Combined decaffeinated green coffee and green tea extracts prevents fibrosis and hypertrophy in high-glucose-treated fibroblasts

1. Introduction

Diabetes mellitus remains as a major health problem, with a continuously increasing burden across the world (International Diabetes Federation, 2021). Chronic hyperglycemia in diabetes not only contributes to metabolic disturbances but also to long-term complications, including tissue fibrosis—a key pathological feature implicated in diabetic nephropathy, cardiomyopathy, and retinopathy (Resnikoff et al., 2022; Salvador et al., 2022). At the tissue level, fibrosis involves abnormal accumulation of extracellular matrix (ECM) proteins which leads to increased tissue stiffness and progressive loss of organ function (Wynn, 2008).

At a cellular level, myofibroblasts are considered as the major effectors of fibrogenesis. In response to cellular stressors such as high glucose, fibroblasts become activated and differentiate into myofibroblasts, marked by increased ECM secretion and upregulation of α-smooth muscle actin (α-SMA), a hallmark of contractile and secretory activity (Tuleta and Frangogiannis, 2021; Younesi et al., 2024). The sustained activation of this process is driven by complex molecular pathways, including transforming growth factor beta/Mothers Against Decapentaplegic Homolog (TGF-β)/SMAD3 signaling pathway, oxidative stress, and pro-apoptotic cascades (Ishikane et al., 2023). Furthermore, accumulating evidence suggests that matrix metalloproteinases (MMPs), particularly MMP-9, contribute to initiation and propagation of fibrogenic cascades in diabetes experimental models by activating latent TGF-β (Dayer and Stamenkovic, 2015; Zhang et al., 2020). Given the central role of fibroblast-to-myofibroblast transition in hyperglycemia-induced fibrosis, targeting this cellular process presents a promising therapeutic strategy for fibrosis prevention in diabetes.

Current strategies to mitigate fibrosis in diabetes focus largely on tight glycemic control and the use of pharmacological agents targeting specific organs, such as Renin-angiotensin system (RAS) inhibitors (Karihaloo, 2012). However, these interventions are often insufficient to prevent or reverse fibrogenesis, urging the need for novel anti-fibrotic agents that specifically inhibit myofibroblast activation pathways. Nevertheless, many of these new anti-fibrotic drug candidates are limited by inconsistent efficacy and potential adverse effects (Ishikane et al., 2023; Siani, 2020). These limitations underscore the urgent need to explore and develop safe and effective antifibrotic agents targetting fibroblast activation, particularly those derived from natural sources with multifunctional bioactivities.

Plant-derived extracts, particularly those obtained from green tea (Camellia sinensis) and coffee (Coffea canephora), have attracted considerable attention due to their wide range of biological activities, including antioxidant, anti-inflammatory, antidiabetic, and anti-metabolic syndrome effects (Nugrahini et al., 2024; Rohman et al., 2021). Previous studies from our group demonstrated that the combination of these two herbs at 320/320 μg/mL (green tea : green coffee) markedly reduced foam cell formation in oxLDL-stimulated Raw264.7 and suppressed 3T3-L1 differentiation into adipocytes (Nugrahini et al., 2024; Rachmawati et al., 2022). Furthermore, this combination was shown to attenuate the expression of key fibrotic markers, including TGF-β1, α-SMA, and collagen type I, in the cardiac tissue of metabolic syndrome rat models (Chomsy et al., 2022; Lukitasari et al., 2021). Epigallocatechin gallate (EGCG), the main cathecin found in green tea, has shown protective activity against hepatic fibrosis by suppressing TGF-β/SMAD signaling activity and enhancing p-PI3K/AKT activation in models of non-alcoholic fatty liver disease (NAFLD) (Xiao et al., 2014). Similarly, chlorogenic acid (CGA), the predominant phenolic compound in green coffee, has shown antifibrotic effects by activating the cyclic guanosine monophosphate (cGMP)/protein kinase G (PKG) pathway, which inhibits the nuclear translocation of p-Smad2/3 and the expression of pro-fibrotic genes in hyperglycemia-induced cardiac fibroblasts (Qin et al., 2021). However, the anti-fibrotic efficacy and mechanistic basis of a combined decaffeinated green coffee and green tea extracts (GCGTE) on fibroblast activation remain unexplored.

Building on this background, the present study aimed to evaluate the protective effects of GCGTE against fibroblast activation under high-glucose condition. In addition, we sought to elucidate the associated cellular and molecular mechanisms by assessing cell viability, relative cell size, αSMA expression, collagen type I secretion, and pAKT expression in NIH-3T3 fibroblasts exposed to high-glucose. The results of this study are expected to provide additional evidence regarding the potency of novel plant-based supplementary treatment for the prevention of fibrosis in the diabetics.

