NAFLD has a sophisticated pathogenesis, starting with fat accumulation and ending up with cirrhosis, fibrosis and hepatocellular carcinoma. [1] Its pathogenesis is complex and multifactorial, involving various genetic, metabolic, and environmental factors. [2, 3] In recent years, scholars have investigated the role of gut-liver axis in NAFLD pathophysiology. It is widely acknowledged that disruptions in the balance of intestinal microorganisms, known as dysbacteriosis, significantly contribute to the development of non-alcoholic fatty liver disease (NAFLD) and its associated metabolic dysregulations.[4] In our previous worked we showed an elevated level of Trimethylamine N-Oxid (TMAO- a gut-liver metabolite in NAFLD patients.[5] Choline, a precursor of TMAO, is catabolized by certain types of gut bacteria. It undergoes microbial conversion within the gastrointestinal to form trimethylamine (TMA) Subsequently, TMA is enzymatically metabolized to trimethylamine N-oxide (TMAO). In the non-alcoholic fatty liver disease (NAFLD), there is an elevated rate of choline conversion into TMA/TMAO, resulting in a choline deficiency. [6, 7] Choline is an essential nutrient involved in lipid metabolism, cell membrane integrity, and the synthesis of phosphatidylcholine, a major component of very low-density lipoproteins (VLDL). [8] Choline is primarily obtained through dietary sources. It can also is produced endogenously via the hepatic phosphatidylethanolamine N-methyltransferase (PEMT) pathway.[9] Several studies have suggested a potential association between choline deficiency and the development or progression of liver steatosis.[10, 11] Choline deficiency leads to impaired VLDL secretion, resulting in lipid accumulation in the liver. Furthermore, alterations in choline metabolism and genetic polymorphisms in the PEMT gene have been implicated in NAFLD pathogenesis.[12] Specific bacterial strains of gut have been identified as choline consumers, contributing to the depletion of choline. These choline-consuming bacterial strains include Anaerococcus hydrogenalis, Clostridium asparagiforme, Prevotella copri, Clostridium hathewayi, and Proteus penneri.[13, 14]The relationship between serum choline levels, gut microbiota composition, and NAFLD development and progression remains poorly understood. The CutC gene is associated with choline utilization in certain bacteria.[15] Specifically, CutC is part of the microbial pathway responsible for the conversion of choline to trimethylamine (TMA); The process involves several steps, and CutC acts as a choline-TMA lyase, catalyzing the conversion of choline to trimethylamine[16]; Hence, we assumed that in the case of sufficient choline or elevated choline uptake in NAFLD patients, increased Cut C expression may responsible for choline deficiency in patients with NAFLD. This study aimed to investigate the correlation of serum choline levels and the frequency of choline-consuming bacterial strains which abundantly expressed Cut C gene with NAFLD. To address this, we assessed serum choline levels, and evaluated the prevalence of choline-consuming bacterial strains as well as Cut C expression in individuals with NAFLD and healthy individuals.

