MicroRNA 33 Potentially Participates in the Development of Goose Fatty Liver via Its Target Gene CROT


 Background

Previous studies indicate that microRNA33 (miR-33) and its target gene, CROT, are implicated in hepatic lipid metabolism, but it is unclear whether miR-33 participates in the development of goose fatty liver via CROT.
Methods

The expression of miR-33 in goose fatty liver, muscle and fat tissues, as well as the mRNA and protein expression of CROT in goose fatty liver was determined by q-PCR or Western-blot. The targeting regulatory relationship between miR-33 and CROT in goose liver cells was validated by miR-33 overexpression and interference assays. The effects of miR-33 mimic and CROT overexpression on lipid deposition and the expression of downstream genes were determined in goose primary hepatocytes. The treatment of high concentrations of glucose and insulin was performed to determine their regulation on the expression of miR-33 and CROT in goose primary hepatocytes.
Results

Here, data showed that miR-33 expression was significantly increased in the liver, muscle and fat tissues of overfed geese. Consistently, miR-33 mimic promoted lipid deposition in goose primary hepatocytes. Moreover, the regulatory targeting relationship between miR-33 and CROT was validated in goose primary hepatocytes. Consistently, the mRNA and protein expression of CROT were significantly reduced in goose fatty liver. Interestingly, CROT overexpression could induce the expression of fatty acid oxidation associated genes including CRAT, PEX5, EHHADH, CAT and ACOT8 in goose primary hepatocytes, but only the expression of PEX5 was significantly inhibited in goose fatty liver. However, it seemed conflicting that CROT overexpression increased lipid deposition and reduced lipid peroxidation in goose primary hepatocytes. Additionally, high glucose inhibited miR-33 expression and induced CROT expression in goose primary hepatocytes.
Conclusions

These findings suggest that miR-33 potentially participates in the development of goose fatty liver via CROT, and that miR-33/CROT may partially mediate the effect of glucose in goose liver cells.

human NAFLD can progress into steatohepatitis, cirrhosis, even cancer [2], while goose fatty liver maintains simple steatosis without deterioration into other overt pathological symptoms, which suggests goose fatty liver is unique in some aspects [3]. Indeed, recent evidence indicates that goose may have a protective mechanism that prevents occurrence of in ammation in goose fatty liver. For example, most complement genes (e.g., complement C3 and C5, the key genes in complement response) and proin ammatory gene, TNF-α, are inhibited in goose fatty liver vs. normal liver, but fatty acid desaturases and adiponectin receptors are induced [4]. This protective mechanism allows goose fatty liver to be fully recovered under certain conditions. Therefore, goose fatty liver, a physiological fatty liver, can be considered as a unique model for NAFLD study.
Compared to human and rodent NAFLD, the mechanism underlying the development of goose fatty liver is much less understood. Previous studies have shown that microRNAs, a class of noncoding RNA molecules with a length of 21-23 nt, play an important role in the development of goose fatty liver and mammalian NAFLD. For example, miR-29c can target COL3A1, SGK1 and INSIG1 genes by binding to the 3'-UTR of their mRNA sequences, thus inhibit their expression contributing to energy homeostasis and cell growth in goose fatty liver [4]. Another example is miR-128-1, which can inhibit the expression of LDLR and ABCA1 genes and thus regulate lipid metabolism in human NAFLD [5]. Recent studies show that miR-33 has a number of target genes including CROT, HADHB and NPC1, and plays a role in hepatic lipid metabolism via CROT and HADHB genes [6 7]. CROT is a member of carnitine acyltransferase family (which includes CPT1, CPT2, CROT and CRAT) and plays an important role in fatty acid metabolism by catalyzing the reversible transport of fatty acyl groups between coenzyme A (CoA) and carnitine and providing a cyclic pathway to transport the medium long chain acyl CoA from peroxisome to the cytosol and mitochondria [8][9][10]. However, it is uncertain whether miR-33 inhibits fatty acid oxidation and thus promotes the development of goose fatty liver by regulating the expression of CROT gene. This study is aimed to validate the targeting regulatory relationship between miR-33 and CROT in goose liver cells, to clarify the association of miR-33 expression with CROT expression in goose fatty liver, to determine the effects of miR-33 mimic and CROT overexpression on lipid deposition in goose primary hepatocytes, as well as to determine the regulation on the expression of miR-33 and CROT by high concentrations of glucose and insulin. The results may provide a foundation for addressing the mechanism by which miR-33 participates in the development of goose fatty liver via CROT gene.