2. Material and Methods

3. Results

3.1 GCGTE exposure did not reduce NIH-3T3 cells viability after 24 hours

GCGTE effect on NIH-3T3 cell viability was evaluated using the MTT assay. Fig. 1 represents the percentage of NIH-3T3 cells viability following 24-hours exposure to different concentrations of GCGTE (40/40, 80/80, 160/160, and 320/320 µg/mL). The results indicate that cell viability remained above 90% across all tested GCGTE concentrations, with no significant cytotoxic effects observed compared to the control (0 µM), suggesting that GCGTE treatment within the tested range does not compromise cell viability. Therefore, GCGTE at 160/160 and 320/320 µg/mL were employed for subsequent experiments.

Close

Fig. 1. The effect of GCGTE on NIH-3T3 cells viability analyzed using MTT assay. NIH-3T3 cells were cultured in 96-well plates and treated for 24 hours under the indicated conditions: (a) complete medium (control); (b-e) GCGTE 40/40, 80/80, 160/160, and 320/320 µg/mL for 24 hours. Viable cells are indicated by the formation of purple formazan crystal (red arrow) (f) The histogram represents the percentages of viable cells relative to untreated controls. Data are expressed as mean ± SD from three technical replicates. GCE: green coffee extract, GTE: green tea extract.

Export to PPT

3.2 GCGTE prevented hypertrophy of HG-treated NIH-3T3 cells

To assess the effect of high glucose on NIH-3T3 cell size, flow cytometric analysis was performed using forward scatter (FSC) as an indicator of relative cell size. Representative FSC histograms for each experimental group are shown in Fig. 2a–c, while the quantitative analysis is presented in Fig. 2d. Cells cultured under high-glucose conditions exhibited an approximately 15% increase in cell size compared with the normal-glucose group. Pre-treatment with metformin resulted in a further increase in cell size (approximately 23–24%). In contrast, pre-treatment with GCGTE at both 160/160 µg/mL and 320/320 µg/mL maintained cell size at levels comparable to the normal-glucose group. These findings indicate that high glucose induces hypertrophic changes in NIH-3T3 fibroblasts, which can be attenuated by GCGTE pre-treatment.

Close

Fig. 2. Effects of high glucose and GCGTE pre-treatment on NIH-3T3 cell size. (a-c) Representative forward scatter (FSC) histograms showing cell size distribution in different treatment groups. (d) Quantification of relative FSC values relative to NG group. Data are presented as mean ± SD of three technical replicates; *p < 0.05 vs. NG, #p < 0.05 vs. HG. NG: normal glucose, HG: high glucose, M: metformin, GCGTE: combined decaffeinated green coffee and green tea extract.

Export to PPT

3.3 GCGTE decreased collagen type-I protein levels in HG-treated NIH-3T3 cells’ supernatant

The concentration of collagen type I protein in NIH-3T3 cell culture supernatants under different treatment conditions was depicted in Fig. 3. The results indicate that the HG group exhibited a significant increase in collagen type I secretion compared to the NG control group. Treatment with metformin and GCGTE at all tested concentrations significantly reduced collagen type I levels compared with HG group, indicating a potential inhibitory effect of GCGTE on excessive collagen secretion induced by high glucose conditions.

Close

Fig. 3. The effect of GCGTE on secreted collagen type-I protein levels in HG-treated NIH-3T3 cells. Normal glucose (NG); the high-glucose (HG) group; 50 µM of metformin (M) group; GCGTE at 160/160 and 320/320 µg/mL. Each value is expressed as mean ± SD of three technical replicates. *p < 0.05 vs. NG; #p < 0.05 vs. HG. NG: normal glucose, HG: high glucose, M: metformin, GCGTE: combined decaffeinated green coffee and green tea extract.

Export to PPT

3.4 GCGTE decreased αSMA protein expression in HG-treated NIH-3T3 cells

αSMA protein expressions were assessed by immunofluorescence staining (Fig. 4a) and measured semi-quantitatively using ImageJ (Fig. 4b). The results show that αSMA expression in the cells cytoplasm was significantly increased in the HG group compared to the NG control. On the contrary, pre-treatment with metformin and GCGTE significantly reduced αSMA expression compared to the HG group. However, no dose-dependent effect was observed.