This study encompassed a selected cohort of 85 individuals diagnosed with non-alcoholic fatty liver disease (NAFLD) and 30 healthy individuals who served as controls. For additional details regarding the characteristics and demographics of this study population, readers are encouraged to refer to a related study conducted by our research group.[5] Briefly, the diagnosis of NAFLD in patients was conducted by skilled subspecialist gastroenterologists, utilizing liver ultrasound or Fibroscan in conjunction with laboratory tests, ensuring a definitive diagnosis of fatty liver disease grade 3. The control group consisted of 30 healthy individuals, matched to the patient group in terms of age, sex, and demographic characteristics, and selected from the general population to ensure unbiased comparisons. All participants were from the Kurdish population of Kurdistan province, Sanandaj, Iran. Inclusion and exclusion criteria were carefully applied: exclusion criteria included the presence of cardiovascular diseases (such as atherosclerosis or history of heart attack or stroke), diabetes, kidney problems, multiple sclerosis, HIV infection, and a long history of alcohol consumption. Individuals who had taken antibiotics, probiotics, or prebiotics in the two months prior to the study were also excluded. Detailed dietary habits were assessed to ensure they did not influence the study outcomes, particularly regarding choline levels, and the study protocol included a thorough review of participants' medication histories to exclude those on multivitamins or other supplements that might affect liver function or choline levels. All participants entered the study with full knowledge of the research project and provided informed written consent. Ethical standards were strictly observed, and the study protocol was approved by the Medical Ethics Committee of Kurdistan University of Medical Sciences. 1.1. Laboratory Tests: Due to the close association of laboratory tests in this study with the pathophysiology of non-alcoholic fatty liver disease, the measurement of these tests was examined. All laboratory tests were performed in a single laboratory. Laboratory tests related to liver function or to non-alcoholic fatty liver disease including alanine aminotransferase (ALT), aspartate aminotransferase (AST), cholesterol (Chol), triglycerides (TG), low-density lipoprotein (LDL), and high-density lipoprotein (HDL) were measured using standard laboratory methods in an automated chemistry analyzer machine. 1.2. Assays for choline-consuming bacterial strains 1.2.1. DNA Extraction DNA was extracted from the fresh stool samples of both patients and healthy individuals using the FavorPrep™ Stool DNA Isolation Mini Kit (FASTI 001-1, Favorgen, TAIWAN). The protocol was followed according to manufacturer's instructions. In brief, approximately 200 milligrams of stool sample were transferred into 1.5ml bead-beating tubes. Subsequently, 300 µL of 1SDE solution were added to each tube. The samples underwent bead-beating at 4000 rpm for 30 seconds, followed by 30 seconds of incubation on ice, and this cycle was repeated five times. Afterward, 20 µL of Proteinase K enzyme were introduced to facilitate the hydrolysis of peptide bonds. Vigorous vortex for 5 minutes ensured proper mixing of the enzyme, followed by incubation at 60 oC for 20 minutes with intermittent vortex. another 5-minute incubation at 95 oC followed, and the tube was then placed on ice for 30 seconds. 100 µL of 2SDE solution was then added and mixed vigorously for 5 minutes, followed by incubation on ice. Subsequently, the tube was centrifuged at 16,000 g for 7 minutes, and the supernatant containing the liquid phase was carefully transferred to fresh 1.5ml tubes without disturbing the pellet. Then, 250 µL of 3SDE solution was added to the transferred supernatant and vortexed thoroughly. After a 2-minute incubation at room temperature, the tube was centrifuged at 16,000 g for 3 minutes. The remaining liquid was transferred to a fresh 1.5ml tube, and 250 microliters of 4SDE solution, along with 250 µL of 100% ethanol, was added. The entire contents were vortexed for 1 minute. Finally, the tube was centrifuged at 16,000 g for 2 minutes at 4 oC, and the supernatant was discarded. the leaving material is the DNA product. 