Targeting of natural killer cells by rabbit antithymocyte globulin and campath-1H: similar effects independent of specificity

Abstract

T cell depleting strategies are an integral part of immunosuppressive regimens widely used in the hematological and solid organ transplant setting. Although it is known to induce lymphocytopenia, little is known about the effects of the polyclonal rabbit antithymocyte globulin (rATG) or the monoclonal anti-CD52 antibody alemtuzumab on Natural Killer (NK) cells in detail. Here, we demonstrate that induction therapy with rATG following kidney/pancreas transplantation results in a rapid depletion of NK cells. Treatment of NK cells with rATG and alemtuzumab in vitro leads to impairment of cytotoxicity and induction of apoptosis even at a 10-fold lower concentration (0.1 microg/ml) compared with T and B cells. By generating Fc-parts of rATG and alemtuzumab we illustrate that their ligation to FcgammaRIII (CD16) is sufficient for the significant induction of degranulation, apoptosis and inflammatory cytokine release (FasL, TNFalpha and IFNgamma) exclusively in CD3(-)CD56(dim) NK cells whereas application of rATG and alemtuzumab F(ab) fragments abolishes these effects. These findings are of general importance as our data suggest that NK cells are also mediators of the clinically relevant cytokine release syndrome and that their targeting by therapeutic antibodies should be considered as they are functionally relevant for the effective clearance of opportunistic viral infections and anti-tumor activity posttransplantation.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Induction therapy of rATG in simultaneous kidney/pancreas transplantation results in a significant decrease of NK cells.

Patients (n = 8) initially received 1.5 mg/kg body weight i.v. rATG (Thymoglobulin, Genzyme GmbH, Neu Isenburg, Germany) starting at day 0 followed by 4 further consecutive days posttransplantation in combination with tacrolimus, mycophenolate mofetil and steroids. We further enrolled nine patients who received a renal allograft as a control group. Control patients received two dosages of basiliximab (20 mg i.v., day 0 and day 4). Asterisks denote significant differences compared to pretransplant levels of CD3−CD56+ NK cells.

Figure 2. Rabbit ATG and alemtuzumab target surface expression of CD16 on NK cells.

Human NK cells cultured with IL-2 (200 IU/ml) and different concentrations of rATG, alemtuzumab or control antibodies such as rIgG and daclizumab (18 hours) were harvested, washed and stained for CD3, CD56, CD8 and CD16. (A) The figure illustrates the average percentages of CD16 and CD8 staining on CD3−CD56+ NK cells after co-incubation with varying concentrations of rATG. Results are displayed as means±SD (n = 5); asterisks (*) denote significant differences compared to untreated controls: ***p<0.001. (B) The mean percentage of CD16 on CD3−CD56+ NK incubated with different concentrations of rATG, alemtuzumab, daclizumab and rIgG is shown. Results are displayed as means±SD of five independent experiments. Asterisks (*) indicate values that showed significantly less CD16 expression compared to untreated controls: ***p<0.001.

Figure 3. Rabbit ATG and alemtuzumab decrease effector mechanisms of peripheral blood CD3−CD56dim NK cells.

(A) Dose-dependent decrease of NK cell cytotoxicity after 18 hours pretreatment with rATG, alemtuzumab, daclizumab and rIgG (0–100 µg). Killing assay for viable and antibody-depleted NK cells was performed using the target cell line K562 (E/T ratio was 10∶1). Analysis of six independent experiments was performed by flow cytometry. Values demonstrate results normalized to untreated cells, which were set at 100%. Asterisks (*) indicate values that showed significantly less cytotoxicity compared to control: *p−CD56dimIFNγ+ NK cells. Values demonstrate the mean of INFγ+/CD56dim NK cells normalized to untreated cells, which were set at 100%. Results are displayed as means±SD for 6 independent experiments. Asterisks (*) indicate values that showed significantly less IFNγ expression compared to untreated controls: *p<0.05, **p<0.01, ***p<0.001.

Figure 4. Rabbit ATG and alemtuzumab influence the degranulation of peripheral blood CD3−CD56dim NK cells.

(A) K562 cells (E/T ratio 2∶1) and CD107a antibody were added to NK cells cultured with IL-2 (200 IU/ml) and varying concentrations of rATG, alemtuzumab, daclizumab and rIgG. NK cells were harvested and stained for CD3 and CD56. For analysis the percentage of CD107a+ NK cells of controls (without K562) was subtracted from the CD107a+ NK cells co-incubated with K562 cells. Values demonstrate CD107a expression on CD56dim NK cells normalized to untreated cells, which were set at 100%. Results are displayed as means±SD for 5 independent experiments. Asterisks (*) indicate values that showed significantly less degranulation compared to untreated controls: *p<0.05, ** p<0.01, *** p<0.001. (B) A degranulation assay was performed (A) without the addition of K562 cells. Results are displayed as the mean of CD107a+ cells on CD56dim NK cells±SD (n = 5). Asterisks (*) indicate values that showed significantly higher degranulation compared to untreated controls: *p<0.05, ** p<0.01, *** p<0.001. (C) Representative FACS dot plots of CD56+CD3− NK cells stained for CD107a. Treatment with 0.1 µg/ml rATG and alemtuzumab produced a higher induction of CD107a in CD56dim NK cells compared to 0.1 µg/ml daclizumab and rIgG.

Figure 5. Induction of apoptosis and necrosis in CD3−CD56dim NK cells after co-incubation with rATG or alemtuzumab.

