SMOX and SMS Responsible for Polyamine Metabolism Enhanced Adipogenesis

Adipose tissue regulating carbohydrate and lipid metabolism had been extensively focused. However, the regulation of amino acid metabolism during adipocyte differentiation remained detailed. Here we applied RNA-Seq technique to establish the transcriptional landscapes of amino acid metabolism during adipogenesis. We totally screened 17 differentially expressed genes (DEGs) for amino acid metabolism at 7, 14, 21, and 28 days during adipogenesis from human mesenchymal stem cells (hMSCs), especially with 13 up-regulated genes most prevalent in our adipogenic anecdote. Small molecule metabolic process was the most enriched biological process following by oxidation-reduction process. Interestingly, the enforced expression of SMOX (spermine oxidase) responsible for polyamine metabolism in arginine and proline metabolism pathway facilitated adipogenesis more than SMS (spermine synthase) using RNA interference (RNAi). The established potential regulatory network further suggested that adipocyte differentiation tightly related with the basal metabolism of amino acid metabolism with the partially conrmed SMS-SMOX-PPARG signaling pathway. It would highlight the effect of adipogenesis on amino acid metabolism in adipocyte biology and provide the potential treatment strategy for the study of fat metabolic diseases.


Introduction
As we known, the consumption of unutilized calories induced a metabolic state promoting the commitment of human mesenchymal stem cells (hMSCs) to become preadipocytes, followed by their differentiation into mature adipocytes known as the primary site for fat storage [1][2][3]. It had been clinically well established for the dysfunction of the endocrine and paracrine signaling usually responsible for the important regulatory functions of adipocytes leading to metabolic disorders and diseases, such as obesity, type 2 diabetes mellitus (T2DM), in ammation, and insulin resistance [4][5][6][7][8]. As a highly orchestrated process, adipocyte differentiation regulated glucose and lipid homeostasis by storing the excess nutrients in lipid droplets and releasing the bioenergetic substrates via lipolysis [9][10].
Recently, the studies involving in the relationship between amino acid and adipogenic differentiation attracted an increasing focus. Amino acids had been con rmed as the metabolic switch fueling adipogenic differentiation, such as branched-chain amino acid (BCAA) [11], homocysteine [12], arginine [13], and polyamine [14]. Especially, it had well documented that the inhibition of BCAA catabolism thoroughly compromised adipogenesis [11]. Hydroxyproline, proline, lysine, glycine, and alanine also showed a potential to induce adipogenic effects in retinal pericytes [15]. However, relatively little is uncovered for the whole pro ling of amino acid metabolism during adipogenesis.
In our previous study, we established the whole transcriptional pro ling of cellular metabolism during adipogenesis [16]. Here, we further concentrated our attention on the transcriptional landscapes of amino acid metabolism during adipogenesis using RNA-Seq technique. Among 17 DEGs for amino acid metabolism at 7, 14, 21, and 28 days, we con rmed that the enforced expression of SMOX (spermine oxidase) responsible for polyamine metabolism in arginine and proline metabolism pathway facilitated adipogenesis more than SMS (spermine synthase) using RNA interference (RNAi) by lentiviruses system. The established potential regulatory network further indicated that adipocyte differentiation tightly related with the basal metabolism of amino acid metabolism with the partially con rmed SMS-SMOX-PPARG signaling pathway. This study would highlight the effect of adipogenesis on amino acid metabolism in adipocyte biology and provide the potential treatment strategy for the study of fat metabolic diseases.
Material And Methods hMSCs isolation and culture Identi ed in our previous study [16], the hMSCs were isolated following the protocols with slight modi cation [17][18]. Adipogenic differentiation assays were developed following the method with the slight modi cations [19][20]. In an Model 3100 series Forma Series Water Jacket CO 2 incubator (Thermo Fisher Scienti c, Ohio, United States), we cultured the hMSCs in 5.0 mL hMSCs Basal Medium (Cyagen bioscience, Inc., Santa Clara, CA, USA) supplemented with FBS (fetal bovine serum), penicillinstreptomycin, and L-glutamine in 25 cm 2 asks (Corning Incorporated, Corning, New York, USA) at 37°C with 5% CO 2 and 95% humidity.

Human adipocyte differentiation
Expanded to passage 6, the hMSCs with 80-90% of the nal con uence were stimulated with adipogenic cocktails of hMSCs Basal Medium including 1.0 µM dexamethasone, 0.01 mg/mL insulin, and 0.5 mM 3isobutyl-1-methyl-xanthine (Sigma, St. Louis, Mo, USA) on day 0. Adipogenic potential of hMSCs was assessed following the documented method with slight modi cation [21][22]. Brie y, the hMSCs were xed with 3.7% formaldehyde for 30 min at room temperature after washed with phosphate buffered saline (PBS) (pH 8.0) buffer twice. And then, the hMSCs were stained with Oil Red O (Cyagen bioscience, Inc., Santa Clara, CA, USA) for 1 h after washing three times with PBS buffer. Cell morphology and the formation of lipid droplets of adipocytes were photographed using a uorescence microscope (IX73, Olympus, Tokyo, Japan).

