**3. Discussion**

Despite decades of good results in the clinical application of DOX in cancer therapy, this drug induces cumulative, dose-dependent adverse effects. Our previous studies have demonstrated that ethanol and water extracts obtained from *M. aquifolium* show good anticancer potential and that berberine and berberine-type alkaloids can be detected in both extracts [14]. Higher content of berberine was detected in MAE extract (2.44%) compared to MAW extract (1.34%) (LC-MC analyses).

Furthermore, after identifying cytotoxic metabolites from *M. aquifolium* using 1H NMR-based metabolomic approach, we concluded that alkaloids with the highest cytotoxicity in our extracts are berberine, palmatine, and the bisbenzylisoquinoline alkaloid berbamine [15]. It has previously been reported that berberine and similar alkaloids can inhibit the growth of cancer cells [9], effectively limiting the toxicity of DOX [4].

Our initial screenings demonstrated that, of the tested cell lines, A549 cells were the least sensitive to the cytotoxic activity of MAE and MAW [14]. Nonsmall cell lung cancer patients often show resistance to therapy [16], and several mechanisms underlying the development of multidrug resistance in lung cancer have been identified, such as overexpression of drug efflux proteins and ATP-binding cassette (ABC) transporters [17]. Many studies have confirmed the presence of ABC transporters, breast cancer-resistant protein (BCRP), and lung resistance-related protein (LRP) in A549 cells, which have been shown to be related to anticancer drug resistance [18–20]; therefore, we selected the A549 cell line for all further experiments. The results obtained here demonstrate that the IC50 concentration of DOX can be multiply reduced (19 to 123 times) when DOX is combined with *M. aquifolium* extracts, suggesting that the same antiproliferative effects can be achieved using much lower concentrations of this drug. Furthermore, DOX and the plant extracts showed strong synergistic effects, clearly demonstrating that the individual anticancer activities of both constituents were preserved. It has been reported that cardiomyophaty, the most important adverse effect of DOX, primarily depends on the applied dose [21]. Doses below 450 mg/m<sup>2</sup> reduce the frequency of on-treatment events, but the cumulative effects lead to the development of late-onset adverse events [22,23]. Based on this, we hereby propose that coadministration of DOX with an additional agen<sup>t</sup> with a synergistic effect may decrease the toxicity of this treatment without affecting the anticancer activity of DOX. Our previous studies have shown that *M. aquifolium* extracts are several times less cytotoxic to normal, healthy cells than to cancer cells in vitro [14], indicating good selectivity and potential for their use in anticancer therapy.

We conducted the experiments on A549 cells, which are the most invasive but also the least sensitive to DOX, which is part of medical therapy in the treatment of lung cancer. Previous research has shown that both berberine and berbamine can inhibit the growth of lung cancer cells in in vivo systems [24,25], and based on these results, we can conclude that the active principles of our extracts have potential to reach this target in the body.

A549 cell cycle analysis has demonstrated that DOX induces a strong G2/M transition block [26] as well as a considerable increase in the percentage of cells in the subG1 phase in contrast to that in the control sample, indicating that treated cells cannot pass through mitosis, which ultimately leads to apoptosis [27,28]. We demonstrated that the percentage of cells in the subG1 phase was similar when cells were treated with IC50 DOX alone or with plant extracts and IC20 DOX combined, confirming that the extracts allowed maintenance of the anticancer activity of DOX even when applying lower drug doses. Furthermore, we observed an increase in the number of cells in the G1 phase following treatment of cells with DOX/plant extract combinations in contrast to the number of cells in the G1 phase treated with DOX alone. As one of the goals of drug discovery efforts today is identifying the agents that target cell cycle checkpoints responsible for the control of progression through the cell cycle [29], we believe that the results obtained may be very important. Cell percentage increase in the G1 and subG1 phases observed in samples treated with DOX/plant extract combinations may sugges<sup>t</sup> that this G1 block is irreversible and that treatment induces apoptosis, which leads to an increase in the number of cells in the subG1 phase. This significantly higher percentage in samples treated with the DOX/MAE combination than in the control sample, as well as the existing DOX-induced G2/M arrest observed after the treatment of cells with both MAE/MAW and DOX combinations, additionally confirms their synergistic effects.

