Regulation of Cell Viability and Anti-inflammatory Tristetraprolin Family Gene Expression in Mouse Macrophages by Cottonseed Extracts

Bioactive plant extracts have been used for the prevention and treatment of various diseases. One of the major classes of bioactive compounds is plant polyphenols. Cottonseed ethanol extracts were determined by HPLC-MS analysis to be essentially free of toxic gossypol. The objective of this study was to investigate the effect of cottonseed ethanol extracts on the cytotoxicity and regulation of anti-inflammatory tristrataprolin (TTP) family gene expression in mouse cells. MTT, qPCR and immunoblotting assays tested the effects of cottonseed extracts in mouse RAW264.7 macrophages and 3T3-L1 adipocytes. No cytotoxicity effect was observed in macrophages treated with extracts from the coat or kernel of glanded and glandless cottonseed. Similarly, the viability of mouse adipocytes was not affected by cottonseed extracts. In contrast, gossypol and lipopolysaccharides were toxic to macrophages but not adipocytes under high concentration or long time treatment. Cottonseed extracts exhibited modest effect on TTP family gene expression in macrophages but glandless cottonseed coat extract significantly increased TTP mRNA and protein levels with a magnitude similar to cinnamon and green tea polyphenol extract and insulin. These results demonstrated that cottonseed extracts are harmless towards the mouse cells and that glandless cottonseed coat extract stimulates TTP gene expression. We propose that glandless cottonseed is a safe source of plant polyphenols with anti-inflammatory property.

Tristetraprolin/zinc finger protein 36 (TTP/ZFP36) and its homologues are anti-inflammatory proteins 23,24 . TTP family consists of four homologues in mice and rats (ZFP36/TTP, ZFP36L1/TIS11B, ZFP36L2/TIS11D, andZFP36L3) 25,26 . TTP binds to some cytokine mRNA AU-rich elements and destabilizes those molecules 27,28 . TTP knockout mice accumulate excessive levels of the proinflammatory cytokines and develop a systemic inflammatory syndrome consisting of arthritis, autoimmunity, and myeloid hyperplasia 29,30 . TTP over-expression decreases inflammation in macrophages 31 . Therefore, chemicals that increase TTP expression may have therapeutic value for inflammation-related disease prevention and/or treatment. However, it is not known whether cottonseed components can regulate TTP family gene expression since no prior work was done in this area.
The aim of current study was to investigate the effects of cottonseed extracts on the viability and regulation of TTP family gene expression in mouse cells. We used MTT, qPCR and immunoblotting assays to investigate cottonseed extract effects on mouse cell viability and the expression of anti-inflammatory TTP family genes 32,33 . Our results showed that cottonseed extracts are harmless towards the mouse cells and that glandless cottonseed coat extract stimulates TTP gene expression. We propose that glandless cottonseed is a safe source of plant polyphenols with anti-inflammatory property.

Results
Effect of cottonseed extracts on macrophage viability. MTT method was used to determine cell viability after being treated with cottonseed extracts for 2-72 h (Fig. 1). The viability of macrophages was not statistically affected by glanded cottonseed coat extract (Fig. 1A). Glanded cottonseed kernel extract also did not show significant effect on macrophage viability (Fig. 1B). Similar experiments were conducted on RAW cell viability using extracts from glandless cottonseed. MTT assays showed that extracts from coat (Fig. 1C) and kernel (Fig. 1D) of glandless cottonseed did not have significant effect on RAW cell viability after 2-72 h treatment with 5-100 µg/mL of the extracts. However, macrophage viability appeared to be reduced slightly, although not significantly, by higher concentration and longer time of the cottonseed extract treatment (Fig. 1A-D).
Effect of cottonseed extracts on adipocyte viability. The viability of mouse adipocytes was also determined with MTT method after being treated with cottonseed extracts for 2-24 h (Fig. 2). The adipocyte viability was not affected statistically by any of the treatments using glanded cottonseed coat or kernel extract up to 100 µg/ mL for 2 or 24 h ( Fig. 2A,B). Similarly, extracts from glandless cottonseed coat (Fig. 2C) and kernel (Fig. 2D) did not have significant effect on adipocyte viability after treatment for 2-24 h with 5-100 µg/mL of the extracts.
