The nutrigenomic metabolite L‑carnitine directly modulates the activity and expression of nuclear receptors in adipocytes, liver and muscle cells CURRENT

Background: L‑carnitine is an indispensable metabolite in eukaryotic cells, which facilitates transport of long‑chain fatty acids into the mitochondrial matrix for subsequent β-oxidation and helps to safeguard the acetyl-CoA level. Additionally, L‑carnitine has been proven to exert a nutrigenomic effect, modulating the expression of numerous target genes. However, the diverging time-dependent effects of short-term and extended L‑carnitine supplementation have not been investigated in more detail yet, especially in the interplay of adipocytes, liver and muscle cells. A cell culture model with conditions of L‑carnitine deficiency and supplementation for these cell types was established to investigate the effects of L‑carnitine on key nuclear receptors and their pathways. Results: L‑carnitine deficiency as well as L‑carnitine supplementation to hepatocytes modulated protein activity of multiple nuclear receptor pathways (RAR, RXR, VDR, PPAR, HNF4, ER, LXR). On the transcriptional level, short‑term L‑carnitine supplementation initially exerted an inhibitory effect on the steady state mRNA levels of PPAR‑α, PPAR‑δ, PPAR-γ, RAR‑β , LXR‑α and RXR‑α in adipocytes, liver and muscle cells. However, extended L‑carnitine supplementation for 24 and 48 hours led to a significant upregulation of PPAR‑α and PPAR‑δ , being key regulators of lipid catabolism, thereby promoting lipolysis and β-oxidation. In addition, significant differences in transcriptional modulation were found between adipocytes, liver and muscle cells. Extended L‑carnitine administration to hepatocytes also modulated mRNA expression levels of nuclear receptor target genes CYP2R1 , ALDH1A1 , HSD11B2 , OGT and HMGCR. Conclusions: These findings show a clear nutrigenomic effect of L‑carnitine on the protein activity and expression levels of selected nuclear receptors in different tissues, promoting lipolytic gene expression as well as decreasing transcription of adipogenic and insulinresistance linked genes. Therefore L‑carnitine supplementation transcription levels as well as transcription levels of key effector genes. The results provide strong evidences that Lcarnitine has a direct effect on nuclear receptor protein activity. Furthermore, we were able to show that Lcarnitine acts as a potent inducer of the transcription of catabolic nuclear receptors and decreases transcription of adipogenic and insulinresistance linked effector genes. Additionally, diverging effects of Lcarnitine supplementation on PPAR transcription levels in different tissues could be shown in the interplay of adipocytes, skeletal muscle cells and

strategy supporting established antihyperlipidemic therapies. Background Lcarnitine (L3hydroxy4N-trimethylaminobutyrate) is a quaternary ammonium compound and an indispensable metabolite in all eukaryotic cells. Lcarnitine is synthesized from the essential amino acids lysine and methionine, with the highest rate of biosynthesis in the kidney and liver [1]. Lcarnitine has a number of essential roles for the metabolism, including transport of activated longchain fatty acids into the mitochondrial matrix for subsequent βoxidation as well as modulation of the acylCoA/CoA ratio in the cell [2].
Lcarnitine itself has also been proven to regulate the expression of numerous important genes for cellular metabolism and has therefore been coined as a nutrigenomic factor [3].
Systematic reviews indicate a significant beneficial impact of Lcarnitine on cardiovascular and metabolic diseases [4,5]. Several studies show that exogenous carnitine administration is associated with a reduction of hyperlipidemia, hypertension, hyperglycemia and insulin resistance [6]. Additionally, Lcarnitine seems to have beneficial impacts on ventricular dysfunction, arrythmia and angina pectoris [7].