1.2.2. PCR Detection of choline-consuming bacterial strains For this purpose, the extracted DNAs were applied in a simple polymerase chain reaction (PCR) assay using a Taq DNA Polymerase Master Mix (180301, Amplyqon, Denmark) and the strain specific primers for Anaerococcus hydrogenalis, Clostridium asparagiforme, Clostridium hathewayi, Providencia rettgeri, and Proteus penneri in a thermal cycler machine (FlexCycler, BYQ602101D-1546, Germany). The characteristic of the primers used for Cut. C gene amplification is given in Table 1. The amplified fragments of targeted genes were confirmed using 2% agarose gel electrophoresis (Supplementary file). Finally, the frequency of bacteria with Cut.C gene in two groups was expressed as a percentage of all samples. 1.2.3. Quantitative PCR (qPCR) for total Cut. C assay To evaluate and compare the total expression of Cut. C in patients and healthy controls, a qPCR assay was performed on the extracted DNAs using a Mix qPCR-HS Blue SYBR kit (SinaColon, IRAN) and the degenerate primers (Table 2) in a Corbett Rotor-Gene 6000 instrument. The Cut. C expression was normalized against 16s rRNA as the internal reference gene. The quality of the qPCR assay was also checked by appropriate melting curves and gel electrophoresis of qPCR products (Supplementary file). Finally, LinRegPCR ver. 2013 software was used to calculate the relative level of the Cut. C. 1.3. FL-HPLC determination of Choline content Measurement: In both patients and healthy individuals, the circulatory level of choline was measured after a derivatization reaction of the hydroxyl group of choline with 1-naphthyl isocyanate (Analytical grade, 170518-5G, sigma-aldrich) to form a stable cationic compound that can be detected by high-performance liquid chromatography device connected to fluorescence detector (FL-HPLC). To prepare the calibration curve, the several serial concentrations of choline (Analytical standard, C7017, sigma-aldrich) in the range of 0.05-50 μM were prepared and derivatizated. They were finally injected into the HPLC system separately (Figure.3). Using different known standard of choline, the standard of choline was drowned (3). The results of this variable were also reported as SD ± Mean. 1.3.1. Instrumentation: An HPLC system with the following components was used (KNAUER, Germany): HPLC pump K-1001 (KNAUER, Germany), Degasser 5000 (KNAUER, Germany), FL-Detector XL A-10RF (KNAUER, Ex: 220 nm, Em: 350 nm), dynamic mixing chamber, column thermostat 5-85 0C , temperature 25 0C, mobile phase (methanol/ buffer 15:85), isocratic run for 15 minutes, flow rate 1 ml/min, column (SXC- A100, 250mm, 4.6 mm, 5μm, Italy), and a personal computer running software chrome gate. The mobile phase was prepared by mixing 10 μL of trimethylamine hydroxide (TMAH) solution (1 M) with 20 mL of aqueous glycolic acid solution. Then, 120 mL of HPLC-grade water was added, followed by the addition of acetonitrile to a final volume of 1 liter. The mixture was thoroughly mixed to remove impurities and excess substances using an ultrasonic filter. 1.3.2. Sample/Standard Preparation and Derivatization for Choline Assay First, 1 mL of acetonitrile was mixed with 20 μL of serum sample or choline standard solution (for standard curve preparation) in a 1.5 mL microcentrifuge tube. The mixture was vortexed, and then 80 mg of magnesium oxide was added. Next, 20 μL of 1-naphthyl isothiocyanate solution (131 μM) was added, and the tube was shaken for 15 minutes at room temperature. After centrifugation at 13,000 rpm for 5 minutes, 400 μL of the supernatant was transferred to a fresh microcentrifuge tube and evaporated under a gentle stream of nitrogen gas. Finally, the residue was reconstituted in 200 μL of mobile phase, vortexed, and then filtered through a 0.22 μm syringe filter prior to injection into the HPLC system. Finally, 100μl of this solution was injected into the HPLC system. For quality assurance, we used to several standard concentrations of choline comprise of 1.25, 2.5, 5, 10 and 50 µM (shown in Fig.1). 1.4. Statistical Analysis: Statistical analysis was performed using R version 4.3.0 (2023-04-21 ucrt). For analyzing bacteria’ frequency in library(openxlsx), chi-square Test was used. For relative gene expression of Cut C gene in library(dplyr), One sample Wilcoxon test applied as well as Pearson correlation was used to correlation between Cut C expression, choline level and bacteria frequency. ROC curve analysis was used to calculated sensitivity and specificity. All graphs were drawn using library(ggplot2). A p-value less than 0.05 was considered statistically significant.

