Elevated Serum Regulator of Calcineurin 2 is Associated With an Increased Risk of Non-Alcoholic Fatty Liver Disease

Background: The promoting effect of the regulator of calcineurin 2 (RCAN2) in hepatic steatosis has been observed in animal studies. However, the association of RCAN2 with non-alcoholic fatty liver disease (NAFLD) in humans remains unclear. This study aimed to evaluate the expression of RCAN2 in the liver of mice with hepatic steatosis and in the serum of NAFLD patients and to explore the relationship between serum RCAN2 levels and NAFLD. Methods: The mRNA and protein expression of RCAN2 were detected by quantitative real-time PCR (qRT-PCR) and Western blot. NAFLD was diagnosed by abdominal ultrasonography. Circulating RCAN2 levels were measured by ELISA kits. The relationship between serum RCAN2 levels and NAFLD was assessed. Results: qRT-PCR and Western blot analysis showed that compared with the corresponding controls, the mRNA and protein expression of RCAN2 were significantly increased in the liver tissues of db/db and mice on a high-fat diet. Serum RCAN2 levels were markedly elevated in NAFLD patients compared with non-NAFLD subjects. Binary logistic regression analysis showed that serum RCAN2 levels were significantly associated with NAFLD. Receiver operation characteristic (ROC) curve analysis showed that serum RCAN2 might act as a predictive biomarker for NAFLD [area under the curve (AUC) = 0.663, 95% CI = 0.623–0.702], and the serum RCAN2/(AST/ALT) ratio displayed improved predictive accuracy (AUC = 0.816, 95% CI = 0.785–0.846). Conclusion: Elevated serum RCAN2 levels were associated with an increased risk of NAFLD. Serum RCAN2, especially the serum RCAN2/(AST/ALT) ratio, might be a candidate diagnostic marker for NAFLD.


INTRODUCTION
Non-alcoholic fatty liver disease (NAFLD) is defined as the presence of steatosis in more than 5% of the hepatocytes without other clear causes of liver fat accumulation, such as viruses, alcohol, drugs, or genetic factors (Chalasani et al., 2012). NAFLD comprises simple hepatic steatosis (fatty liver) alone and nonalcoholic steatohepatitis (NASH), a more serious process with inflammation, hepatocyte damage, or fibrosis that can eventually progress to liver cirrhosis and hepatocellular carcinoma (HCC) (Chalasani et al., 2018;Parthasarathy et al., 2020). In recent decades, NAFLD has become a major public health concern. The number of NAFLD patients is dramatically increasing worldwide, and the global prevalence of NAFLD has been estimated to be about 25% (Younossi et al., 2016;Negro, 2020). NAFLD has emerged as one of the predominant causes of chronic liver diseases worldwide. Its pathogenesis is complicated, including increased de novo synthesis of liver fatty acids, hepatocellular injury, liver insulin resistance, intestinal dysbiosis, and fibrogenesis (Friedman et al., 2018). However, these mechanisms have not been well elucidated, and there are currently no effective drugs for the treatment of NAFLD or approved biomarkers to assess disease progression. Although some studies have evaluated potential biomarkers, they have been rarely verified and seldom used in clinical practice. Therefore, there is still a great need to identify and validate other effective biomarkers to predict NAFLD.
The regulator of calcineurin 2 (RCAN2) [also termed thyroid hormone-responsive protein ZAKI-4, down syndrome candidate region 1-like 1 (DSCR1L1), or myocyte-enriched calcineurininteracting protein 2 (MCIP2)] (Davies et al., 2007) was initially identified as a thyroid hormone (triiodothyronine, T3)responsive gene in human skin fibroblasts (Miyazaki et al., 1996) and then reported to inhibit calcineurin-dependent transcriptional responses by binding to the catalytic subunit A of calcineurin (Fuentes et al., 2000;Kingsbury and Cunningham, 2000). ZAKI-4 has three transcripts, α (also called RCAN2-3), β1, and β2, in which β1 and β2 encode an identical protein product β (also called RCAN2-1) (Cao et al., 2002). Mouse RCAN2 isoforms are highly homologous with human transcripts. The RCAN2 mouse gene, located on chromosome 17, corresponds exactly to the gene containing the corresponding human homolog on chromosome 6 (Strippoli et al., 2000). The tissue distribution of RCAN2 isoforms in mice is highly consistent with that in humans. RCAN2-3 is expressed only in the brain, while RCAN2-1 is ubiquitously expressed in the brain, heart, muscle, kidney, and liver (Cao et al., 2002;Mizuno et al., 2004). RCAN2 is involved in many pathophysiological processes. In striated muscles, RCAN2 modulates hypertrophic growth and selective programs of gene expression (Rothermel et al., 2000). RCAN2 plays a vital role in the brain development and function by regulating the function of calcineurin (Siddiq et al., 2001). RCAN2 is expressed in endothelial cells and restricts angiogenesis by inhibiting calcineurin activity (Gollogly et al., 2007). RCAN2 is also involved in the development, proliferation, or pro-tumorigenic network of cancers (Niitsu et al., 2016;Hattori et al., 2019;Mammarella et al., 2021). Interestingly, the knockout of RCAN2 in the whole body can significantly reduce the liver weight of high-fat diet-induced obese mice and did not exhibit liver steatosis (Sun et al., 2011;Zhao et al., 2016). This suggests that RCAN2 plays an important role in NAFLD in mice; however, there is still no evidence that RCAN2 is involved in the development of human NAFLD or that it may be a diagnostic marker for NAFLD.
In the present study, we aimed to conduct the first study to evaluate serum RCAN2 levels in NAFLD patients and non-NAFLD subjects, and the association of serum RCAN2 with NAFLD risk.

