p < .001, Figure 3b; ALT: p < .001, Figure 3c; total and direct bilirubin: p < .001, Figure 3d,e).
Treating with 40 mg/kg AGM significantly decreased ALP (p < .001, the BDL rats (p < .001, Figure 3f). Elevated AST/ALT ratio is associated with increased risk of developing fibrosis.

Assessment of locomotor activity
Impairments of locomotor activity and muscle strength are associated with several liver diseases and have been reported in BDL model of cirrhotic and HE. In this study, we evaluated these parameters by using open field, rotarod, and wire grip test and obtained the following results.

Open field
In the open-field test, grooming behavior in BDL rats was not change (F (3,28) = 4.01, p = .32, Figure 4a), whereas vertical activity, including climbing and rearing in BDL rats, was less than sham-operated rats

F I G U R E 2
Yellow ears, skin, paws, tail, and liver, plus enlarged abdomen in bile duct ligation (BDL) rats compared to sham rats.

F I G U R E 4
Locomotor activity following bile duct ligation (BDL) and agmatine treatment (40 and 80 mg/kg) in male rats. Grooming and vertical activity (a and b); time spent in the central (c) and peripheral (d) areas; total distance moved (e), and velocity (f). ***p < .001. Compared with sham group, ## p < .01, ### p < .001 compared with BDL. Data are shown as box plots (Whiskers: min to max), (n = 8).

Rotarod
One-way ANOVA analysis showed a statistically significant difference between the groups in the mean latency of three trials on accelerating rotarod (time on rod: F (3,28) = 11.75, p < .001, Figure 5a), and Tukey's post hoc test showed a significant decrease in this parameter following BDL-induced HE compared with sham animals (p < .001, Repeated measure ANOVA showed that only sham animals were able to learn the rotarod task and indicated a significant improvement in task acquisition (the first trial compared to the second [p < .05] and third trial [0.001]; F (3,21) = 7.8, p = .001 for sham groups, Figure 5b).
There was a significant difference in the performance of different

Wire grip
The muscle strength and balance in the wire grip test were significant among different groups (Kruskal-Wallis statistic = 14.19, p = .002).
Our results demonstrate a significant decrease in latency to fall in BDL rats (p = .01, BDL vs. sham, Figure 6). There was no significant difference among the BDL+ 40 mg/kg AGM animals and sham rats (p > .05).

Assessment of learning and memory
Learning and memory were assessed using the NOR test. Animals had no preference for either object in training sessions. Three animals in the BDL group did not move at all and were removed from the analysis.

Nissl staining
To determine the effects of BDL and AGM on the neuronal damage in the CA1 region of hippocampus and cerebellum, Nissl staining F I G U R E 6 Mean latency to fall using wire grip test following bile duct ligation (BDL) and agmatine treatment (40 and 80 mg/kg) in male rats. **p < 0.01 versus sham, data are shown as box plots (Whiskers: min to max), (n = 8).

F I G U R E 8
The percentages of damaged hippocampal CA1 (a) and cerebellar Purkinje neurons per field (b) were observed using the Nissl staining. **p <0.01, ***p < 0.001 versus sham; # p < .05, ##p <0.01 versus bile duct ligation (BDL). Data are shown as box plots (Whiskers: min to max). Four animals/group, two slides/rat, and four fields/slide. In the CA1 region of the hippocampus, caspase-3 positive cells were not seen in most subjects except only a few BDL rats ( Figure 12).

DISCUSSION
Several studies have revealed that AGM, as a metabolite of arginine, can offer protection against nicotine-induced hepatic damage. There is a study that introduces AGM as an effective drug for the treatment and prevention of liver ischemia reperfusion injury (Han et al., 2020). In the mentioned studies, reduced oxidative stress, nitric oxide, and inflammatory mediators were the proposed mechanism of action of AGM (Kim et al., 2016;Nissim et al., 2002).
It is important to note that arginine therapy in HE was recommended because of it stimulates ureagenesis, but it seems that arginine stimulates ureagenesis through metabolites. AGM, the decarboxylation product of arginine, irritates the hepatic content of NAG and ureagenesis and with respect to safety and efficacy in the treatment of cognitive and motor dysfunction can be a target for treating the HE (Nissim et al., 2002). Today, it is known as a popular dietary supplement to increase muscle mass in athletes and also for therapeutic purposes in combination medicine (Molderings & Haenisch, 2012).
In the current study, we did not observe any difference between BDL and sham-operated rats in the NOR test. In consist with the results of our study, the results of the study by Cho et al. (2020) and Leke et al. (2013) showed that chronic cholestasis-induced by BDL cannot change the performance of rodents on NOR test. NOR test is relatively less stressful, and animals do not receive an aversive stimulus. In the other words, it requires no external motivation, reward, or punishment and rodents explore the novel object as their innate desire for novelty.
Furthermore, locomotor deficits induced by BDL may also be an important confounding factor in the NOR test. Our suggestion is that the test should not be used in future studies of HE or chronic BDL ().
Neuroprotection is another beneficial feature of AGM that makes it a therapeutic target in neurodegenerative diseases. The neuroprotective effects have been reported in numerous studies in both in vitro and vivo experimental models against the excitotoxicity induced by glutamate, glucocorticoids, amyloid β, and also in the animal model of spinal cord injury and traumatic brain injury (Neis et al., 2017;Zhu et al., 2008). Caspase-3 results revealed that CA1 neurons are more resistant to BDL-induced apoptosis than Purkinje cells; only a few apoptotic neurons were observed in this area. In the Nissl stained sections, the morphology of neurons in the hippocampal CA1 region and cerebellum was significantly different in BDL rats. AGM restored the CA1 neurons to normal at 40 mg/kg while recovered Purkinje cells at 80 mg/kg.
The cerebellum histological results confirmed the motor coordination outcomes on the rotarod test.
There were some important methodological limitations in our study to estimate AGM levels in the brain and further study will be needed to determine whether the neuroprotective effects of AGM are related to increased AGM levels in the brain or other direct and indirect causes.
It is worth noting that, in our study, the AGM dosages were selected based on the existing literature of animal studies examining the influence of AGM following cognitive or liver deficits. In most of these studies, AGM (at dosages 20-100 mg/kg) was administered intraperitoneally or orally to adult rats. Moreover, previous preclinical studies showed that daily oral delivery of AGM is safe and increases brain and plasma AGM levels (Aglawe et al., 2021;Barua et al., 2019;Bergin et al., 2019;Miski et al., 2021).
Taken together, the present study suggests AGM as a potential therapeutic adjunct for HE. The beneficial effects of AGM used in this study on HE rats were most likely due to its beneficial effects on enhancing liver functions. The authors recommend future research examining the effects of AGM on NAG and comparing the effect of AGM versus arginine in HE animal models.

ACKNOWLEDGMENTS
This research article is a part of the first author's thesis. We would like to thank the Kerman University of Medical Sciences, Kerman, Iran, for supporting this research article financially.

CONFLICT OF INTEREST STATEMENT
The authors report that there are no conflict of interest to declare.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.