The effect of apigenin and chemotherapy combination treatments on apoptosis-related genes and proteins in acute leukaemia cell lines

Apigenin is a dietary polyphenol found abundantly in fruit and vegetables, which sensitizes leukaemia cells to topoisomerase inhibitor agents (e.g., etoposide), and alkylating agents (e.g., cyclophosphamide), reducing ATP levels and inducing apoptosis; whilst being protective to control haematopoietic stem cells. This study analysed the expression profiles of intrinsic and extrinsic apoptosis-related genes and proteins to help elucidate the mechanisms of action of apigenin when used in combination with etoposide or cyclophosphamide in lymphoid and myeloid leukaemia cell lines (Jurkat and THP-1). Expression of apoptosis-related genes were measured using a TaqMan® Human Apoptosis Array and the StepOne Plus RT-qPCR System, whilst apoptosis-related proteins were determined using a protein profiler™-human apoptosis array and the LI-COR OdysseyR Infrared Imaging System. Apigenin when combined with etoposide or cyclophosphamide-induced apoptosis via the mitochondrial pathway, increasing the expression of pro-apoptotic cytochrome c, SMAC/DIABLO, and HTRA2/OMI, which promoted caspase-9 and -3 activation. Targeting anti-apoptotic and/or pro-apoptotic members of the apoptotic pathways is a promising strategy to induce cancer cell death and improve sensitivity to chemotherapy agents. Here the apoptotic pathways induced by apigenin in combination with etoposide or cyclophosphamide were identified within human leukaemia cell lines, such applications could provide combination therapies for the treatment of leukaemia.

). A subsequent analysis was made of 35 apoptosis-related proteins (Supplementary Table S2) using the Proteome Profiler™ -Human Apoptosis Array Kit. These analyses were utilised to investigate potential gene and protein regulation by apigenin alone and in combination with chemotherapy agents.
Leukaemia cell line culture. Two human leukaemia cell lines were selected for this study: one acute myeloid leukaemia cell line [THP-1 (monocytic leukaemia)] (ATCC, TIB-202) and one acute lymphoid leukaemia cell line [Jurkat (peripheral blood T cell leukaemia)] (ATCC, TIB-152). Cells were cultured as described in Mahbub et al., 2013 55 , and incubated under standard cell culture conditions at 37 °C with a 5% CO 2 atmosphere. Furthermore, cells were tested for mycoplasma contamination using the MycoAlert™ mycoplasma detection kit (Lonza, LT07-318) and were negative throughout all experiments.
Treatment regimes. Cells were seeded at 1 × 10 6 per well in 12-well plates and treated with the lowest significant doses (LSDs) of apigenin (AP) (Sigma Aldrich, 10798) at 10 µM and 50 µM, etoposide (ETP) (Sigma Aldrich, E1383) at 0.01 µM and cyclophosphamide (CYCLO) (Sigma Aldrich, PHR1404) at 2 µM and 10 µM, alone or in combination; along with a vehicle control for 24 h. The therapeutic agents: AP, ETP, and CYCLO were prepared as described previously 55,63,64 . LSDs capable of inducing apoptosis were selected based on our earlier published data 55,63,64 which caused a 10% to 20% increase in caspase-3 activity and induced apoptosis in THP-1 and Jurkat cells, when compared to the vehicle control.
