13-Methylberberine, a berberine analogue with stronger anti-adipogenic effects on mouse 3T3-L1 cells

Lipid metabolism modulation is a main focus of metabolic syndrome research, an area in which many natural and synthetic chemicals are constantly being screened for in vitro and in vivo activity. Berberine, a benzylisoquinoline plant alkaloid, has been extensively investigated for its anti-obesity effects and as a potential cholesterol and triglyceride-lowering drug. We screened 11 protoberberine and 2 benzophenanthridine alkaloids for their anti-adipogenic effects on 3T3-L1 adipocytes and found that 13-methylberberine exhibited the most potent activity. 13-Methylberberine down-regulated the expression of the main adipocyte differentiation transcription factors, peroxisome proliferator-activated receptor gamma (PPARγ) and CCAAT enhancer binding protein alpha (C/EBPα), as well as their target genes. PPARγ, C/EBPα, and sterol regulatory element binding protein 1 (SREBP-1) protein levels were reduced, and this lipid-reducing effect was attenuated by an AMP-activated protein kinase (AMPK) inhibitor, indicating that the effect of this compound requires the AMPK signaling pathway. Decreased Akt phosphorylation suggested reduced de novo lipid synthesis. C-13 methyl substitution of berberine increased its accumulation in treated cells, suggesting that 13-methylberberine has improved absorption and higher accumulation compared to berberine. Our findings suggest that 13-methylberberine has potential as an anti-obesity drug.

Many studies have reported that PPARγ pathway inhibition and AMP-activated protein kinase (AMPK) pathway activation are the mechanisms by which berberine reduces lipid levels 3,[11][12][13] . As 13-methylberberine is a close analogue of berberine, we examined its effects on PPARγ activation and on the upstream AMPK and Akt signaling. An alkyl substitution at position 13 enhanced the anti-adipogenic activity of 13-methylberberine; thus, we compared the structure-activity relationship of the benzylisoquinoline alkaloids in this study to berberine. The potential of 13-methylberberine as a candidate for metabolic syndrome treatment and its cytotoxicity are discussed.

13-Methylberberine inhibits adipogenesis and the expression of AMPK signaling pathway genes.
We used RT2 Profiler PCR Arrays kits (QIAGEN) designed for pathway-focused gene expression analyses to characterize the molecular mechanisms by which 13-methylberberine and berberine exert their anti-adipogenic effects. We compared the expression profiles of genes involved in adipogenesis and the AMPK pathway in cells treated with berberine or 13-methylberberine with the non-treated control. (c) Triglyceride content in 3T3-L1 adipocytes on Day 12; the results are normalized to the control (0.1% DMSO). All compounds were tested at 5 μ M (containing 0.1% DMSO). n ≥ 9, error bar = SE. *p < 0.05, **p < 0.01, ***p < 0.001 two-tailed Student's t-test. (d) Triglyceride content in 3T3-L1 adipocytes after treatment with various concentrations of berberine or 13-methylberberine. The results are normalized to the control (0.1% DMSO). n ≥ 6, error bar = SE. ***p < 0.001 two-tailed Student's t-test.
Scientific RepoRts | 6:38129 | DOI: 10.1038/srep38129 signaling genes, 13-methylberberine more strongly down-regulated lipogenesis-related gene expression in the AMPK signaling pathway than berberine. Activation of AMPK signaling by 13-methylberberine. The lipid-reducing effect of berberine is reported to act via AMPK activation. Here, we examined the effects of 13-methylberberine and berberine on AMPK signaling components at the translational and transcriptional levels. The immunoblot analysis showed increased phosphorylation of AMPK and its downstream target, acetyl-CoA carboxylase (ACC), after alkaloid treatment (Figs 3 and S5), especially 13-methylberberine treatment. AMPK phosphorylation was observed after 24-48 h of treatment, and ACC phosphorylation was observed after 6-48 h of treatment. In our study, we found that ACC was phosphorylated earlier than AMPK. However, ACC protein levels were substantially decreased after the 24-96 h treatment, suggesting that both alkaloids affected lipid biosynthesis by increasing ACC phosphorylation and degradation to inhibit its activity.
Quantitative real-time PCR and immunoblot analyses showed that both alkaloids affected the PPARγ and AMPK pathways in a similar manner. However, 13-methylberberine showed a stronger effect than berberine at the same concentration.