Animal Models and Diet
Six-week-old male C57BL/6 mice, db/db mice, and heterozygous control db/m mice were purchased from SPF Biotechnology Co., Ltd. (Beijing, China). All mice were housed in the IVC system with a 12-h light/dark cycles and ad libitum access to food and water. After a 7-d acclimation, 7-week-old male C57BL/6 mice were fed a high-fat diet (HFD) for 16 weeks (containing 60.0% fat, 20.6% carbohydrate, and 19.4% protein; TP23300) to induce fatty liver or low-fat diet (LFD; TP23302) as controls. HFD and LFD were purchased from Trophic Animal Feed High-Tech Co., Ltd., Nantong, China. The db/db mice served as another fatty liver model and were fed a normal diet for 6 weeks. All animals were sacrificed under anesthesia and the liver tissues were excised and frozen at −80°C for analyzing the mRNA and protein levels. All animal protocols were approved by the Experimental Animal Ethics Committee of Southwest Medical University (Permission no. 2020827).

Liver Histological Analysis, Quantitative Real-Time PCR, and Western Blot Analyses
The liver was fixed overnight with 10% phosphate-buffered formalin acetate at 4°C and embedded in paraffin. The paraffin sections were mounted on glass slides for hematoxylin and eosin (H&E) staining. The total RNA was extracted from the liver tissues using TRIzol reagent (15596026, Thermo Fisher), and then the total RNA was reverse transcribed to cDNA using the first strand cDNA synthesis kit Toyobo). qRT-PCR of cDNA was performed using SYBR Green PCR Master Mix (208056, Qiagen) in a StepOnePlus Real-Time PCR System (qTOWER 3 G, Germany). The relative mRNA expression levels of the target genes were normalized to GAPDH expression. The primer sequences are shown in Supplementary Table S1. Western blots were used for protein expression, as described previously (Gao et al., 2019). Primary antibodies were anti-RCAN2 (1:3,000, Thermo Fisher Scientific, PA5-112914) and anti-tubulin (1:5,000, Beyotime, China, AF5012). Western blots were visualized by a chemiluminescence system (ChemStudio_815, Germany) and quantified by using the ImageJ software (National Institutes of Health, Bethesda, MD).

