Synthesis and Anticancer Activity of Novel 9-O-Substituted Berberine Derivatives

Berberine is a bioactive isoquinoline alkaloid derived from many plants. Although berberine has been shown to inhibit growth and induce apoptosis of several tumor cell lines, its poor absorption and moderate activity hamper its full therapeutic potential. Here, we describe the synthesis of a series of 9-O-substituted berberine derivatives with improved antiproliferative and apoptosis-inducing activities. An analysis of novel berberine derivatives by EPR spectroscopy confirmed their similar photosensitivity and analogous behavior upon UVA irradiation as berberine, supporting their potential to generate ROS. Improved antitumor activity of novel berberine derivatives was revealed by MTT assay, by flow cytometry and by detection of apoptotic DNA fragmentation and caspase-3 activation, respectively. We showed that novel berberine derivatives are potent inhibitors of growth of HeLa and HL-60 tumor cell lines with IC50 values ranging from 0.7 to 16.7 µM for HL-60 cells and 36 to >200 µM for HeLa cells after 48 h treatment. Further cell cycle analysis showed that the observed inhibition of growth of HL-60 cells treated with berberine derivatives was due to arresting these cells in the G2/M and S phases. Most strikingly, we found that berberine derivative 3 (9-(3-bromopropoxy)-10-methoxy-5,6-dihydro-[1,3]dioxolo[4,5-g]isoquino[3,2-a] isoquinolin-7-ylium bromide) possesses 30-fold superior antiproliferative activity with an IC50 value of 0.7 µM and 6-fold higher apoptosis-inducing activity in HL-60 leukemia cells compared to berberine. Therefore, further studies are merited of the antitumor activity in leukemia cells of this berberine derivative.

Detailed studies have shown that berberine can induce apoptosis of tumor cells through its modulation of activity of several pro-apoptotic and anti-apoptotic genes. For example, it alters the Bcl-2/Bax ratio and decreases the mitochondrial membrane potential of selected tumor cells [14]. In addition, it might induce apoptosis through the mitochondrial/caspase pathway, activate caspase-3, and caspase-8, and release the cytochrome c [3,8]. It has been also shown that berberine has a variety of effects on the cell cycle and its lower concentrations arrest tumor cells in the G 1 phase [15,16]. In contrast, higher concentrations of berberine induce the arrest of tumor cells in the G 2 /M phase [17]. Interestingly, it has been reported that berberine might also induce apoptosis of tumor cells by a mechanism that involves the production of a reactive oxygen species (ROS) [18][19][20][21]. Most importantly, several reports have demonstrated that berberine could be combined with radiotherapy or chemotherapy drugs to neutralize their toxicity and enhance their therapeutic activities, thereby improving therapeutic outcomes [22][23][24][25][26][27][28].
Several studies have shown that berberine selectively inhibits tumor cell growth and induces apoptosis while eliciting less cytotoxic effects in normal cells [29,30]. Its poor absorption and moderate activity, however, hamper its full antitumor therapeutic potential [31]. As a result, many laboratories have focused their efforts towards the design and synthesis of novel berberine derivatives with improved biological activities, which has led to the synthesis of numerous berberine derivatives with various modifications at the C8, C9, C10, C12 and C13 positions of the berberine structure [32][33][34][35][36][37][38]. These studies have shown that modifications of berberine at C8 or C13 positions mainly increase its antimicrobial activity, which is closely related to the length of the substituent chains. On the other hand, the modifications to the C9 position of berberine usually increased its antitumor activities [39].
Over the past years, several novel 9-O-substituted berberine derivatives, including the 9-O-pyrazole alkyl group [40], 9-O-chlorohexyl group [41], or 9-O-alkyl-and 9-O-terpenyl group [35] modified berberine derivatives with improved aqueous solubility and enhanced absorption have been synthesized and tested for their anticancer activity. These studies have shown berberine derivatives to be more potent antitumor agents with improved apoptosis-inducing activity than that of berberine. Additionally, it has been shown that the substitution of berberine on C9 with lipophilic group substitutes, such as ibuprofen and naproxen [42], or cinnamic acid [36], significantly increased the anti-inflammatory and enhanced the hypoglycemic activity of berberine, respectively. These studies point out that modification of berberine on the C9 position can significantly improve its biological activities. Several recent reviews have summarized and outlined in more detail the most important findings on the relationship between various structural modifications of berberine and its improved pharmacological activities [38,43,44].
