Recombinant human lactoferrin carrying humanized glycosylation exhibits antileukemia selective cytotoxicity, microfilament disruption, cell cycle arrest, and apoptosis activities

Summary

Lactoferrin has gained extensive attention due to its ample biological properties. In this study, recombinant human lactoferrin carrying humanized glycosylation (rhLf-h-glycan) expressed in the yeast Pichia pastoris SuperMan5, which is genetically glycoengineered to efficiently produce functional humanized glycoproteins inclosing (Man)₅(GlcNAc)₂ Asn-linked glycans, was analyzed, inspecting its potential toxicity against cancer cells. The live-cell differential nuclear staining assay was used to quantify the rhLf-h-glycan cytotoxicity, which was examined in four human cell lines: acute lymphoblastic leukemia (ALL) CCRF-CEM, T-cell lymphoblastic lymphoma SUP-T1, cervical adenocarcinoma HeLa, and as control, non-cancerous Hs27 cells. The defined CC₅₀ values of rhLf-h-glycan in CCRF-CEM, SUP-T1, HeLa, and Hs27 cells were 144.45 ± 4.44, 548.47 ± 64.41, 350 ± 14.82, and 3359.07 ±164 μg/mL, respectively. The rhLf-h-glycan exhibited a favorable selective cytotoxicity index (SCI), preferentially killing cancer cells: 23.25 for CCRF-CEM, 9.59 for HeLa, and 6.12 for SUP-T1, as compared with Hs27 cells. Also, rhLf-h-glycan showed significant antiproliferative activity (P < 0.0001) at 24, 48, and 72 h of incubation on CCRF-CEM cells. Additionally, it was observed via fluorescent staining and confocal microscopy that rhLf-h-glycan elicited apoptosis-associated morphological changes, such as blebbing, nuclear fragmentation, chromatin condensation, and apoptotic bodies in ALL cells. Furthermore, rhLf-h-glycan-treated HeLa cells revealed shrinkage of the microfilament structures, generating a speckled/punctuated pattern and also caused PARP-1 cleavage, a hallmark of apoptosis. Moreover, in ALL cells, rhLf-h-glycan altered cell cycle progression inducing the G2/M phase arrest, and caused apoptotic DNA fragmentation. Overall, our findings revealed that rhLf-h-glycan has potential as an anticancer agent and therefore deserves further in vivo evaluation.

Introduction

Lactoferrin (Lf) has recently attracted worldwide interest due to its ample range of biological and therapeutic properties [1–4]. Lf is a natural monomeric globular glycoprotein with the iron-binding capability and a member of the transferrin family [5, 6]. Previously known as lactotransferrin, Lf is a ~ 80 kDa glycoprotein, comprising 698–720 amino acid residues, and it is highly conserved between mammalian proteins [7]. The human Lf (hLf) consists of a single 711 amino acids peptide [8] and is present in human secretions such as maternal milk, human colostrum, amniotic fluid, blood plasma, saliva, tears, vaginal fluids, semen, mucosal and other secretions [9]. In addition, Lf is a critical component of the first-line defense of the innate immune system in mammals [10]. Moreover, Lf is considered as a nutraceutical [6] and multifaceted protein due to its immunomodulatory properties, antimicrobial, anti-inflammatory, anticancer, and antioxidant activities [11–14]. For years feeding bovine and recombinant human Lf has been used as a dietary supplement and for the treatment of stomach, duodenal ulcers, diarrhea, and several bacterial and viral infections [15–18]. Furthermore, Lf has extended significant attention as a safer anticancer chemopreventive and therapeutic agent [4, 19].

Cancer is a significant public health problem worldwide, encompassing a group of more than 100 different types of diseases [20] and is the second leading cause of death in the United States [21]. It is estimated that 1.8 million new cancer cases diagnosed and 606,520 cancer deaths will occur in the United States in 2020 [21]. The World Health Organization (WHO) reported the death of 9.6 million people from cancer in 2018 worldwide [22]. It is projected that in 2020, 11,050 new cases of cancer will be diagnosed among children aged 0 to 14 in the USA, and 1,190 children will die of this disease [21]. Furthermore, leukemia is the most prevalent childhood cancer, accounting for around 25% of all cancer arising prior to age 20 [23].

Nevertheless, presently, the most common cancer treatments are surgery, radiation therapy, chemotherapy, targeted therapy, hormonal and immune therapy [24]. Although there are many advances in cancer treatment, chemotherapy remains the primary option for the treatment of most cancers [25]. However, by using chemotherapy against cancer, harmful side effects emerge, mainly due to the non-specific and systemic damage inflicted on healthy tissue, provoking nausea, vomiting, hair loss, fatigue, etc. [26]. An additional serious problem is the acquisition of drug resistance by tumor cells, which is believed to be the reason for chemotherapy failure in more than 90% of patients with metastatic cancer [27]. Therefore, the development of more specific and efficient anticancer drugs is crucial to reduce the current high mortality of cancer.

The cytotoxic mechanism of lactoferrin has not been entirely elucidated, but prior studies have revealed that it has anticancer activity [4]. Extracellular effects are reported, mainly related to the interaction of lactoferrin with the cell membrane and membrane receptors [28], as well as intracellular effects primarily related to apoptosis, inhibition of cell cycle progression and immunomodulatory action by activating immune system cells [4, 8, 29–32]. Nevertheless, recent advances in recombinant DNA technology have paved the way for obtaining purified recombinant proteins with anticancer properties [6, 33], which are becoming increasingly popular and valuable as therapeutic options [34]. Many currently approved therapeutic proteins require suitable glycosylation to ensure adequate biological properties [35] since glycosylation can influence a variety of their physiological processes [36]. Moreover, the potential human application of our lactoferrin can circumvent the immune response induced by non-humanized (i.e., yeast) types of glycosylation.

In the present study, a genetic construct containing the recombinant human lactoferrin (rhLf) gene was expressed in P. pastoris SuperMan5, a specially designed strain to efficiently produce functional humanized (h) glycoproteins (glycan) by adding (Man)5(GlcNAc)2 Asn-linked glycan (rhLf-h-glycan) via post-translational modification. This rhLf-h-glycan purified glycoprotein exerted an antiproliferative activity with a favorable CC₅₀, and selective cytotoxicity index on acute lymphoblastic leukemia (ALL) cells, with minimal damage to non-cancerous cells. Besides, the results suggest that rhLf-h-glycan inflicted its cytotoxic activity by inducing cellular morphological changes associated with apoptosis, disruption of microfilaments (F-actin), PARP-1 cleavage, alteration of cell-cycle progression, and DNA fragmentation. Taken together, all these findings suggest that the rhLf-h-glycan induces the apoptotic pathway to inflict cell death. Additionally, this novel rhLf-h-glycan protein could be considered as a potential member of the new generation recombinant proteins as an anticancer agent.

