The restriction enzymes, including BamHI, NdeI, XhoI, and Pfu DNA polymerase used in the present study were purchased from Tiangen Biotech Co., Ltd. (Beijing, China). The T4 DNA ligase and the ligase buffer were purchased from Fermentas. All primers used in the PCR experiments were synthesized at Invitrogen Co., Ltd. (Shanghai, China), and the recombinant plasmids were also sequenced at Invitrogen. The protein marker was purchased from GenStar Biosolutions Co., Ltd. (Beijing, China). Streptavidin was purchased from Tianjin Heowns, Biochemical Technology Co., Ltd. (Tianjin, China). Dynabeads^® M-280 Streptavidin was purchased from Invitrogen Co., Ltd. (Shanghai, China). X-VIVO™ 15 medium was purchased from Sartorius Stedim Biotech GmbH. The FITC-conjugated anti-human CD3 monoclonal antibody was purchased from Biolegend, and APC-conjugated anti-human CD56 monoclonal antibody was purchased from Miltenyi Biotec. Human AB serum was purchased from Andygene Co. (USA). Fetal bovine serum (FBS) was purchased from Molecular Devices (USA).

The plasmids pGEX-6p-1 and pET-28a and E. coli strains DH5α and BL21 (DE3) were obtained from our laboratory. We isolated PBMCs from the concentrated human leukocyte blood of healthy donors provided at the Beijing Red Cross Blood Center. The cell line K562 transformed into luciferase plasmid (K562-luc cells) was also obtained from our laboratory.

To generate the plasmid construct pGEX-6p-1-h4-1BBL-BirA-tag, we fused the BirA-tag gene to the N-terminus of the 184-amino acid extracellular domain of human 4-1BBL. There is a (GGGGS)3 linker between h4-1BBL and BirA-tag. Overlap PCR was used for target gene amplification, and we inserted the target gene between the BamHI and XhoI restriction sites in the plasmid pGEX-6p-1. In addition, we amplified the human IL-21 gene without the signal peptide and generated a C-terminal BirA-tagged fusion protein, with a linker bridging IL-21 and the BirA-tag, as described above. The target gene was also amplified using overlap PCR, however, the target gene was inserted between the NdeI and XhoI restriction sites in pET-28a to generate the recombinant plasmid pET-28a-hIL-21-BirA-tag. The primers used to PCR amplify the target genes are shown in Table I.

To express the proteins h4-1BBL and hIL-21, the recombinant plasmids pGEX-6p-1-h4-1BBL-BirA-tag and pET-28a-hIL-21-BirA-tag were transformed into E. coli strain BL21 (DE3) through heat shock transformation. To express h4-1BBL, the cells were grown for 8 h at 37°C from a single colony in 5 ml LB medium supplemented with 50 μg/ml of ampicillin. Subsequently, the cells were diluted 100-fold into 300 ml of fresh medium (supplemented as above) and 40-fold into two liters of fresh medium (supplemented as above) and incubated at 37°C with shaking at 200 rpm until reaching an OD600 of 0.6. Protein expression was induced upon the addition of 1 mM IPTG. The cells were grown at 25°C for 5 h and subsequently harvested through centrifugation. To express hIL-21, the single colony propagation was performed as described for h4-1BBL, but the medium was supplemented with 50 μg/ml of kanamycin. When the OD600 reached 0.8, the culture was induced with 1 mM IPTG and subsequently incubated overnight at 37°C. The protein expression was examined using 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).

To purify h4-1BBL-BirA-tag protein, the harvested cells were lysed using ultrasonication. The GST-h4-1BBL fusion protein was harvested from the supernatant using Glutathione Sepharose 4 Fast Flow according to the manufacturer’s instructions. The GST-tag was cleaved using precision protease (PSP) on the column at 4°C overnight, and the target protein was purified through fast protein liquid chromatography (FPLC). To purify hIL-21-BirA-tag protein, we obtained the inclusion bodies after lysing the cells, and refolded the proteins using the rapid dilution method. Subsequently, we concentrated the protein from the refolding buffer using Ultrafiltration concentration tubes (10 kDa), followed by FPLC.

Several solutions were used for protein biotinylation: biotin-protein ligase (3 mg/ml BirA enzyme); Solution A (0.5 M bicine buffer, pH 8.3); Solution B (100 mM ATP, 100 mM MgOAc, and 200 mM biotin); extra D-biotin (500 mM biotin); pepstatin (2 mg/ml in DMSO) and leupeptin (2 mg/ml in dH₂O). The biotinylation reaction system included 700 μl of target protein (1–2 mg), 100 μl Solution A, 100 μl Solution B, 100 μl extra D-biotin, 50 μl BirA enzyme, 2 μl pepstatin and 2 μl leupeptin. The reaction was incubated overnight at 25°C. For biotinylated h4-1BBL-BirA-tag (h4-1BBL-BirA-tag-biotin) protein was directly purified through FPLC. However, the biotinylated hIL-21-BirA-tag (hIL-21-BirA-tag-biotin) protein was purified using Ni-NTA, followed by FPLC.

