The evergreen neem tree (Azadirachta indica) exhibits several activities useful for agricultural and medicinal applications known in India for thousands of years. The plant or its preparations are extensively reported to exert insecticidal, pesticidal, and agrochemical properties. Additionally, its constituents are applied in alternative (Ayurveda, Unani, Homeopathy) and modern medicine, e.g., for the treatment of diverse infectious, metabolic, or cancer diseases [11, 28] (for chemical structures, see Scheme 1).

Scheme 1
scheme 1

Chemical structures and biological activity of natural products isolated from the neem tree

In this context, azadirachtin 1, a tetranortriterpenoid extracted from neem seeds over 40 years ago [12], was shown to act as a potent phagorepellent and growth inhibitor of insects [53]. Neem tree constituents exhibit various multi-targeted biological activities, such as antibacterial [63], hypoglycemic [39], anti-ulcer [45], antimalarial [33], hepatoprotective [8], spermicidal [4], anti-inflammatory [9, 44, 66], chemopreventive [15, 62], and chemotherapeutic [23, 24, 41] properties. Additionally, the nutritional potential of neem oil was evaluated by Rukimini [50].

Related to cancer, Arivazahagan and coworkers suggest that neem leaf extract may exert its chemopreventive effects by down-regulation of lipid peroxidation, simultaneously increasing the level of glutathione (GSH) and GSH-dependent enzymes [6]. The strong chemopreventive activity of the extract could additionally be confirmed in vivo in rodents [22, 58]. Neem extract was able to induce apoptosis and to inhibit proliferation in prostate cancer [29]. This anti-proliferative effect might partially be due to nimbolide, a triterpene extracted from neem flower that was highly effective in several cancer cell lines, e.g., U937, HL-60, THP1, B16, and BoWe [23, 48]. Moreover, the leaf extract stimulates the immune function by the activation of killer cells [10]. Antioxidative activities were described for gallic acid 3, catechin 5, and epicatechin 6 extracted from the bark [66]. The antioxidative effect is mainly relied on the total content of phenolic compounds (e.g., flavonoids such as quercetin 4) [59]. Molecular mechanisms largely remain to be elucidated.

The NF-κB transcription factor plays an important role in cancer and related diseases [56]. This transcription factor is localized in the cytosol and is blocked by inhibitor of κB (IκB). Upon activation by pro-inflammatory cytokines or viruses, phosphorylation of IκB by IκB kinase complex (IKK) is observed. After ubiquitinylation of IκB, the latter is degraded resulting in the translocation of NF-κB to the nucleus. Here, the activated factor regulates the transcription of multiple genes responsible for regulation of cell cycle, apoptosis, or cellular invasion [17, 27]. Moreover, inhibition of NF-κB-regulated genes potentiates apoptosis [3, 5]. This study gives first mechanistic insights into the inhibitory effect of a methanolic neem leaf extract (NLE) on the NF-κB signaling pathway and subsequent induction of apoptosis in human leukemia cancer cell lines.

Materials and methods

Materials and cell culture

Tumor necrosis factor-α (TNF-α) purchased from Sigma was dissolved in 10 μg/ml in PBS (1×) supplemented with 0.5% (w/v) BSA according to manufacturer’s instructions. Quercetin, gallic acid, catechin, and epicatechin were purchased from Sigma. Salannin and nimbin were purchased from Trifolio GmbH. K562 (human chronic myeloid leukemia in blast crisis) and Jurkat (human T-cell leukemia) and U937 (human histiocytic lymphoma) cells were obtained from (DSMZ) Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH and were maintained in RPMI medium (Bio-Whittaker) with 10% (v/v) fetal calf serum (FCS) (Bio-Whittaker) and 1% (v/v) antibiotic–antimycotic solution (Sigma) in an incubator (37°C, 5% CO2).

Healthy blood samples were kindly donated as buffy coats by the Red Cross (Luxembourg, Luxembourg). By applying diluted (1/3) blood onto a Ficoll layer followed by centrifugation (400g, 20 min), mononuclear cells were isolated and collected. The isolated peripheral blood mononuclear cells (PBMCs) were kept in culture at 37°C and 5% CO2 for 24 h before they were subjected to treatments.