Effect of gossypol and lipopolysaccharides on cell cytotoxicity. As controls for the experiments, cell viability was measured in macrophages and adipocytes treated with the cytotoxic compound gossypol (known to be accumulated in the glanded cottonseed which causes male infertility) and the well-known endotoxin lipopolysaccharides (LPS). MTT assays showed that gossypol exhibited significant inhibitory effect on RAW macrophage viability under high concentration or long treatment time (Fig. 3A). Mitochondrial activity in RAW macrophages was almost completely inhibited by gossypol after treatment at 5-50 µg/mL for 24-72 h or 100 µg/mL for 2-72 h (Fig. 3A). MTT assays also showed that LPS had significant inhibitory effect on RAW macrophage survival after 100-1000 ng/mL treatment for 72 h (Fig. 3B). However, MTT assays showed that gossypol and LPS did not have significant effect on adipocyte survival after 2-24 h treatments (Fig. 3C,D).
Effect of cottonseed extracts on TTP/ZFP36 gene expression. The above results suggest that cottonseed extracts were not toxic to mouse macrophages and adipocytes under the experimental conditions. To provide evidence for potential health and nutritional benefits of this abundant resource of plant polyphenols, the effect of cottonseed extracts on anti-inflammatory TTP gene expression was investigated using mouse macrophages. qPCR showed that the effect of glanded cottonseed coat extract on TTP mRNA levels was minimal to modest in mouse RAW macrophages (Fig. 4A). TTP mRNA levels were increased less than two-fold without statistical significance in macrophages treated for 2, 8 and 24 h. Similarly, the kernel extract from glanded cottonseed showed less than three-fold of increases in TTP mRNA levels after 24 h treatment (Fig. 4B). The coat extract from glandless cottonseed did not have significant effects on TTP mRNA levels after 2-8 h treatment. However, TTP mRNA levels were significantly increased up to 7-fold in 24 h treated-macrophages by glandless coat extract (Fig. 4C). In contrast, the effect of kernel extract from glandless cottonseed on TTP mRNA levels was minimal with no statistical significance in the mouse RAW macrophages (Fig. 4D).
Immunoblotting was used to confirm if increased TTP mRNA levels by glandless cottonseed coat extract could result in increased TTP protein levels (Fig. 5). TTP polyclonal antibodies were used to detect TTP protein in macrophages treated with cottonseed extracts using LPS as a positive control, a well-known agent to induce TTP protein expression and phosphorylation in mouse macrophages 14,34 . Immune-reactive band(s) between 37 and 50 kDa corresponding to the predicted sizes of TTP and its phosphorylated forms were detected in the cells treated for 2 h with 100 ng/mL of LPS (Fig. 5A, lane 1 and Fig. 5B, lane 2). Similar sizes of immune-reactive bands with less intensity were detected by anti-MBP-mTTP antibodies in macrophages treated for 24 h with 5, 10, 20, 30, 40 and 100 µg/mL of glandless cottonseed coat extract (Fig. 5A). Similar immune-reactive bands were detected by synthetic TTP peptide antibodies in macrophages treated with 100 µg/mL of glandless cottonseed coat extract (lanes 7-10), and their levels were more than those treated with glanded cottonseed coat extract (lanes 3-6) (Fig. 5B).
Effect of cottonseed extracts on ZFP36L1 gene expression. Analyses of the effects of cottonseed extracts on gene expression were extended to the other three homologues of TTP in mouse macrophages. The effect of coat or kernel extracts from glanded or glandless cottonseed on ZFP36L1 gene mRNA levels in mouse RAW macrophages was minimal, mostly without statistical significance (Table 1). ZFP36L1 mRNA levels were increased approximately two-fold in macrophages treated for 2 h with most of the glanded coat extract   concentrations but the effects were reduced to minimal or even less than the control after 8-24 h treatment ( Table 1). The effect of kernel extract from glanded cottonseed on ZFP36L1 mRNA levels showed less than the control after 2-8 h treatment and returned to normal in 24 h treatment ( Table 1). The effect of coat extract from glandless cottonseed on ZFP36L1 mRNA levels in mouse RAW macrophages was also minimal except an increase of ZFP36L1 mRNA levels in 30 µg/mL treated macrophages ( Table 1). The effect of kernel extract from glandless cottonseed on ZFP36L1 mRNA levels showed similar pattern of glanded cottonseed kernel extract with less than the control after 2-8 h treatment and returned to normal in 24 h treatment ( Table 1).