RNA isolation and RNA-Seq sequencing and analysis
Following the manufacturer's instructions, the total RNA isolated from the adipogenic samples using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) was stored in liquid nitrogen for the further analysis. The freshly isolated RNA samples were for RNA sequencing after puri ed using the NucleoSpin RNA clean-up kit (Macherey-Nagel, Düren, Germany) and quali ed by Bioanalyzer 2200 (Aligent Technologies, Santa Clara, CA, USA).
In our previous study [23], RNA-Seq sequencing was developed by NovelBio Bio-Pharm Technology Co., Ltd, Shanghai, China. Additional remarks, the Ion Total RNA-Seq kit v2.0 (Life Technologies, Santa Clara, CA, USA) following the manufacturer's instructions was used to prepare RNA of RIN (RNA integrity number) > 8.0 to construct the complementary DNA (cDNA) libraries for single-end sequencing of protonsequencing process.
Here, we systematically established the transcriptional landscapes of amino acid metabolism during adipogenesis depending on KEGG pathway. We de ned an absolute value of log 2 ratio ≥ 2.0 as the threshold of the differentially expressed genes (DEGs), and the signi cant DEGs simultaneously met a false discovery rate (FDR) < 0.001 and an absolute value of log 2 ratio ≥ 2.0. We set the cutoff of p_value < 0.05 as the differential expressed biological process for GO (Gene Ontology) analysis.

qRT-PCR assays
Here, the relative gene expression level was determined using qRT-PCR (quantitative real-time PCR). 2.0 µg RNA was used to synthesize cDNA using Oligo (dT 18 Table 1. ACTB (actin beta) was used as the internal control [24]. All experiments were carried out in triplicates. In order to investigate gene function of our screened target genes, we strengthened the expression of SMS and SMOX differentially down-regulated during adipogenesis with small interfering RNA (siRNA) using a lentivector expression system. Lentiviral vector pGV492 digested with BamHI/AgeI was used to overexpress SMS (NM_004595) and SMOX (NM_175839). The primer sequences of SMS were as follows: 5'-AGGTCGACTCTAGAGGATCCCGCCACCATGGCAGCAGCACGGCACAGCACG-3' and 5'-TCCTTGTAGTCCATACCGGGTTTAGCTTTCTTCCAAACAG-3', and those of SMOX were as follows: 5'-AGGTCGACTCTAGAGGATCCCGCCACCATGCAAAGTTGTGAATCCAGTGG-3' and 5'-TCCTTGTAGTCCATACCGGTCCCCTGCTGGAAGAGGTCTCG-3'. All the lentiviruses, including SMS-siRNA, SMOX-siRNA, and the negative control, were prepared by GeneChem Co., LTD (Shanghai, China).
After expanded to passage 6 approximately with 20% con uence, the hMSCs were separately transfected with SMS-siRNA and SMOX-siRNA lentiviruses using a multiplicity of infection (MOI) of 10 and HitransG Transfection Reagent P (Genechem Inc., Shanghai, China). Transfection e ciency was con rmed according to uorescence intensity of GFP (Green uorescent protein) expression under a uorescence microscope (IX73, Olympus, Tokyo, Japan).

Adipogenic differentiation
Here, we assessed adipogenic potential of hMSCs via Oil Red O staining assays (Fig. 1). No lipid droplet appeared at 0 day, and the most obvious phenotypic changes was at 7 days after hMSCs differentiating to adipocytes with lipid-droplets accumulating ( Fig. 1a; 1b). Specially, the expansion of size and number of lipid droplets increased from 7 to 28 days (Fig. 1b-f). It indicated our hMSCs differentiating along adipogenesis.