Preparation of the neem leaf extract

Dried leaves of the neem tree were grinded in a mortar. The resulting green solid powder was dissolved in methanol, and the suspension was mixed in a shaker at 25°C for 24 h. Then, remaining powder particles were removed by filtration, and the solvent of the green methanolic solution was evaporated to dryness yielding a green crude extract (yield: 13.2% m/m of the starting material). This crude extract was dissolved in dimethyl sulfoxide (DMSO) at a concentration of 80 mg/ml and stored at −20°.

HPLC analysis

An Agilent 1200 HPLC system with a DAD UV–VIS detector was used. The separation was performed on a reversed-phase column (ZORBAX C18 21.2 × 150 mm, 5 μm) eluting with methanol–water (0:100–100:0) at a flow rate of 1 ml/min and UV detection (280 nm). The active compound quercetin in methanolic leaf extract was identified by comparison with pure standard.

Measurement of NF-κB-inhibitory activity

The inhibitory activity against NF-κB was determined using a Dual-Glo™ Luciferase assay system from Promega. Electroporation of K562 cells was realized as described previously [18]. Equal amounts (each 5 μg per pulse (2.5 × 106 cells)) of an NF-κB plasmid (Stratagene, p5xNF-κB containing five repeats of a consensus NF-κB site) and of a Renilla plasmid (phRG-TK, Promega) were added to each pulse. After transient transfection, cells were transferred to culture medium. After 24 h of incubation, transfected K562 cells were solubilized at a concentration of 1 * 106 cells per ml in RPMI 1640 medium containing 0.1% FCS and 1% antibiotic–antimycotic solution and incubated with different concentrations of the neem leaf extract solution. After 2 h of treatment, TNF-α was added (at 20 ng/ml). After 6 additional hours of incubation, expression of NF-κB was assessed according to the provider’s protocol by luminescence measurement using a Berthold Orion Luminometer (integration time 10 s).

Viability assay

The effect of neem leaf extract on cell viability was assessed using the CellTiter-Glo® Luminescent Cell Viability Assay Kit from Promega. The reduction in the number of viable cells in percent was quantified in comparison with a vehicle-treated control according to the manufacturer’s instructions.

Electrophoretic mobility shift assay (EMSA)

K562 cells were preincubated with indicated concentrations of neem leaf extract solution. After 2 h of incubation, 20 ng/ml of TNF-α was added for additional 6 hours of incubation (37°C, 5% CO2). Then, nuclear proteins were extracted according to the method of Müller and coworkers [38] and stored at −80°C.

The nucleotide 5′-CGCTTGATGACTCAGCCGGAA-3′ (consensus NF-κB site) and the complementary sequence were used as probe. Ten micrograms of each nuclear protein extract was incubated for 20 min at 4°C with the 32P-labeled probe in a mixture of 15 μl containing 35 mM of spermidine, 4 μM of BSA, and 2.5 mM of poly-dIdC. In case of a super-shift experiment, the nuclear extract was first incubated with the antibody in the mixture for 15 min at 4°C, before the labeled probe was added, followed by an incubation period of 30 min at 4°C. Each reaction mixture was then loaded with 5 μl of loading buffer in a well of the gel, which was pre-run for at least 30 min. The electrophoresis was realized at 16 mA per gel for at least 2–3 h. The gel was then dried, and the imaging of the signals was performed using a Kodak BioMax XAR film.

Western blot analysis

After treatment of Jurkat cells with NLE (300 μg/ml) and TNF-α (20 ng/ml), both IκBα degradation and subsequent translocation of p50 or p65 to the nucleus were analyzed in cytosolic and nuclear protein extracts, respectively, by Western blot. In a second set of experiments, K562 cells were treated with different concentrations of NLE. Caspase activation and up-regulation of anti-apoptotic proteins were evaluated through the analysis of total protein extracts. For caspase activation, a positive control was realized by using heteronemin or cisplatin [55, 60]. Cytosolic and nuclear extracts or total protein extraction of the cell lines was performed, and protein concentration of these extracts was determined by Bradford assay and analyzed by Western blot as previously described [13, 18]. Antibodies were from Santa Cruz Biotechnology (p50, p65, IκB, lamin B, β-actin) Cell Signaling Technologies (Bid, Caspases -3, -7, -8, -9), BD Pharmingen (Bcl-XL, XIAP) and Calbiochem (α-tubulin).