13-Methylberberine anti-adipogenic effects are dependent on AMPK signaling. Because our results showed that 13-methylberberine reduced the mRNA and protein levels of intermediates in the PPARγ and AMPK pathways, we investigated if its lipid-reducing activity required AMPK signaling. When adipocytes were co-treated with 13-methylberberine and compound C, an AMPK inhibitor, the lipid-reducing effect was suppressed. Lipid droplet accumulation was observed (Fig. 4a), and triglyceride levels were significantly increased ( Fig. 4b) in cells co-treated with compound C and 13-methylberberine. Although AMPK phosphorylation was not affected by compound C treatment at 6 h, ACC phosphorylation level showed trend of decrease (Fig. 4c). Qualitative and quantitative triglyceride measurements indicated that AMPK inhibition by compound C attenuated the anti-adipogenic effect of 13-methylberberine on 3T3-L1 cells.
Akt is involved in adipocyte differentiation, and its activation induces 3T3-L1 preadipocyte differentiation 19,20 . Akt also regulates the insulin signaling pathway in adipocytes. Insulin-stimulated Akt phosphorylation is critical for insulin-induced glucose metabolism, glucose transport, and adipocyte differentiation 21,22 . Akt reduction inhibited mouse embryonic fibroblasts differentiation into mature adipocytes 23 . The Akt signaling pathway activates PPARγ and C/EBPα during 3T3-L1 adipocyte differentiation, inducing adipogenesis. When we investigated Akt activation by analyzing Akt phosphorylation levels with immunoblots in control and 13-methylberberine-treated adipocytes, 13-methylberberine treatment showed trend of decreased phospho-Akt levels (Fig. 5b).
In vivo accumulation and cytotoxicity of 13-methylberberine in 3T3-L1 cells. When we compared the effects of berberine and 13-methylberberine on adipogenesis, 13-methylberberine clearly exhibited stronger activity, including reduction in lipid droplet accumulation, triglyceride levels (Fig. 1b,d), and the mRNA (Fig. 2) and protein levels (Figs 3 and 5) of lipogenesis-related enzymes. AMPK was more strongly activated by 13-methylberberine treatment.
Next, we investigated the metabolic fate of both alkaloids in 3T3-L1 cells. Cells were treated with 10 μ M berberine or 13-methylberberine for 48 h after which the cells and culture medium were collected for extraction of alkaloids and analysis using LC-MS. Studies have reported four main metabolites of berberine found in rat plasma: berberrubine (m/z 322), thalifendine (m/z 322), demethyleneberberine (m/z 324) and jatrorhizine (m/z 338) 24 . In our berberine-treated cell extracts, berberine (m/z 336) and m/z 338 were detected. In the 13-methylberberine-treated sample, only 13-methylberberine (m/z 350) was detected (Supplementary Fig. S2). In the cell culture medium extracts, berberine (m/z 336), m/z 284, 384, and 380 molecular ion peaks were detected, whereas 13-methylberberine (m/z 350), m/z 284, 366, 384, and 380 were detected in respective sample ( Supplementary Fig. S3). We quantified the amount of berberine and 13-methylberberine using both alkaloids as standards. 49 pmole/mg protein of berberine and 121 pmole/mg protein of 13-methylberberine were detected in respective cell extract sample. In the cell culture medium, 0.048 μ M berberine and 0.69 μ M 13-methylberberine were detected (Fig. 6a). These results show 13-methylberberine accumulated at higher levels in the cell extracts and culture medium compared to berberine, suggesting that 13-methylberberine exhibits higher uptake and accumulation in vivo, which likely contributes to the higher potency of 13-methylberberine compared to berberine.
When cytotoxicity was evaluated, both alkaloids showed weak cytotoxicity in 3T3-L1 adipocytes. Treatment with 30 μ M decreased cell viability by approximately 25% (Fig. 6b). However, the cytotoxicity of these alkaloids at the effective concentrations for lipid modulation was marginal and resulted in a 15 to 20% decrease in cell viability.

Structure-activity relationship of the anti-adipogenic effects of benzylisoquinoline alkaloids.