Close

Fig. 4. Expression of αSMA in NIH3T3 cells was detected using immunofluorescence (magnification ×200). (a) Representative immunofluorescence images showing αSMA expression under different treatment conditions (original magnification ×200). (b) Quantification of relative fluorescence intensity. Data are presented as mean ± SD of five representative images obtained from three technical replicates per group. *p < 0.05 vs. NG; #p < 0.05 vs. HG.

Export to PPT

3.5 GCGTE decreased MMP-9 protein expression in HG-treated NIH-3T3 cells

MMP-9 protein expressions were assessed by immunofluorescence staining (Fig. 5a). The semi-quantitative analysis of MMP-9 expression (Fig. 5b) demonstrated that MMP-9 expression was significantly increased in the cytoplasm of NIH-3T3 from HG group compared to the control, whereas both GCGTE at 160/160 and 320/320 µg/mL significantly reduced MMP-9 expression compared to the HG group. However, treatment with metformin did not affect MMP-9 expression.

Close

Fig. 5. Expression of MMP-9 in NIH3T3 cells was detected using immunofluorescence (magnification ×200) (a) Representative immunofluorescence images showing MMP-9 expression under different treatment conditions (original magnification ×200). (b) Quantification of relative fluorescence intensity. Data are presented as mean ± SD of five representative images obtained from three technical replicates per group. *p < 0.05 vs. NG; #p < 0.05 vs. HG.

Export to PPT

3.6 GCGTE increased pAKT protein expression in HG-treated NIH-3T3 cells

The expression of pAKT protein in the cytoplasm was evaluated by immunofluorescence staining (Fig. 6a) and measured semi-quantitatively using ImageJ (Fig. 6b). The results indicated that pAKT expression was significantly reduced in the HG group compared to the control, reflecting that HG treatment downregulated AKT activation. In contrast, pre-treatment with metformin and GCGTE at 160/160 and 320/320 µg/mL significantly abolished this effect.

Close

Fig. 6. Expression of pAKT in NIH3T3 cells was detected using immunofluorescence (magnification ×200) (a) Representative immunofluorescence images showing pAKT expression under different treatment conditions (original magnification ×200). (b) Quantification of relative fluorescence intensity. Data are presented as mean ± SD of five representative images obtained from three technical replicates per group. *p < 0.05 vs. NG; #p < 0.05 vs. HG.

Export to PPT

4. Discussion

In this study, we evaluated the protective effects of GCGTE on NIH-3T3 fibroblasts exposed to high-glucose conditions, by measuring cell viability, collagen type I secretion, the expression of αSMA, MMP-9, and pAKT proteins, as well as cell size. Our results demonstrate that high glucose exposure significantly increased collagen type I secretion and upregulated αSMA and MMP-9 expression, decreased pAKT expression, and induced cellular hypertrophy, collectively indicating fibroblast activation toward the myofibroblast phenotype. Moreover, pretreatment with GCGTE at two different concentrations effectively attenuated these effects. These findings suggest that GCGTE can inhibit hyperglycemia-induced fibrotic signaling and fibroblast activation, which to our knowledge, have not been reported by previous studies.

Hyperglycemia is a major contributor to diabetic fibrosis, largely through the activation of pro-fibrotic pathways in fibroblasts (Tuleta and Frangogiannis, 2021). The 24-hours HG exposure used in this study was selected to capture the early molecular and phenotypic events associated with fibroblast activation, rather than late-stage ECM accumulation. Several studies have demonstrated that 24 h of HG exposure is sufficient to induce measurable increases in key pro-fibrotic markers, including αSMA and collagen type I. For instance, exposure of cardiac fibroblasts to HG for 24 hours significantly increases αSMA protein expression, indicating the initiation of myofibroblast differentiation, while concomitantly upregulating collagen I expression at mRNA level (Shamhart et al., 2014). Similarly, induction of collagen I and TGF-β1 mRNA expression has been reported in human peritoneal fibroblasts following 24 hours of HG exposure (Saed and Diamond, 2003). Furthermore, neonatal rat cardiac fibroblasts exposed to HG for 24 hours exhibit significant increases in collagen type I and III protein levels in the conditioned media, as assessed by ELISA (Tang et al., 2007). Collectively, these findings support the suitability of a 24 h HG exposure for detecting early pro-fibrotic responses in fibroblast models.