IKK kinase activity

The K-LISA IKKβ Inhibitor Screening kit from Calbiochem was used for the measurement of the kinase activity. IKK-2 inhibitor IV ([5-(p-fluorophenyl)-2-ureido]thiophene-3-carboxamide) (Calbiochem) was used as a positive control. The assay was performed as indicated in the manufacture’s protocol. The absorbance at 450 nm (with a reference wave length at 590 nm) was read using a SpectraCount UV-spectrometer (Canberra-Packard).

Analysis of nuclear fragmentation

The percentage of apoptotic cells was determined through the fraction of apoptotic nuclei via fluorescence microscopy (Leica-DM IRB microscope, Lecuit, Luxemburg) after staining with the DNA-specific dye Hoechst 33342 (Sigma, Bornem, Belgium), as previously described [13]. The fraction of cells with nuclear apoptotic morphology was counted (at least 300 cells in at least three independent fields), and the images were analyzed using the Image J software (

Flow cytometric analysis

K562 cells were treated with indicated concentrations of NLE for an incubation period of 8 h; then, the cells were assayed for phosphatidylserine exposure, by using the AnnexinV–FITC Apoptosis Detection Kit I® (Becton–Dickinson Biosciences, Erembodegem, Belgium) as written in the manufacturer’s instructions. Stained samples were analyzed by FACS (FACSCalibur, Becton–Dickinson, San José, CA, USA). The data were recorded via the CellQuest software ( for further analysis.

Results and discussion

Neem leaf extract inhibited TNF-α-activated NF-κB pathway

NF-κB assays of the pure neem standards quercetin, gallic acid catechin, epicatechin, nimbin, and salannin showed that only the product quercetin exerted high NF-κB inhibition potential at a concentration of 100 μM (Fig. 1). HPLC analysis was done to identify quercetin in methanolic neem leaf fractions. HPLC chromatograms are shown in Fig. 2a. Our results reveal that NLE contains quercetin 4 (52 min) Fig. 2b. This result confirms previous observations as quercetin has been reported to inhibit NF-κB signaling in murine intestinal epithelial cells [49].

Fig. 1
figure 1

Effect of pure neem products on the inhibition of TNF-α-induced κNF-κB activation assessed by a Dual-Glo™ Luciferase assay system. K562 cells were pretreated with pure compounds at concentrations of 10 or 100 μM and incubated for 2 h, followed by TNF-α stimulation for an additional incubation period of 6 h. The results of the luciferase NF-κB reporter gene assay represented the ratio of the measured luminescence of the firefly luciferase vector divided by the measured luminescence of the Renilla plasmid. An untreated cell solution was the negative control; positive control was a cell solution treated with TNF-α only. Results were presented as mean and the standard deviation of 4 individual measurements. Experiments were performed in triplicate

Fig. 2
figure 2

HPLC diagrams of crude neem leaf extract and pure Quercetin (Fig. 2a, b, respectively)

Neem leaf extract (NLE) acts an inhibitor of the TNF-α-triggered induction of NF-κB in our model (Fig. 3) as the activation of the transcription factor was inhibited for more than 80 percent at a dose of 240 μg/ml of the leaf extract. An IC50 of 220 ± 20 μg/ml (K562) or 250 ± 20 μg/ml (Jurkat) was calculated using three independent experiments.