Gene expression analysis using a TaqMan® Human Apoptosis Array based on the RT-qPCR reaction. Treated cells were collected and the RNA was extracted using a PureLink™ RNA Mini Kit (Thermofisher Scientific-Invitrogen, 12183018A) following the manufacturer's instructions. The purified RNA was then transcribed to cDNA using High-Capacity RNA-to-cDNA™ Kit (Thermofisher Scientific-Applied Biosystems, 4387406) according to the manufacturer's instructions. Following preparation of the cDNA, a 10 µl of PCR reaction mix was prepared with the 5 µl of cDNA and 5 µl of TaqMan™ Fast Advanced Master Mix (Thermofisher Scientific-Applied Biosystems, 4444557) following the manufacturer's instructions and loaded onto the TaqMan Array Human Apoptosis Fast 96-Well Plate Pathway (Thermofisher Scientific-Applied Biosystems, 4418762). Next, the plate was loaded into the StepOnePlus™ Real-Time PCR system (Applied Biosystems, Sheffield Hallam University, UK). Thermal cycling conditions consisted of a hold step for 20 s at 95 °C, and then amplification for 40 cycles starting with an initial denaturation step at 95 °C for 3 s and an annealing/extending step at 60 °C for 30 s. The relative gene expression was determined using the 2 −∆∆CT method 66 . Normalizing against the average housekeeping genes of 4 endogenous control genes: transferrin receptor (TFRC), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), hypoxanthine phosphoribosyltransferase 1 (HPRT1), and Beta glucuronidase (GUSB) (Supplementary Table S1); and untreated controls. Three arrays were processed for each treatment in THP-1 and Jurkat cell lines, in three technical experiments. The results are expressed as medians and ranges, of three independent repeats.
Protein expression analysis using a proteome profiler™: human apoptosis array. Treated cells were collected and washed twice with PBS. Cells were lysed with CelLytic M (Sigma Aldrich, C2978, 10 ml/g) supplemented with a protease inhibitor cocktail (Sigma Aldrich, P2714, 1:100) as described in the manufacturer's instructions. The total protein content was quantified using the Pierce™ BCA protein assay kit (Thermofisher Scientific, 23227), following the manufacturer's instructions. The expression of 35 apoptosis-related proteins was assessed using the Proteome Profiler™-Human Apoptosis Array Kit (Biotechne-R&D Systems, ARY009) (Supplementary Table S2). In brief, the Human Apoptosis Proteome Profiler™ array was composed of a nitrocellulose membrane with duplicate spots for each apoptosis-related protein. Protein expression analysis was performed in duplicate on arrays (each array contained two spots for each protein analysed) for duplicate technical repeats for each treatment in each type of cell, thus a total of 4 repeats per protein per treatment group. All analyses were performed according to the manufacturer's instructions.
Cell lysates were analysed using the proteome profiler human apoptosis antibody array kit (Biotechne-R&D Systems, ARY009) in accordance with the manufacturer's instructions, utilising 300 µg of cell lysate per membrane. Thereafter, the array membranes were scanned using the LI-COR Odyssey R Infrared Imaging System (LI-COR). The pixel density of each duplicated array spot was quantified using Image Studio™ Software version 5.2 (LI-COR). Following the manufacturer's instructions (Biotechne-R&D Systems, ARY009), the mean pixel density of each duplicated protein spot was determined and subtracted from the mean density of the background/ reference spots. Reference spots are included in each array to align the transparency overlay template and to demonstrate that the array has been incubated with IRDye® 800CW Streptavidin (LI-COR Biosciences UK, 926-32230) during the assay procedure. The results were expressed as a median and range of two independent experiments.
Statistical analysis. The median and range was determined for each assay. Stats Direct software (Stats Direct Ltd, Altrincham, UK) was used to determine if the data followed a normal distribution using a Shapiro-Wilk test. As data was identified as non-parametric, a Kruskal-Wallis with a Conover-Inman post-hoc test was used to determine statistical significance. Significance was set at P ≤ 0.05. Statistical significance of individual agents was determined firstly in comparison to the vehicle control (VC) (Supplementary Tables S3-S6). The statistical significance of combined apigenin and chemotherapy treatments was determined in comparison to the vehicle control and the individual treatments alone (Supplementary Tables S3-S6)  Individual treatment effects. A significant increase or decrease following individual treatments on apoptotic gene and protein expression was determined when compared to the vehicle control (VC) (P ≤ 0.05) (*). A significant increase is shown in Table 1 as an upward arrow in green (↑), whilst a significant decrease is shown as a downward arrow in red (↓). Where there were no significant changes in gene/protein expression compared to the vehicle control (VC); this was represented by a dash (-).