Here, we report the synthesis and evaluation of biological activity of four novel 9-O-substituted berberine derivatives. To improve the pharmacological properties and increase the therapeutic efficiency of berberine, we synthesized its 9-O-substituted derivatives bearing 2-chloro-and 2-bromo-acetoxy substituents (derivate 1 and 2), 3-bromopropoxy substituent (derivate 3) and ciprofloxacin (derivate 4) (Figure 1, . Ciprofloxacin, as a quinolone antibiotic, was selected for the synthesis of berberine derivative 4 as it was reported to have high intrinsic lipophilicity [45] and various biological activity, including the activity against multidrug resistant bacteria Klebsiella pneumoniae [46,47].
Berberrubine was then treated with 1,3-dibromopropane (DBP) to prepare the berberine derivative 3 (Scheme 3). No base was used here to avoid cyclisation to nitrogen. Various conditions for the synthesis of berberine derivative 3 were tested and obtained yields were summarized in Table  1. Table 1. Reaction conditions for the preparation of berberine derivative 3 (1 equivalent of substrate, no base used).

Reaction DBP (Equivalents) Solvent Temperature (°C) Time (h)
Yield (%)  R1  3  CH3CN  60  2  48  R2  10  CH3CN  60  2  55  R3  10  DMF  80  2  59  R4  58  DMF  80  2  74 As presented, we found that the high excess of alkylating agent DBP was the most important factor in increasing the yield of the reaction. Similar to the alkylation of berberrubine by DBP, the excess of the haloacetylhalogenides was limiting factor for derivatization of berberrubine to haloacetylesters 1 and 2 (Scheme 2). It should be noted that, under these conditions, the separation of berberine derivatives from reactions was simple because the products were soluble in polar aprotic solvents, thus distinguishing them from the berberrubine. Concerning the synthesis of the berberineciprofloxacin conjugate (referred as berberine derivative 4), the first step of its synthesis was the derivatization of ciprofloxacin to its acylated form (Scheme 4) [51,52]. Prepared carbonyl chloride of ciprofloxacin was further mixed with berberrubine and the reaction was microwave irradiated to produce the berberine derivative 4 in 31% yield (Scheme 5). Importantly, no berberine derivative 4 was obtained without microwave irradiation. It is widely accepted that microwave irradiation is an unconventional energy source with great potential in synthetic organic chemistry [53]. In this respect, We demonstrate that novel berberine derivatives maintain similar photosensitivity and analogous behavior upon UVA irradiation to that of berberine, supporting their potential to induce apoptosis through generation of the reactive oxygen species (ROS). We show that novel berberine derivatives are more potent antiproliferative and apoptosis-inducing agents compared to berberine. Moreover, these derivatives arrest HL-60 leukemia cells in the G 2 /M and S phases, with the most promising derivative 3 showing superior apoptosis-inducing activity.
Berberrubine was then treated with 1,3-dibromopropane (DBP) to prepare the berberine derivative 3 (Scheme 3). No base was used here to avoid cyclisation to nitrogen. Various conditions for the synthesis of berberine derivative 3 were tested and obtained yields were summarized in Table 1. As presented, we found that the high excess of alkylating agent DBP was the most important factor in increasing the yield of the reaction. Similar to the alkylation of berberrubine by DBP, the excess of the haloacetylhalogenides was limiting factor for derivatization of berberrubine to haloacetylesters 1 and 2 (Scheme 2). It should be noted that, under these conditions, the separation of berberine derivatives from reactions was simple because the products were soluble in polar aprotic solvents, thus distinguishing them from the berberrubine. Concerning the synthesis of the berberine-ciprofloxacin conjugate (referred as berberine derivative 4), the first step of its synthesis was the derivatization of ciprofloxacin to its acylated form (Scheme 4) [51,52]. Prepared carbonyl chloride of ciprofloxacin was further mixed with berberrubine and the reaction was microwave irradiated to produce the berberine derivative 4 in 31% yield (Scheme 5). Importantly, no berberine derivative 4 was obtained without 4 of 18 microwave irradiation. It is widely accepted that microwave irradiation is an unconventional energy source with great potential in synthetic organic chemistry [53]. In this respect, microwave irradiation has been applied with success to synthesize berberine derivatives under mild reaction conditions and with short reaction times [49,50,54].