Although, as described, multiple lines of evidence suggest that Lcarnitine plays an important role in modulating metabolic functions in humans, few studies have elucidated the regulatory role of Lcarnitine on the molecular level. Artificial carnitine depletion in vitro has been shown to decrease mRNA expression of carnitine acetyltransferase (CRAT) and carnitine palmitoyl transferases (CPT1A and CPT2), whereas following carnitine supplementation leads to a full reversal of the downregulation [8]. In addition to this effect, a close connection between Lcarnitine and the peroxisome proliferatoractivated receptor system (PPAR) seems to exist. In adipocytes, supplementation of Lcarnitine suppresses the expression of PPARγ, which is involved in adipogenesis, and induces lipolytic gene expression [9]. Supplementation of Lcarnitine to hepatocytes seems to upregulate PPARα expression, which participates in lipid catabolism, as well as PPARαregulated genes such as CPT1 and acylcoenzyme A oxidase (ACOX) [10].
However, the diverging shortterm and longterm effects of Lcarnitine supplementation have not been investigated in more detail yet, especially in the interplay of adipocytes, liver and muscle cells. The major goal of this study was to investigate the regulatory mechanisms of Lcarnitine on nuclear receptor expression and activity. The nuclear hormone receptor superfamily of ligandactivated transcription factors represents the most important group of cellular nutrient sensors, enabling the organism to rapidly adapt to metabolic changes by inducing appropriate genes and pathways [11,12]. We hypothesized that Lcarnitine has distinct time-dependent effects on nuclear receptor activity and expression. Therefore, the effects on the protein activity of subfamily I (RAR, LXR, VDR, PPAR), subfamily II (RXR, HNF4) and subfamily III nuclear receptors (ER, GR, AR, PR) were measured in hepatocytes. In addition we hypothesized that Lcarnitine stimulates nuclear receptor expression associated with lipid catabolism (PPARα and PPARδ) accompanied with a suppression of adipogenic nuclear receptors (PPARγ) [13]. Finally, key genes of these pathways were further examined by qPCR, namely CYP27A1, CYP2R1, ALDH1A1, HSD11B2, OGT and HMGCR.

Materials
The human epithelial liver cell line WRL68 and the murine preadipocyte cell line 3T3L1 were obtained from the American Type Culture Collection (ATCC). Primary skeletal muscle cells (SKMC) were acquired from Lonza. Cignal TM Finder Nuclear Receptors 10Pathway Reporter Arrays were obtained from Qiagen. All oligonucleotides for qPCR experiments were synthesized and purchased from Microsynth Austria.
Cell culture All cells were maintained at 37°C in a humidified atmosphere containing 5% CO 2 .
WRL68 cells were resuspended in 7600µl OptiMEM and 400µl 10%FCS or dialyzed 10%FCS and counted with the help of a CASY cell counter. According to the number of cells, the volume for the media containing 6x10 5 cells was calculated. 50µl of this solution were added to each well and incubated for 24 hours at 37°C, 5% CO2. The next day, the medium in the wells was completely removed and 150µl of the respective transfection medium were added to each well and again incubated for 4 or 24 hours at 37°C, 5% CO2.
Luciferase activity was determined by using the DualLuciferase® Reporter Assay System (Promega).

Reverse transcription and qPCR experiments
Total RNA was extracted from cells using the RNeasy® Plus Mini Kit (Qiagen) according to the manufacturer's protocol. RNA concentration and purity were measured using a NanoDrop spectrophotometer (Thermo Scientific).
One µg of total RNA was converted to cDNA using the LunaScript TM Reverse Transcriptase SuperMix (New England Biolabs).

Results
Lcarnitine directly effects protein activity of nuclear receptor subfamily I members Our first experiments examined the regulatory effect of Lcarnitine on nuclear receptor protein activity in WRL68 cells via reporter gene assay. The relative firefly luciferase activity of all subfamily I nuclear receptor pathways was increased after cultivation of cells with dialyzed FCS (Fig.1). In the case of the retinoid acid receptor pathway (RAR), shortterm induction with Lcarnitine (for 4 h) led to a further induction of the activity, whereas extended Lcarnitine supplementation for 24 hours slightly decreased activity again to 1.29fold of normal untreated cells (Fig. 1A). A similar effect was observed for the PPAR pathway (Fig. 1B). In this case, short term Lcarnitine induction even led to a 2.85fold induction of the luciferase activity.