Effect of cottonseed extracts on ZFP36L2 gene expression. ZFP36L2 mRNA levels in mouse RAW
macrophages were minimally to modestly affected by the extracts from coat or kernel extracts from glanded or glandless cottonseed (Table 2). ZFP36L2 mRNA levels were not affected by the coat extract from glanded cottonseed in mouse RAW macrophages treated for 2, 8 or 24 h except at 100 µg/ml treatment (Table 2). However, ZFP36L2 mRNA levels in macrophages were significantly increased by glanded kernel extract up to 3-fold (20 µg/mL for 8 h) and 4-fold (10 µg/mL for 24 h) ( Table 2). The coat extract from glandless cottonseed increased ZFP36L2 mRNA levels less than two-fold but with statistical significance in 24 h treated macrophages ( Table 2). The effect of kernel extract from glandless cottonseed on ZFP36L2 mRNA levels in mouse RAW macrophages was similar to glanded cottonseed kernel extract with less than the control in 2-8 h treated cells and less effective than the kernel extract of glanded cottonseed in 24 h treated cells ( Table 2).
Effect of cottonseed extracts on ZFP36L3 gene expression. ZFP36L3 mRNA levels were modestly affected by the coat and kernel extracts from glanded and glandless cottonseed in mouse RAW macrophages (Table 3). ZFP36L3 mRNA levels were only significantly increased to 2-fold by the kernel extract from glanded cottonseed after 8 h treatment (Table 3). qPCR did not show any significant effect of the coat and kernel extracts from glandless cottonseed on ZFP36L3 mRNA levels in mouse RAW macrophages (Table 3).

Discussion
We explored the potential of using cottonseed extracts as a safe source of plant polyphenols so that cottonseed value could be increased because plant polyphenols have been used for the prevention and treatment of various diseases. We recently showed that cottonseed extracts regulate DGAT and HuR gene expression in mouse macrophages 21,22 . In this study, we examined the effects of cottonseed extracts on the viability and mRNA levels of anti-inflammatory TTP family genes in mouse cells.
The major finding is that cottonseed extracts are harmless towards the cultured macrophages and adipocytes. As shown by MTT assays, the extracts from the coat and kernel of glanded and glandless cottonseed did not have any significant effect on cell viability after treatment for 2-24 h with up to 100 µg/mL of the extracts in culture medium. We recently developed a protocol for isolating bioactive extracts from seed coat and kernel of glanded and glandless cottonseed 20 . These cottonseed extracts are essentially free of the toxic compound gossypol with only 0.82, 0.03, 0.37 and 0 ng of gossypol per mg of the extracts from glanded coat, glanded kernel, glandless coat and glandless kernel, respectively 20 . These results suggest that cottonseed extracts are probably safe for consumption.
Another major finding is that gossypol and LPS exhibited significant inhibition on the viability of RAW macrophage but not adipocytes; the higher the concentration or the longer the treatment resulted in more severe www.nature.com/scientificreports www.nature.com/scientificreports/ reduction of mitochondrial activity. Our results that gossypol and LPS inhibited RAW cell viability are in contrast to the conclusion from a previous study 35 . It is unclear why the effect of gossypol and LPS on RAW cell viability in the two studies is different. One possible reason is that their study used a much higher cell density (4 × 10 5 cells/mL) in 96-well plate and ours used ¼ of their cells in 24-well plate. Another potential reason is that the final concentrations of gossypol used in their studies were probably lower than those used in our study; but it is difficult to know since they did not report the total volume of the culture medium in their study 35 . The cytotoxic compound gossypol is known to be accumulated in the glanded cottonseed which causes male infertility 36 but minimally present in glandless cottonseed 37 . Our results suggest that gossypol may affect immunity when over-consumption and accumulation of this toxic compound in the body.