Study Participants
All participants were recruited from the physical examination center of The Affiliated Hospital of Southwest Medical University. Fatty liver can be diagnosed by abdominal ultrasonography with two of the following three abnormalities: 1) diffusely increased liver tissue near the field ultrasonic echo; 2) kidney echo weaker than liver; and 3) vascular blurring and the gradual attenuation of the far-field ultrasonic echo (Fan et al., 2011). Fatty liver can be divided into mild (ultrasonic echo attenuation does not exceed onethird of the liver), moderate (ultrasonic echo attenuation exceeds one-third and less than two-thirds of the liver), and severe (ultrasonic echo attenuation exceeds two-thirds of the liver). The subjects were excluded if they had any of the following: liver dysfunction caused by other reasons (e.g., alcohol, drugs, toxins, virus, or genetic factors) other than NAFLD, cancer, kidney dysfunction, cardiovascular or cerebrovascular disease, history of thyroid disease, HIV infection, pregnancy or breastfeeding, systemic corticosteroid treatment, anti-inflammatory therapy, and hypoglycemic, lipid-lowering, or antihypertensive treatment. All human investigations followed the ethical guidelines of the 1964 Declaration of Helsinki and were approved by the Human Research Ethics Committee of the Affiliated Hospital of Southwest Medical University (Permission no. KY2021086). Informed consent was obtained from all participants in this study.

Measurements of Serum RCAN2
Serum RCAN2 levels were measured with a commercial enzymelinked immunosorbent assay (ELISA) kit (Human ELISA kit, AE22621HU; Mouse ELISA Kit, AE22620MO, Abebio, China) according to the manufacturer's protocol. The pilot experiments showed that RCAN2 can be detected in our 10-fold diluted human serum samples and in 100 μL mouse serum samples. Therefore, all human serum samples were diluted 10-fold before detection, while mouse serum samples were tested with 100 μL. The intra-and inter-assay variations were <8% and <12%, respectively.

Statistical Analysis
SPSS 22.0 and GraphPad Prism 8.0 were used for all statistical analysis and graphics. Data were expressed as number (%) for categorical variables or as mean ± standard deviation (SD) for continuous variables. For continuous variables, Student's t-test or the Mann-Whitney U-test was used for the differences between the two groups, and one-way analysis of variance (ANOVA) or the Kruskal-Wallis test was used for the differences among more than two groups. Comparisons of categorical variables were made using the Chi-square test. The correlation between serum RCAN2 and other clinical parameters was determined by partial correlation coefficients. Variables independently associated with serum RCAN2 were analyzed by multiple linear regressions. Binary logistic regression analyses were used to analyze the association between the serum RCAN2 levels and NAFLD risk. The diagnostic value of serum RCAN2 or serum RCAN2/(AST/ALT) ratio for the NAFLD was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC). A p-value < 0.05 was considered statistically significant.

Hepatic RCAN2 Expression Was Up-Regulated in Db/db and HFD-induced Hepatic Steatosis Mice
As shown in the Supplementary Figures S1, S2, we successfully constructed a mouse model of hepatic steatosis. In order to explore the potential role of RCAN2 in the development of fatty liver, we first analyzed its expression in the liver tissues of mice. Our data showed that compared with db/m mice, the mRNA and protein expression levels were significantly increased in the liver tissues of db/db mice ( Figures 1A,B). Elevated RCAN2 mRNA and protein expression was also observed in the liver tissues of HFD-induced fatty liver mice ( Figures 1C,D). We detected the RCAN2 in the serum of NAFLD mice and found a trend of increase but without a statistical difference (Supplementary Figure S3). This may be due to the insufficient number of mice.

