HMGN2, a new anti-tumor effector molecule of CD8⁺ T cells

PHA is a lectin found in plants, especially legumes, and is a mitogen that triggers T-lymphocyte cell division and activation. Supernatants of PBMCs that have been stimulated by PHA contain a series of effector proteins. In our experiments, PBMCs were isolated from healthy donors and stimulated with 20 μg/ml PHA or 100 IU/ml IL-2 for 72 hours. Supernatants were collected and HMGN2 concentrations were analyzed by ELISA. The results showed that PBMCs release high levels of HMGN2 (771.33 ± 123.12 ng/ml) following 20 μg/ml PHA stimulation for 72 hours. Compared with the medium control (289.67 ± 98.5 ng/ml) and IL-2 stimulated (397.67 ± 134.15 ng/ml), HMGN2 concentration significantly increased (Figure 1A). In order to make sure that HMGN2 was released by activated T cells, we used surface and indirect intracellular staining to analyze HMGN2 expression in CD3⁺ T cells. Consistent with the ELISA results (Figure 1A), compared with the medium control (9.06 ± 3.8%) and IL-2 stimulated (18.85 ± 8.71%), the percentage of HMGN2⁺CD3⁺ cell population significantly increased after PHA stimulation (47.79 ± 11.87%) (Figure 1B and C).

To determine which T cell population expressed HMGN2, we stained the PHA pre-activated PBMCs with CD4-PE or CD8-PE surface staining. We then stained intracellular HMGN2 with rabbit anti-human HMGN2/goat anti-rabbit IgG-FITC. Results showed that HMGN2 expression significantly increased in both CD4⁺ and CD8⁺ T cell populations (Figure 2A, B, C). In addition, HMGN2 expression in CD8⁺ T cells (50.71 ± 10.34%) was significantly higher than that in CD4⁺ T cells (16.67 ± 5.61%) after PHA stimulation (Figure 2D). Finally, we used a MoFlo XDP high-speed flow cytometry sorter to purify PHA pre-activated CD3⁺CD8⁺ T cells, CD3⁺CD8^- T cells and CD3^- mix cells, followed by culturing in complete medium with 2000 IU/ml IL-2 for 5 days and analyzed the expression of HMGN2 by ELISA and intracellular staining. Results showed that HMGN2 was still expressed at high levels in CD8⁺ T cells (539.00 ± 118 ng/ml; 68.37 ± 15.21%). On the other hand, the expression of HMGN2 in the CD3⁺CD8^- T cells (mainly CD4⁺ T cells) (307.67 ± 97.34 ng/ml; 36.23 ± 11.52%) and the CD3^- mixed PBMCs population (386.67 ± 105.46 ng/ml; 35.73 ± 13.71%) was significantly lower in both supernatants and intercellular (Figure 3A, B and C). These results confirmed that activated CD8 + T cells expressed high levels of HMGN2.

In order to ensure that HMGN2 was an effector protein of tumor antigen activated T cells, we used tumor full protein as tumor antigen (T-Ag) to stimulate PBMCs for 7 days. Supernatants were collected for assaying HMGN2 levels by ELISA and cells were collected for HMGN2 intracellular staining. Results showed that there was no significant change of HMGN2 expression after stimulation with T-Ag compared with the medium control (Figure 4). In order to ensure that T cells were activated by T-Ag, we used CD44^high as the activated marker of T cells. T-Ag stimulated PBMCs were stained with CD8-PE/CD44-APC surface and HMGN2 indirect intracellular staining. CD44^high was used as the activated T cell population (Figure 5Aa, gate R2) and CD44^low was used as the naive T cell population (Figure 5Aa, gate R6). T-Ag induced only 18.35 ± 6.20% PBMCs activated (Figure 5Aa, gate R2 and5B). There was about 43.69 ± 12.51% of activated CD8⁺ T cells expressed HMGN2 (Figure 5Ab and C), where only 18.71 ± 6.34% naive CD8⁺ T cells expressed HMGN2 (Figure 5Ac and C).

We used a Beckman Coulter MoFlo XDP high-speed flow cytometry sorter to isolate T-Ag-activated the CD44^highCD8⁺ T cell population (Figure 6A, Gate R4). The purified cells were cultured with 2000 IU/ml IL-2 for 5 days and then the cells were identified with CD44-APC and CD8-FITC staining (Figure 6A). Cells were collected and stained with CD8-PE surface and HMGN2 intracellular staining. A total of 69.35 ± 12.13% of CD8⁺ T cells still expressed HMGN2 (Figure 6B).