Combination treatments effects:. The effect of combination treatments on apoptotic gene and protein expression were classified either as: • Synergistic-here there is a significant increase or decrease in gene/protein expression compared to the vehicle control (VC) and drugs alone (P ≤ 0.05) (*). Synergistic increases in gene/protein expression are represented in the bar charts in dark green, whilst a synergistic decrease is represented in the bar charts in dark red. • Non-interactive, but with significant effect-here there is a significant increase or decrease in gene/protein expression compared to the vehicle control (VC) and/or one of the individual drugs alone (P ≤ 0.05) (*). www.nature.com/scientificreports/ The significant increase in gene/protein expression is represented in the bar charts in light green, whilst a significant decrease is represented in the bar charts in pink. • No effect-here there is no significant difference in gene/protein expression compared to the vehicle control (VC) and/or one of the individual drugs. This effect is represented in the bar charts in grey.
Etoposide. Treatment of THP-1 and Jurkat cells with the LSD of etoposide (0.01 µM) for 24 h significantly increased the expression of the pro-apoptotic genes: TNFR1/TNFRSF1A, CASP-9, and CASP-3 compared to the vehicle control (P ≤ 0.05) (Table 1), plus there was an increase of BAX expression in the Jurkat lymphoid leukaemia cells when compared to the vehicle control (P ≤ 0.05) ( Table 1).
Etoposide treatment also caused a significant decrease of the anti-apoptotic factor BCL2 gene expression and protein levels in THP-1 myeloid leukaemia cells, when compared to the vehicle control (P ≤ 0.05) (Table 1). However, there was no significant effect on the remaining investigated apoptotic genes or proteins in both cell lines (Table 1).
Cyclophosphamide. Following treatment with the LSD of cyclophosphamide for THP-1 myeloid (2 µM) and Jurkat lymphoid (10 µM) leukaemia cells for 24 h, this caused differential effects on the apoptosis genes and proteins (Table 1). In THP-1 acute myeloid leukaemia cells, there was a significant decrease in the expression of BCL2, BCLX, and BAD genes and proteins; with a significant increase in TNFR1/TNFRSF1A, CASP-9, and CASP-3 gene expression, when compared to the vehicle control (P ≤ 0.05) ( Table 1).
In Jurkat acute lymphoid leukaemia cells, there was a significant increase in the expression of BAD, BAX, and CYT c genes and proteins, as well as CASP-9 and CASP-3 genes when compared to the vehicle control (P ≤ 0.05) (Table 1). However, there were no significant effects on the remaining apoptotic genes or proteins in either cell line ( Table 1).
Effect of combination treatments on expression of BCL-2 family anti-apoptotic genes and proteins (BCL2 and BCLX). When apigenin was used in combination with etoposide, there was a synergistic decrease of BCL2 gene expression in both myeloid (THP-1) and lymphoid (Jurkat) leukaemia cells compared to both apigenin alone and etoposide treatment alone (Fig. 1A). Similarly, when apigenin was used in combination with cyclophosphamide there was a synergistic decrease on BCL2 gene expression in Jurkat lymphoid leukaemia cells only (Fig. 1A), when compared to the vehicle control and drugs alone, (P ≤ 0.05), with no further decrease seen in THP-1 cells when cyclophosphamide was combined with apigenin compared to cyclophosphamide alone (Fig. 1A). However, whilst protein expression for BCL2, was significantly decreased following apigenin and etoposide combination treatments compared to vehicle controls (P ≤ 0.05), in both THP-1 and Jurkat cells a synergistic response was not observed ( Apigenin when used in combination with etoposide was shown to synergistically decrease the expression of BCLX gene and protein levels in Jurkat lymphoid leukaemia cells, compared to the vehicle control and drugs alone (P ≤ 0.05) (Fig. 1A). Similarly, apigenin when used in combination with cyclophosphamide, also caused a significant decrease in the expression of BCLX gene and protein in Jurkat cells (Fig. 1A,B) (Supplementary Figs. S1, S2) when compared to the vehicle control and one of the individual treatments, but not both and thus failed to show synergistic responses (P ≤ 0.05). In contrast, in THP-1 myeloid cells, apigenin when combined with cyclophosphamide and etoposide caused a significant reduction in BCLX gene expression compared to apigenin treatment alone, but failed to result in a synergistic response (Fig. 1A), whilst protein levels remained unchanged ( Fig. 1B) (Supplementary Figs. S1, S2).