The synthetic routes and characterization of novel berberine derivatives 1-4 by 1 H NMR and 13 C NMR spectroscopy are described in detail in the Materials and Methods section (Schemes 2-5).
It was previously shown that berberine, like many other bioactive molecules, exhibits sensitivity towards light and its photochemical behavior is strongly affected by the solvent and the presence of molecular oxygen [55]. As a photosensitive agent, berberine was previously studied, and the photoinduced generation of ROS, via both electron and energy transfer mechanisms (Types I and II), was evidenced by EPR spectroscopy and direct detection of singlet oxygen phosphorescence [20,56,57]. The investigation of the cytotoxic/phototoxic effect of berberine on murine fibroblast non-cancer NIH-3T3 and Ehrlich ascites carcinoma (EAC) cells showed a statistically significant degree of UVA induced increase of the cytotoxic effect only in cell line EAC. The cell cycle profile analysis revealed that the DNA damage triggered by berberine in combination with UVA irradiation led to a significant blockage of cells in the S and G 2 /M phases with no sub-G 0 cell fraction detected [20]. Consequently, to provide a more complex view on the newly prepared berberine derivatives, their ability to produce ROS upon UVA irradiation (λ max = 365 nm) was tested using the in situ EPR spin trapping technique.
The UV/vis absorption spectra of the prepared derivatives were recorded to follow the effect of the structural changes on the absorption maxima ( Figure 2). We found that berberine derivative 3 exhibits an analogous spectrum to that of berberine with the characteristic absorption maxima at 348 nm and 421 nm ( Figure 2B), while berberine derivatives 1 and 2 reveal a slight bathochromic shift of these bands (355 and 464 nm) together with the presence of other bands in the 300-600 nm region ( Figure 2A). The absorption bands reaching higher wavelength up to 600 nm reflect the extended conjugation in berberine derivatives 1 and 2 due to the presence of 1-chloro-acetyl or 1-bromo-acetyl substituents. A combination of berberine with ciprofloxacin (berberine derivative 4) led to a slight bathochromic shift of the absorption maxima (325 and 337 nm) compared to the ciprofloxacin (321 and 334 nm) and the appearance of broad shoulders at higher wavelength (370 and 490 nm), expanding the absorption more towards the visible region ( Figure 2C). As novel berberine derivatives reveal a suitable overlap of their absorption spectra and the emission of UVA light, a source with a maximum at 365 nm was applied in further photochemical experiments.
The ROS generation initiated by UVA photoexcitation of berberine derivatives was investigated using indirect techniques of EPR spectroscopy.