Materials and methods

Bacterial and yeast strains and their culture conditions

The Escherichia coli BL21(DE3) (Invitrogen) strain was used as a host for the maintenance and propagation of the pIGF5 genetic construct. This novel pIGF5 plasmid contains the human lactoferrin gene (Access No. M83202.1) under the control of the AOX1 gene promoter (Alcohol oxidase 1), a selection marker gene that confers resistance to zeocin, an αF signal peptide from Saccharomyces cerevisiae that allows the secretion of proteins into the medium, a bacterial replication origin, and a six-histidine amino acid sequence [4]. The Luria-Bertani (LB) media (10% peptone, 10% NaCl and 5% yeast extract, pH 7.0) supplemented with zeocin (25 μg/mL), as the selective marker, was used for cultivation and maintenance of bacteria at 37 °C. The Pichia pastoris (P. pastoris) SuperMan5 (HIS⁺) expression strain acquired from BioGrammatics Inc. (Carlsbad, CA, USA) was used for the expression of rhLf-h-glycan glycoprotein gene. This P. pastoris strain was cultured with shaking at 30 °C in YPD media (1% yeast extract, 2% peptone and 2% dextrose, pH 6.5) supplemented with 200 μg/mL zeocin as the selective marker. Furthermore, to induce the rhLf-h-glycan expression, the buffered minimal methanol-complex medium, BMMY (1% yeast extract, 2% peptone, 100 mM potassium phosphate 1.34% YNB, 0.0004% biotin, and 0.5% methanol, at pH 6.0) was used.

P. pastoris SuperMan5 transformation

The pIGF5 plasmid DNA was isolated from E. coli using the commercial Quantum Prep plasmid Kit (Bio-Rad) following the vendor’s instructions to obtain high-purity DNA. Next, this plasmid was linearized by using the Sac I restriction enzyme (Invitrogen), and 8 μg of linearized plasmid DNA was used to transform P. pastoris SuperMan5 electrocompetent cells. The transformation was carried out via electroporation in a MicroPulser Electroporator (Bio-Rad), according to the pre-established parameters for P. pastoris (1500 V, 200 Ω, and 25 mF; [37]. After 96 h of incubation at 30 °C in the presence of the 200 μg/mL zeocin selective antibiotic, zeocin-resistant clones were isolated for analysis of the integration of the hLf gene by PCR.

Expression of the rhLf-h-glycan protein in glycoengineered P. pastoris

A yeast colony surviving the two selection requisites was grown in YPD medium supplemented with 200 μg/mL zeocin at 30 °C at 250 rpm for 24 h. Then, the cells were harvested by centrifugation and incubated at a seeding density of 0.1–0.2 optical density in BMMY medium at 30 °C to induce the expression of rhLf-h-glycan. Methanol was added to a final concentration of 0.5% (v/v) every 24 h for 96 h to maintain the induction. After 96 h, the cells were collected by centrifugation at 5000 rpm for 10 min at 4 °C. The cell-free yeast culture supernatant was dialyzed against PBS pH 7.4 for 16 h. After dialysis, the protein extracts were purified using a high-affinity Ni-Charged Resin (GenScript) according to the manufacturer directions to obtain His-tagged rhLf-h-glycan recombinant proteins. The derived purified protein was dialyzed and concentrated simultaneously with ultracentrifugation filters (Amicon, Sigma Aldrich) following the vendor’s conditions. The content of endotoxin levels in the purified rhLf-h-glycan was measured with ToxinSensor Chromogenic LAL Endotoxin Assay Kit (GenScript) based on the manufacturer’s guidelines.

Western blot analysis

Purified rhLf-h-glycan was boiled with 1 × loading buffer sample for 5 min and separated by SDS-PAGE using 10% resolving and 8% stacking acrylamide gel. Proteins separated via electrophoresis were stained with Coomassie Brilliant Blue R-250. A mixture of prestained protein molecular weight marker was used for western blot analysis (20–250 kDa, Invitrogen). After electrophoresis, the separated proteins were transferred from the gels onto polyvinylidene fluoride (PVDF) microporous membrane (Pierce, Rockford, IL, USA) and blocked overnight at 4 °C with blocking solution (5% w/v nonfat dry milk in TBS-T). The blot was then incubated for 2 h at 4 °C with mouse anti-6x-His primary monoclonal antibody (Clontech) at a dilution of 1:2000 in 5% w/v of skim milk in TBS-T (0.1% Tween 20). Unbound antibody was washed away with TBS-T three times for 15 min each, followed by the addition of a secondary antibody (polyclonal goat anti-mouse conjugated to horseradish peroxidase; Thermo Scientific) at a dilution of 1:5000 in TBS-T and additionally incubated for 2 h at room temperature. Unbound antibodies were again washed away with TBS-T 3 times 15 min each. Subsequently, the membrane was developed by chemiluminescence reagent using the commercial Pierce ECL Western Blotting Substrate (Thermo Scientific), and digital images were capture by using the iBright FL1500 imaging system (Invitrogen).

Cell lines and culture conditions

The acute lymphoblastic leukemia (ALL) CCRF-CEM cell line (ATCC CCL-119) and the T cell lymphoblastic lymphoma SUP-T1 cell line (ATCC CRL-1942), which grow in suspension, were cultured under similar conditions in RPMI-1640 medium (HyClone) supplemented with 10% fetal bovine serum (Sigma-Aldrich), 100 U/mL penicillin and 100 μg/mL streptomycin (Life Technologies). Cervical adenocarcinoma HeLa cell line (ATCC CCL-2), as well as the non-cancer human foreskin fibroblasts Hs27 cell line (ATCC CRL-1634), were cultured in DMEM 1X medium (HyClone) supplemented with 10% fetal bovine serum (Sigma-Aldrich), 100 U/mL penicillin and 100 μg/mL streptomycin (Life Technologies). Consistently, the incubation conditions were maintained at 37 °C in a humidified 5% CO₂ atmosphere. Cells were prepared as previously recommended to guarantee high viability [38].

Differential Nuclear Staining (DNS) assay

The differential nuclear staining (DNS) assay was used [38–40] to analyze the potential cytotoxic activity of rhLf-h-glycan. For this assay, three cancer cell lines (Table 1) were tested. Adherent HeLa cells were detached by the addition of 0.25% trypsin solution (Invitrogen, Carlsbad, CA), diluted in serum-free DMEM medium, and incubated for approximately 10 min at 37 °C. Consistently, before preparing any experimental plate, the cell viability was monitored by using PI and flow cytometry, as previously detailed [41, 42]. After harvesting, adherent and non-adherent cells were counted using a hemocytometer and seeded at a cell density of 10,000 cells in 100 μL/well of complete media in a 96-multi-well tissue culture plate and incubated overnight at 37 °C. Subsequently, cells were treated with 50 and 200 μg/mL of rhLf-h-glycan for 48 h. The following controls were also included; PBS as solvent control, 2 mM H₂O₂ as a positive control of cell death, and untreated cells to determine the background of toxicity due to cell manipulation and to intrinsic factors commonly associated with the culture protocol. Two hours before completing the 48 h incubation time, a mixture of two fluorescent DNA intercalators was added to each well, Hoechst 33,342 (Invitrogen, Eugene, OR) and propidium iodide (PI; MP Biomedicals, Solon, OH); at a final concentration of 5 μg/mL each. Due to its high permeability, Hoechst stains the nucleus of living and dead cells, thus helping to label the total number of cells. In contrast, PI only enters cells with compromised membranes, serving as a selective marker of dead cells. Next, images were captured by using an IN Cell 2000 bioimaging system (GE Healthcare, Chicago, IL). Montages (2 × 2) from four adjacent image fields were acquired using a 10X objective directly from each well of the cell culture plates, to assess an adequate number of regions of interest (ROI, number of nuclei/cells). Data analysis to determine the percentage of dead cells from the individual wells was accomplished using an IN Cell Investigator software (GE Healthcare). Each experimental data point, as well as controls, were evaluated in triplicate. The 50% cytotoxic concentration (CC₅₀) value, which is defined as the concentration of rhLf-h-glycan that causes 50% of the cell death population, as compared to PBS-treated cells, was calculated based on an interpolation equation (https://www.johndcook.com/interpolator.html).