Fig. 3
figure 3

Effect of neem leaf extract on the inhibition of TNF-α-induced NF-κB activation assessed by a Dual-Glo™ Luciferase assay system. K562 (a) or Jurkat (b) cells were pretreated with neem leaf extract at various concentrations of 160–300 μg/ml and incubated for 2 h, followed by TNF-α stimulation for an additional incubation period of 6 h. Results were presented as mean and the standard deviation of 4 individual measurements. Experiments were performed in triplicate. * and **, significantly different from positive control (P < 0.05 and 0.01, respectively)

Furthermore, binding of NF-κB to DNA was suppressed after TNF-α-triggered activation of the transcription factor when NLE was applied at concentrations between 160 and 320 μg/ml to the CML cell line K562, as shown by EMSA where we observe a decrease in the p50/p65-dimer binding to a consensus NF-κB binding site (Fig. 4a). Immunodepletion experiments proved the identity of the p50/p65 and p50/p50 lanes. (Fig. 4c). Similar results were obtained for neem-treated Jurkat cell lines (Fig. 5).

Fig. 4
figure 4

a Effect of neem leaf extract on the binding affinity of NF-κB assessed by an EMSA. K562 cells were pretreated during 2 h with different concentrations of neem leaf extract (80–320 μg/ml), followed by TNF-α addition (at 20 ng/ml). After additional 6 h of incubation, the cells were extracted by a method described by Müller and coworkers. [38]. The DNA binding affinity is determined by an incubation of the nuclear cell extract (10 μg) with a labeled oligonucleotide probe containing the NF-κB binding site C-κB. b A quantification of one representative experiment is shown. c In case of a super-shift/immunodepletion experiment, the nuclear extract was first incubated with the antibody in the mixture for 15 min, before the labeled probe was added, followed by an incubation period of 30 min. Positive control was a cell solution treated with TNF-α only. Negative control was a cell solution with neither TNF-α nor natural compound. The data shown here were representative for three independent experiments with similar results

Fig. 5
figure 5

a Effect of neem leaf extract on the DNA binding affinity of NF-κB assessed by an EMSA on Jurkat cell line. b A quantification of one representative experiment is shown

Over the last years, the use of plant extracts allowed the discovery of various bioactive compounds. Indeed, Moon et al. [37] observed that an aqueous extract of Benincasa hispida Cogniaux reduced NF-κB promoter activity in glucose-induced vascular inflammation of human umbilical vein endothelial cells. The extract of the tropical plant Knema laurina (Myristicaceae) was used for centuries for the treatment of digestive and inflammatory diseases and inhibited NF-κB-translocation in brain tissue after inflammatory damage [21]. Lampronti et al. [30] evaluated the inhibitory effects of Bangladeshi medicinal plant extracts on interactions between selected transcription factors and target DNA sequences and discovered that extracts from Terminalia arjuna and Saraca asoka appeared to be most efficient in inhibiting NF-κB binding activities. Interestingly, a purified extract called CML-1 was generated from a mixture of 13 oriental herbs widely used for the treatment of inflammatory diseases in Asia (Achyranthis Radix, Angelicae Gigantis Radix, Cinnamomi Cortex Spissus, Eucommiae Cortex, Glycyrrhizae Radix, Hoelen, Lycii Fructus, Paeoniae Radix, Rehmanniae Radix Preparata and Atractylodis Rhizoma, Zingiberis Rhizoma, Zizyphi Semen, Acori Graminei Rhizoma). This complex mixture inhibits TNF-α-induced IκB kinase activation, subsequent degradation of IκBα, and nuclear translocation of NK-κB that may explain the ability of CML-1 to suppress inflammation [36]. Finally, an extract from stem bark of Cinnamomum cassia Blume (Lauraceae) was discovered to exert an inhibitory effect on lipopolysaccharide (LPS)-induced NF-ΚB transcriptional activity. Following activity-guided fractionation, trans-cinnamaldehyde and 2-methoxycinnamaldehyde were identified as the NF-κB inhibitors from C cassia with IC50 values of 43 μM and 31 μM, respectively [46].

