Almonertinib was provided by Jiangsu Hausen Pharmaceutical Co., Ltd. Baicalein (CAS:491-67-8, MCE) was obtained from MedChemExpress (Shanghai, China). Cell counting kit-8 (CCK-8) was procured from APExBIO (Houston, USA). The annexin V-FITC/PI staining kit was purchased from Bestbio (Shanghai, China). The EdU, BCA, and reactive oxygen species (ROS) kits were supplied by Beyotime (Shanghai, China). The antibody against PARP (#T40050) was purchased from Abmart. The antibodies of GAPDH (60004-1-Ig) and β-tubulin (10094-1-AP) were purchased from Proteintech. The antibody against PI3K (BS6052) was obtained from Bioward. The antibodies of p-PI3K (#4249), Akt (#4691), p-Akt (#4060), Caspase-3 (#14220), Caspase-9 (#9502), cleaved Caspase-3 (#9661), and cleaved Caspase-9 (#7237) were supplied by Cell Signaling Technology.

H1975 and HCC827 cells in the logarithmic growth phase (with cell fusion of 70%–80%) were collected and exposed to 2 μM of almonertinib-containing medium initially. After 24 h, the medium was discarded and cells were rinsed twice with phosphate-buffered saline (PBS) before switching to a normal medium without almonertinib. Once the cells resumed their growth, they were subcultured. When the cells achieved a certain level of fusion, 2 μM of almonertinib was reapplied. Once the cells stably grew at this drug concentration, the concentration of almonertinib was gradually increased for continued cultivation. Cells were typically cultured for 1–2 weeks at each concentration level. As the drug concentrations were continuously increased, constant monitoring of the cell sensitivity was required to ensure that the cells retained their drug resistance. The drug induction process lasted for 6 months. The H1975 and HCC827 parent cells and their almonertinib-resistant equivalents grew in RPMI-1640 medium with 1% penicillin/streptomycin solution and 10% fetal bovine serum (FBS) in a constant-temperature incubator at 37°C and 5% CO₂.

The cells were seeded in 96-well plates at 5,000 cells per well. Twenty-four hours later, the cells were subjected to 0, 1, 2, 4, 8, 16, and 32 μM of almonertinib or 0, 20, 40, 60, 80, and 100 μM of baicalein; a blank group and a control group were also set up simultaneously. After 24 h of cultivation, the cell media were discarded carefully and the CCK-8 kit was added to the wells; then, the cells were incubated in the dark at 37°C for 30 min. The optical density (OD) value (at 450 nm) was obtained, and the cell viability was analyzed according to the OD value as described previously (Lei et al., 2023). The half maximal inhibitory concentration (IC₅₀) value was acquired based on the OD value using the GraphPad Prism 8 software.

The concentration range of almonertinib or baicalein was premeasured via the IC₅₀ value of each agent, and the cell viability was determined based on the concentration of the two drugs (almonertinib: 1, 2, 4, 8, 16, and 32 μM; baicalein: 40, 60, 80, 100, and 120 μM). Twenty-four hours later, the cell viabilities were analyzed with the CCK-8 kit. The online software SynergyFinder (https://synergyfinder.fimm.fi) was used to obtain the zero potency (ZIP) based on the OD value. Synergism was considered when the ZIP synergy scores exceeded 0, and strong synergistic effects were assumed when the scores exceeded 10.

Each well of a 6-well plate was inoculated with 1×10³ cells. After 3 days in the incubator, fresh media containing almonertinib or/and baicalein were added for another 10–12 days. When the colonies were uniformly formed, the cells were processed as mentioned previously, and the counts of the cell colonies were analyzed using ImageJ software.

Each well of a 24-well plate was seeded with 5×10⁴ cells. The cells were processed according to the manufacturer’s instructions and examined with a fluorescence microscope. The ImageJ software was applied to evaluate the percentage of EdU-positive cells.