Table 1.

The rhLf-h-glycan Cytotoxic Concentration 50% (CC₅₀) values from four human cell lines after 48 h of exposure, and its Selective Cytotoxicity Index (SCI)

Cell linea	CC50 (μg/ml)b	SCIc
CCRF-CEMd	144.45 ±4.44	23.25
HeLae	350± 14.82	9.59
Sup-T1f	548.47± 64.41	6.12
Hs-27*	3359.07 ±164‡

^aIn-depth information on each specific cell line is available online from the ATCC webpage (http://www.atcc.org/)

^bThe cytotoxic concentration 50% (CC₅₀)

^cSCI values were calculated, dividing the CC₅₀ of the non-cancerous Hs-27 cell line by the CC₅₀ of the cancer cell line

^dPeripheral blood-derived leukemia cells, from a 4-year-old female “±” = values refer to standard deviations

^eTumor’s epithelial cells derived from an adult with cervical adenocarcinoma

^fT-lymphoblast from an 8-year-old male with T-cell lymphoblastic lymphoma

^*Non-cancerous foreskin fibroblast from a newborn boy

^‡Maximum toxicity observed experimentally was less than 20%

Determination of the selective cytotoxicity index (SCI)

The selective cytotoxicity index (SCI) for rhLf-h-glycan, which denotes the ability of an experimental compound to kill cancer cells efficiently with minimal toxicity to non-cancer cells, was calculated by the ratio of the non-cancer cells CC₅₀ to the cancer cell CC₅₀ [39].

Antiproliferative assay

The ALL CCRF-CEM cells were seeded in a 24-well plate at 5×10⁴ cells/well in 1 mL media and incubated at 37 °C overnight. Then, the cells were exposed to 200 μg/mL of rhLf-h-glycan, and the absolute number of cells per milliliter was quantified every 24 h for 72 h, using an inverted microscope and a hemocytometer [43]. Since CCRF-CEM cells grow in suspension, they were gently homogenized directly in the well, and an aliquot (~ 30 μL) was loaded onto a hemocytometer, and the cell count was performed. The rhLf-h-glycan antiproliferative activity was consistently compared to the PBS-treated cells, the diluent of the rhLf-h-glycan. Also, the PBS-treated cells were considered as 100% of cell growth and used as a reference to calculate the percentage of rhLf-h-glycan-treated cell proliferation. The results are expressed as the average of eight independent measurements [43]. Also, untreated cells were included to observe the viability of the cells during the whole incubation time monitoring the inherent effects of cell manipulations in the cell cultures. Additionally, cells exposed to 1 μM paclitaxel (PTX) were also included as an antiproliferation positive control.

Analysis of morphological changes associated with apoptosis

The ALL CCRF-CEM cells growing in 96-well plates, as described above, were treated with 200 μg/mL of rhLf-h-glycan and incubated for 48 h. The staining was performed with Hoechst 33,342 and PI (DNS assay) to observe the cellular morphological changes induced by rhLf-h-glycan. The live-cell images were attained via confocal microscopy (LSM 700; Zeiss, New York, NY) using an EC plan-Neofluar 40X / 1.30 oil DIC objective. Images from PBS-treated cells (solvent control) were obtained and included for comparison purposes. Three individual images were captured as follows: both the brightfield and Hoechst (blue) images were acquired simultaneously by using the 405 nm laser, whereas the 555 nm laser was used to excite the PI (red). Typically, the pinholes in each channel were of 1 Airy Unit (AU), for image acquisition purposes. Next, high-resolution digital fluorescent and brightfield images were analyzed by using the ZEN 2009 software (Zeiss, New York, NY).

PARP-1 cleavage

Next, the potential cleavage of poly (ADP-ribose) polymerase (PARP)-1 induced by rhLf-h-glycan was examined as an additional marker of apoptotic cell death. To detect the subcellular localization of the large cleavage 89 kDa PARP-l fragment, HeLa cells were seeded in a 24-well plate at a cell density of 100,000 cells in 1 mL/well of culture medium and exposed to 200 μg/mL rhLf-h-glycan for 48 h. At the end of the incubation time and without removing the culture medium, freshly prepared 8% formaldehyde was added to each well as a fixation solution to obtain a final concentration of 4%, and the plate was incubated for 20 min at room temperature. The fixative solution was removed, and 200 μL of PBS containing 0.1% Tween 20 was added to each well for washing and permeabilization purposes and incubated for 10 min at room temperature. Two additional washes were performed, and the permeabilizing solution was removed. Next, 200 μL of 5% BSA in TBS-T was added to each well and incubated for 1 h at room temperature for blocking purposes. Then cells were stained using 50 μL PBS per well containing 0.1% Tween 20, 5 μg/mL DAPI and a 1:50 dilution of Alexa Fluor 488 conjugated polyclonal anti-PARP-1 antibody (clone DM1A; Millipore Sigma), and incubating for 1 h at room temperature in the dark. Finally, the cells were washed three times, using 200 μL of the permeabilization solution as described above. Images were captured via two fluorescence channels (Alexa-488 and DAPI), as well as a brightfield channel, using a confocal microscope (LSM-700; Zeiss) equipped with an EC PLan-Neofluar 40X/1.30 oil DIC objective.