The most significant finding was that only coat extract from glandless cottonseed significantly increased anti-inflammatory TTP mRNA levels up to 7-fold after 24 h treatment, resulted in increased TTP protein levels. This provides evidence for the potential health benefits of cottonseed extracts because plant extracts such as green tea and cinnamon extracts that can induce TTP gene expression have nutritional and therapeutic value for the prevention and/or treatment of inflammation-related diseases 11,14 . Other cottonseed extracts from the coat and kernel of glanded cottonseed and kernel of glandless cottonseed only had modestly effects on TTP mRNA levels in the macrophages. All four cottonseed extracts exhibited minor effects on the mRNA levels of ZFP36L1, ZFP36L2, and ZFP36L3 genes in the macrophage. It was shown previously that the anti-depressive compound quercetin was only identified in the glandless seed 20 . We showed here that TTP was only significantly induced by The effect of ethanol extracts from glandless cottonseed on stimulating TTP gene expression up to 7-fold is similar to those of insulin (7 fold with 100 nM for 30 min in mouse adipocytes) 38 , cinnamon extract (2 fold with 100 µg/mL for 2 h in macrophages and 10 fold for 1.5 h in adipocytes) 13,14,16 and green tea extract (50-140% in rat liver and muscle) 11 . Since anti-inflammatory TTP family proteins target TNF and a few other key cytokine mRNAs for their destruction by binding to the unstable 3′ AU-rich sequence 34,[39][40][41][42] , these results suggest that cottonseed extract from glandless seed may have potential health benefits for inflammatory diseases. However, the fact that no effect of the cottonseed extract on the viability of macrophages or adipocytes require further in-depth studies of its inflammatory effects. It also requires additional studies to elucidate the molecular mechanism of cottonseed extract in potential regulation of inflammation via anti-inflammatory TTP action.
In conclusion, this study demonstrated that no cytotoxicity effect was observed in mouse cells treated with cottonseed extracts which are essentially gossypol-free, suggesting that cottonseed extracts are probably safe to use. We also showed that the extract from glandless cottonseed coat significantly increased anti-inflammatory TTP gene expression in macrophages with a magnitude similar to cinnamon polyphenol extract, green tea extract and insulin, suggesting health and nutritional benefits for inflammation-related diseases. We propose that glandless cottonseed can be a safe source of bioactive polyphenols with anti-inflammatory property.  Table 2. Effect of Cottonseed Extracts on ZFP36L2 mRNA Levels. The data represent the mean and standard deviation of three independent samples. Different lower case letters displayed on the right side of the data in each column are significantly different between the cottonseed extract concentrations at p < 0.05. Values with statistical significance in gene expression are underlined.  www.nature.com/scientificreports www.nature.com/scientificreports/ seed coat), 0.03 (glanded seed kernel), 0.37 (glandless seed coat) and 0 ng (glandless seed kernel) of gossypol per mg of the extracts 20 .
Macrophage culture and treatment. Mouse RAW264.7 cells were maintained in DMEM+ medium containing DMEM, fetal bovine serum (10%, v/v), penicillin (100 U/mL), streptomycin (100 μg/mL), and L-glutamine (2 mM) in a water jacket CO 2 incubator at 37 °C with 5% CO 2 as described 34 . RAW macrophages were detached from the flask with a cell scraper and counted with a TC20 Automatic Cell Counter after stained with trypsin blue dye. For MTT and qPCR assays, the cells (0.5 mL, 1 × 10 5 cells/mL) were subcultured in 24-well cell culture plate. Raw macrophages were treated with cottonseed extracts, gossypol, LPS and 1% DMSO as the control. For immunoblotting, RAW macrophages were cultured in 75-mL culture flask with 20 mL DMEM. The cells were treated with cottonseed extracts for 24 h before preparing extracts as described below.
Adipocyte culture and treatment. Mouse 3T3-L1 preadipocytes were cultured at 37 °C in DMEM+ as described in macrophage culture. Preadipocytes were differentiated into adipocytes with a medium containing DMEM+, recombinant human insulin (1 μg/mL), dexamethasone (0.25 μM), and 1-isobutyl-3-methylxanthine (250 μM) 13 . After 48 h incubation, the medium was replaced with DMEM+ containing only of insulin (1 μg/mL). The medium was replaced with DMEM+ after incubation for additional 48 h, and the cells were grown for another 4-6 days. The cells were starved in DMEM without any supplementation for 4 h before being treated with cottonseed extracts and DMSO (the vehicle control, 1%).