Animals and sample collection
All animal protocols were approved by the Animal Care and Use Committee of Yangzhou University.A total of 48 male healthy Landes geese (purchased from Yangzhou RuiNong Farm) with similar body weight at 65 days old were randomly divided into a control group and an overfeeding group (24 geese per group). A 5-day-long pre-overfeeding was performed in the overfeeding group before the 19 days of formal overfeeding. The overfeeding procedures and dietary regimes were carried out as previously described [12]. The control group was allowed to feed and water ad libitum. Six geese from each group were sacri ced on the 7th, 14th and 19th days of overfeeding, respectively. The liver, abdominal fat and pectoral muscle samples were collected, snap frozen in liquid nitrogen and stored at -70°C.

Isolation and treatment of goose primary hepatocytes
Goose primary hepatocytes were isolated from Landes goose embryos on the 23rd day of hatching as previously described [11 12]. The isolated primary hepatocytes were cultured and treated with high level of glucose or insulin, respectively, according to the protocols previously described [4]. In brief, the isolated cells were seeded at the density of 1 × 10 6 cells/well and cultured in complete culture media (containing high glucose DMEM culture medium plus 10% fetal bovine serum, 1% Penicillin-Streptomycin solution (100 IU/mL), and 10 µL EGF (20 ng/mL)) overnight, followed by treating the cells for 14 h with fresh complete culture media supplemented with or without 100 mmol/L glucose and 100 nmol/L insulin, respectively. For transfection assays, the cells at 70-90% con uent were transfected with miR-33 mimics (sense: 5'-GUGCAUUGUAGUUGCAUUGC-3'; antisense: 5'-AAUGCAACUACAAUGCACUU-3'), overexpression vector, inhibitor, and their negative controls (sense: 5'-UUCUCCGAACGUGUCACGUTT-3'; antisense: 5'-ACGUGACACGUUCGGAGAATT-3'), respectively, using Lipofectamine 2000 (Cat. No. 11668-027, Invitrogen, USA) according to the manufacturer's instruction. Brie y, for each well of 6-well plate, 4 µL Lipofectamine 2000 was diluted with 96 µL Opti-MEM, followed by dissolving 2 µg mimics, inhibitors, negative controls, or vectors (overexpression vector or empty vector) separately into 100 µL Opti-MEM. After that, the two solutions were mixed and sit for 20 min at room temperature. The mixture was then added to the cultured cells. After 24 h of transfection, the cells were rinsed with phosphate buffer saline (PBS) twice, followed by harvesting the cells with 1 mL TRIzol reagent (Cat. No. 15596026, Life, USA) per well. The transfection of CROT overexpression vector or empty vector followed the same procedures. The miR-33 mimics, overexpression vector, inhibitor, and their controls were provided by GenePharma Co., Ltd. (Shanghai, China). The miR-33 overexpression vector containing the precursor sequence of goose miR-33 was constructed by PCR-based ampli cation of miR-33 precursor sequence with goose genomic DNA and the pair of primers (Pri-miR-33, see Table 1), restriction of the PCR product with HindIII and XhoI enzymes, and cloning into the pcDNA3.1 (+) vector (Cat. No. V79020, Thermo Fisher scienti c, Shanghai, China) according to the previously published methods [12]. CROT overexpression vector was constructed with pcDNA3.1 (+) vector by Thermo Fisher Scienti c Inc. (Cat. No. V79020, Shanghai, China).  [14]. Goose glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was used as an internal control. The sequences of the primers for qPCR analysis were listed in Table 1. For RT-qPCR analysis of miR-33, TaqMan™ MicroRNA Assay Kit (Cat. No. 4427975, Thermo Fisher Scienti c, Shanghai, China) was used according to the manufacturer's instruction. The U6 gene was used as an internal control. The cycle threshold (Ct) was determined with the supplied software. The expression of the target genes was calculated using 2-ΔΔCt [15] and presented as fold change over control.

Protein assay and immunoblot analysis
Protein concentration of each tissue or cell sample was determined as previously described. The protocol for immunoblot analysis was also previously described [16]. The following antibodies were used in this study: CROT antibody (Cat. No. bs-5048R, Beijing Biosynthesis Biotechnology Co., Ltd. Beijing, China) and β-Actin (Cat. No. sc-47778; Santa Cruz Biotechnology, Inc., Santa Cruz, CA).