The application of spin trapping agent 5,5-dimethyl-1-pyrroline N-oxide (DMPO) enabled us to follow the formation of reactive radical intermediates upon in situ photoexcitation of berberine derivatives as the corresponding spin-adducts. The EPR spin trapping experiments confirmed that berberine derivatives preserved the photosensitivity of the parent berberine molecule, since they exhibited the photoinduced formation of DMPO spin-adducts upon UVA irradiation, reaching the highest EPR signal intensity for derivative 3 ( Figure 3). substituents. A combination of berberine with ciprofloxacin (berberine derivative 4) led to a slight bathochromic shift of the absorption maxima (325 and 337 nm) compared to the ciprofloxacin (321 and 334 nm) and the appearance of broad shoulders at higher wavelength (370 and 490 nm), expanding the absorption more towards the visible region ( Figure 2C). As novel berberine derivatives reveal a suitable overlap of their absorption spectra and the emission of UVA light, a source with a maximum at 365 nm was applied in further photochemical experiments.  The ROS generation initiated by UVA photoexcitation of berberine derivatives was investigated using indirect techniques of EPR spectroscopy. The application of spin trapping agent 5,5-dimethyl-1-pyrroline N-oxide (DMPO) enabled us to follow the formation of reactive radical intermediates upon in situ photoexcitation of berberine derivatives as the corresponding spin-adducts. The EPR spin trapping experiments confirmed that berberine derivatives preserved the photosensitivity of the parent berberine molecule, since they exhibited the photoinduced formation of DMPO spin-adducts upon UVA irradiation, reaching the highest EPR signal intensity for derivative 3 ( Figure 3). The dominant twelve-line signal corresponding to berberine derivative 3 in the spectrum was assigned to the • DMPO-O2 − (spin-Hamiltonian parameters, aN = 1.279 mT, aH β = 1.025 mT; aH γ = 0.137 mT; g = 2.0059) based on the simulation analysis. This evidences the effective electron transfer from the photoexcited berberine derivative 3 to the molecular oxygen generating the superoxide radical anion stabilized in the aprotic DMSO solvent. Further spin-adduct recognizable in the spectrum belongs to the spin-adduct • DMPO-OCH3 (aN = 1.303 mT, aH β = 0.842 mT; aH γ = 0.147 mT; g = 2.0059) originating from the • OCH3 produced by the interaction of O2 •− with DMSO. Even though the absorption of the berberine derivatives is shifted towards the visible region, only a negligible concentration of DMPO spin-adducts upon the visible-light irradiation (λ > 420 nm) was found.
The EPR spectra monitored during the continuous irradiation of the berberine and its derivatives 1-4 in a mixed water/DMSO (2:3, v:v) solvent containing DMPO are more complex and represent superposition of signals attributable to four individual DMPO spin-adducts, as shown in the  The EPR spectra monitored during the continuous irradiation of the berberine and its derivatives 1-4 in a mixed water/DMSO (2:3, v:v) solvent containing DMPO are more complex and represent superposition of signals attributable to four individual DMPO spin-adducts, as shown in the berberine derivative 2 in Figure 4A. Based on the spin-Hamiltonian parameters elucidated from the simulation analysis, the signals were assigned to the spin- [58] and to the two spin-adducts of the carbon-centered radicals with slightly different parameters: · DMPO-CR1 (a N = 1.551 mT, a H β = 2.197 mT; g = 2.0056) and · DMPO-CR2 (a N = 1.506 mT, a H β = 2.121 mT; g = 2.0057). The last two spin-adducts were produced via interaction of photogenerated ROS with DMSO or berberine molecules. Moreover, a low-intensity single-line signal (g~2.0043, asterisk in Figure 4A) with unresolved hyperfine structure was found upon UVA photoexcitation of the berberine derivative 2 in water/DMSO (2:3, v:v) solution under given experimental conditions, which was most probably produced by the electron transfer reactions of the photoexcited berberine derivative 2, and its g-value is compatible with the oxygen-centered radicals [59].  Upon the continuous UVA irradiation of berberine derivatives 1-4 in the presence of stericallyhindered amine 4-oxo-2,2,6,6-tetramethylpiperidine (TMPO) in DMSO, we only monitored the generation of low-intensity EPR signal of 4-oxo-2,2,6,6-tetramethylpiperidine N-oxyl (Tempone) produced via TMPO oxidation by singlet oxygen. The generated nitroxide radical Tempone is most likely transformed to the diamagnetic products by the interaction of >NO • moiety with other radical intermediates [60], effectively formed upon the UVA photoexcitation of berberine derivatives 1-4.
Overall, the EPR spin trapping experiments demonstrated analogous behaviour of novel berberine derivatives upon UVA irradiation involving effective ROS generation to that of berberine. It will be interesting to examine these berberine derivatives for their dose-dependent accumulation of intracellular ROS to find out if generated ROS are involved in their cytotoxicity.