Although extracts did not affect DOX uptake, we have demonstrated that they induce the retention of DOX up to 20% more than in untreated samples. Increasing drug dose to overcome drug resistance in cancer therapy is not feasible due to numerous potential side effects [30], and alternative approaches include improving accumulation, prolonging retention of drugs in cancer cells [31], and reducing drug exposure time [32]. The inhibition of drug efflux transporters p-glycoprotein (Pgp) and BCRP restores the intracellular levels of drug in DOX-resistant osteosarcoma cells and leads to the retention of DOX [33]. Our research suggests that the investigated extracts inhibit one or more of these proteins and induce the retention of DOX. Furthermore, this may be a mechanism underlying the activation of apoptosis and increase in the subG1 phase cell numbers after the DOX/extract treatment [33]. Our further studies should clarify the mechanisms of DOX retention.

The inhibitors of matrix metalloproteinases are considered potential novel agents able to inhibit tumor growth and metastases, but they were shown to be unsuccessful in several clinical trials, which may result from the dual role of matrix metalloproteinases during cancer cell invasion and metastases [34]. These enzymes can degrade extracellular matrix as well as promote cancer cell invasion, migration, and neovascularization [35], but, on the other hand, they are able to reduce cancer growth and vascularization by inducing the generation of angiogenesis inhibitors (angiostatin and tumstatin) [35–38]. Scientists [39] concluded that MMP-2 and MMP-9 drive metastatic pathways, migration, viability, and secretion of angiogenic factors in two cell lines representing the metastatic and nonmetastatic forms of retinoblastoma cells. The observed inhibition of MMP9 expression in cells treated with the extracts was maintained even after combining these extracts with DOX. Chen et al. [34] showed that the increase in plasma levels of MMP9 promotes tumorigenicity in vivo, and these tumors are smaller and less vascularized compared with those grown in mice with lower MMP9 levels, which is explained by MMP9-induced angiostatin synthesis. MMP9 inhibitors lead to a decrease in the number of tumor colonies, but tumors in vivo are larger and more vascularized, which may provide a rationale for the coadministration of MMP inhibitors and antiangiogenic agents [34]. Our previous study [14] showed good antiangiogenic potential of *M. aquifolium* extracts, especially MAE, and this effect may indicate that these agents are suitable for overcoming the aforementioned issue of tumor vascularization and growth at lower MMP9 levels, while lower MMP9 expression in cells treated with extracts may reduce the metastatic potential of cancer cells. However, further experiments with a broader range of extract doses must be carried out to establish the proapoptotic activity of M. *aquifolium*.

Berberine, the main plant alkaloid of the genus *Mahonia* and the constituent of both *M. aquifolium* extracts involved [14], exhibits antimetastatic potential as well by blocking Wnt/β-catenin signaling pathway [40–42]. Furthermore, berberine can activate ZO-1 (Zonula Occludance-1), which participates in the formation of cell-tight junctions and indirectly reduces cell mobility [43]. We have investigated the influence of DOX, extracts, and their combinations on the expression of genes that participate in cell adherence and tight junction formation. *CTNNB1* encodes β-catenin, while OCLN encodes the occludin protein, one of the main components of tight junctions. Taken together, the results obtained

here sugges<sup>t</sup> that a decrease in *MMP9* and an increase in *OCLN* expression levels following treatment with a combination of DOX and the investigated extracts may lead to inhibition of cell migration and reduction of the metastatic potential of the treated cells. We have also examined the e ffects of plant extracts on cell migration, showing that both extracts, alone or in combination with DOX, inhibit cell migration, unlike DOX alone. Colony formation analysis results support our observations that the investigated plant extracts work together with DOX, enhancing its anticancer e ffects.