Overall, Jurkat lymphoid cells were more susceptible to combination treatments, which caused a reduction in anti-apoptotic genes and proteins; whilst the THP-1 myeloid cells were unresponsive (Fig. 1A

Effect of combination treatments on expression of BCL-2 family pro-apoptotic genes and proteins (BAD and BAX).
When apigenin was combined with chemotherapy agents in THP-1 cells the induction of BAD which was seen following treatment with apigenin alone, was inhibited in combination treatments (Fig. 2B). However, when Jurkat cells were treated with a combination of apigenin and cyclophosphamide there was a synergistic increase in the expression of BAD gene and protein levels compared to the vehicle controls and drugs alone, (P ≤ 0.05) ( Fig Effect of combination treatments on expression of caspase genes and proteins (CASP-8, CASP-9, CASP-3). In both myeloid and lymphoid leukaemia cell lines, all combination treatments caused a synergistic increase in CASP-9 and CASP-3 gene expression when compared to the vehicle control and drugs alone (P ≤ 0.05) (Fig. 7A). Whilst CASP-8 was significantly increased compared to the vehicle control and chem- www.nature.com/scientificreports/ otherapy treatments alone (P ≤ 0.05), but not significantly different from that seen from apigenin alone (Fig. 7A). CASP-3 protein expression was also significantly increased in both THP-1 and Jurkat cells when combination treatments were performed compared to the vehicle control and chemotherapy alone (P ≤ 0.05), but no difference was seen from apigenin treatments alone (Fig. 7B) (Supplementary Figs. S1, S2).

Discussion
Our earlier work has shown that polyphenols can decrease cell proliferation, cause cell cycle arrest, and induce apoptosis in leukaemia cell lines 23,55 . Furthermore, we have shown that the polyphenol, apigenin acts synergistically with chemotherapy agents such as etoposide (a topoisomerase inhibitor) and cyclophosphamide (an alkylating agent), causing a reduction in ATP and glutathione levels, DNA damage and induction of apoptosis in acute myeloid (THP-1) and lymphoid (Jurkat) leukaemia cell lines as shown by induction of caspase activity and morphological evidence of apoptosis 63,64 . To progress our understanding of the therapeutic potential of apigenin, it is essential to elucidate their molecular mechanism of action during apoptosis. Here, we investigated the molecular events that trigger apoptosis in response to apigenin alone, and in combination with etoposide and cyclophosphamide in both acute myeloid (THP-1) and lymphoid (Jurkat) leukaemia cells, following treatment for 24 h. There are two core pathways involved in inducing apoptosis: (1) The intrinsic or mitochondrial-mediated pathway, and (2) the extrinsic or death receptor-mediated pathway 5,16 . Initiation of either pathway ultimately results in a caspase activation cascade and cellular death 5,16,17  This study demonstrated apigenin increased gene and protein expression for BAX, BAD, SMAC/DIABLO and CASP-9, but had no effect on the expression of anti-apoptotic BCL2 and BCLX genes and proteins. In addition, apigenin also induced TRAILR1/DR4, TNFR1/TNFRSF1A, and FAS/CD95 and CASP-8, in both acute leukaemia cell lines. Demonstrating the increase in CASP-3 activity was induced via both intrinsic and extrinsic apoptotic pathways. Apigenin, has previously been shown to induce apoptosis in leukaemia cells (HL60, THP-1, U937) via mitochondrial-dependent mechanisms, with the release of cytochrome c to the cytosol and the activation of caspase-9 and -3 [51][52][53][54] . Similarly, Wang et al., 1999 reported that apigenin could induce apoptosis through cytochrome c release and activation of caspase 9 and 3 in HL60 cells 51 .