The pathway, which was affected the most by Lcarnitine depletion and supplementation was the VitaminDreceptor pathway ( Fig. 1C). Whilst the supplementation of Lcarnitine for 4 hours led to 2.88fold increase in activity, supplementing for 24 hours resulted in a reduction significantly below the normal physiological level. Contrary, in the case of the liverxreceptor pathway (LXR) the activity was further upregulated when Lcarnitine was supplemented for 24 hours (Fig. 1D).
Lcarnitine directly effects protein activity of nuclear receptor subfamily II members Subfamily II nuclear receptor pathways were as well upregulated after induction of Lcarnitine deficiency via dialysis in WRL68 cells (Fig. 2). After shortterm Lcarnitine supplementation, the retinoid X receptor (RXR) activity was further increased 1.93fold compared to untreated cells ( Fig. 2A). Extended supplementation for 24 hours led to a subsequent decrease of luciferase activity to normal physiological levels.
In the case of hepatocyte nuclear factor 4 (HNF4), a similar pattern was observed (Fig.   2B). Lcarnitine depletion and subsequent supplementation for 4 hours led to an increase in the activity of the pathway, whereas extended supplementation led to a decrease to physiological activity with rising Lcarnitine levels.
Lcarnitine only effects estrogen receptor pathway, but no other subfamily III nuclear receptors In WRL68 cells under Lcarnitine deficiency, a 2.12fold increase of the estrogen receptor (ER) activity was observed (Fig. 3A). Addition of Lcarnitine for 4 hours led to a further upregulation of the activity. When cells were subsequently induced with 80µM or 120µM Lcarnitine for 24 hours, representing supraphysiological conditions, activity slightly decreased again.
Contrasting, androgen receptor (AR), glucocorticoid receptor (GR) and progesterone receptor (PR) activity were not significantly changed either under conditions of Lcarnitine deficiency or supplementation ( Figure 3BD). induced after dialysis in WRL68 and downregulated in SKMC and 3T3L1 cells (Fig. 4).
PPARγ levels remained unchanged after dialysis in WRL68 and SKMC but were decreased in 3T3L1 cells. Alcohol dehydrogenase 1A1 (ALDH1A1) expression was decreased significantly after induction of Lcarnitine deficiency (Fig. 6C). Both shortterm Lcarnitine supplementation as well as extended Lcarnitine supplementation could not upregulate transcription, and the expression levels remained at 15% compared to normal untreated cells.
11βhydroxysteroid dehydrogenase type 2 (HSD11B2) expression rose slightly 1.42fold when cells were grown under Lcarnitine deficiency condition (Fig. 6D) hormonesensitive lipase, as well as a significant downregulation of PPARγ [9]. Since PPARγ was proven to be involved in adipogenesis, this would suggest a role of Lcarnitine as a stimulator of lipolysis and energy dissipation. A corresponding effect was found in the case of PPARα in rats, where in vivo substitution of Lcarnitine for 28 days led to a significant upregulation of PPARα expression [14]. PPARα is already known to play a crucial role as an enhancer of fatty acid oxidation, therefore substantiating the role of Lcarnitine as an important stimulator of the catabolism.
By reportergene assays, we were able to prove a direct effect of Lcarnitine on nuclear receptor protein activity ( Fig. 1-3). The protocol for induction of Lcarnitine deficiency with the use of dialyzed FCS is an established procedure and has been previously used in publications from our research group [8,10]. Lcarnitine deficiency, achieved via dialysis, in WRL68 cells led to an initial increase of seven examined nuclear receptor pathways (RAR, RXR, VDR, PPAR, HNF4, ER, LXR). No changes could be observed in the AR, PR and GR pathway, where the relative firefly luciferase activity remained at the level of physiological cultivation conditions. Subsequent shortterm Lcarnitine supplementation for 4 hours resulted in a further stimulation of the seven above mentioned pathways. Short term Lcarnitine pulses for 4 hours can only address preformed transcription complexes, because no significant de novo protein synthesis can take place in this limited time period. Therefore, Lcarnitine substitution obviously has the potential to interact with preformed nuclear receptor complexes, thereby increasing pathway signaling activity.