Several groups reported anticancer, anti-inflammatory, and antioxidant activities of neem tree extracts [20, 34, 54, 57] without providing in-depth analysis of the molecular mechanisms leading to NLE-induced inhibition of NF-κB activation. Accordingly, we compared Jurkat cells treated with TNF-α with or without pretreatment by NLE. Results show in control cells that IκBα is degraded after 15 min of cytokine treatment, followed by the translocation of p50 and p65 to the nucleus (Fig. 6 left panel). However, pretreatment of Jurkat cells with 300 μg/ml of NLE resulted in a complete abrogation of IκBα degradation and a significant reduction in nuclear translocation of p50/p65 (Fig. 6 right panel) confirming previous results.

Fig. 6
figure 6

Effect of neem leaf extract on the degradation of IκBα and translocation of p65 and p50 to the nucleus. Jurkat cells were incubated with neem leaf extract (at 300 μg/ml) for 2 h, followed by a TNF-α (20 ng/ml) treatment for the indicated time periods. Cytoplasmic and nuclear extracts were prepared, fractionated on a 10% SDS–page gel, transferred to a membrane, and then tested for IκBα, p50, p65 (cytosolic) or p50 and p65 (nuclear). Equality of protein loading and purity of nuclear/cytosolic extracts was checked by lamin B and α-tubulin Western blots. The data shown here were representative for three independent experiments with similar results

Furthermore, IKKβ assay showed that NLE significantly inhibited the kinase activity of IKKβ at a concentration from 160 to 240 μg/ml. Moreover, the inhibitory activity of NLE is comparable to the prototypical IKK inhibitor ([5-(p-fluorophenyl)-2-ureido]thiophene-3-carboxamide) provided as a positive control by the supplier (Fig. 7) again confirming previous results.

Fig. 7
figure 7

Effect of neem leaf extract at the indicated concentrations on the kinase activity of IKKβ for an incubation period of 30 min at 30°C. “no enzyme” was performed in the absence of IKKβ, and “negative control” was evaluated without any test compound, however, in the presence of IKKβ. IKK-2 inhibitor IV (Calbiochem) was used at the recommended concentration of 1 μM. Results are shown as representative mean ± SD of three individual measurements. * represents P < 0.01 to the negative control

A number of extracts were described in the literature to efficiently inhibit IKK activity. No results exist so far for neem extracts, but the aqueous extract of Dichroa febrifuga, a medicinal plant well known in China and Korea, inhibits interleukin-1β and interleukin-6 production in LPS-stimulated mouse peritoneal macrophages. These effects are mediated by the inhibition of the activity of IKK/IκB/NF-κB signaling cascade [42]. Moreover, water-soluble extracts of Panellus serotinus (Mukitake) powder showed an inhibitory effect on IKKβ, whose activation is required for NF-κB-mediated inflammatory response [40]. Finally, medicinal value of Marasmius oreades was also described as a source for bioactive substances able to induce a direct blockage of NF-κB activation at the IKK level [43].

Neem leaf extract induced apoptosis in leukemia cells

As there is a clearly documented link between inhibition of the NF-κB pathway and the induction of apoptotic cell death [1, 2], we assessed whether NLE induces apoptosis in human leukemia cell.

Accordingly, we evaluated the percentage of induction and nature of cell death and found that NLE induced cell death in K562 cells with a LC50 of 300 ± 20 μg/ml after 8 h of incubation. At a concentration of 400 μg/ml, NLE induced 75 ± 5% of cell death to the K562 cell line (Fig. 8a). Interestingly, at concentrations that inhibit NF-kB or induce cell death in cancer cells, healthy PBMCs were not significantly affected by the treatment.

Fig. 8
figure 8

Viability assay of K562 a and normal PBMCs b treated with indicated concentrations of neem leaf extract for an incubation period of 8 h. Results are shown as representative measurement ± SD of three independent experiments

In a next step, we performed a long-term viability assay over 3 days on three different leukemias (K562, Jurkat, and U937) and one prostate cancer (PC-3) cell lines (Fig. 9). K562 showed to be the most resistant cell line, whereas U937 was the most sensitive one (Table 1). At a concentration of 320 μg/ml of neem leaf extract, almost all PC-3 cells and K562 cells were killed, in contrast to Jurkat and U937 cells, for which a concentration of 200 μg/ml was sufficient to reach the same target. These concentrations seem to be very high in contrast to known cancer drugs such as taxol [67]; however, neem leaves are known to be nontoxic to humans, and the pure compound of this extract responsible for this activity will have a lower IC50 (in μM) than that of the crude extract.