The Annexin V-FITC/PI staining was used to observe the apoptosis. Each well of a 6-well plate was inoculated with 2×10⁵ cells and supplied with almonertinib (8 μM) or/and baicalein (80 μM or 60 μM) for 24 h. The cells were processed according to the manufacturer’s instructions and detected using a flow cytometer, then subsequent determination with FlowJo software.

The cells were placed in culture plates with appropriate cell densities. After incubation with almonertinib (8 μM) or/and baicalein (80 μM or 60 μM) for 24 h, the cells were lysed on ice in RIPA buffer. After measuring the protein concentration via the bicinchoninic acid protein assay kit, SDS-PAGE was performed to separate the equivalent proteins from all groups; the proteins were moved to membranes that were then blocked with 5% skim milk, treated with primary antibodies, reacted with secondary antibodies, visualized using ECL chemiluminescent reagent, photographed with a gel imaging instrument, and quantified for the bands using ImageJ software. GAPDH or β-tubulin was used as the loading reference.

The fluorescent probe 2′,7′-dichlorofluorescin diacetate (DCFH-DA) was employed to evaluate the intracellular ROS level. The cells grown in 6-well plates and processed with almonertinib (8 μM) or/and baicalein (80 μM or 60 μM) for 24 h were subsequently treated with DCFH-DA (10 μM) for 20 min, and the level of DCF fluorescence of the cells was determined via flow cytometry. To prevent ROS production, a ROS remover like N-acetyl-L-cysteine (NAC) (2 mM) was applied for 1 h before treatment with almonertinib or/and baicalein to the cells. The results were then analyzed with FlowJo software.

Five-week-old female BALB/c nude mice of 18–20 g body weight were supplied by Hangzhou Ziyuan Experimental Animal Technology Co., Ltd. (experimental animal quality certificate of license No. SCXK2019-0004). All mice were fed under specific pathogen-free conditions, and food and water were made freely available in the cage. The temperature and humidity were maintained at (25 ± 2)°C and (50 ± 5)%, respectively. The animal study was approved by the Ethics Committee of Bengbu Medical University (approval no. 2022A170). A total of 5×10⁶ HCC827/AR cells were inoculated into each of the nude mice subcutaneously. Seven days later, each animal was assigned to one of four groups at random: control group, baicalein (20 mg/kg) group, almonertinib (15 mg/kg) group, and baicalein (20 mg/kg) combined with almonertinib (15 mg/kg) group. The body weight of each mouse and its tumor size were determined once every 3 days, and the tumor volume was obtained as follows: V = length × width × width/2. Twenty-four days later, tumor samples were collected from the mice for the TUNEL assay and immunohistochemistry (IHC) analysis; the liver, kidney, heart, spleen, and lung of each mouse were obtained for hematoxylin and eosin (HE) staining; blood was also drawn for detecting the levels of alanine aminotransferase (ALT), aspartate transaminase (AST), blood urea nitrogen (BUN), and creatinine (Cr).

The tumor slides were processed with protease K and reacted with Alexa Fluor 488-12-dUTP in the dark. The stained slides were dyed with DAPI solution and subsequently observed under a microscope.

The tissue specimens were made transparent using xylene before cooling in melted paraffin and solidifying to wax blocks that were cut into slices of thickness 4 μm. The slices were subject to HE staining as reported previously (Ma et al., 2024) and observed and photographed under a microscope.

The tumor sections were subject to 3% hydrogen peroxide and blocked with serum; subsequently, they were probed overnight with anti-Ki67, anti-cleaved Caspase-3, anti-p-PI3K, and anti-p-Akt antibodies before being reacted with secondary antibodies, followed by DAB staining and examination under a confocal microscope.

The data were evaluated and plotted with GraphPad Prism 8 and SPSS 27 software; the numerical data were exhibited as mean ± SD, and the one-way ANOVA and LSD t-test were carried out to observe the differences among the groups. The value of p < 0.05 was used to indicate differences with statistical significance.