Statistical analysis
The data were expressed as the means ± SE. SPSS 18.0 (SPSS China, Shanghai, China) was used to perform the Student's t-test for statistical signi cance of differences between control and treatment. P < 0.05 was considered statistically signi cant.

Results
MiR-33 was involved in the formation of goose fatty liver As expected, similar to previous study, on the 19th day of overfeeding, the gain of body weight, the liver weight and the ratio of liver to body weights were much greater in the overfed geese (treatment group) than the normally fed geese (control group) (Tab. S1), and the livers became milk-white in the treatment group while the livers remain normal dark red, suggesting that fatty liver was successfully induced in the overfed geese.
RT-qPCR analysis indicates that the expression of miR-33 was signi cantly induced in goose fatty liver vs. normal liver on the 7th,14th and 19th days of overfeeding (Fig. 1A). This induction was also shown in other lipid metabolism related tissues including abdominal fat and pectoral muscle of the overfed geese on the 19th day of overfeeding (Fig. 1A). These results indicated that miR-33 was involved in the development of goose fatty liver.
To test whether miR-33 promotes lipid accumulation in goose liver cells, goose primary hepatocytes were transfected with miR-33 overexpression vectors or empty vectors (as control). ORO staining assay indicated that the cells overexpressing miR-33 (Fig. 1B) had more lipid deposition than the control (Fig. 1C).

Validation of the targeting regulatory relationship between miR-33 and CROT gene in goose liver cells
Previous study has demonstrated that miR-33 can target CROT gene at two sites of its 3'UTR in CHO cells ( Fig. 2A) [6], but this targeting regulatory relationship has not been validated in goose liver cells yet. To address this, miR-33 mimic, overexpression vector and inhibitor transfection assays were performed. Data showed that miR-33 mimic could signi cantly inhibit the expression of CROT gene in goose primary hepatocytes compared with the hepatocytes transfected with negative control (Fig. 2B). The inhibitory effect of miR-33 on the expression of CROT was also con rmed by transfecting goose hepatocytes with miR-33 overexpression vectors vs. empty vectors (Fig. 2B). In contrast, the expression of CROT gene in goose hepatocytes was induced by miR-33 inhibitor vs. negative control (Fig. 2B). These ndings indicated that CROT expression could be regulated by miR-33 in goose liver cells.

Induction of CROT gene in goose fatty liver vs. normal liver
In consistent with the targeting regulatory relationship between miR-33 and CROT genes in goose liver cells, the expression of CROT gene was signi cantly inhibited in goose fatty liver vs. normal liver on the 7th, 14th and 19th days of overfeeding (Fig. 3A). In particular, the protein level of CROT gene was also signi cantly lower in goose fatty liver vs. normal liver on the 19th day of overfeeding (Fig. 3B and 3C). The results indicated that there was an association between the induction of miR-33 and inhibition of CROT in goose fatty liver vs. normal liver.
Potential function of CROT in the formation of goose fatty liver At present, the function of CROT in goose liver cells is still unclear. As previous study shows that overexpression of CROT in HepG2 cells induces the expression of CRAT gene, a gene with known function in β-oxidation of fatty acids, we speculated that CROT also could regulate the expression of other genes associated with β-oxidation of fatty acids. To validate this, we overexpressed CROT in goose liver cells (Fig. 4A), and indeed, compared to goose primary hepatocytes transfected with empty vector, the expression of CRAT,PEX5 EHHADH, CAT and ACOT8 was signi cantly increased in the cells transfected with CROT overexpression vector (Fig. 4B). The expression of these genes in goose fatty liver vs. normal liver was also determined. Data showed that only the expression of PEX5 gene in goose fatty liver was signi cantly lower than that in normal liver on the 19th day of overfeeding (Fig. 4C).
Moreover, to determine the effect of CROT gene expression on fatty acid oxidation, ORO staining assay (which determines the deposition of neutral lipid such as triacyglycerols) and MDA assay (which determines the lipid peroxidation mainly due to reactive oxygen species generated by mitochondria) were performed in goose liver cells transfected with CROT overexpression vector vs. empty vector. The results showed that the overexpression of CROT increased lipid deposition and reduced lipid peroxidation in the cells ( Fig. 4D and 4E), which indicated that CROT expression was associated with fatty acid oxidation in goose hepatocytes.