Detection of the rhLf-h-glycan effect on the cytoskeleton architecture

For this series of experiments, a confocal microscope, the HeLa and Hs27 cell lines were used. Cells were processed as described above and seeded at a density of 2,500 cells in 100 μL of medium per well. Optically transparent 96-well plates (BD Falcon) were used. The microplates were incubated overnight to promote cell adhesion, and cells were exposed to 144.45 μg/mL (CC₅₀) of rhLf-h-glycan for 2 h. At the end of the incubation time and without removing the culture medium, freshly prepared 8% formaldehyde was added to each well as a fixation solution to obtain a final concentration of 4%, and the plate was incubated for 20 min at room temperature. The fixation solution was removed, and 200 μL of 0.1% Tween 20 in PBS was added to each well for washing and permeabilization purposes and incubated for 10 min at room temperature. Two additional washes were performed, and the permeabilizing solution was removed. 200 μL of 5% BSA in TBS-T was added to each well and incubated for 1 h at room temperature. Cells were then stained with 50 μL per well of a 0.1% Tween 20 PBS solution containing 5 μg/mL of DAPI, 0.156 μM of Alexa Fluor 568-conjugated phalloidin (Invitrogen), and 0.5 μg/mL of Alexa Fluor 488-conjugated anti-α-tubulin monoclonal antibody (clone DM1A; Thermo Fisher Scientific, Rochester, NY) incubating 1 h at room temperature. All the staining stages were executed in the dark to diminish fluorescence photobleaching effects [44]. Finally, the cells were washed three times, using 200 μL of the permeabilization solution as previously detailed [4]. Images were captured on three fluorescent channels (Alexa-488, Alexa-568, and DAPI) using a confocal microscope (LSM-700; Zeiss) equipped with an EC Plan-Neofluar 40X / 1.30 oil DIC objective and using a 1-Airy Unit (AU) pinhole setting for each channel consistently. For the acquisition and analysis of confocal images, the ZEN 2009 software was utilized (Zeiss, New York, NY).

The rhLf-h-glycan influence on the cell cycle profile

Analysis of the changes in the cell cycle profile was carried out by measuring the cellular DNA content using a Gallios flow cytometer (Beckman Coulter). This flow cytometer is equipped with a solid-state 405 nm violet laser capable of exciting DAPI (4’, 6-diamidino-2-phenylindole, dihydrochloride), which was used to label DNA. CCRF-CEM cells were seeded in a 24-well plate at a cell density of 25,000 cells in 1 mL of culture medium per well and exposed to increasing doses of rhLf-h-glycan (28.89 μg/mL, 57.78 μg/mL and 86.67 μg/mL) for 72 h. The controls included in this series of experiments were those described for the above examinations. After the incubation time, the cells were harvested in flow cytometry tubes, placed on ice, and centrifuged at 262 g for 5 min. The resulting cell pellets were gently resuspended in 100 μL of PBS, and 200 μL of Nuclear Isolation Medium (NIM)-DAPI solution (NPE Systems, Inc. and Beckman Coulter) were added; cells were immediately analyzed by flow cytometry [39, 45]. The NIM-DAPI reagent is a combination of DAPI and NP-40 detergent that is capable of permeabilizing and staining the nuclei of the cells, allowing their DNA content to be quantified using an FL-9 detector and the 405 nm laser for excitation purposes (Gallios). This FL-9 detector captures the fluorescence signal emitted when DAPI is intercalated with DNA (461 nm), skipping the emission signal when DAPI complexes with RNA (~ 500 nm). Approximately 10,000 events per sample were acquired to obtain a well-defined cell cycle distribution profile and analyzed using Kaluza flow cytometry software (Beckman Coulter). Each experimental point, as well as its corresponding controls, were processed simultaneously and evaluated via three independent measurements. The percentages of cells with different DNA content distributions were determined from the histogram with gates for sub-G0/G1, hypodiploid; G0/G1, diploid; S, hyperdiploid; and G2/M, tetraploid. The S phase population was defined as the percentage of cells with a DNA content between G0/G1 (diploid) and G2/M (tetraploid) [43].

Statistical analysis

Each experimental point represents at least three independent measurements. The results are depicted as the average of the multiple measurements with their corresponding standard deviations, to indicate experimental variability. The statistical significance of the differences between two experimental samples was performed using a two-tailed Student’s t-test, and a P-value equal to or less than 0.05 was deemed significant.

Results

The rhLf-h-glycan protein purification

After P. pastoris SuperMan5 strain transformation with the pIGF5 plasmid carrying the human Lf gene, a selected colony was cultured for 96 h using the induction medium in the presence of methanol to promote the expression of rhLf-h-glycan. Total protein extracts were recovered from the supernatant culture medium and extensively dialyzed. Subsequently, the rhLf-h-glycan protein was purified by using a high-affinity Ni-charged resin affinity column, followed by SDS-PAGE analysis and immunodetection (Fig. 1a) via western blot (Fig. 1b). Both SDS-PAGE gels stained with Coomassie brilliant blue, as well as the signal obtained by chemiluminescence from the western blot, confirmed the presence of a band at approximately 80 kDa (see arrows in Fig. 1a–b), which is consistent with the theoretical inferred molecular weight of native human lactoferrin ranging from 75 to 95 kDa [4, 46–48]. These findings indicate that the expression and purification of free-endotoxin rhLf-h-glycan protein synthesized in P. pastoris SuperMan5 strain were accomplished, and it was suitable to perform further biological experiments.

Fig. 1.

Purification and detection of 6x Histidine tagged rhLf-h-glycan protein expressed in P. pastoris SuperMan5 strain. a SDS-PAGE of purified rhLf-h-glycan from yeast supernatant using immobilized nickel affinity chromatography visualized by Coomassie brilliant blue staining: 1, molecular weight marker, kDa; 2, non-purified protein sample; 3, purified rhLf-h-glycan protein. b Western blot immunodetection with anti-6x His of purified rhLf-h-glycan identified via chemiluminescence. The arrows indicated the presence of a ~ 80 kDa protein, (A) SDS-PAGE and (B) western blot

The rhLf-h-glycan cytotoxic concentration 50 (CC₅₀) on human cancer and non-cancerous cells and its selective cytotoxicity index (SCI)

The rhLf-h-glycan cytotoxic effects on three cancer and one non-cancerous cell lines were determined by the differential nuclear staining (DNS) assay [38], which uses two DNA Intercalating fluorescent dyes, Hoechst and propidium iodide (PI). For each cell line, dose-response graphs were created to assess the rhLf-h-glycan CC₅₀, assisted by a linear interpolation approach [49]; Table 1; Fig. 2; [50, 51]; a representative graph is depicted in Fig. 2a. Also, some representative Hoechst (blue) and PI (red) merged images used to quantify the rhLf-h-glycan cytotoxicity percentages are displayed (Fig. 2b–e). Results indicated that the human acute lymphoblastic leukemia CCRF-CEM cells were the most sensitive to the rhLf-h-glycan treatment, followed by HeLa, SUP-T, and Hs27 cells, with a CC₅₀ of 144.45 ±4.44, 350 ±14.82, 548.47 ± 64.41, and 3359.07 ±164 μg/mL, respectively (Table 1). The CCRF-CEM cells treated with PBS as solvent control and untreated cells exhibited less than 5% of death cells (Fig. 2a, c, d). In contrast and as expected, cells exposed to 2 mM of H₂O₂ displayed the highest percentage of cytotoxicity (Fig. 2a, e). Moreover, rhLf-h-glycan revealed tumor-selective cytotoxicity as evidenced by favorable SCI values on all the three cell lines tested, when compared with non-cancerous Hs27 cells. The SCI values for CCRF-CEM, HeLa, and SUP-T cells were 23.25, 9.59, and 6.12, respectively (Table 1). Thus, based on the CC50 and SCI values, rhLf-h-glycan exhibited its best cytotoxicity and selectivity on the human leukemia CCRF-CEM cells (144.45 μg/mL and 23.25 respectively), was selected to perform most of the following experiments (Table 1). Additionally, the HeLa cells were also used to accomplish the PARP-1 cleavage and the cytoskeleton structure experiments, since they typically grow as monolayer adherent cells having larger cytoplasm network than ALL CCRF-CEM cells.