Only 3 (berberine, 13-methylberberine, and coptisine) of the 11 protoberberines tested in this study reduced triglyceride levels in 3T3-L1 adipocytes (Fig. 1c). Dihydroprotoberberines were less active than the oxidized form. 13-methyldihydroberberine only slightly reduced lipid accumulation, and dihydroberberine had no effect. A comparison of the methylene dioxy-ring-type and dimethoxy-type alkaloids indicated the importance of the methylene dioxy-ring for the lipid-reducing effect, because the 2-3 dimethoxy-type alkaloids palmatine, corydaline, and dehydrocorydaline showed no lipid reduction. Additionally, the loss of lipid-lowering activity in berberrubine indicated the importance of the 9-methoxy group.
Alkylation at position 8 or 13 of protoberberine alkaloids increases relative lipophilicity and substituent size and affects cytotoxic activity in cancer cell lines 25,26 . Derivatives of 13-substituted quaternary coptisine were more effective at inhibiting human tumor cell growth. These derivatives were suggested to have increased lipophilicity, which in turn enhanced absorption into bodily fluids 27 . In our study, C-13 methyl-substitution of berberine increased lipid-reducing activity, suggesting that the increased lipophilicity enhanced absorption into the cells. This result is consistent with the increased accumulation of 13-methylberberine in treated cells (Fig. 6a).

Discussion
In this study, we examined the effects of several berberine analogues on adipogenesis in mouse 3T3-L1 cells. Oil Red O staining and triglyceride measurement results showed that 13-methylberberine treatment robustly inhibited lipid droplet accumulation. Further characterization of the molecular mechanism of the anti-adipogenic effect indicated that 13-methylberberine reduced the expression of the main adipogenic transcription factors, PPARγ and C/EBPα , at both the transcriptional and translational levels. qRT-PCR results showed the mRNA levels of PPARγ target genes and lipogenesis-related genes were more strongly suppressed in 13-methylberberinethan berberine-treated cells. We also investigated the involvement of AMPK and Akt signaling pathways, which are upstream of PPARγ and C/EBPα . Immunoblotting results showed that 13-methylberberine activated AMPK which directly reduced ACC levels via phosphorylation and indirectly via suppression of SREBP-1. ACC catalyzes the conversion of acetyl-CoA to malonyl-CoA, which is used in de novo fatty acid synthesis. Thus, reduction in ACC levels by 13-methylberberine inhibited lipid droplet accumulation in 3T3-L1 adipocytes. Furthermore, inhibition of AMPK by compound C attenuated the anti-adipogenic effect of 13-methylberberine, indicating that the lipid-reducing activity of 13-methylberberine is directly dependent on AMPK signaling. Akt signaling induces adipogenesis by activating the key transcription factors involved in the process, i.e. PPARγ and C/EBPα . Our immunoblot results demonstrated that Akt phosphorylation and activation were reduced by 13-methylberberine treatment, indicating that 13-methylberberine also modulates the Akt signaling pathway. All these results suggest 13-methylberberine acts on multiple molecular targets that contribute to its anti-adipogenic effect.
Scientific RepoRts | 6:38129 | DOI: 10.1038/srep38129 When we examined the cytotoxicity of berberine and 13-methylberberine, we found that the viability of 13-methylberberine-treated cells was slightly reduced compared to berberine at same concentration. However, the lipid-reducing activity of 13-methylberberine is stronger. When the amount of both alkaloids were quantified, higher concentration of 13-methylberberine was found in the cells and culture medium. The higher uptake and retention of 13-methylberberine in vivo could likely increase its potency compared to berberine.
An examination of the structure-activity relationship among the benzylisoquinoline alkaloids used in this study indicated the importance of the 2-3 methylene dioxy-ring, full oxidation of the protoberberine ring, and the methoxy residue at the 9 position in modulating 3T3-L1 adipocyte differentiation. These structural requirements would be useful for the development of new candidate chemicals with anti-adipogenic effects. Additionally, our investigation indicated the importance of methyl substitution at C-13 in enhancing uptake into the 3T3-L1 cells. The higher absorption and accumulation of 13-methylberberine in the cell contributed to its stronger activity at low doses compared to berberine.
Although it has not yet been clarified how these molecular structures affect lipid-reducing activity, berberine was reported to inhibit 3T3-L1 adipocytes differentiation through PPARγ pathways 28 . We found 13-methylberberine inhibited the PPARγ pathway more potently than berberine. Previous studies also suggested that berberine activates AMPK via AMP accumulation, which is induced by mitochondrial respiratory complex I inhibition in rats 29 . Although we have not performed experiments to determine the inhibitory activity of these chemicals on respiration, our finding that dihydroberberine, which inhibited respiratory complex I in rat, was ineffective in mouse 3T3-L1 cells suggests that further characterization of the action mechanism of berberine analogues is needed.