Effects of glucose and insulin on the expression of miR-33 and CROT in goose primary hepatocytes
As the development of NAFLD and goose fatty liver is usually accompanied with hyperglycemia and hyperinsulinemia [17 18], goose primary hepatocytes were treated with high levels of glucose (100 mmol/L) and insulin (100 nmol/L), respectively. The data showed that the expression of miR-33 was signi cantly inhibited in goose primary hepatocytes treated with 100 nmol/L glucose, and accordingly, the expression of CROT was signi cantly increased (Fig. 5A). However, 100 nmol/L insulin signi cantly induced the expression of miR-33 in goose primary hepatocytes but has no effect on the expression of CROT (Fig. 5B). The results indicated that the effect of high level of glucose on the expression pattern of miR-33/CROT was consistent with that in goose fatty liver.

Discussion
MicroRNAs, as a class of small noncoding RNA molecules, can bind to the 3'UTRs of the mRNAs of their target genes, thus inhibit translation of the genes and/or promote destabilization and degradation of the mRNAs [19]. As a key posttranscriptional repressor of gene expression, microRNAs are implicated in almost all physiological and pathological processes [20]. For example, miR-33, embedded in the 16th intron of sterol response element binding transcription factor (SREBF) gene [21], is co-transcribed with SREBF and can target a number of genes including CROT, HADHB, NPC1, CPT1A and AMPK, etc. [22]. CROT is required for the transport of medium and long chain acyl CoA molecules from the peroxisome to the cytoplasm and mitochondria. It mainly exists in the peroxisome and mitochondria, and functions as an auxiliary enzyme in fatty acid metabolism [23]. HADHB encodes the beta subunit of mitochondrial trifunctional protein, which catalyzes the last three steps of long-chain fatty acid β-oxidation in mitochondria [25]. Therefore, induction of CROT and HADHB expression may promote fatty acid oxidation, while inhibition of their expression may promote fatty acid accumulation. Indeed, miR-33 knockdown induces the expression of CROT, CPT1A, HADHB and AMPK, and thus promotes fatty acid oxidation [24]. In contrary, overexpression of miR-33 can inhibit fatty acid oxidation in liver cells [22]. These effects of miR-33 may be mediated by regulating the expression of its target genes, including CROT and HADHB.
Although miR-33 is implicated in fatty acid metabolism, it is unclear whether miR-33 participates in the development of goose fatty liver via CROT gene. Goose fatty liver is characterized by deposition of large amount of fat, which sometimes accounts for about 60% of liver weight. The weight of fatty liver can reach 8-10 times that of normal liver in a short period of time (3-4 wk) [25]. These features re ect the marvelous capacity of goose liver in lipid accumulation. Revealing the mechanism underlying this capacity may help improve the production of goose fatty liver (or foie gras), as well as provide new ideas to develop therapeutic approach to NAFLD in human and other domestic animals. Considering the role of miR-33 and CROT genes in lipid metabolism, we speculate that miR-33 and CROT genes contribute to this unique mechanism that goose liver owns. In this study, data showed that the expression of miR-33 in the liver, muscle and fat tissues of the overfed geese was signi cantly upregulated compared with the normally fed control geese. The targeting regulatory relationship between miR-33 and CROT genes was also subsequently validated in goose liver cells. As expected, the mRNA and protein expressions of CROT gene were signi cantly downregulated in goose fatty liver vs. normal liver, which is contrary to the expression of miR-33. These ndings suggest that, although CROT gene expression could be regulated by multiple factors, miR-33 is a major factor contributing to the downregulation of CROT gene in the development of goose fatty liver. As miR-33 is an intronic sequence of the major regulator of lipid metabolism, SREBF, it may enhance the role of SREBF by being co-transcribed with SREBF and suppressing its target genes including CROT. Interestingly, upregulation of miR-33 expression also occurred in muscle and fat tissues of the overfed geese vs. control geese, thus lipid accumulation in these tissues may be partially attributed to the upregulation of miR-33. Consistently, miR-33 mimic promoted lipid deposition in goose primary hepatocytes. However, it seemed con icting that CROT overexpression increased lipid deposition and reduced lipid peroxidation in goose primary hepatocytes.