The HCC827/AR and H1975/AR cells were derived from the HCC827 and H1975 cells (Wuhan Punosai Technology Co., Ltd.), respectively, by supplying almonertinib at increasing concentrations for about 6 months (Figure 1A). We found that both the HCC827 and H1975 parent cells, as well as the almonertinib-resistant cells, adhered to the walls. However, the HCC827/AR and H1975/AR cells were larger than their parent cells and displayed cluster-like growths, with some cells having elongated pseudopodia and irregular edges (Figure 1B). Moreover, CCK-8 assay was used to detect the resistances of the HCC827/AR and H1975/AR cells to almonertinib. The findings indicated that almonertinib dramatically decreased the survival rates of the HCC827 and H1975 cells and that high concentrations of almonertinib inhibited the survival rates of the HCC827/AR and H1975/AR cells. The IC₅₀ values of the HCC827 and HCC827/AR cells to almonertinib were 2.59 ± 0.30 μM and 14.9 ± 1.38 μM, respectively, and the resistance index (RI) was 5.78 ± 0.65; the IC₅₀ values of the H1975 and H1975/AR cells to almonertinib were 2.79 ± 0.49 μM and 12.67 ± 0.83 μM, respectively, and the RI was 4.63 ± 0.81 (Figure 1C). These findings implied that the drug-resistant cells were established successfully.

In an effort to explore the impact of baicalein on cell viability, different concentrations of baicalein were used to process the HCC827/AR and H1975/AR cells, and the cell viabilities were dramatically restrained in a concentration-dependent manner (Figure 2A). To evaluate the effects of baicalein combined with almonertinib, baicalein (40, 60, 80, 100, and 120 μM) and almonertinib (0, 1, 2, 4, 8, 16, and 32 μM) were used to treat the cells; in the HCC827/AR and H1975/AR cells, cell proliferation was found to be remarkably suppressed. ZIP synergy scores were obtained with the SynergyFinder software, and the scores for the HCC827/AR and H1975/AR cells were 14.228 and 13.987, respectively (Figure 2B), indicating that almonertinib and baicalein had highly synergistic effects in suppressing cancer cell proliferation. Meanwhile, baicalein combined with almonertinib dramatically elevated the inhibition of cell viability (Figure 2C), and the EdU-positive cells (Figure 2D) were obviously reduced compared to the group treated with almonertinib or baicalein alone. Furthermore, the results of the colony formation assay suggested that baicalein and almonertinib in combination markedly declined the formation of colonies (Figure 2E). The above results indicated that baicalein and almonertinib in combination could obviously suppress the proliferation of almonertinib-resistant NSCLC cells.

We further explored the combined influences of baicalein and almonertinib on apoptosis in the HCC827/AR and H1975/AR cells, and the apoptosis rates of the cells were detected with the annexin-V/PI staining assay. Thus, the apoptosis rates of the cells processed with baicalein and almonertinib in combination were remarkably higher than that of the cells processed with a single drug (Figure 3A). Then, we conducted Western blotting to investigate the levels of several apoptosis-associated proteins and found that baicalein combined with almonertinib enhanced the cleaved Caspase-3, cleaved PARP, and cleaved Caspase-9 expressions in both H1975/AR and HCC827/AR cells, while decreasing the Caspase-3, PARP, and Caspase-9 expressions (Figure 3B). The above findings implied that baicalein and almonertinib triggered the activation of Caspase and then synergistically induce apoptosis.