The antiproliferative activity of berberine derivatives was carried out by the 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and rated in half-maximal inhibitory concentration, which is a measure of the effectiveness of an agent in inhibiting the growth and viability of cells [61]. The antiproliferative activity of berberine derivatives in tumor HeLa and HL-60 cell lines is summarized in Table 2.  Upon the continuous UVA irradiation of berberine derivatives 1-4 in the presence of sterically-hindered amine 4-oxo-2,2,6,6-tetramethylpiperidine (TMPO) in DMSO, we only monitored the generation of low-intensity EPR signal of 4-oxo-2,2,6,6-tetramethylpiperidine N-oxyl (Tempone) produced via TMPO oxidation by singlet oxygen. The generated nitroxide radical Tempone is most likely transformed to the diamagnetic products by the interaction of >NO • moiety with other radical intermediates [60], effectively formed upon the UVA photoexcitation of berberine derivatives 1-4.
Overall, the EPR spin trapping experiments demonstrated analogous behaviour of novel berberine derivatives upon UVA irradiation involving effective ROS generation to that of berberine. It will be interesting to examine these berberine derivatives for their dose-dependent accumulation of intracellular ROS to find out if generated ROS are involved in their cytotoxicity.
The antiproliferative activity of berberine derivatives was carried out by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and rated in half-maximal inhibitory concentration, which is a measure of the effectiveness of an agent in inhibiting the growth and viability of cells [61]. The antiproliferative activity of berberine derivatives in tumor HeLa and HL-60 cell lines is summarized in Table 2. We observed an increase in antiproliferative activity of berberine derivatives 1-4 against tested tumor cell lines compared to berberine or ciprofloxacin. Cytotoxic activity was observed in berberine derivatives 1, 2 and 4 with IC 50 values ranging from 95.0 to >200 µM for HeLa cells and from 11. As cell membrane blebbing, observed in sub-population of berberine derivative-treated HeLa and HL-60 cells, is one of the defining features of apoptosis [62,63], next, we analyzed the potential of novel berberine derivatives to induce apoptosis. We compared their ability to induce apoptosis by employing DNA laddering and caspase-3 activation assays. As presented, after incubation of HL-60 cells with 10 µM concentrations of berberine derivatives for 48 h, only berberine derivative 3 treated HL-60 cells were positive for apoptotic DNA fragmentation ( Figure 6A) and had active caspase-3 ( Figure 6B). Comparing the intensities of apoptotic DNA fragmentation and activation of caspase-3 in HL-60 cells treated with 30 µM berberine and 10 µM berberine derivative 3, berberine derivative 3 clearly exhibited 30-fold superiority to that of berberine.
shown, untreated HeLa cells had normal epithelial morphology. In contrast, in some tumor cells, exposure to 10 µM concentrations of berberine or its derivatives 1, 2 and 3 induced the formation of cell membrane blebbing after 24 and 48 h of treatment. However, HeLa cells treated with 10 µM concentrations of ciprofloxacin and berberine derivative 4 did not show any morphological changes. On the other hand, compared to untreated HL-60 cells, which have the rounded morphology characteristic of leukemia cells, the treatment of HL-60 cells with 10 µM concentrations of berberine, ciprofloxacin and berberine derivatives 1-4 resulted in profound formation of cell membrane blebs. As cell membrane blebbing, observed in sub-population of berberine derivative-treated HeLa and HL-60 cells, is one of the defining features of apoptosis [62,63], next, we analyzed the potential of novel berberine derivatives to induce apoptosis. We compared their ability to induce apoptosis by employing DNA laddering and caspase-3 activation assays. As presented, after incubation of HL-60 cells with 10 µM concentrations of berberine derivatives for 48 h, only berberine derivative 3 treated HL-60 cells were positive for apoptotic DNA fragmentation ( Figure 6A) and had active caspase-3 ( Figure 6B). Comparing the intensities of apoptotic DNA fragmentation and activation of caspase-3 in HL-60 cells treated with 30 µM berberine and 10 µM berberine derivative 3, berberine derivative 3 clearly exhibited 30-fold superiority to that of berberine.  To further extend our cytotoxicity studies, we analyzed the effect of novel berberine derivatives on the cell cycle of HL-60 leukemia cells. We found that in contrast to berberine, which at 30 µM concentrations arrests HL-60 cells in the G 0 /G 1 phase, the novel berberine derivatives 1 and 2 (30 µM) and 3 (10 µM) arrest HL-60 cells at the G 2 /M phase with a proportional decrease of cells in the G 0 /G 1 and S phases, while berberine derivative 4 induces G 2 /M and S phases-arrest (Figure 7).