Fig. 2.

The rhLf-h-glycan exerted cytotoxicity on human ALL CCRF-CEM cells in a dose-dependent manner. After 48 h of incubation, the toxicity was observed in ALL cells by a differential nuclear staining (DNS) assay. a Percentage of cytotoxicity (y-axis). Each bar represents the average of four independent measurements, and error bars correspond to standard deviation. Cytotoxic concentration 50% (CC₅₀) was calculated via linear interpolation. Representative fluorescent images that merge the Hoechst and propidium iodide (PI) channels used to obtain the percentage of cytotoxicity are illustrated in b-e. b Cells treated with 200 μg/mL of hLf-h-glycan; c Cells treated with the solvent control (PBS); d Untreated (Unt) cells to determine cell viability during the entire incubation period; e Cells exposed to 2 mM hydrogen peroxide (H₂O₂) as a positive control of cytotoxic activity. The live and dead cell nuclei stained with Hoechst are shown in blue and comprise the total number of cells, whereas dead cells nuclei stained with PI are denoted in red. Statistical significance between samples was calculated using Student’s t-test (P-value). When experimental samples were compared with each other, the P-value was 0.00059. Whereas when 50 μg/mL and 200 μg/mL rhLf-h-glycan-treated cells compared with PBS-treated and untreated cells, the P-value was consistently P < 0.0001(* and ‡) in all data sets, respectively

rhLf-h-glycan inhibited cell proliferation significantly on human acute lymphoblastic leukemia CCRF-CEM cells

The rhLf-h-glycan antiproliferative activity was analyzed on the ALL cells, quantifying the total number of cells per milliliter every 24 h for 72 h. Cells treated with rhLf-h-glycan showed a statistically significant decrease in cell proliferation in a time-dependent manner (Fig. 3); when compared 24 h to 48 h, P = 0.0021; and when compared 24 h with 72 h, P < 0.0001. After normalization with the PBS-treated cells, considered as 100% of proliferation for each incubation time, the proliferation percentages of rhLf-h-glycan-treated cells were calculated (Fig. 3). The PBS-treated cells (solvent control), and the untreated cells, exhibited similar growth pattern, increasing the number of cells concomitantly to the measured incubation time (from 24 to 72 h). Furthermore, the paclitaxel (PTX)-treated cell, used as an antiproliferative control, resembled the pattern exhibited by the rhLf-h-glycan-treated cells, in particular after 72 h of incubation (Fig. 3). Findings indicate a significant difference in the antiproliferation of rhLf-h-glycan-treated cells (P <0.0001), as compared to the PBS-treated cells (Fig. 3). In particular, after 48 h and 72 h of treatment, cell proliferation was reduced by 71.89% and 88.16%, respectively (Fig. 3; data table). These results suggest that rhLf-h-glycan has the ability to suppress significant cell proliferation in human acute lymphoblastic leukemia CCRF-CEM cells in a time-dependent modality.

Fig. 3.

The rhLf-h-glycan exhibited antiproliferative activity on human ALL CCRF-CEM cells in a time-dependent manner. Cells were exposed to 200 μg/mL of rhLf-h-glycan. The total numbers of cells per milliliter (y-axis) were quantified using a Neubauer chamber (hemocytometer) every 24 h for 72 h of treatment. PBS-treated cells, solvent control, were also examined and used as 100% of growth. Untreated cells (Unt) were used as an indicator of cellular viability during all incubation periods. As a positive control of an antiproliferative effect, 1 μM paclitaxel (PTX) was included. Each bar represents the average of eight independent measurements, and the error bars represent the standard deviation. The statistical significance of the differences between incubation periods of rhLf-h-glycan-treated cells is of P = 0.0021 for 24 h and 48 h, and P < 0.0001 for 24 h and 72 h, respectively. Whereas rhLf-h-glycan-treated cells compared with PBS-treated cells using the same incubation times for both cells, the P-value was consistently < 0.0001(*, †, ‡) in all data sets (data table)

Cellular morphological changes caused by rhLf-h-glycan on ALL CCRF-CEM cells

Next, the CCRF-CEM cells were incubated for 48 h with 200 μg/mL of rhLf-h-glycan to explore whether the cytotoxic effect was related to its ability to induce apoptosis-associated morphological changes [4]. For this purpose, cells growing in 96-well plates were dual stained with Hoechst, and PI and confocal images were captured in a live-cell manner. The results revealed that rhLf-h-glycan induced cell death, eliciting apoptosis-related morphological changes in CCRF-CEM cells, like blebbing, cell shrinkage, nuclear condensation, nuclear fragmentation, and formation of apoptotic bodies (Fig. 4a–d). Also, live-cell images from untreated control CCRF-CEM cells are shown in Fig. 4e–h for comparative purposes. Images from untreated cells exhibit only healthy cells, PI negative, without the alterations described above (Fig. 4e–h). Thus, based on changes in cellular morphology, it is likely that rhLf-h-glycan protein induces apoptosis.

Fig. 4.

The rhLf-h-glycan induces cellular shrinkage, blebbing, nuclear fragmentation, and chromatin condensation, which are apoptosis-associated morphological changes in CCRF-CEM cells. Cells were exposed to 200 μg/mL of rhLf-h-glycan for 48 h and stained with Hoechst and PI. Representative live-cell confocal microscope images, captured using Hoechst channel (a, e), bright field differential interference contrast (DIC; b, f), propidium iodide (PI) channel (c, g). Merged images, including the preceding three images, are depicted (d, h). The white arrows are denoting nuclear fragmentation, cell shrinkage, chromatin condensation, and apoptotic bodies (a-d). The black arrows are indicating bleb formation (b and d). Also, the white arrows are signaling a cell with a non-condensed nucleus, non-cell shrinkage, and non-blebs formation, as well as PI negative, therefore a healthy living cell (e, f, and h). Scale bar = 10 μm

The rhLf-h-glycan induces PARP-1 cleavage in HeLa cells

To determine whether rhLf-h-glycan induces proteolytic steps indicative of apoptosis, HeLa cells were analyzed for PARP-1 cleavage, a biochemical hallmark of apoptosis [52]. To achieve this objective, a specific antibody that binds to the large 89 kDa fragment of PARP-1 and confocal microscopy were used. After treatment with rhLf-h-glycan, PARP-1 was cleaved in HeLa cells (Fig. 5), as evidenced by the presence of an abundant nuclear green fluorescence signal that is colocalized with the DAPI nuclear counterstaining signal, used as a control (Fig. 5a–d). Moreover, during this analysis was confirmed that rhLf-h-glycan produced nuclear condensation and fragmentation (Fig. 5a), also as indicated above in Fig. 4a–d. In contrast, the absence of the green fluorescence signal from untreated control HeLa cells was noted (Fig. 5e–h). Thus, these outcomes support the notion that rhLf-h-glycan protein is activating the apoptosis pathway to induce cell death.