We used 3 EP tubes (1.5 ml) to conduct this assay. Two tubes were filled with either moderate biotinylated protein or SA, and equal amounts of both biotinylated protein and SA were added to the third tube. PBS was added to the tubes containing only protein or SA to obtain equal reaction volumes in all tubes. Subsequently, the tubes were incubated on ice for 0.5–1 h after mixing. The shift results were assessed using 12% SDS-PAGE.

Appropriate amounts of Dynabeads M-280 Streptavidin were washed with PBST (0.05% Tween-20 in PBS) and PBS, and resuspended in 500 μl PBS. Excessive and equimolar h4-1BBL-BirA-tag-biotin and hIL-21-BirA-tag-biotin proteins were mixed with the prepared beads. The ligation reaction system was performed on a rotary mixer at 4°C for at least 6 h. Mouse anti-His IgG-PE was added to a sample of the ligation reaction to detect the ligation efficiency of hIL-21-BirA-tag-biotin and the beads, while rabbit anti-His IgG-FITC was added to detect the ligation efficiency of h4-1BBL-BirA-tag-biotin and the beads, and mouse anti-His IgG-PE and rabbit anti-His IgG-FITC were consecutively used to detect the ligation efficiency of the two proteins and the beads. The same batch beads were used to repeat the experiments described above, but phosphate-buffered saline (PBS) was substituted for mouse anti-His IgG-PE, 4-1BB-His-tag protein and rabbit anti-His IgG-FITC as important controls to exclude the influence of non-specific ligation.

PBMCs were isolated from human concentrated leukocyte blood, resuspended in X-VIVO™ 15 medium supplemented with 10% human AB serum, 10 mmol/l HEPES, 1 mmol/l sodium pyruvate, 2 mmol/l L-glutamine, 1% MEM NEAA, 1% penicillin-streptomycin and 100 IU/ml IL-2, seeded at a density of 1×10⁶ cells/ml onto 24-well plates at 1 ml/well, and subsequently incubated at 37°C in 5% CO₂. The following experimental groups were designated: two proteins (4-1BBL-IL-21-beads), single fixed and single soluble proteins (4-1BBL-beads+sIL-21, s4-1BBL+IL-21-beads), the mixture of the two soluble proteins (s4-1BBL+sIL-21), and the non-irritant control group. Each group was prepared with two parallel wells. On the first day, each group was co-incubated with PBMCs at 3:1. Repeated stimulation was performed weekly for 3 weeks. We should appropriately enlarge the culture system to ensure the viable cell density of 0.5–2×10⁶/ml during this stimulus.

After stimulation, we collected the cells of each group, separated the beads from the culture system with magnetic separation rack, and stained the cells with an LDS/PI mixture for cell counting using Guava flow cytometry. Approximately 5×10⁵ cells were washed twice with PBS, followed by incubation with the appropriate antibodies (FITC-conjugated anti-human CD3 or APC-conjugated anti-human CD56) in a 100-μl reaction volume for 30 min at 4°C. Subsequently, the cells were washed twice as described above and resuspended in 300 μl of PBS. The data were acquired using a FACSCalibur flow cytometer (BD Biosciences) and analyzed using FlowJo software (Ashland, OR, USA).

K562-luc cells cultured in the basic RPMI-1640 medium supplemented with 10% FBS, 1% penicillin-streptomycin were used as the target cells. The target cells were diluted to 1×10⁵ cells/ml in basic medium and plated onto 96-well plates at 100 μl/well. Subsequently, effector cells were added to the 96-well plates and incubated with the target cells for 4 h in 5% CO₂ at 37°C at ratios of 8:1, 4:1, 2:1, and 1:1, with 4 parallel experiment groups for each effector/target (E/T) ratio. Subsequently, we added luciferin and measured the total fluorescence in each well. The percent cytotoxicity was calculated using the following formula: cytotoxicity (%) = (control group fluorescence value-experimental group fluorescence value)/control group fluorescence value × 100%.