Etoposide and cyclophosphamide alone induced apoptosis in both cell lines associated with increased CASP-3, however gene and protein regulation differed between THP-1 and Jurkat cells suggesting differential regulation between myeloid and lymphoid cells. In THP-1 cells, both etoposide and cyclophosphamide reduced antiapoptotic BCL2 and/or BCLX, and an increase in the expression of TNFR1/TNFRSF1A, and increased CASP-9 and -3. In Jurkat lymphoid cells, etoposide increased BAX, TNFR1/TNFRSF1A, CASP-9, and -3 genes. Whereas cyclophosphamide was much more effective in Jurkat cells and seemed to specifically target intrinsic apoptosis; Combination treatments of apigenin and chemotherapy agents (etoposide or cyclophosphamide), induced apoptosis via a synergistic induction of the intrinsic apoptosis pathway, combined with an apigenin-induced extrinsic apoptosis, resulting in an overall enhanced induction of apoptosis. Specifically, apigenin enhanced the pro-apoptotic activity of chemotherapy agents, through the synergistic increase in expression of BAX, CYT c, SMAC/DIABLO, HTRA2/OMI, CASP-9, and -3 gene and protein levels, whilst decreasing anti-apoptotic BCL2 gene expression (Figs. 8 and 9).
Hassanpour et al., 2018 reported that the dysfunction of the intrinsic apoptotic pathway is a key strategy to avoid apoptosis in leukaemia cells 2 . Acute leukaemia is commonly associated with a decrease in cell sensitivity to pro-apoptotic signals due to overexpression of anti-apoptotic proteins of the BCL-2 family such as BCL2 and BCLX 3,9,10,16,69 ; and/or with low levels of pro-apoptotic members such as BAX 11  Overexpression of anti-apoptotic proteins BCL2 and BCLX inhibit apoptosis; and are reported to be correlated with the pathogenesis of acute leukaemia (mainly AML), and can induced resistance to chemotherapeutics 16,17,65 . Here, the effective targeting of anti-apoptotic members BCL2 and BCLX, by the use of apigenin with standard chemotherapy agents; seem to vary dependent on the type of leukaemia cell line and chemotherapy agent used,  (Figs. 8 and 9), whilst in Jurkat cells when treated with apigenin and cyclophosphamide, this was accompanied by a synergistic increase of BAD (Fig. 9). In a similar study, when apigenin was used in combination with the BCL2 inhibitor navitoclax for 48 h, there was an upregulation of BAX and BIM gene expression and protein levels, which led to apoptotic cell death in colon cancer (HTC-116) cells 60 . Likewise, Chen et al., 2019 showed that apigenin could inhibit tumorigenesis and induce apoptosis in cisplatin-resistant colon cancer cells (HT-295), when grown in vitro and in vivo as mouse xenografts, through the upregulation of pro-apoptotic BAX and downregulation of anti-apoptotic BCL2 71 .
Conventional chemotherapy agents used clinically, are known to indirectly activate BAX 72 . Liu et al., 2016 and Walensky et al., 2019 reported that the direct activation of BAX, holds great promise for cancer therapies, with the advantages of specificity, and the potential of overcoming chemo-and radio-resistance; and that this is a possible target for small-molecule drug discovery 72,73 . Here, the synergistic increase of BAX gene and protein expression in both THP-1 and Jurkat cells with all combination treatments, was associated with a synergistic increase in CYT c and SMAC/DIABLO, and/or HTRA2/OMI (Figs. 8 and 9).