Fig. 5.

The rhLf-h-glycan provoked proteolytic PARP-1 cleavage on HeLa cells, a biochemical hallmark of apoptosis. Cells were treated with rhLf-h-glycan and examined after 24 h of treatment by using a specific antibody (Alexa Fluor 488 dye conjugated; green) binding to the human PARP-1 large 89 kDa fragment, and confocal microscopy. Representative live-cell confocal microscope images, captured using DAPI channel (nuclei; a, e), differential interference contrast (DIC; b, f), Alexa-488 dye conjugated (c, g) and merging of the preceding three images (d, h), are depicted

rhLf-h-glycan induces the cytoskeleton disruption in HeLa and Hs27 cells

To further characterize the effect of rhLf-h-glycan-induced apoptosis, we analyzed the potential of this glycoprotein on the perturbation of the cytoskeleton structure in both HeLa and Hs27 cells. Since the adherent HeLa and Hs27 cells are easily visualized for alterations in their cytoskeleton structure, they were chosen over the leukemia cell lines for these analyses. Two fluorescently conjugated reagents were used to label both microfilaments (F-actin) and microtubules (polymerized tubulin). Immunofluorescence images were captured using a confocal microscope (LSM 700: Zeiss). As detailed above, cells were seeded in 96-well plate format and exposed to 200 μg/mL rhLf-h-glycan for 2 h, followed by formaldehyde fixation, Tween 20 permeabilization, and BSA blocking. Next, cells were stained with phalloidin-Alexa 568 (red; microfilaments), anti-tubulin-Alexa 488 (green; microtubules), and DAPI (blue; nuclei) [4]. High-resolution digital fluorescent images were captured and analyzed, assisted with ZEN 2009 software (Zeiss, New York). Based on immunofluorescence images, abundant microfilaments (F-actin) were visualized through the cell bodies in PBS-treated HeLa and Hs27 cells (control), displaying a comparatively well-organized structure (Figs. 6a–c and 7a–c. Moreover, actin microfilaments are also useful to delimitate the periphery of the cells. In contrast, results from both rhLf-h-glycan-exposed HeLa (Fig. 6d–f) and Hs27 (Fig. 7d–f) cells displayed perturbation in the disposition of microfilaments (F-actin). Since the filaments were no longer evident, and the red signal formed an abundant speckled/punctuated pattern finely distributed throughout the cytoplasm of the cells (Figs. 6d–f and 7d–f). Concomitantly, the overall size of the rhLf-h-glycan-exposed cells decreased markedly, denoting cell shrinkage (Figs. 6d–f and 7d–f), as compared to PBS-treated control cells (Figs. 6a–c and 7a–c). In contrast, the filamentous shape structures of the microtubules were conserved, suggesting that they were not affected considerably. These results indicate that rhLf-h-glycan caused an alteration in the cytoskeleton F-actin microfilament network, as evidenced by the appearance of a stippled pattern in both HeLa and Hs27 cells. Thus, the rhLf-h-glycan induced microfilament disruption as an additional biochemical event to inflict its cytotoxicity.

Fig. 6.

The rhLf-h-glycan disrupted the microfilaments (F-actin) of the cytoskeleton organization in HeLa cells. Cells treated for 2 h with rhLf-h-glycan were triple stained with Alexa-568-conjugated phalloidin, Alexa-488 anti-tubulin, and DAPI. Representative immunofluorescence confocal microscopy images of HeLa cells displaying the microfilaments (F-actin; Alexa-568 red channel) and microtubules (tubulin; Alexa-488 green channel) and nucleus (blue channel; DAPI), respectively. The left column of images (a, d) correspond to the F-actin microfilaments (red), the center column (b, e) microtubules (green) and the right column (c, f) merged images of the Alexa-568 and Alexa-488 channels with DAPI (blue; nuclei) channel. (d) White arrows indicate the speckled/punctuated pattern of the F-actin induced by rhLf-h-glycan. Scale bar =10 μm

Fig. 7.

The rhLf-h-glycan disturbed the structure of F-actin microfilaments in Hs27 cells. Cells were exposed for 2 h to rhLf-h-glycan and triple stained with Alexa-568-conjugated phalloidin, Alexa-488-conjugated anti-tubulin, and DAPI. Representative immunofluorescence confocal microscopy images of Hs27 cells showing the microfilaments (F-actin; Alexa-568 red channel) and microtubules (tubulin; Alexa-488 green channel) organization and nucleus (blue channel; DAPI), respectively. The left column of images (a, d) correspond to the F-actin microfilaments (red), the center column (b, e) microtubules (green), and the right column (c, f) merged images of the Alexa-568 and Alexa-488 channels with DAPI (blue; nuclei) channel. (d) White arrows are indicating the speckled/punctuated pattern of the F-actin induced by rhLf-h-glycan. Notice that the scale bar size (10 μm) is larger in panels d-f, as compared with panels a-c

rhLf-h-glycan induces DNA fragmentation and cell cycle G2/M arrest on acute lymphoblastic leukemia CCRF-CEM cells

To investigate whether rhLf-h-glycan was able to modify the cell cycle profile on ALL CCRF-CEM cells, as well as to elucidate its potential DNA fragmentation activity, a flow cytometer was utilized [4, 39, 43, 45, 50]. To accomplish this task, cells were exposed for 72 h to a low concentrations gradient of rhLf-h-glycan to avoid massive cell destruction; 28.89 μg/mL (CC₁₀), 57.78 μg/mL (CC₂₀) and 86.67 μg/mL (CC₃₀). After treatment with 28.89 μg/mL and 57.78 μg/mL rhLf-h-glycan, the percentage of the sub-G0/G1 subpopulation of cells were similar to PBS-treated and untreated control cells (Fig. 8a and f). In contrast, a significant increase in the sub-G0/G1 subpopulation (hypodiploid; P =0.001), as compared to those treated with PBS, was found in cells treated with 86.67 μg/mL of rhLf-h-glycan (Fig. 8a and g). An increase in the subpopulation frequencies of cells in both G0/G1 (diploid) and S (hyperdiploid) phase was undetectable (Fig. 8b and c). Conversely, a significant increment (P = 0.001) in the G2/M subpopulation was detected from rhLf-h-glycan-treated cells, when compared to PBS-treated and untreated control cells (Fig. 8d and g). Also, in H₂O₂-treated cells, used as an inducer for DNA fragmentation, large quantities of cells were found in the sub-G0/G1 subpopulation (75.0 ± 3.86%; Fig. 8a and h). These results indicate that rhLf-h-glycan-treated CCRF-CEM cells exhibit both apoptosis-induced DNA fragmentation (P = 0.018 and P =0.009) and slowed down cell cycle progression, as revealed by arrest at the G2/M phase, in a dose-dependent manner.