Cell cytotoxicity assay. Cell cytotoxicity was evaluated with MTT method with the In Vitro Toxicology Assay Kit essentially as described previously 20 . Cells in 500 µL medium were treated with cottonseed extracts, gossypol and LPS and incubated at 37 °C, 5% CO 2 for 2, 5, 24 and 72 h. MTT assay reagent (50 µL thiazolyl blue tetrazolium bromide) was added to the medium, and incubated at 37 °C, 5% CO 2 for 2 h before adding 500 µL MTT solubilization solution. The color density at A570 nm was measured by microplate spectrophotometer (Epoch) and SmartSpec plus Spectrophotometer (BioRad).
Cell extracts, protein determination, sds-page and immunoblotting. Cell extracts were prepared according to a previously described procedure 34 . Briefly, RAW macrophages were scraped into 0.9% NaCl, transferred into falcon tube and centrifuged at 1,000 g for 5 min. Cells were lysed in a lysis buffer containing NaH 2 PO 4 (50 mM), pH 7.6, NaCl (250 mM), Nonidet P-40 (0.5%), phenylmethylsulfonyl fluoride (1 mM) and protease inhibitor cocktail (Sigma, cat #P8340,1:100 dilution). The cell lysate was centrifuged at 10,000 g for 10 min at 4 °C. Protein concentrations in the supernatant were estimated with the Bradford method using the Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories) after the extracts were denatured with 0.5 M NaOH 44 . Polypeptides (100 µg of protein per lane) were separated by 4-20% SDS-PAGE (Life Technologies) following the standard protocol. Proteins were transferred onto nitrocellulose membrane in transfer buffer containing 0.1% SDS with iBlot Gel Transfer System (Invitrogen). The membrane was blocked with nonfat dry milk in TTBS buffer (5%) and incubated with anti-MBP-mTTP antibodies (1:5,000 dilution in the blocking buffer for 18 h) produced against recombinant MBP-mTTP in rabbits 34 . Alternatively, immunoblotting was performed using rabbit polyclonal antibodies against synthetic peptides derived from human TTP (Abcam product number ab83579, 1:500 dilution in blocking buffer). After washed with TTBS buffer, the membrane was incubated with second antibodies (goat anti-rabbit IgG (H + L) horseradish peroxidase conjugate, Bio-Rad, 1:5,000 dilution in TTBS buffer, 2-4 h). Following washing with TTBS buffer, the membrane was incubated with Amersham ECL Prime Western Blotting Detection Reagent (GE Life sciences) for 5 min and chemiluminescent intensity was captioned by ChemiDoc Touch Image system (BioRad). RNA extraction, cDNA synthesis and real-time qPCR analysis. RNAs were isolated from macrophages using TRIzol reagent according to a previous procedure 14 . The total RNAs were used to synthesize cDNAs by a DNA Engine Gradient Cycler essentially as described 14 . SYBR Green qPCR assays were identical to those described 45 . The reaction mixture contained total RNA-derived cDNAs (5 ng), forward primer and reverse primer (200 nM each), and 1× iQ SYBR Green Supermix and performed with CFX96 real-time system-C1000 Thermal Cycler. The thermal cycle consisted of 3 min at 95 °C, 40-50 cycles at 95 °C for 10 s, 65 °C for 30 s, and 72 °C for 30 s. qPCR Data analysis. The fold change in gene expression was estimated by the ΔΔC T method 16,46 . The cycle of threshold (C T ) was obtained from 3-6 independent samples. The first delta C T value (ΔC T ) equals to the C T value of the target mRNA mius the C T value of the internal reference control (mouse 60 S ribosome protein 32, Rpl32) (ΔC T = C TTarget − C Tref ). The second delta C T value (ΔΔC T ) equals to the ΔC T of the target mRNA minus the ΔC T of the calibrator (DMSO control) (ΔΔC T = ΔC TTarget − ΔC Tcal ). The fold change in expression equals to 2 −ΔΔCT .
Statistics. The data represent the mean and standard deviation (n = 3-6). Statistics was conducted with ANOVA with SigmaStat 3.1 software (Systat Software). Multiple comparisons among the treatments were performed with Student-Newman-Keuls Method 16,21 . Significant difference between the treatment concentrations at p < 0.05 are indicated by the different lower case letters above each of the treatment time on the figures or in the table columns.