The pathways which were affected the most by 4 hours of Lcarnitine supplementation were the VDR and PPAR pathway. Interestingly, a link between these two pathways has already been revealed, with 1,25dihydroxyvitamin D being able to upregulate the PPAR pathway and thereby stimulating lipid metabolism in diabetic rats [15]. The PPAR signaling pathway is already known to be influenced by numerous metabolites, thereby acting as a key regulator of lipid homeostasis [16,17]. Since Lcarnitine serves as an essential In addition to this effect, we were able to reveal a further close connection between the PPARsystem and Lcarnitine on the transcriptional level in different tissues. Therefore, primary human skeletal muscle cells were investigated together with the WRL68 cell line, representative of mature human hepatocytes, as well as differentiated 3T3L1 cells, representative of mature murine adipocytes. The combination of human hepatocyte cell lines and murine adipocyte cell lines has already been successfully used in previous studies [18,19].
PPARγ, PPARα and PPARδ steady state mRNA levels in WRL68 cells were upregulated in the absence of Lcarnitine via the use of dialyzed FCS, according to the results of the reporter gene assays (Fig. 4). In contrast, in SKMC and 3T3L1 cells, dialysis led to a downregulation of the whole PPAR system. In addition, transcript amounts of all PPAR members were downregulated in all three observed cell lines in the course of shortterm Lcarnitine supplementation for 4 hours, whilst the reporter gene assay showed an increase of PPAR activity after short-term supplementation. This observation further substantiates the hypothesis that Lcarnitine is able to rapidly interact with preformed protein complexes on the nuclear level.
In contrast to the short time pulse, extended Lcarnitine supplementation for 24 and 48 hours led to a slight increase of PPARγ transcription levels in WRL68 cells. PPARδ and PPAR-α mRNA amounts in WRL68 were markedly increased 6.7-and 7.7-fold, respectively.
In contrast, in the reporter gene assay for PPAR protein signaling, activity decreased down Different to PPARγ, the transcript levels of PPARα were significantly increased in all the cell lines after extended L-carnitine supplementation, concordant with previous studies [9]. PPAR-δ mRNA amounts remained relatively lowered in SKMC and 3T3-L1 cells after prolonged L-carnitine supplementation but were as well significantly increased 6.7-fold in WRL68 cells. PPARγ upregulation is known to lead to an increase of lipogenesis and a decrease of lipolysis in adipose tissue [21]. In contrast, PPARα and PPARδ promote an increase of βoxidation in skeletal muscles and hepatocytes as well as a decrease of de novo lipogenesis in the liver [13]. Therefore, based on our results, extended Lcarnitine supplementation seems to be able to potently promote catabolic pathways, by inducing expression of catabolic PPARs and thus enhancing βoxidation and reducing lipogenesis.
The additional promoter active factors LXRα and RARβ were downregulated following the absence of Lcarnitine in all cell lines. In SKMC and WRL68 cells, mRNA levels remained significantly lowered after both short-term and extended supplementation. In WRL68 cells, steady state expression levels even dropped down to 2% under dialysis and remained at 15% even after extended supplementation, compared to normal growth conditions. These results show a clear disparity between decreasing mRNA abundance and the increase of LXR-α and RAR-β protein activity in WRL68 cells observed in the reporter gene assays ( Fig.   1-2). This phenomenon could be due to several reasons. Whilst transcription is downregulated, the pool of already synthesized mRNA might however be more efficiently translated. Another reason could be an increase of the proteins' halflife due to a reduced rate of degradation via posttranslational modifications. Lcarnitine administration has already been shown to modulate posttranslational modifications, leading to an increase of phosphorylation of AMPactivated protein kinase (AMPK), PI3K, Akt and mTOR [22][23][24].