To further extend our cytotoxicity studies, we analyzed the effect of novel berberine derivatives on the cell cycle of HL-60 leukemia cells. We found that in contrast to berberine, which at 30 µM concentrations arrests HL-60 cells in the G0/G1 phase, the novel berberine derivatives 1 and 2 (30 µM) and 3 (10 µM) arrest HL-60 cells at the G2/M phase with a proportional decrease of cells in the G0/G1 and S phases, while berberine derivative 4 induces G2/M and S phases-arrest (Figure 7).  This finding indicates that novel berberine derivatives are more active molecules than berberine and ciprofloxacin, which in the same concentration have no effect or induce G 0 /G 1 cell cycle arrest of HL-60 leukemia cells [15][16][17]. The observed G 2 /M cell cycle arrest of HL-60 cells treated with berberine derivatives 1-4 supports their higher antiproliferative activity. Interestingly, berberine derivative 3 induced a significant increase in the number of cells defined as the sub-G 0 cell population (1.8% in untreated vs. 21.9% in berberine derivative 3 treated cells, Figure 7B), confirming its apoptotis-inducing potential.
In conclusion, we have synthesized a series of four novel 9-O-substituted berberine derivatives and evaluated their antiproliferative activity in human HeLa and HL-60 tumor cell lines. We found these derivatives to be more active antitumor agents than their parental molecules, berberine and ciprofloxacin. Most importantly, we showed that berberine derivative 3 (9-(3-bromopropoxy)-10-methoxy-5,6-dihydro- [1,3]dioxolo [4,5-g]isoquino[3,2-a]isoquinolin-7-ylium bromide) possesses the highest cytotoxic activity and ability to induce apoptosis of HL-60 leukemia cells. Therefore, this compound could become a promising candidate for anticancer therapy, and has been selected for the follow-up studies to reveal the molecular mechanisms of its antitumor action.

Electron Paramagnetic Resonance (EPR) Spectroscopy Studies of Berberine Derivatives
The UV/visible spectra of berberine, ciprofloxacin and newly synthesized berberine derivatives dissolved in the mixed solvent DMSO/water (1:100, v:v) were recorded using a UV-3600 UV-vis-NIR spectrophotometer (Shimadzu, Kyoto, Japan) with a 1-cm square quartz cell.
The formation of paramagnetic intermediates of the studied berberine derivatives upon their UVA irradiation was monitored in situ using the EPR spin trapping technique. The photoinduced production of singlet oxygen during the excitation of berberine derivatives was followed by the oxidation of a sterically-hindered amine 4-oxo-2,2,6,6-tetramethylpiperidine (TMPO). The solution of the berberine derivatives was mixed either with the spin trapping agent 5,5-dimethyl-1-pyrroline N-oxide (DMPO) or TMPO solution, and immediately after a careful air or argon saturation using a slight gas stream, the prepared sample was transferred to a small quartz flat cell (WG 808-Q, optical cell length 0.04 cm; Wilmad-LabGlass, Vineland, NJ, USA) optimized for the TE102 cavity (Bruker, Rheinstetten, Germany) of the X-band EPR spectrometer (EMX Plus, Bruker, Rheinstetten, Germany). The photoexcitation of the samples took place directly in the EPR resonator at 295 K and the EPR spectra were recorded in situ during the continuous irradiation or after a defined exposure. A UV LED monochromatic radiator (λ max = 365 nm; Bluepoint LED, Hönle UV Technology, Gräfelfing/München, Germany) reaching the irradiance value of 13 mW·cm −2 within the EPR cavity, determined using a UVX radiometer (UVP, Upland, CA, USA) or the visible-light source (KL 1600LED (T = 5600 K; Schott)) were used as the irradiation sources. The g-values were determined using a built-in nuclear magnetic resonance teslameter (ER 036TM, Bruker) and an integrated frequency counter. The EPR spectra were analyzed and simulated using Bruker software WinEPR and EasySpin toolbox working under Matlab software (MathWorks, Natick, MA, USA) [64].