Fig. 8.

The rhLf-h-glycan caused DNA fragmentation and cell-cycle G2/M arrest on ALL CCRF-CEM cells in a dose-dependent manner. After 72 h of incubation with rhLf-h-glycan, cells were collected, permeabilized, and DAPI-stained in a single step to be analyzed via flow cytometry. a-d The quantification of event/cell frequency percentages are included in the y-axis, and the different treatments are plotted along the x-axis. e-h Representative single-parameter histograms, including four gates that encompass the percentage of cell frequency per phase of the cell cycle. (e-h) The gates from left to right are as follows: sub-G0/G1 (hypodiploid), counted as DNA fragmentation-apoptotic cells; G0/G1, S, and G2/M phases. Event (cells) counts are plotted along the y-axis, whereas DNA content is along the x-axis. Controls included in this series of experiments were: rhLf-h-glycan solvent control, PBS (e), and 2 mM H₂O₂ as a positive control of DNA fragmentation inducer (h). Each bar represents an average of three independent measurements, and the error bars represent the corresponding standard deviation. The significance of the differences between CC₁₀ rhLf-h-glycan-treated cells as compared to PBS-treated cells, and also, with untreated cells, was of P =0.0038 (*) and P = 0.0044 (‡), respectively

Discussion

Cancer is the second leading cause of death worldwide; about 1 in 6 deaths is due to cancer [22]. Therefore, in this study, the rhLf-h-glycan protein was tested for its capability to kill three different cancer cell lines in vitro. Additionally, the cell death mechanism induced by rhLf-h-glycan was investigated. Moreover, by including a human non-cancerous cell line, it was possible to examine whether rhLf-h-glycan possesses selective cytotoxicity (SCI) against cancer cells.

The live-cell differential nuclear staining (DNS) assay, validated for High-Throughput Screening (HTS) initiatives, was used to quantify the rhLf-h-glycan cytotoxicity [38, 39]. This DNS assay has been demonstrated to be robust, reliable, and suitable for both primary and secondary screens initiatives searching for potential cytotoxic experimental compounds, with Z’ factors of 0.86 for 96-well plate format [38, 39, 53]. In addition, the DNS live-cell staining and imaging strategy circumvent the time-consuming fixing, permeabilizing, washing, and incubation phases linked with fluorophore-conjugated antibody and other reagents created for cell labeling purposes. For ease of detection, a 96-well plate reader was used for this purpose (bioimager; IN Cell 2000; GE Healthcare). Our results revealed that rhLf-h-glycan efficiently killed all three-cancer cell lines tested with encouraging CC50 values for CCRF-CEM, HeLa, and SUP-T cells (144.45 ± 4.44, 350 ± 14.82, and 548.47 ± 64.41 μg/mL, respectively), with much less noticeable toxicity on non-cancerous Hs27 cells (CC₅₀ of 3359.07 ± 164 μg/mL). Other authors have reported different CC₅₀ values from those obtained in this work. For example, Kim et al. [54] reported caprine Lf CC₅₀ values of 27.5 μg/mL for ZR-75–1 breast cancer cells, 717.2 μg/mL for cervical cancer HeLa cells, and 258.9 μg/mL for colon cancer HT-29 cells. In addition, Ma et al. [55] reported a CC₅₀ of 11.86 mg/mL of Lf purchased from Seebio Company (Shanghai, China) in colon cancer Caco-2 cells. In this study, of the three cell lines tested, the rhLf-h-glycan protein resulted more efficient in killing ALL CCRF-CEM cells.

A potential chemotherapeutic agent must have the ability to kill cancer cells selectively, without harming normal healthy cells. Therefore, the development of chemotherapeutic drugs with higher selective cytotoxicity (SCI) towards cancer cells, with little, if any, activity on non-cancerous cells is highly desirable. Selective cytotoxicity index (SCI) values of 1 or less indicate a compound lack of selectivity to kill cancer cells. In contrast, SCI values greater than 1 are indicative of selective compound toxicity to cancer cells, as compared to non-cancer cells. So any experimental compound to be considered as a potential anticancer drug requires high SCI values [39]. It has been reported that bovine Lf possesses selective cytotoxicity in breast cancer cells (T-47D, MDA-MB-231, Hs578T, and MCF-7), compared to non-cancerous MCF-10A cells, and also in gastric cancer AGS cells, compared to non-cancerous human kidney embryonic cells HEK; nevertheless, the SCI was not obtained in both publications [32, 56]. In contrast, Iglesias-Figueroa et al. [4] demonstrated the selective effect of rhLf on cancer cells, reporting a favorable selective cytotoxicity index (SCI) on six breast cancer cell lines. The best SCI values were of 11.68 and 13.99 from the MDA-MB-231 and MDA-MB-231-LM2–4 cell lines, respectively [4]. On the other hand, current clinical cancer therapy, including the anticancer paclitaxel (PTX), and doxorubicin, lacks selectivity for cancer cells, resulting in severe toxicity for cancer patients undergoing these treatments [57, 58]. In this study, rhLf-h-glycan exhibited selective cytotoxicity on human leukemia CCRF-CEM cells, cervical cancer HeLa cells, and lymphoma SUP-T1 cells (23.25, 9.59, and 6.12, respectively). These findings make rhLf-h-glycan protein an attractive potential anticancer agent, that causes little harm on non-cancerous cells, unlike the most widely used current standard clinical anticancer drugs, which causes damage to both cancer and non-cancerous cells [59].

Proliferation inhibition has been demonstrated in cancer cells treated with human lactoferrin (hLf), bovine (bLf) and caprine (cLf) [54, 60, 61]. It was reported that bLf caused an 80% reduction in cell proliferation in HeLa cells after 48 hours of exposure (Luzi et al. [60]. Also, Gibbons et al. [61] demonstrated that both forms of bLf, iron-free and iron-saturated, reduced cell proliferation by up to 74.86% in breast cancer MDA-MB-231 cell line after 48 h of treatment. Moreover, Duarte et al. [62] obtained decreases in the proliferation rate of 35.5 and 52.5% for breast cancer HS578T and T47D cells, respectively. Also, 5 mg/mL of camel milk Lf caused inhibition of 56% of growth in colon cancer HCT-116 cells, after 48 h of treatment, [63]; however, the camel Lf concentration used was very high. In our study, 200 μg/mL of rhLf-h-glycan reduced cell proliferation by 71.89% and 88.16% in human leukemia CCRF-CEM cells, after 48 h and 72 h of treatment, respectively (Fig. 3; data table). These findings suggest that rhLf-h-glycan protein is a potentially attractive antileukemia therapeutic agent.