One explanation is that the function of CROT is to mobilize fatty acids by promoting β-oxidation of medium-and long-chain fatty acids and transport of the product of β-oxidation from peroxisome to the cytosol. Whether CROT promotes or inhibits lipid deposition, it depends on the context where other genes join and decide the fate of fatty acids, i.e., entering mitochondria to be further degraded or forming triacylgycerols in the cytosol. In other words, lipid deposition in the cell is not only determined by CROT, but also is subjected to regulation by multiple proteins and processes, which may be applicable to the regulation of lipid deposition by miR-33 in goose fatty liver. In line with this explanation, lipid peroxidation inhibited by CROT overexpression in goose primary hepatocytes suggested that fatty acid oxidation in mitochondria was somehow suppressed. Lipid peroxidation is mainly caused by mitochondria-derived reactive oxygen species. Recent studies indicate that CROT gene may participate in the regulation of fatty acid composition [27][28], thus it is worthwhile to determine if CROT gene has dual functions (i.e., involving in β-oxidation of fatty acids, and modulating fatty acid composition) in the development of goose fatty liver.
In addition, previous study shows that CROT overexpression in HepG2 cells induces CRAT expression, a gene playing a key role in fatty acid oxidation [26]. Consistently, this study showed that CROT overexpression in goose primary hepatocytes could induce the expression of PEX5,EHHADH CAT and ACOT8 genes in addition to CRAT gene. It is known that all the genes are associated with fatty acid oxidation. PEX5 (peroxisome biogenic factor 5) plays an important role in peroxisome protein input by binding to the C-terminal PTS1 type tripeptide peroxisome targeting signal (SKL type) and thus is necessary for protein assembly of functional peroxisome. EHHADH (enol COA hydratase and 3hydroxyacyl CoA dehydrogenase) is one of the four enzymes involved in peroxisomal β-oxidation pathway as its N-terminal contains enol CoA hydratase activity and C-terminal contains 3-hydroxyacyl CoA dehydrogenase activity [27]. ACOT8 is a peroxisomal thioesterase, which is involved in the oxidation of fatty acids. As peroxisomes play a key role in fatty acid oxidation and the genes are involved in assembly and function of peroxisome, CROT gene may promote fatty acid oxidation via these genes. As upregulation of fatty acid oxidation increases release of reactive oxidative species, it is reasonable that CROT gene overexpression induced the expression of CAT gene, a key antioxidant enzyme in the cell, which provides an explanation why CROT overexpression inhibited lipid peroxidation in goose primary hepatocytes. Similarly, the induction of these genes only indicates mobilization but not degradation of fatty acids as suggested by CROT overexpression assay. Induction of CRAT, PEX5, EHHADH and ACOT8 genes may be due to the increased level of product generated by CROT. It is found that signi cant reduction in PEX5 expression was concomitant with the inhibition of CROT expression in goose fatty liver vs. normal liver. This is probably due to the regulation of other factors on the expression of these genes during the development of goose fatty liver, which needs to be further investigated.
Finally, this study showed that high level of glucose could inhibit miR-33 expression and induce CROT expression in goose primary hepatocytes, whereas high level of insulin could induce miR-33 expression without changing CROT expression. As the development of fatty liver is usually accompanied with hyperglycemia and hyperinsulinemia, the results suggest that hyperglycemia and hyperinsulinemia are involved in the regulation of miR-33 in the development of goose fatty liver, but only hyperglycemia is involved in the expression of CROT gene in goose fatty liver. It is, however, unknown how miR-33 expression is regulated by glucose and insulin. It is likely that glucose and insulin regulate the expression of miR-33 by regulating the expression of SREBP as previous studies have shown that glucose and insulin can regulate the expression of SREBF [ 18 28]. Moreover, other factors may also regulate the expression of miR-33 by regulating the expression of SREBF, such as thyroid hormone, which is known to be able to stimulate the expression of SREBF [29]. These speculations warrant further investigation.
It is noteworthy that, although the regulatory relationship between goose miR-33 and CROT has been published , the relationship is demonstrated only in CHO cells (from Chinese hamster ovary) other than goose liver cells. This study provides some new insights into the relationship between miR-33 and CROT and the functions of the genes in goose liver cells, including con rmation of the reciprocal relationship between miR-33 and CROT in goose fatty liver and primary hepatocytes, the regulation of miR-33 and CROT expression by insulin and glucose, and the induction of lipid deposition by miR-33 mimics in goose primary hepatocytes.

Conclusions
MiR-33 participates in the lipid accumulation in the development of goose fatty liver, potentially via CROT gene, which contributes to the mechanism by which goose fatty liver is quickly developed. This work provides a foundation to further study the role of miR-33 and the related mechanism in the development of goose fatty liver.