(A) Magnetically isolated NK, T and B cells from healthy volunteer blood donors were treated with different concentrations of rATG for 18 hours. Annexin V+ and PI− cells were shown as apoptotic cells and Annexin V+ and PI+ cells as necrotic cells. Rabbit ATG induced apoptosis and necrosis of NK cells at lower concentrations (0.1 µg) compared with T and B cells. Values demonstrate the results normalized to untreated cells and are displayed as means of five independent experiments. Asterisks (*) indicate significant values: **p<0.01, ***p<0.001. (B) NK cells were cultured with IL-2 (200 IU/ml) and different concentrations of rATG, alemtuzumab, daclizumab and rIgG for 18 hours. Both rATG and alemtuzumab led to a rapid and significant induction of apoptosis of NK cells even at the low concentration of 0.1 µg/ml. In contrast, rIgG resulted in NK cell apoptosis at higher concentrations (e.g. 50 µg/ml). Values demonstrate the results normalized to untreated cells and are displayed as means of five independent experiments. Asterisks (*) indicate significant values compared to untreated controls: *p<0.05, ***p<0.001. (C) CD56+CD3− NK cells were FACS-sorted into CD56dim and CD56bright subpopulations and cultured with IL2 (200 IU) and 1 µg/ml rATG for 18 hours. Determination of apoptotic and necrotic cells was performed as for (A). Rabbit ATG induced apoptosis and necrosis exclusively in CD56dim NK cells (left-hand bar). A degranulation assay with K562 cells was performed and the mean reduction of CD107a staining after treatment of the sorted populations with 1 µg/ml rATG is illustrated in the right-hand graph (n = 2). Values demonstrate the results normalized to untreated cells and are displayed as means of two independent experiments.

Figure 6. Rabbit ATG and alemtuzumab increase FasL, TNFα and IFNγ mRNA in NK cells.

(A) IL-2 (200 IU/ml) pre-activated NK cells cultured in the presence of rATG were analyzed for FasL, TNFα and IFNγ mRNA after 1, 2, 3 and 6 hours of co-culture. rATG induced a rapid and dose-dependent induction of FasL, TNFα and IFNγ mRNA in NK cells which decreased after 6 hours of co-incubation. (B) Similar to rATG, the application of alemtuzumab results in a rapid FasL, TNFα and IFNγ mRNA induction within the first hour of co-incubation. Values demonstrate the results relativized to untreated controls (2−ΔΔct) and are displayed as means of six independent experiments. Asterisks (*) indicate significant values compared to untreated controls: **p<0.01, ***p<0.001. (C) Induction of cytokines by rATG and alemtuzumab (10 µg/ml) was further confirmed for TNFα and IFNγ at the protein level, illustrating a significant induction over time. Asterisks (*) indicate significant values: ***p<0.001 compared to untreated controls.

Figure 7. The IgG1 Fc-part of rATG and alemtuzumab is sufficient to induce enhanced cytokine expression, apoptosis and degranulation.

IL-2 (200 IU/ml) preactivated NK cells were incubated for 1 hour with either 1 µg/ml whole IgG antibodies, Fc-parts or F(ab) fragments of rATG or alemtuzumab. Cells treated with anti-CD3 (OKT3, IgG2a) or rabbit IgG served as controls. (A) Values for FasL, TNFα and IFNγ mRNA demonstrate the results relativized to untreated controls (2−ΔΔct) and are displayed as means±SD (n = 5): ***p>0.001. (B) Preactivated NK cells with IL-2 (200 IU/ml) were incubated with either 10 µg/ml intact antibody, Fc-parts or F(ab) fragments of rATG and alemtuzumab, or OKT3 for 1 hour. The antibodies daclizumab and rIgG served as controls. FACS dot plots illustrate staining for Annexin V and PI of treated CD56+CD3− NK cells. One representative of four independent experiments is shown. (C) Preactivated NK cells with IL-2 (200 IU/ml) were incubated with 10 µg/ml antibody preparations and anti-CD107a mAb for 3 hours. FACS dot plots illustrate staining for CD107a of CD56+CD3− cells. One representative of four independent experiments is shown.

Figure 8. Application of F(ab) rATG and alemtuzumab fragments abolishes apoptosis and targeting of CD16 on CD56+CD16+ NK cells.

NK cells were treated either with 1 µg/ml rATG or alemtuzumab or with 1 µg/ml F(ab) fragments of rATG or alemtuzumab for 1, 2, 3 and 6 hours. Controls remained untreated. Values of CD56+CD16+ NK cells are displayed as means±SD (n = 5); asterisks (*) indicate significant values compared to controls (***p<0.001, left panel). F(ab) fragments of rATG or alemtuzumab did not induce apoptosis and did not target CD16. Annexin V+/PI− CD56+CD3− cells were normalized to untreated cells. Values are displayed as means±SD of five independent experiments. P values are related to the rATG/alemtuzumab treatment: ***p<0.001.

Figure 9. The Fc-part of rATG and alemtuzumab is sufficient to induce cytokine release in whole blood samples.

Whole blood samples were treated for 2 hours with 1, 10 and 50 µg/ml of intact antibody or Fc-parts of rATG and alemtuzumab. After red blood cell lysis induction of TNFα, IFNγ and FasL mRNA was observed. Values demonstrate the results relativized to untreated controls (2−ΔΔct) and are displayed as means of four independent experiments. Asterisks (*) indicate significant values compared to untreated controls: * p<0.05, **p<0.01, ***p<0.001.