Consistent with these reports, our study demonstrated increased cell size, αSMA expression, and collagen type I secretion under HG conditions, confirming fibroblast activation toward a highly secretory and contractile myofibroblast phenotype. Previous reports have shown that high glucose promotes fibroblast hypertrophy, characterized by increased cell size and protein content, which is often followed by excessive deposition of ECM proteins relevant to fibrosis progression (Chao et al., 2010; Wu and Derynck, 2009). Morphologically, myofibroblasts can be distinguished from fibroblasts by their larger cell size, with earlier studies on 1000 individual cells reporting that myofibroblast area may exceed that of fibroblasts by more than five-fold (Hillsley et al., 2022). Consistently, we observed a ∼15% increase in NIH-3T3 cell size under high-glucose conditions compared with normal glucose, representing a modest but statistically significant effect.

In parallel, αSMA a well-established marker of fibroblast activation that mediates contractility and tissue stiffening in fibrosis (Younesi et al., 2024) was upregulated, further confirming the acquisition of a myofibroblast-like phenotype. Together, these findings validate our experimental model of hyperglycemia-induced fibroblast activation and support the concept that hyperglycemia promotes fibrosis by inducing fibroblast activation into myofibroblast through canonical TGF-β/Smad signaling (Li et al., 2003). This is consistent with earlier studies in other fibroblasts cell lines, including cardiac fibroblasts, renal fibroblasts, and hepatic stellate cells, showing that high-glucose environments upregulate αSMA expression and subsequent collagen production via TGF-β signaling and oxidative stress pathways (Lam et al., 2003; Shamhart et al., 2014; Vivar et al., 2021; Weiskirchen, 2025).

To further explore potential upstream mechanisms underlying high-glucose–induced pro-fibrotic changes, this study evaluated the expression of MMP-9 and pAKT. Under HG conditions, MMP-9 expression was markedly increased, whereas pAKT levels were significantly reduced. Although MMP-9 is classically recognized as an ECM–degrading enzyme, accumulating evidence indicates that its role in fibrosis is multifaceted. In fibroblasts, MMP-9 has been shown to facilitate the activation of latent TGFβ, thereby amplifying pro-fibrotic signaling through the canonical TGF-β/Smad pathway (Kobayashi et al., 2014; Wu and Derynck, 2009). Consistently, pharmacological inhibition of MMP-9 has been reported to reduce TGF-β activation by approximately 60–70% under HG conditions (Wu and Derynck, 2009). In addition, both the active and pro-enzyme forms of MMP-9 have been implicated in the upregulation of αSMA expression, suggesting a more direct association with fibroblast activation (Dayer and Stamenkovic, 2015). Concurrently, suppression of pAKT observed under HG conditions provides additional mechanistic insight. Several studies have demonstrated that hyperglycemic conditions suppress PI3K/AKT signaling, which is associated with increased FoxO1 activity and subsequent upregulation of connective tissue growth factor (CTGF) expression, which promotes the activation of the Smad pathway through the non-canonical signaling axis (Vivar et al., 2021). Accordingly, the concurrent increase in MMP-9 and decrease in pAKT expression observed in the present study might indicate that the pro-fibrotic milieu under HG conditions is associated with the modulation of the MMP-9/TGF-β/Smad and PI3K/AKT/FoxO1/CTGF-related pathway.

Conversely, our findings indicate that GCGTE exerts anti-fibrotic effects, as reflected by attenuation of fibroblast hypertrophy, reduced αSMA expression, and decreased secretion of collagen type I. The observed reduction in cell size may reflect preservation of a more quiescent fibroblast phenotype, characterized by diminished cellular capacity to express αSMA and secrete ECM proteins, particularly collagen type I (Chao et al., 2010; D’Urso and Kurniawan, 2020; Hillsley et al., 2022). These effects are likely to be associated with modulation of the TGF-β/Smad signaling axis or its upstream regulatory mechanisms. In support of this notion, MMP-9 expression was reduced in GCGTE-treated groups, which may be linked to decreased activation of latent TGF-β and subsequent attenuation of downstream pro-fibrotic responses, including αSMA expression and collagen secretion (Wu and Derynck, 2009). In parallel, increased pAKT expression observed in GCGTE-treated groups suggest normalization of PI3K/AKT signaling. Increased AKT phosphorylationis expected to promote FoxO1 phosphorylation, thereby limiting its nuclear localization and suppress transcription of pro-fibrotic mediators such as CTGF (Vivar et al., 2021). Collectively, these findings suggest that GCGTE is associated with concurrent modulation of MMP-9/TGF-β/Smad and PI3K/AKT/FoxO1/CTGF–related pathways, which may underlie its observed anti-fibrotic effects under high-glucose conditions.