Evaluation of Antiproliferative Activities of Berberine Derivatives
Cell Cultures and Determination of Cell Viability and Morphology HL-60 (human promyelocytic leukemia) and HeLa (human cervix epithelioid carcinoma) cells were grown in a humidified atmosphere containing 5% CO 2 and 95% air at 37 • C. HeLa cells were maintained in a DMEM medium and HL-60 cells were maintained in RPMI 1640 medium, both supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS), 50 U/mL penicillin, and 50 µg/mL streptomycin. Stock solutions of berberine and its derivatives were prepared in DMSO, except for a stock solution of ciprofloxacin, which was prepared in 0.1 M HCl. In all experiments, the final concentration of DMSO was kept ≤ 0.1% (v/v) and 0.1% DMSO alone was used as vehicle control. Cell viability was determined using a 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay [61]. HL-60 cells (2 × 10 4 cells/well) or HeLa cells (4 × 10 4 cells/well) were seeded in 96-well plates. Cells were treated with various concentrations of berberine, ciprofloxacin and berberine derivatives (0-100 µM) for 24 and 48 h. At the end of the treatment, MTT solution (final concentration 1 mg/mL) was added to each well and the plates were incubated for another 4 h. The precipitated formazan was dissolved in DMSO and its concentration was measured at 595 nm using a microplate reader (xMark, BioRad, Tokyo, Japan). Based on the curve fitting using nonlinear regression (Origin 7.0, Microcal), the median inhibitory concentrations IC 50 values (the concentration resulting in 50% inhibition of the cell proliferation that was recorded in control experiments) were determined separately for each experiment. The values were calculated from three independent experiments. Data are presented as a mean ± SD (n = 3).
The morphology of cells treated with tested drugs was evaluated using light microscopy (Zeiss Jenalumar, Jena, Germany) [65].
Detection of Induction of Apoptosis by DNA Laddering Assay Cells (2 × 10 6 ) treated with various concentrations of tested derivatives were washed with ice-cold PBS containing 10 mM EDTA. Pellets of cells were resuspended in 250 µL of lysis buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM EDTA, 5% Triton-X, 0.25% SDS) supplemented with 0.4 mg/mL DNase-free RNase A, incubated at 65 • C for 15 min, cooled to RT for 10 min, and incubated at 37 • C for 90 min. After that, Proteinase K was added to a final concentration of 0.2 mg/mL and samples were incubated at 50 • C for an additional 1 h. The DNA was extracted with 200 µL of phenol:chloroform:isoamyl alcohol (25:24:1) for 1 min and centrifuged at 14,500× g for 5 min. The DNA was precipitated from the aqueous phase with 3 volumes of chilled EtOH containing 0.3 M sodium acetate. The DNA pellets were washed with 70% EtOH, dried, and dissolved in 100 µL of TE buffer. The DNA was separated on 1.5% agarose gel and visualized under UV illumination after being stained with GelRed Nucleic Acid Gel Stain (254 nm, Ultra-Lum Electronic UV Transilluminator, Paramount, CA, USA) [66].

Determination of Cell Cycle Distribution
Cells (0.5 × 10 6 ) treated with berberine derivatives (0-30 µM) for 24 and 48 h were washed with PBS and fixed in ice-cold 70% ethanol. Next, the cells were washed twice with PBS and resuspended in 0.1% Triton-X in PBS containing 50 µg/mL RNase A and incubated for 30 min at 37 • C. Afterwards, DNA was stained using PI (50 µg/mL) for 15 min at 4 • C. The cell cycle distribution was analyzed using a FACS Canto II flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). A minimum of 10,000 cells per sample was analyzed at a flow rate of 200 cells/s [67].

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.