General Clinical Characteristics and Serum RCAN2 Concentration in the Study Population
In order to investigate the expression of serum RCAN2 in NAFLD patients, 346 non-NAFLD subjects and 369 NAFLD patients were enrolled in this study. The general characteristics of the study subjects are displayed in Table 1. The average age of all the subjects was 40.52 ± 10.58 years, including 464 males (64.90%) and 251 females (35.10%). Compared with the non-NAFLD group, the NAFLD group had a higher male/female ratio, BW, BMI, WC, HC, WHR, SBP, DBP, FBG, WBC, NEU, HGB, PLT, ALT, AST, TP, ALB, GGT, ALP, UA, CREA, TC, TG, LDL, HCY, and FLI (all p < 0.05) but lower AST/ALT, DBIL, eGFR, and HDL-C (all p < 0.05). No significant differences were found in age, GLO, A/G, TBIL, IBIL, and UREA (all p > 0.05).
In all study participants, the distribution of serum RCAN2 levels was from 2.44-47.73 ng/mL. As shown in Table 1 and Figure 2A, the serum RCAN2 levels were significantly higher in the NAFLD group than those in the non-NAFLD group (11.47 ± 5.42 ng/ml vs. 8.79 ± 2.85 ng/ml, p = 0.000).

Clinical Characteristics and Prevalence of NAFLD by Tertiles of Serum RCAN2 in all Study Subjects
In order to explore the changes of clinical parameters and prevalence of NAFLD under different RCAN2 concentration gradients in all study subjects, serum RCAN2 levels were divided into tertiles. As shown in Table 2 and Figure 2B, GGT, TC, TG, and the prevalence of NAFLD were gradually increased among the lowest tertile 1, median tertile 2, and the highest tertile 3 groups (all p < 0.05). The subjects in the highest tertile 3 had a higher male/female ratio, BW, BMI, WC, WHR, SBP, DBP, FBG, HGB, ALT, AST, TP, ALB, UA, LDL-C, and FLI than the lowest tertile 1 group, and a higher BW, BMI, SBP, and ALB than the median tertile 2 group (all p < 0.05).
Liner regression analysis was performed to explore variables independently associated with serum RCAN2 levels. As shown in Figure 3, all factors enter and stepwise regression analyses showed that group, GGT, and TG were independent factors associated with serum RCAN2 levels.

Diagnostic Value of Serum RCAN2 Levels and RCAN2/(AST/ALT) Ratio for NAFLD Risk
Finally, ROC curve analysis was performed to investigate the predictive value of serum RCAN2 levels for NAFLD. The results revealed that best cutoff value for serum RCAN2 to predict NAFLD was 9.11 ng/ml (AUC = 0.663, 95% CI = 0.623-0.702, p = 0.000, sensitivity = 63.7%, specificity = 61.3%) ( Figure 4A). In order to improve the diagnostic value, the factors (AST/ALT, A/G, DBIL, HDL-C) negatively correlated with serum RCAN2 levels in all study subjects were obtained by partial correlation analysis. Then, the ratios of serum RCAN2 levels to AST/ALT, A/G, DBIL, and HDL-C were calculated. ROC curve analysis was performed again to explore the predictive value of these ratios for NAFLD. As displayed in Figure 4B, the serum RCAN2/(AST/ALT) ratio showed improved predictive accuracy (AUC = 0.816, 95% CI = 0.785-0.846, p = 0.000, sensitivity = 70.2%, specificity = 78.9%).