The xenograft model experiments were performed with nude mice to ascertain the findings of the in vitro experiments. To determine the combined influences of baicalein and almonertinib on almonertinib-resistant NSCLC cells in vivo, a mouse xenograft model was established using the HCC827/AR cells. Compared to baicalein or almonertinib alone, the combination of two drugs significantly inhibited tumor growth (Figures 4A–C), even as no statistical differences were observed in the body weight of the nude mice (Figure 4D). To further observe the toxicity of liver and kidney functions in the nude mice, ELISA was applied to determine the liver and kidney function indicators, and the results suggested that both single and combined administrations of baicalein and almonertinib did not lead to kidney or liver toxicity (Figure 4E). Furthermore, the TUNEL staining results showed that baicalein combined with almonertinib dramatically elevated apoptosis in the tumor tissues in comparison with the groups receiving baicalein or almonertinib alone (Figure 4F). Otherwise, HE staining was employed to observe the structures and morphologies of the liver, spleen, heart, lung, and kidney, which displayed that these structures were intact without pathological changes (Figure 4G). Finally, we examined the level of proliferation indicator Ki67 and apoptosis indicator cleaved Caspase-3; in the tumor tissues, the findings implied that baicalein combined with almonertinib inhibited Ki67 expression and elevated the activation of Caspase-3 (Figure 4H), which were similar to the results acquired from the in vitro experiments. These findings showed that baicalein synergistically elevated the anticancer effects of almonertinib in vivo and that no significant hepatorenal toxicity was observed in any of the groups.

To explore the associations between baicalein and lung cancer, a total of 190 baicalein targets and 6,737 non-duplicate lung cancer targets were intersected to create a Venn diagram (Figure 5A), resulting in 157 intersecting target genes. A bar graph of the top 10 GO entries involving biological process (BP), cellular component (CC), and molecular function (MF) was created (Figure 5B). BP primarily covered reactions to oxidative stress, cellular responses to oxidative stress, and responses to ROS. Generally, ROS were produced and accumulated under oxidative stress, and the intracellular ROS affected various biological processes, including proliferation and apoptosis, by altering the equilibrium between cell survival and death processes. ROS generation was detected via DCFH-DA dyeing, followed by flow cytometry analysis of the ROS levels in the H1975/AR and HCC827/AR cells. As displayed in Figure 5C, the intracellular ROS levels in the combination group were dramatically higher than those in the single-drug groups.

To discuss the influences of ROS on proliferation inhibition and apoptosis induction in almonertinib-resistant NSCLC cells by baicalein combined with almonertinib, we hypothesized that ROS was a critical upstream mediator. The ROS inhibitor N-acetylcysteine (NAC) was used for cell pretreatment, and we found that NAC could partially reverse the combined influences of baicalein and almonertinib on proliferation inhibition (Figure 6A) and apoptosis induction (Figure 6B) in the HCC827/AR and H1975/AR cells. Meanwhile, the cleavages of PARP, Caspase-3, and Caspase-9 were also eliminated by NAC (Figure 6C). In summary, baicalein combined with almonertinib resulted in proliferation inhibition and apoptosis induction depending on ROS production.

The P13K/Akt pathway broadly exists in a wide range of cancers and exhibits essential modulating roles in cell survival, angiogenesis, and proliferation. As described in Section 3.5, we found 157 intersecting targets between baicalein and lung cancer. The KEGG and PPI pathways were employed to find the targets of baicalein for the treatment of lung carcinoma, and we found that the PI3K/Akt pathway played an essential role in treatment with baicalein in lung carcinoma (Figure 7A); the PPI network was created, as displayed in Figure 7B. Therefore, we speculate that the PI3K/Akt pathway serves as a vital pathway in the anticancer activity of baicalein combined with almonertinib in almonertinib-resistant NSCLC cells. To confirm this hypothesis, the levels of this pathway-associated protein in HCC827/AR and H1975/AR cells processed with baicalein and/or almonertinib were investigated. The findings revealed that the p-Akt and p-PI3K expressions were markedly downregulated in the groups treated with baicalein and/or almonertinib (Figure 7C). To further confirm the regulatory effects of baicalein combined with almonertinib on the PI3K/Akt pathway, IHC analysis was carried out to detect p-Akt and p-PI3K in the tumor tissues. The results showed that baicalein combined with almonertinib diminished the levels of p-Akt and p-PI3K, which were similar to the in vitro experimental results (Figure 7D). Furthermore, after pretreatment with NAC, the activation of PI3K and Akt was repressed, indicating that ROS participated in the PI3K/Akt pathway (Figure 7E). The above findings revealed that baicalein combined with almonertinib obviously suppressed activation of the PI3K/Akt pathway through ROS accumulation.