Based on the obtained results, the mean levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) enzymes were significantly higher in the patient group compared to the control group. However, no significant difference was found in the plasma LDL levels between the two groups in the evaluation of the lipid profile. There was a significant difference in the plasma HDL levels between the two groups. Moreover, the plasma cholesterol and triglyceride levels were clearly higher in the patient group compared to the healthy individuals (Table 3). 1.1. Results Related to the Analysis of Intestinal Microbiota in the Study Groups The presence or absence of the studied bacteria in the study groups was reported as positive (present) or negative (absent) using specific primers targeting gene. The frequency of the targeted bacteria's presence and the difference between the two groups were examined using Chi-square test between the patient and healthy groups. Interestingly, Anarococcus hydrogenalis DSM7454 (p = 0.003), Clostridium asparagiform DSM15981 (p = 0.007) were significantly abundant in patients in comparison with healthy individuals. While others, Providencia rettgeri DSM 1131 (p = 0.637), Clostridium hathewayi DSM 13749 (p = 0.296), and Proteus penneri DSM 35198 (p = 0.130) did not show significant differences between study groups (shown in Fig.2). 1.2. Results Related to Quantitative Measurement of the cutC Gene The quantitative measurement of the cutC gene in the patient and healthy groups was performed using the real-time PCR (qPCR) method. The average expression of the cutC gene in patients was higher compared to healthy individuals. Statistical analysis indicated that there was a significant difference between patients and healthy individuals (p <0.0001) (shown in Fig 3 A). Receiver Operating Characteristic (ROC) curve analysis was conducted to evaluate the discriminatory power of Cut C expression in distinguishing patients with non-alcoholic fatty liver disease (NAFLD) from healthy individuals. The results revealed that Cut C expression exhibited a high sensitivity of 76.63% and specificity of 75.23% (P value < 0.0001, Fig 3 B). These findings underscore the potential utility of Cut C expression as a biomarker for identifying individuals with NAFLD, highlighting its diagnostic significance in clinical settings. 1.3. Results related to Choline: Our study showed significant differences in plasma choline levels between patients and healthy individuals) p<0.0001, shown in Fig.4 A). The mean of plasma level of choline was lower in patients compared to healthy individuals. ROC curve analysis was performed to assess the discriminatory capacity of choline levels in distinguishing patients with non-alcoholic fatty liver disease (NAFLD) from healthy individuals. The results revealed that choline levels exhibited a sensitivity of 82.33% and a specificity of 65.54% (P value < 0.0001, Fig 4 B). These findings suggest that choline levels hold promise as a potential biomarker for detecting NAFLD, indicating its potential value in clinical diagnosis and management 1.4. Correlation Analysis: Correlation analysis was used to determine the associations among choline levels, Cut C gene expression, and various characteristics of the study population, including age, sex, BMI, and dietary habits. Specifically, Pearson correlation coefficients were calculated to assess the strength and direction of relationships between choline levels Cut C expression, bacterial frequency and dietary sources and potential influencing factors. 1.4.1. Correlation of Choline level and Cut c gene Expression with characteristics of study population In this study, the dietary habits of the population, including the consumption of red meat, chicken meat, fish meat, and eggs as sources of choline, alongside age, sex and BMI were examined for their correlation with choline levels. Notably, age exhibited a negative correlation with serum level of choline (P value <0.01). There was no significant correlation between choline level and its dietary sources (P value >0.05). (Figure 5). 1.4.2. Investigating the Relationship Between Choline Levels, Cut C Expression, and Bacterial Frequency Given the role of the Cut C gene in choline consumption within the studied bacteria, correlation analysis was undertaken to explore the associations among choline levels, Cut C expression, and bacterial frequency. Statistical analysis revealed choline level negatively had a significant correlation with Anarococcus Hydrogenalis and Clostridium asparagiforme which were significantly abundant in patients. Interestingly, Cut C expression also showed a significant negative correlation with choline level, indicating higher expression of Cut C in Anarococcus Hydrogenalis and Clostridium asparagiforme result in choline shortage in patients. Furthermore, Cut C expression positively was correlated with Anarococcus Hydrogenalis and Clostridium asparagiforme and Providencia Rettgeri (P value <0.05), adding further evidence this gene is highly expressed in these bacterial species which might contribute higher choline consumption by these bacteria (Figure 6).

In conclusion, we propose an interaction between choline deficiency and specific gut bacteria, notably Enterococcus hydrolyticus and Clostridium asparagiforme, in the development of NAFLD. The heightened presence of these choline-utilizing bacteria among NAFLD patients could result in a depletion of choline levels, further exacerbated by the increased activity of the Cut C gene responsible for choline conversion to trimethylamine (TMA) in studied bacterial strain. Our study underscores the pivotal roles played by the disturbances in the gut microbiome in the initiation and advancement of NAFLD. Moreover, analysis of choline levels and Cut C expression offers promising avenues for early detection and diagnosis of the disease. By leveraging the sensitivity of choline levels and the specificity of Cut C expression as biomarkers, healthcare providers can improve the accuracy of NAFLD screening and facilitate timely interventions for affected individuals.