Mechanistically, these effects are likely attributable to the bioactive polyphenols present in green tea and green coffee extracts. Green tea contains high level of EGCG, which has been shown to suppress TGF-β/Smad signaling and enhances PI3K/AKT activity in fibrotic models (Xiao et al., 2014). Other report also revealed that EGCG decreased collagen type I secretion and CTGF expression in dermal fibroblasts derived from systemic sclerosis patients (Dooley et al., 2010). This anti-fibrotic effect seems to be mediated, at least partially, by the activating of G-protein coupled estrogen receptor (GPER), which subsequently blocks TGF-β/Smad signaling (Garate-Carrillo et al., 2021). Furthermore, chlorogenic acid (CGA), the dominant polyphenol in green coffee, has been reported to inhibit p-Smad2/3 nuclear translocation via activation of the cGMP/PKG pathway. Besides, it also attenuates pro-fibrotic gene expression as well as increasing nitric oxide and endothelial nitric oxide synthase production in cardiac fibroblast of hyperglycemia mice model (Qin et al., 2021). Additionally, anti-fibrotic effect of CGA may involve many signaling pathways, including miR-21/Smad7/TGF-β1/Interleukin-13 (IL-13) Toll-Like Receptor 4 (TLR-4)/HIF-α/miR-122 (Nwafor et al., 2022). Together, these results indicate that green coffee and green tea extracts may synergistically modulate multiple signaling cascades involved in fibroblast activation which prompts their potential as supplementary treatment to prevent fibrosis progression in diabetes mellitus patients.

In present study, metformin pre-treatment also exerted anti-fibrotic effects, as indicated by reduced α-SMA expression and collagen type I secretion, accompanied by restoration of pAKT levels. However, metformin did not significantly reduce MMP-9 expression, suggesting that its anti-fibrotic effects in this model may occur independently of the MMP-9/TGF-β axis. Instead, metformin may primarily act through modulation of PI3K/AKT/FoxO1/CTGF signaling or by directly interfering with canonical TGF-β/Smad pathways. Under hyperglycemic conditions, reduced AKT phosphorylation enhances FoxO1 activity and CTGF expression; metformin has been shown to restore AKT phosphorylation, thereby inactivating FoxO1 and attenuating CTGF-driven fibrotic signaling (Kominato et al., 2022). In addition, metformin has been reported to directly bind the receptor-binding domain of TGF-β1, impairing its downstream signaling (Xiao et al., 2016).

Interestingly, despite these molecular anti-fibrotic effects, metformin increased fibroblast cell size beyond that observed with HG alone. This finding contrasts with previous reports describing anti-hypertrophic effects of metformin mediated via activated protein kinase (AMPK) activation and subsequent inhibition of mechanistic Target of Rapamycin (mTOR)-dependent cell growth (Bai et al., 2013). Under hyperglycemic conditions, however, metabolic stress and redox imbalance may reduce AMPK sensitivity, thereby permitting persistent mTOR activity which contribute to cellular hypertrophy. Moreover, HG–induced fibroblast hypertrophy may occur independently of phosphoinositide 3-kinase/protein kinase B (PI3K/AKT)/mTOR signaling and involve alternative growth-related pathways such as extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK) or p38 mitogen-activated protein kinase (MAPK), which can negatively regulate AMPK activity (Yuan et al., 2020). Collectively, these findings suggest that fibroblast hypertrophy and fibrotic activation are partially exclusive processes, and that metformin can attenuate fibrotic signaling without necessarily preventing cell hypertrophy in this experimental context.

Lastly, several limitations of present study must be acknowledged. First, the experiments were conducted using a short-term in vitro hyperglycemia model on murine fibroblast cell line NIH-3T3. Although this model is appropriate for investigating fibroblast responses to acute glucose stress and has been shown to induce significant increases in pro-fibrotic markers in previous studies, it does not fully capture the chronic and complex nature of diabetic fibrosis in vivo. Consequently, this approach may underestimate the magnitude and complexity of ECM accumulation observed under prolonged diabetic conditions. Second, while the study demonstrated modulation of key fibrotic markers, no specific pathway inhibitors were employed; thus, causal relationships between the observed molecular changes and fibroblast activation could not be firmly established. Third, while GCGTE appears to have protective effect, no dose-dependent response was observed. This limitation may be attributable to the relatively narrow range of concentrations tested in the present study.

5. Conclusions

Taken together, our data demonstrate that GCGTE attenuates high-glucose-induced fibroblast activation in NIH-3T3 cells, which is associated with concurrent modulation of key pro-fibrotic signaling pathways, including MMP-9/TGF-β/Smad and PI3K/AKT.