4-1BBL (4-1BB ligand or CD137 ligand) is a member of the tumor necrosis factor (TNF) ligand family. This 254-amino acid protein contains a 28-amino acid cytoplasmic domain, a 21-amino acid transmembrane domain, and a 205-amino acid extracellular domain (31). In the present study, the extracellular domain was cloned and the recombinant expression cassette is shown in Fig. 1A. Because 4-1BBL is a type II transmembrane protein (31), we generated an N-terminal BirA-tagged fusion protein to ensure that the protein functions normally, and the theoretical molecular weight of the recombinant protein is ~22.1 kDa. However, for human IL-21, a secreted protein belonging to the IL-15/IL-2 family, there is no significant functional difference between the N- and C-terminal fusion proteins (32). Therefore, we generated a C-terminal BirA-tagged fusion for this protein (Fig. 1A), the theoretical molecular weight of which is ~20.3 kDa. These constructs were confirmed through gel electrophoresis of the PCR products for targeted cassettes (Fig. 1B).

The constructs were expressed in E. coli strain BL21 (DE3), and subsequently the h4-1BBL-BirA-tag, which forms a dimer in solution, was cut using PSP, and the recombinant proteins were purified using size-exclusion chromatography (Fig. 1D). The hIL-21-BirA-tag recombinant protein was obtained through denaturation and refolding (33,34) and subsequently purified using size-exclusion chromatography (Fig. 1C).

Following biotinylation, h4-1BBL-BirA-tag-biotin protein was purified using FPLC (Fig. 1D). SDS-PAGE was used to identify the purified target protein, and the results showed that peak 1 was h4-1BBL-BirA-tag-biotin protein (Fig. 1D). However, for hIL-21-BirA-tag-biotin protein, the target protein could not be purified from the other proteins in the first experiment (data not shown). His-tag in hIL-21-BirA-tag protein was used to purify this protein using Ni-NTA beads, followed by further purification through FPLC (data not shown). Next, we used the SA electrophoretic mobility shift assay (35) to measure the biotinylation efficiency of these two proteins. The data revealed that these two types of protein were nearly 100% shifted with SA after conjugation, suggesting that both proteins were markedly and efficiently biotinylated (Fig. 2A). To ensure that the stimulation experiment could be effectively conducted, the bead-protein ligation efficiency was also examined. The FACS analysis demonstrated that each of two proteins could efficiently ligate with the beads, with an efficiency of 96.4% for h4-1BBL-BirA-tag-beads (Fig. 2B-b) and 97.8% for hIL-21-BirA-tag-beads (Fig. 2B-a). Moreover, these proteins could efficiently be simultaneously bound to the beads, with ~80% efficiency (Fig. 2B-c).

After stimulation for 3 weeks, the expanded cells were harvested, purified from the culture system with magnetic separation rack and stained with LDS/PI to evaluate cell viability using flow cytometry. Appropriate amounts of the cells from each group were sampled to detect the expression of the NK cell surface molecular markers CD3 and CD56 via flow cytometry and to determine the NK cell purity in each group. The lymphocyte profile before stimulation is shown in Fig. 3, indicating that the sample was obtained from a normal source, with 10–15% NK cells, consistent with previous studies, and this information was used to calculate the fold-expansion of each subset in each group of cells. Compared with the NK purity before stimulation in the non-irritant control group, the NK purity in each of the stimulation groups was markedly increased, particularly in the 4-1BBL-IL-21-beads stimulation group, obtaining 95% NK purity, which was the highest level detected among all groups (Fig. 3). This result suggests that 4-1BBL-IL-21-beads could elicit high-purity NK cells. Furthermore, according to the statistical and calculated results shown in Table II, the 4-1BBL-IL-21-bead-stimulation group achieved the highest fold-expansion of NK cells, with ~140-fold higher expansion than the other groups. Moreover, the NKT cells in each group showed a small increase in the ratio of lymphocytes, but a high proportion of T cells was no longer observed (Fig. 3), potentially reflecting the fact that 4-1BBL-IL-21-beads were beneficial for the expansion of NK cells but not T cells. Thus, 4-1BBL-IL-21-beads induced significant NK cell expansion.

Luciferase quantitative assay, a simple and sensitive method, has been increasingly and widely used to measure the cytolytic effects of various effector cells (36–38). In the present study, we used this method to test the cytotoxicity of stimulated cells against K562-luc cells. The results showed that compared with the control group and other stimulation groups, stimulation with 4-1BBL-IL-21-beads achieved the best cytotoxicity against K562-luc cells at each E/T ratio, and this cell group exhibited high cytotoxicity even at an E/T ratio of 1:1, with nearly 100% cytotoxicity at an E/T ratio of 8:1 (Fig. 4). In contrast, the cells under single fixed and single soluble cytokine stimulation or mixed soluble cytokine stimulation showed higher killing activity than the cells in the control group, and cytotoxicity increased with increasing E/T ratio. Moreover, the control group cells also exhibited killing activity in relation to the small number of NK cells in the control group in the last graph of Fig. 3. Briefly, the cells stimulated with 4-1BBL-IL-21-beads showed the highest cytotoxicity against the K562-luc cells, indicating that NK cells could be highly activated using 4-1BBL-IL-21-beads.