Apoptosis is an essential physiological mechanism occurring during homeostasis or in response to toxic aggression and is defined by a very distinct set of biochemical and morphological features [64]. Apoptosis was evolutionarily developed to remove dying cells at any given time, preventing unwanted immune or/and inflammatory responses in vivo [65]. Moreover, apoptosis includes a complex series of dramatic alterations to cellular architecture. Therefore, to elucidate whether the rhLf-h-glycan was capable of inducing the complex cellular morphological changes related to apoptosis, CCRF-CEM cells, together with the live-cell DNS staining and confocal microscopy, were utilized [4]. An earlier report has revealed that bLf was incapable of eliciting morphological changes related to apoptosis in MDA-MB-231 cells [32]. Nevertheless, in this study, findings indicated that rhLf-h-glycan induced plasma membrane blebbing, cell shrinkage, nuclear fragmentation, nuclear condensation, and apoptotic bodies in CCRF-CEM cells (Fig. 4). Overall, our results suggest that rhLf-h-glycan protein is able to cause cellular morphological changes attributed to apoptosis.

A previous report indicated that the same rhLf gene used in this study, but expressed in a non-humanized P. pastoris strain, was able to induce phosphatidylserine externalization in triple-negative breast cancer MDA-MB-231 cells; an early apoptotic occurrence [4]. Therefore, it was decided to investigate whether the rhLf-h-glycan was capable of provoking the poly (ADP-ribose) polymerase (PARP)-1 cleavage, a downstream apoptotic hallmark. PARP-1 is an essential enzyme functioning to repair DNA damage and to preserve genomic integrity [66]. The occurrence of PARP-1 cleavage is a crucial step in cellular destruction during apoptosis [67]. Furthermore, under apoptotic conditions, PARP-1 hyperactivation enhances the depletion of ATP from cellular energy stores promoting cell death [68]. Both PARP-1 activation and cleavage play an essential and complex role in apoptosis [69]. During apoptosis, once caspase-3/7 (cysteine proteases) are activated, their primary downstream substrate is PARP-1. The whole PARP-1 protein is an abundantly confined nuclear enzyme (116 kDa), and after caspase-3/7-mediated cleavage, it releases a 24 kDa and 89 kDa protein fragments [52]. Subsequently, we examined whether rhLf-h-glycan-treated HeLa cells were able to experience PARP-1 cleavage by using a specific antibody (Alexa Fluor 488 conjugated; green signal) binding to the large fragment (89 kDa) of PARP-1. Thus, rhLf-h-glycan was able to induce PARP-1 cleavage in HeLa cells, as evidenced by the presence of abundant nuclear green fluorescence signal colocalizing with the DAPI signal used to stain the nuclei in blue (Fig. 5). PARP-1 cleavage is an additional biochemical evidence indicating that rhLf-h-glycan protein uses the apoptotic pathway to inflict cell death.

The cytoskeleton comprises a complex network of polymeric proteins, which determines the shape of the cell [70]. Eukaryotic cells mainly contain three distinct types of polymerized proteins that serve as structural elements in the cell’s cytoskeleton; microfilaments (F-actin), intermediate filaments, and microtubules (polymerized tubulin; [71]. Thus, F-actin provides the highest resistance to deformation, and the microtubules act together with the intermediate filaments to stabilize the cytoskeleton [72]. To validate whether the cytotoxicity induced by rhLf-h-glycan involved the cytoskeleton disorganization on HeLa and Hs27 cells, fluorescence confocal microscopy analyses were conducted. Results revealed that rhLf-h-glycan caused uncoupling of the typical cytoskeleton organization in both HeLa and Hs27 cells by shrinking the microfilaments (F-actin) network, associated with the formation of an irregular cytoplasmic punctuated pattern (aggresomes). The cytoskeleton alteration commonly leads to the creation of abnormal aggregates of peptides or proteins due to incorrect folding or unusual protein-protein interactions [4, 73]. Prior studies have revealed that bLf caused alterations in the structure and dynamics of the cytoskeleton in breast cancer cells (Pereira et al. [74]. Other reports have shown a correlation between the reorganization of cytoskeleton proteins and apoptosis [75, 76]. Also, Grzanka et al. [77] reported a marked pattern of F-actin at the apoptotic body formation site. Whereas Levee et al. [78] also demonstrated that apoptotic cells contained a discrete microfilament network concentrated at the apoptotic body formation site. Reorganization of the microfilament network at the start of the apoptosis process, in combination with F-actin depolymerization, is an essential element for the formation of apoptotic bodies. Thus, the ability of rhLf to cause a change in the cytoskeleton structures was explored, and our results indicated that rhLf-h-glycan caused microfilament (F-actin) shrinkage, destroying their typical filamentous structure creating a punctuated cytoplasmic pattern (Figs. 6 and 7). These findings are consistent with the observed cell blebbing, which is considered to be originated by the disconnection of the cytoskeleton from the cytoplasmic membrane. Thus, this microfilament perturbation is an additional biochemical occurrence suggesting that rhLf-h-glycan protein is activating the apoptotic pathway to inflict its cytotoxicity.

Lactoferrin’s anticancer activity is also based on its ability to trigger cell cycle arrest [79, 80], and this effect of Lf has been reported to be dependent on the cell type. For instance, Xiao et al. [81] reported that rhLf treatment of three of the four head and neck cancer cell lines studied resulted in significant cell cycle G0-G1 arrest. bLf treatment was also shown to arrest breast cancer MDA-MB-231 cells at the G2 phase of the cell cycle, whereas, on MCF-7 cells, the arrest was at the G1 or G2 phase after using low or high bLf concentrations, respectively [32]. In contrast, by exposing triple-negative breast MDA-MB-231 cells to rhLf, a cell cycle arrest at the S phase was reported [4]. In this report, we confirmed that rhLf-h-glycan indeed causes the cell cycle arrest at the G2/M phase in leukemia CCRF-CEM cells. Additionally, during this analysis, it was noted that rhLf-h-glycan was also able to elicit DNA fragmentation, as evidenced by an increase of the sub-G0/G1 (hypodiploid) subpopulation. Thus, our results suggest that rhLf-h-glycan causes cell cycle G2/M arrest and induces DNA fragmentation in a dose-dependent modality on ALL cells.

Conclusions

Our results demonstrate that rhLf-h-glycan protein elicited efficient toxicity on three cancer cells, with favorable SCI values, as compared with non-cancerous cells. In addition, rhLf-h-glycan caused cell shrinkage, nuclear condensation, nuclear fragmentation, the formation of apoptotic bodies, and plasma membrane blebbing, which are morphological alterations associated with apoptosis. Furthermore, rhLf-h-glycan caused microfilaments shrinkage, probably linked to the observed membrane blebbing. Moreover, rhLf-h-glycan induced PARP-1 cleavage and DNA fragmentation both considered biochemical hallmarks of apoptosis. Furthermore, rhLf-h-glycan caused cell cycle G2/M arrest. These promising findings strongly support the anticancer properties of rhLf-h-glycan glycoprotein. In view of the importance of the type of glycosylation carried by a potential application in humans, our rhLf-h-glycan is an attractive human-compatible glycosylated protein, which can avoid induction of detrimental immune responses in humans receiving this type treatment in the future.