Fig. 9
figure 9

Long-term Viability assay of Jurkat, K562, PC-3, and U937 cells treated with indicated concentrations of neem leaf extract for incubation periods of 24, 48 and 72 h. Results are shown as mean ± SD of three independent experiments

Table 1 IC50 of neem leaf extract in various cell lines after 48 h of treatment

In order to determine the precise modality of cell death and mechanisms, we proceeded with nuclear morphology analysis of cells, stained with Hoechst, via fluorescence microscopy. Treated cells clearly displayed apoptotic fragmentation after 8 h of treatment with 240 μg/ml. Figure 10 shows the kinetic evolution of apoptotic cell death induced by NLE and confirmed the data of the viability assay. The flow cytometry analysis clearly indicated that NLE induced a slight level of apoptotic cell death to the K562 cell line with a maximum of 18 percent at a concentration of 240 μg/ml (Fig. 11) with a high number of cells in late apoptosis. Next, we evaluated whether NLE activated a caspase-dependent apoptosis by using Western blot analysis (Fig. 12). Caspase-3, -7, -8, and -9 activation could be observed after 6 h of treatment at a concentration of 200 μg/ml; the sesterterpene heteronemin and the known cancer drug cisplatin were used as positive control [55, 60]. Additionally, the level of the anti-apoptotic proteins Bcl-xL, Bid, and XIAP was down-regulated after the identical treatment. These results proved that NLE progressively induced apoptotic cell death most likely via the mitochondrial cell death pathway.

Fig. 10
figure 10

K562 cells stained with Hoechst to detect nuclear morphology, untreated (left figure) vs. treated with 120 μg/ml of neem extract for 8 h (right picture), the latter presenting typical fragmented apoptotic nuclei. The graph shows a kinetic analysis of apoptosis induced by neem extract, evaluated as the fraction of cells with apoptotic nuclei stained with Hoechst. The data are the mean of n = 3 independent experiments ± SD

Fig. 11
figure 11

Apoptotic cells were detected after annexin V-FITC/propidium iodide (PI) staining by FACS. The cells in the lower right quadrant, which are only positive for annexin V, represent early apoptotic cells. One of three independent experiments with similar results is shown

Fig. 12
figure 12

Analysis of caspase-3, -7, -8, and -9 cleavage and BCL-xL, Bid and XIAP activation by Western blot. Cells were incubated for 8 h with different concentrations of neem leaf extract, and the activation of the protein was evaluated after 8 h in comparison with nontreated control (−), a positive control (heteronemin or cisplatin), and a vehicle control (DMSO). Beta-actin was used as a control for protein loading. One representative experiment of three independent ones was shown

Our results corroborate with published data of an ethanolic neem extract able to induce apoptosis in cancer cells from rodents and men [29, 57]. Interestingly, to the best knowledge of the authors, no effect of neem extract on XIAP was reported in literature; recently, the inhibition of XIAP emerged as an important cancer drug target [14, 19, 52]. Similar pro-apoptotic effects were described for extracts from various plants including Physalis minima L. [31], Cassia tora Linn (Leguminacea) [47], Serenoa repens [7], Uncaria rhynchophylla [26], Hibiscus [32], Smilax glabra Roxb. [51], Arnebia nobilis [64], Cimicifuga foetida [65], Brassica oleracea [35], Calendula officinalis [25], Pereskia bleo (Kunth) DC. (Cactaceae) [61], and Catha edulis (Khat) [16].


Altogether, we show here that NLE acted on different levels of the NF-κB pathway and that it induced apoptosis in leukemia cancer cell. This is, to the best of our knowledge, the first report of these TNF-α-induced, anti-inflammatory, and anti-proliferative treatment properties of NLE in human leukemia cancer cell lines. Since the observed effects are beneficial, the extract and its constituents including quercetin might be promising candidates for future studies.