Scholarlink

<p>Identifying natural compounds that exhibit selective toxicity toward cancer cells while sparing healthy tissues remains a critical challenge in pharmacology. The primary aim of this study is to investigate the selective cytotoxic and oxidative potential of <i>Ceratophyllum demersum</i> on hepatocellular carcinoma (HepG2) and normal hepatocyte (THLE2) cells, using machine learning models to decode the underlying oxidative drivers of this selectivity. Lactate dehydrogenase (LDH) release assays revealed dose-dependent cytotoxicity in both cell lines. The water extract caused a 79.10% increase in LDH activity at 400&#xa0;µg/mL in HepG2 cells, compared to a 67.28% increase in THLE2 cells, indicating selective toxicity. The methanol extract exhibited less selectivity, with LDH increases of 69.04% and 55.28% in HepG2 and THLE2 cells, respectively, at the same concentration. IC<sub>50</sub> values for the water extract were calculated as 95.29&#xa0;µg/mL for HepG2 and 165.44&#xa0;µg/mL for THLE2. For methanol extract, IC<sub>50</sub> values were 196.82&#xa0;µg/mL (HepG2) and 271.87&#xa0;µg/mL (THLE2). Total antioxidant status (TAS) levels peaked at low doses (12.5–50&#xa0;µg/mL). Conversely, total oxidant status (TOS) levels rose sharply, with maximum oxidative stress observed at 200&#xa0;µg/mL water extract in HepG2 cells. A strong positive correlation (r = 0.97) was found between LDH and TOS, while LDH and TAS showed a negative correlation (r = –0.80) for HepG2 cells. Heatmap and hierarchical clustering analyses further highlighted extract-specific cellular responses. Machine learning analyses using random forest regression, gradient boosting regression, and extreme gradient boosting identified extract concentration and TOS as key predictors of LDH levels, with SHapley Additive exPlanations analysis supporting these findings. The integration of machine learning provided deeper insight into the relationship between oxidative stress and cytotoxicity, highlighting its potential as a powerful tool in phytochemical bioactivity research.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning-assisted evaluation of the cytotoxic and oxidative effects of Ceratophyllum demersum extracts on HepG2 and THLE2 cells

  • 期刊论文

  • 发表日期：

    2026年6月3日

查看全文
  • Mustafa Ari,
  • Bugrahan Emsen,
  • Muhammet Dogan,
  • Kagan Sarlar

摘要

Identifying natural compounds that exhibit selective toxicity toward cancer cells while sparing healthy tissues remains a critical challenge in pharmacology. The primary aim of this study is to investigate the selective cytotoxic and oxidative potential of Ceratophyllum demersum on hepatocellular carcinoma (HepG2) and normal hepatocyte (THLE2) cells, using machine learning models to decode the underlying oxidative drivers of this selectivity. Lactate dehydrogenase (LDH) release assays revealed dose-dependent cytotoxicity in both cell lines. The water extract caused a 79.10% increase in LDH activity at 400 µg/mL in HepG2 cells, compared to a 67.28% increase in THLE2 cells, indicating selective toxicity. The methanol extract exhibited less selectivity, with LDH increases of 69.04% and 55.28% in HepG2 and THLE2 cells, respectively, at the same concentration. IC50 values for the water extract were calculated as 95.29 µg/mL for HepG2 and 165.44 µg/mL for THLE2. For methanol extract, IC50 values were 196.82 µg/mL (HepG2) and 271.87 µg/mL (THLE2). Total antioxidant status (TAS) levels peaked at low doses (12.5–50 µg/mL). Conversely, total oxidant status (TOS) levels rose sharply, with maximum oxidative stress observed at 200 µg/mL water extract in HepG2 cells. A strong positive correlation (r = 0.97) was found between LDH and TOS, while LDH and TAS showed a negative correlation (r = –0.80) for HepG2 cells. Heatmap and hierarchical clustering analyses further highlighted extract-specific cellular responses. Machine learning analyses using random forest regression, gradient boosting regression, and extreme gradient boosting identified extract concentration and TOS as key predictors of LDH levels, with SHapley Additive exPlanations analysis supporting these findings. The integration of machine learning provided deeper insight into the relationship between oxidative stress and cytotoxicity, highlighting its potential as a powerful tool in phytochemical bioactivity research.