Additionally, Lcarnitine supplementation has been proven to exert a decrease in proteasome activity and downregulate genes of the ubiquitin proteasome system [25,26].
A similar effect has been described for the nutrigenomic metabolite Niacin, where supplementation reduced ubiquitination of hepatic ACOX1 and CYP4A1, thereby increasing protein levels without affecting mRNA transcription [27].
Interestingly, 3T3-L1 was the only cell line, where an increase of LXR-α and RAR-β transcription was observed after extended L-carnitine supplementation (Fig. 5). After 48 hours of supplementation, mRNA levels increased up to 2-fold compared to the normal untreated cells. LXR activation in both murine and human adipocytes has already been shown to shift substrate oxidation towards utilization of lipids and upregulate mitochondrial β-oxidation [28]. Concordant with these findings, RAR activation in adipocytes seems to suppress adipogenesis and to down-regulate mRNA expression of PPAR-γ, a key regulator of adipocyte differentiation [29,30]. These results further substantiate our hypothesis, that L-carnitine acts as potent promoter of catabolic pathways via increase of lipolysis and decrease of lipid storage in multiple tissues.
Interestingly, the expression levels of RXR-α increased as well only in 3T3-L1 cells after extended supplementation. In contrast, in SKMC and WRL68 cells, RXR-α mRNA amounts declined under Lcarnitine deficiency but remained relatively unchanged in all states of supplementation, averaging at two-thirds of the expression levels observed in normal growth conditions. A possible reason for this could be the fact that RXRα acts as an obligate hetero-dimerization partner for a number of other nuclear receptors [31]. Since a strong decrease in RXRα levels would also impair the effectiveness of those other receptors, it could be hypothesized that cells must maintain relatively high amounts of RXRα in order to preserve their capacity to modulate transcription of essential target genes.
In addition, the effect of Lcarnitine on the expression levels of key effector genes in WRL68 cells was measured (Fig. 6). ALDH1A1 is a major gene of the oxidative pathway of alcohol metabolism and has been shown to act as a promoter of adipogenesis [29,30].
Lcarnitine supplementation led to a significant decrease of ALDH1A1 expression, substantiating the hypothesis that Lcarnitine substitution effectively decreases adipogenesis. Olinked βNacetylglucosamine transferase (OGT) catalyzes the addition of Nacetylglucosamine in Oglycosidic linkage to serine or threonine residues [32].
Additionally, hepatic OGT overexpression has been shown to impair the expression of insulinresponsive genes and contribute to insulin resistance and dyslipidemia [33]. Beyond that, in a chipscreen study performed by our lab, we observed that several hundreds of genes throughout the whole genome showed an increased or decreased transcription due to Lcarnitine supplementation, underlining the importance of this nutrigenomic metabolite [54]. Similar intense effects on the transcriptome, which resulted in altered expression of hundreds of genes, were already observed for the nutrigenomic metabolites niacin, vitamin D and leucine [55][56][57]. In the case of niacin, similar to our results, a tissuespecific pattern of effects has been observed as well. Niacin specifically altered expression of a group of genes in adipose tissue, but not in the liver, heart or skeletal muscle [57]. For a more detailed representation, Table 2 shows a comparison of Lcarnitine with other important nutrigenomic metabolites exerting similar effects.  [58]. Therefore, future studies on the complex effects of Lcarnitine on cellular pathways are necessary to extend our understanding of the key molecular mechanisms behind its apparent beneficial effects. Availability of data and materials

Declarations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.       Means without asterisk show no statistical significance (p > 0.05); (p-values of asterisk marked means: *p < 0.05, **p < 0.01, ***p < 0.001).