Tca8113 cells were seeded at a density of 1 × 10³ cells per well in 96-well plates. After overnight growth, the medium was replaced with maintenance medium containing the desired concentrations (v/v) of T-Ag activated CD8⁺ T cell supernatants. To one group, anti-HMGN2 antibody (10 μg/ml) was added, in order to block HMGN2. Human HMGN2 protein was used as the positive control, and the same volume of normal medium as the negative control. Cell viability was assessed after 72 hours using the CCK8 colorimetric assay. The supernatant was able to kill tumor cells in a dose-dependent manner; 20% supernatant could significantly kill tumor cells, as could the HMGN2 positive control (Figure 6C). The effect of killing tumor cells by the supernatant could be significantly inhibited by the anti-HMGN2 antibody (Figure 6D).

To confirm that the HMGN2 in the supernatant of T-Ag activated CD8⁺ T cells could be transmembrane transported into tumor cells, we used fluorescence FITC to label the proteins in the supernatant, before adding to the medium. Human FITC labeled HMGN2 protein was used as the positive control. Results showed that the protein in the supernatant could effectively be transported into the tumor cells, as could the human HMGN2 protein control (Figure 7A, Figure 7B b&e, Figure 7C b&d). After HMGN2-depleted by anti-HMGN2, the number of FITC-positive cells visibly decreased comparing with HMGN2 undepleted samples analyzed by both fluorescent microscope (Figure 7B b vs c, e vs f) and Flow Cytometry (Figure 7C b vs c, d vs e).

The human tongue squamous cell carcinoma cell line Tca8113 cells were obtained from the State Key Laboratory of Oral Disease (Sichuan University, Chengdu, China). Cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 100 IU/ml penicillin, 100 μg/ml streptomycin, 3% L-glutamine, and 7.5% sodium bicarbonate (GIBCO Life Technologies). Cells were maintained as a monolayer in 25 cm plastic tissue culture flasks at 37°C in a humidified atmosphere containing 5% CO₂. Tumor full protein (Tumor antigen, T-Ag) was prepared by lysing Tca8113 cells in PBS, the lysate was collected and stored at -80°C as T-Ag.

PBMCs from healthy human donors were isolated using Human Lymphocyte Separation tube (Beijing Dakewe Biotech Company Limited, China). PBMCs were plated at a density of 1 × 10⁷/well in six-well plates and cultured in RPMI 1640 medium containing 10% fetal bovine serum, 100 IU/ml penicillin, and 100 μg/ml streptomycin). The cells were stimulated with 150 μg/ml T-Ag for 7 days, or 20 μg/ml Phytohemagglutinin (PHA, Sigma, USA), 100 IU/ml IL-2 for 72 hours to activate T lymphocytes. Supernatants were collected and HMGN2 concentration was analyzed, along with its effect on tumor cell survival. Cells were removed to analyze HMGN2 expression by intracellular staining.

PHA stimulated PBMCs were stained with CD8-PE/CD3-FITC at 4°C for 30 minutes. Cells were washed three times with sterilized PBS. The CD8⁺CD3⁺T cells, CD8^-CD3⁺T cells, and CD3^- Mix cells populations were gated and isolated using MoFlo XDP high-speed flow cytometry sorter (Beckman). The purified T cell populations were cultured in complete medium with 2000 IU/ml IL-2 for 5 days. Cells were collected for analyzing HMGN2 expression.

T-Ag-stimulated PBMCs were stained with CD8-FITC/CD44-APC at 4°C for 30 minutes. CD44^highCD8⁺ cells were gated as the activated CD8⁺ T cell population. CD44^highCD8⁺ activated CD8⁺ T cells were isolated using a MoFlo XDP high-speed flow cytometry sorter (Beckman). The purified T cells were cultured in complete medium with 2000 IU/ml IL-2 for 5 days. Supernatants and cells were collected for analyzing anti-tumor effect and HMGN2 expression.