Transcriptional landscapes of amino acid metabolism during adipogenesis
Here, RNA-Seq technique was performed to establish the transcriptional landscapes of amino acid metabolism during adipogenesis. 17 DEGs were identi ed during adipogenesis (Fig. 2). Furthermore, 13 up-regulated genes were most prevalent at four sampling points in our adipogenic anecdote, and thus suggested adipogenic differentiation facilitated the gene expression of amino acid metabolism pathway.
As amino acid metabolism was a high energy demanding process, the facilitation of amino acid metabolism at transcriptional level would be the supply process of the essential energy.
Nine DEGs (MAOA, ALDH1A3, AOC2, AOC3, SMS, ALDH4A1, ASS1, GLUL, and SMOX) got involved in more metabolic pathways of amino acid metabolism, including arginine and proline metabolism, glycine, serine and threonine metabolism, histidine metabolism, phenylalanine metabolism, tryptophan metabolism, tyrosine metabolism, cysteine and methionine metabolism, aspartate and glutamate metabolism, and arginine biosynthesis. For the DEG of MAOA (monoamine oxidase A), it related with six amino acid metabolism pathways, such as arginine and proline metabolism, glycine, serine and threonine metabolism, histidine metabolism, phenylalanine metabolism, tryptophan metabolism, and tyrosine metabolism. Seven DEGs (ALDH1A3, AOC2, AOC3, SMS, GGT1, GGT5, and SMOX) were the most abundant for the other amino acids metabolism, including beta-alanine metabolism, glutathione metabolism, taurine and hypotaurine metabolism, and beta-alanine metabolism was the most prevalent. Above all, the gene transcriptional landscapes of amino acid metabolism were strong in response to adipogenesis.
The above DEGs were also matched along with biological process by GO analysis (Fig. 3). Based on one gene corresponding to multiple GO, we identi ed the differential expressed biological processes (p_value small molecule metabolic process (GO:0044281) was the most enriched biological process following by oxidation-reduction process (GO:0055114). As we known, all amino acids derived from the intermediates in glycolysis, the tricarboxylic acid cycle (TCA), or pentose phosphate pathway (PPP). The two differential biological processes would be the key way for the energy supply.

SMS and SMOX enhanced adipogenesis
The two genes SMS and SMOX were subjected to the further additional validation experiments. Here, we focused on the investigation of effect of overexpression SMS and SMOX on adipogenesis.
We rstly con rmed the successful overexpression of the hMSCs following the reporter gene GFP expressing via transfection using lentiviral system ( Fig. 4a; 4b). With SMS and SMOX overexpressed, respectively, the relative expression level of adipogenic biomarkers, including CEBPA (CCAAT enhancer binding protein alpha), CEBPG (CCAAT enhancer binding protein gamma), FASN (fatty acid synthase), LPL (lipoprotein lipase), and PPARG (peroxisome proliferator activated receptor gamma) (Fig. 4c), was raised by SMOX more than SMS at transcriptional level. Protein level was brought into correspondence with the increase of PPARG gene expression (Fig. 4d). Compared with the negative control, the lipid droplets accumulation of SMOX overexpression was also more than that of SMS ( Fig. 4e; 4f). Above all, SMOX overexpression enhanced adipogenesis at transcriptional, translational, and cellular phenotype level of lipid droplets accumulation more than SMS.
Here, we found SMOX overexpression fuels adipogenesis more than SMS using a lentivector expression system. Firstly, polyamine, ubiquitous positively charged amines, comprising agmatine, putrescine, spermidine, and spermine was essential for life, critical for adipogenesis from stem cell differentiation [14,[25][26][27][28]. Specially, spermidine was indispensable in differentiation of 3T3-L1 broblasts to adipocytes [29]. Exogenous polyamine also reciprocally regulated adipogenic differentiation [30]. Secondly, far fewer effects brought about by SMS enhancement was agreed with the study in transgenic mice displaying just slight increases in spermine and decreases in spermidine levels [31]. And it had proved that the expression of SMS was regulated primarily depending on substrate availability under the normal conditions and up-regulated when exposured to polyamines and their analogues [32]. As we known, spermidine was catalyzed by SMS to spermine, and spermine was inversely oxidized by SMOX to spermidine, 3-aminopropanal and hydrogen peroxide in polyamine pathway [33].
The predicted regulatory network between adipogenic biomarkers and amino acid metabolism Here, except for BHMT2 for cysteine and methionine metabolism and ID2 for adipogenic biomarkers, we established the regulatory network between the other 16 DEGs of amino acid metabolism and 9 adipogenic biomarkers at transcriptional level using the Integrative Genomic Viewer (IGV) software Especially, with SMS and SMOX overexpressed, PPARG was also partially con rmed to increase at transcriptional and translational level ( Fig. 4c; 4d). In all, there was the potential regulatory network, and SMS-SMOX-PPARG was a potential and positive regulation signaling pathway. Although we found some evidence to support SMS-SMOX-PPARG signaling pathway working for amino acid metabolism during adipogenesis, further work on additional interaction mechanisms between the upstream and downstream in SMS-SMOX-PPARG signaling pathways should be detailed.

Conclusion
In summary, we rstly uncovered the transcriptional pro ling of amino acid metabolism during adipogenesis by deep sequencing. We totally screened 17 DEGs for amino acid metabolism at 7, 14, 21, and 28 days during adipogenesis from hMSCs, especially with 13 up-regulated genes most prevalent in our adipogenic anecdote. Interestingly, the enforced expression of SMOX (spermine oxidase) responsible for polyamine metabolism in arginine and proline metabolism pathway facilitated adipogenesis more than SMS (spermine synthase). The established potential regulatory network further suggested that adipocyte differentiation tightly related with the basal metabolism of amino acid metabolism with the partially con rmed SMS-SMOX-PPARG signaling pathway. It would highlight the effect of adipogenesis on amino acid metabolism in adipocyte biology and provide the potential treatment strategy for the study of fat metabolic diseases.