Cytochrome c (CYT c) is considered to be a critical mediator and biomarker in mitochondria-mediated apoptosis [74][75][76] . It is located on the inner surface of the mitochondria, it is a heme-containing metalloprotein and multifunctional enzyme, that is involved in cell apoptosis 74 . Functionally, overexpression of cytochrome c effectively suppresses the proliferation of cancer cells and induces cell apoptosis, whilst the knockdown of cytochrome c, reverses these effects 74 . SMAC/DIABLO and HTRA2/OMI are important also to neutralize the Death receptor related genes and proteins were shown to be increased in this study by apigenin, however these failed to demonstrate synergistic responses when combined with chemotherapy agents, with the only synergistic increases seen in TNFR1/TNFRSF1A in THP-1 myeloid leukaemia cells, when treated with apigenin and etoposide (Fig. 8); and in TRAILR1/DR4 in Jurkat lymphoid leukaemia cells when treated with apigenin and cyclophosphamide (Fig. 9). Apigenin alone was capable of inducing these death receptors which chemotherapy agents were not able to induce alone. Suggesting combined therapies which maintained the effect induced by apigenin alone could enhance apoptosis induction from chemotherapy alone via activation of additional apoptosis pathways. This was further supported by the induction of caspase 8 activity in cells treated with apigenin, but not chemotherapy. Apigenin has previously been shown to target the extrinsic pathway of apoptosis when used in combination with: (1) TRAIL in Jurkat cells via upregulation of the expression of the DR5 gene 70 ; and (2) 5-fluorouracil and cisplatin in head and neck carcinoma cells (SCC25 and A431) via upregulation of the expression of TNFR and TRAILR genes 58 . Our previous study showed that polyphenols, including apigenin synergistically enhanced the action of chemotherapeutic agents, including etoposide 63 and cyclophosphamide 64 in leukaemia cells (including Jurkat and THP-1), through a reduction of glutathione levels. Furthermore, these studies found that these polyphenol/ chemotherapy combination treatments synergistically reduced cell ATP levels, arrested cell cycle, caused DNA damage, and induced apoptosis 63,64 . Interestingly, many researchers reported that there is a strong correlation between glutathione depletion and restored apoptosis induction 75,77 . Franco et al., 2009 reported that a depletion in the glutathione content can directly activate the intrinsic apoptotic pathway, either by: activating the initiator BAX, releasing the cytochrome c from the mitochondria, or by forming the apoptosome; which can be oxidised for its pro-apoptotic action. This may explain why the investigated combination agents here, synergistically increased the expression of mitochondrial BAX and cytochrome c. Franco et al., 2009 reported that the depletion of glutathione, restored apoptosis via the intrinsic apoptotic pathway; and suggested that this could be a highly effective way to increase the efficacy of chemotherapy or anti-cancer agents 75,77 . In addition, Traverso et al., 2013 reported that high glutathione levels are commonly found in cancer cells and the efflux of glutathione is one of the major key mechanisms in the development of multi-drug resistance in cancer 77 . Our earlier work has found that the basal glutathione levels of the leukaemia cell lines were linked with the sensitivity to the treatments of   63 . It was demonstrated that Jurkat lymphoid leukaemia cell lines had lower basal glutathione levels than the THP-1 myeloid cell lines and non-tumour control cells 63,64 . This could explain why the Jurkat acute lymphoid cell lines are more sensitive and susceptible.
In conclusion, the combination of apigenin with etoposide and cyclophosphamide induced apoptosis by the synergistic increase of BAX expression (Figs. 8 and 9), the permeabilization of the mitochondrial membrane, and the release of CYT c, SMAC/DIABLO, and HTRA2/OMI, which then promotes caspase-9 and -3 activation (Figs. 8 and 9). Furthermore, the induction of the extrinsic apoptosis pathway with induction of death receptors and ligands and caspase 8 by apigenin, activation of which were not seen in chemotherapy stimulations alone could further enhance the total apoptosis seen in combination therapies. Thus, the utilisation of combined therapies targeting apoptosis could have therapeutic potential in the treatment of leukaemia.