ELISA plates were coated with 100 μl of supernatants, standards made from different concentrations of human HMGN2 protein, or PBS (negative control). The plates were left at 4°C overnight. After washing three times with wash buffer, 100 μl of rabbit anti-human HMGN2 antibody (Proteintech Group, USA) (1:500) were added and plates were incubated at 37°C for 1 hour. Then, 100 μl of HRP-conjugated anti-rabbit IgG secondary antibody (1:1000) were added after thorough washing of the first antibody, and plates were incubated at 37°C for 1 hour. Add 100 μl TMB substrate solution to each well. After 20 min incubation, reactions were stopped with 2 N H₂SO₄ and measured at 490 nm in an ELISA plate reader (VARIOSKAN FLASH, Thermo Fisher Scientific, Vantaa, Finland).

PHA or T-Ag stimulated PBMCs or purified T cell populations were collected and stained with fluorescence-labeled CD4, CD8, or CD44 surface marker or msIgG-PE/FITC isotypes (Biolegend, USA) control. After washing three times with wash buffer, cells were analyzed for HMGN2 expression by using an intracellular staining kit (Invitrogen, USA). Briefly, a volume of 100 μl fixation buffer was added to fix cells, which were then left at 4°C overnight, followed by washing with permeabilization buffer three times in order to permeabilize cells. All of the samples were divided into two tubes, rabbit-anti-human HMGN2 antibody (1 μg/ml) was added into one tube and the same volume PBS was added into the other as the 2^nd-Ab-FITC control. The samples were incubated at 4°C for 1 hour. The cells were washed three times with permeabilization buffer and then incubated with goat anti-rabbit-IgG-FITC secondary antibody (2^nd-Ab-FITC) at 4°C for 1 hour. Finally, the cells were washed with Flow Cytometry buffer and runed on a Beckman coulter FC500 Flow cytometry. Dates were analyzed by using Submit 5.2 software (Beckman Coulter, USA) after gate lymphocytes group on dot plot graph.

Tca8113 cells were seeded at a density of 1 × 10³ cells per well in 96-well plates. After overnight growth, the medium was replaced with maintenance medium containing the desired concentrations (v/v) of supernatant of T-Ag-activated CD8⁺ T cells. Human HMGN2 protein was used as the positive control and medium as the negative control. Blocking of HMGN2 was achieved by adding 10 μg/ml anti-HMGN2 antibody to 20% (v/v) supernatants. Cell viability was assessed after 72 hours using the CCK8 colorimetric assay. Briefly, the cells were washed with 300 μl of PBS, followed by incubation with 100 μl of 5 mg/ml CCK8 in RPMI 1640 at 37°C for 1 hour, and quantified by measuring the optical absorbance (OA) at 450 nm in a plate reader (VARIOSKAN FLASH, Thermo Fisher Scientific, Vantaa, Finland).

Human HMGN2 protein (2 μg/ml) and the supernatants of T-Ag activated CD8⁺ T cells were labeled with FITC. Briefly, FITC was added to 2 mg/ml protein solution and incubated for 5 hours at room temperature. The unconjugated FITC was removed using a 3 kDa filter. Half of these FITC-labeled samples were depleted of HMGN2 by using anti-human HMGN2 antibody adsorption in 96-well plates. Briefly, 10ug anti-human HMGN2 antibody was coated in 96-well plates at 4°C for overnight. The free antibody was removed by washing the wells with PBS three times. The FITC labeled samples were added into the wells and incubated at 37°C for 1 hour. The samples were collected and store at -80°C as the HMGN2 depleted samples.

Tca8113 cells were seeded at a density of 3 × 10⁴ per well in 24-well plates. After overnight growth, 2 μg FITC-labeled HMGN2, 20% (v/v) FITC-labeled CD8⁺ T cells supernatants and the same volume HMGN2-depleted samples were added to the mediums, respectively. Plates were incubated at 37°C in a humidified atmosphere containing 5% CO₂ for 1 hour. Nuclear staining control of Tca8113 cells was measured using Hoechst 33258 (Promega Corporation, Madison, WI, USA). Staining was performed according to the manufacturer's instructions. Cells were analyzed under a fluorescence microscope and pictures were taken. Then, all the cells were removed with 0.25% trypsin and analyzed FITC positive cells on a Beckman coulter FC500 using Submit 5.2 software. Untreated Tca8113 cells were used as the negative control.

All values were expressed as means ± SEM. Data were analyzed by one-way analysis of variance (ANOVA) followed by Bonferroni test. P < 0.05 was considered to indicate a statistically significant difference.