DISCUSSION
This study was the first to investigate the association between serum RCAN2 levels and NAFLD in humans. The main innovation of the current study is that we first demonstrated that RCAN2 was highly expressed in the liver of mice with fatty liver by qRT-PCR and Western blot, and subsequently confirmed that serum RACN2 levels were also significantly elevated in NAFLD patients. Meanwhile, the results of the present study showed that serum RCAN2 levels were positively correlated with the prevalence of NAFLD. Systemic knockout of RCAN2 (RCAN2 −/− ) has been reported to prevent hepatic steatosis in mice fed a high-fat diet (Sun et al., 2011;Zhao et al., 2016).
Recently, RCAN2 was found to be a target gene of peroxisome  proliferator-activated receptor gamma (PPARγ) (Yuan et al., 2021), which regulates hepatic lipid deposition in mice with diet-induced obesity (Chen et al., 2022). Taken together, RCAN2 may be a pathogenic driver of NAFLD. It has been reported that the global knockout of RCAN2 can significantly reduce the liver weight in obese mice induced by a high-fat diet by reducing the food intake. This is probably the main reason that RCAN2−/−mice did not develop typical features of fatty liver on a high-fat diet (Sun et al., 2011;Zhao et al., 2016). Both RCAN2-3 and RCAN2-1 were knocked out in these studies, and it was mainly the downregulation of RCAN2-3 rather than RCAN2-1 that caused the reduction of food intake. RCAN2-3 was expressed only in the brain, while RCAN2-1 was widely expressed in the heart, muscle, kidney, and liver (Cao et al., 2002;Mizuno et al., 2004). Systemic knockout of RCAN2-3 and RCAN2-1 in the whole organism may obscure the role of RCAN2-1 in specific tissues. In the current study, we found that RCAN2-1 was significantly upregulated in the liver tissues of the genetic and HFD-induced steatosis mouse model. Whether liver specificknockout of RCAN2-1 is involved in the development of NAFLD deserves further study.
RCAN2 was first recognized as a thyroid hormone responsive gene (Miyazaki et al., 1996). In subsequent studies, only RCAN2-3 was regulated by T3 in a human skin fibroblast cell line (Cao et al., 2002). Hypothyroidism significantly decreased the mRNA expression of RCAN2-3 in the brain and had no effect on RCAN2-1, while hyperthyroidism had no influence on RCAN2-1 and RCAN2-3. In the heart, hyperthyroidism upregulated the expression of RCAN2-1, whereas hypothyroidism has no effect on RCAN2-1. However, in the liver, exogenous T3 had no effect on the expression of RCAN2-1 (Mizuno et al., 2004). In animal or cell experiments, T3 was found to reduce hepatocyte lipid storage by promoting fibroblast growth factor 21 (Adams et al., 2010). In the present study, T3 was within the normal range in all the subjects tested for T3 (data not shown). Moreover, partial correlation analysis (age and sex-adjusted) (r = 0.042, p = 0.603) and multiple linear regression analysis= 0.768, 95% CI = −2.261-3.796, p = 0.617) also showed that serum RCAN2 levels were not correlated with T3. Therefore, these results suggested that elevated serum RCAN2 levels in NAFLD patients may not be associated with T3.
In all study subjects, partial correlation analysis showed that serum RCAN2 levels were positively correlated with BW, BMI, WC, WHR, SBP, DBP, FBG, WC, ALT, AST, TP, GLO, GGT, ALP, UA, TC, TG, and LDL-C, and negatively associated with AST/ALT ratio, A/G ratio, DBIL, and HDL-C. In non-NAFLD participants, the partial correlation analysis showed that serum RCAN2 levels were positively associated with TP, ALB, TC, TG, and LDL-C and negatively associated with HC. In NAFLD patients, after age and sex adjustment, serum RCAN2 was still positively correlated with SBP, GLO, ALP, UA, TC, and TG. Based on these results, we found that RCAN2 was significantly positively correlated with TC and TG in all participants, non-NAFLD subjects, and NAFLD patients, and had a closer correlation with TG, especially in all participants (r = 0.475, p = 0.000) and NAFLD patients (r = 0.458, p = 0.000). Additionally, the close relationship between serum RCAN2 and TG was further confirmed by multiple linear regression analysis. It is well known that NAFLD is based on an abnormal increase in the hepatic lipid content (mainly TG), followed by pathological processes such as hepatocellular injury, inflammation, or fibrosis (Donnelly et al., 2005;Friedman et al., 2018). From these findings, we hypothesized that RCAN2 promotes the occurrence of NAFLD by affecting the storage of TG in the liver, but more studies are needed to verify this relationship and clarify the specific mechanism. Given that NASH is a more serious process with inflammation, hepatocyte damage, or fibrosis, it has also been reported that upregulation of RCAN2 promotes inflammation and apoptosis (Luo et al., 2022). Therefore, future studies are necessary to explore the relationship between RCAN2 and fibrosis and inflammatory factors such as C-reactive protein.