This study was aimed to screen plant benzylisoquinoline alkaloids for anti-adipogenic activity in 3T3-L1 cells which are extensively used to study adipogenesis and biochemistry of adipocytes. Our results indicated that 13-methylberberine inhibited adipogenesis more potently than berberine. This finding suggests 13-methylberberine is a potential drug lead for anti-obesity. Future studies using whole animal models would be crucial to evaluate its efficacy and safety for anti-obesity therapy.

Materials and Methods
Chemicals. Compound C was obtained from Sigma-Aldrich; dimethyl sulfoxide (DMSO) and all other chemicals were obtained from Wako Pure Chemical Industries, Ltd., unless indicated otherwise.
Cell culture. 3T3-L1 cells range from Passage 9 to 11 (received from Dr. Masaya Nagao, Kyoto University) were cultured in Dulbecco's Modified Eagle Medium (Wako) with 10% fetal bovine serum (Corning) at 37 °C and 5% CO 2 . Cells were passaged twice before used in assays to allow cells to re-establish normal cell cycle. Cell differentiation was induced at 2 days post-confluence (designated as Day 0) by adding 5 μ g/mL insulin (Sigma), 500 μ M isobutylmethylxanthine (Sigma), and 0.25 μ M dexamethasone 30 and cultured for two days. Subsequently, the cells were maintained in DMEM, 10% FBS and 5 μ g/mL insulin and the medium was changed every two days. Alkaloids were added into the medium on Day 4 unless otherwise stated.
Oil Red O staining. 3T3-L1 adipocytes (Day 12) grown in cell culture plates were rinsed with phosphate buffered saline (PBS) and fixed in formalin for 30 min at room temperature. The formalin was removed, and the cells were rinsed twice with PBS. A 0.3% w/v Oil Red O (Sigma) solution was added at room temperature to stain the cells. After 1 h, the cells were rinsed in PBS twice, and lipid droplet accumulation was assessed under a microscope.
Triglyceride and protein measurements. 3T3-L1 adipocytes (Day 12) grown in cell culture plates were rinsed twice with PBS. Cell lysis buffer (1 M Tris-HCl pH 7.5, 1 M MgCl 2 , and 10% Triton X100) was added to each well, and the cells were harvested into an Eppendorf tube using a cell scraper. The cells were disrupted with an ultrasonicator, and the triglyceride and protein levels in the cell lysates were measured. The triglyceride concentrations were determined using the Triglyceride E test kit (Wako), and the absorbance was measured at 595 nm using a PowerScan4 (Biotek) plate reader. Protein concentration was determined using Bio-Rad DC Protein Assay (Bio-Rad) reagents, and the absorbance was measured at 595 nm using a PowerScan4 (Biotek) plate reader. The triglyceride content of each sample was normalized to its corresponding protein content.
Gene expression analysis. 3T3-L1 adipocytes (Day 4) were treated without or with alkaloids for 48 h and collected as described above. RNA was extracted using the RNeasy Mini Kit (QIAGEN). Reverse transcription was performed using 2 μ g total RNA, and real-time PCR was performed using the RT 2 Profiler PCR Arrays format D (QIAGEN) kit according to the manufacturer's instructions. The data were analyzed using the Δ Δ C T method, and the relative transcript level was standardized using Gapdh as an internal control. The fold change between samples was normalized to the control (0.1% DMSO).
Alkaloids were then added and incubated with the cells for an additional 6 h. The cells were collected and lysed as described above.
Alkaloid concentrations were calculated based on LC peak area at 280 nm relative to standards peak area and those values were normalized to the protein content in cell extracts. (Supplementary Fig. S4).
Cytotoxicity assay. 3T3-L1 cells were cultured at density of 1.5 × 10 4 cells/mL in 96-well plates for 16 h. The culture medium was then replaced with fresh medium, treated with alkaloids, and cultured for an additional 24 h. Cell viability was determined using Cell Counting Kit-8 (Dojinbo). The cells were incubated with the reagent for 1 h, and the absorbance of the living cells was measured using a PowerScan4 (Biotek) plate reader at 450 nm.